CC2540/41 HW User Guide

CC253x System-on-Chip Solution for 2.4-GHz
IEEE 802.15.4 and ZigBee® Applications
A
CC2540/41 System-on-Chip Solution for 2.4GHz Bluetooth® low energy Applications
User's Guide
Literature Number: SWRU191D
April 2009 – Revised March 2013
Contents
......................................................................................................................................
Introduction ......................................................................................................................
1.1
Overview ....................................................................................................................
1.1.1 CPU and Memory ................................................................................................
1.1.2 Clocks and Power Management ................................................................................
1.1.3 Peripherals ........................................................................................................
1.1.4 Radio ...............................................................................................................
1.2
Applications ................................................................................................................
8051 CPU ..........................................................................................................................
2.1
8051 CPU Introduction ....................................................................................................
2.2
Memory .....................................................................................................................
2.2.1 Memory Map ......................................................................................................
2.2.2 CPU Memory Space .............................................................................................
2.2.3 Physical Memory .................................................................................................
2.2.4 XDATA Memory Access .........................................................................................
2.2.5 Memory Arbiter ...................................................................................................
2.3
CPU Registers .............................................................................................................
2.3.1 Data Pointers ......................................................................................................
2.3.2 Registers R0–R7 .................................................................................................
2.3.3 Program Status Word ............................................................................................
2.3.4 Accumulator .......................................................................................................
2.3.5 B Register .........................................................................................................
2.3.6 Stack Pointer ......................................................................................................
2.4
Instruction Set Summary .................................................................................................
2.5
Interrupts ....................................................................................................................
2.5.1 Interrupt Masking .................................................................................................
2.5.2 Interrupt Processing ..............................................................................................
2.5.3 Interrupt Priority ...................................................................................................
Debug Interface .................................................................................................................
3.1
Debug Mode ...............................................................................................................
3.2
Debug Communication ...................................................................................................
3.3
Debug Commands ........................................................................................................
3.3.1 Debug Configuration .............................................................................................
3.3.2 Debug Status ......................................................................................................
3.3.3 Hardware Breakpoints ...........................................................................................
3.4
Flash Programming .......................................................................................................
3.4.1 Lock Bits ...........................................................................................................
3.5
Debug Interface and Power Modes .....................................................................................
3.6
Registers ....................................................................................................................
Power Management and Clocks ..........................................................................................
4.1
Power Management Introduction ........................................................................................
4.1.1 Active and Idle Mode .............................................................................................
4.1.2 PM1 ................................................................................................................
4.1.3 PM2 ................................................................................................................
Preface
14
1
17
2
3
4
2
Contents
18
21
21
21
23
23
24
25
25
25
27
28
33
33
34
34
35
35
36
36
36
36
40
41
45
47
50
51
51
53
55
55
56
57
57
57
59
60
61
62
62
62
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
www.ti.com
4.2
4.3
4.4
4.5
4.6
5
Reset
5.1
5.2
6
6.3
6.4
6.5
7.7
7.8
7.9
7.10
7.11
8
............................................................................................................................... 70
Flash
Flash
6.2.1
6.2.2
6.2.3
6.2.4
Flash
6.3.1
6.3.2
Flash
Flash
I/O Ports
7.1
7.2
7.3
7.4
7.5
7.6
................................................................................................................ 72
Memory Organization ..............................................................................................
Write .................................................................................................................
Flash-Write Procedure ...........................................................................................
Writing Multiple Times to a Word ...............................................................................
DMA Flash Write .................................................................................................
CPU Flash Write ..................................................................................................
Page Erase .........................................................................................................
Performing Flash Erase From Flash Memory ................................................................
Different Flash Page Size on CC2533 .........................................................................
DMA Trigger ........................................................................................................
Controller Registers ................................................................................................
73
73
73
74
74
75
75
76
76
76
76
........................................................................................................................... 78
Unused I/O Pins ...........................................................................................................
Low I/O Supply Voltage ...................................................................................................
General-Purpose I/O ......................................................................................................
General-Purpose I/O Interrupts ..........................................................................................
General-Purpose I/O DMA ...............................................................................................
Peripheral I/O ..............................................................................................................
7.6.1 Timer 1 .............................................................................................................
7.6.2 Timer 3 .............................................................................................................
7.6.3 Timer 4 .............................................................................................................
7.6.4 USART 0 ...........................................................................................................
7.6.5 USART 1 ...........................................................................................................
7.6.6 ADC ................................................................................................................
7.6.7 Operational Amplifier and Analog Comparator ...............................................................
Debug Interface ............................................................................................................
32-kHz XOSC Input .......................................................................................................
Radio Test Output Signals ...............................................................................................
Power-Down Signal MUX (PMUX) ......................................................................................
I/O Registers ...............................................................................................................
DMA Controller
8.1
8.2
62
62
63
66
66
66
67
67
69
69
Power-On Reset and Brownout Detector .............................................................................. 71
Clock-Loss Detector ....................................................................................................... 71
Flash Controller
6.1
6.2
7
4.1.4 PM3 ................................................................................................................
Power-Management Control .............................................................................................
Power-Management Registers ..........................................................................................
Oscillators and Clocks ....................................................................................................
4.4.1 Oscillators .........................................................................................................
4.4.2 System Clock .....................................................................................................
4.4.3 32-kHz Oscillators ................................................................................................
4.4.4 Oscillator and Clock Registers ..................................................................................
Timer Tick Generation ....................................................................................................
Data Retention .............................................................................................................
79
79
79
79
80
80
81
81
82
82
82
83
83
83
83
84
84
84
................................................................................................................. 92
DMA Operation ............................................................................................................
DMA Configuration Parameters .........................................................................................
8.2.1 Source Address ...................................................................................................
8.2.2 Destination Address ..............................................................................................
8.2.3 Transfer Count ....................................................................................................
8.2.4 VLEN Setting ......................................................................................................
8.2.5 Trigger Event ......................................................................................................
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Contents
93
95
95
95
95
96
96
3
www.ti.com
8.3
8.4
8.5
8.6
8.7
8.8
9
Timer 1 (16-Bit Timer)
9.1
9.2
9.3
9.4
9.5
9.6
9.7
9.8
9.9
9.10
9.11
9.12
9.13
10
10.3
10.4
10.5
10.6
10.7
10.8
12
104
104
104
105
105
105
106
106
111
111
111
112
113
113
113
113
114
119
..................................................................................... 120
8-Bit Timer Counter ......................................................................................................
Timer 3/Timer 4 Mode Control .........................................................................................
10.2.1 Free-Running Mode ...........................................................................................
10.2.2 Down Mode .....................................................................................................
10.2.3 Modulo Mode ...................................................................................................
10.2.4 Up/Down Mode .................................................................................................
Channel Mode Control ..................................................................................................
Input Capture Mode .....................................................................................................
Output Compare Mode ..................................................................................................
Timer 3 and Timer 4 Interrupts .........................................................................................
Timer 3 and Timer 4 DMA Triggers ...................................................................................
Timer 3 and Timer 4 Registers .........................................................................................
Sleep Timer
121
121
121
121
121
121
121
122
122
122
123
123
..................................................................................................................... 128
11.1
11.2
11.3
11.4
General ....................................................................................................................
Timer Compare ...........................................................................................................
Timer Capture ............................................................................................................
Sleep Timer Registers ...................................................................................................
ADC
............................................................................................................................... 132
ADC Introduction ......................................................................................................... 133
ADC Operation ........................................................................................................... 133
12.1
12.2
4
....................................................................................................... 103
16-Bit Counter ............................................................................................................
Timer 1 Operation ........................................................................................................
Free-Running Mode .....................................................................................................
Modulo Mode .............................................................................................................
Up/Down Mode ...........................................................................................................
Channel-Mode Control ..................................................................................................
Input Capture Mode .....................................................................................................
Output Compare Mode ..................................................................................................
IR Signal Generation and Learning ....................................................................................
9.9.1 Introduction ......................................................................................................
9.9.2 Modulated Codes ...............................................................................................
9.9.3 Non-Modulated Codes .........................................................................................
9.9.4 Learning ..........................................................................................................
9.9.5 Other Considerations ...........................................................................................
Timer 1 Interrupts ........................................................................................................
Timer 1 DMA Triggers ...................................................................................................
Timer 1 Registers ........................................................................................................
Accessing Timer 1 Registers as Array ................................................................................
Timer 3 and Timer 4 (8-Bit Timers)
10.1
10.2
11
8.2.6 Source and Destination Increment ............................................................................. 96
8.2.7 DMA Transfer Mode .............................................................................................. 97
8.2.8 DMA Priority ....................................................................................................... 97
8.2.9 Byte or Word Transfers .......................................................................................... 97
8.2.10 Interrupt Mask .................................................................................................... 97
8.2.11 Mode 8 Setting ................................................................................................... 97
DMA Configuration Setup ................................................................................................ 97
Stopping DMA Transfers ................................................................................................. 98
DMA Interrupts ............................................................................................................. 98
DMA Configuration Data Structure ...................................................................................... 98
DMA Memory Access ..................................................................................................... 98
DMA Registers ........................................................................................................... 101
Contents
129
129
129
130
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
www.ti.com
12.2.1
12.2.2
12.2.3
12.2.4
12.2.5
12.2.6
12.2.7
12.2.8
12.2.9
12.2.10
13
Battery Monitor
13.1
13.2
13.3
14
14.3
16
17.2
17.3
17.4
17.5
17.6
17.7
17.8
144
144
144
144
144
145
............................................................................................................ 146
AES Operation ...........................................................................................................
Key and IV ................................................................................................................
Padding of Input Data ...................................................................................................
Interface to CPU .........................................................................................................
Modes of Operation ......................................................................................................
CBC-MAC .................................................................................................................
CCM Mode ................................................................................................................
AES Interrupts ............................................................................................................
AES DMA Triggers .......................................................................................................
AES Registers ............................................................................................................
147
147
147
147
147
147
148
150
150
150
.............................................................................................................. 152
Watchdog Mode .......................................................................................................... 153
Timer Mode ............................................................................................................... 153
Watchdog Timer Register ............................................................................................... 153
USART
17.1
18
.............................................................................................. 143
Introduction ...............................................................................................................
Random-Number-Generator Operation ...............................................................................
14.2.1 Pseudorandom Sequence Generation ......................................................................
14.2.2 Seeding .........................................................................................................
14.2.3 CRC16 ...........................................................................................................
Random-Number-Generator Registers ................................................................................
Watchdog Timer
16.1
16.2
16.3
17
................................................................................................................ 139
AES Coprocessor
15.1
15.2
15.3
15.4
15.5
15.6
15.7
15.8
15.9
15.10
133
134
134
134
135
135
135
135
135
136
Functionality and Usage of the Battery Monitor ...................................................................... 140
Using the Battery Monitor for Temperature Monitoring ............................................................. 140
Battery Monitor Registers ............................................................................................... 141
Random-Number Generator
14.1
14.2
15
ADC Inputs ......................................................................................................
ADC Conversion Sequences .................................................................................
Single ADC Conversion .......................................................................................
ADC Operating Modes ........................................................................................
ADC Conversion Results .....................................................................................
ADC Reference Voltage ......................................................................................
ADC Conversion Timing ......................................................................................
ADC Interrupts .................................................................................................
ADC DMA Triggers ............................................................................................
ADC Registers ................................................................................................
............................................................................................................................ 155
UART Mode ...............................................................................................................
17.1.1 UART Transmit .................................................................................................
17.1.2 UART Receive .................................................................................................
17.1.3 UART Hardware Flow Control ................................................................................
17.1.4 UART Character Format ......................................................................................
SPI Mode ..................................................................................................................
17.2.1 SPI Master Operation .........................................................................................
17.2.2 SPI Slave Operation ...........................................................................................
SSN Slave-Select Pin ...................................................................................................
Baud-Rate Generation ..................................................................................................
USART Flushing .........................................................................................................
USART Interrupts ........................................................................................................
USART DMA Triggers ...................................................................................................
USART Registers ........................................................................................................
Operational Amplifier
156
156
156
156
157
157
157
158
158
158
159
159
159
159
....................................................................................................... 164
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Contents
5
www.ti.com
Description ................................................................................................................
Calibration .................................................................................................................
Clock Source .............................................................................................................
Registers ..................................................................................................................
165
165
165
165
..........................................................................................................
19.1 Description ................................................................................................................
19.2 Register ...................................................................................................................
I2C ..................................................................................................................................
20.1 Operation ..................................................................................................................
20.1.1 I2C Initialization and Reset ....................................................................................
20.1.2 I2C Serial Data ..................................................................................................
20.1.3 I2C Addressing Modes .........................................................................................
20.1.4 I2C Module Operating Modes .................................................................................
20.1.5 I2C Clock Generation and Synchronization .................................................................
20.1.6 Bus Error ........................................................................................................
20.1.7 I2C Interrupt .....................................................................................................
20.1.8 I2C Pins ..........................................................................................................
20.2 I2C Registers ..............................................................................................................
USB Controller ................................................................................................................
21.1 USB Introduction .........................................................................................................
21.2 USB Enable ...............................................................................................................
21.3 48-MHz USB PLL ........................................................................................................
21.4 USB Interrupts ............................................................................................................
21.5 Endpoint 0 ................................................................................................................
21.6 Endpoint-0 Interrupts ....................................................................................................
21.6.1 Error Conditions ................................................................................................
21.6.2 SETUP Transactions (IDLE State) ..........................................................................
21.6.3 IN Transactions (TX State) ...................................................................................
21.6.4 OUT Transactions (RX State) ................................................................................
21.7 Endpoints 1–5 ............................................................................................................
21.7.1 FIFO Management .............................................................................................
21.7.2 Double Buffering ...............................................................................................
21.7.3 FIFO Access ....................................................................................................
21.7.4 Endpoint 1–5 Interupts ........................................................................................
21.7.5 Bulk/Interrupt IN Endpoint ....................................................................................
21.7.6 Isochronous IN Endpoint ......................................................................................
21.7.7 Bulk/Interrupt OUT Endpoint .................................................................................
21.7.8 Isochronous OUT Endpoint ...................................................................................
21.8 DMA .......................................................................................................................
21.9 USB Reset ................................................................................................................
21.10 Suspend and Resume ..................................................................................................
21.11 Remote Wake-Up ........................................................................................................
21.12 USB Registers ...........................................................................................................
Timer 2 (MAC Timer) ........................................................................................................
22.1 Timer Operation ..........................................................................................................
22.1.1 General ..........................................................................................................
22.1.2 Up Counter ......................................................................................................
22.1.3 Timer Overflow .................................................................................................
22.1.4 Timer Delta Increment .........................................................................................
22.1.5 Timer Compare .................................................................................................
22.1.6 Overflow Count .................................................................................................
22.1.7 Overflow-Count Update .......................................................................................
166
18.1
18.2
18.3
18.4
19
20
21
22
6
Analog Comparator
Contents
167
167
168
169
170
170
171
171
177
178
178
178
178
181
182
182
182
183
183
183
184
184
184
185
185
185
186
187
187
188
188
188
188
189
189
189
190
190
197
198
198
198
198
198
198
198
199
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
www.ti.com
22.2
22.3
22.4
22.5
23
22.1.8 Overflow-Count Overflow .....................................................................................
22.1.9 Overflow-Count Compare .....................................................................................
22.1.10 Capture Input .................................................................................................
22.1.11 Long Compare (CC2541 Only) .............................................................................
Interrupts ..................................................................................................................
Event Outputs (DMA Trigger and Radio Events) ....................................................................
Timer Start/Stop Synchronization ......................................................................................
22.4.1 General ..........................................................................................................
22.4.2 Timer Synchronous Stop ......................................................................................
22.4.3 Timer Synchronous Start .....................................................................................
Timer 2 Registers ........................................................................................................
CC253x Radio
199
199
199
199
199
200
200
200
200
201
202
.................................................................................................................. 208
RF Core ...................................................................................................................
23.1.1 Interrupts ........................................................................................................
23.1.2 Interrupt Registers .............................................................................................
23.2 FIFO Access ..............................................................................................................
23.3 DMA .......................................................................................................................
23.4 Memory Map ..............................................................................................................
23.4.1 RXFIFO .........................................................................................................
23.4.2 TXFIFO ..........................................................................................................
23.4.3 Frame-Filtering and Source-Matching Memory Map ......................................................
23.5 Frequency and Channel Programming ................................................................................
23.6 IEEE 802.15.4-2006 Modulation Format ..............................................................................
23.7 IEEE 802.15.4-2006 Frame Format ...................................................................................
23.7.1 PHY Layer ......................................................................................................
23.7.2 MAC Layer ......................................................................................................
23.8 Transmit Mode ...........................................................................................................
23.8.1 TX Control ......................................................................................................
23.8.2 TX State Timing ................................................................................................
23.8.3 TXFIFO Access ................................................................................................
23.8.4 Retransmission .................................................................................................
23.8.5 Error Conditions ................................................................................................
23.8.6 TX Flow Diagram ..............................................................................................
23.8.7 Transmitted Frame Processing ..............................................................................
23.8.8 Synchronization Header .......................................................................................
23.8.9 Frame-Length Field ............................................................................................
23.8.10 Frame Check Sequence .....................................................................................
23.8.11 Interrupts ......................................................................................................
23.8.12 Clear-Channel Assessment .................................................................................
23.8.13 Output Power Programming ................................................................................
23.8.14 Tips and Tricks ...............................................................................................
23.9 Receive Mode ............................................................................................................
23.9.1 RX Control ......................................................................................................
23.9.2 RX State Timing ................................................................................................
23.9.3 Received Frame Processing .................................................................................
23.9.4 Synchronization Header and Frame-Length Fields .......................................................
23.9.5 Frame Filtering .................................................................................................
23.9.6 Source Address Matching ....................................................................................
23.9.7 Frame-Check Sequence ......................................................................................
23.9.8 Acknowledgement Transmission ............................................................................
23.10 RXFIFO Access ..........................................................................................................
23.10.1 Using the FIFO and FIFOP .................................................................................
23.10.2 Error Conditions ..............................................................................................
23.1
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Contents
209
209
209
213
213
213
214
214
214
215
215
217
217
217
218
218
218
218
219
219
219
221
221
221
221
222
222
222
222
222
222
223
223
224
224
227
230
230
232
232
233
7
www.ti.com
23.11
23.12
23.13
23.14
23.15
23.10.3 RSSI ............................................................................................................
23.10.4 Link Quality Indication .......................................................................................
Radio Control State Machine ..........................................................................................
Random-Number Generation ..........................................................................................
Packet Sniffing and Radio Test Output Signals .....................................................................
Command Strobe/CSMA-CA Processor ..............................................................................
23.14.1 Instruction Memory ...........................................................................................
23.14.2 Data Registers ................................................................................................
23.14.3 Program Execution ...........................................................................................
23.14.4 Interrupt Requests ............................................................................................
23.14.5 Random Number Instruction ................................................................................
23.14.6 Running CSP Programs .....................................................................................
23.14.7 Registers ......................................................................................................
23.14.8 Instruction Set Summary ....................................................................................
23.14.9 Instruction Set Definition .....................................................................................
Registers ..................................................................................................................
23.15.1 Register Settings Update ....................................................................................
23.15.2 Register Access Modes .....................................................................................
23.15.3 Register Descriptions ........................................................................................
233
234
234
236
237
238
238
239
239
239
239
239
240
241
243
255
256
256
257
26
........................................................................ 275
.................................................................................................................. 276
CC2541 Proprietary Mode Radio ........................................................................................ 278
25.1 RF Core ................................................................................................................... 279
25.2 Interrupts .................................................................................................................. 279
25.2.1 Interrupt Registers ............................................................................................. 279
25.3 RF Core Data Memory .................................................................................................. 280
25.3.1 FIFOs ............................................................................................................ 281
25.3.2 DMA ............................................................................................................. 284
25.3.3 RAM-Based Registers ......................................................................................... 285
25.3.4 Variables in RAM Page 5 ..................................................................................... 291
25.4 Bit-Stream Processor .................................................................................................... 291
25.4.1 Whitening ....................................................................................................... 291
25.4.2 CC2500-Compatible PN9 Whitening ........................................................................ 292
25.4.3 CRC .............................................................................................................. 293
25.4.4 Coprocessor Mode ............................................................................................. 295
25.5 Frequency and Channel Programming ................................................................................ 296
25.6 Modulation Formats ...................................................................................................... 296
25.7 Receiver ................................................................................................................... 296
25.8 Packet Format ............................................................................................................ 297
25.8.1 Rx FIFO Packet Organization ................................................................................ 299
25.8.2 Tx FIFO Packet Organization ................................................................................ 300
25.8.3 Tx Buffers for ACK Payload .................................................................................. 300
25.9 Link Layer Engine ........................................................................................................ 301
25.9.1 Command Register ............................................................................................ 302
25.9.2 Radio Tasks .................................................................................................... 302
25.9.3 RF Test Commands ........................................................................................... 317
25.10 Random Number Generation .......................................................................................... 317
25.11 Packet Sniffing ........................................................................................................... 318
25.12 Registers .................................................................................................................. 319
25.12.1 Register Overview ............................................................................................ 319
25.12.2 Register Settings Update .................................................................................... 320
25.12.3 SFR Register Descriptions .................................................................................. 321
Voltage Regulator ............................................................................................................ 341
8
Contents
24
CC2540/CC2541 Bluetooth low energy Radio
24.1
25
Registers
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
www.ti.com
27
Available Software
........................................................................................................... 342
SmartRF™ Software for Evaluation (www.ti.com/smartrfstudio) ..................................................
RemoTI™ Network Protocol (www.ti.com/remoti) ...................................................................
SimpliciTI™ Network Protocol (www.ti.com/simpliciti) ..............................................................
TIMAC Software (www.ti.com/timac) ..................................................................................
Z-Stack™ Software (www.ti.com/z-stack) ............................................................................
BLE Stack Software .....................................................................................................
343
343
344
344
345
345
..................................................................................................................
B
Additional Information ......................................................................................................
B.1
Texas Instruments Low-Power RF Web Site .........................................................................
B.2
Low-Power RF Online Community .....................................................................................
B.3
Texas Instruments Low-Power RF Developer Network .............................................................
B.4
Low-Power RF eNewsletter .............................................................................................
C
References ......................................................................................................................
Revision History .......................................................................................................................
346
27.1
27.2
27.3
27.4
27.5
27.6
A
Abbreviations
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Contents
349
350
350
350
350
351
352
9
www.ti.com
List of Figures
1-1.
CC253x Block Diagram ................................................................................................... 18
1-2.
CC2540 Block Diagram
1-3.
2-1.
2-2.
2-3.
2-4.
3-1.
3-2.
3-3.
3-4.
3-5.
4-1.
6-1.
8-1.
8-2.
9-1.
9-2.
9-3.
9-4.
9-5.
9-6.
9-7.
9-8.
9-9.
11-1.
12-1.
14-1.
15-1.
15-2.
15-3.
15-4.
19-1.
20-1.
20-2.
20-3.
20-4.
20-5.
20-6.
20-7.
20-8.
21-1.
21-2.
23-1.
23-2.
23-3.
23-4.
10
.................................................................................................. 19
CC2541 Block Diagram ................................................................................................... 20
XDATA Memory Space (Showing SFR and DATA Mapping) ....................................................... 26
CODE Memory Space .................................................................................................... 26
CODE Memory Space for Running Code From SRAM .............................................................. 26
Interrupt Overview ......................................................................................................... 43
External Debug Interface Timing ........................................................................................ 51
Transmission of One Byte ................................................................................................ 51
Typical Command Sequence—No Extra Wait for Response ........................................................ 52
Typical Command Sequence. Wait for Response .................................................................... 53
Burst Write Command (First 2 Bytes)................................................................................... 55
Clock System Overview .................................................................................................. 65
Flash Write Using DMA ................................................................................................... 75
DMA Operation ............................................................................................................ 94
Variable Length (VLEN) Transfer Options ............................................................................. 96
Free-Running Mode ..................................................................................................... 104
Modulo Mode ............................................................................................................. 105
Up/Down Mode ........................................................................................................... 105
Output Compare Modes, Timer Free-Running Mode ............................................................... 108
Output Compare Modes, Timer Modulo Mode ....................................................................... 109
Output Compare Modes, Timer Up/Down Mode ..................................................................... 110
Block Diagram of Timers in IR-Generation Mode.................................................................... 112
Modulated Waveform Example ........................................................................................ 112
IR Learning Board Diagram ............................................................................................ 113
Sleep Timer Capture (Example Using Rising Edge on P0_0) ..................................................... 130
ADC Block Diagram ..................................................................................................... 133
Basic Structure of the Random-Number Generator ................................................................. 144
Message Authentication Phase Block B0 ............................................................................ 148
Authentication Flag Byte ................................................................................................ 148
Message Encryption Phase Block ..................................................................................... 149
Encryption Flag Byte .................................................................................................... 149
Analog Comparator ...................................................................................................... 167
Block Diagram of the I2C Module ...................................................................................... 169
I2C Bus Connection Diagram ........................................................................................... 170
I2C Module Data Transfer ............................................................................................... 170
Bit Transfer on I2C Bus .................................................................................................. 171
I2C Module 7-Bit Addressing Format .................................................................................. 171
I2C Module Addressing Format With Repeated START Condition ................................................ 171
Arbitration Procedure Between Two Master Transmitters .......................................................... 177
Synchronization of Two I2C Clock Generators During Arbitration ................................................. 177
USB Controller Block Diagram ......................................................................................... 182
IN/OUT FIFOs ............................................................................................................ 186
Modulation ................................................................................................................ 216
I/Q Phases When Transmitting a Zero-Symbol Chip Sequence, tC = 0.5 μs ..................................... 216
Schematic View of the IEEE 802.15.4 Frame Format [1]........................................................... 217
Format of the Frame Control Field (FCF) ............................................................................. 217
List of Figures
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
www.ti.com
23-5.
Frame Data Written to the TXFIFO .................................................................................... 219
23-6.
TX Flow.................................................................................................................... 220
23-7.
Single Transmitted Frame .............................................................................................. 221
23-8.
Transmitted Synchronization Header.................................................................................. 221
23-9.
FCS Hardware Implementation ........................................................................................ 222
.............................................
SFD Signal Timing .......................................................................................................
Filtering Scenarios (Exceptions Generated During Reception) ....................................................
Matching Algorithm for Short and Extended Addresses ............................................................
Interrupts Generated by Source Address Matching .................................................................
Data in RXFIFO for Different Settings .................................................................................
Acknowledge Frame Format ...........................................................................................
Acknowledgment Timing ................................................................................................
Command Strobe Timing ...............................................................................................
Behavior of FIFO and FIFOP Signals .................................................................................
Main FSM .................................................................................................................
FFT of the Random Bytes ..............................................................................................
Histogram of 20 Million Bytes Generated With the RANDOM Instruction ........................................
Running a CSP Program ................................................................................................
Example Hardware Structure for the R* Register Access Mode ..................................................
Mapping of Radio Memory to MCU XDATA Memory Space .......................................................
FIFO Pointers .............................................................................................................
PN7 Whitening ...........................................................................................................
CC2500-Compatible Whitening ........................................................................................
CRC Module ..............................................................................................................
Air Interface Packet Format for Basic Mode..........................................................................
Air Interface Packet Format for Auto Mode...........................................................................
Bits of 9-Bit Header ......................................................................................................
Bits of 10-Bit Header ....................................................................................................
Structure of Packets in the Rx FIFO...................................................................................
Structure of Packets in the Tx FIFO ...................................................................................
Timing of Packets in Rx Tasks .........................................................................................
Timing of Packets in Tx Tasks .........................................................................................
Complete Appended Packet ............................................................................................
23-10. Single Received Frame and Transmitted Acknowledgement Frame
23-11.
23-12.
23-13.
23-14.
23-15.
23-16.
23-17.
23-18.
23-19.
23-20.
23-21.
23-22.
23-23.
23-24.
25-1.
25-2.
25-3.
25-4.
25-5.
25-6.
25-7.
25-8.
25-9.
25-10.
25-11.
25-12.
25-13.
25-14.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
List of Figures
223
224
226
228
229
230
230
231
231
233
235
236
236
240
256
281
281
292
293
294
297
298
298
298
299
300
315
316
319
11
www.ti.com
List of Tables
CC253x Family Overview
0-2.
Register Bit Conventions ................................................................................................. 16
2-1.
SFR Overview.............................................................................................................. 29
2-2.
Overview of XREG Registers ............................................................................................ 32
2-3.
Instruction Set Summary ................................................................................................. 37
2-4.
Instructions That Affect Flag Settings
2-5.
Interrupts Overview........................................................................................................ 41
2-6.
Priority Level Setting ...................................................................................................... 48
2-7.
Interrupt Priority Groups .................................................................................................. 48
2-8.
Interrupt Polling Sequence ............................................................................................... 49
3-1.
Debug Commands
3-2.
Debug Configuration ...................................................................................................... 55
3-3.
3-4.
Debug Status............................................................................................................... 55
Relation Between PCON_IDLE and PM_ACTIVE ......................................................................... 56
3-5.
Flash Lock-Protection Bit Structure Definition ......................................................................... 57
4-1.
Power Modes
6-1.
Example Write Sequence
7-1.
8-1.
8-2.
9-1.
9-2.
10-1.
13-1.
13-2.
17-1.
20-1.
20-2.
20-3.
20-4.
20-5.
20-6.
21-1.
21-2.
22-1.
23-1.
23-2.
23-3.
23-4.
23-5.
23-6.
23-7.
25-1.
25-2.
25-3.
25-4.
25-5.
12
................................................................................................
0-1.
..................................................................................
........................................................................................................
15
40
53
.............................................................................................................. 61
................................................................................................ 74
Peripheral I/O Pin Mapping .............................................................................................. 81
DMA Trigger Sources ..................................................................................................... 98
DMA Configuration Data Structure ...................................................................................... 99
Initial Compare Output Values (Compare Mode) .................................................................... 107
Frequency Error Calculation for 38-kHz Carrier ..................................................................... 111
Initial Compare Output Values (Compare Mode) .................................................................... 122
Values Showing How Different Temperatures Relate to BATTMON_VOLTAGE for a Typical Device ....... 140
Values for A and B (for a Typical Device) When Using the Battery monitor for Temperature Monitoring ... 141
Commonly Used Baud-Rate Settings for 32 MHz System Clock .................................................. 158
Slave Transmitter Mode ................................................................................................. 172
Slave Receiver Mode .................................................................................................... 173
Master Transmitter Mode ............................................................................................... 175
Master Receiver Mode .................................................................................................. 176
Miscellaneous States .................................................................................................... 178
Clock Rates Defined at 32 MHz........................................................................................ 179
USB Interrupt Flags Interrupt-Enable Mask Registers .............................................................. 183
FIFO Sizes for EP 1–5 .................................................................................................. 186
Internal Registers ........................................................................................................ 203
Frame Filtering and Source Matching Memory Map ................................................................ 214
IEEE 802.15.4-2006 Symbol-to-Chip Mapping....................................................................... 216
FSM State Mapping ..................................................................................................... 236
Instruction Set Summary ................................................................................................ 242
Register Overview ....................................................................................................... 255
Registers That Require Update From Their Default Value ......................................................... 256
Register-Bit Access Modes ............................................................................................. 256
Radio RAM Pages ....................................................................................................... 280
Commands to FIFO via RFST Register............................................................................... 283
Access to FIFO Registers............................................................................................... 284
RAM-Based Registers ................................................................................................... 285
Address Structure for Auto Mode ...................................................................................... 289
List of Tables
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
www.ti.com
25-6.
Address Structure for Basic Mode ..................................................................................... 290
25-7.
RAM-Based Registers in RAM Page 5................................................................................ 291
25-8.
Register Settings for Different CRCs .................................................................................. 294
25-9.
Register Settings for Some Commonly Used CRCs, Assuming Initialization With All 1s ...................... 295
25-10. Supported Modulation Formats, Data Rates, and Deviations ...................................................... 296
25-11. Segments for Holding ACK Payload for Each Address Entry ...................................................... 300
25-12. Commands From MCU to LL Engine via RFST Register
..........................................................
302
25-13. Timer 2 Capture Settings ............................................................................................... 304
25-14. End-of-Task Causes ..................................................................................................... 304
25-15. Recommended RAM Register Settings for Start Tone.............................................................. 306
25-16. Interrupt and Counter Operation for Received Messages .......................................................... 307
25-17. Interrupt and Counter Operation for Received Messages .......................................................... 308
25-18. End-of-Receive Tasks ................................................................................................... 310
25-19. Interrupt and Counter Operation for Received ACK Packets ...................................................... 312
25-20. End-of-Transmit Tasks .................................................................................................. 313
25-21. Additional Reasons for End-of-Transmit on Clear-Channel Tasks ................................................ 315
25-22. Packet-Sniffer Modes of Operation .................................................................................... 318
25-23. XREG Register Overview ............................................................................................... 319
25-24. Registers That Should Be Updated From Their Default Value, Bit Rates 1 Mbps and Lower ................. 320
25-25. Registers That Should Be Updated From Their Default Value, Bit Rate 2 Mbps
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
...............................
List of Tables
321
13
Preface
SWRU191D – April 2009 – Revised March 2013
Read This First
About This Manual
The CC2540/41 is a cost-effective, low-power, and true system-on-chip (SoC) solution for Bluetooth lowenergy applications. It enables robust BLE master or slave nodes to be built with very low total bill-ofmaterial costs. The CC2540/41 combines the excellent performance of a leading RF transceiver with an
industry-standard enhanced 8051 MCU, in-system programmable flash memory, 8-KB RAM, and many
other powerful supporting features and peripherals. The CC2540/41 is suited for systems where very low
power consumption is required. Very low-power sleep modes are available. Short transition times between
operating modes further enable low power consumption.
The CC2540 comes in two different versions: CC2540F128 and CC2540F256, with 128 KB and 256 KB of
flash memory, respectively.
The CC2541 comes in two different versions: CC2541F128 and CC2541F256, with 128 KB and 256 KB of
flash memory, respectively.
The CC2541F128/F256 comes in two different versions: CC2541F128/F256, with 128 and 256 KB of flash
memory, respectively.
Combined with the Bluetooth low-energy protocol stack from Texas Instruments, the
CC2540F128/CC2540F256 and CC2541F128/CC2541F256 constitute the market’s most comprehensive
single-mode Bluetooth low energy solution.
The CC253x System-on-Chip solution for 2.4 GHz is suitable for a wide range of applications. These can
easily be built on top of the IEEE 802.15.4 based standard protocols ( RemoTI™ network protocol, TIMAC
software, and Z-Stack™ software for ZigBee® compliant solutions) or on top of the proprietary SimpliciTI™
network protocol. The usage is, however, not limited to these protocols alone. The CC253x family is, e.g.
also suitable for 6LoWPAN and Wireless HART implementations.
Each chapter of this manual describes details of a module or peripheral; however, not all features are
present on all devices. To see the differences regarding features, see Table 0-1 in the Devices section.
For detailed technical numbers, such as power consumption and RF performance, see the device-specific
data sheet (Appendix C).
Related Documentation and Software From Texas Instruments
Related documentation (e.g., the CC2530 data sheet http://www-s.ti.com/sc/techlit/swrs081 and CC2540
data sheet http://www-s.ti.com/sc/techlit/swrs084) can be found in Appendix C.
For more information regarding software that can be used with the CC253x/CC2540/41 System-on-Chip
solution (e.g., SmartRF™ software for radio performance and functionality evaluation), see Chapter 27,
which also contains more information regarding the RemoTI network protocol, the SimpliciTI network
protocol, the TIMAC software, the Z-Stack software, and the BLE stack software.
SmartRF, RemoTI, SimpliciTI, Z-Stack are trademarks of Texas Instruments.
Bluetooth is a registered trademark of Bluetooth SIG, Inc.
Microsoft, Windows are trademarks of Microsoft Corporation.
ZigBee is a registered trademark of ZigBee Alliance.
14
Read This First
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
FCC Warning
www.ti.com
FCC Warning
This equipment generates, uses, and can radiate radio frequency energy and has not been tested for
compliance with the limits of computing devices pursuant to subpart J of part 15 of FCC rules, which are
designed to provide reasonable protection against radio frequency interference. Operation of this
equipment in other environments may cause interference with radio communications, in which case the
user at his own expense will be required to take whatever measures may be required to correct this
interference.
If You Need Assistance
All technical support is channeled through the TI Product Information Centers (PIC) - www.ti.com/support.
To send an E-mail request, please enter your contact information, along with your request at the following
link – PIC request form.
Also visit the Low Power RF, ZigBee, and Bluetooth low-energy sections of the TI E2E Community
(www.ti.com/lprf-forum), where you can easily get in touch with other CC253x and CC2540/41 users and
find FAQs, Design Notes, Application Notes, Videos, and so forth.
Glossary
Abbreviations used in this user guide can be found in Appendix A.
Devices
The CC253x System-on-Chip solution family consists of several devices. The following table provides a
device overview and points out the differences regarding memory sizes and peripherals. For a complete
feature list of any of the devices, see the corresponding data sheet (Appendix C).
Table 0-1. CC253x Family Overview
Feature
CC2530F32/F64/
F128/F256
CC2531F128/F256
CC2533F32/F64/F96
CC2540F128/
F256
CC2541F128/
F256
FLASH_SIZE
32 KB/64 KB/128
KB/256 KB
128 KB/256 KB
32 KB/64 KB/96 KB
128 KB/256 KB
128 KB/256 KB
SRAM_SIZE
8 KB/8 KB/8 KB/8
KB
8 KB/8 KB
4 KB/4 KB/6 KB
8 KB
8 kB
Not included
USB
Not included
Included
Not included
Included
ADC
Included
Included
Not included
Included
Included
Battery monitor
Not included
Not included
Included
Not included
Not included
I2C
Not included
Not included
Included
Not included
Included
Operational
amplifier
Included
Included
Not included
Included
Not included
Analog comparator
Included
Included
Not included
Included
Included
Legend:
FLASH_SIZE – The size of the flash
SRAM_SIZE – The size of the SRAM
Register Conventions
Each SFR and XREG register is described in a separate table, where each table title contains the
following information in the format indicated:
For SFR registers: REGISTER NAME (SFR address) – register description
For XREG registers: REGISTER NAME (XDATA address) – register description
Each table has five columns to describe the different register fields as described in the following:
Column 1 – Bit: Denotes which bits of the register are described/addressed in the specific row
Column 2 – Name: Specific name of the register field
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Read This First
15
Register Conventions
www.ti.com
Column 3 – Reset: Reset/initial value of the register field
Column 4 – R/W: Key indicating the accessibility of the bits in the field (see Table 0-2 for more details)
Column 5 – Description: More details about the register field, and often a description of the functions of
the different values
In the register descriptions, each register field is shown with a symbol (R/W) indicating the access mode of
the register field. The register values are always given in binary notation unless prefixed by 0x, which
indicates hexadecimal notation.
Table 0-2. Register Bit Conventions
SYMBOL
16
Read This First
ACCESS MODE
R/W
Read/write
R
Read-only
R0
Read as 0
R1
Read as 1
W
Write-only
W0
Write as 0
W1
Write as 1
H0
Hardware clear
H1
Hardware set
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 1
SWRU191D – April 2009 – Revised March 2013
Introduction
As mentioned in the preface, the CC253x/CC2540/41 device family provides solutions for a wide range of
applications. In order to help the user to develop these applications, this user's guide focuses on the
usage of the different building blocks of the CC253x/CC2540/41 device family. For detailed device
descriptions, complete feature lists, and performance numbers, see the device-specific data sheet
(Appendix C).
In order to provide easy access to relevant information, the following subsections guide the reader to the
different chapters in this guide.
Topic
1.1
1.2
...........................................................................................................................
Page
Overview .......................................................................................................... 18
Applications ..................................................................................................... 23
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Introduction
17
Overview
1.1
www.ti.com
Overview
The block diagrams in Figure 1-1 and Figure 1-2 show the different building blocks of the CC253x and
CC2540/41 device family. Not all features and functions of all modules or peripherals are present on all
devices of the CC253x/CC2540/41; hence, see the device-specific data sheet for a device-specific block
diagram.
VDD (2 V–3.6 V)
ON-CHIP VOLTAGE
REGULATOR
POWER-ON RESET
BROWNOUT
BATTERY MONITOR (CC2533 ONLY)
WATCHDOG
TIMER
RESET
XOSC_Q2
32-MHz
CRYSTAL OSC
XOSC_Q1
P2_4
32.768-kHz
CRYSTAL OSC
P2_3
P2_2
DEBUG
INTERFACE
P2_1
SLEEP TIMER
POWER MANAGEMENT CONTROLLER
CLOCK MUX
and
CALIBRATION
HIGHSPEED
RC-OSC
SFR Bus
RESET_N
32-kHz
RC-OSC
CC2531
1-KB
FIFO SRAM
DP
DM
USB
PHY
USB
P2_0
P1_7
DCOUPL
PDATA
P1_6
8051 CPU
CORE
P1_5
P1_4
P1_3
8-KB SRAM
XRAM
IRAM
SFR
MEMORY
ARBITER
32/64/128/256-KB
FLASH
P1_2
P1_1
DMA
P1_0
P0_7
IRQ CTRL
FLASH CTRL
P0_6
P0_5
ANALOG COMPARATOR
P0_4
RADIO REGISTERS
CSMA/CA STROBE PROCESSOR
CC2530/CC2531
SFR Bus
12-BIT D-S
ADC
RADIO DATA INTERFACE
AES
ENCRYPTION
AND
DECRYPTION
DEMODULATOR
AND AGC
USART 0
RECEIVE
USART 1
MODULATOR
FIFO and FRAME CONTROL
P0_0
OP-AMP
SYNTH
P0_1
FREQUENCY
SYNTHESIZER
P0_2
I/O CONTROLLER
P0_3
TRANSMIT
TIMER 1 (16-Bit)
TIMER 2
(IEEE 802.15.4 MAC TIMER)
RF_P
DIGITAL
RF_N
TIMER 3 (8-Bit)
ANALOG
MIXED
TIMER 4 (8-Bit)
B0301-03
Figure 1-1. CC253x Block Diagram
18
Introduction
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Overview
www.ti.com
XOSC_Q2
32-MHz
CRYSTAL OSC
XOSC_Q1
P2_4
32.768-kHz
CRYSTAL OSC
P2_3
P2_2
HIGHSPEED
RC-OSC
DEBUG
INTERFACE
P2_1
DCOUPL
POWER-ON RESET
BROWN OUT
CLOCK MUX
and
CALIBRATION
SFR Bus
RESET
VDD (2 V–3.6 V)
ON-CHIP VOLTAGE
REGULATOR
WATCHDOG
TIMER
RESET_N
SLEEP TIMER
32-kHz
RC-OSC
POWER MANAGEMENT CONTROLLER
P2_0
PDATA
P1_7
P1_6
XRAM
8051 CPU
CORE
P1_5
IRAM
P1_4
SFR
MEMORY
ARBITRATOR
FLASH
FLASH
P1_3
P1_2
DMA
P1_1
UNIFIED
P1_0
IRQ CTRL
FLASH CTRL
P0_7
P0_6
ANALOG COMPARATOR
P0_5
1 KB SRAM
FIFOCTRL
Radio Arbiter
P0_4
OP-AMP
P0_2
RADIO REGISTERS
AES
ENCRYPTION
AND
DECRYPTION
DS
ADC
AUDIO/DC
Link Layer Engine
SFR Bus
P0_0
I/O CONTROLLER
P0_1
DEMODULATOR
SYNTH
P0_3
MODULATOR
USART 0
RECEIVE
TIMER 1 (16-Bit)
TIMER 2
(BLE LL TIMER)
FREQUENCY
SYNTHESIZER
USART 1
TRANSMIT
TIMER 3 (8-Bit)
RF_P
TIMER 4 (8-Bit)
USB_N
USB_P
RF_N
DIGITAL
ANALOG
USB
PHY
USB
MIXED
B0301a-055
Figure 1-2. CC2540 Block Diagram
The modules can be roughly divided into one of three categories: CPU and memory related modules;
modules related to peripherals, clocks, and power management; and radio-related modules.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Introduction
19
Overview
www.ti.com
RESET
XOSC_Q2
WATCHDOG TIMER
XOSC_Q1
CLOCK MUX and
CALIBRATION
32.768-kHz
CRYSTAL OSC
P2_4
P2_3
P2_2
DEBUG
INTERFACE
P2_1
DCOUPL
POWER-ON RESET
BROWN OUT
32-MHZ
CRYSTAL OSC
HIGH SPEED
RC-OSC
SFR bus
RESET_N
VDD (2 V–3.6 V)
ON-CHIP VOLTAGE
REGULATOR
SLEEP TIMER
POWER MGT. CONTROLLER
32-kHz
RC-OSC
P2_0
PDATA
P1_7
P1_6
P1_5
RAM
SRAM
FLASH
FLASH
XRAM
8051 CPU
CORE
IRAM
P1_4
SFR
MEMORY
ARBITRATOR
P1_3
P1_2
UNIFIED
DMA
P1_1
P1_0
IRQ
CTRL
ANALOG COMPARATOR
P0_4
P0_3
P0_2
P0_1
P0_0
FIFOCTRL
OP-
DS ADC
AUDIO / DC
Radio Arbiter
P0_5
I/O CONTROLLER
P0_6
AES
ENCRYPTION
and
DECRYPTION
Link Layer Engine
2
I C
SCL
SFR bus
DEMODULATOR
SDA
1-KB SRAM
RADIO
REGISTERS
SYNTH
P0_7
FLASH CTRL
MODULATOR
USART 1
RECEIVE
TIMER 1 (16-Bit)
TIMER 2
(BLE LL TIMER)
FREQUENCY
SYNTHESIZER
USART 0
TRANSMIT
TIMER 3 (8-bit)
RF_P RF_N
TIMER 4 (8-bit)
DIGITAL
ANALOG
MIXED
Figure 1-3. CC2541 Block Diagram
20
Introduction
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Overview
www.ti.com
1.1.1 CPU and Memory
The 8051 CPU core is a single-cycle 8051-compatible core . It has three different memory access buses
(SFR, DATA, and CODE/XDATA) with single-cycle access to SFR, DATA, and the main SRAM. It also
includes a debug interface and an 18-input extended interrupt unit. The detailed functionality of the CPU
and the memory is addressed in Chapter 2.
The interrupt controller services a total of 18 interrupt sources, divided into six interrupt groups, each of
which is associated with one of four interrupt priorities. Any interrupt service request is serviced also when
the device is in idle mode by going back to active mode. Some interrupts can also wake up the device
from sleep mode (when in sleep mode, the device is in one of the three low-power modes PM1, PM2, or
PM3); see Chapter 4 for more details.
The memory arbiter is at the heart of the system, as it connects the CPU and DMA controller with the
physical memories and all peripherals through the SFR bus. The memory arbiter has four memory access
points, access of which can map to one of three physical memories: SRAM, flash memory, and
XREG/SFR registers. It is responsible for performing arbitration and sequencing between simultaneous
memory accesses to the same physical memory.
The 4/6/8 KB SRAM maps to the DATA memory space and to parts of the XDATA memory spaces. The
SRAM is an ultralow-power SRAM that retains its contents in all power modes . This is an important
feature for low-power applications.
The 32/64/96/128/256 KB flash block provides in-circuit programmable non-volatile program memory for
the device, and maps into the CODE and XDATA memory spaces. In addition to holding program code
and constants, the non-volatile memory allows the application to save data that must be preserved such
that it is available after restarting the device. Using this feature one can, e.g., use saved network-specific
data to avoid the need for a full start-up and network find-and-join process.
1.1.2 Clocks and Power Management
The digital core and peripherals are powered by a 1.8-V low-dropout voltage regulator (Chapter 26).
Additionally, the CC253x/CC2540/41 contains a power management functionality that allows the use of
different low-power modes (PM1, PM2, and PM3) for low-power applications with a long battery life (see
Chapter 4 for more details). Five different reset sources exist to reset the device; see Chapter 5 for more
details.
1.1.3 Peripherals
The CC253x/CC2540/41 includes many different peripherals that allow the application designer to develop
advanced applications. Not all peripherals are present on all devices. See Table 0-1 for a listing of which
peripherals are present on each device.
The debug interface (Chapter 3) implements a proprietary two-wire serial interface that is used for incircuit debugging. Through this debug interface, it is possible to perform an erasure of the entire flash
memory, control which oscillators are enabled, stop and start execution of the user program, execute
supplied instructions on the 8051 core, set code breakpoints, and single-step through instructions in the
code. Using these techniques, it is possible to perform in-circuit debugging and external flash
programming elegantly.
The device contains flash memory for storage of program code. The flash memory is programmable from
the user software and through the debug interface (as mentioned previously). The flash controller
(Chapter 6) handles writing and erasing the embedded flash memory. The flash controller allows pagewise erasure and 4-bytewise programming.
The I/O controller (Chapter 7) is responsible for all general-purpose I/O pins. The CPU can configure
whether peripheral modules control certain pins or whether they are under software control, and if so,
whether each pin is configured as an input or output and if a pullup or pulldown resistor in the pad is
connected. CPU interrupts can be enabled on each pin individually. Each peripheral that connects to the
I/O pins can choose between two different I/O pin locations to ensure flexibility in various applications.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Introduction
21
Overview
www.ti.com
A versatile five-channel DMA controller (Chapter 8) is available in the system, accesses memory using
the XDATA memory space, and thus has access to all physical memories. Each channel (trigger, priority,
transfer mode, addressing mode, source and destination pointers, and transfer count) is configured with
DMA descriptors anywhere in memory. Many of the hardware peripherals (AES core, flash controller,
USARTs, timers, ADC interface) achieve highly efficient operation by using the DMA controller for data
transfers between SFR or XREG addresses and flash/SRAM.
Timer 1 (Chapter 9) is a 16-bit timer with timer/counter/PWM functionality. It has a programmable
prescaler, a 16-bit period value, and five individually programmable counter/capture channels, each with a
16-bit compare value. Each of the counter/capture channels can be used as a PWM output or to capture
the timing of edges on input signals. It can also be configured in IR generation mode, where it counts
Timer 3 periods and the output is ANDed with the output of Timer 3 to generate modulated consumer IR
signals with minimal CPU interaction (see Section 9.9).
Timer 2 (MAC Timer) (Chapter 22) is specially designed for supporting an IEEE 802.15.4 MAC or other
time-slotted protocol in software. The timer has a configurable timer period and a 24-bit overflow counter
that can be used to keep track of the number of periods that have transpired. A 40-bit capture register is
also used to record the exact time at which a start-of-frame delimiter is received/transmitted or the exact
time at which transmission ends, as well as two 16-bit output compare registers and two 24-bit overflow
compare registers that can send various command strobes (start RX, start TX, etc.) at specific times to the
radio modules.
Timer 3 and Timer 4 (Chapter 10) are 8-bit timers with timer/counter/PWM functionality. They have a
programmable prescaler, an 8-bit period value, and one programmable counter channel with an 8-bit
compare value. Each of the counter channels can be used as a PWM output.
The Sleep Timer (Chapter 11) is an ultralow-power timer that counts 32-kHz crystal oscillator or 32-kHz
RC oscillator periods. The Sleep Timer runs continuously in all operating modes except power mode 3
(PM3). Typical applications of this timer are as a real-time counter or as a wake-up timer for coming out of
power mode 1 (PM1) or power mode 2 (PM2).
The ADC (Chapter 12) supports 7 bits (30 kHz bandwidth) to 12 bits (4 kHz bandwidth) of resolution. DC
and audio conversions with up to eight input channels (Port 0) are possible. The inputs can be selected as
single-ended or differential. The reference voltage can be internal, AVDD, or a single-ended or differential
external signal. The ADC also has a temperature-sensor input channel. The ADC can automate the
process of periodic sampling or conversion over a sequence of channels.
The battery monitor (Chapter 13) (CC2533 only) enables simple voltage monitoring in devices that do
not include an ADC. It is designed such that it is accurate in the voltage areas around 2 V, with lower
resolution at higher voltages.
The random-number generator (Chapter 14) uses a 16-bit LFSR to generate pseudorandom numbers,
which can be read by the CPU or used directly by the command strobe processor. It can be seeded with
random data from noise in the radio ADC.
The AES coprocessor (Chapter 15) allows the user to encrypt and decrypt data using the AES algorithm
with 128-bit keys. The core is able to support the security operations required by IEEE 802.15.4 MAC
security, the ZigBee network layer, and the application layer.
A built-in Watchdog Timer (Chapter 16) allows the device to reset itself in case the firmware hangs.
When enabled by software, the Watchdog Timer must be cleared periodically; otherwise, it resets the
device when it times out. It can alternatively be configured for use as a general 32-kHz timer.
USART 0 and USART 1 (Chapter 18) are each configurable as either a SPI master/slave or a UART.
They provide double buffering on both RX and TX and hardware flow control, and are thus well suited to
high-throughput full-duplex applications. Each has its own high-precision baud-rate generator, thus leaving
the ordinary timers free for other uses.
The I2C module (Chapter 20) (CC2533 and CC2541) provides a digital peripheral connection with two pins
and supports both master and slave operation.
The USB 2.0 controller (Chapter 21) (CC2531 and CC2540) operates at Full-Speed, 12 Mbps transfer
rate. The controller has five bidirectional endpoints in addition to control endpoint 0. The endpoints support
Bulk, Interrupt, and Isochronous operation for implementation of a wide range of applications. The 1024
bytes of dedicated, flexible FIFO memory combined with DMA access ensures that a minimum of CPU
involvement is needed for USB communication.
22
Introduction
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Applications
www.ti.com
The operational amplifier (Chapter 18) (CC2530, CC2531, and CC2540) is intended to provide front-end
buffering and gain for the ADC. Both the inputs as well as the output are available on pins, so the
feedback network is fully customizable. A chopper-stabilized mode is available for applications that need
good accuracy with high gain.
The ultralow-power analog comparator (Chapter 19) (CC2530, CC2531, CC2540, and CC2541) enables
applications to wake up from PM2 or PM3 based on an analog signal. Both inputs are brought out to pins;
the reference voltage must be provided externally. The comparator output is mapped into the digital I/O
port and can be treated by the MCU as a regular digital input.
1.1.4 Radio
The CC2540/41 provides a Bluetooth low-energy-compliant radio transceiver. The RF core which controls
the analog and digital radio modules is only indirectly accessible through API commands to the BLE stack.
More details about the CC2540/41 BLE radio can be found in Chapter 24. The CC2541 can also be run in
proprietary modes; more details can be found in Chapter 25.
The CC253x device family provides an IEEE 802.15.4-compliant radio transceiver. The RF Core
controls the analog radio modules. In addition, it provides an interface between the MCU and the radio
which makes it possible to issue commands, read status, and automate and sequence radio events. The
radio also includes a packet-filtering and address-recognition module. More details about the CC253x
radio can be found in Chapter 23.
1.2
Applications
As shown in the overview (Section 1.1), this user's guide focuses on the functionality of the different
modules that are available to build different types of applications based on the CC253x/CC2540/41 device
family. When looking at the complete application development process, additional information is useful.
However, as this information and help is not device-specific (i.e., not unique for the CC253x/CC2540/41
device family), see the additional information sources in the following paragraphs.
The first step is to set up the development environment (HW, tools, etc.) by purchasing a development kit
(see the device-specific product Web site to find links to the relevant development kits). The development
kits come with an out-of-the-box demo and information on how to set up the development environment;
install required drivers (done easily by installing the SmartRF software, Section 27.1), set up the compiler
tool chain, etc. As soon as one has installed the development environment, one is ready to start the
application development.
The easiest way to write the application software is to base the application on one of the available
standard protocols (RemoTI network protocol, Section 27.2; TIMAC software, Section 27.4; Z-Stack
software for ZigBee-compliant solutions, Section 27.5); BLE stack software for Bluetooth low-energycompliant solutions Section 27.6; or the proprietary SimpliciTI network protocol (Section 27.3). They all
come with several sample applications.
For the hardware layout design of the user-specific HW, the designer can find reference designs on the
different product pages (Section B.1). By copying these designs, the designer achieves optimal
performance. The developed HW can then be tested easily using the SmartRF Studio software
(Section 27.1).
In case the final system should not have the expected performance, it is recommended to try out the
developed software on the development kit hardware and see how it works there. To check the userspecific HW, it is a good first step to use SmartRF Studio software to compare the development kit
performance versus the user-specific HW using the same settings.
The user can also find additional information and help by joining the Low-Power RF Online Community
(Section B.2) and by subscribing to the Low-Power RF eNewsletter (Section B.4).
To contact a third-party to help with development or to use modules, check out the Texas Instruments
Low-Power RF Developer Network (Section B.3).
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Introduction
23
Chapter 2
SWRU191D – April 2009 – Revised March 2013
8051 CPU
The System-on-Chip solution is based on an enhanced 8051 core. More details regarding the core,
memory map, instruction set, and interrupts are described in the following subsections.
Topic
...........................................................................................................................
2.1
2.2
2.3
2.4
2.5
24
8051 CPU
8051 CPU Introduction .......................................................................................
Memory ............................................................................................................
CPU Registers ..................................................................................................
Instruction Set Summary ....................................................................................
Interrupts .........................................................................................................
Page
25
25
34
36
40
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU Introduction
www.ti.com
2.1
8051 CPU Introduction
The enhanced 8051 core uses the standard 8051 instruction set. Instructions execute faster than the
standard 8051 due to the following:
• One clock per instruction cycle is used as opposed to 12 clocks per instruction cycle in the standard
8051.
• Wasted bus states are eliminated.
Because an instruction cycle is aligned with memory fetch when possible, most of the single-byte
instructions are performed in a single clock cycle. In addition to the speed improvement, the enhanced
8051 core also includes architectural enhancements:
• A second data pointer
• An extended 18-source interrupt unit
The 8051 core is object-code-compatible with the industry-standard 8051 microcontroller. That is, object
code compiled with an industry-standard 8051 compiler or assembler executes on the 8051 core and is
functionally equivalent. However, because the 8051 core uses a different instruction timing than many
other 8051 variants, existing code with timing loops may require modification. Also, because the peripheral
units such as timers and serial ports differ from those on other 8051 cores, code which includes
instructions using the peripheral-unit SFRs does not work correctly.
Flash prefetching is not enabled by default, but improves CPU performance by up to 33%. This is at the
expense of slightly increased power consumption, but in most cases improves energy consumption as it is
faster. Flash prefetching can be enabled in the FCTL register.
2.2
Memory
The 8051 CPU architecture has four different memory spaces. The 8051 has separate memory spaces for
program memory and data memory. The 8051 memory spaces are the following (see Section 2.2.1 and
Section 2.2.2 for details):
CODE. A read-only memory space for program memory. This memory space addresses 64 KB.
DATA. A read/write data memory space that can be directly or indirectly accessed by a single-cycle CPU
instruction. This memory space addresses 256 bytes. The lower 128 bytes of the DATA memory space
can be addressed either directly or indirectly, the upper 128 bytes only indirectly.
XDATA. A read/write data memory space, access to which usually requires 4–5 CPU instruction cycles.
This memory space addresses 64 KB. Access to XDATA memory is also slower than DATA access, as
the CODE and XDATA memory spaces share a common bus on the CPU core, and instruction prefetch
from CODE can thus not be performed in parallel with XDATA accesses.
SFR. A read/write register memory space which can be directly accessed by a single CPU instruction.
This memory space consists of 128 bytes. For SFR registers whose address is divisible by eight, each bit
is also individually addressable.
The four different memory spaces are distinct in the 8051 architecture, but are partly overlapping in the
device to ease DMA transfers and hardware debugger operation.
How the different memory spaces are mapped onto the three physical memories (flash program memory,
SRAM, and memory-mapped registers) is described in Section 2.2.1 and Section 2.2.2.
2.2.1 Memory Map
The memory map differs from the standard 8051 memory map in two important aspects, as described in
the following paragraphs.
First, in order to allow the DMA controller access to all physical memory and thus allow DMA transfers
between the different 8051 memory spaces, parts of SFR and the DATA memory space are mapped into
the XDATA memory space.
Second, two alternative schemes for CODE memory space mapping can be used. The first scheme is the
standard 8051 mapping where only the program memory (i.e., flash memory) is mapped to CODE
memory space. This mapping is the default after a device reset.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
25
Memory
www.ti.com
The second scheme is used for executing code from SRAM. In this mode, the SRAM is mapped into the
region of 0x8000 through (0x8000 + SRAM_SIZE – 1). The map is shown in Figure 2-2. Executing code
from SRAM improves performance and reduces power consumption.
The upper 32 KB of XDATA is a read-only area called XBANK. Any of the available 32 KB flash banks can
be mapped in here. This gives software access to the whole flash memory. This area is typically used to
store additional constant data.
Details about mapping of all 8051 memory spaces are given in Section 2.2.2.
The memory map showing how the different physical memories are mapped into the CPU memory spaces
is given in Figure 2-1 through Figure 2-3. The number of available flash banks depends on the flash size
option.
0 xFFFF
XBANK
(SELECTABLE 32KB FLASH BANK)
0x8000
0x7 FFF
INFORMATION PAGE
(2KB)
0x 7800
8051 SFR SPACE
8051 DATA SPACE
0x70FF
0x 7080
SFR (128B)
0x63FF
0x 6000
XREG (1KB)
SRAM_SIZE – 1
SRAM SIZE – 256
SRAM
(SRAM_SIZE Bytes)
0x0000
M0097-02
Figure 2-1. XDATA Memory Space (Showing SFR and DATA Mapping)
26
8051 CPU
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Memory
www.ti.com
0 xFFFF
0 xFFFF
Banks 0–7
(Upper 24KB FLASH)
Bank 0–7
(32KB FLASH)
0x8000 + SRAM_SIZE
0x8000 + SRAM_SIZE – 1
SRAM
0x 8000
0x7FFF
0x 8000
0x7FFF
Common Area/Bank 0
(32KB FLASH)
Common Area/Bank 0
(32KB FLASH)
0x 0000
0x 0000
M0098-02
Figure 2-2. CODE Memory Space
M0099-04
Figure 2-3. CODE Memory Space for Running Code
From SRAM
2.2.2 CPU Memory Space
XDATA memory space. The XDATA memory map is given in Figure 2-1.
The SRAM is mapped into address range of 0x0000 through (SRAM_SIZE – 1).
The XREG area is mapped into the 1 KB address range (0x6000–0x63FF). These registers are additional
registers, effectively extending the SFR register space. Some peripheral registers and most of the radio
control and data registers are mapped in here.
The SFR registers are mapped into address range (0x7080–0x70FF).
The flash information page (2 KB) is mapped into the address range (0x7800–0x7FFF). This is a read-only
area and contains various information about the device.
The upper 32 KB of the XDATA memory space (0x8000–0xFFFF) is a read-only flash code bank (XBANK)
and can be mapped to any of the available flash banks using the MEMCTR.XBANK[2:0] bits.
The mapping of flash memory, SRAM, and registers to XDATA allows the DMA controller and the CPU
access to all the physical memories in a single unified address space.
Writing to unimplemented areas in the memory map (shaded in the figure) has no effect. Reading from
unimplemented areas returns 0x00. Writes to read-only regions, i.e., flash areas, are ignored.
CODE memory space. The CODE memory space is 64 KB and is divided into a common area
(0x0000–0x7FFF) and a bank area (0x8000–0xFFFF) as shown in Figure 2-2. The common area is
always mapped to the lower 32 KB of the physical flash memory (bank 0). The bank area can be mapped
to any of the available 32-KB flash banks (from 0 to 7). The number of available flash banks depends on
the flash size option. Use the flash-bank-select register, FMAP, to select the flash bank. On 32-KB
devices, no flash memory can be mapped into the bank area. Reads from this region return 0x00 on these
devices.
To allow program execution from SRAM, it is possible to map the available SRAM into the lower range of
the bank area from 0x8000 through (0x8000 + SRAM_SIZE – 1). The rest of of the currently selected bank
is still mapped into the address range from (0x8000 + SRAM_SIZE) through 0xFFFF). Set the
MEMCTR.XMAP bit to enable this feature.
DATA memory space. The 8-bit address range of DATA memory is mapped into the upper 256 bytes of
the SRAM, i.e., the address range from (SRAM_SIZE – 256) through (SRAM_SIZE – 1).
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
27
Memory
www.ti.com
SFR memory space. The 128-entry hardware register area is accessed through this memory space. The
SFR registers are also accessible through the XDATA address space at the address range
(0x7080–0x70FF). Some CPU-specific SFR registers reside inside the CPU core and can only be
accessed using the SFR memory space and not through the duplicate mapping into XDATA memory
space. These specific SFR registers are listed in SFR Registers.
2.2.3 Physical Memory
RAM. All devices contain static RAM. At power on, the content of RAM is undefined. RAM content is
retained in all power modes.
Flash Memory. The on-chip flash memory is primarily intended to hold program code and constant data.
The flash memory has the following features:
• Page size: 1 KB or 2 KB (details are given in the data sheet of the device.)
• Flash-page erase time: 20 ms
• Flash-chip (mass) erase time: 20 ms
• Flash write time (4 bytes): 20 μs
• Data retention (at room temperature): 100 years
• Program/erase endurance: 20,000 cycles
The flash memory is organized as a set of 1 or 2 KB pages. The 16 bytes of the upper available page
contain page-lock bits and the debug-lock bit. There is one lock bit for each page, except the lock-bit page
which is implicitly locked when not in debug mode. When the lock bit for a page is 0, it is impossible to
erase/write that page. When the debug lock bit is 0, most of the commands on the debug interface are
ignored. The primary purpose of the debug lock bit is to protect the contents of the flash against read-out.
The Flash Controller is used to write and erase the contents of the flash memory.
When the CPU reads instructions and constants from flash memory, it fetches the instructions through a
cache. Four bytes of instructions and four bytes of constant data are cached, at 4-byte boundaries. That
is, when the CPU reads from address 0x00F1 for example, bytes 0x00F0–0x00F3 are cached. A separate
prefetch unit is capable of prefetching 4 additional bytes of instructions. The cache is provided mainly to
reduce power consumption by reducing the amount of time the flash memory is accessed. The cache may
be disabled with the FCTL.CM[1:0] register bits. Doing so increases power consuption and is not
recommended. The execution time from flash is not cycle-accurate when using the default cache mode
and the cache mode with prefetch, i.e., one cannot determine exactly the number of clock cycles a set of
instructions takes. To obtain cycle-accurate execution, enable the real-time cache mode and ensure all
DMA transfers have low priority. The prefetch mode improves performance by up to 33%, at the expense
of increased power consumption due to wasted flash reads. Typically, performance improves by
15%–20%. Total energy, however, may decrease (depending on the application) due to fewer wasted
clock cycles waiting for the flash to return instructions/data. This is very application-dependent and
requires the use of power modes to be effective.
The Information Page is a 2 KB read-only region that stores various device information. Among other
things, it contains for IEEE 802.15.4 or Bluetooth low energy compliant devices a unique IEEE address
from the TI range of addresses. For CC253x, this is a 64-bit IEEE address stored with least-significant
byte first at XDATA address 0x780C. For CC2540/41, this is a 48-bit IEEE address stored with leastsignificant byte first at XDATA address 0x780E.
SFR Registers. The special function registers (SFRs) control several of the features of the 8051 CPU
core and/or peripherals. Many of the 8051 core SFRs are identical to the standard 8051 SFRs. However,
there are additional SFRs that control features that are not available in the standard 8051. The additional
SFRs are used to interface with the peripheral units and RF transceiver.
Table 2-1 shows the addresses of all SFRs in the device. The 8051 internal SFRs are shown with gray
background, whereas the other SFRs are the SFRs specific to the device.
NOTE: All internal SFRs (shown with gray background in Table 2-1), can only be accessed through
SFR space, as these registers are not mapped into XDATA space. One exception is the port
registers (P0, P1, and P2) which are readable from XDATA.
28
8051 CPU
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Memory
www.ti.com
Table 2-1. SFR Overview
Register
Name
SFR
Address
Module
Description
ADCCON1
0xB4
ADC
ADC control 1
ADCCON2
0xB5
ADC
ADC control 2
ADCCON3
0xB6
ADC
ADC control 3
ADCL
0xBA
ADC
ADC data low
ADCH
0xBB
ADC
ADC data high
RNDL
0xBC
ADC
Random number generator data low
RNDH
0xBD
ADC
Random number generator data high
ENCDI
0xB1
AES
Encryption/decryption input data
ENCDO
0xB2
AES
Encryption/decryption output data
ENCCS
0xB3
AES
Encryption/decryption control and status
P0
0x80
CPU
Port 0. Readable from XDATA (0x7080)
SP
0x81
CPU
Stack pointer
DPL0
0x82
CPU
Data pointer 0 low byte
DPH0
0x83
CPU
Data pointer 0 high byte
DPL1
0x84
CPU
Data pointer 1 low byte
DPH1
0x85
CPU
Data pointer 0 high byte
PCON
0x87
CPU
Power mode control
TCON
0x88
CPU
Interrupt flags
P1
0x90
CPU
Port 1. Readable from XDATA (0x7090)
DPS
0x92
CPU
Data pointer select
S0CON
0x98
CPU
Interrupt flags 2
IEN2
0x9A
CPU
Interrupt enable 2
S1CON
0x9B
CPU
Interrupt flags 3
P2
0xA0
CPU
Port 2. Readable from XDATA (0x70A0)
IEN0
0xA8
CPU
Interrupt enable 0
IP0
0xA9
CPU
Interrupt priority 0
IEN1
0xB8
CPU
Interrupt enable 1
IP1
0xB9
CPU
Interrupt priority 1
IRCON
0xC0
CPU
Interrupt flags 4
PSW
0xD0
CPU
Program status Word
ACC
0xE0
CPU
Accumulator
IRCON2
0xE8
CPU
Interrupt flags 5
B
0xF0
CPU
B register
DMAIRQ
0xD1
DMA
DMA interrupt flag
DMA1CFGL
0xD2
DMA
DMA channel 1–4 configuration address low
DMA1CFGH
0xD3
DMA
DMA channel 1–4 configuration address high
DMA0CFGL
0xD4
DMA
DMA channel 0 configuration address low
DMA0CFGH
0xD5
DMA
DMA channel 0 configuration address high
DMAARM
0xD6
DMA
DMA channel armed
DMAREQ
0xD7
DMA
DMA channel start request and status
—
0xAA
—
Reserved
—
0x8E
—
Reserved
—
0x99
—
Reserved
—
0xB0
—
Reserved
—
0xB7
—
Reserved
—
0xC8
—
Reserved
P0IFG
0x89
IOC
Port 0 interrupt status flag
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
29
Memory
www.ti.com
Table 2-1. SFR Overview (continued)
Register
Name
30
SFR
Address
Module
Description
P1IFG
0x8A
IOC
Port 1 interrupt status flag
P2IFG
0x8B
IOC
Port 2 interrupt status flag
PICTL
0x8C
IOC
Port pins interrupt mask and edge
P0IEN
0xAB
IOC
Port 0 interrupt mask
P1IEN
0x8D
IOC
Port 1 interrupt mask
P2IEN
0xAC
IOC
Port 2 interrupt mask
P0INP
0x8F
IOC
Port 0 input mode
PERCFG
0xF1
IOC
Peripheral I/O control
APCFG
0xF2
IOC
Analog peripheral I/O configuration
P0SEL
0xF3
IOC
Port 0 function select
P1SEL
0xF4
IOC
Port 1 function select
P2SEL
0xF5
IOC
Port 2 function select
P1INP
0xF6
IOC
Port 1 input mode
P2INP
0xF7
IOC
Port 2 input mode
P0DIR
0xFD
IOC
Port 0 direction
P1DIR
0xFE
IOC
Port 1 direction
P2DIR
0xFF
IOC
Port 2 direction
PMUX
0xAE
IOC
Power-down signal mux
MPAGE
0x93
MEMORY
Memory page select
MEMCTR
0xC7
MEMORY
Memory system control
FMAP
0x9F
MEMORY
Flash-memory bank mapping
RFIRQF1
0x91
RF
RF interrupt flags MSB
RFD
0xD9
RF
RF data
RFST
0xE1
RF
RF command strobe
RFIRQF0
0xE9
RF
RF interrupt flags LSB
RFERRF
0xBF
RF
RF error interrupt flags
ST0
0x95
ST
Sleep Timer 0
ST1
0x96
ST
Sleep Timer 1
ST2
0x97
ST
Sleep Timer 2
STLOAD
0xAD
ST
Sleep-timer load status
SLEEPCMD
0xBE
PMC
Sleep-mode control command
SLEEPSTA
0x9D
PMC
Sleep-mode control status
CLKCONCMD
0xC6
PMC
Clock control command
CLKCONSTA
0x9E
PMC
Clock control status
T1CC0L
0xDA
Timer 1
Timer 1 channel 0 capture/compare value low
T1CC0H
0xDB
Timer 1
Timer 1 channel 0 capture/compare value high
T1CC1L
0xDC
Timer 1
Timer 1 channel 1 capture/compare value low
T1CC1H
0xDD
Timer 1
Timer 1 channel 1 capture/compare value high
T1CC2L
0xDE
Timer 1
Timer 1 channel 2 capture/compare value low
T1CC2H
0xDF
Timer 1
Timer 1 channel 2 capture/compare value high
T1CNTL
0xE2
Timer 1
Timer 1 counter low
T1CNTH
0xE3
Timer 1
Timer 1 counter high
T1CTL
0xE4
Timer 1
Timer 1 control and status
T1CCTL0
0xE5
Timer 1
Timer 1 channel 0 capture/compare control
T1CCTL1
0xE6
Timer 1
Timer 1 channel 1 capture/compare control
T1CCTL2
0xE7
Timer 1
Timer 1 channel 2 capture/compare control
T1STAT
0xAF
Timer 1
Timer 1 status
8051 CPU
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Memory
www.ti.com
Table 2-1. SFR Overview (continued)
Register
Name
SFR
Address
Module
Description
T2CTRL
0x94
Timer 2
Timer 2 control
T2EVTCFG
0x9C
Timer 2
Timer 2 event configuration
T2IRQF
0xA1
Timer 2
Timer 2 interrupt flags
T2M0
0xA2
Timer 2
Timer 2 multiplexed register 0
T2M1
0xA3
Timer 2
Timer 2 multiplexed register 1
T2MOVF0
0xA4
Timer 2
Timer 2 multiplexed overflow register 0
T2MOVF1
0xA5
Timer 2
Timer 2 multiplexed overflow register 1
T2MOVF2
0xA6
Timer 2
Timer 2 multiplexed overflow register 2
T2IRQM
0xA7
Timer 2
Timer 2 interrupt mask
T2MSEL
0xC3
Timer 2
Timer 2 multiplex select
T3CNT
0xCA
Timer 3
Timer 3 counter
T3CTL
0xCB
Timer 3
Timer 3 control
T3CCTL0
0xCC
Timer 3
Timer 3 channel 0 compare control
T3CC0
0xCD
Timer 3
Timer 3 channel 0 compare value
T3CCTL1
0xCE
Timer 3
Timer 3 channel 1 compare control
T3CC1
0xCF
Timer 3
Timer 3 channel 1 compare value
T4CNT
0xEA
Timer 4
Timer 4 counter
T4CTL
0xEB
Timer 4
Timer 4 control
T4CCTL0
0xEC
Timer 4
Timer 4 channel 0 compare control
T4CC0
0xED
Timer 4
Timer 4 channel 0 compare value
T4CCTL1
0xEE
Timer 4
Timer 4 channel 1 compare control
T4CC1
0xEF
Timer 4
Timer 4 channel 1 compare value
TIMIF
0xD8
TMINT
Timers 1/3/4 joint interrupt mask/flags
U0CSR
0x86
USART 0
USART 0 control and status
U0DBUF
0xC1
USART 0
USART 0 receive/transmit data buffer
U0BAUD
0xC2
USART 0
USART 0 baud-rate control
U0UCR
0xC4
USART 0
USART 0 UART control
U0GCR
0xC5
USART 0
USART 0 generic control
U1CSR
0xF8
USART 1
USART 1 control and status
U1DBUF
0xF9
USART 1
USART 1 receive/transmit data buffer
U1BAUD
0xFA
USART 1
USART 1 baud-rate control
U1UCR
0xFB
USART 1
USART 1 UART control
U1GCR
0xFC
USART 1
USART 1 generic control
WDCTL
0xC9
WDT
Watchdog Timer control
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
31
Memory
www.ti.com
XREG Registers. The XREG registers are additional registers in the XDATA memory space. These
registers are mainly used for radio configuration and control. For more details regarding each register, see
the corresponding module/peripheral chapter. Table 2-2 gives a descriptive overview of the register
address space.
Table 2-2. Overview of XREG Registers
XDATA Address
Description
—
Radio registers (see CC253x Radio Section 23.15 or CC2540
Radio Section 24.1 or CC2541 Radio Section 25.12 for
complete list)
MONMUX
Battery monitor MUX (CC2533)
OPAMPMC
Operational amplifier mode control (CC2530/CC2531)
0x61AD
OPAMPMC
Operational amplifier mode control (CC2540)
0x6200–0x622B
—
USB registers (see Section 21.12 for complete list)
0x6230
I2CCFG
I2C control
0x6231
I2CSTAT
I2C status
0x6232
I2CDATA
I2C data
0x6233
I2CADDR
I2C own slave address
0x6234
I2CWC
Wrapper control
0x6235
I2CIO
GPIO
0x6243
OBSSEL0
Observation output control register 0
0x6244
OBSSEL1
Observation output control register 1
0x6245
OBSSEL2
Observation output control register 2
0x6246
OBSSEL3
Observation output control register 3
0x6247
OBSSEL4
Observation output control register 4
0x6248
OBSSEL5
Observation output control register 5
0x6249
CHVER
Chip version
0x624A
CHIPID
Chip identification
0x624B
TR0
Test register 0
0x6260
DBGDATA
Debug interface write data
0x6262
SRCRC
Sleep reset CRC
0x6264
BATTMON
Battery monitor
0x6265
IVCTRL
Analog control register
0x6270
FCTL
Flash control
0x6271
FADDRL
Flash address low
0x6272
FADDRH
Flash address high
0x6273
FWDATA
Flash write data
0x6276
CHIPINFO0
Chip information byte 0
0x6277
CHIPINFO1
Chip information byte 1
0x6281
IRCTL
Timer 1 IR generation control
0x6290
CLD
Clock-loss detection
0x62A0
T1CCTL0
Timer 1 channel 0 capture/compare control (additional XREG
mapping of SFR register)
0x62A1
T1CCTL1
Timer 1 channel 1 capture/compare control (additional XREG
mapping of SFR register)
0x62A2
T1CCTL2
Timer 1 channel 2 capture/compare control (additional XREG
mapping of SFR register)
0x62A3
T1CCTL3
Timer 1 channel 3 capture/compare control
0x62A4
T1CCTL4
Timer 1 channel 4 capture/compare control
0x62A6
T1CC0L
Timer 1 channel 0 capture/compare value low (additional XREG
mapping of SFR register)
0x62A7
T1CC0H
Timer 1 channel 0 capture/compare value high (additional
XREG mapping of SFR register)
0x6000–0x61FF
0x61A6
32
Register Name
8051 CPU
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Memory
www.ti.com
Table 2-2. Overview of XREG Registers (continued)
XDATA Address
Register Name
Description
0x62A8
T1CC1L
Timer 1 channel 1 capture/compare value low (additional XREG
mapping of SFR register)
0x62A9
T1CC1H
Timer 1 channel 1 capture/compare value high (additional
XREG mapping of SFR register)
0x62AA
T1CC2L
Timer 1 channel 2 capture/compare value low (additional XREG
mapping of SFR register)
0x62AB
T1CC2H
Timer 1 channel 2 capture/compare value high (additional
XREG mapping of SFR register)
0x62AC
T1CC3L
Timer 1 channel 3 capture/compare value low
0x62AD
T1CC3H
Timer 1 channel 3 capture/compare value high
0x62AE
T1CC4L
Timer 1 channel 4 capture/compare value low
0x62AF
T1CC4H
Timer 1 channel 4 capture/compare value high
0x62B0
STCC
Sleep Timer capture control
0x62B1
STCS
Sleep Timer capture status
0x62B2
STCV0
Sleep Timer capture value byte 0
0x62B3
STCV1
Sleep Timer capture value byte 1
0x62B4
STCV2
Sleep Timer capture value byte 2
0x62C0
OPAMPC
Operational amplifier control
0x62C1
OPAMPS
Operational amplifier status
0x62D0
CMPCTL
Analog comparator control and status
2.2.4 XDATA Memory Access
The MPAGE register is used during instructions MOVX A,@Ri and MOVX @Ri,A. MPAGE gives the 8 mostsignificant address bits, whereas the register Ri gives the 8 least-significant bits.
In some 8051 implementations, this type of XDATA access is performed using P2 to give the mostsignificant address bits. Existing software may therefore have to be adapted to make use of MPAGE
instead of P2.
MPAGE (0x93) – Memory Page Select
Bit
7:0
Name
MPAGE[7:0]
Reset
0x00
R/W
R/W
Description
Memory page, high-order bits of address in MOVX instruction
2.2.5 Memory Arbiter
The memory arbiter handles CPU and DMA access to all physical memory except the CPU internal
registers. When an access conflict between the CPU and DMA occurs, the memory arbiter stalls one of
the bus masters so that the conflict is resolved.
The control registers MEMCTR and FMAP are used to control various aspects of the memory subsystem.
The MEMCTR and FMAP registers are described as follows.
MEMCTR.XMAP must be set to enable program execution from RAM.
The flash-bank map register, FMAP, controls mapping of physical 32-KB code banks to the program
address region 0x8000–0xFFFF in CODE memory space.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
33
CPU Registers
www.ti.com
MEMCTR (0xC7) – Memory Arbiter Control
Bit
7:4
3
—
XMAP
Name
2:0
XBANK[2:0]
Reset
0000
0
R/W
R0
R/W
000
R/W
Reset
0000 0
001
R/W
R0
R/W
Description
Reserved
XDATA map to code. When this bit is set, the SRAM XDATA region, from 0x0000
through (SRAM_SIZE – 1) is mapped into the CODE region from 0x8000 through
(0x8000 + SRAM_SIZE – 1). This enables execution of program code from RAM.
0: SRAM map into CODE feature disabled
1: SRAM map into CODE feature enabled
XDATA bank select. Controls which code bank of the physical flash memory is
mapped into the XDATA region (0x8000–0xFFFF). When set to 0, the root bank is
mapped in.
Valid settings depend on the flash size for the device. Writing an invalid setting is
ignored, i.e., no update to XBANK[2:0] is performed.
32-KB version: 0 only (i.e, the root bank is always mapped in.)
64-KB version: 0–1
96-KB version: 0–2
128-KB version: 0–3
256-KB version: 0–7
FMAP (0x9F) – Flash Bank Map
Bit
7:3
2:0
2.3
Name
—
MAP[2:0]
Description
Reserved
Flash bank map. Controls which bank is mapped into the bank area of the CODE
memory space (0x8000–0xFFFF). When set to 0, the root bank is mapped in. Valid
settings depend on the flash size for the device. Writing an invalid setting is ignored,
i.e., no update to MAP[2:0] is performed.
32-KB version: No value can be written. Bank area is only used for running program
code from SRAM. See MEMCTR.XMAP.
64-KB version: 0–1
96-KB version: 0–2
128-KB version: 0–3
256-KB version: 0–7
CPU Registers
This section describes the internal registers found in the CPU.
2.3.1 Data Pointers
Two data pointers, DPTR0 and DPTR1, exist to accelerate the movement of data blocks to/from memory.
The data pointers are generally used to access CODE or XDATA space. For example:
MOVC A,@A+DPTR
MOV A,@DPTR.
The data pointer select bit, bit 0 in the data pointer select register DPS, chooses which data pointer is the
active one during execution of an instruction that uses the data pointer, e.g., in one of the preceding
instructions.
The data pointers are two bytes wide, consisting of the following SFRs:
• DPTR0–DPH0:DPL0
• DPTR1–DPH1:DPL1
DPH0 (0x83) – Data Pointer-0 High Byte
Bit
7:0
Name
DPH0[7:0]
Reset
0x00
R/W
R/W
Data pointer-0, high byte
Description
R/W
R/W
Data pointer-0, low byte
DPL0 (0x82) – Data Pointer-0 Low Byte
Bit
7:0
34
Name
DPL0[7:0]
8051 CPU
Reset
0x00
Description
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CPU Registers
www.ti.com
DPH1 (0x85) – Data Pointer-1 High Byte
Bit
7:0
Name
DPH1[7:0]
Reset
0x00
R/W
R/W
Data pointer-1, high byte
Description
R/W
R/W
Data pointer-1, low byte
DPL1 (0x84) – Data Pointer-1 Low Byte
Bit
7:0
Name
DPL1[7:0]
Reset
0x00
Description
DPS (0x92) – Data-Pointer Select
Bit
7:1
0
Name
–
DPS
Reset
0000 000
0
R/W
R0
R/W
Description
Reserved
Data pointer select. Selects active data pointer.
0: DPTR0
1: DPTR1
2.3.2 Registers R0–R7
There are four register banks (not to be confused with CODE memory space banks that only apply to flash
memory organization) of eight registers each. These register banks are mapped in the DATA memory
space at addresses 0x00–0x07, 0x08–0x0F, 0x10–0x17, and 0x18–0x1F. Each register bank contains the
eight 8-bit registers R0–R7. The register bank to be used is selected through the program status word
PSW.RS[1:0]. Register bank 0 uses flip-flops internally for storing the values (SRAM is
bypassed/unused), whereas banks 1–3 use SRAM for storage. This is done to save power. Typically, the
current consumption goes down by approximately 200 μA by using register bank 0 instead of register
banks 1–3.
2.3.3 Program Status Word
The program status word (PSW) contains several bits that show the current state of the CPU. The PSW is
accessible as an SFR, and it is bit-addressable. The PSW is shown as follows and contains the carry flag,
auxiliary carry flag for BCD operations, register-select bits, overflow flag, and parity flag. Two bits in the
PSW are uncommitted and can be used as user-defined status flags.
PSW (0xD0) – Program Status Word
Bit
7
CY
Reset
0
R/W
R/W
6
AC
0
R/W
F0
RS[1:0]
0
00
R/W
R/W
2
OV
0
R/W
1
0
F1
P
0
0
R/W
R/W
5
4:3
Name
Description
Carry flag. Set to 1 when the last arithmetic operation resulted in a carry (during
addition) or borrow (during subtraction); otherwise, cleared to 0 by all arithmetic
operations.
Auxiliary carry flag for BCD operations. Set to 1 when the last arithmetic operation
resulted in a carry into (during addition) or borrow from (during subtraction) the highorder nibble, otherwise cleared to 0 by all arithmetic operations.
User-defined, bit-addressable
Register bank select bits. Selects which set of R7–R0 registers to use from four
possible banks in DATA space.
00: Register bank 0, 0x00–0x07
01: Register bank 1, 0x08–0x0F
10: Register bank 2, 0x10–0x17
11: Register bank 3, 0x18–0x1F
Overflow flag, set by arithmetic operations. Set to 1 when the last arithmetic
operation is a carry (addition), borrow (subtraction), or overflow (multiply or divide).
Otherwise, the bit is cleared to 0 by all arithmetic operations.
User-defined, bit-addressable
Parity flag, parity of accumulator set by hardware to 1 if it contains an odd number of
1s; otherwise it is cleared to 0.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
35
Instruction Set Summary
www.ti.com
2.3.4 Accumulator
ACC is the accumulator. This is the source and destination of most arithmetic instructions, data transfers,
and other instructions. The mnemonic for the accumulator (in instructions involving the accumulator) is A
instead of ACC.
ACC (0xE0) – Accumulator
Bit
7:0
Name
ACC[7:0]
Reset
0x00
R/W
R/W
Description
Accumulator
2.3.5 B Register
The B register is used as the second 8-bit argument during execution of multiply and divide instructions.
When not used for these purposes, it may be used as a scratchpad register to hold temporary data.
B (0xF0) – B Register
Bit
7:0
Name
B[7:0]
Reset
0x00
R/W
R/W
Description
B register. Used in MUL/DIV instructions
2.3.6 Stack Pointer
The stack resides in DATA memory space and grows upwards. The PUSH instruction first increments the
stack pointer (SP) and then copies the byte into the stack. The SP is initialized to 0x07 after a reset, and it
is incremented once to start from location 0x08, which is the first register (R0) of the second register bank.
Thus, in order to use more than one register bank, the SP should be initialized to a different location not
used for data storage.
SP (0x81) – Stack Pointer
Bit
7:0
2.4
Name
SP[7:0]
Reset
0x07
R/W
R/W
Description
Stack pointer
Instruction Set Summary
The 8051 instruction set is summarized in Table 2-3. All mnemonics copyrighted © Intel Corporation,
1980.
The following conventions are used in the instruction set summary:
• Rn – Register R7–R0 of the currently selected register bank
•
•
•
•
•
•
•
•
36
Direct – 8-bit internal data-location address. This can be DATA area (0x00–0x7F) or SFR area
(0x80–0xFF).
@Ri – 8-bit internal data location, DATA area (0x00–0xFF) addressed indirectly through register R1 or
R0
#data – 8-bit constant included in instruction
#data16 – 16-bit constant included in instruction
addr16 – 16-bit destination address. Used by LCALL and LJMP. A branch can be anywhere within the
64 KB CODE memory space.
addr11 – 11-bit destination address. Used by ACALL and AJMP. The branch is within the same 2 KB
page of program memory as the first byte of the following instruction.
rel – Signed (2s-complement) 8-bit offset byte. Used by SJMP and all conditional jumps. Range is –128
to 127 bytes relative to first byte of the following instruction.
bit – Direct addressed bit in DATA area or SFR
8051 CPU
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Instruction Set Summary
www.ti.com
The instructions that affect CPU flag settings located in PSW are listed in Table 2-4. Note that operations
on the PSW register or bits in PSW also affect the flag settings. Also note that the cycle count for many
instructions assumes single-cycle access to the memory element being accessed, i.e., the best-case
situation. This is not always the case. Reads from flash may take 1–3 cycles, for example.
Table 2-3. Instruction Set Summary
Mnemonic
Description
Hex Opcode
Bytes
Cycles
ARITHMETIC OPERATIONS
ADD A,Rn
Add register to accumulator
28–2F
1
1
ADD A,direct
Add direct byte to accumulator
25
2
2
ADD A,@Ri
Add indirect RAM to accumulator
26–27
1
2
ADD A,#data
Add immediate data to accumulator
24
2
2
ADDC A,Rn
Add register to accumulator with carry flag
38–3F
1
1
ADDC A,direct
Add direct byte to A with carry flag
35
2
2
ADDC A,@Ri
Add indirect RAM to A with carry flag
36–37
1
2
ADDC A,#data
Add immediate data to A with carry flag
34
2
2
SUBB A,Rn
Subtract register from A with borrow
98–9F
1
1
SUBB A,direct
Subtract direct byte from A with borrow
95
2
2
SUBB A,@Ri
Subtract indirect RAM from A with borrow
96–97
1
2
SUBB A,#data
Subtract immediate data from A with borrow
94
2
2
INC A
Increment accumulator
04
1
1
INC Rn
Increment register
08–0F
1
2
INC direct
Increment direct byte
05
2
3
INC @Ri
Increment indirect RAM
06–07
1
3
INC DPTR
Increment data pointer
A3
1
1
DEC A
Decrement accumulator
14
1
1
DEC Rn
Decrement register
18–1F
1
2
DEC direct
Decrement direct byte
15
2
3
DEC @Ri
Decrement indirect RAM
16–17
1
3
MUL AB
Multiply A and B
A4
1
5
DIV A
Divide A by B
84
1
5
DA A
Decimal adjust accumulator
D4
1
1
LOGICAL OPERATIONS
ANL A,Rn
AND register to accumulator
58–5F
1
1
ANL A,direct
AND direct byte to accumulator
55
2
2
ANL A,@Ri
AND indirect RAM to accumulator
56–57
1
2
ANL A,#data
AND immediate data to accumulator
54
2
2
ANL direct,A
AND accumulator to direct byte
52
2
3
ANL direct,#data
AND immediate data to direct byte
53
3
4
ORL A,Rn
OR register to accumulator
48–4F
1
1
ORL A,direct
OR direct byte to accumulator
45
2
2
ORL A,@Ri
OR indirect RAM to accumulator
46–47
1
2
ORL A,#data
OR immediate data to accumulator
44
2
2
ORL direct,A
OR accumulator to direct byte
42
2
3
ORL direct,#data
OR immediate data to direct byte
43
3
4
XRL A,Rn
Exclusive OR register to accumulator
68–6F
1
1
XRL A,direct
Exclusive OR direct byte to accumulator
65
2
2
XRL A,@Ri
Exclusive OR indirect RAM to accumulator
66–67
1
2
XRL A,#data
Exclusive OR immediate data to accumulator
64
2
2
XRL direct,A
Exclusive OR accumulator to direct byte
62
2
3
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
37
Instruction Set Summary
www.ti.com
Table 2-3. Instruction Set Summary (continued)
Bytes
Cycles
XRL direct,#data
Mnemonic
Exclusive OR immediate data to direct byte
Description
63
Hex Opcode
3
4
CLR A
Clear accumulator
E4
1
1
CPL A
Complement accumulator
F4
1
1
RL A
Rotate accumulator left
23
1
1
RLC A
Rotate accumulator left through carry
33
1
1
RR A
Rotate accumulator right
03
1
1
RRC A
Rotate accumulator right through carry
13
1
1
SWAP A
Swap nibbles within the accumulator
C4
1
1
MOV A,Rn
Move register to accumulator
E8–EF
1
1
MOV A,direct
Move direct byte to accumulator
E5
2
2
MOV A,@Ri
Move indirect RAM to accumulator
E6–E7
1
2
MOV A,#data
Move immediate data to accumulator
74
2
2
MOV Rn,A
Move accumulator to register
F8–FF
1
2
MOV Rn,direct
Move direct byte to register
A8–AF
2
4
MOV Rn,#data
Move immediate data to register
78–7F
2
2
MOV direct,A
Move accumulator to direct byte
F5
2
3
MOV direct,Rn
Move register to direct byte
88–8F
2
3
MOV direct1,direct2
Move direct byte to direct byte
85
3
4
MOV direct,@Ri
Move indirect RAM to direct byte
86–87
2
4
MOV direct,#data
Move immediate data to direct byte
75
3
3
MOV @Ri,A
Move accumulator to indirect RAM
F6–F7
1
3
MOV @Ri,direct
Move direct byte to indirect RAM
A6–A7
2
5
MOV @Ri,#data
Move immediate data to indirect RAM
76–77
2
3
MOV DPTR,#data16 Load data pointer with a 16-bit constant
90
3
3
MOVC A,@A+DPTR Move code byte relative to DPTR to accumulator
93
1
3
MOVC A,@A+PC
Move code byte relative to PC to accumulator
83
1
3
MOVX A,@Ri
Move external RAM (8-bit address) to A
E2–E3
1
3
MOVX A,@DPTR
Move external RAM (16-bit address) to A
E0
1
3
MOVX @Ri,A
Move A to external RAM (8-bit address)
F2–F3
1
4
MOVX @DPTR,A
Move A to external RAM (16-bit address)
F0
1
4
PUSH direct
Push direct byte onto stack
C0
2
4
POP direct
Pop direct byte from stack
D0
2
3
XCH A,Rn
Exchange register with accumulator
C8–CF
1
2
XCH A,direct
Exchange direct byte with accumulator
C5
2
3
XCH A,@Ri
Exchange indirect RAM with accumulator
C6–C7
1
3
XCHD A,@Ri
Exchange low-order nibble indirect. RAM with A
D6–D7
1
3
DATA TRANSFERS
PROGRAM BRANCHING
38
ACALL addr11
Absolute subroutine call
xxx11
2
6
LCALL addr16
Long subroutine call
12
3
6
RET
Return from subroutine
22
1
4
RETI
Return from interrupt
32
1
4
AJMP addr11
Absolute jump
xxx01
2
3
LJMP addr16
Long jump
02
3
4
SJMP rel
Short jump (relative address)
80
2
3
JMP @A+DPTR
Jump indirect relative to the DPTR
73
1
2
JZ rel
Jump if accumulator is zero
60
2
3
8051 CPU
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Instruction Set Summary
www.ti.com
Table 2-3. Instruction Set Summary (continued)
Bytes
Cycles
JNZ rel
Mnemonic
Jump if accumulator is not zero
Description
70
Hex Opcode
2
3
JC rel
Jump if carry flag is set
40
2
3
JNC
Jump if carry flag is not set
50
2
3
JB bit,rel
Jump if direct bit is set
20
3
4
JNB bit,rel
Jump if direct bit is not set
30
3
4
JBC bit,direct rel
Jump if direct bit is set and clear bit
10
3
4
CJNE A,direct rel
Compare direct byte to A and jump if not equal
B5
3
4
CJNE A,#data rel
Compare immediate to A and jump if not equal
B4
3
4
CJNE Rn,#data rel
Compare immediate to reg. and jump if not equal
B8–BF
3
4
CJNE @Ri,#data rel
Compare immediate to indirect and jump if not equal
B6–B7
3
4
DJNZ Rn,rel
Decrement register and jump if not zero
D8–DF
1
3
DJNZ direct,rel
Decrement direct byte and jump if not zero
D5
3
4
NOP
No operation
00
1
1
Boolean VARIABLE OPERATIONS
CLR C
Clear carry flag
C3
1
1
CLR bit
Clear direct bit
C2
2
3
SETB C
Set carry flag
D3
1
1
SETB bit
Set direct bit
D2
2
3
CPL C
Complement carry flag
B3
1
1
CPL bit
Complement direct bit
B2
2
3
ANL C,bit
AND direct bit to carry flag
82
2
2
ANL C,/bit
AND complement of direct bit to carry
B0
2
2
ORL C,bit
OR direct bit to carry flag
72
2
2
ORL C,/bit
OR complement of direct bit to carry
A0
2
2
MOV C,bit
Move direct bit to carry flag
A2
2
2
MOV bit,C
Move carry flag to direct bit
92
2
3
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
39
Interrupts
www.ti.com
Table 2-4. Instructions That Affect Flag Settings
Instruction
CY
OV
AC
ADD
x
x
x
ADDC
x
x
x
SUBB
x
x
x
MUL
0
x
–
DIV
0
x
–
DA
x
–
–
RRC
x
–
–
RLC
x
–
–
SETB C
1
–
–
CLR C
x
–
–
CPLC
x
–
–
ANL C,bit
x
–
–
ANL C,/bit
x
–
–
ORL C,bit
x
–
–
ORL C,/bit
x
–
–
MOV C,bit
x
–
–
CJNE
x
–
–
(1)
2.5
(1)
0 = set to 0, 1 = set to 1, x = set to 0/1, – = not affected
Interrupts
The CPU has 18 interrupt sources. Each source has its own request flag located in a set of interrupt flag
SFR registers. Each interrupt requested by the corresponding flag can be individually enabled or disabled.
The definitions of the interrupt sources and the interrupt vectors are given in Table 2-5.
The interrupts are grouped into a set of priority-level groups with selectable priority levels.
The interrupt-enable registers are described in Section 2.5.1 and the interrupt priority settings are
described in Section 2.5.3.
40
8051 CPU
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Interrupts
www.ti.com
2.5.1 Interrupt Masking
Each interrupt can be individually enabled or disabled by the interrupt-enable bits in the interrupt-enable
SFRs IEN0, IEN1, and IEN2. The CPU interrupt-enable SFRs are described as follows and summarized
in Table 2-5.
Note that some peripherals have several events that can generate the interrupt request associated with
that peripheral. This applies to Port 0, Port 1, Port 2, Timer 1, Timer 2, Timer 3, Timer 4, DMA controller,
and Radio. These peripherals have interrupt mask bits for each internal interrupt source in the
corresponding SFR or XREG registers.
In
1.
2.
3.
4.
order to enable any of the interrupts, the following steps must be taken:
Clear interrupt flags.
Set individual interrupt-enable bit in the peripherals SFR register, if any.
Set the corresponding individual interrupt-enable bit in the IEN0, IEN1, or IEN2 register to 1.
Enable global interrupt by setting the EA bit in IEN0 to 1.
5. Begin the interrupt service routine at the corresponding vector address of that interrupt. See Table 2-5
for addresses.
Figure 2-4 gives a complete overview of all interrupt sources and associated control and state registers.
Shaded boxes in Figure 2-4 are interrupt flags that are automatically cleared by hardware when the
interrupt service routine is called.
indicates a one-shot, either due to the level source or due to edge
shaping. Interrupts missing this are to be treated as level-triggered (apply to ports P0, P1, and P2). The
switch boxes are shown in the default state, and
or
indicates rising- or falling-edge detection, i.e., at
what time instance the interrupt is generated. As a general rule for pulsed or edge-shaped interrupt
sources, one should clear CPU interrupt flag registers prior to clearing the source flag bit, if available, for
flags that are not automatically cleared. For level sources, one must clear the source prior to clearing the
CPU flag.
Note that when clearing source interrupt flags in a register that contains several flags, interrupts may be
lost if a read-modify-write operation is done (even in a single assembly instruction), as it will also clear
interrupt flags that became active between the read and write operation. The source interrupt flags (with
the exception of the USB controller interrupt flags) have the access mode R/W0. This means that writing 1
to a bit has no effect, so 1 should be written to an interrupt flag that is not to be cleared. For instance, to
clear the TIMER2_OVF_PERF bit (bit 3) of T2IRQF in C code, one should do:
T2IRQF = ~(1 <<
3);
and not:
T2IRQF &= ~(1 << 3); //
wrong!
Table 2-5. Interrupts Overview
Interrupt
Number
(2)
Interrupt
Name
Interrupt
Vector
Interrupt Mask,
CPU
Interrupt Flag, CPU
0
RF core error situation
RFERR
0x03
IEN0.RFERRIE
TCON.RFERRIF (1)
1
ADC end of conversion
ADC
0x0B
IEN0.ADCIE
TCON.ADCIF (1)
2
USART 0 RX complete
URX0
0x13
IEN0.URX0IE
TCON.URX0IF (1)
3
USART 1 RX complete
URX1
0x1B
IEN0.URX1IE
TCON.URX1IF (1)
4
AES encryption/decryption complete
ENC
0x23
IEN0.ENCIE
S0CON.ENCIF
5
Sleep Timer compare
ST
0x2B
IEN0.STIE
IRCON.STIF
6
Port-2 inputs/USB/I2C
P2INT
0x33
IEN2.P2IE
IRCON2.P2IF (2)
7
USART 0 TX complete
UTX0
0x3B
IEN2.UTX0IE
IRCON2.UTX0IF
8
DMA transfer complete
DMA
0x43
IEN1.DMAIE
IRCON.DMAIF
9
Timer 1 (16-bit) capture/compare/overflow
T1
0x4B
IEN1.T1IE
IRCON.T1IF (1)
10
(1)
Description
Timer 2
T2
0x53
IEN1.T2IE
IRCON.T2IF
(2)
(1) (2)
Hardware-cleared when interrupt service routine is called
Additional IRQ mask and IRQ flag bits exist.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
41
Interrupts
www.ti.com
Table 2-5. Interrupts Overview (continued)
Interrupt
Number
11
42
Description
Timer 3 (8-bit) capture/compare/overflow
Interrupt
Name
T3
Interrupt
Vector
0x5B
Interrupt Mask,
CPU
IEN1.T3IE
Interrupt Flag, CPU
IRCON.T3IF (1)
(2)
(1) (2)
12
Timer 4 (8-bit) capture/compare/overflow
T4
0x63
IEN1.T4IE
IRCON.T4IF
13
Port 0 inputs
P0INT
0x6B
IEN1.P0IE
IRCON.P0IF (2)
14
USART 1 TX complete
UTX1
0x73
IEN2.UTX1IE
IRCON2.UTX1IF
15
Port 1 inputs
P1INT
0x7B
IEN2.P1IE
IRCON2.P1IF (2)
16
RF general interrupts
RF
0x83
IEN2.RFIE
S1CON.RFIF (2)
17
Watchdog overflow in timer mode
WDT
0x8B
IEN2.WDTIE
IRCON2.WDTIF
8051 CPU
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
P0[7:0]
P1[7:0]
USBCIF
USBOIF
USBIIF
P2[4:0]
USB_DP
0
1
1
0
0
0
1
PICTL.P0ICON
1
PICTL.P1ICON
USBCIE
USBOIE
USBIIE
2
PICTL.P2ICON
1
0
T3OVFIF
T4CH0IF
T4OVFIF
P0IFG
P1IFG
5
4
3
T4CH1IF
TIMIF
2
T3CH1IF
T3CH0IF
7:0
7:0
P0IEN
P1IEN
T4CTL.OVFIM
T4CCTL0.IM
T4CCTL1.IM
T3CTL.OVFIM
T3CCTL0.IM
T3CCTL1.IM
5:0
T2IRQM
7:0
7:0
P2IEN[4:0]
P2IEN[5]
TIMIF.T1OVFIM
P2IFG[4:0]
TIMIF
RFIRQM1
RFIRQM0
T1CCTL{0-4}.IM
7:0
7:0
P2IFG.DPIF
T2IRQF
T1STAT.OVFIF
T1STAT[4:0]
RFIRQF1
RFIRQF0
WDT
P0INT
ST
P1INT
T4
ENC
UTX1
T3
URX1
UTX0
T2
URX0
P2INT
T1
ADC
DMA
RF
RFERR
IT1
IT0
STIF
ircon.6
ENCIF_1
ENCIF_0
RFIF_1
RFIF_0
WDTIF
P0IF
P1IF
T4IF
UTX1IF
T3IF
URX1IF
UTX0IF
T2IF
URX0IF
P2IF
T1IF
ADCIF
DMAIF
RFERRIF
EA
WDTIE
P0IE
STIE
P1IE
T4IE
ENCIE
UTX1IE
T3IE
URX1IE
UTX0IE
T2IE
URX0IE
P2IE
T1IE
ADCIE
DMAIE
RFIE
RFERRIE
IP1_5
IP0_5
IP1_4
IP0_4
IP1_3
IP0_3
IP1_2
IP0_2
IP1_1
IP0_1
IP1_0
IP0_0
Interrupt Priority Bits
B0302-02
www.ti.com
Interrupts
Figure 2-4. Interrupt Overview
8051 CPU
43
Interrupts
www.ti.com
IEN0 (0xA8) – Interrupt Enable 0
Bit
7
Name
EA
Reset
0
R/W
R/W
6
5
–
STIE
0
0
R0
R/W
4
ENCIE
0
R/W
3
URX1IE
0
R/W
2
URX0IE
0
R/W
1
ADCIE
0
R/W
RFERRIE
0
R/W
0
Description
Disables all interrupts.
0: No interrupt is acknowledged.
1: Each interrupt source is individually enabled or disabled by setting its
corresponding enable bit.
Reserved. Read as 0
Sleep Timer interrupt enable
0: Interrupt disabled
1: Interrupt enabled
AES encryption/decryption interrupt enable
0: Interrupt disabled
1: Interrupt enabled
USART 1 RX interrupt enable
0: Interrupt disabled
1: Interrupt enabled
USART0 RX interrupt enable
0: Interrupt disabled
1: Interrupt enabled
ADC interrupt enable
0: Interrupt disabled
1: Interrupt enabled
RF core error interrupt enable
0: Interrupt disabled
1: Interrupt enabled
IEN1 (0xB8) – Interrupt Enable 1
Bit
7:6
5
–
P0IE
Reset
00
0
R/W
R0
R/W
4
T4IE
0
R/W
3
T3IE
0
R/W
2
T2IE
0
R/W
1
T1IE
0
R/W
DMAIE
0
R/W
0
44
Name
8051 CPU
Description
Reserved. Read as 0
Port 0 interrupt enable
0: Interrupt disabled
1: Interrupt enabled
Timer 4 interrupt enable
0: Interrupt disabled
1: Interrupt enabled
Timer 3 interrupt enable
0: Interrupt disabled
1: Interrupt enabled
Timer 2 interrupt enable
0: Interrupt disabled
1: Interrupt enabled
Timer 1 interrupt enable
0: Interrupt disabled
1: Interrupt enabled
DMA transfer interrupt enable
0: Interrupt disabled
1: Interrupt enabled
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Interrupts
www.ti.com
IEN2 (0x9A) – Interrupt Enable 2
Bit
7:6
5
–
WDTIE
Reset
00
0
R/W
R0
R/W
4
P1IE
0
R/W
3
UTX1IE
0
R/W
2
UTX0IE
0
R/W
1
P2IE
0
R/W
RFIE
0
R/W
0
Name
Description
Reserved. Read as 0
Watchdog Timer interrupt enable
0: Interrupt disabled
1: Interrupt enabled
Port 1 interrupt enable
0: Interrupt disabled
1: Interrupt enabled
USART 1 TX interrupt enable
0: Interrupt disabled
1: Interrupt enabled
USART 0 TX interrupt enable
0: Interrupt disabled
1: Interrupt enabled
Port 2 and USB interrupt enable
0: Interrupt disabled
1: Interrupt enabled
RF general interrupt enable
0: Interrupt disabled
1: Interrupt enabled
2.5.2 Interrupt Processing
When an interrupt occurs, the CPU vectors to the interrupt-vector address as shown in Table 2-5. Once an
interrupt service has begun, it can be interrupted only by a higher-priority interrupt. The interrupt service is
terminated by an RETI (return-from-interrupt instruction). When an RETI is performed, the CPU returns to
the instruction that would have been next when the interrupt occurred.
When the interrupt condition occurs, the CPU also indicates this by setting an interrupt flag bit in the
interrupt flag registers. This bit is set regardless of whether the interrupt is enabled or disabled. If the
interrupt is enabled when an interrupt flag is set, then on the next instruction cycle, the interrupt is
acknowledged by hardware, forcing an LCALL to the appropriate vector address.
Interrupt response requires a varying amount of time, depending on the state of the CPU when the
interrupt occurs. If the CPU is performing an interrupt service with equal or greater priority, the new
interrupt is pending until it becomes the interrupt with highest priority. In other cases, the response time
depends on current instruction. The fastest possible response to an interrupt is seven machine cycles.
This includes one machine cycle for detecting the interrupt and six cycles to perform the LCALL.
NOTE:
If an interrupt is disabled and the interrupt flag is polled, the 8051 assembly instruction JBC
must not be used to poll the interrupt flag and clear it when set. If the JBC instruction is
used, the interrupt flag may be re-asserted immediately.
NOTE:
If the assembly instruction XCH A, IEN0 is used to clear the global interrupt enable flag
EA, the CPU may enter the interrupt routine on the cycle following this instruction. If that
happens, the interrupt routine is executed with EA set to 0, which may delay the service of
higher-priority interrupts.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
45
Interrupts
www.ti.com
TCON (0x88) – Interrupt Flags
Bit
7
Name
URX1IF
Reset
0
R/W
R/W
H0
6
5
–
ADCIF
0
0
R/W
R/W
H0
4
3
–
URX0IF
0
0
R/W
R/W
H0
2
IT1
1
R/W
1
RFERRIF
0
R/W
H0
0
IT0
1
R/W
Reset
0000 00
0
R/W
R/W
R/W
0
R/W
Description
USART 1 RX interrupt flag. Set to 1 when USART 1 RX interrupt occurs and cleared
when CPU vectors to the interrupt service routine.
0: Interrupt not pending
1: Interrupt pending
Reserved
ADC interrupt flag. Set to 1 when ADC interrupt occurs and cleared when CPU
vectors to the interrupt service routine.
0: Interrupt not pending
1: Interrupt pending
Reserved
USART 0 RX interrupt flag. Set to 1 when USART 0 interrupt occurs and cleared
when CPU vectors to the interrupt service routine.
0: Interrupt not pending
1:Interrupt pending
Reserved. Must always be set to 1. Setting a zero enables low-level interrupt
detection, which is almost always the case (one-shot when interrupt request is
initiated).
RF core error interrupt flag. Set to 1 when RFERR interrupt occurs and cleared
when CPU vectors to the interrupt service routine.
0: Interrupt not pending
1: Interrupt pending
Reserved. Must always be set to 1. Setting a zero enables low-level interrupt
detection, which is almost always the case (one-shot when interrupt request is
initiated).
S0CON (0x98) – Interrupt Flags 2
Bit
7:2
1
0
Name
–
ENCIF_1
ENCIF_0
Description
Reserved
AES interrupt. ENC has two interrupt flags, ENCIF_1 and ENCIF_0. Setting one of
these flags requests interrupt service. Both flags are set when the AES coprocessor
requests the interrupt.
0: Interrupt not pending
1: Interrupt pending
AES interrupt. ENC has two interrupt flags, ENCIF_1 and ENCIF_0. Setting one of
these flags requests interrupt service. Both flags are set when the AES coprocessor
requests the interrupt.
0: Interrupt not pending
1: Interrupt pending
S1CON (0x9B) – Interrupt Flags 3
Bit
7:2
1
–
RFIF_1
Reset
0000 00
0
R/W
R/W
R/W
0
RFIF_0
0
R/W
46
Name
8051 CPU
Description
Reserved
RF general interrupt. RF has two interrupt flags, RFIF_1 and RFIF_0. Setting one of
these flags requests interrupt service. Both flags are set when the radio requests the
interrupt.
0: Interrupt not pending
1: Interrupt pending
RF general interrupt. RF has two interrupt flags, RFIF_1 and RFIF_0. Setting one of
these flags requests interrupt service. Both flags are set when the radio requests the
interrupt.
0: Interrupt not pending
1: Interrupt pending
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Interrupts
www.ti.com
IRCON (0xC0) – Interrupt Flags 4
Bit
7
STIF
Name
Reset
0
R/W
R/W
6
5
–
P0IF
0
0
R/W
R/W
4
T4IF
0
R/W
H0
3
T3IF
0
R/W
H0
2
T2IF
0
R/W
H0
1
T1IF
0
R/W
H0
0
DMAIF
0
R/W
Description
Sleep Timer interrupt flag
0: Interrupt not pending
1: Interrupt pending
Must be written 0. Writing a 1 always enables the interrupt source.
Port 0 interrupt flag
0: Interrupt not pending
1: Interrupt pending
Timer 4 interrupt flag. Set to 1 when Timer 4 interrupt occurs and cleared when
vectors to the interrupt service routine.
0: Interrupt not pending
1: Interrupt pending
Timer 3 interrupt flag. Set to 1 when Timer 3 interrupt occurs and cleared when
vectors to the interrupt service routine.
0: Interrupt not pending
1: Interrupt pending
Timer 2 interrupt flag. Set to 1 when Timer 2 interrupt occurs and cleared when
vectors to the interrupt service routine.
0: Interrupt not pending
1: Interrupt pending
Timer 1 interrupt flag. Set to 1 when Timer 1 interrupt occurs and cleared when
vectors to the interrupt service routine.
0: Interrupt not pending
1: Interrupt pending
DMA-complete interrupt flag
0: Interrupt not pending
1: Interrupt pending
CPU
CPU
CPU
CPU
IRCON2 (0xE8) – Interrupt Flags 5
Bit
7:5
4
–
WDTIF
Name
Reset
000
0
R/W
R/W
R/W
3
P1IF
0
R/W
2
UTX1IF
0
R/W
1
UTX0IF
0
R/W
0
P2IF
0
R/W
Description
Reserved
Watchdog Timer interrupt flag
0: Interrupt not pending
1: Interrupt pending
Port 1 interrupt flag
0: Interrupt not pending
1: Interrupt pending
USART 1 TX interrupt flag
0: Interrupt not pending
1: Interrupt pending
USART 0 TX interrupt flag
0: Interrupt not pending
1: Interrupt pending
Port 2 interrupt flag
0: Interrupt not pending
1: Interrupt pending
2.5.3 Interrupt Priority
The interrupts are grouped into six interrupt priority groups, and the priority for each group is set by
registers IP0 and IP1. In order to assign a higher priority to an interrupt, i.e., to its interrupt group, the
corresponding bits in IP0 and IP1 must be set as shown in Table 2-6.
The interrupt priority groups with assigned interrupt sources are shown in Table 2-7. Each group is
assigned one of four priority levels. While an interrupt service request is in progress, it cannot be
interrupted by a lower- or same-level interrupt.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8051 CPU
47
Interrupts
www.ti.com
In the case when interrupt requests of the same priority level are received simultaneously, the polling
sequence shown in Table 2-8 is used to resolve the priority of each request. Note that the polling
sequence in Figure 2-4 is the algorithm found in Table 2-8, not that polling is among the IP bits as listed in
the figure.
IP1 (0xB9) – Interrupt Priority 1
Bit
7:6
5
4
3
2
1
0
Name
–
IP1_IPG5
IP1_IPG4
IP1_IPG3
IP1_IPG2
IP1_IPG1
IP1_IPG0
Reset
00
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Reserved
Interrupt group
Interrupt group
Interrupt group
Interrupt group
Interrupt group
Interrupt group
Description
Reset
00
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Reserved
Interrupt group
Interrupt group
Interrupt group
Interrupt group
Interrupt group
Interrupt group
5, priority
4, priority
3, priority
2, priority
1, priority
0, priority
control
control
control
control
control
control
bit
bit
bit
bit
bit
bit
1, see
1, see
1, see
1, see
1, see
1, see
Table
Table
Table
Table
Table
Table
2-7: Interrupt Priority
2-7: Interrupt Priority
2-7: Interrupt Priority
2-7: Interrupt Priority
2-7: Interrupt Priority
2-7: Interrupt Priority
Groups
Groups
Groups
Groups
Groups
Groups
2-7: Interrupt Priority
2-7: Interrupt Priority
2-7: Interrupt Priority
2-7: Interrupt Priority
2-7: Interrupt Priority
2-7: Interrupt Priority
Groups
Groups
Groups
Groups
Groups
Groups
IP0 (0xA9) – Interrupt Priority 0
Bit
7:6
5
4
3
2
1
0
Name
–
IP0_IPG5
IP0_IPG4
IP0_IPG3
IP0_IPG2
IP0_IPG1
IP0_IPG0
Description
5, priority
4, priority
3, priority
2, priority
1, priority
0, priority
control
control
control
control
control
control
bit
bit
bit
bit
bit
bit
0, see
0, see
0, see
0, see
0, see
0, see
Table
Table
Table
Table
Table
Table
Table 2-6. Priority Level Setting
IP1_x
IP0_x
Priority Level
0
0
0 – lowest
0
1
1
1
0
2
1
1
3 – highest
Table 2-7. Interrupt Priority Groups
Group
48
8051 CPU
Interrupts
IPG0
RFERR
RF
DMA
IPG1
ADC
T1
P2INT
IPG2
URX0
T2
UTX0
IPG3
URX1
T3
UTX1
IPG4
ENC
T4
P1INT
IPG5
ST
P0INT
WDT
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Interrupts
www.ti.com
Table 2-8. Interrupt Polling Sequence
Interrupt Number
Interrupt Name
0
RFERR
16
RF
8
DMA
1
ADC
9
T1
2
URX0
10
T2
3
URX1
11
T3
4
ENC
12
T4
5
ST
13
P0INT
6
P2INT
7
UTX0
14
UTX1
15
P1INT
17
WDT
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Polling sequence
↓
8051 CPU
49
Chapter 3
SWRU191D – April 2009 – Revised March 2013
Debug Interface
The two-wire debug interface allows programming of the on-chip flash, and it provides access to memory
and register contents and debug features such as breakpoints, single-stepping, and register modification.
The debug interface uses I/O pins P2.1 and P2.2 as debug data and debug clock, respectively, during
debug mode. These I/O pins can be used as general-purpose I/O only while the device is not in debug
mode. Thus, the debug interface does not interfere with any peripheral I/O pins.
Topic
3.1
3.2
3.3
3.4
3.5
3.6
50
...........................................................................................................................
Debug Mode .....................................................................................................
Debug Communication ......................................................................................
Debug Commands .............................................................................................
Flash Programming ...........................................................................................
Debug Interface and Power Modes ......................................................................
Registers ..........................................................................................................
Debug Interface
Page
51
51
53
57
57
59
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Mode
www.ti.com
3.1
Debug Mode
Debug mode is entered by forcing two falling-edge transitions on pin P2.2 (debug clock) while the
RESET_N input is held low. When RESET_N is set high, the device is in debug mode.
On entering debug mode, the CPU is in the halted state with the program counter reset to address
0x0000.
While in debug mode, pin P2.1 is the debug-data bidirectional pin, and P2.2 is the debug-clock input pin.
NOTE: Note that the debugger cannot be used with a divided system clock. When running the
debugger, the value of CLKCONCMD.CLKSPD should be set to 000 when CLKCONCMD.OSC = 0 or
to 001 when CLKCONCMD.OSC = 1.
3.2
Debug Communication
The debug interface uses a SPI-like two-wire interface consisting of the P2.1 (debug data) and P2.2
(debug clock) pins. Data is driven on the bidirectional debug-data pin at the positive edge of the debug
clock, and data is sampled on the negative edge of this clock.
The direction of the debug-data pin depends on the command being issued. Data is driven on the positive
edge of the debug clock and sampled on the negative edge. Figure 3-1 shows how data is sampled.
Debug Clock
Debug Data
Data is sampled by the
receiver on the falling
edge of debug clock.
Data is set up on the
rising edge of debug clock.
T0302-01
Figure 3-1. External Debug Interface Timing
The data is byte-oriented and is transmitted MSB-first. A sequence of one byte is shown in Figure 3-2.
Start of Byte
End of Byte
Time
Debug Clock
Debug Data
Bit 7
Bit 5
Bit 6
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
T0303-01
Figure 3-2. Transmission of One Byte
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Interface
51
Debug Communication
www.ti.com
A debug command sequence always starts with the host transmitting a command through the serial
interface. This command encodes the number of bytes containing further parameters to follow, and
whether a response is required. Based on this command, the debug module controls the direction of the
debug data pad. A typical command sequence is shown in Figure 3-3. Note that the debug-data signal is
simplified for the clarity of the figure, not showing each individual bit change. The direction is not explicitly
indicated to the outside world, but must be derived by the host from the command protocol.
Start of
Command
Sequence
Start to
Change
Direction
End of
Command
Sequence
Pad is
Output
Time
Debug
Clock
Debug
Data
Data Pad
Direction
Cmd Byte
Data Byte 1
Data Byte 2
Output Byte
Input
Input
Output
tdir_change
The Level is
Sampled by the
External Device
(Asynchronously)
T0304-01
Figure 3-3. Typical Command Sequence—No Extra Wait for Response
For commands that require a response, there must be a small idle period between the command and the
response to allow the pad to change direction. After the minimum waiting time (tdir_change), the chip indicates
whether it is ready to deliver the response data by pulling the data pad low. The external debugger, which
is sampling the data pad, detects this and begins to clock out the response data. If the data pad is high
after the waiting time, it is an indication to the debugger that the chip is not ready yet. Figure 3-4 shows
how the wait works.
52
Debug Interface
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Commands
www.ti.com
Start of
Command
Sequence
Start to
Change
Direction
Pad Is Output, But
Chip Is Not Ready to
Respond
End of
Command
Sequence
Time
8 Cycles
Debug
Clock
Debug
Data
Data Pad
Direction
Data Byte 1
Cmd Byte
Data Byte 2
Output Byte
Input
Input
Output
tdir_change
tsample_wait
The Level Is Sampled.
Result = Ready
Chip Is Ready to
Provide Response
The Level Is Sampled.
Result = Not Ready
T0305-01
Figure 3-4. Typical Command Sequence. Wait for Response
If the debug interface indicates by pulling the data line high that it is not ready to return data, the external
device must issue exactly eight clock pulses before it samples the ready level again. This must be
repeated until the level is low. The wait cycle is equivalent to reading a byte from the debug interface, but
ignoring the result. Note that the pad starts to change direction on the falling edge of the debug clock.
Thus, the pad driver drives against the driver in the programmer until the programmer changes pad
direction. This duration should be minimized in a programmer implementation.
3.3
Debug Commands
The debug commands are shown in Table 3-1. Some of the debug commands are described in further
detail in the following subsections.
The 3 least-significant bits (the Xs) are don't care values.
Table 3-1. Debug Commands
Command
Instruction
Byte
Additi
onal
Input
Bytes
Output
Bytes
Description
CHIP_ERASE
00010XXX
0
1
Perform flash chip erase (mass erase) and clear lock bits. If any other
command except READ_STATUS is issued, then the use of CHIP_ERASE
is disabled.
Input byte: none
Output byte: Debug status byte. See Table 3-3.
WR_CONFIG
00011XXX
1
1
Write debug configuration data.
Input byte: See Table 3-2 for details.
Output byte: Debug status byte. See Table 3-3.
RD_CONFIG
00100XXX
0
1
Read debug configuration data.
Input byte: none.
Output byte: Returns value set by WR_CONFIG command. See Table 3-2.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Interface
53
Debug Commands
www.ti.com
Table 3-1. Debug Commands (continued)
Command
Instruction
Byte
Additi
onal
Input
Bytes
Output
Bytes
Description
GET_PC
00101XXX
0
2
Return value of 16-bit program counter.
Input byte: none
Output bytes: Returns 2 bytes.
READ_STATUS
00110XXX
0
1
Read status byte.
Input byte: none
Output byte: Debug status byte. See Table 3-3.
SET_HW_BRKPNT 00111XXX
3
1
Set hardware breakpoint.
Input bytes: See Section 3.3.3 for details.
Output byte: Debug status byte. See Table 3-3.
HALT
01000XXX
0
1
Halt CPU operation
Input byte: none
Output byte: Debug status byte. See Table 3-3. If the CPU was already
halted, the output is undefined.
RESUME
01001XXX
0
1
Resume CPU operation. The CPU must be in the halted state for this
command to be run.
Input byte: none
Output byte: Debug status byte. See Table 3-3.
DEBUG_INSTR
01010Xyy
1–3
1
Run debug instruction. The supplied instruction is executed by the CPU
without incrementing the program counter. The CPU must be in halted
state for this command to be run. Note that yy is number of bytes following
the command byte, i.e., how many bytes the CPU instruction has (see
Table 2-3).
Input byte(s): CPU instruction
Output byte: The resulting accumulator register value after the instruction
has been executed
STEP_INSTR
01011XXX
0
1
Step CPU instruction. The CPU executes the next instruction from
program memory and increments the program counter after execution. The
CPU must be in the halted state for this command to be run.
Input byte: none
Output byte: The resulting accumulator register value after the instruction
has been executed
GET_BM
01100XXX
0
1
This command does the same thing as GET_PC, except that it returns the
memory bank. It returns one byte, where the 3 least-significant bits are the
currently used memory bank.
Input byte: none
Output byte: Memory bank (current value of FMAP.MAP)
GET_CHIP_ID
01101XXX
0
2
Return value of 16-bit chip ID and version number.
Input byte: none.
Output bytes: The CHIPID and CHVER register values
BURST_WRITE
10000kkk
2–2049
1
This command writes a sequence of 1–2048 bytes to the DBGDATA
register. Each time the register is updated, a DBG_BW DMA trigger is
generated.
The number of parameters to the BURST_WRITE command is variable.
The number of data bytes in the burst is indicated using the 3 last bits of
the command byte (kkk), and the whole next byte. The command
sequence is shown in Figure 3-5. The burst length is indicated by an 11-bit
value (b10–b0). After these two bytes, the given number of data bytes
must be appended. The value 0 means 2048 data bytes; thus, the smallest
number of bytes to transfer is 1.
Input bytes: Command sequence
Output byte: Debug status byte. See Table 3-3.
54
Debug Interface
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Commands
www.ti.com
Parameter
BURST_WRITE Command
0
1
0
0
0
b10
b9
b8
b7
b6
b5
b4
b3
b2
b1
b0
T0306-01
Figure 3-5. Burst Write Command (First 2 Bytes)
3.3.1 Debug Configuration
The commands WR_CONFIG and RD_CONFIG are used to access the debug-configuration data byte.
The format and description of this configuration data are shown in Table 3-2.
Table 3-2. Debug Configuration
Bit
Name
Reset
7:6 –
Description
00
Reserved
5
SOFT_POWER_MODE
1
When set, the digital regulator is not turned off during PM2 and PM3. If this bit is
cleared, the debug interface is reset during PM2 and PM3.
4
–
0
Reserved
3
TIMERS_OFF
0
Disable timers. Disable timer operation. This overrides the TIMER_SUSPEND bit and its
function.
0: Do not disable timers
1: Disable timers
2
1
DMA_PAUSE
DMA pause. The DMA registers must not be accessed while this bit is set.
0: Enable DMA transfers
1: Pause all DMA transfers
1
1
TIMER_SUSPEND
Suspend timers.
Suspend timers when the chip is halted. The timers are also suspended during debug
instructions. When executing a STEP, the timers receive exactly (or as close as
possible) as many ticks as they would if the program were free-running.
0: Do not suspend timers
1: Suspend timers
0
–
0
Reserved. Always write 0.
3.3.2 Debug Status
A debug-status byte is read using the READ_STATUS command. The format and description of this
debug status is shown in Table 3-3.
The READ_STATUS command is, for example, used for:
• Polling the status of the chip erase (CHIP_ERASE_BUSY) after a CHIP_ERASE command.
• Checking whether the oscillator is stable (OSCILLATOR_STABLE); required for debug commands
HALT, RESUME, DEBUG_INSTR, STEP_REPLACE, and STEP_INSTR.
Table 3-3. Debug Status
Bit
7
Name
CHIP_ERASE_BUSY
Reset
Description
0
Flash chip erase busy
The signal is only high when a chip erase is in progress. It goes high immediately after a
CHIP_ERASE command is received and returns to low when the flash is fully erased.
0: –
1: Chip erase in progress
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Interface
55
Debug Commands
www.ti.com
Table 3-3. Debug Status (continued)
Bit
6
Name
Reset
0
PCON_IDLE
Description
PCON idle. See also Table 3-4.
0: CPU is running. Chip in operational mode controlled by debugger.
1: CPU is not running. Chip is in power mode defined by SLEEPCMD.MODE register
setting. See Section 4.1 - Section 4.3 for details.
5
0
CPU_HALTED
CPU was halted
0: CPU is running.
1: CPU was halted from a breakpoint or from a HALT debug command.
4
0
PM_ACTIVE
Chip is active. Note that PM0 and PM1 are not supported in debug mode. See also
Table 3-4.
0: Chip is in normal operation with CPU running (if not halted).
1: Chip is out of normal operation (active mode) and either in transition up or down from
power mode or stable in the power mode defined by SLEEPCMD.MODE register
setting. See Section 4.1 - Section 4.3 for details.
3
0
HALT_STATUS
Halt status. Returns cause of last CPU halt
0: CPU was halted by HALT debug command
1: CPU was halted by hardware breakpoint
2
Debug interface is locked. Returns value of DBGLOCK bit. See Section 3.4.1.
0
DEBUG_LOCKED
0: Debug interface is not locked.
1: Debug interface is locked.
1
0
OSCILLATOR_STABLE
System clock oscillator stable.
0: Oscillators not stable
1: Oscillators stable
0
0
STACK_OVERFLOW
Stack overflow. This bit indicates when the CPU writes to DATA memory space at
address 0xFF, which is possibly a stack overflow.
0: No stack overflow
1: Stack overflow
Table 3-4. Relation Between
PCON_IDLE
and PM_ACTIVE
Description
PCON_IDLE
PM_ACTIVE
0
0
Chip in normal operation with CPU running (if not halted)
0
1
Chip in transition to start-up from power mode
1
0
Chip in transition to enter power mode
1
1
Chip stable in power mode
3.3.3 Hardware Breakpoints
The debug command SET_HW_BRKPNT is used to set one of the four available hardware breakpoints.
When a hardware breakpoint is enabled, it compares the CPU address bus with the breakpoint. When a
match occurs, the CPU is halted.
When issuing the SET_HW_BRKPNT, the external host must supply three data bytes that define the
hardware breakpoint. The hardware breakpoint itself consists of 19 bits, whereas three bits are used for
control purposes. The format of the three data bytes for the SET_HW_BRKPNT command is as follows.
The first data byte consists of the following:
• Bits 7–6: Unused
• Bits 5–4: Breakpoint number, 0–3
• Bit 3: 1 = enable, 0 = disable
• Bits 2–0: Memory bank bits. Bits 18–16 of hardware breakpoint.
The second data byte consists of bits 15–8 of the hardware breakpoint.
56
Debug Interface
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Flash Programming
www.ti.com
The third data byte consists of bits 7–0 of the hardware breakpoint. Thus, the second and third data bytes
set the CPU CODE address at which to stop execution.
3.4
Flash Programming
Programming of the on-chip flash is performed via the debug interface. The external host must initially
send instructions using the DEBUG_INSTR debug command to perform the flash programming with the
flash controller.
3.4.1 Lock Bits
For software and/or access protection, a set of lock bits can be written to the upper available flash
page—the lock-bit page. The lock-bit structure consists of 128 bits where the first (FLASH_PAGES-1)
each corresponds to the first flash pages available in the device. The last bit (at the highest address) is
the debug lock bit (see Table 3-5). The structure starts at address 0x7FF0 ( address 0xFFF0 in XDATA )
when the upper flash bank is mapped in, and occupies 16 bytes. The rest of the lock-bit page can be used
to store code/constants, but cannot be changed without entering debug mode.
The PAGELOCK[FLASH_PAGES-2:0] lock-protect bits are used to enable erase and write protection for
individual flash memory pages (2 KB; 1 KB on CC2533). There is one bit for each available page.
When the debug-lock bit, DBGLOCK, is set to 0 (see Table 3-5), all debug commands except
CHIP_ERASE, READ_STATUS, and GET_CHIP_ID are disabled. The status of the debug-lock bit can be
read using the READ_STATUS command (see Section 3.3.2).
Note that after the debug-lock bit has changed due to a write to the lock-bit page or a CHIP_ERASE
command, the device must be reset to lock/unlock the debug interface.
Issuing a CHIP_ERASE command is the only way to clear the debug-lock bit, thereby unlocking the debug
interface.
Table 3-5 defines the 16-byte structure containing the flash lock-protection bits. Bit 0 of the first byte
contains the lock bit for page 0, bit 1 of the first byte contains the lock bit for page 1, and so on. Bit 7 of
the last byte in the flash is the DBGLOCK bit (bit 127 in the structure).
Table 3-5. Flash Lock-Protection Bit Structure Definition
Bit
3.5
Name
Description
127
DBGLOCK
Debug-lock bit
0: Disable debug commands
1: Enable debug commands
126:FLASH_PAGES-1
FREE SPACE
On devices with less than 256 KB memory: Code space available for
storing code or constants.
Note: Can only be changed while in debug mode.
FLASH_PAGES-2:0
PAGELOCK[FLASH_PAGES-2:0] Page-lock bits. There is one bit for each of the up to 128 pages.
Page-lock bits for unavailable pages are not used.
0: Page locked
1: Page not locked
Debug Interface and Power Modes
Power modes PM2 and PM3 may be handled in two different ways when the chip is in debug mode. The
default behavior is never to turn off the digital voltage regulator. This emulates power modes while
maintaining debug mode operation. The clock sources are turned off as in ordinary power modes. The
other option is to turn off the 1.8-V internal digital power. This leads to a complete shutdown of the digital
part, which disables debug mode. When the chip is in debug mode, the two options are controlled by
configuration bit 5 (SOFT_POWER_MODE).
The debug interface still responds to a reduced set of commands while in one of the power modes. The
chip can be woken up from sleep mode by issuing a HALT command to the debug interface. The HALT
command brings the chip up from sleep mode in the halted state. The RESUME command must be issued
to resume software execution.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Interface
57
Debug Interface and Power Modes
www.ti.com
The debug status may be read when in power modes. The status must be checked when leaving a power
mode by issuing a HALT command. The time needed to power up depends on which power mode the
chip is in, and must be checked in the debug status. The debug interface only accepts commands that are
available in sleep mode before the chip is operational.
NOTE:
58
Debugging in Idle mode and PM1 is not supported. It is recommended to use active mode
or another power mode when debugging.
Debug Interface
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
3.6
Registers
DBGDATA (0x6260) – Debug Data
Bit
7:0
Name
BYTE[7:0]
Reset
0
R/W
R
Reset
Chip
dependent
R/W
R
Reset
R/W
Chip
dependent
R
Description
Debug data from BURST_WRITE command
This register is updated each time a new byte has been transferred to the debug
interface using the BURST_WRITE command. A DBG_BW DMA trigger is
generated when this byte is updated. This allows the DMA controller to fetch the
data.
CHVER (0x6249) – Chip Version
Bit
7:0
Name
VERSION[7:0]
Description
Chip revision number
CHIPID (0x624A) – Chip ID
Bit
7:0
Name
CHIPID[7:0]
Description
Chip identification number.
CC2530: 0xA5
CC2531: 0xB5
CC2533: 0x95
CC2540: 0x8D
CC2541: 0x41
CHIPINFO0 (0x6276) – Chip Information Byte 0
Bit
7
6:4
3
2
1:0
Name
–
FLASHSIZE[2:0]
USB
–
–
Reset
0
Chip
dependent
Chip
dependent
1
00
R/W
R0
R
R
Description
Reserved. Always 0.
Flash Size. 001 – 32 KB, 010 – 64 KB, 011 – 128 KB (for CC2533: 011 – 96 KB),
100 – 256 KB
1 if chip has USB, 0 otherwise
R1
R0
Reserved. Always 1
Reserved. Always 00
CHIPINFO1 (0x6277) – Chip Information Byte 1
Bit
7:3
–
Name
2:0
SRAMSIZE[2:0]
Reset
Chip
dependent
Chip
dependent
R/W
R
R
Description
Reserved.
SRAM size in KB minus 1. For example, a 4-KB device has this field set to 011. Add
1 to the number to get the number of KB available.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Interface
59
Chapter 4
SWRU191D – April 2009 – Revised March 2013
Power Management and Clocks
Low-power operation is enabled through different operating modes (power modes). The various operating
modes are referred to as active mode, idle mode, and power modes 1, 2, and 3 (PM1–PM3). Ultralowpower operation is obtained by turning off the power supply to modules to avoid static (leakage) power
consumption and also by using clock gating and turning off oscillators to reduce dynamic power
consumption.
Topic
4.1
4.2
4.3
4.4
4.5
4.6
60
...........................................................................................................................
Power Management Introduction .........................................................................
Power-Management Control ...............................................................................
Power-Management Registers ............................................................................
Oscillators and Clocks .......................................................................................
Timer Tick Generation ........................................................................................
Data Retention ..................................................................................................
Power Management and Clocks
Page
61
62
63
66
69
69
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power Management Introduction
www.ti.com
4.1
Power Management Introduction
Different operating modes, or power modes, are used to allow low-power operation. Ultralow-power
operation is obtained by turning off the power supply to modules to avoid static (leakage) power
consumption and also by using clock gating and turning off oscillators to reduce dynamic power
consumption.
The five various operating modes (power modes) are called active mode, idle mode, PM1, PM2, and PM3
(PM1/PM2/PM3 are also referred to as sleep modes). Active mode is the normal operating mode, whereas
PM3 has the lowest power consumption. The impact of the different power modes on system operation is
shown in Table 4-1, together with voltage regulator and oscillator options.
Table 4-1. Power Modes
Power Mode
Configuration
High-Frequency Oscillator
Low-Frequency Oscillator
A
32-MHz XOSC
C
32-kHz XOSC
B
16-MHz RCOSC
D
32-kHz RCOSC
Voltage Regulator
(Digital)
Active / idle mode
A or B
C or D
PM1
None
C or D
ON
ON
PM2
None
C or D
OFF
PM3
None
None
OFF
Active mode: The fully functional mode. The voltage regulator to the digital core is on, and either the 16MHz RC oscillator or the 32-MHz crystal oscillator or both are running. Either the 32-kHz RCOSC or the
32-kHz XOSC is running.
Idle mode: Identical to active mode, except that the CPU core stops operating (is idle).
PM1: The voltage regulator to the digital part is on. Neither the 32-MHz XOSC nor the 16-MHz RCOSC is
running. Either the 32-kHz RCOSC or the 32-kHz XOSC is running. The system goes to active mode on
reset, an external interrupt, or when the Sleep Timer expires.
PM2: The voltage regulator to the digital core is turned off. Neither the 32-MHz XOSC nor the 16-MHz
RCOSC is running. Either the 32-kHz RCOSC or the 32-kHz XOSC is running. The system goes to active
mode on reset, an external interrupt, or when the Sleep Timer expires.
PM3: The voltage regulator to the digital core is turned off. None of the oscillators is running. The system
goes to active mode on reset or an external interrupt.
The POR is active in PM2/PM3, but the BOD is powered down, which gives a limited voltage supervision.
If the supply voltage is lowered to below 1.4 V during PM2/PM3, at temperatures of 70°C or higher, and
then brought back up to good operating voltage before active mode is re-entered, registers and RAM
contents that are saved in PM2/PM3 may become altered. Hence, care should be taken in the design of
the system power supply to ensure that this does not occur. The voltage can be periodically supervised
accurately by entering active mode, as a BOD reset is triggered if the supply voltage is below
approximately 1.7 V.
The CC2533 and CC2541 have functionality to perform automatically a CRC check of the retained
configuration register values in PM2/PM3 to check that the device state was not altered during sleep. The
bits in SRCRC.CRC_RESULT indicate whether there were any changes, and by enabling
SRCRC.CRC_RESET_EN, the device immediately resets itself with a watchdog reset if
SRCRC.CRC_RESULT is not 00 (= CRC of retained registers passed) after wakeup from PM2/PM3. The
SRCRC register also contains the SRCRC.FORCE_RESET bit that can be used by software to immediately
trigger a watchdog reset to reboot the device.
For CC2533 and CC2541, additional analog reset architecture adds another brownout detector (the
3VBOD) that senses on the unregulated voltage. The purpose of this 3VBOD is to reduce the current
consumption of the device when supplied with voltages well below the operating voltage.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power Management and Clocks
61
Power-Management Control
www.ti.com
4.1.1 Active and Idle Mode
Active mode is the fully functional mode of operation where the CPU, peripherals, and RF transceiver are
active. The digital voltage regulator is turned on.
Active mode is used for normal operation. By enabling the PCON.IDLE bit while in active mode
(SLEEPCMD.MODE = 0x00), the CPU core stops operating and the idle mode is entered. All other
peripherals function normally, and any enabled interrupt wakes up the CPU core (to transition back from
idle mode to active mode).
4.1.2 PM1
In PM1, the high-frequency oscillators are powered down (32-MHz XOSC and 16-MHz RCOSC). The
voltage regulator and the enabled 32-kHz oscillator are on. When PM1 is entered, a power-down
sequence is run.
PM1 is used when the expected time until a wakeup event is relatively short (less than 3 ms), because
PM1 uses a fast power-down/up sequence.
4.1.3 PM2
PM2 has the second-lowest power consumption. In PM2, the power-on reset, external interrupts, selected
32-kHz oscillator, and Sleep Timer peripherals are active. I/O pins retain the I/O mode and output value
set before entering PM2. All other internal circuits are powered down. The voltage regulator is also turned
off. When PM2 is entered, a power-down sequence is run.
PM2 is typically entered when using the Sleep Timer as the wakeup event, and also combined with
external interrupts. PM2 should typically be choosen, compared to PM1, when expected sleep time
exceeds 3 ms. Using less sleep time does not reduce system power consumption compared to using
PM1.
4.1.4 PM3
PM3 is used to achieve the operating mode with the lowest power consumption. In PM3, all internal
circuits that are powered from the voltage regulator are turned off (basically all digital modules; the only
exceptions are interrupt detection and POR level sensing). The internal voltage regulator and all oscillators
are also turned off.
Reset (POR or external) and external I/O port interrupts are the only functions that operate in this mode.
I/O pins retain the I/O mode and output value set before entering PM3. A reset condition or an enabled
external I/O interrupt event wakes the device up and places it into active mode (an external interrupt starts
from where it entered PM3, whereas a reset returns to start-of-program execution). The content of RAM
and registers is partially preserved in this mode (see Section 4.6). PM3 uses the same power-down/up
sequence as PM2.
PM3 is used to achieve ultralow power consumption when waiting for an external event. It should be used
when expected sleep time exceeds 3 ms.
4.2
Power-Management Control
The required power mode is selected by the MODE bits in the SLEEPCMD control register and the
PCON.IDLE bit. Setting the SFR register PCON.IDLE bit enters the mode selected by SLEEPCMD.MODE.
An enabled interrupt from port pins or Sleep Timer or a power-on reset wakes the device from other power
modes and brings it into active mode.
When PM1, PM2, or PM3 is entered, a power-down sequence is run. When the device is taken out of
PM1, PM2, or PM3, it starts at 16 MHz and automatically changes to 32 MHz if CLKCONCMD.OSC was 0
when entering the power mode (setting PCON.IDLE). If CLKCONCMD.OSC was 1 when PCON.IDLE was
set, when entering the power mode, it continues to run at 16 MHz.
62
Power Management and Clocks
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power-Management Registers
www.ti.com
The instruction that sets the PCON.IDLE bit must be aligned in a certain way for correct operation. The
first byte of the assembly instruction immediately following this instruction must not be placed on a 4-byte
boundary. Furthermore, cache must not be disabled (see CM in the FCTL register description in
Chapter 6). Failure to comply with this requirement may cause higher current consumption. Provided this
requirement is fulfilled, the first assembly instruction after the instruction that sets the PCON.IDLE bit is
performed before the ISR of the interrupt that caused the system to wake up, but after the system woke
up. If this instruction is a global interrupt disable, it is possible to have it followed by code for execution
after wakeup, but before the ISR is serviced.
An example of how this can be done in the IAR compiler is shown as follows. The command for setting
PCON to 1 is placed in a function written in assembly code. In a C file calling this function, a declaration
such as extern void
EnterSleepModeDisableInterruptsOnWakeup(void); is used. The
RSEG NEAR_CODE:CODE:NOROOT(2) statement ensures that the MOV PCON,#1 instruction is placed on
a 2-byte boundary. It is a 3-byte instruction, so the following instruction is not placed on a 4-byte
boundary, as required. In the following example, this instruction is CLR EA, which disables all interrupts.
That means that the ISR of the interrupt that woke up the system is not executed until after the IEN0.EA
bit has been set again later in the code. If this functionality is not wanted, the CLR EA instruction can be
replaced by a NOP.
PUBLIC EnterSleepModeDisableInterruptsOnWakeup FUNCTION
EnterSleepModeDisableInterruptsOnWakeup,0201H RSEG NEAR_CODE:CODE:NOROOT(2)
EnterSleepModeDisableInterruptsOnWakeup: MOV PCON,#1 CLR EA RET
4.3
Power-Management Registers
This section describes the power-management registers. All register bits retain their previous values when
entering PM2 or PM3.
SRCRC (0x6262) – Sleep Reset CRC (CC2533 and CC2541 only)
Bit
7
6
5
4
3:2
1
0
Name
XOSC_AMP_DET_EN
Reset
0
R/W
R/W
–
FORCE_RESET
0
0
R0
R/W
–
CRC_RESULT
0
00
R
R/W0
–
CRC_RESET_EN
0
0
R
R/W
Description
0: Disable
1: Enable the amplitude detector for the 32-MHz XOSC, CC2533 only
Reserved. Always read 0.
0: No action
1: Force watchdog reset.
Reserved
00: CRC of retained registers passed
01: Low CRC value failed
10: High CRC value failed
11: Both CRC values failed
Reserved
0: Disable reset of chip due to CRC.
1: Enable reset of chip if CRC_RESULT != 00 after wakeup from PM2/PM3.
PCON (0x87) – Power Mode Control
Bit
7:1
0
Name
–
IDLE
Reset
0000 000
0
R/W
R/W
R0/W
H0
Description
Reserved, always write as 0000 000.
Power mode control. Writing 1 to this bit forces the device to enter the power mode
set by SLEEPCMD.MODE (note that MODE = 0x00 AND IDLE = 1 stops the CPU core
activity). This bit is always read as 0.
All enabled interrupts clear this bit when active, and the device re-enters active
mode.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power Management and Clocks
63
Power-Management Registers
www.ti.com
SLEEPCMD (0xBE) – Sleep-Mode Control Command
Bit
7
Name
OSC32K_CALDIS
Reset
0
R/W
R/W
6:3
2
1:0
–
–
MODE[1:0]
000 0
1
00
R0
R/W
R/W
Description
Disable 32-kHz RC oscillator calibration
0: 32-kHz RC oscillator calibration is enabled.
1: 32-kHz RC oscillator calibration is disabled.
This setting can be written at any time, but does not take effect before the chip has
been running on the 16-MHz high-frequency RC oscillator.
Reserved
Reserved. Always write as 1
Power-mode setting
00: Active / Idle mode
01: Power mode 1 (PM1)
10: Power mode 2 (PM2)
11: Power mode 3 (PM3)
SLEEPSTA (0x9D) – Sleep-Mode Control Status
Bit
7
Name
OSC32K_CALDIS
6:5
4:3
2:1
0
64
Reset
0
R/W
R
–
RST[1:0]
00
XX
R
R
–
CLK32K
00
0
R
R
Power Management and Clocks
Description
32-kHz RC oscillator calibration status
SLEEPSTA.OSC32K_CALDIS shows the current status of disabling of the 32-kHz
RC calibration. The bit is not set to the same value as
SLEEPCMD.OSC32K_CALDIS before the chip has been run on the 32-kHz RC
oscillator.
Reserved
Status bit indicating the cause of the last reset. If there are multiple resets, the
register only contains the last event.
00: Power-on reset and brownout detection
01: External reset
10: Watchdog Timer reset
11: Clock loss reset
Reserved
The 32-kHz clock signal (synchronized to the system clock)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power-Management Registers
www.ti.com
CLKCONCMD.OSC
0
System Clock
SLEEPCMD.MODE[1:0]
1
XTAL1
XOSC_ STB
32-MHz Crystal Oscillator
SLEEPCMD.MODE[1:0]
16-MHz RC Oscillator
HFRC_STB
CLKCONCMD.OSC32K
0
SLEEPCMD.MODE[1:0]
1
32-kHz Clock
Sleep Timer
Watchdog Timer
XTAL2
32-MHz Crystal Oscillator
SLEEPCMD.MODE[1:0]
32-kHz RC Oscillator
SLEEPCMD.OSC32K_CALDIS
B0303-02
Figure 4-1. Clock System Overview
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power Management and Clocks
65
Oscillators and Clocks
4.4
www.ti.com
Oscillators and Clocks
The device has one internal system clock, or main clock. The source for the system clock can be either
the 16-MHz RC oscillator or the 32-MHz crystal oscillator. Clock control is performed using the
CLKCONCMD SFR register.
There is also one 32-kHz clock source that can either be an RC oscillator or a crystal oscillator, also
controlled by the CLKCONCMD register.
The CLKCONSTA register is a read-only register used for getting the current clock status.
The choice of oscillator allows a trade-off between high accuracy in the case of the crystal oscillator and
low power consumption when the RC oscillator is used. Note that operation of the RF transceiver requires
that the 32-MHz crystal oscillator is used.
In the CC2533, CC2540 and CC2541, an additional module for detection of 32-MHz XOSC stability is
available. This amplitude detector can be useful in environments with significant noise on the power
supply, to ensure that the clock source is not used until the clock signal is stable. In the CC2533, this
module can be enabled by setting the SRCRC.XOSC_AMP_DET_EN bit, and this adds around 20 μs to the
32-MHz XOSC startup time. In the CC2540/41, the module is always enabled.
4.4.1 Oscillators
Figure 4-1 gives an overview of the clock system with available clock sources.
Two high-frequency oscillators are present in the device:
• 32-MHz crystal oscillator
• 16-MHz RC oscillator
The 32-MHz crystal-oscillator start-up time may be too long for some applications; therefore, the device
can run on the 16-MHz RC oscillator until the crystal oscillator is stable. The 16-MHz RC oscillator
consumes less power than the crystal oscillator, but because it is not as accurate as the crystal oscillator it
cannot be used for RF transceiver operation.
Two low-frequency oscillators are present in the device:
• 32-kHz crystal oscillator
• 32-kHz RC oscillator.
The 32-kHz XOSC is designed to operate at 32.768 kHz and provide a stable clock signal for systems
requiring time accuracy. The 32-kHz RCOSC runs at 32.753 kHz when calibrated. The calibration can only
take place when the 32-MHz XOSC is enabled, and this calibration can be disabled by enabling the
SLEEPCMD.OSC32K_CALDIS bit. The 32-kHz RCOSC should be used to reduce cost and power
consumption compared to the 32-kHz XOSC solution. The two 32-kHz oscillators cannot be operated
simultaneously.
4.4.2 System Clock
The system clock is derived from the selected system clock source, which is the 32-MHz XOSC or the 16MHz RCOSC. The CLKCONCMD.OSC bit selects the source of the system clock. Note that to use the RF
transceiver, the 32-MHz crystal oscillator must be selected and stable.
Note that changing the CLKCONCMD.OSC bit does not cause the system clock to change instantly. The
clock source change first takes effect when CLKCONSTA.OSC = CLKCONCMD.OSC. This is due to the
requirement to have stable clocks prior to actually changing the clock source. Also note that the
CLKCONCMD.CLKSPD bit reflects the frequency of the system clock and thus is a mirror of the
CLKCONCMD.OSC bit.
The 16 MHz RC oscillator is calibrated once after the 32-MHz XOSC has been selected and is stable, i.e,.
when the CLKCONSTA.OSC bit switches from 1 to 0.
66
Power Management and Clocks
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Oscillators and Clocks
www.ti.com
NOTE: The change from the 16-MHz clock source to the 32-MHz clock source (and vice versa)
aligns with the CLKCONCMD.TICKSPD setting. A slow CLKCONCMD.TICKSPD setting
when CLKCONCMD.OSC is changed results in a longer time before the actual source
change takes effect. The fastest switching is obtained when CLKCONCMD.TICKSPD equals
000.
NOTE: After coming up from PM1, PM2, or PM3, the CPU must wait for CLKCONSTA.OSC to be 0
before operations requiring the system to run on the 32-MHz XOSC (such as the radio) are
started.
4.4.3 32-kHz Oscillators
Two 32-kHz oscillators are present in the device as clock sources for the 32-kHz clock:
• 32-kHz XOSC
• 32-kHz RCOSC
By default, after a reset, the 32-kHz RCOSC is enabled and selected as the 32-kHz clock source. The
RCOSC consumes less power, but is less accurate compared to the 32-kHz XOSC. The chosen 32-kHz
clock source drives the Sleep Timer, generates the tick for the Watchdog Timer, and is used as a strobe in
Timer 2 to calculate the Sleep Timer sleep time. The CLKCONCMD.OSC32K register bit selects the
oscillator to be used as the 32-kHz clock source. This bit does not give an indication of the stability of the
32-kHz XOSC.
The CLKCONCMD.OSC32K register bit can be written at any time, but does not take effect before the 16MHz RCOSC is the active system clock source. When system clock is changed from the 16-MHz RCOSC
to the 32-MHz XOSC (CLKCONCMD.OSC from 1 to 0), calibration of the 32-kHz RCOSC starts up and is
performed once if the 32-kHz RCOSC is selected. During calibration, a divided version of the 32-MHz
XOSC is used. The result of the calibration is that the 32-kHz RSOSC is running at 32.753 kHz. The 32kHz RCOSC calibration may take up to 2 ms to complete. Calibration can be disabled by setting
SLEEPCMD.OSC32K_CALDIS to 1. At the end of the calibration, an extra pulse may occur on the 32-kHz
clock source, which causes the sleep timer to be incremented by 1.
Note that after having switched to the 32-kHz XOSC and when coming up from PM3 with the 32-kHz
XOSC enabled, the oscillator requires up to 500 ms to stabilize on the correct frequency. The Sleep
Timer, Watchdog Timer and clock-loss detector should not be used before the 32-kHz XOSC is stable.
4.4.4 Oscillator and Clock Registers
This section describes the oscillator and clock registers. All register bits retain their previous values when
entering PM2 or PM3.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power Management and Clocks
67
Oscillators and Clocks
www.ti.com
CLKCONCMD (0xC6) – Clock Control Command
Bit
7
6
5:3
Name
OSC32K
OSC
TICKSPD[2:0]
Reset
1
R/W
R/W
1
R/W
001
R/W
Description
32-kHz clock-source select. Setting this bit initiates a clock-source change only.
CLKCONSTA.OSC32K reflects the current setting. The 16-MHz RCOSC must be
selected as system clock when this bit is to be changed. This bit does not give an
indication of the stability of the 32-kHz XOSC.
0: 32 kHz XOSC
1: 32 kHz RCOSC
System clock-source select. Setting this bit initiates a clock-source change only.
CLKCONSTA.OSC reflects the current setting.
0: 32 MHz XOSC
1: 16 MHz RCOSC
Timer ticks output setting. Cannot be higher than system clock setting given by OSC
bit setting.
000: 32 MHz
001: 16 MHz
010: 8 MHz
011: 4 MHz
100: 2 MHz
101: 1 MHz
110: 500 kHz
111: 250 kHz
Note that CLKCONCMD.TICKSPD can be set to any value, but the effect is
limited by the CLKCONCMD.OSC setting; i.e., if CLKCONCMD.OSC = 1 and
CLKCONCMD.TICKSPD = 000, CLKCONSTA.TICKSPD reads 001, and the
real TICKSPD is 16 MHz.
2:0
CLKSPD
001
R/W
Clock speed. Cannot be higher than system clock setting given by the OSC bit
setting. Indicates current system-clock frequency
000: 32 MHz
001: 16 MHz
010: 8 MHz
011: 4 MHz
100: 2 MHz
101: 1 MHz
110: 500 kHz
111: 250 kHz
Note that CLKCONCMD.CLKSPD can be set to any value, but the effect is
limited by the
CLKCONCMD.OSC setting; i.e., if
CLKCONCMD.OSC = 1 and CLKCONCMD.CLKSPD = 000,
CLKCONSTA.CLKSPD reads 001, and the real CLKSPD is 16 MHz.
Note also that the debugger cannot be used with a divided system clock. When
running the debugger, the value of CLKCONCMD.CLKSPD should be set to 000 when
CLKCONCMD.OSC = 0 or to 001 when CLKCONCMD.OSC = 1.
68
Power Management and Clocks
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer Tick Generation
www.ti.com
CLKCONSTA (0x9E) – Clock Control Status
Bit
7
6
Name
OSC32K
OSC
Reset
1
R/W
R
1
R
5:3
TICKSPD[2:0]
001
R
2:0
CLKSPD
001
R
4.5
Description
Current 32-kHz clock source selected:
0: 32-kHz XOSC
1: 32-kHz RCOSC
Current system clock selected:
0: 32-MHz XOSC
1: 16-MHz RCOSC
Current timer ticks output setting
000: 32 MHz
001: 16 MHz
010: 8 MHz
011: 4 MHz
100: 2 MHz
101: 1 MHz
110: 500 kHz
111: 250 kHz
Current clock speed
000: 32 MHz
001: 16 MHz
010: 8 MHz
011: 4 MHz
100: 2 MHz
101: 1 MHz
110: 500 kHz
111: 250 kHz
Timer Tick Generation
The value of the CLKCONCMD.TICKSPD register controls a global prescaler for Timer 1, Timer 3, and
Timer 4. The prescaler value can be set to a value from 0.25 MHz to 32 MHz. It should be noted that if
CLKCONCMD.TICKSPD indicates a higher frequency than the system clock, the actual prescaler value
indicated in CLKCONSTA.TICKSPD is the same as the system clock.
4.6
Data Retention
In power modes PM2 and PM3, power is removed from most of the internal circuitry. However, SRAM
retains its contents, and the content of internal registers is also retained in PM2 and PM3.
All CPU, RF, and peripheral registers retain their contents in PM2 and PM3, except the AES, I2C, and
USB registers, OBSSEL0–OBSSEL5, TR0, and in the CC2541, LLECTRL.
Switching to the PM2 or PM3 low-power modes appears transparent to software. Note that the value of
the Sleep Timer is not preserved in PM3.
All registers retain their values in PM1.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power Management and Clocks
69
Chapter 5
SWRU191D – April 2009 – Revised March 2013
Reset
The device has five reset sources. The following events generate a reset:
• Forcing the RESET_N input pin low
• A power-on reset condition
• A brownout reset condition
• Watchdog Timer reset condition
• Clock-loss reset condition
The initial conditions after a reset are as follows:
• I/O pins are configured as inputs with pullups (P1.0 and P1.1 are inputs, but do not have
pullup/pulldown)
• CPU program counter is loaded with 0x0000 and program execution starts at this address
• All peripheral registers are initialized to their reset values (see register descriptions)
• Watchdog Timer is disabled
• Clock-loss detetector is disabled
During reset, the I/O pins are configured as inputs with pullups (P1.0 and P1.1 are inputs, but do not have
pullup/pulldown). The RESET_N input is always configured as an input with pull-up.
In the CC2533 and CC2541, a watchdog reset can be generated immediately in software by writing the
SRCRC.FORCE_RESET bit to 1 (see Section 4.3 for the register description). In the other devices in the
family, a watchdog reset can be triggered from software by enabling the watchdog timer with the shortest
time-out and waiting for it to trigger.
Topic
5.1
5.2
70
Reset
...........................................................................................................................
Page
Power-On Reset and Brownout Detector .............................................................. 71
Clock-Loss Detector .......................................................................................... 71
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Power-On Reset and Brownout Detector
www.ti.com
5.1
Power-On Reset and Brownout Detector
The device includes a power-on reset (POR), providing correct initialization during device power on. It also
includes a brownout detector (BOD) operating on the regulated 1.8-V digital power supply only. The BOD
protects the memory contents during supply voltage variations which cause the regulated 1.8-V power to
drop below the minimum level required by digital logic, flash memory, and SRAM.
When power is initially applied, the POR and BOD hold the device in the reset state until the supply
voltage rises above the power-on-reset and brownout voltages.
The cause of the last reset can be read from the register bits SLEEPSTA.RST. It should be noted that a
BOD reset is read as a POR reset.
5.2
Clock-Loss Detector
The clock-loss detector can be used in safety-critical systems to detect that one of the XOSC clock
sources (32-MHz XOSC or 32-kHz XOSC) has stopped. This can typically happen due to damage to the
external crystal or supporting components. When the clock-loss detector is enabled, the two clocks
monitor each other continously. If one of the clocks stops toggling, a clock-loss detector reset is generated
within a certain maximum time-out period. The time-out depends on which clock stops. If the 32-kHz clock
stops, the time-out period is 0.5 ms. If the 32-MHz clock stops, the time-out period is 0.25 ms. When the
system comes up again from reset, software can detect the cause of the reset by reading
SLEEPSTA.RST[1:0]. After a reset, the internal RC oscillators are used. Thus, the system is able to start
up again and can then be powered down gracefully. The clock-loss detector is enabled/disabled with the
CLD.EN bit. It is assumed that the 32-MHz XOSC is selected as system clock source when using the
clock-loss detector. The 32-kHz clock can be 32-kHz RCOSC (should be calibrated for accurate reset
timeout) or 32-kHz XOSC.
In power modes 1 and 2, the clock-loss detector is automatically stopped and restarted when the clocks
start up again.
Before entering power mode 3, switch to the 16-MHz RCOSC and disable the clock-loss detector. When
entering active mode again, turn on the clock-loss detector and then switch back to the 32-MHz XOSC.
CLD (0x6290) – Clock-Loss Detection
Bit
7:1
0
Name
–
EN
Reset
0000 000
0
R/W
R0
R/W
Description
Reserved
Clock-loss detector enable
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Reset
71
Chapter 6
SWRU191D – April 2009 – Revised March 2013
Flash Controller
The device contains flash memory for storage of program code. The flash memory is programmable from
the user software and through the debug interface.
The flash controller handles writing and erasing the embedded flash memory. The embedded flash
memory consists of up to 128 pages of 2048 bytes (CC2530/CC2531/CC2540/CC2541) or 1024 bytes
(CC2533) each.
The flash controller has the following features:
• 32-bit word programmable
• Page erase
• Lock bits for write protection and code security
• Flash-page erase timing 20 ms
• Flash-chip erase timing 20 ms
• Flash-write timing (4 bytes) 20 μs
Topic
6.1
6.2
6.3
6.4
6.5
72
...........................................................................................................................
Flash
Flash
Flash
Flash
Flash
Flash Controller
Memory Organization ................................................................................
Write .......................................................................................................
Page Erase ..............................................................................................
DMA Trigger .............................................................................................
Controller Registers ..................................................................................
Page
73
73
75
76
76
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Flash Memory Organization
www.ti.com
6.1
Flash Memory Organization
The flash memory is divided into 2048-byte or 1024-byte flash pages. A flash page is the smallest
erasable unit in the memory, whereas a 32-bit word is the smallest writable unit that can be written to the
flash.
When performing write operations, the flash memory is word-addressable using a 16-bit address written to
the address registers FADDRH:FADDRL.
When performing page-erase operations, the flash memory page to be erased is addressed through the
register bits FADDRH[7:1] (CC2530/CC2531/CC2540/CC2541) or FADDRH[6:0] (CC2533).
Note the difference in addressing the flash memory; when accessed by the CPU to read code or data, the
flash memory is byte-addressable. When accessed by the flash controller, the flash memory is wordaddressable, where a word consists of 32 bits.
The following sections describe the procedures for flash write and flash page-erase in detail.
6.2
Flash Write
The flash is programmed serially with a sequence of one or more 32-bit words (4 bytes), starting at the
start address (set by FADDRH:FADDRL). In general, a page must be erased before writing can begin. The
page-erase operation sets all bits in the page to 1. The chip-erase command (through the debug interface)
erases all pages in the flash. This is the only way to set bits in the flash to 1. When writing a word to the
flash, the 0-bits are programmed to 0 and the 1-bits are ignored (leaves the bit in the flash unchanged).
Thus, bits are erased to 1 and can be written to 0. It is possible to write multiple times to a word. This is
described in Section 6.2.2.
6.2.1 Flash-Write Procedure
The flash-write sequence algorithm is as follows:
1. Set FADDRH:FADDRL to the start address. (This is the 16 MSBs of the 18-bit byte address).
2. Set FCTL.WRITE to 1. This starts the write-sequence state machine.
3. Write four times to FWDATA within 20 μs (since the last time FCTL.FULL became 0, if not first
iteration). LSB is written first. (FCTL.FULL goes high after the last byte.)
4. Wait until FCTL.FULL goes low. (The flash controller has started programming the 4 bytes written in
step 3 and is ready to buffer the next 4 bytes).
5. Optional status check step:
• If the 4 bytes were not written fast enough in step 3, the operation has timed out and FCTL.BUSY
(and FCTL.WRITE) are 0 at this stage.
• If the 4 bytes could not be written to the flash due to the page being locked, FCTL.BUSY (and
FCTL.WRITE) are 0 and FCTL.ABORT is 1.
6. If this was the last 4 bytes then quit, otherwise go to step 3.
The write operation is performed using one of two methods:
• Using DMA transfer (preferred method)
• Using CPU, running code from SRAM
The CPU cannot access the flash, e.g., to read program code, while a flash-write operation is in progress.
Therefore, the program code executing the flash write must be executed from RAM. See Section 2.2.1 for
a description of how to run code from RAM.
When a flash-write operation is executed from RAM, the CPU continues to execute code from the next
instruction after initiation of the flash-write operation (FCTL.WRITE = 1).
Power mode 1, 2, or 3 must not be entered while writing to the flash. Also, the system clock source
(XOSC/RCOSC) must not be changed while writing. Note that setting CLKCONSTA.CLKSPD to a high
value makes it impossible to meet the timing requirement of 20-μs write timing. With CLKCONSTA.CLKSPD
= 111, the clock period is only 4 μs. It is therefore recommended to keep CLKCONSTA.CLKSPD at 000 or
001 while writing to the flash.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Flash Controller
73
Flash Write
www.ti.com
6.2.2 Writing Multiple Times to a Word
The following rules apply when writing multiple times to a 32-bit word between erase:
• Writing 0 to a bit within a 32-bit flash word, which has been set to 1 by the last erase operation,
changes the state of the bit to 0, subject to the last bullet below.
• It is possible to write 0 to a bit within a 32-bit word repeatedly (subject to the last bullet below) once the
bit has been written with 0. This does not change the state of the bit.
• Writing 1 to a bit does not change the state of the bit, subject to the last bullet below.
• The following limitations apply to writes subsequent to the last page erase:
– A 0 must not be written more than two times to a single bit.
– A 32-bit word shall not be written more than 8 times.
– A page must not be written more than 1024 times.
The state of any bit of a 32-bit flash word is nondeterministic if these limitations are violated.
This makes it possible to write up to 4 new bits to a 32-bit word 8 times. One example write sequence to a
word is shown in Table 6-1. Here bn represents the 4 new bits written to the word for each update. This
technique is useful to maximize the lifetime of the flash for data-logging applications.
Table 6-1. Example Write Sequence
Step
Value Written
FLASH Contents After Writing
1
(page erase)
0xFFFFFFFF
The erase sets all bits to 1.
Comment
2
0xFFFFFFFb0
0xFFFFFFFb0
Only the bits written 0 are set to 0, whereas all bits
written 1 are ignored.
3
0xFFFFFFb1F
0xFFFFFFb1b0
Only the bits written 0 are set to 0, whereas all bits
written 1 are ignored.
4
0xFFFFFb2FF
0xFFFFFb2b1b0
Only the bits written 0 are set to 0, whereas all bits
written 1 are ignored.
5
0xFFFFb3FFF
0xFFFFb3b2b1b0
Only the bits written 0 are set to 0, whereas all bits
written 1 are ignored.
6
0xFFFb4FFFF
0xFFFb4b3b2b1b0
Only the bits written 0 are set to 0, whereas all bits
written 1 are ignored.
7
0xFFb5FFFFF
0xFFb5b4b3b2b1b0
Only the bits written 0 are set to 0, whereas all bits
written 1 are ignored.
8
0xFb6FFFFFF
0xFb6b5b4b3b2b1b0
Only the bits written 0 are set to 0, whereas all bits
written 1 are ignored.
9
0xb7FFFFFFF
0xb7b6b5b4b3b2b1b0
Only the bits written 0 are set to 0, whereas all bits
written 1 are ignored.
6.2.3 DMA Flash Write
When using DMA write operations, the data to be written into flash is stored in the XDATA memory space
(RAM or registers). A DMA channel is configured to read the data to be written from the memory source
address and write this data to the flash write-data register (FWDATA) fixed destination address, with the
DMA trigger event FLASH (TRIG[4:0] = 1 0010 in DMA configuration) enabled. Thus, the flash controller
triggers a DMA transfer when the flash write-data register, FWDATA, is ready to receive new data. The
DMA channel should be configured to perform single-mode, byte-size transfers with the source address
set to start-of-data block and destination address to fixed FWDATA (note that the block size, LEN in
configuration data, must be divisible by 4; otherwise, the last word is not written to the flash). High priority
should also be ensured for the DMA channel, so it is not interrupted in the write process. If interrupted for
more than 20 μs, the write operation may time out, and the write bit, FCTL.WRITE, is set to 0.
When the DMA channel is armed, starting a flash write by setting FCTL.WRITE to 1 triggers the first DMA
transfer (DMA and flash controller handle the reset of the transfer).
Figure 6-1 shows an example of how a DMA channel is configured and how a DMA transfer is initiated to
write a block of data from a location in XDATA to flash memory.
74
Flash Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Flash Page Erase
www.ti.com
Set up DMA channel:
SRCADDR=<XDATA location>
DESTADDRR=FWDATA
VLEN=0
LEN=<block size>
WORDSIZE=byte
TMODE=single mode
TRIG=FLASH
SRCINC=1 byte
DESTINC=0 bytes
IRQMASK=yes
M8=0
PRIORITY=high
Set up flash address
Arm DMA Channel
Start flash write
F0031-01
Figure 6-1. Flash Write Using DMA
6.2.4 CPU Flash Write
To write to the flash using the CPU, a program executing from SRAM must implement the steps outlined
in the procedure described in Section 6.2.1. Disable interrupts to ensure the operation does not time out.
6.3
Flash Page Erase
The flash page-erase operation sets all bits in the page to 1.
A page erase is initiated by setting FCTL.ERASE to 1. The page addressed by FADDRH[7:1]
(CC2530/CC2531/CC2540/CC2541) or FADDRH[6:0] (CC2533) is erased when a page erase is
initiated . Note that if a page erase is initiated simultaneously with a page write, i.e., FCTL.WRITE is set to
1, the page erase is performed before the page-write operation starts. The FCTL.BUSY bit can be polled
to see when the page erase has completed.
Power mode 1, 2, or 3 must not be entered while erasing a page. Also, the system clock source
(XOSC/RCOSC) must not be changed while erasing.
NOTE: If a flash page-erase operation is performed from within flash memory and the Watchdog
Timer is enabled, a Watchdog Timer interval must be selected that is longer than 20 ms, the
duration of the flash page-erase operation, so that the CPU can clear the Watchdog Timer.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Flash Controller
75
Flash DMA Trigger
www.ti.com
6.3.1 Performing Flash Erase From Flash Memory
Note that while executing program code from within flash memory, when a flash erase or write operation is
initiated the CPU stalls, and program execution resumes from the next instruction when the flash controller
has completed the operation.
The following code example of how to erase one flash page in the CC2530 is given for use with the IAR
compiler:
#include <ioCC2530.h>
unsigned char erase_page_num = 3;
/* page number to erase, here: flash page #3 */
/* Erase one flash page */
EA = 0;
while (FCTL & 0x80);
FADDRH = erase_page_num << 1;
FCTL |= 0x01;
while (FCTL & 0x80);
EA = 1;
/*
/*
/*
/*
/*
/*
disable interrupts */
poll FCTL.BUSY and wait until flash controller is ready */
select the flash page via FADDRH[7:1] bits */
set FCTL.ERASE bit to start page erase */
optional: wait until flash write has completed (~20 ms) */
enable interrupts */
6.3.2 Different Flash Page Size on CC2533
The flash page size has been reduced from 2 KB (2048 bytes) on CC2530/CC2531/CC2540/CC2541 to 1
KB (1024 bytes) on CC2533. When performing page-erase operations on the flash memory, the page to
be erased is addressed with the register bits FADDRH[6:0] on CC2533 as opposed to FADDRH[7:1] on
CC2530/CC2531/CC2540. The page lock bits are still placed in the upper 16 bytes of the last accessible
flash page.
6.4
Flash DMA Trigger
The flash DMA trigger is activated when flash data written to the FWDATA register has been written to the
specified location in the flash memory, thus indicating that the flash controller is ready to accept new data
to be written to FWDATA. Four trigger pulses are generated. In order to start the first transfer, one must set
the FCTL.WRITE bit to 1. The DMA and the flash controller then handle all transfers automatically for the
defined block of data (LEN in DMA configuration). It is further important that the DMA is armed prior to
setting the FCTL.WRITE bit, that the trigger source is set to FLASH (TRIG[4:0] = 10010), and that
the DMA has high priority so the transfer is not interrupted. If interrupted for more than 20 μs, the write
operation times out and FCTL.WRITE bit is cleared.
6.5
Flash Controller Registers
The flash controller registers are described in this section.
76
Flash Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Flash Controller Registers
www.ti.com
FCTL (0x6270) – Flash Control
Bit
7
BUSY
6
FULL
5
ABORT
0
–
CM[1:0]
0
01
1
WRITE
0
0
ERASE
0
4
3:2
Name
Reset
0
R/W
R
Description
Indicates that write or erase is in operation. This flag is set when the WRITE or
ERASE bit is set.
0: No write or erase operation active
1: Write or erase operation activated
R/H0 Write buffer-full status. This flag is set when 4 bytes have been written to FWDATA
during flash write. The write buffer is then full and does not accept more data; i.e,
writes to FWDATA are ignored when the FULL flag is set. The FULL flag is cleared
when the write buffer again is ready to receive 4 more bytes. This flag is only
needed when the CPU is used to write to the flash.
0: Write buffer can accept more data.
1: Write buffer full
R/H0 Abort status. This bit is set when a write operation or page erase is aborted. An
operation is aborted when the page accessed is locked. The abort bit is cleared
when a write or page erase is started.
R
Reserved
R/W Cache mode
00: Cache disabled
01: Cache enabled
10: Cache enabled, prefetch mode
11: Cache enabled, real-time mode
Cache mode. Disabling the cache increases the power consumption and reduces
performance. Prefetching, for most applications, improves performance by up to
33% at the expense of potentially increased power consumption. Real-time mode
provides predictable flash-read access time; the execution time is equal to that in
cache-disabled mode, but the power consumption is lower.
Note: The value read always represents the current cache mode. Writing a new
cache mode starts a cache mode-change request that may take several clock cycles
to complete. Writing to this register is ignored if there is a current cache-change
request in progress.
R/W1/ Write. Start writing word at location given by FADDRH:FADDRL. The WRITE bit stays
H0
at 1 until the write completes. The clearing of this bit indicates that the erase has
completed, i.e., it has timed out or aborted.
If ERASE is also set to 1, a page erase of the whole page addressed by
FADDRH[7:1] is performed before the write. Setting WRITE to 1 when ERASE is 1
has no effect.
R/W1/ Page erase. Erase the page that is given by FADDRH[7:1]
H0
(CC2530/CC2531/CC2540/CC2541) or FADDRH[6:0] (CC2533) . The ERASE
bit stays at 1 until the erase completes. The clearing of this bit indicates that the
erase has completed successfully or aborted.
Setting ERASE to 1 when WRITE is 1 has no effect.
FWDATA (0x6273) – Flash Write Data
Bit
7:0
Name
FWDATA[7:0]
Reset
0x00
R/W
R0/W
Description
Flash write data. This register can only be written to when FCTL.WRITE is 1.
FADDRH (0x6272) – Flash-Address High Byte
Bit
7:0
Name
FADDRH[7:0]
Reset
0x00
R/W
R/W
Description
Page address/high byte of flash word address
Bits [7:1] (CC2530/CC2531/CC2540/CC2541) or bits [6:0] (CC2533) select
which page to access.
FADDRL (0x6271) – Flash-Address Low Byte
Bit
7:0
Name
FADDRL[7:0]
Reset
0x00
R/W
R/W
Description
Low byte of flash word address
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Flash Controller
77
Chapter 7
SWRU191D – April 2009 – Revised March 2013
I/O Ports
There are 21 digital input/output pins that can be configured as general-purpose digital I/O or as peripheral
I/O signals connected to the ADC, timers, or USART peripherals. The use of the I/O ports is fully
configurable from user software through a set of configuration registers.
The I/O ports have the following key features:
• 21 digital input/output pins
• General-purpose I/O or peripheral I/O
• Pullup or pulldown capability on inputs
• External interrupt capability
The external interrupt capability is available on all 21 I/O pins. Thus, external devices may generate
interrupts if required. The external interrupt feature can also be used to wake the device up from sleep
mode (power modes PM1/PM2/PM3).
Topic
...........................................................................................................................
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
7.10
7.11
78
I/O Ports
Unused I/O Pins ................................................................................................
Low I/O Supply Voltage ......................................................................................
General-Purpose I/O ..........................................................................................
General-Purpose I/O Interrupts ...........................................................................
General-Purpose I/O DMA ...................................................................................
Peripheral I/O ....................................................................................................
Debug Interface .................................................................................................
32-kHz XOSC Input ............................................................................................
Radio Test Output Signals ..................................................................................
Power-Down Signal MUX (PMUX) ........................................................................
I/O Registers .....................................................................................................
Page
79
79
79
79
80
80
83
83
84
84
84
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Unused I/O Pins
www.ti.com
7.1
Unused I/O Pins
Unused I/O pins should have a defined level and not be left floating. One way to do this is to leave the pin
unconnected and configure the pin as a general-purpose I/O input with pullup resistor. This is also the
state of all pins during and after reset (except P1.0 and P1.1, which do not have pullup/pulldown
capability). Alternatively, the pin can be configured as a general-purpose I/O output. In either case, the pin
should not be connected directly to VDD or GND, in order to avoid excessive power consumption.
7.2
Low I/O Supply Voltage
In applications where the digital I/O power supply voltage pins, DVDD1 and DVDD2, are below 2.6 V, the
register bit PICTL.PADSC should be set to 1 in order to obtain the output dc characteristics specified in
the DC Characteristics table in the device data sheet (Appendix C).
7.3
General-Purpose I/O
When used as general-purpose I/O, the pins are organized as three 8-bit ports, Port 0, Port 1, and Port 2;
denoted P0, P1, and P2. P0 and P1 are complete 8-bit-wide ports, whereas P2 has only five usable bits.
All ports are both bit- and byte-addressable through the SFR registers P0, P1, and P2. Each port pin can
individually be set to operate as a general-purpose I/O or as a peripheral I/O.
The output drive strength is 4 mA on all outputs, except for the two high-drive outputs, P1.0 and P1.1,
which each have 20-mA output drive strength.
The registers PxSEL, where x is the port number 0–2, are used to configure each pin in a port as either a
general-purpose I/O pin or as a peripheral I/O signal. By default, after a reset, all digital input/output pins
are configured as general-purpose input pins.
To change the direction of a port pin, the registers PxDIR are used to set each port pin to be either an
input or an output. Thus, by setting the appropriate bit within PxDIR to 1, the corresponding pin becomes
an output.
When reading the port registers P0, P1, and P2, the logic values on the input pins are returned regardless
of the pin configuration. This does not apply during the execution of read-modify-write instructions. The
read-modify-write instructions are: ANL, ORL, XRL, JBC, CPL, INC, DEC, DJNZ, MOV, CLR, and SETB.
Operating on a port register, the following is true: When the destination is an individual bit in port register
P0, P1, or P2, the value of the register, not the value on the pin, is read, modified, and written back to the
port register.
When used as an input, the general-purpose I/O port pins can be configured to have a pullup, pulldown or
three-state mode of operation. By default, after a reset, inputs are configured as inputs with pullup. To
deselect the pullup or pulldown function on an input, the appropriate bit within the PxINP must be set to 1.
The I/O port pins P1.0 and P1.1 do not have pullup/pulldown capability. Note that pins configured as
peripheral I/O signals do not have pullup/pulldown capability, even if the peripheral function is an input.
In power modes PM1, PM2, and PM3, the I/O pins retain the I/O mode and output value (if applicable) that
was set when PM1/PM2/PM3 was entered.
7.4
General-Purpose I/O Interrupts
General-purpose I/O pins configured as inputs can be used to generate interrupts. The interrupts can be
configured to trigger on either a rising or falling edge of the external signal. Each of the P0, P1, and P2
ports has port interrupt-enable bits common for all bits within the port located in the IEN1–IEN2 registers
as follows:
• IEN1.P0IE: P0 interrupt enable
• IEN2.P1IE: P1 interrupt enable
• IEN2.P2IE: P2 interrupt enable
In addition to these common interrupt enables, the bits within each port have individual interrupt enables
located in SFR registers P0IEN, P1IEN, and P2IEN. Even I/O pins configured as peripheral I/O or
general-purpose outputs have interupts generated when enabled.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Ports
79
General-Purpose I/O DMA
www.ti.com
When an interrupt condition occurs on one of the I/O pins, the interrupt status flag in the corresponding
P0–P2 interrupt flag register, P0IFG, P1IFG, or P2IFG, is set to 1. The interrupt status flag is set
regardless of whether the pin has its interrupt enable set. When an interrupt is serviced, the interrupt
status flag is cleared by writing a 0 to that flag. This flag must be cleared prior to clearing the CPU port
interrupt flag (PxIF). This is illustrated in Figure 2-4: There is an edge detect between the input line and
PxIFG, but no edge detect or one-shot between PxIFG and PxINT. The practical impact of this is what is
written in Section 2.5.1
The SFR registers used for interrupts are described later in this section. The registers are summarized as
follows:
• P0IEN: P0 interrupt enables
• P1IEN: P1 interrupt enables
• P2IEN: P2 interrupt enables
• PICTL: P0, P1, and P2 edge configuration
• P0IFG: P0 interrupt flags
• P1IFG: P1 interrupt flags
• P2IFG: P2 interrupt flags
7.5
General-Purpose I/O DMA
When used as general-purpose I/O pins, the P0 and P1 ports are each associated with one DMA trigger.
These DMA triggers are IOC_0 for P0 and IOC_1 for P1, as shown in Table 8-1.
The IOC_0 trigger is activated when an interrupt occurs on the P0 pins. The IOC_1 trigger is activated
when an interrupt occurs on the P1 pins.
7.6
Peripheral I/O
This section describes how the digital I/O pins are configured as peripheral I/Os. For each peripheral unit
that can interface with an external system through the digital input/output pins, a description of how
peripheral I/Os are configured is given in the following subsections.
For USART and timer I/O, setting the appropriate PxSEL bits to 1 is required for the output signals on a
digital I/O pin to be controlled by the peripheral. For peripheral inputs from digital I/O pins, this is optional.
PxSEL = 1 overrides the pullup/pulldown settings of a pin, so to be able to control pullup/pulldown with
the PxINP bits, the PxSEL bit should be set to 0 for that pin.
Note that peripheral units have two alternative locations for their I/O pins; see Table 7-1. Priority can be
set between peripherals if conflicting settings regarding I/O mapping are present (using the
P2SEL.PRIxP1 and P2DIR.PRIP0 bits). All combinations not causing conflicts can be used.
Note that a peripheral normally is present at the selected location even if it is not used, and another
peripheral that is to use the pins must be given higher priority. The exception is the RTS and CTS pins of
a USART in UART mode with flow control disabled and the SSN pin of a USART configured in SPI master
mode.
Note also that peripheral units that have input pins receive an input from the pin regardless of the PxINP
setting, and this may influence on the state of the peripheral unit. For instance, a UART should be flushed
before use if there may have been activity on the RX pin prior to taking it in use as a UART pin.
80
I/O Ports
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Peripheral I/O
www.ti.com
Table 7-1. Peripheral I/O Pin Mapping
Periphery/
Function
ADC
P0
P1
7
6
5
4
3
2
1
0
A7
A6
A5
A4
A3
A2
A1
A0
O
–
+
Operational
amplifier
Analog
comparator
+
–
USART 0 SPI
C
SS
MO
7
6
RT
CT
TX
RX
MI
M0
C
SS
Alt. 2
USART 1
UART
RX
TX
RT
CT
3
2
1
0
Alt. 2
TIMER 1
4
Alt. 2
3
P2
3
2
M0
MI
C
SS
TX
RX
RT
CT
MI
M0
C
SS
RX
TX
RT
CT
4
0
TIMER 3
1
Alt. 2
1
0
4
3
2
1
1
1
2
1
0
0
0
TIMER 4
Alt. 2
1
32-kHz XOSC
Q1
DEBUG
0
Q2
DC
OBSSEL
0
MI
Alt. 2
USART 1 SPI
4
T
Alt. 2
USART 0
UART
5
5
4
3
2
1
DD
0
7.6.1 Timer 1
PERCFG.T1CFG selects whether to use alternative 1 or alternative 2 locations.
In
•
•
•
•
•
Table 7-1, the Timer 1 signals are shown as the following:
0: Channel 0 capture/compare pin
1: Channel 1 capture/compare pin
2: Channel 2 capture/compare pin
3: Channel 3 capture/compare pin
4: Channel 4 capture/compare pin
P2DIR.PRIP0 selects the order of precedence when assigning several peripherals to Port 0. When set to
10, Timer 1 channels 0–1 have precedence, and when set to 11, Timer 1 channels 2–3 have precedence.
To have all Timer 1 channels visible in the alternative 1 location, move both USART 0 and USART 1 to
the alternative 2 location.
P2SEL.PRI1P1 and P2SEL.PRI0P1 select the order of precedence when assigning several peripherals
to Port 1. The Timer 1 channels have precedence when the former is set low and the latter is set high.
7.6.2 Timer 3
PERCFG.T3CFG selects whether to use alternative 1 or alternative 2 locations.
In Table 7-1, the Timer 3 signals are shown as the following:
• 0: Channel 0 capture/compare pin
• 1: Channel 1 capture/compare pin
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Ports
81
Peripheral I/O
www.ti.com
P2SEL.PRI2P1 and P2SEL.PRI3P1 select the order of precedence when assigning several peripherals
to Port 1. The Timer 3 channels have precedence when both bits are set high. If P2SEL.PRI2P1 is set
high and P2SEL.PRI3P1 is set low, the Timer 3 channels have precedence over USART 1, but USART 0
has precedence over the Timer 3 channels as well as over USART 1.
7.6.3 Timer 4
PERCFG.T4CFG selects whether to use alternative 1 or alternative 2 locations.
In Table 7-1, the Timer 4 signals are shown as the following:
• 0: Channel 0 capture/compare pin
• 1: Channel 1 capture/compare pin
P2SEL.PRI1P1 selects the order of precedence when assigning several peripherals to Port 1. The Timer
4 channels have precedence when the bit is set.
7.6.4 USART 0
The SFR register bit PERCFG.U0CFG selects whether to use alternative 1 or alternative 2 locations.
In Table 7-1, the USART 0 signals are shown as follows:
UART:
• RX: RXDATA
• TX: TXDATA
• RT: RTS
• CT: CTS
SPI:
• MI: MISO
• MO: MOSI
• C: SCK
• SS: SSN
P2DIR.PRIP0 selects the order of precedence when assigning several peripherals to Port 0. When set to
00, USART 0 has precedence. Note that if UART mode is selected and hardware flow control is disabled,
USART 1 or Timer 1 has precedence to use ports P0.4 and P0.5.
P2SEL.PRI3P1 and P2SEL.PRI0P1 select the order of precedence when assigning several peripherals
to Port 1. USART 0 has precedence when both are set to 0. Note that if UART mode is selected and
hardware flow control is disabled, Timer 1 or Timer 3 has precedence to use ports P1.2 and P1.3.
7.6.5 USART 1
The SFR register bit PERCFG.U1CFG selects whether to use alternative 1 or alternative 2 locations.
In Table 7-1, the USART 1 signals are shown as follows:
UART:
• RX: RXDATA
• TX: TXDATA
• RT: RTS
• CT: CTS
SPI:
• MI: MISO
• MO: MOSI
• C: SCK
• SS: SSN
82
I/O Ports
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Debug Interface
www.ti.com
P2DIR.PRIP0 selects the order of precedence when assigning several peripherals to Port 0. When set to
01, USART 1 has precedence. Note that if UART mode is selected and hardware flow control is disabled,
USART 0 or Timer 1 has precedence to use ports P0.2 and P0.3.
P2SEL.PRI3P1 and P2SEL.PRI2P1 select the order of precedence when assigning several peripherals
to Port 1. USART 1 has precedence when the former is set to 1 and the latter is set to 0. Note that if
UART mode is selected and hardware flow control is disabled, USART 0 or Timer 3 has precedence to
use ports P1.4 and P1.5.
7.6.6 ADC
In
•
•
•
•
•
•
•
•
•
Table 7-1, the ADC signals are shown as follows:
A0: ADC input 0
A1: ADC input 1
A2: ADC input 2
A3: ADC input 3
A4: ADC input 4
A5: ADC input 5
A6: ADC input 6
A7: ADC input 7
T: ADC external trigger pin
When using the ADC, Port 0 pins must be configured as ADC inputs. Up to eight ADC inputs can be used.
To configure a Port 0 pin to be used as an ADC input, the corresponding bit in the APCFG register must be
set to 1. The default values in this register select the Port 0 pins as non-ADC input, i.e., digital
input/outputs.
The settings in the APCFG register override the settings in P0SEL.
The ADC can be configured to use the general-purpose I/O pin P2.0 as an external trigger to start
conversions. P2.0 must be configured as a general-purpose I/O in input mode when being used for ADC
external trigger.
7.6.7 Operational Amplifier and Analog Comparator
When using the operational amplifier and analog comparator, the corresponding Port 0 pins must be
configured as ADC inputs (see Table 7-1). To configure a Port 0 pin to be used as an ADC input, the
corresponding bit in the APCFG register must be set to 1. The default values in this register select the Port
0 pins as non-ADC input, i.e., digital input/outputs.
The settings in the APCFG register override the settings in P0SEL.
7.7
Debug Interface
Ports P2.1 and P2.2 are used for debug data and clock signals, respectively. These are shown as DD
(debug data) and DC (debug clock) in Table 7-1. When in debug mode, the debug interface controls the
direction of these pins. Pullup/pulldown is disabled on these pins while in debug mode.
7.8
32-kHz XOSC Input
Ports P2.3 and P2.4 can be used to connect an external 32-kHz crystal. These port pins are used by the
32-kHz XOSC when CLKCONCMD.OSC32K is low, regardless of register settings. The port pins are set in
analog mode when CLKCONCMD.OSC32K is low.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Ports
83
Radio Test Output Signals
7.9
www.ti.com
Radio Test Output Signals
By using the OBSSELx registers (OBSSEL0–OBSSEL5) the user can output different signals from the RF
Core to GPIO pins. These signals can be useful for debugging of low-level protocols or control of external
PA, LNA, or switches. The control registers OBSSEL0–OBSSEL5 can be used to override the standard
GPIO behavior and output RF Core signals (rfc_obs_sig0, rfc_obs_sig1, and rfc_obs_sig2) on
the pins P1[0:5]. For a list of available signals, see the respective RFC_OBS_CTRLx registers in
Section 23.15.3 for CC253x or Section 24.1 for CC2540 or Chapter 25 for CC2541.
7.10 Power-Down Signal MUX (PMUX)
The PMUX register can be used to output the 32-kHz clock and/or the digital regulator status.
The selected 32-kHz clock source can be output on one of the P0 pins. The enable bit CKOEN enables the
output on P0, and the pin of P0 is selected using the CKOPIN (see the PMUX register description for
details). When CKOEN is set, all other configurations for the selected pin are overridden. The clock is
output in all power modes; however, in PM3 the clock stops (see PM3 in Chapter 4).
Furthermore, the digital regulator status can be output on one of the P1 pins. When the DREGSTA bit is
set, the status of the digital regulator is output. DREGSTAPIN selects the P1 pin (see the PMUX register
description for details). When DREGSTA is set, all other configurations for the selected pin are overridden.
The selected pin outputs 1 when the 1.8-V on-chip digital regulator is powered up (chip has regulated
power). The selected pin outputs 0 when the 1.8-V on-chip digital regulator is powered down, i.e., in PM2
and PM3.
7.11 I/O Registers
The registers for the I/O ports are described in this section. The registers are:
• P0: Port 0
• P1: Port 1
• P2: Port 2
• PERCFG: Peripheral-control register
• APCFG: Analog peripheral I/O configuration
• P0SEL: Port 0 function-select register
• P1SEL: Port 1 function-select register
• P2SEL: Port 2 function-select register
• P0DIR: Port 0 direction register
• P1DIR: Port 1 direction register
• P2DIR: Port 2 direction register
• P0INP: Port 0 input-mode register
• P1INP: Port 1 input-mode register
• P2INP: Port 2 input-mode register
• P0IFG: Port 0 interrupt-status flag register
• P1IFG: Port 1 interrupt-status flag register
• P2IFG: Port 2 interrupt-status flag register
• PICTL: Interrupt edge register
• P0IEN: Port 0 interrupt-mask register
• P1IEN: Port 1 interrupt-mask register
• P2IEN: Port 2 interrupt-mask register
• PMUX: Power-down signal-mux register
• OBSSEL0: Observation output control register 0
• OBSSEL1: Observation output control register 1
• OBSSEL2: Observation output control register 2
84
I/O Ports
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Registers
www.ti.com
•
•
•
OBSSEL3: Observation output control register 3
OBSSEL4: Observation output control register 4
OBSSEL5: Observation output control register 5
P0 (0x80) – Port 0
Bit
7:0
Name
P0[7:0]
Reset
0xFF
R/W
R/W
Description
Port 0. General-purpose I/O port. Bit-addressable from SFR. This CPU-internal register is readable,
but not writable, from XDATA (0x7080).
Reset
0xFF
R/W
R/W
Description
Port 1. General-purpose I/O port. Bit-addressable from SFR. This CPU-internal register is readable,
but not writable, from XDATA (0x7090).
Reset
000
1 1111
R/W
R0
R/W
Description
Reserved
Port 2. General-purpose I/O port. Bit-addressable from SFR. This CPU-internal register is readable,
but not writable, from XDATA (0x70A0).
P1 (0x90) – Port 1
Bit
7:0
Name
P1[7:0]
P2 (0xA0) – Port 2
Bit
7:5
4:0
Name
–
P2[4:0]
PERCFG (0xF1) – Peripheral Control
Bit
7
6
Name
–
T1CFG
Reset
0
0
R/W
R0
R/W
5
T3CFG
0
R/W
4
T4CFG
0
R/W
3:2
1
–
U1CFG
00
0
R/W
R/W
0
U0CFG
0
R/W
Description
Reserved
Timer 1 I/O location
0:
Alternative 1 location
1:
Alternative 2 location
Timer 3 I/O location
0:
Alternative 1 location
1:
Alternative 2 location
Timer 4 I/O location
0:
Alternative 1 location
1:
Alternative 2 location
Reserved
USART 1 I/O location
0:
Alternative 1 location
1:
Alternative 2 location
USART 0 I/O location
0:
Alternative 1 location
1:
Alternative 2 location
APCFG (0xF2) – Analog Peripheral I/O Configuration
Bit
7:0
Name
APCFG[7:0]
Reset
0x00
R/W
R/W
Description
Analog Perpheral I/O configuration . APCFG[7:0] select P0.7–P0.0 as analog I/O.
0: Analog I/O disabled
1: Analog I/O enabled
P0SEL (0xF3) – Port 0 Function Select
Bit
7:0
Name
SELP0_[7:0]
Reset
0x00
R/W
R/W
Description
P0.7 to P0.0 function select
0: General-purpose I/O
1: Peripheral function
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Ports
85
I/O Registers
www.ti.com
P1SEL (0xF4) – Port 1-Function Select
Bit
7:0
Name
SELP1_[7:0]
Reset
0x00
R/W
R/W
Description
P1.7 to P1.0 function select
0: General-purpose I/O
1: Peripheral function
P2SEL (0xF5) – Port 2 Function Select and Port 1 Peripheral Priority Control
Bit
7
6
Name
–
PRI3P1
Reset
0
0
R/W
R0
R/W
5
PRI2P1
0
R/W
4
PRI1P1
0
R/W
3
PRI0P1
0
R/W
2
SELP2_4
0
R/W
1
SELP2_3
0
R/W
0
SELP2_0
0
R/W
Description
Reserved
Port 1 peripheral priority control. This bit determines
modules are assigned to the same pins.
0:
USART 0 has priority.
1:
USART 1 has priority.
Port 1 peripheral priority control. This bit determines
assigns USART 1 and Timer 3 to the same pins.
0:
USART 1 has priority.
1:
Timer 3 has priority.
Port 1 peripheral priority control. This bit determines
assigns Timer 1 and Timer 4 to the same pins.
0:
Timer 1 has priority.
1:
Timer 4 has priority.
Port 1 peripheral priority control. This bit determines
assigns USART 0 and Timer 1 to the same pins.
0:
USART 0 has priority.
1:
Timer 1 has priority.
P2.4 function select
0:
General-purpose I/O
1:
Peripheral function
P2.3 function select
0:
General-purpose I/O
1:
Peripheral function
P2.0 function select
0:
General-purpose I/O
1:
Peripheral function
which module has priority in the case when
the order of priority in the case when PERCFG
the order of priority in the case when PERCFG
the order of priority in the case when PERCFG
P0DIR (0xFD) – Port 0 Direction
Bit
7:0
Name
Reset
DIRP0_[7:0] 0x00
R/W
R/W
Description
P0.7 to P0.0 I/O direction
0:
Input
1:
Output
P1DIR (0xFE) – Port 1 Direction
Bit
7:0
86
Name
Reset
DIRP1_[7:0] 0x00
I/O Ports
R/W
R/W
Description
P1.7 to P1.0 I/O direction
0:
Input
1:
Output
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Registers
www.ti.com
P2DIR (0xFF) – Port 2 Direction and Port 0 Peripheral Priority Control
Bit
7:6
Name
PRIP0[1:0]
Reset
00
R/W
R/W
Description
Port 0 peripheral priority control. These bits determine the order of priority in the case
when PERCFG assigns several peripherals to the same pins.
Detailed priority list:
00:
1st priority: USART 0
2nd priority: USART 1
3rd priority: Timer 1
01:
1st priority: USART 1
2nd priority: USART 0
3rd priority: Timer 1
10:
1st priority: Timer 1 channels 0-1
2nd priority: USART 1
3rd priority: USART 0
4th priority: Timer 1 channels 2-3
5
4:0
–
DIRP2_[4:0]
0
0 0000
11:
1st priority: Timer 1 channels 2-3
2nd priority: USART 0
3rd priority: USART 1
4th priority: Timer 1 channels 0-1
Reserved
P2.4 to P2.0 I/O direction
0:
Input
1:
Output
R0
R/W
P0INP (0x8F) – Port 0 Input Mode
Bit
7:0
Name
MDP0_[7:0]
Reset
0x00
R/W
R/W
Description
P0.7 to P0.0 I/O input mode
Pullup/pulldown [see P2INP
0:
1:
(0xF7) – Port 2 input mode]
3-state
P1INP (0xF6) – Port 1 Input Mode
Bit
7:2
1:0
Name
MDP1_[7:2]
–
Reset
R/W
0000 00 R/W
00
Description
P1.7 to P1.2 I/O input mode
Pullup/pulldown [see P2INP
0:
(0xF7) – Port 2 input mode]
1:
3-state
Reserved
R0
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Ports
87
I/O Registers
www.ti.com
P2INP (0xF7) – Port 2 Input Mode
Bit
7
Name
PDUP2
Reset
0
R/W
R/W
6
PDUP1
0
R/W
5
PDUP0
0
R/W
4:0
MDP2_[4:0]
0 0000
R/W
Description
Port 2 pullup/pulldown select. Selects function for all Port 2 pins configured as
pullup/pulldown inputs.
0:
Pullup
1:
Pulldown
Port 1 pullup/down select. Selects function for all Port 1 pins configured as pullup/pulldown
inputs.
0:
Pullup
1:
Pulldown
Port 0 pullup/pulldown select. Selects function for all Port 0 pins configured as
pullup/pulldown inputs.
0:
Pullup
1:
Pulldown
P2.4 to P2.0 I/O input mode
0:
Pullup/pulldown
1:
3-state
P0IFG (0x89) – Port 0 Interrupt Status Flag
Bit
7:0
Name
Reset
P0IF[7:0] 0x00
R/W
R/W0
Description
Port 0, inputs 7 to 0 interrupt status flags. When an input port pin has an interrupt request
pending, the corresponding flag bit is set.
P1IFG (0x8A) – Port 1 Interrupt Status Flag
Bit
7:0
Name
Reset
P1IF[7:0] 0x00
R/W
R/W0
Description
Port 1, inputs 7 to 0 interrupt status flags. When an input port pin has an interrupt request
pending, the corresponding flag bit is set.
P2IFG (0x8B) – Port 2 Interrupt Status Flag
Bit
7:6
5
Name
–
DPIF
4:0
P2IF[4:0] 0 0000
88
I/O Ports
Reset
00
0
R/W
R0
R/W0
R/W0
Description
Reserved
USB D+ interrupt-status flag. This flag is set when the D+ line has an interrupt request pending
and is used to detect USB resume events in USB suspend state. This flag is not set when the
USB controller is not suspended.
Port 2, inputs 4 to 0 interrupt status flags. When an input port pin has an interrupt request
pending, the corresponding flag bit is set.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Registers
www.ti.com
PICTL (0x8C) – Port Interrupt Control
Bit
7
Name
PADSC
Reset
0
R/W
R/W
6:4
3
–
P2ICON
000
0
R0
R/W
2
P1ICONH
0
R/W
1
P1ICONL
0
R/W
0
P0ICON
0
R/W
Description
Drive strength control for I/O pins in output mode. Selects output drive strength enhancement to
account for low I/O supply voltage on pin DVDD (this to ensure the same drive strength at lower
voltages as at higher).
0:
Minimum drive strength enhancement. DVDD1/2 equal to or greater than 2.6 V
1:
Maximum drive strength enhancement. DVDD1/2 less than 2.6 V
Reserved
Port 2, inputs 4 to 0 interrupt configuration. This bit selects the interrupt request condition for Port 2
inputs 4 to 0.
0:
Rising edge on input gives interrupt.
1:
Falling edge on input gives interrupt.
Port 1, inputs 7 to 4 interrupt configuration. This bit selects the interrupt request condition for the high
nibble of Port 1 inputs.
0:
Rising edge on input gives interrupt.
1:
Falling edge on input gives interrupt
Port 1, inputs 3 to 0 interrupt configuration. This bit selects the interrupt request condition for the low
nibble of Port 1 inputs.
0:
Rising edge on input gives interrupt.
1:
Falling edge on input gives interrupt.
Port 0, inputs 7 to 0 interrupt configuration. This bit selects the interrupt request condition for all Port
0 inputs.
0:
Rising edge on input gives interrupt.
1:
Falling edge on input gives interrupt.
P0IEN (0xAB) – Port 0 Interrupt Mask
Bit
7:0
Name
P0_[7:0]IEN
Reset
0x00
R/W
R/W
Description
Port P0.7 to P0.0 interrupt enable
0:
Interrupts are disabled.
1:
Interrupts are enabled.
P1IEN (0x8D) – Port 1 Interrupt Mask
Bit
7:0
Name
P1_[7:0]IEN
Reset
0x00
R/W
R/W
Description
Port P1.7 to P1.0 interrupt enable
0:
Interrupts are disabled.
1:
Interrupts are enabled.
P2IEN (0xAC) – Port 2 Interrupt Mask
Bit
7:6
5
Name
–
DPIEN
Reset
00
0
R/W
R0
R/W
4:0
P2_[4:0]IEN
0 0000 R/W
Description
Reserved
USB D+ interrupt enable
0:
USB D+ interrupt disabled.
1:
USB D+ interrupt enabled.
Port P2.4 to P2.0 interrupt enable
0:
Interrupts are disabled.
1:
Interrupts are enabled.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Ports
89
I/O Registers
www.ti.com
PMUX (0xAE) – Power-Down Signal Mux
Bit
7
Name
CKOEN
Reset
0
R/W
R/W
6:4
3
CKOPIN[2:0]
DREGSTA
000
0
R/W
R/W
2:0
DREGSTAPIN[2:0]
000
R/W
Description
Clock Out Enable. When this bit is set, the selected 32-kHz clock is output on one of the P0
pins. CKOPIN selects the pin to use. This overrides all other configuration for the selected pin.
The clock is output in all power modes; however, in PM3 the clock stops (see PM3 in
Chapter 4).
Clock Out Pin. Selects which P0 pin is to be used to output the selected 32-kHz clock.
Digital Regulator Status. When this bit is set, the status of the digital regulator is output on one
of the P1 pins. DREGSTAPIN selects the pin. When DREGSTA is set, all other configurations for
the selected pin are overridden. The selected pin outputs 1 when the 1.8-V on-chip digital
regulator is powered up (chip has regulated power). The selected pin outputs 0 when the 1.8V on-chip digital regulator is powered down.
Digital Regulator Status Pin. Selects which P1 pin is to be used to output DREGSTA signal.
Note that registers OBSSEL0 through OBSSEL5 do not retain data in states PM2 and PM3.
OBSSEL0 (0x6243) – Observation Output Control Register 0
Bit
7
Name
EN
Reset
0
R/W
R/W
6:0
SEL[6:0]
000 0000 R/W
Description
Bit controlling the observation output 0 on P1[0].
0 – Observation output disabled
1 – Observation output enabled
Note: If enabled, this overwrites the standard GPIO behavior of P1.0.
Select output signal on observation output 0
111 1011 (123): rfc_obs_sig0
111 1100 (124): rfc_obs_sig1
111 1101 (125): rfc_obs_sig2
Others: Reserved
OBSSEL1 (0x6244) – Observation Output Control Register 1
Bit
7
Name
EN
Reset
0
R/W
R/W
6:0
SEL[6:0]
000 0000 R/W
Description
Bit controlling observation output 1 on P1[1].
0 – Observation output disabled
1 – Observation output enabled
Note: If enabled, this overwrites the standard GPIO behavior of P1.1.
Select output signal on observation output 1
111 1011 (123): rfc_obs_sig0
111 1100 (124): rfc_obs_sig1
111 1101 (125): rfc_obs_sig2
Others: Reserved
OBSSEL2 (0x6245) – Observation Output Control Register 2
Bit
7
Name
EN
Reset
0
R/W
R/W
6:0
SEL[6:0]
000 0000 R/W
Description
Bit controlling observation output 2 on P1[2].
0 – Observation output disabled
1 – Observation output enabled
Note: If enabled, this overwrites the standard GPIO behavior of P1.2.
Select output signal on observation output 2
111 1011 (123): rfc_obs_sig0
111 1100 (124): rfc_obs_sig1
111 1101 (125): rfc_obs_sig2
Others: Reserved
OBSSEL3 (0x6246) – Observation Output Control Register 3
Bit
7
Name
EN
Reset
0
6:0
SEL[6:0]
000 0000 R/W
90
I/O Ports
R/W
R/W
Description
Bit controlling observation output 3 on P1[3].
0 – Observation output disabled
1 – Observation output enabled
Note: If enabled, this overwrites the standard GPIO behavior of P1.3.
Select output signal on observation output 3
111 1011 (123): rfc_obs_sig0
111 1100 (124): rfc_obs_sig1
111 1101 (125): rfc_obs_sig2
Others: Reserved
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Registers
www.ti.com
OBSSEL4 (0x6247) – Observation Output Control Register 4
Bit
7
Name
EN
Reset
0
R/W
R/W
6:0
SEL[6:0]
000 0000 R/W
Description
Bit controlling observation output 4 on P1[4].
0 – Observation output disabled
1 – Observation output enabled
Note: If enabled, this overwrites the standard GPIO behavior of P1.4.
Select output signal on observation output 4
111 1011 (123): rfc_obs_sig0
111 1100 (124): rfc_obs_sig1
11 11101 (125): rfc_obs_sig2
Others: Reserved
OBSSEL5 (0x6248) – Observation Output Control Register 5
Bit
7
Name
EN
Reset
0
R/W
R/W
6:0
SEL[6:0]
000 0000 R/W
Description
Bit controlling the observation output 5 on P1[5].
0 – Observation output disabled
1 – Observation output enabled
Note: If enabled, this overwrites the standard GPIO behavior of P1.5.
Select output signal on observation output 5
111 1011 (123): rfc_obs_sig0
111 1100 (124): rfc_obs_sig1
111 1101 (125): rfc_obs_sig2
Others: Reserved
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I/O Ports
91
Chapter 8
SWRU191D – April 2009 – Revised March 2013
DMA Controller
The Direct Memory Access (DMA) Controller can be used to relieve the 8051 CPU core of handling data
movement operations, thus achieving high overall performance with good power efficiency. The DMA
controller can move data from a peripheral unit such as ADC or RF transceiver to memory with minimum
CPU intervention.
The DMA controller coordinates all DMA transfers, ensuring that DMA requests are prioritized
appropriately relative to each other and to CPU memory access. The DMA controller contains a number of
programmable DMA channels for memory-memory data movement.
The DMA controller controls data transfers over the entire address range in XDATA memory space.
Because most of the SFR registers are mapped into the DMA memory space, these flexible DMA
channels can be used to unburden the CPU in innovative ways, e.g., to feed a USART with data from
memory or periodically to transfer samples between ADC and memory, etc. Use of the DMA can also
reduce system power consumption by keeping the CPU in a low-power mode without having to wake up to
move data to or from a peripheral unit (see Section 4.1.1 for CPU low-power mode). Note that
Section 2.2.3 describes the SFR registers that are not mapped into XDATA memory space.
The main features of the DMA controller are as follows:
• Five independent DMA channels
• Three configurable levels of DMA channel priority
• 32 configurable transfer trigger events
• Independent control of source and destination address
• Single, block and repeated transfer modes
• Supports length field in transfer data, setting variable transfer length
• Can operate in either word-size or byte-size mode
Topic
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
92
...........................................................................................................................
Page
DMA Operation ................................................................................................. 93
DMA Configuration Parameters ........................................................................... 95
DMA Configuration Setup ................................................................................... 97
Stopping DMA Transfers .................................................................................... 98
DMA Interrupts .................................................................................................. 98
DMA Configuration Data Structure ...................................................................... 98
DMA Memory Access ......................................................................................... 98
DMA Registers ................................................................................................ 101
DMA Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Operation
www.ti.com
8.1
DMA Operation
There are five DMA channels available in the DMA controller, numbered channel 0 through channel 4.
Each DMA channel can move data from one place within the DMA memory space to another, i.e.,
between XDATA locations.
In order to use a DMA channel, it must first be configured as described in Section 8.2 and Section 8.3.
Figure 8-1 shows the DMA state diagram.
Once a DMA channel has been configured, it must be armed before any transfers are allowed to be
initiated. A DMA channel is armed by setting the appropriate bit in the DMA channel-arm register DMAARM.
When a DMA channel is armed, a transfer begins when the configured DMA trigger event occurs. Note
that the time to arm one channel (i.e., get configuration data) takes nine system clocks; thus, if the
corresponding DMAARM bit is set and a trigger appears within the time it takes to configure the channel, the
wanted trigger is lost. If two or more DMA channels are armed simultaneously, the time for all channels to
be configured is longer (sequential read from memory). If all five are armed, it takes 45 system clocks, and
channel 1 is ready first, then channel 2, and lastly channel 0 (all within the last eight system clocks). There
are 32 possible DMA trigger events (see Table 8-1), e.g., UART transfer, timer overflow. The trigger event
to be used by a DMA channel is set by the DMA channel configuration; thus, no knowledge of this is
available until after the configuration has been read. The DMA trigger events are listed in Table 8-1.
In addition to starting a DMA transfer through the DMA trigger events, the user software may force a DMA
transfer to begin by setting the corresponding DMAREQ bit.
It should be noted that if the previously configured trigger source generates trigger events while DMA is
being configured, these are counted as missed events, and as soon as the DMA channel is ready, the
transfer is started. This occurs even though the new trigger source is not the same as the previous one. In
some situations, this leads to errors in the transfer. In order to account for this, trigger source 0 should be
the source between reconfigurations. This is achieved by setting up dummy source and destination
addresses, using fixed length of one byte, block transfer, and trigger source 0. Enabling a software trigger
(DMAREQ) clears missed-trigger counting, and no new triggers are generated while a new configuration is
fetched from memory (unless software writes to DMAREQ for this channel).
A DMAREQ bit is cleared only when the corresponding DMA transfer occurs. The DMAREQ bit is not cleared
when the channel is disarmed.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Controller
93
DMA Operation
www.ti.com
Initialization
Yes
Write DMA Channel
Configuration
No
Reconfigure?
DMA Channel Idle
Setting DMAARM.ABORT = 1 aborts all
channels where the DMAARMn bit is set
simultaneously.
that is , setting DMAARM = 0x85 aborts
channel 1 and channel 3.
No
DMAARM.DMAARMn
= 1?
Yes
Load DMA Channel
Configuration
DMA Channel Armed
Trigger or
DMAREQ.DMAREQn
= 1?
No
Yes
Transfer One Byte or
Word When Channel
is Granted Access
Modify Source/Destination
Address
Yes
Yes
Reached Transfer
Count?
Set Interrupt Flag
(DMAIRQ.DMAIFn = 1;
If IRQMASK == 1 then
IRCON.DMAIF = 1)
No
Repetitive Transfer
Mode?
DMAARMn = 0
No
Yes
Block Transfer
Mode?
No
F0033-01
Figure 8-1. DMA Operation
94
DMA Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Configuration Parameters
www.ti.com
8.2
DMA Configuration Parameters
Setup and control of the DMA operation is performed by the user software. This section describes the
parameters which must be configured before a DMA channel can be used. Section 8.3 describes how the
parameters are set up in software and passed to the DMA controller.
The behavior of each of the five DMA channels is configured with the following parameters:
Source address: The first address from which the DMA channel should read data.
Destination address: The first address to which the DMA channel should write the data read from the
source address. The user must ensure that the destination is writable.
Transfer count: The number of transfers to perform before rearming or disarming the DMA channel
and alerting the CPU with an interrupt request. The length can be defined in the configuration or it can
be defined as described next for the VLEN setting.
VLEN setting: The DMA channel is capable of variable-length transfers, using the first byte or word to
set the transfer length. When doing this, various options are available regarding how to count the
number of bytes to transfer.
Priority: The priority of the DMA transfers for the DMA channel with respect to the CPU and other
DMA channels and access ports.
Trigger event: All DMA transfers are initiated by so-called DMA trigger events. This trigger either
starts a DMA block transfer or a single DMA transfer. In addition to the configured trigger, a DMA
channel can always be triggered by setting its designated DMAREQ.DMAREQx flag. The DMA trigger
sources are described in Table 8-1.
Source and destination increment: The source and destination addresses can be controlled to
increment or decrement or not change.
Transfer mode: The transfer mode determines whether the transfer should be a single transfer or a
block transfer, or repeated versions of these.
Byte or word transfers: Determines whether each DMA transfer should be 8-bit (byte) or 16-bit
(word).
Interrupt mask: An interrupt request is generated on completion of the DMA transfer. The interrupt
mask bit controls whether the interrupt generation is enabled or disabled.
M8: Decide whether to use seven or eight bits per byte byte for transfer length. This is only applicable
when doing byte transfers.
A detailed description of all configuration parameters is given in Section 8.2.1 through Section 8.2.11.
8.2.1 Source Address
The address in XDATA memory where the DMA channel starts to read data. This can be any XDATA
address – in RAM, in the mapped flash bank (see MEMCTR.XBANK), XREG, or XDATA addressed SFR.
8.2.2 Destination Address
The first address to which the DMA channel should write the data read from the source address. The user
must ensure that the destination is writable. This can be any XDATA address – in RAM, XREG, or XDATA
addressed SFR.
8.2.3 Transfer Count
The number of bytes/words that must be transferred for the DMA transfer to be complete. When the
transfer count is reached, the DMA controller rearms or disarms the DMA channel and alerts the CPU with
an interrupt request. The transfer count can be defined in the configuration or it can be defined as
variable-length, as described in Section 8.2.4.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Controller
95
DMA Configuration Parameters
www.ti.com
8.2.4 VLEN Setting
The DMA channel is capable of using the first byte or word (for word, bits 12:0 are used) in source data as
the transfer length. This allows variable-length transfers. When using variable-length transfer, various
options regarding how to count number of bytes to transfer is given. In any case, the transfer-count (LEN)
setting is used as a maximum transfer count. If the transfer length specified by the first byte or word is
greater than LEN, then LEN bytes/words are transferred. When using variable-length transfers, then LEN
should be set to the largest allowed transfer length plus one.
Note that the M8 bit (Section 8.2.11) is only used when byte-size transfers are chosen.
Options which can be set with VLEN are the following:
1. Transfer number of bytes/words commanded by first byte/word + 1 (transfers the length byte/word, and
then as many bytes/words as dictated by the length byte/word)
2. Transfer number of bytes/words commanded by first byte/word
3. Transfer number of bytes/words commanded by first byte/word + 2 (transfers the length byte/word, and
then as many bytes/words as dictated by the length byte/word + 1)
4. Transfer number of bytes/words commanded by first byte/word + 3 (transfers the length byte/word, and
then as many bytes/words as dictated by the length byte/word + 2)
Figure 8-2 shows the VLEN options.
Byte/Word n + 2
Byte/Word n + 1
Time
Byte/Word n
Byte/Word n + 1
Byte/Word n
Byte/Word n
Byte/Word n – 1
Byte/Word n – 1
Byte/Word n – 1
Byte/Word n – 1
•
•
•
Byte/Word 3
•
•
•
Byte/Word 3
•
•
•
Byte/Word 3
•
•
•
Byte/Word 3
Byte/Word 2
Byte/Word 2
Byte/Word 2
Byte/Word 2
Byte/Word 1
Byte/Word 1
Byte/Word 1
Byte/Word 1
LENGTH = n
LENGTH = n
LENGTH = n
LENGTH = n
VLEN = 001
VLEN = 010
VLEN = 011
VLEN = 100
M0103-02
Figure 8-2. Variable Length (VLEN) Transfer Options
8.2.5 Trigger Event
Each DMA channel can be set up to sense on a single trigger. This field determines which trigger the
DMA channel senses.
8.2.6 Source and Destination Increment
When the DMA channel is armed or rearmed, the source and destination addresses are transferred to
internal address pointers. The possibilities for address increment are:
• Increment by zero. The address pointer remains fixed after each transfer.
• Increment by one. The address pointer increments one count after each transfer.
• Increment by two. The address pointer increments two counts after each transfer.
• Decrement by one. The address pointer decrements one count after each transfer.
where a count equals 1 byte in byte mode and 2 bytes in word mode.
96
DMA Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Configuration Setup
www.ti.com
8.2.7 DMA Transfer Mode
The transfer mode determines how the DMA channel behaves when it starts transferring data. There are
four transfer modes described as follows:
Single: On a trigger, a single DMA transfer occurs, and the DMA channel awaits the next trigger. After
the number of transfers specified by the transfer count is completed, the CPU is notified, and the DMA
channel is disarmed.
Block: On a trigger, the number of DMA transfers specified by the transfer count is performed as
quickly as possible, after which the CPU is notified and the DMA channel is disarmed.
Repeated single: On a trigger, a single DMA transfer occurs, and the DMA channel awaits the next
trigger. After the number of transfers specified by the transfer count is completed, the CPU is notified,
and the DMA channel is rearmed.
Repeated block: On a trigger, the number of DMA transfers specified by the transfer count is
performed as quickly as possible, after which the CPU is notified and the DMA channel is rearmed.
8.2.8 DMA Priority
A DMA priority is configurable for each DMA channel. The DMA priority is used to determine the winner in
the case of multiple simultaneous internal memory requests, and whether the DMA memory access should
have priority or not over a simultaneous CPU memory access. In case of an internal tie, a round-robin
scheme is used to ensure access for all. There are three levels of DMA priority:
High: Highest internal priority. DMA access always prevails over CPU access.
Normal: Second-highest internal priority. DMA access prevails over the CPU on at least every second
try.
Low: Lowest internal priority. DMA access always defers to a CPU access.
8.2.9 Byte or Word Transfers
Determines whether 8-bit (byte) or 16-bit (word) transfers are done.
8.2.10 Interrupt Mask
On completing a DMA transfer, the channel can generate an interrupt to the processor. This bit masks the
interrupt.
8.2.11 Mode 8 Setting
This field determines whether to use 7 or 8 bits per byte for transfer length. Only applicable when doing
byte transfers.
8.3
DMA Configuration Setup
The DMA channel parameters such as address mode, transfer mode, and priority, described in the
previous section, must be configured before a DMA channel can be armed and activated. The parameters
are not configured directly through SFR registers, but instead they are written in a special DMA
configuration data structure in memory. Each DMA channel in use requires its own DMA configurationdata structure. The DMA configuration data structure consists of eight bytes and is described in
Section 8.6. A DMA configuration-data structure may reside at any location decided on by the user
software, and the address location is passed to the DMA controller through a set of SFRs,
DMAxCFGH:DMAxCFGL. Once a channel has been armed, the DMA controller reads the configuration data
structure for that channel, given by the address in DMAxCFGH:DMAxCFGL.
It is important to note that the method for specifying the start address for the DMA configuration data
structure differs between DMA channel 0 and DMA channels 1–4 as follows:
DMA0CFGH:DMA0CFGL gives the start address for the DMA channel 0 configuration data structure.
DMA1CFGH:DMA1CFGL gives the start address for the DMA channel 1 configuration data structure,
followed by the channel 2–4 configuration data structures.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Controller
97
Stopping DMA Transfers
www.ti.com
Thus, the DMA controller expects the DMA configuration data structures for DMA channels 1–4 to lie in a
contiguous area in memory starting at the address held in DMA1CFGH:DMA1CFGL and consisting of 32
bytes.
8.4
Stopping DMA Transfers
Ongoing DMA transfers or armed DMA channels are aborted using the DMAARM register to disarm the
DMA channel.
One or more DMA channels are aborted by writing a 1 to the DMAARM.ABORT register bit, and at the same
time selecting which DMA channels to abort by setting the corresponding DMAARM.DMAARMx bits to 1.
When setting DMAARM.ABORT to 1, the DMAARM.DMAARMx bits for nonaborted channels must be written
as 0.
No DMA interrupt is generated when aborting an ongoing DMA transfer (disarming a DMA channel).
8.5
DMA Interrupts
Each DMA channel can be configured to generate an interrupt to the CPU on completing a DMA transfer.
This is accomplished with the IRQMASK bit in the channel configuration. The corresponding interrupt flag
in the DMAIRQ SFR register is set when the interrupt is generated.
Regardless of the IRQMASK bit in the channel configuration, the corresponding interrupt flag in the
DMAIRQ register is set on DMA channel completion. Thus, software should always check (and clear) this
register when rearming a channel with a changed IRQMASK setting. Failure to do so could generate an
interrupt based on the stored interrupt flag.
If a DMA transfer is aborted prior to its completion, the corresponding bit in the DMAIRQ register is not set,
and an interrupt is not generated.
8.6
DMA Configuration Data Structure
For each DMA channel, the DMA configuration data structure consists of eight bytes. The configuration
data structure is described in Table 8-2.
8.7
DMA Memory Access
The DMA data transfer is affected by endian convention. Note that the DMA descriptors follow big-endian
convention while the other registers follow little-endian convention. This must be accounted for in
compilers.
Table 8-1. DMA Trigger Sources
DMA Trigger
Number
98
Name
Functional Unit
Description
0
NONE
DMA
No trigger, setting the DMAREQ.DMAREQx bit starts transfer.
1
PREV
DMA
DMA channel is triggered by completion of previous channel.
2
T1_CH0
Timer 1
Timer 1, compare, channel 0
3
T1_CH1
Timer 1
Timer 1, compare, channel 1
4
T1_CH2
Timer 1
Timer 1, compare, channel 2
5
T2_EVENT1
Timer 2
Timer 2, event pulse 1
6
T2_EVENT2
Timer 2
Timer 2, event pulse 2
7
T3_CH0
Timer 3
Timer 3, compare, channel 0
8
T3_CH1
Timer 3
Timer 3, compare, channel 1
9
T4_CH0
Timer 4
Timer 4, compare, channel 0
10
T4_CH1
Timer 4
Timer 4, compare, channel 1
11
ST
Sleep Timer (not in
CC2540/41)
Sleep Timer compare
RADIO1
Radio (CC2541)
Radio DMA trigger 1 (see Section 25.3.2)
DMA Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Memory Access
www.ti.com
Table 8-1. DMA Trigger Sources (continued)
DMA Trigger
Number
Functional Unit
Name
Description
12
IOC_0
I/O controller
Port 0 I/O pin input transition (1)
13
IOC_1
I/O controller
Port 1 I/O pin input transition (1)
14
URX0
USART 0
USART 0 RX complete
15
UTX0
USART 0
USART 0 TX complete
16
URX1
USART 1
USART 1 RX complete
17
UTX1
USART 1
USART 1 TX complete
18
FLASH
Flash controller
Flash data write complete
19
RADIO
Radio (not in CC2540)
CC253x: RF packet byte received (see Section 23.3)
CC2541: Radio DMA trigger 0 (see Section 25.3.2)
20
ADC_CHALL
ADC
ADC end of a conversion in a sequence, sample ready
21
ADC_CH11
ADC
ADC end of conversion channel 0 in sequence, sample ready
22
ADC_CH21
ADC
ADC end of conversion channel 1 in sequence, sample ready
23
ADC_CH32
ADC
ADC end of conversion channel 2 in sequence, sample ready
24
ADC_CH42
ADC
ADC end of conversion channel 3 in sequence, sample ready
25
ADC_CH53
ADC
ADC end of conversion channel 4 in sequence, sample ready
26
ADC_CH63
ADC
ADC end of conversion channel 5 in sequence, sample ready
27
ADC_CH74
ADC
ADC end of conversion channel 6 in sequence, sample ready
28
ADC_CH84
ADC
ADC end of conversion channel 7 in sequence, sample ready
29
ENC_DW
AES
AES encryption processor requests download of input data
30
ENC_UP
AES
AES encryption processor requests upload of output data
DBG_BW
Debug interface
Debug interface burst write
31
(1)
Using this trigger source must be aligned with port interrupt-enable bits. Note that all interrupt-enabled port pins generate a
trigger.
Table 8-2. DMA Configuration Data Structure
Byte
Offset
Bit
Name
Description
0
7:0
SRCADDR[15:8]
DMA channel source address, high
1
7:0
SRCADDR[7:0]
DMA channel source address, low
2
7:0
DESTADDR[15:8]
DMA channel destination address, high. Note that flash memory is not directly writable.
3
7:0
DESTADDR[7:0]
DMA channel destination address, low. Note that flash memory is not directly writable.
4
7:5
VLEN[2:0]
Variable-length transfer mode. In word mode, bits 12:0 of the first word are considered as
the transfer length.
4
4:0
000:
Use LEN for transfer count
001:
Transfer the number of bytes/words specified by the first byte/word + 1 (up to a
maximum specified by LEN). Thus, the transfer count excludes the length
byte/word.
010:
Transfer the number of bytes/words specified by the first byte/word (up to a
maximum specified by LEN). Thus, the transfer count includes the length byte/word.
011:
Transfer the number of bytes/words specified by the first byte/word + 2 (up to a
maximum specified by LEN).
100:
Transfer the number of bytes/words specified by the first byte/word + 3 (up to a
maximum specified by LEN).
101:
Reserved
110:
Reserved
111:
Alternative for using LEN as the transfer count
The DMA channel transfer count
LEN[12:8]
Used as the maximum allowable length when VLEN differs from 000 and 111. The DMA
channel counts in words when in WORDSIZE mode, and in bytes otherwise.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Controller
99
DMA Memory Access
www.ti.com
Table 8-2. DMA Configuration Data Structure (continued)
Byte
Offset
5
Bit
7:0
Name
LEN[7:0]
Description
The DMA channel transfer count
Used as the maximum allowable length when VLEN differs from 000 and 111. The DMA
channel counts in words when in WORDSIZE mode, and in bytes otherwise.
6
7
WORDSIZE
Selects whether each DMA transfer is 8-bit (0) or 16-bit (1).
6
6:5
TMODE[1:0]
The DMA channel transfer mode
Single
01:
Block
10:
Repeated single
11:
Repeated block
6
4:0
TRIG[4:0]
Selects one of the triggers shown in Table 8-1
7
7:6
SRCINC[1:0]
Source address increment mode (after each transfer):
7
7
7
7
100
00:
5:4
3
2
1:0
DMA Controller
DESTINC[1:0]
IRQMASK
M8
PRIORITY[1:0]
00:
0 bytes/words
01:
1 byte/word
10:
2 bytes/word
11:
–1 byte/word
Destination address increment mode (after each transfer):
00:
0 bytes/words
01:
1 byte/word
10:
2 bytes/words
11:
–1 byte/word
Interrupt mask for this channel.
0:
Disable interrupt generation
1:
Enable interrupt generation on DMA channel done
Mode of 8th bit for VLEN transfer length; only applicable when WORDSIZE = 0 and VLEN
differs from 000 and 111.
0:
Use all 8 bits for transfer count
1:
Use 7 LSB for transfer count
The DMA channel priority:
00:
Low, CPU has priority.
01:
Assured, DMA at least every second try
10:
High, DMA has priority
11:
Reserved
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Registers
www.ti.com
8.8
DMA Registers
This section describes the SFR registers associated with the DMA controller.
DMAARM (0xD6) – DMA Channel Arm
Bit
7
6:5
4
3
2
1
0
Name
ABORT
Reset
0
–
DMAARM4
00
0
R/W
R/W1
0
DMAARM3
R/W1
0
DMAARM2
R/W1
0
DMAARM1
R/W1
0
DMAARM0
R/W
R0/W
R/W1
Description
DMA abort. This bit is used to stop ongoing DMA transfers. Writing a 1 to this bit aborts all
channels which are selected by setting the corresponding DMAARM bit to 1.
0:
Normal operation
1:
Abort all selected channels
Reserved
DMA arm channel 4
This bit must be set in order for any DMA transfers to occur on
transfer modes, the bit is automatically cleared on completion.
DMA arm channel 3
This bit must be set in order for any DMA transfers to occur on
transfer modes, the bit is automatically cleared on completion.
DMA arm channel 2
This bit must be set in order for any DMA transfers to occur on
transfer modes, the bit is automatically cleared on completion.
DMA arm channel 1
This bit must be set in order for any DMA transfers to occur on
transfer modes, the bit is automatically cleared on completion.
DMA arm channel 0
This bit must be set in order for any DMA transfers to occur on
transfer modes, the bit is automatically cleared on completion.
the channel. For nonrepetitive
the channel. For nonrepetitive
the channel. For nonrepetitive
the channel. For nonrepetitive
the channel. For nonrepetitive
DMAREQ (0xD7) – DMA Channel Start Request and Status
Bit
7:5
4
3
2
1
0
Name
–
DMAREQ4
DMAREQ3
DMAREQ2
DMAREQ1
DMAREQ0
Reset
000
0
0
0
0
0
R/W
R0
R/W1 H0
R/W1 H0
R/W1 H0
R/W1 H0
R/W1 H0
Description
Reserved
DMA transfer request, channel 4
When set to 1, activate the DMA channel
is cleared when DMA transfer is started.
DMA transfer request, channel 3
When set to 1, activate the DMA channel
is cleared when DMA transfer is started.
DMA transfer request, channel 2
When set to 1, activate the DMA channel
is cleared when DMA transfer is started.
DMA transfer request, channel 1
When set to 1, activate the DMA channel
is cleared when DMA transfer is started.
DMA transfer request, channel 0
When set to 1, activate the DMA channel
is cleared when DMA transfer is started.
(has the same effect as a single trigger event). This bit
(has the same effect as a single trigger event). This bit
(has the same effect as a single trigger event). This bit
(has the same effect as a single trigger event). This bit
(has the same effect as a single trigger event). This bit
DMA0CFGH (0xD5) – DMA Channel-0 Configuration Address High Byte
Bit
7:0
Name
DMA0CFG[15:8]
Reset
0x00
R/W
R/W
Description
The DMA channel-0 configuration address, high-order
DMA0CFGL (0xD4) – DMA Channel-0 Configuration Address Low Byte
Bit
7:0
Name
DMA0CFG[7:0]
Reset
0x00
R/W
R/W
Description
The DMA channel 0 configuration address, low-order
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA Controller
101
DMA Registers
www.ti.com
DMA1CFGH (0xD3) – DMA Channel 1–4 Configuration Address High Byte
Bit
7:0
Name
Reset
DMA1CFG[15:8] 0x00
R/W
R/W
Description
The DMA channel 1–4 configuration address, high-order
DMA1CFGL (0xD2) – DMA Channel 1–4 Configuration Address Low Byte
Bit
7:0
Name
Reset
DMA1CFG[7:0] 0x00
R/W
R/W
Description
The DMA channel 1–4 configuration address, low-order
DMAIRQ (0xD1) – DMA Interrupt Flag
Bit
7:5
4
3
2
1
0
102
Name
–
DMAIF4
DMAIF3
DMAIF2
DMAIF1
DMAIF0
Reset
000
0
0
0
0
0
DMA Controller
R/W
R0
R/W0
R/W0
R/W0
R/W0
R/W0
Description
Reserved
DMA channel-4 interrupt flag
0:
DMA channel transfer
1:
DMA channel transfer
DMA channel-3 interrupt flag
0:
DMA channel transfer
1:
DMA channel transfer
DMA channel-2 interrupt flag
0:
DMA channel transfer
1:
DMA channel transfer
DMA channel-1 interrupt flag
0:
DMA channel transfer
1:
DMA channel transfer
DMA channel-0 interrupt flag
0:
DMA channel transfer
1:
DMA channel transfer
not complete
complete/interrupt pending
not complete
complete/interrupt pending
not complete
complete/interrupt pending
not complete
complete/interrupt pending
not complete
complete/interrupt pending
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 9
SWRU191D – April 2009 – Revised March 2013
Timer 1 (16-Bit Timer)
Timer 1 is an independent 16-bit timer which supports typical timer/counter functions such as input
capture, output compare, and PWM functions. The timer has five independent capture/compare channels.
The timer uses one I/O pin per channel. The timer is used for a wide range of control and measurement
applications, and the availability of up/down count mode with five channels allows, for example,
implementation of motor-control applications.
The features of Timer 1 are as follows:
• Five capture/compare channels
• Rising, falling, or any-edge input capture
• Set, clear or toggle output compare
• Free-running, modulo, or up/down counter operation
• Clock prescaler for divide by 1, 8, 32, or 128
• Interrupt request generated on each capture/compare and terminal count
• DMA trigger function
Topic
9.1
9.2
9.3
9.4
9.5
9.6
9.7
9.8
9.9
9.10
9.11
9.12
9.13
...........................................................................................................................
16-Bit Counter .................................................................................................
Timer 1 Operation ............................................................................................
Free-Running Mode .........................................................................................
Modulo Mode ..................................................................................................
Up/Down Mode ................................................................................................
Channel-Mode Control .....................................................................................
Input Capture Mode .........................................................................................
Output Compare Mode .....................................................................................
IR Signal Generation and Learning ....................................................................
Timer 1 Interrupts ............................................................................................
Timer 1 DMA Triggers ......................................................................................
Timer 1 Registers ............................................................................................
Accessing Timer 1 Registers as Array ................................................................
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
Page
104
104
104
105
105
105
106
106
111
113
113
114
119
103
16-Bit Counter
9.1
www.ti.com
16-Bit Counter
The timer consists of a 16-bit counter that increments or decrements at each active clock edge. The
period of the active clock edges is defined by the register bits, CLKCONCMD.TICKSPD, which set the
global division of the system clock, giving a variable clock-tick frequency from 0.25 MHz to 32 MHz (given
the use of the 32-MHz XOSC as clock source). This frequency is further divided in Timer 1 by the
prescaler value set by T1CTL.DIV. This prescaler value can be 1, 8, 32, or 128. Thus, the lowest clock
frequency used by Timer 1 is 1953.125 Hz and the highest is 32 MHz when the 32 MHz XOSC is used as
system clock source. When the 16-MHz RCOSC is used as system clock source, then the highest clock
frequency used by Timer 1 is 16 MHz.
The counter operates as a free-running counter, a modulo counter, or an up/down counter for use in
center-aligned PWM.
It is possible to read the 16-bit counter value through the two 8-bit SFRs, T1CNTH and T1CNTL, containing
the high-order byte and low-order byte, respectively. When T1CNTL is read, the high-order byte of the
counter at that instant is buffered in T1CNTH so that the high-order byte can be read from T1CNTH. Thus,
T1CNTL must always be read first, before reading T1CNTH.
All write accesses to the T1CNTL register reset the 16-bit counter.
The counter produces an interrupt request when the terminal count value (overflow) is reached. It is
possible to start and halt the counter with T1CTL control register settings. The counter is started when a
value other than 00 is written to T1CTL.MODE. If 00 is written to T1CTL.MODE, the counter halts at its
present value.
9.2
Timer 1 Operation
In general, control register T1CTL is used to control the timer operation. The status register T1STAT holds
the interrupt flags. The various modes of operation are described as follows.
9.3
Free-Running Mode
In the free-running mode of operation, the counter starts from 0x0000 and increments at each active clock
edge. When the counter reaches 0xFFFF (overflow), the counter is loaded with 0x0000 and continues
incrementing its value as shown in Figure 9-1. When the terminal count value 0xFFFF is reached, the
interrupt flag T1STAT.OVFIF is set. An interrupt request is generated if enabled, see Section 9.10 for
details. The free-running mode can be used to generate independent time intervals and output signal
frequencies.
FFFFh
0000h
OVFL
OVFL
T0308-01
Figure 9-1. Free-Running Mode
104
Timer 1 (16-Bit Timer)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Modulo Mode
www.ti.com
9.4
Modulo Mode
When the timer operates in modulo mode, the 16-bit counter starts at 0x0000 and increments at each
active clock edge. After the counter has reached the period value T1CC0, held in registers
T1CC0H:T1CC0L, the counter is reset to 0x0000 and continues to increment. If the timer is started with a
value above T1CC0, the interrupt flag T1STAT.OVFIF is set when the terminal count value (0xFFFF) is
reached, after which the counter wraps to 0x0000. An interrupt request is generated if enabled, see
Section 9.10 for details. If a periodic interrupt is wanted at the period value, this can be obtained by
enabling an output compare interrupt on channel 0, as explained in Section 9.8. The modulo mode can be
used for applications where a period other then 0xFFFF is required. The counter operation is shown in
Figure 9-2.
T1CC0
0000h
T0309-02
Figure 9-2. Modulo Mode
9.5
Up/Down Mode
In the up/down timer mode, the counter repeatedly starts from 0x0000 and counts up until the value held
in T1CC0H:T1CC0L is reached, and then the counter counts down until 0x0000 is reached, as shown in
Figure 9-3. This timer mode is used when symmetrical output pulses are required with a period other than
0xFFFF, and therefore allows implementation of center-aligned PWM output applications. The interrupt
flag T1STAT.OVFIF is set when the counter value reaches 0x0000 in the up/down mode. An interrupt
request is generated if enabled, see Section 9.10 for details..
T1CC0
0000h
OVFL
OVFL
T0310-01
Figure 9-3. Up/Down Mode
9.6
Channel-Mode Control
The channel mode is set for each channel with its control and status register, T1CCTLn. The settings
include input capture and output compare modes.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
105
Input Capture Mode
9.7
www.ti.com
Input Capture Mode
When a channel is configured as an input capture channel, the I/O pin associated with that channel is
configured as an input. After the timer has been started, a rising edge, falling edge, or any edge on the
input pin triggers a capture of the 16-bit counter contents into the associated capture register. Thus, the
timer is able to capture the time when an external event takes place.
NOTE: Before an I/O pin can be used by the timer, the required I/O pin must be configured as a
Timer 1 peripheral pin.
The channel input pin is synchronized to the internal system clock. Thus, pulses on the input pin must
have a minimum duration greater than the system clock period.
The content of the 16-bit capture register is read out from registers T1CCnH:T1CCnL.
When the capture takes place, the interrupt flag for the channel, T1STAT.CHnIF (n is the channel
number), is set. An interrupt request is generated if enabled, see Section 9.10 for details.
9.8
Output Compare Mode
In output compare mode, the I/O pin associated with a channel is set as an output. After the timer has
been started, the contents of the counter are compared with the contents of the channel compare register
T1CCnH:T1CCnL. If the compare register equals the counter contents, the output pin is set, reset, or
toggled, according to the compare output mode setting of T1CCTLn.CMP. Note that all edges on output
pins are glitch-free when operating in a given output compare mode. Writing to the compare register
T1CCnL is buffered, so that a value written to T1CCnL does not take effect until the corresponding highorder register, T1CCnH, is written. Writing to compare registers T1CCnH:T1CCnL does not take effect on
the output compare value until the counter value is 0x00.
Note that channel 0 has fewer output compare modes because T1CC0H:T1CC0L has a special function in
modes 6 and 7, meaning these modes would not be useful for channel 0.
When a compare occurs, the interrupt flag for the channel, T1STAT.CHnIF (n is the channel number), is
set. An interrupt request is generated if enabled, see Section 9.10 for details.
Examples of output compare modes in various timer modes are given in the following figures.
Edge-aligned: PWM output signals can be generated using the timer modulo mode and channels 1 and 2
in output compare mode 6 or 7 (defined by the T1CCTLn.CMP bits, where n is 1 or 2) as shown in
Figure 9-4. The period of the PWM signal is determined by the setting in T1CC0, and the duty cycle is
determined by T1CCn, where n is the PWM channel, 1 or 2.
The timer free-running mode may also be used. In this case, CLKCONCMD.TICKSPD and the prescaler
divider value in the T1CTL.DIV bits set the period of the PWM signal. The polarity of the PWM signal is
determined by whether output compare mode 6 or 7 is used.
PWM output signals can also be generated using output compare modes 4 and 5 as shown in Figure 9-4,
or by using modulo mode as shown in Figure 9-5. Using output compare mode 4 or 5 is preferred for
simple PWM.
Center-aligned: PWM outputs can be generated when the timer up/down mode is selected. The channel
output compare mode 4 or 5 (defined by T1CCTLn.CMP bits, where n is 1 or 2) is selected, depending on
the required polarity of the PWM signal. The period of the PWM signal is determined by T1CC0, and the
duty cycle for the channel output is determined by T1CCn, where n is the PWM channel, 1 or 2.
The center-aligned PWM mode is required by certain types of motor-drive applications, and typically less
noise is produced than in the edge-aligned PWM mode, because the I/O pin transitions are not lined up on
the same clock edge.
In some types of applications, a defined delay or dead time is required between outputs. Typically, this is
required for outputs driving an H-bridge configuration to avoid uncontrolled cross-conduction in one side of
the H-bridge. The delay or dead-time can be obtained in the PWM outputs by using T1CCn as shown in
the following:
106
Timer 1 (16-Bit Timer)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Output Compare Mode
www.ti.com
Assuming that channel 1 and channel 2 are used to drive the outputs using timer up/down mode and the
channels use output compare modes 4 and 5, respectively, then the timer period (in Timer 1 clock
periods) is:
tP = T1CC0 × 2
and the dead time, i.e., the time when both outputs are low, (in Timer 1 clock periods) is given by:
tD = T1CC1 – T1CC2
A compare output pin is initialized to the value listed in Table 9-1 when:
• a value is written to T1CNTL (all Timer 1 channels)
• 0x7 is written to T1CCTLn.CMP (channel n)
Table 9-1. Initial Compare Output Values (Compare Mode)
Compare Mode (T1CCTLn.CMP)
Initial Compare Output
Set output on compare (000)
0
Clear output on compare (001)
1
Toggle output on compare (010)
0
Set output on compare-up, clear on compare down in up-down mode (011)
0
In other modes than up-down mode, set output on compare, clear on 0 (011)
0
Clear output on compare-up, set on compare down in up-down mode (100)
1
In other modes than up-down mode, clear output on compare, set on 0 (100)
1
Clear when equal T1CC0, set when equal T1CCn (101)
0
Set when equal T1CC0, clear when equal T1CCn (110)
1
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
107
Output Compare Mode
www.ti.com
FFFFh
T1CC0
T1CCn
0000h
0 - Set Output on Compare
1 - Clear Output on Compare
2 - Toggle Output on Compare
3 - Set Output on Compare-Up,
Clear on 0
4 - Clear Output on Compare-Up,
Set on 0
5 - Clear When T1CC0, Set When T1CCn
6 - Set When T1CC0, Clear When T1CCn
T1CCn
T1CC0
T1CCn
T1CC0
T0311-01
Figure 9-4. Output Compare Modes, Timer Free-Running Mode
108
Timer 1 (16-Bit Timer)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Output Compare Mode
www.ti.com
T1CC0
0000h
0 - Set Output on Compare
1 - Clear Output on Compare
2 - Toggle Output on Compare
3 - Set Output on Compare-Up,
Clear on 0
4 - Clear Output on Compare-Up,
Set on 0
5 - Clear When T1CC0, Set When T1CCn
6 - Set When T1CC0, Clear When T1CCn
T1CCn
T1CC0
T1CCn
T1CC0
T0312-01
Figure 9-5. Output Compare Modes, Timer Modulo Mode
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
109
Output Compare Mode
www.ti.com
T1CC0
T1CCn
0000h
0 - Set Output on Compare
1 - Clear Output on Compare
2 - Toggle Output on Compare
3 - Set Output on Compare-Up,
Clear on Compare-Down
4 - Clear Output on Compare-Up,
Set on Compare-Down
5 - Clear When T1CC0, Set When T1CCn
6 - Set When T1CC0, Clear When T1CCn
T1CCn
T1CCn
T1CC0
T1CCn
T1CCn
T1CC0
T0313-01
Figure 9-6. Output Compare Modes, Timer Up/Down Mode
110
Timer 1 (16-Bit Timer)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
IR Signal Generation and Learning
www.ti.com
9.9
IR Signal Generation and Learning
This section describes how Timer 1 can be configured in IR generation mode, where it counts Timer 3
periods and the output is ANDed with the output of Timer 3 to generate modulated consumer IR signals
with minimal CPU interaction.
9.9.1 Introduction
Generation of IR signals for remote control is generally done in one of two ways:
• Modulated codes
• Non-modulated codes (C-codes, flash codes)
The device includes flexible timer functionality to implement generation and learning of both types of IR
signals with minimal CPU interaction. Most IR protocols can be implemented with only one CPU
intervention per command.
9.9.2 Modulated Codes
Modulated codes can be generated using Timer 1 (16-bit) and Timer 3 (8-bit). Timer 3 in modulo mode is
used to generate the carrier. Timer 3 has an individual prescaler for its input. Its period is set using
T3CC0. Timer 3 channel 1 is used for PWM output. The duty cycle of the carrier is set using T3CC1.
Channel 1 uses compare mode: Clear output on compare, set on 0x00 (T3CCTL1.CMP = 100). Table 9-2
shows the frequency error calculation for a 38-kHz carrier using Timer 3.
Table 9-2. Frequency Error Calculation for 38-kHz
Carrier
Description
System clock frequency
IR carrier frequency
Value
32,000 kHz
38 kHz
System clock period
0.00003125 ms
IR carrier period
0.026315789 ms
Timer prescaler
4
Timer period
0.000125 ms
Ideal timer value
210.5263158
True timer value
211
True timer period
0.026375 ms
True timer frequency
37.91469194 kHz
Period error
59.21052632 ns
Frequency error
85.30805687 Hz
Frequency error %
0.2245%
The IRCTL.IRGEN register bit enables IR generation mode in Timer 1. When the IRGEN bit is set, Timer
1 takes the output of the Timer 3 channel 1 compare signal as tick instead of the system tick. The Timer 1
period is set using T1CC0 with Timer 1 in modulo mode (T1CTL.MODE = 10) and channel 0 in compare
mode (T1CCTL0.MODE = 1). Channel 1 compare mode Clear output on compare, set on 0x0000
(T1CCTL1.CMP = 100) is used for output of the gating signal.
The number of mark carrier periods is set by T1CC1. T1CC1 must be updated every Timer 1 period by the
DMA or CPU. Note that an update to T1CC1 is buffered and does not take effect before Timer 1 reaches
0x0000.
The number of space carrier periods is set by T1CC0. Its value should be set to the total number of mark
and space carrier periods wanted. The compare values are buffered until the timer hits 0x0000.
The output of Timer 1 channel 1 is ANDed with that of Timer 3 channel 1 to form the IR output as shown
in Figure 9-7
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
111
IR Signal Generation and Learning
www.ti.com
Timer 3 Ch 1 Output
Timer 3
AND
Gate
Timer 3 Ch 0 Compare
IR OUT
Timer 1 Ch 1 Output
Timer 1
B0358-01
Figure 9-7. Block Diagram of Timers in IR-Generation Mode
The timing of the Timer 3 channel 1 output and Timer 1 channel 1 output signals is synchronized such that
no glitches are produced on the IR Out signal.
When the IRGEN bit is set, the IR out signal is routed to pins instead of the normal Timer 1 channel 1
output (see also Section 7.6.1).
Timer 1 Ch 1 Compare
Timer 3 Ch 0 Compare
Timer 3 Ch 1 Compare
Start Timers
Figure 9-8 shows the example of Timer 3 being initialized to a 33% duty cycle (T3CC0 = 3 × T3CC1).
Timer 1 has been initialized to 3.
Timer 3 Output
Timer 1 Output
IR Out
T0440-01
Figure 9-8. Modulated Waveform Example
To achieve a period of space only, T1CC1 should be set to 0x00.
9.9.3 Non-Modulated Codes
To generate non-modulated IR codes, Timer 1 is used in modulo mode. The period of the signal is given
by T1CC0, and the pulse duration is given by T1CC1. T1CC1 gives the length of the mark period, and
T1CC0 gives the total number of mark and space periods. The compare values are buffered until the timer
hits 0x0000. The compare values must be updated once every period by the DMA or CPU if they are not
to be kept the same.
112
Timer 1 (16-Bit Timer)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 Interrupts
www.ti.com
9.9.4 Learning
Learning is done by using the capture function of Timer 1 (16-bit) and Timer 3 (8-bit). Timer 3 can handle
the carrier frequency detection and Timer 1 can handle the code learning from the demodulated signal.
The circuit could be set up as described in Figure 9-9
IR
Demod
Timer 1 Ch 2
CC253x
CC2540
CC2541
PIN
Diode
Timer 3 Ch 1
B0359-01
Figure 9-9. IR Learning Board Diagram
9.9.4.1
Carrier Frequency Detection
Timer 3 is used to capture and detect the carrier frequency with input directly from the IR PIN diode. The
timer should sample the carrier a limited number of times. If a carrier is detected, the frequency detected
should contribute to the average number, which is what can be stored in the database.
9.9.4.2
Demodulated Code Learning
The output from the IR PIN diode is demodulated by an appropriate circuit. The output from this circuit is
used as input to one of the Timer 1 channels in capture mode.
9.9.5 Other Considerations
The IR output pin should be placed in the high-impedance state or pulled down during reset to avoid
unnecessary power consumption from illuminating the IR LED. Note that only the P1.1 output for Timer 1
channel 1 is placed in the high-impedance state with no pullup during and after reset.
9.10 Timer 1 Interrupts
One interrupt vector is assigned to the timer. An interrupt request is generated when one of the following
timer events occurs:
• Counter reaches terminal count value (overflow, or turns around zero).
• Input capture event
• Output compare event
The status register, T1STAT, contains the source interrupt flags for the terminal-count value event and the
five channel compare/capture events. A source interrupt flag is set when the corresponding event occurs,
regardless of interrupt mask bits. The CPU interrupt flag IRCON.T1IF is set when one of the events
occurs if the corresponding interrupt mask bit is equal to 1. The interrupt mask bits are T1CCTLn.IM for
the five channels and TIMIF.T1OVFIM for the overflow event. The CPU interrupt flag IRCON.T1IF is
also set when a Timer 1 source interrupt flag is being cleared and one or more other Timer 1 source
interrupt flags are still set while the corresponding interrupt mask bit is set. An interrupt request is
generated when IRCON.T1IF goes from 0 to 1 if IEN1.T1IEN and IEN0.EA are both equal to 1.
9.11 Timer 1 DMA Triggers
There are three DMA triggers associated with Timer 1. These are DMA triggers T1_CH0, T1_CH1, and
T1_CH2, which are generated on timer compare events as follows:
• T1_CH0 – Channel 0 compare
• T1_CH1 – Channel 1 compare
• T1_CH2 – Channel 2 compare
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
113
Timer 1 Registers
www.ti.com
There are no triggers associated with channels 3 and 4.
9.12 Timer 1 Registers
This section describes the Timer 1 registers, which consist of the following registers:
• T1CNTH – Timer 1 count high
• T1CNTL – Timer 1 count low
• T1CTL – Timer 1 control
• T1STAT – Timer 1 status
• T1CCTLn – Timer 1 channel n capture/compare control
• T1CCnH – Timer 1 channel n capture/compare value high
• T1CCnL – Timer 1 channel n capture/compare value low
The TIMIF.T1OVFIM register bit resides in the TIMIF register, which is described together with the
Timer 3 and Timer 4 registers.
T1CNTH (0xE3) – Timer 1 Counter High
Bit
7:0
Name
Reset
CNT[15:8 0x00
]
R/W
R
Description
Timer count high-order byte. Contains the high byte of the 16-bit timer counter buffered at the time
T1CNTL is read
T1CNTL (0xE2) – Timer 1 Counter Low
Bit
7:0
Name
Reset
CNT[7:0] 0x00
R/W
R/W
Description
Timer count low-order byte. Contains the low byte of the 16-bit timer counter. Writing anything to
this register results in the counter being cleared to 0x0000 and initializes all output pins of
associated channels.
T1CTL (0xE4) – Timer 1 Control
Bit
7:4
3:2
Name
–
DIV[1:0]
Reset
0000
00
R/W
R0
R/W
1:0
MODE
[1:0]
00
R/W
Description
Reserved
Prescaler divider value. Generates the active clock edge used to update the counter as follows:
00: Tick frequency/1
01: Tick frequency/8
10: Tick frequency/32
11: Tick frequency/128
Timer 1 mode select. The timer operating mode is selected as follows:
00:
01:
10:
11:
114
Timer 1 (16-Bit Timer)
Operation is suspended.
Free-running, repeatedly count from 0x0000 to 0xFFFF.
Modulo, repeatedly count from 0x0000 to T1CC0.
Up/down, repeatedly count from 0x0000 to T1CC0 and from T1CC0 down to 0x0000.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 Registers
www.ti.com
T1STAT (0xAF) – Timer 1 Status
Bit
7:6
5
Name
–
OVFIF
Reset
00
0
R/W
R0
R/W0
4
CH4IF
0
R/W0
3
CH3IF
0
R/W0
2
CH2IF
0
R/W0
1
CH1IF
0
R/W0
0
CH0IF
0
R/W0
Description
Reserved
Timer 1 counter-overflow interrupt flag. Set when the counter reaches the terminal count value in freerunning or modulo mode, and when zero is reached counting down in up-down mode. Writing a 1 has
no effect.
Timer 1 channel 4 interrupt flag. Set when the channel 4 interrupt condition occurs. Writing a 1 has no
effect.
Timer 1 channel 3 interrupt flag. Set when the channel 3 interrupt condition occurs. Writing a 1 has no
effect.
Timer 1 channel 2 interrupt flag. Set when the channel 2 interrupt condition occurs. Writing a 1 has no
effect.
Timer 1 channel 1 interrupt flag. Set when the channel 1 interrupt condition occurs. Writing a 1 has no
effect.
Timer 1 channel 0 interrupt flag. Set when the channel 0 interrupt condition occurs. Writing a 1 has no
effect.
T1CCTL0 (0xE5) – Timer 1 Channel 0 Capture/Compare Control
Bit
7
6
5:3
Name
RFIRQ
IM
CMP[2:0]
Reset
0
1
000
2
MODE
0
1:0
CAP[1:0] 00
R/W
R/W
R/W
R/W
R/W
R/W
Description
When set, use RF interrupt for capture instead of regular capture input.
Channel 0 interrupt mask. Enables interrupt request when set.
Channel 0 compare-mode select. Selects action on output when timer value equals compare value in
T1CC0
000:
Set output on compare
001:
Clear output on compare
010:
Toggle output on compare
011:
Set output on compare-up, clear on 0
100:
Clear output on compare-up, set on 0
101:
Reserved
110:
Reserved
111:
Initialize output pin. CMP[2:0] is not changed.
Mode. Select Timer 1 channel 0 capture or compare mode
0:
Capture mode
1:
Compare mode
Channel 0 capture-mode select
00:
No capture
01:
Capture on rising edge
10:
Capture on falling edge
11:
Capture on all edges
T1CC0H (0xDB) – Timer 1 Channel 0 Capture/Compare Value, High
Bit
7:0
Name
Reset
T1CC0[15:8] 0x00
R/W
R/W
Description
Timer 1 channel 0 capture/compare value high-order byte. Writing to this register when
T1CCTL0.MODE = 1 (compare mode) causes the T1CC0[15:0] update to the written value to be
delayed until T1CNT = 0x0000.
T1CC0L (0xDA) – Timer 1 Channel 0 Capture/Compare Value, Low
Bit
7:0
Name
T1CC0[7:0]
Reset
0x00
R/W
R/W
Description
Timer 1 channel 0 capture/compare value low-order byte. Data written to this register is stored
in a buffer but not written to T1CC0[7:0] until, and at the same time as, a later write to T1CC0H
takes effect.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
115
Timer 1 Registers
www.ti.com
T1CCTL1 (0xE6) – Timer 1 Channel 1 Capture/Compare Control
Bit
7
6
5:3
Name
RFIRQ
IM
CMP[2:0]
Reset
0
1
000
R/W
R/W
R/W
R/W
Description
When set, use RF interrupt for capture instead of regular capture input.
Channel 1 interrupt mask. Enables interrupt request when set.
Channel 1 compare-mode select. Selects action on output when timer value equals compare value in
T1CC1.
000:
001:
010:
011:
100:
2
MODE
0
1:0
CAP[1:0] 00
R/W
R/W
Set output on compare
Clear output on compare
Toggle output on compare
Set output on compare-up, clear on compare-down in up-down mode. Otherwise set output on
compare, clear on 0.
Clear output on compare-up, set on compare-down in up-down mode. Otherwise clear output
on compare, set on 0.
Clear when equal T1CC0, set when equal T1CC1
Set when equal T1CC0, clear when equal T1CC1
Initialize output pin. CMP[2:0] is not changed.
101:
110:
111:
Mode. Select Timer 1 channel 1 capture or compare mode
0:
Capture mode
1:
Compare mode
Channel 1 capture-mode select
00:
No capture
01:
Capture on rising edge
10:
Capture on falling edge
11:
Capture on all edges
T1CC1H (0xDD) – Timer 1 Channel 1 Capture/Compare Value, High
Bit
7:0
Name
Reset
T1CC1[15:8] 0x00
R/W
R/W
Description
Timer 1 channel 1 capture/compare value high-order byte. Writing to this register when
T1CCTL1.MODE = 1 (compare mode) causes the T1CC1[15:0] update to the written value to be
delayed until T1CNT = 0x0000.
T1CC1L (0xDC) – Timer 1 Channel 1 Capture/Compare Value, Low
Bit
7:0
116
Name
T1CC1[7:0]
Reset
0x00
Timer 1 (16-Bit Timer)
R/W
R/W
Description
Timer 1 channel 1 capture/compare value low-order byte. Data written to this register is stored in
a buffer but not written to T1CC1[7:0] until, and at the same time as, a later write to T1CC1H
takes effect.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 Registers
www.ti.com
T1CCTL2 (0xE7) – Timer 1 Channel 2 Capture/Compare Control
Bit
7
6
5:3
Name
RFIRQ
IM
CMP[2:0]
Reset
0
1
000
R/W
R/W
R/W
R/W
Description
When set, use RF interrupt for capture instead of regular capture input.
Channel 2 interrupt mask. Enables interrupt request when set.
Channel 2 compare mode select. Selects action on output when timer value equals compare value in
T1CC2.
000:
001:
010:
011:
100:
2
MODE
0
1:0
CAP[1:0] 00
R/W
R/W
Set output on compare
Clear output on compare
Toggle output on compare
Set output on compare-up, clear on compare-down in up-down mode. Otherwise set output on
compare, clear on 0.
Clear output on compare-up, set on compare-down in up-down mode. Otherwise clear output
on compare, set on 0.
Clear when equal T1CC0, set when equal T1CC2
Set when equal T1CC0, clear when equal T1CC2
Initialize output pin. CMP[2:0] is not changed.
101:
110:
111:
Mode. Select Timer 1 channel 2 capture or compare mode
0:
Capture mode
1:
Compare mode
Channel 2 capture-mode select
00:
No capture
01:
Capture on rising edge
10:
Capture on falling edge
11:
Capture on all edges
T1CC2H (0xDF) – Timer 1 Channel 2 Capture/Compare Value, High
Bit
7:0
Name
Reset
T1CC2[15:8] 0x00
R/W
R/W
Description
Timer 1 channel 2 capture/compare value high-order byte. Writing to this register when
T1CCTL2.MODE = 1 (compare mode) causes the T1CC2[15:0] update to the written value to be
delayed until T1CNT = 0x0000.
T1CC2L (0xDE) – Timer 1 Channel 2 Capture/Compare Value, Low
Bit
7:0
Name
T1CC2[7:0]
Reset
0x00
R/W
R/W
Description
Timer 1 channel 2 capture/compare value low-order byte. Data written to this register is stored in
a buffer but not written to T1CC2[7:0] until, and at the same time as, a later write to T1CC2H
takes effect.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
117
Timer 1 Registers
www.ti.com
T1CCTL3 (0x62A3) – Timer 1 Channel 3 Capture/Compare Control
Bit
7
6
5:3
Name
RFIRQ
IM
CMP[2:0]
Reset
0
1
000
R/W
R/W
R/W
R/W
Description
When set, use RF interrupt for capture instead of regular capture input.
Channel 3 interrupt mask. Enables interrupt request when set.
Channel 3 compare mode select. Selects action on output when timer value equals compare value in
T1CC3.
000:
001:
010:
011:
100:
2
MODE
0
1:0
CAP[1:0] 00
R/W
R/W
Set output on compare
Clear output on compare
Toggle output on compare
Set output on compare-up, clear on compare-down in up-down mode. Otherwise set output on
compare, clear on 0.
Clear output on compare-up, set on compare down in up-down mode. Otherwise clear output
on compare, set on 0.
Clear when equal T1CC0, set when equal T1CC3
Set when equal T1CC0, clear when equal T1CC3
Initialize output pin. CMP[2:0] is not changed.
101:
110:
111:
Mode. Select Timer 1 channel 3 capture or compare mode
0:
Capture mode
1:
Compare mode
Channel 3 capture-mode select
00:
No capture
01:
Capture on rising edge
10:
Capture on falling edge
11:
Capture on all edges
T1CC3H (0x62AD) – Timer 1 Channel 3 Capture/Compare Value, High
Bit
7:0
Name
Reset
T1CC3[15:8] 0x00
R/W
R/W
Description
Timer 1 channel 3 capture/compare value high-order byte. Writing to this register when
T1CCTL3.MODE = 1 (compare mode) causes the T1CC3[15:0] update to the written value to be
delayed until T1CNT = 0x0000.
T1CC3L (0x62AC) – Timer 1 Channel 3 Capture/Compare Value, Low
Bit
7:0
118
Name
T1CC3[7:0]
Reset
0x00
Timer 1 (16-Bit Timer)
R/W
R/W
Description
Timer 1 channel 3 capture/compare value low-order byte. Data written to this register is stored in
a buffer but not written to T1CC3[7:0] until, and at the same time as, a later write to T1CC3H
takes effect.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Accessing Timer 1 Registers as Array
www.ti.com
T1CCTL4 (0x62A4) – Timer 1 Channel 4 Capture/Compare Control
Bit
7
6
5:3
Name
RFIRQ
IM
CMP[2:0]
Reset
0
1
000
R/W
R/W
R/W
R/W
Description
When set, use RF interrupt for capture instead of regular capture input.
Channel 4 interrupt mask. Enables interrupt request when set.
Channel 4 compare mode select. Selects action on output when timer value equals compare value in
T1CC4.
000:
001:
010:
011:
100:
2
MODE
0
1:0
CAP[1:0] 00
R/W
R/W
Set output on compare
Clear output on compare
Toggle output on compare
Set output on compare-up, clear on compare down in up-down mode. Otherwise set output on
compare, clear on 0.
Clear output on compare-up, set on compare down in up-down mode. Otherwise clear output
on compare, set on 0.
Clear when equal T1CC0, set when equal T1CC4
Set when equal T1CC0, clear when equal T1CC4
Initialize output pin. CMP[2:0] is not changed.
101:
110:
111:
Mode. Select Timer 1 channel 4 capture or compare mode
0:
Capture mode
1:
Compare mode
Channel 4 capture-mode select
00:
No capture
01:
Capture on rising edge
10:
Capture on falling edge
11:
Capture on all edges
T1CC4H (0x62AF) – Timer 1 Channel 4 Capture/Compare Value, High
Bit
7:0
Name
Reset
T1CC4[15:8] 0x00
R/W
R/W
Description
Timer 1 channel 4 capture/compare value high-order byte. Writing to this register when
T1CCTL4.MODE = 1 (compare mode) causes the T1CC4[15:0] update to the written value to be
delayed until T1CNT = 0x0000.
T1CC4L (0x62AE) – Timer 1 Channel 4 Capture/Compare Value, Low
Bit
7:0
Name
T1CC4[7:0]
Reset
0x00
R/W
R/W
Description
Timer 1 channel 4 capture/compare value low-order byte. Data written to this register is stored in
a buffer but not written to T1CC4[7:0] until, and at the same time as, a later write to T1CC4H
takes effect.
IRCTL (0x6281) – Timer 1 IR Generation Control
Bit
7:1
0
Name
–
IRGEN
Reset
R/W
0000 000 R/W
0
R/W
Description
Reserved
When this bit is set, a connection between Timer 3 channel 1 and Timer 1 tick input is made so
that the timers can be used to generate modulated IR codes (see also Section 9.9).
9.13 Accessing Timer 1 Registers as Array
The Timer 1 capture/compare channel registers can be accessed as a contiguous region in the XDATA
memory space. This facilitates accessing the registers as a simple indexed structure. The five
capture/compare control registers are mapped to 0x62A0–0x62A4. The 16-bit capture/compare values are
mapped to 0x62A6–0x62AF; 0x62A5 is unused.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 1 (16-Bit Timer)
119
Chapter 10
SWRU191D – April 2009 – Revised March 2013
Timer 3 and Timer 4 (8-Bit Timers)
Timer 3 and Timer 4 are two 8-bit timers. Each timer has two independent capture/compare channels,
each using one I/O pin per channel.
Features of Timer 3 and Timer 4 are as follows:
• Two capture/compare channels
• Set, clear, or toggle output compare
• Clock prescaler for divide by 1, 2, 4, 8, 16, 32, 64, 128
• Interrupt request generated on each capture/compare and terminal-count event
• DMA trigger function
Topic
10.1
10.2
10.3
10.4
10.5
10.6
10.7
10.8
120
...........................................................................................................................
8-Bit Timer Counter ..........................................................................................
Timer 3/Timer 4 Mode Control ...........................................................................
Channel Mode Control ......................................................................................
Input Capture Mode .........................................................................................
Output Compare Mode .....................................................................................
Timer 3 and Timer 4 Interrupts ..........................................................................
Timer 3 and Timer 4 DMA Triggers ....................................................................
Timer 3 and Timer 4 Registers ...........................................................................
Timer 3 and Timer 4 (8-Bit Timers)
Page
121
121
121
122
122
122
123
123
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
8-Bit Timer Counter
www.ti.com
10.1 8-Bit Timer Counter
All timer functions are based on the main 8-bit counter found in Timer 3 and Timer 4. The counter
increments or decrements at each active clock edge. The period of the active clock edges, as defined by
the register bits CLKCONCMD.TICKSPD[2:0], is further multiplied (the frequency is divided) by the
prescaler value set by TxCTL.DIV[2:0] (where x refers to the timer number, 3 or 4). The counter
operates as either a free-running counter, a down counter, a modulo counter, or an up/down counter.
It is possible to read the 8-bit counter value through the SFR register TxCNT, where x refers to the timer
number, 3 or 4.
The possibility to clear and halt the counter is given with TxCTL control register settings. The counter is
started when a 1 is written to TxCTL.START. If a 0 is written to TxCTL.START, the counter halts at its
present value.
10.2 Timer 3/Timer 4 Mode Control
In general, the control register TxCTL is used to control the timer operation.
10.2.1 Free-Running Mode
In the free-running mode of operation, the counter starts from 0x00 and increments at each active clock
edge. When the counter reaches 0xFF, the counter is loaded with 0x00 and continues incrementing its
value. When the terminal count value 0xFF is reached (i.e., an overflow occurs), the interrupt flag
TIMIF.TxOVFIF is set. An interrupt request is generated if enabled, see Section 10.6 for details. The
free-running mode can be used to generate independent time intervals and output-signal frequencies.
10.2.2 Down Mode
In the down mode, after the timer has been started, the counter is loaded with the contents in TxCC0. The
counter then counts down to 0x00. The interrupt flag TIMIF.TxOVFIF is set when 0x00 is reached. An
interrupt request is generated if enabled, see Section 10.6 for details. The timer down mode can generally
be used in applications where an event timeout interval is required.
10.2.3 Modulo Mode
When the timer operates in modulo mode, the 8-bit counter starts at 0x00 and increments at each active
clock edge. After the count has reached the period value held in register TxCC0, the counter is reset to
0x00 and continues to increment. If the timer started with a value above TxCC0, the interrupt flag
TIMIF.TxOVFIF is set when the terminal value (0xFF) is reached, after which the counter wraps to 0x00.
An interrupt request is generated if enabled, see Section 10.6 for details. If a periodic interrupt is wanted
at the period value, this can be obtained by enabling an output compare interrupt on channel 0, as
explained in Section 10.5. The modulo mode can be used for applications where a period other than 0xFF
is required.
10.2.4 Up/Down Mode
In the up/down timer mode, the counter repeatedly starts from 0x00 and counts up until the value held in
TxCC0 is reached, and then the counter counts down until 0x00 is reached. This timer mode is used when
symmetrical output pulses are required with a period other than 0xFF, allowing implementation of centeraligned PWM output applications. The interrupt flag TIMIF.TxOVFIF is set when the counter value
reaches 0x00 in the up/down mode. An interrupt request is generated if enabled, see Section 10.6 for
details.
Clearing the counter by writing to TxCTL.CLR also resets the count direction to the count-up-from-0x00
mode.
10.3 Channel Mode Control
The channel modes for each channel, 0 and 1, are set by the control and status registers TxCCTLn,
where n is the channel number, 0 or 1. The settings include capture and compare modes.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 3 and Timer 4 (8-Bit Timers)
121
Input Capture Mode
www.ti.com
10.4 Input Capture Mode
When a channel is configured as an input capture channel, the I/O pin associated with that channel is
configured as an input. After the timer has been started, a rising edge, falling edge, or any edge on the
input pin triggers a capture of the 8-bit counter contents into the associated capture register. Thus, the
timer is able to capture the time when an external event takes place.
NOTE: Before an I/O pin can be used by the timer, the required I/O pin must be configured as a
Timer 3/Timer 4 peripheral pin.
The channel input pin is synchronized to the internal system clock. Thus, pulses on the input pin must
have a minimum duration greater than the system clock period.
The content of the 8-bit capture register for channel n is read out from register T3CCn/T4CCn.
When the capture takes place, the interrupt flag for the channel, TIMIF.TxCHnIF (x is 3 or 4, n is the
channel number), is set. An interrupt request is generated if enabled, see Section 10.6 for details.
10.5 Output Compare Mode
In output-compare mode, the I/O pin associated with a channel must be set to an output. After the timer
has been started, the content of the counter is compared with the contents of channel compare register
TxCC0n. If the compare register equals the counter contents, the output pin is set, reset, or toggled
according to the compare output mode setting of TxCCTL.CMP1:0. Note that all edges on output pins are
glitch-free when operating in a given compare output mode.
For simple PWM use, output compare modes 4 and 5 are preferred.
Writing to compare register TxCC0 or TxCC1 does not take effect on the output compare value until the
counter value is 0x00.
When the capture takes place, the interrupt flag for the channel, TIMIF.TxCHnIF (x is 3 or 4, n is the
channel number), is set. An interrupt request is generated if enabled, see Section 10.6 for details.
A compare output pin is initialized to the value listed in Table 9-1 when:
• a 1 is written to TxCNTR.CLR (All Timer x channels)
• 0x7 is written to TxCCTLn.CMP (Timer x, channel n)
Table 10-1. Initial Compare Output Values (Compare Mode)
Compare Mode (TxCCTLn.CMP)
Initial Compare Output
Set output on compare (000)
0
Clear output on compare (001)
1
Toggle output on compare (010)
0
Set output on compare-up, clear on compare-down in up-down mode (011)
0
In other modes than up-down mode, set output on compare, clear on 0 (011)
0
Clear output on compare-up, set on compare-down in up-down mode (100)
1
In other modes than up-down mode, clear output on compare, set on 0 (100)
1
Set output on compare, clear on 0xFF (101)
0
Clear output on compare, set on 0x00 (110)
1
10.6 Timer 3 and Timer 4 Interrupts
One interrupt vector is assigned to each of the timers. These are T3 and T4. An interrupt request is
generated when one of the following timer events occurs:
• Counter reaches terminal count value.
• Compare event
• Capture event
122
Timer 3 and Timer 4 (8-Bit Timers)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 3 and Timer 4 DMA Triggers
www.ti.com
The SFR register TIMIF contains all interrupt flags for Timer 3 and Timer 4. The register bits
TIMIF.TxOVFIF and TIMIF.TxCHnIF contain the source interrupt flags for the two terminal-count value
events and the four channel-compare events, respectively. A source interrupt flag is set when the
corresponding event occurs, regardless of interrupt mask bits. The CPU interrupt flag IRCON.T3IF or
IRCON.T4IF is set when one of the events occurs if the corresponding interrupt mask bit is equal to 1.
The interrupt mask bits are TxCCTLn.IM for the four channels and TxCTL.OVFIM for the overflow events.
The CPU interrupt flag IRCON.T3IF or IRCON.T4IF is also set when a Timer 3 or Timer 4 source
interrupt flag is being cleared and one or more other source interrupt flags for the same timer are still set
while the corresponding interrupt mask bit is set. An interrupt request is generated when IRCON.TxIF
goes from 0 to 1 if IEN1.TxIEN and IEN0.EA are both equal to 1 (x is 3 or 4).
10.7 Timer 3 and Timer 4 DMA Triggers
Two DMA triggers are associated with Timer 3, and two DMA triggers are associated with Timer 4.
• T3_CH0: Timer 3 channel 0 capture/compare
• T3_CH1: Timer 3 channel 1 capture/compare
• T4_CH0: Timer 4 channel 0 capture/compare
• T4_CH0: Timer 4 channel 1 capture/compare
10.8 Timer 3 and Timer 4 Registers
T3CNT (0xCA) – Timer 3 Counter
Bit
7:0
Name
Reset
CNT[7:0] 0x00
R/W
R
Description
Timer count byte. Contains the current value of the 8-bit counter
T3CTL (0xCB) – Timer 3 Control
Bit
7:5
Name
DIV[2:0]
Reset
000
R/W
R/W
4
3
START
OVFIM
0
1
R/W
R/W
2
CLR
0
R0/W1
1:0
MODE[1:0]
00
R/W
Description
Prescaler divider value. Generates the active clock edge used to clock the timer from
CLKCONCMD.TICKSPD as follows:
000: Tick frequency/1
001: Tick frequency/2
010: Tick frequency/4
011: Tick frequency/8
100: Tick frequency16
101: Tick frequency /32
110: Tick frequency/64
111: Tick frequency/128
Start timer. Normal operation when set, suspended when cleared
Overflow interrupt mask
0:
Interrupt is disabled.
1:
Interrupt is enabled.
Clear counter. Writing a 1 to CLR resets the counter to 0x00 and initializes all output pins of
associated channels. Always read as 0.
Timer 3 mode. Select the mode as follows:
00:
Free-running, repeatedly count from 0x00 to 0xFF
01:
Down, count from T3CC0 to 0x00
10:
Modulo, repeatedly count from 0x00 to T3CC0
11:
Up/down, repeatedly count from 0x00 to T3CC0 and down to 0x00
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 3 and Timer 4 (8-Bit Timers)
123
Timer 3 and Timer 4 Registers
www.ti.com
T3CCTL0 (0xCC) – Timer 3 Channel 0 Capture/Compare Control
Bit
7
6
Name
–
IM
Reset
0
1
R/W
R0
R/W
5:3
CMP[2:0]
000
R/W
2
MODE
0
R/W
1:0
CAP[1:0]
00
R/W
Description
Reserved
Channel 0 interrupt mask
0:
Interrupt is disabled.
1:
Interrupt is enabled.
Channel 0 compare output mode select. Specified action occurs on output when timer value
equals compare value in T3CC0
000: Set output on compare
001: Clear output on compare
010: Toggle output on compare
011: Set output on compare-up, clear on 0
100: Clear output on compare-up, set on 0
101: Set output on compare, clear on 0xFF
110: Clear output on compare, set on 0x00
111: Initialize output pin. CMP[2:0] is not changed.
Mode. Select Timer 3 channel 0 mode
0:
Capture mode
1:
Compare mode
Capture mode select
00:
No capture
01:
Capture on rising edge
10:
Capture on falling edge
11:
Capture on both edges
T3CC0 (0xCD) – Timer 3 Channel 0 Capture/Compare Value
Bit
7:0
124
Name
Reset
VAL[7:0] 0x00
R/W
R/W
Description
Timer capture/compare value channel 0. Writing to this register when T3CCTL0.MODE=1 (compare
mode) causes the T3CC0.VAL[7:0] update to the written value to be delayed until
T3CNT.CNT[7:0]=0x00.
Timer 3 and Timer 4 (8-Bit Timers)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 3 and Timer 4 Registers
www.ti.com
T3CCTL1 (0xCE) – Timer 3 Channel 1 Capture/Compare Control
Bit
7
6
Name
–
IM
Reset
0
1
R/W
R0
R/W
5:3
CMP[2:0]
000
R/W
Description
Reserved
Channel 1 interrupt mask
0:
Interrupt is disabled.
1:
Interrupt is enabled.
Channel 1 compare output-mode select. Specified action on output when timer value equals
compare value in T3CC1
000:
001:
010:
011:
2
MODE
0
R/W
1:0
CAP[1:0]
00
R/W
Set output on compare
Clear output on compare
Toggle output on compare
Set on compare-up, clear on compare-down in up-down mode. Otherwise, set output on
compare, clear on 0.
100: Clear output on compare-up, set on compare-down in up-down mode. Otherwise clear
output on compare, set on 0.
101: Set output on compare, clear on 0xFF
110: Clear output on compare, set on 0x00
111: Initialize output pin. CMP[2:0] is not changed
Mode. Select Timer 3 channel 1 mode
0:
Capture mode
1:
Compare mode
Capture mode select
00:
No capture
01:
Capture on rising edge
10:
Capture on falling edge
11:
Capture on both edges
T3CC1 (0xCF) – Timer 3 Channel 1 Capture/Compare Value
Bit
7:0
Name
Reset
VAL[7:0] 0x00
R/W
R/W
Description
Timer 3 capture/compare value, channel 1. Writing to this register when T3CCTL1.MODE = 1
(compare mode) causes the T3CC1.VAL[7:0] update to the written value to be delayed until
T3CNT.CNT[7:0] = 0x00.
T4CNT (0xEA) – Timer 4 Counter
Bit
7:0
Name
Reset
CNT[7:0] 0x00
R/W
R
Description
Timer count byte. Contains the current value of the 8-bit counter
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 3 and Timer 4 (8-Bit Timers)
125
Timer 3 and Timer 4 Registers
www.ti.com
T4CTL (0xEB) – Timer 4 Control
Bit
7:5
Name
DIV[2:0]
Reset
000
R/W
R/W
4
3
2
START
OVFIM
CLR
1
0
R/W
R/W
R0/W1
1:0
MODE[1:0]
0
R/W
Description
Prescaler divider value. Generates the active clock edge used to clock the timer from
CLKCONCMD.TICKSPD as follows:
000:
Tick frequency/1
001:
Tick frequency/2
010:
Tick frequency/4
011:
Tick frequency/8
100:
Tick frequency/16
101:
Tick frequency/32
110:
Tick frequency/64
111:
Tick frequency/128
Start timer. Normal operation when set, suspended when cleared
Overflow interrupt mask
Clear counter. Writing a 1 to CLR resets the counter to 0x00 and initialize all output pins of
associated channels. Always read as 0.
Timer 4 mode. Select the mode as follows:
00:
Free running, repeatedly count from 0x00 to 0xFF
01:
Down, count from T4CC0 to 0x00
10:
Modulo, repeatedly count from 0x00 to T4CC0
11:
Up/down, repeatedly count from 0x00 to T4CC0 and down to 0x00
T4CCTL0 (0xEC) – Timer 4 Channel 0 Capture/Compare Control
Bit
7
6
5:3
Name
–
IM
CMP[2:0]
Reset
0
1
000
R/W
R0
R/W
R/W
2
MODE
0
R/W
1:0
CAP[1:0]
00
R/W
Description
Reserved
Channel 0 interrupt mask
Channel 0 compare output-mode select. Specified action occurs on output when timer value
equals compare value in T4CC0.
000: Set output on compare
001: Clear output on compare
010: Toggle output on compare
011: Set output on compare-up, clear on 0
100: Clear output on compare-up, set on 0
101: Set output on compare, clear on 0xFF
110: Clear output on compare, set on 0x00
111: Initialize output pin. CMP[2:0] is not changed
Mode. Select Timer 4 channel 0 mode
0:
Capture mode
1:
Compare mode
Capture mode select. 00 – No capture, 01 – Capture on rising edge, 10 – Capture on falling
edge, 11 – Capture on both edges
T4CC0 (0xED) – Timer 4 Channel 0 Capture/Compare Value
Bit
7:0
126
Name
Reset
VAL[7:0] 0x00
R/W
R/W
Description
Timer 4 capture/compare value, channel 0. Writing to this register when T4CCTL0.MODE = 1
(compare mode) causes the T4CC0.VAL[7:0] update to the written value to be delayed until
T4CNT.CNT[7:0] = 0x00.
Timer 3 and Timer 4 (8-Bit Timers)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 3 and Timer 4 Registers
www.ti.com
T4CCTL1 (0xEE) – Timer 4 Channel 1 Capture/Compare Control
Bit
7
6
5:3
Name
–
IM
CMP[2:0]
Reset
0
1
000
R/W
R0
R/W
R/W
Description
Reserved
Channel 1 interrupt mask
Channel 1 compare output-mode select. Specified action on output when timer value equals
compare value in T4CC1
000:
001:
010:
011:
2
MODE
0
R/W
1:0
CAP[1:0]
00
R/W
Set output on compare
Clear output on compare
Toggle output on compare
Set on compare-up, clear on compare-down in up-down mode. Otherwise, set output on
compare, clear on 0.
100: Clear output on compare-up, set on compare-down in up-down mode. Otherwise clear
output on compare, set on 0.
101: Set output on compare, clear on 0xFF
110: Clear output on compare, set on 0x00
111: Initialize output pin. CMP[2:0] is not changed.
Mode. Select Timer 4 channel 1 mode
0:
Capture mode
1:
Compare mode
Capture mode select. 00 – No Capture, 01 – Capture on rising edge, 10 – Capture on falling
edge, 11 – Capture on both edges
T4CC1 (0xEF) – Timer 4 Channel 1 Capture/Compare Value
Bit
7:0
Name
Reset
VAL[7:0] 0x00
R/W
R/W
Description
Timer capture/compare value, channel 1. Writing to this register when T4CCTL1.MODE = 1
(compare mode) causes the T4CC1.VAL[7:0] update to the written value to be delayed until
T4CNT.CNT[7:0] = 0x00.
TIMIF (0xD8) – Timer 1/3/4 Interrupt Mask/Flag
Bit
7
6
5
Name
–
T1OVFIM
T4CH1IF
Reset
0
1
0
R/W
R0
R/W
R/W0
4
T4CH0IF
0
R/W0
3
T4OVFIF
0
R/W0
2
T3CH1IF
0
R/W0
1
T3CH0IF
0
R/W0
0
T3OVFIF
0
R/W0
Description
Reserved
Timer 1 overflow interrupt mask
Timer 4 channel 1 interrupt flag
0:
No interrupt is pending.
1:
Interrupt is pending.
Timer 4 channel 0 interrupt flag
0:
No interrupt is pending.
1:
Interrupt is pending.
Timer 4 overflow interrupt flag
0:
No interrupt is pending.
1:
Interrupt is pending.
Timer 3 channel 1 interrupt flag
0:
No interrupt is pending.
1:
Interrupt is pending.
Timer 3 channel 0 interrupt flag
0:
No interrupt is pending.
1:
Interrupt is pending.
Timer 3 overflow interrupt flag
0:
No interrupt is pending.
1:
Interrupt is pending.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 3 and Timer 4 (8-Bit Timers)
127
Chapter 11
SWRU191D – April 2009 – Revised March 2013
Sleep Timer
The Sleep Timer is used to set the period during which the system enters and exits low-power modes
PM1 and PM2. The Sleep Timer is also used to maintain timing in Timer 2 when entering power mode
PM1 or PM2.
The main features of the Sleep Timer are the following:
• 24-bit timer up-counter operating at 32-kHz clock rate
• 24-bit compare with interrupt and DMA trigger
• 24-bit capture
Topic
11.1
11.2
11.3
11.4
128
...........................................................................................................................
General ...........................................................................................................
Timer Compare ................................................................................................
Timer Capture .................................................................................................
Sleep Timer Registers ......................................................................................
Sleep Timer
Page
129
129
129
130
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
General
www.ti.com
11.1 General
The Sleep Timer is a 24-bit timer running on the 32-kHz clock (either RCOSC or XOSC). The timer starts
running immediately after a reset and continues to run uninterrupted.
The current value of the timer can be read from SFR registers ST2:ST1:ST0. When ST0 is read, the
current value of the 24-bit counter is latched. Thus, the ST0 register must be read before ST1 and ST2 to
read a correct Sleep Timer count value.
The Sleep Timer is running when operating in all power modes except PM3. The value of the Sleep Timer
is not preserved in PM3. When returning from PM1 or PM2 (where the system clock is shut down), the
Sleep Timer value in ST2:ST1:ST0 is not up-to-date until a positive edge on the 32-kHz clock has been
detected after the system clock restarted. To ensure an updated value is read, wait for a positive transition
on the 32-kHz clock by polling the SLEEPSTA.CLK32K bit, before reading the Sleep Timer value.
Note that if supply voltage drops below 2 V while in PM2, the sleep interval might be affected.
11.2 Timer Compare
A timer compare event occurs when the timer value is equal to the 24-bit compare value and there is a
positive edge on the 32-kHz clock. The compare value is set by writing to registers ST2:ST1:ST0. Writing
to ST0 while STLOAD.LDRDY is 1 initiates loading of the new compare value, i.e., the most-recent values
written to the ST2, ST1, and ST0 registers. This means that when writing a compare value, ST2 and ST1
must be written before ST0. STLOAD.LDRDY is 0 during the load, and software must not start a new load
until STLOAD.LDRDY has flipped back to 1.
When setting a new compare value, the value should be at least 5 more than the current sleep timer
value. Otherwise, the timer compare event may be lost.
The interrupt enable bit for the ST interrupt is IEN0.STIE, and the interrupt flag is IRCON.STIF. When a
timer compare event occurs, the interrupt flag IRCON.STIF is asserted.
In PM1 and PM2, the Sleep Timer compare event may be used to wake up the device and return to active
operation in active mode. The default value of the compare value after reset is 0xFF FFFF.
For all devices except the CC2540/41, the Sleep Timer compare event can also be used as a DMA trigger
(DMA trigger 11 in Table 8-1).
Note that if supply voltage drops below 2 V while in PM2, the sleep interval might be affected.
11.3 Timer Capture
The timer capture occurs when the interrupt flag for a selected I/O pin is set and this event has been
dectected by the 32-kHz clock. Sleep Timer capture is enabled by setting STCC.PORT[1:0] and
STCC.PIN[2:0] to the I/O pin that is to be used to trigger the capture. When STCS.VALID goes high,
the capture value in STCV2:STCV1:STCV0 can be read. The captured value is one more than the value
at the instant for the event on the I/O pin. Software should therefore subtract one from the captured value
if abolute timing is required. To enable a new capture, follow these steps:
1. Clear STCS.VALID.
2. Wait until SLEEPSTA.CLK32K is low.
3. Wait until SLEEPSTA.CLK32K is high.
4. Clear the pin interrupt flag in the P0IFG/P1IFG/P2IFG register.
This sequence, using the rising edge on P0.0 as an example, is shown in Figure 11-1. Failure to follow the
procedure may cause the capture functionality to stop working until a chip reset.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Sleep Timer
129
Sleep Timer Registers
www.ti.com
Clear P0IFG[0] After
Having Detected a
Rising Edge On
SLEEPSTA.CLK32K
Read STCV[23:0],
Then Clear
STCS.VALID
P0_0
P0IFG[0]
STCS.VALID
SLEEPSTA.CLK32K
STCV[23:0]
Timer Value is Captured
T0412-01
Figure 11-1. Sleep Timer Capture (Example Using Rising Edge on P0_0)
It is not possible to switch the input-capture pin while capture is enabled. Capture must be disabled before
a new input-capture pin can be selected. To disable capture, follow these steps (note that interrupts will be
disabled for up to half of a 32-kHz cycle, or 15.26 µs):
1. Disable interrupts.
2. Wait until SLEEPSTA.CLK32K is high.
3. Set STCC.PORT[1:0] to 3. This disables capture.
11.4 Sleep Timer Registers
The registers used by the Sleep Timer are:
• ST2 – Sleep Timer 2
• ST1 – Sleep Timer 1
• ST0 – Sleep Timer 0
• STLOAD – Sleep Timer load status
• STCC – Sleep Timer capture control
• STCS – Sleep Timer capture status
• STCV0 – Sleep Timer capture value byte 0
• STCV1 – Sleep Timer capture value byte 1
• STCV2 – Sleep Timer capture value byte 2
ST2 (0x97) – Sleep Timer 2
Bit
7:0
130
Name
Reset
ST2[7:0] 0x00
Sleep Timer
R/W
R/W
Description
Sleep Timer count/compare value. When read, this register returns the high bits [23:16] of the Sleep
Timer count. When writing, this register sets the high bits [23:16] of the compare value. The value
read is latched at the time of reading register ST0. The value written is latched when ST0 is written.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Sleep Timer Registers
www.ti.com
ST1 (0x96) – Sleep Timer 1
Bit
7:0
Name
Reset
ST1[7:0] 0x00
R/W
R/W
Description
Sleep Timer count/compare value. When read, this register returns the middle bits [15:8] of the
Sleep Timer count. When writing, this register sets the middle bits [15:8] of the compare value. The
value read is latched at the time of reading register ST0. The value written is latched when ST0 is
written.
R/W
R/W
Description
Sleep Timer count/compare value. When read, this register returns the low bits [7:0] of the Sleep
Timer count. When writing, this register sets the low bits [7:0] of the compare value. Writes to this
register are ignored unless STLOAD.LDRDY is 1.
ST0 (0x95) – Sleep Timer 0
Bit
7:0
Name
Reset
ST0[7:0] 0x00
STLOAD (0xAD) – Sleep Timer Load Status
Bit
7:1
Name
–
0
LDRDY
Reset
R/W
0000 00 R0
0
1
R
Description
Reserved
Load ready. This bit is 0 while the Sleep Timer loads the 24-bit compare value and 1 when the
Sleep Timer is ready to start loading a new compare value.
STCC (0x62B0) – Sleep Timer Capture Control
Bit
7:5
4:3
2:0
Name
–
PORT[1:0
]
PIN[2:0]
Reset
000
11
R/W
R0
R
111
Description
Reserved
Port select. Valid settings are 0–2. Capture is disabled when set to 3, i.e. an invalid setting is
selected.
Pin select. Valid settings are 0–7 when PORT[1:0] is 0 or 1, 0–5 when PORT[1:0] is 2. Capture is
disabled when an invalid setting is selected.
STCS (0x62B1) – Sleep Timer Capture Status
Bit
7:1
Name
–
0
VALID
Reset
R/W
0000 00 R0
0
0
R/W0
Description
Reserved
Capture valid flag. Set to 1 when capture value in STCV has been updated. Clear explicitly to allow
new capture.
STCV0 (0x62B2) – Sleep Timer Capture Value Byte 0
Bit
7:0
Name
Reset
STCV[7:0 0x00
]
R/W
R
Description
Bits [7:0] of Sleep Timer capture value
STCV1 (0x62B3) – Sleep Timer Capture Value Byte 1
Bit
7:0
Name
Reset
STCV[15: 0x00
8]
R/W
R
Description
Bits [15:8] of Sleep Timer capture value
STCV2 (0x62B4) – Sleep Timer Capture Value Byte 2
Bit
7:0
Name
Reset
STCV[23: 0x00
16]
R/W
R
Description
Bits [23:16] of Sleep Timer capture value
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Sleep Timer
131
Chapter 12
SWRU191D – April 2009 – Revised March 2013
ADC
The ADC (in the CC2530/CC2531/CC2540/CC2541) supports 14-bit analog-to-digital conversion with up
to 12 effective number of bits (ENOB). It includes an analog multiplexer with up to eight individually
configurable channels and a reference voltage generator. Conversion results can be written to memory
through DMA. Several modes of operation are available.
Topic
12.1
12.2
132
ADC
...........................................................................................................................
Page
ADC Introduction ............................................................................................. 133
ADC Operation ................................................................................................ 133
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
ADC Introduction
www.ti.com
12.1 ADC Introduction
The ADC supports up to 14-bit analog-to-digital conversion with up to 12 bits ENOB (Effective Number Of
Bits). It includes an analog multiplexer with up to eight individually configurable channels and a reference
voltage generator. Conversion results can be written to memory through DMA. Several modes of operation
are available.
The main features of the ADC are as follows:
• Selectable decimation rates which also set the effective resolution (7 to 12 bits).
• Eight individual input channels, single-ended or differential
• Reference voltage selectable as internal, external single-ended, external differential, or AVDD5
• Interrupt request generation
• DMA triggers at end of conversions
• Temperature sensor input
• Battery measurement capability
...
AIN0
AIN7
VDD/3
Input
Mux
Sigma-Delta
Modulator
Decimation
Filter
TMP_ SENSOR
Internal Reference Voltage
AIN7
Ref
Mux
AVDD
Clock Generation
and
Control
AIN6–AIN7
B0304-01
Figure 12-1. ADC Block Diagram
12.2 ADC Operation
This section describes the general setup and operation of the ADC and describes the use of the ADC
control and status registers accessed by the CPU.
12.2.1 ADC Inputs
The signals on the port-0 pins can be used as ADC inputs. In the following, these port pins are referred to
as the AIN0–AIN7 pins. The input pins AIN0–AIN7 are connected to the ADC.
It is possible to configure the inputs as single-ended or differential inputs. In the case where differential
inputs are selected, the differential inputs consist of the input pairs AIN0–AIN1, AIN2–AIN3, AIN4–AIN5,
and AIN6–AIN7. Note that no negative supply can be applied to these pins, nor a supply higher than VDD
(unregulated power). It is the difference between the pins of each pair that is converted in differential
mode.
In addition to the input pins AIN0–AIN7, the output of an on-chip temperature sensor can be selected as
an input to the ADC for temperature measurements. In order to do so, the registers TR0.ADCTM and
ATEST.ATESTCTRL must be set as described in the register descriptions in Section 12.2.10 and
Section 23.15.3 (CC253x) or Section 24.1 (CC2540), or Section 25.12 (CC2541).
It is also possible to select a voltage corresponding to AVDD5 divided by 3 as an ADC input. This input
allows the implementation of, e.g., a battery monitor in applications where this feature is required. Note
that the reference in this case must not be dependent on the battery voltage; for instance, the AVDD5
voltage must not be used as a reference.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
ADC
133
ADC Operation
www.ti.com
The single-ended inputs AIN0 through AIN7 are represented by channel numbers 0 to 7. Channel
numbers 8 through 11 represent the differential inputs consisting of AIN0–AIN1, AIN2–AIN3, AIN4–AIN5,
and AIN6–AIN7. Channel numbers 12 through 15 represent GND (12), temperature sensor (14), and
AVDD5/3 (15), with channel 13 being reserved. These values are used in the ADCCON2.SCH and
ADCCON3.SCH fields.
The ADC input is a switched capacitance stage which draws current during the conversion. As an
example, the equivalent input impedance of a typical device was found to be 176 kΩ when used with an
input voltage of 3 V, a 512× decimation rate, and the internal reference.
To enable the temperature sensor as an input to the ADC, the TR0.ADCTM bit must be set to 1 before
setting the ATEST.ATESTCTRL bit to 1. When disabling the temperature sensor as an input, the
ATEST.ATESTCTRL bit must be set to 0 before clearing the TR0.ADCTM bit. The TR0 register does not
have any retention in PM2 or PM3, so ATEST and TR0 must be cleared in the correct manner before
entering these power modes.
12.2.2 ADC Conversion Sequences
The ADC can perform a sequence of conversions and move the results to memory (through DMA) without
any interaction from the CPU.
The conversion sequence can be influenced with the APCFG register (see Section 7.6.6), in that the eight
analog inputs to the ADC come from I/O pins that are not necessarily programmed to be analog inputs. If
a channel should normally be part of a sequence, but the corresponding analog input is disabled in the
APCFG register, then that channel is skipped. When using differential inputs, both pins in a differential pair
must set as analog input pins in the APCFG register.
The ADCCON2.SCH register bits are used to define an ADC conversion sequence from the ADC inputs. If
ADCCON2.SCH is set to a value less than 8, the conversion sequence contains a conversion from each
channel from 0 up to and including the channel number programmed in ADCCON2.SCH. When
ADCCON2.SCH is set to a value between 8 and 12, the sequence consists of differential inputs, starting at
channel 8 and ending at the programmed channel. For ADCCON2.SCH greater than or equal to 12, the
sequence consists of the selected channel only.
12.2.3 Single ADC Conversion
In addition to this sequence of conversions, the ADC can be programmed to perform a single conversion
from any channel. Such a conversion is triggered by writing to the ADCCON3 register. The conversion
starts immediately unless a conversion sequence is already ongoing, in which case the single conversion
is performed as soon as that sequence is finished.
12.2.4 ADC Operating Modes
This section describes the operating modes and initialization of conversions.
The ADC has three control registers: ADCCON1, ADCCON2, and ADCCON3. These registers are used to
configure the ADC and to report status.
The ADCCON1.EOC bit is a status bit that is set high when a conversion ends and cleared when ADCH is
read.
The ADCCON1.ST bit is used to start a sequence of conversions. A sequence starts when this bit is set
high, ADCCON1.STSEL is 11, and no conversion is currently running. When the sequence is completed,
this bit is automatically cleared.
The ADCCON1.STSEL bits select the event that starts a new sequence of conversions. The options which
can be selected are rising edge on external pin P2.0, end of previous sequence, a Timer 1 channel 0
compare event, or ADCCON1.ST is 1.
The ADCCON2 register controls how the sequence of conversions is performed.
ADCCON2.SREF is used to select the reference voltage. The reference voltage should only be changed
when no conversion is running.
134
ADC
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
ADC Operation
www.ti.com
The ADCCON2.SDIV bits select the decimation rate, thereby also the resolution and time required to
complete a conversion, and hence the sample rate. The decimation rate should only be changed when no
conversion is running.
The last channel of a sequence is selected with the ADCCON2.SCH bits as described previoiusly.
The ADCCON3 register controls the channel number, reference voltage, and decimation rate for a single
conversion. The single conversion takes place immediately after the ADCCON3 register is written to, or if a
conversion sequence is ongoing, immediately after the sequence has ended. The coding of the register
bits is exactly as for ADCCON2.
12.2.5 ADC Conversion Results
The digital conversion result is represented in 2s-complement form. For single-ended configurations, the
result can be expected to be positive. This is because the result is the difference between the input signal
and ground, which is always positively signed (Vconv = Vinp – Vinn, where Vinn = 0 V). The maximum
value is reached when the input signal is equal to VREF, the selected voltage reference. For differential
configurations, the difference between two pins is converted, and this difference can be negatively signed.
For example, with a decimation rate of 512 using only the 12 MSBs of the digital conversion result
register, the maximum value of 2047 is reached when the analog input (Vconv) is equal to VREF, and
minimum value of –2048 is reached when the analog input is equal to –VREF .
The digital conversion result is available in ADCH and ADCL when ADCCON1.EOC is set to 1. Note that the
conversion result always resides in the MSB section of the combined ADCH and ADCL registers.
When the ADCCON2.SCH bits are read, they indicate the channel on which conversion is ongoing. The
results in ADCL and ADCH normally apply to the previous conversion. If the conversion sequence has
ended, ADCCON2.SCH has a value of one more than the last channel number, but if the channel number
last written to ADCCON2.SCH was 12 or more, the same value is read back.
12.2.6 ADC Reference Voltage
The positive reference voltage for analog-to-digital conversions is selectable as either an internally
generated voltage, the AVDD5 pin, an external voltage applied to the AIN7 input pin, or a differential
voltage applied to the AIN6–AIN7 inputs.
The accuracy of the conversion results depend on the stability and noise properties of the reference
voltage. Offset from the wanted voltage introduces a gain error in the ADC proportional to the ratio of the
wanted voltage and the actual voltage. Noise on the reference must be lower than quantization noise of
the ADC to ensure the specified SNR is achieved.
12.2.7 ADC Conversion Timing
The ADC should only be used with the 32-MHz XOSC, and no system clock division should be
implemented by the user. The actual ADC sampling frequency of 4 MHz is generated by fixed internal
division. The time required to perform a conversion depends on the selected decimation rate. In general,
the conversion time is given by:
Tconv = (decimation rate + 16) × 0.25 μs.
12.2.8 ADC Interrupts
The ADC generates an interrupt when a single conversion triggered by writing to ADCCON3 has completed.
No interrupt is generated when a conversion from the sequence is completed.
12.2.9 ADC DMA Triggers
The ADC generates a DMA trigger every time a conversion from the sequence has completed. When a
single conversion completes, no DMA trigger is generated.
There is one DMA trigger for each of the eight channels defined by the first eight possible settings for
ADCCON2.SCH. The DMA trigger is active when a new sample is ready from the conversion for the
channel. The DMA triggers are named ADC_CHsd in Table 8-1, where s is single-ended channel and d is
differential channel.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
ADC
135
ADC Operation
www.ti.com
In addition, one DMA trigger, ADC_CHALL, is active when new data is ready from any of the channels in
the ADC conversion sequence.
12.2.10 ADC Registers
This section describes the ADC registers.
ADCL (0xBA) – ADC Data, Low
Bit
7:2
1:0
Name
ADC[5:0]
–
Reset
0000 00
00
R/W
R
R0
Description
Least-significant part of ADC conversion result
Reserved. Always read as 0
R/W
R
Description
Most-significant part of ADC conversion result
ADCH (0xBB) – ADC Data, High
Bit
7:0
Name
ADC[13:6]
Reset
0x00
ADCCON1 (0xB4) – ADC Control 1
Bit
7
Name
EOC
Reset
0
R/W
R/H0
6
ST
0
R/W1/
H0
5:4
STSEL[1:0]
11
R/W
3:2
–
00
R/W
1:0
–
11
R/W
136
ADC
Description
End of conversion. Cleared when ADCH has been read. If a new conversion is completed
before the previous data has been read, the EOC bit remains high.
0:
Conversion not complete
1:
Conversion completed
Start conversion. Read as 1 until conversion has completed
0:
No conversion in progress
1:
Start a conversion sequence if ADCCON1.STSEL = 11 and no sequence is running.
Start select. Selects the event that starts a new conversion sequence
00:
External trigger on P2.0 pin
01:
Full speed. Do not wait for triggers
10:
Timer 1 channel 0 compare event
11:
ADCCON1.ST = 1
Controls the 16-bit random-number generator. See ADCCON1 (0xB4) – ADC Control 1
description in Section 14.3.
Reserved. Always set to 11
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
ADC Operation
www.ti.com
ADCCON2 (0xB5) – ADC Control 2
Bit
7:6
Name
SREF[1:0]
Reset
00
R/W
R/W
5:4
SDIV[1:0]
01
R/W
3:0
SCH[3:0]
0000
R/W
Description
Selects reference voltage used for the sequence of conversions
00:
Internal reference
01:
External reference on AIN7 pin
10:
AVDD5 pin
11:
External reference on AIN6–AIN7 differential input
Sets the decimation rate for channels included in the sequence of conversions. The decimation
rate also determines the resolution and time required to complete a conversion.
00:
64 decimation rate (7 bits ENOB setting)
01:
128 decimation rate (9 bits ENOB setting)
10:
256 decimation rate (10 bits ENOB setting)
11:
512 decimation rate (12 bits ENOB setting)
Sequence channel select. Selects the end of the sequence. A sequence can either be from AIN0
to AIN7 (SCH ≤ 7) or from differential input AIN0–AIN1 to AIN6–AIN7 (8 ≤ SCH ≤ 11). For other
settings, only one conversions is performed.
When read, these bits indicate the channel number on which a conversion is ongoing.
0000: AIN0
0001: AIN1
0010: AIN2
0011: AIN3
0100: AIN4
0101: AIN5
0110: AIN6
0111: AIN7
1000: AIN0–AIN1
1001: AIN2–AIN3
1010: AIN4–AIN5
1011: AIN6–AIN7
1100: GND
1101: Reserved
1110: Temperature sensor
1111: VDD/3
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
ADC
137
ADC Operation
www.ti.com
ADCCON3 (0xB6) – ADC Control 3
Bit
7:6
Name
EREF[1:0]
Reset
00
R/W
R/W
5:4
EDIV[1:0]
00
R/W
3:0
ECH[3:0]
0000
R/W
Description
Selects reference voltage used for the extra conversion
00:
Internal reference
01:
External reference on AIN7 pin
10:
AVDD5 pin
11:
External reference on AIN6–AIN7 differential input
Sets the decimation rate used for the extra conversion. The decimation rate also determines the
resolution and the time required to complete the conversion.
00:
64 decimation rate (7 bits ENOB)
01:
128 decimation rate (9 bits ENOB)
10:
256 decimation rate (10 bits ENOB)
11:
512 decimation rate (12 bits ENOB)
Single channel select. Selects the channel number of the single conversion that is triggered by
writing to ADCCON3.
0000:
0001:
0010:
0011:
0100:
0101:
0110:
0111:
1000:
1001:
1010:
1011:
1100:
1101:
1110:
1111:
AIN0
AIN1
AIN2
AIN3
AIN4
AIN5
AIN6
AIN7
AIN0–AIN1
AIN2–AIN3
AIN4–AIN5
AIN6–AIN7
GND
Reserved
Temperature sensor
VDD/3
TR0 (0x624B) – Test Register 0
Bit
7:1
0
Name
–
ADCTM
138
ADC
Reset
0000 000
0
R/W
R0
R/W
Description
Reserved. Write as 0.
Set to 1 to connect the temperature sensor to the SOC_ADC. See also ATEST register description
to enable the temperature sensor in Section 23.15.3 (CC253x) or Section 24.1 (CC2540) or
Chapter 25 (CC2541).
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 13
SWRU191D – April 2009 – Revised March 2013
Battery Monitor
The battery monitor (in the CC2533 only) enables simple voltage monitoring in the devices that do not
include an ADC. It is designed such that it is accurate in the voltage areas around 2 V, with lower
resolution at higher voltages. The registers BATTMON and MONMUX are used to access and control the
functionality of the battery monitor.
The battery monitor can also be used to do simple temperature monitoring by connecting it to the chip
internal temperature sensor instead of the supply voltage. The input is controlled using the MONMUX
register.
Topic
13.1
13.2
13.3
...........................................................................................................................
Page
Functionality and Usage of the Battery Monitor ................................................... 140
Using the Battery Monitor for Temperature Monitoring ........................................ 140
Battery Monitor Registers ................................................................................. 141
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Battery Monitor
139
Functionality and Usage of the Battery Monitor
www.ti.com
13.1 Functionality and Usage of the Battery Monitor
The battery monitor makes it possible to check whether the supply voltage (AVDD5) is above or below a
certain programmable level. Its usage is controlled by the BATTMON register in the following manner:
BATTMON_VOLTAGE is used to set the trigger point for the battery monitor. Note the fact that the step size
is different for different voltage ranges (see the register description in Section 13.3 for details). This is
done to achieve good accuracy in the voltage areas around 2 V, with lower resolution at higher voltages.
BATTMON_PD is used to enable/disable the battery monitor.
After enabling the battery monitor by setting BATTMON_PD = 0 and waiting for at least 2 µs, the value of
BATTMON_OUT indicates whether the voltage is above or below the trigger point (set by
BATTMON_VOLTAGE).
NOTE: One should turn the battery monitor off (BATTMON_PD = 1) after reading the measurement
BATTMON_OUT in order to save power, as the battery monitor consumes power when
enabled ( = 0).
Recommended usage of the battery monitor can be summarized in the following way:
1. Set BATTMON_VOLTAGE to the value to be monitored.
2. Enable the battery monitor by setting BATTMON_PD = 0.
3. Wait for at least 2 µs.
4. Read the BATTMON_OUT result to see whether the voltage level is above or below the value set in
BATTMON_VOLTAGE.
5. Disable the battery monitor (BATTMON_PD = 1) to avoid unnecessary current consumption.
13.2 Using the Battery Monitor for Temperature Monitoring
The battery monitor can also be used to do some simple temperature monitoring. When the battery
monitor is connected to the internal temperature sensor instead of the supply voltage AVDD5 (see the
description of MONMUX in Section 13.3), it can indicate whether the temperature is above or below a
certain level. This is done by comparing the voltage coming from the temperature sensor to the voltage
trigger point of the battery monitor. The controls for this measurement are the same as for the normal use
of the battery monitor (see the description of BATTMON in Section 13.3).
It is important to understand that due to the nature of the battery monitor (optimized for voltages arround 2
V) and the output voltage range of the temperature sensor, there are only about 8 temperature trigger
values in the temperature range of –40°C to 125°C (see Table 13-1). As a result, the battery monitor gives
only a rough indication of the temperature range, but this is useful for doing temperature compensation on
analog components in a system. See the device's data sheet (Appendix C) for performance characteristics
details.
Table 13-1. Values Showing How Different
Temperatures Relate to BATTMON_VOLTAGE for a
Typical Device
140
Battery Monitor
Temperature
BATTMON_VOLTAGE
-40°C
22
-26°C
21
-11°C
20
7°C
19
25°C
18
47°C
17
70°C
16
97°C
15
128°C
14
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Battery Monitor Registers
www.ti.com
The temperature sensor is inversely proportional to BATTMON_VOLTAGE. The temperature (in °C)
corresponding to a given BATTMON_VOLTAGE is given by:
A
Temp =
-B
BATTMON_VOLTAGE<4:0>
(1)
Assuming BATTMON_VOLTAGE < 27, and only valid for –40°C < Temp < 125°C, A and B for a typical
device are given in Table 13-2.
Table 13-2. Values for A and B (for a Typical Device)
When Using the Battery monitor for Temperature
Monitoring
Constant
Typ
A
6470
B
334
Note that A should be relatively constant for all devices, but B is not. Information that can be used to
calculate B for a given chip is included in the chip's information page (see Section 2.2.3 for info about the
information page).
Example:
Find the BATTMON_VOLTAGE setting that tells whether the temperature is above or below 75°C.
6470
BATTMON_VOLTAGE<4:0> =
= 15.82
75 + 334
(2)
The closest setting is 16, which corresponds to ~70°C (see Table 13-1). By writing 16 to
BATTMON_VOLTAGE, an output of BATTMON_OUT = 1 tells that the temperature is above 70°C, whereas
BATTMON_OUT = 0 tells that it is below 70°C.
13.3 Battery Monitor Registers
This section describes the battery monitor registers.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Battery Monitor
141
Battery Monitor Registers
www.ti.com
BATTMON (0x6264) – Battery Monitor
Bit
7
6
Name
–
BATTMON_OUT
Reset
0
0
R/W
R0
R
5:1
BATTMON_VOLTAGE
11100 R/W
0
BATTMON_PD
1
R/W
Description
Reserved. Always read 0
Result from the battery monitor.
1: Voltage is above value set in BATTMON_VOLTAGE.
0: Voltage is below the value set in BATTMON_VOLTAGE.
Note that the value of BATTMON_OUT is undefined except when BATTMON_PD is 0 and has
been 0 for 2 us.
Controls the trigger point for the battery monitor. The step size is 24 mV for the first 23 settings,
and then 169 mV (unless temperature-sense mode is enabled; see Section 13.2 for details).
Range to be used: 3–31
3: 1.93 V
4: 1.93 V + (4 – 3) × 0.024 V = 1.954 V
5: 1.93 V + (5 – 3) × 0.024 V = 1.978 V
6: 1.93 V + (6 – 3) × 0.024 V = 2.002 V
7: 1.93 V + (7 – 3) × 0.024 V = 2.026 V
8: 1.93 V + (8 – 3) × 0.024 V = 2.050 V
9: 1.93 V + (9 – 3) × 0.024 V = 2.074 V
10: 1.93 V + (10 – 3) × 0.024 V = 2.098 V
11: 1.93 V + (11 – 3) × 0.024 V = 2.122 V
12: 1.93 V + (12 – 3) × 0.024 V = 2.146 V
13: 1.93 V + (13 – 3) × 0.024 V = 2.170 V
14: 1.93 V + (14 – 3) × 0.024 V = 2.194 V
15: 1.93 V + (15 – 3) × 0.024 V = 2.218 V
16: 1.93 V + (16 – 3) × 0.024 V = 2.242 V
17: 1.93 V + (17 – 3) × 0.024 V = 2.266 V
18: 1.93 V + (18 – 3) × 0.024 V = 2.290 V
19: 1.93 V + (19 – 3) × 0.024 V = 2.314 V
20: 1.93 V + (20 – 3) × 0.024 V = 2.338 V
21: 1.93 V + (21 – 3) × 0.024 V = 2.362 V
22: 1.93 V + (22 – 3) × 0.024 V = 2.386 V
23: 1.93 V + (23 – 3) × 0.024 V = 2.410 V
24: 1.93 V + (24 – 3) × 0.024 V = 2.434 V
25: 1.93 V + (25 – 3) × 0.024 V = 2.458 V
26: 1.93 V + (26 – 3) × 0.024 V = 2.482 V
27: 2.482 V + (27 – 26) × 0.169 V = 2.651 V
28: 2.482 V + (28 – 26) × 0.169 V = 2.820 V
29: 2.482 V + (29 – 26) × 0.169 V = 2.989 V
30: 2.482 V + (30 – 26) × 0.169 V = 3.158 V
31 : 2.48 2V + (31 – 26) × 0.169 V = 3.327 V
Turns on the battery monitor. Wait at least 2 µs before reading BATTMON_OUT.
0: Enable the battery monitor.
1: Disable the battery monitor.
One should turn the battery monitor off (BATTMON_PD = 1) after reading out the measurement
BATTMON_OUT in order to save power, as the battery monitor consumes power when enabled
(BATTMON_PD = 0).
MONMUX (0x61A6) – Monitor MUX
Bit
7:1
0
Name
–
BATTMON_INPUT
Reset
0
R/W
R/W
Description
Reserved
Determines the input to the battery moitor:
0: Supply voltage (AVDD5)
1: Voltage from the temperature sensor, which needs to be enabled using the
ATEST.ATESTCTRL register; described in Section 23.15.3
142
Battery Monitor
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 14
SWRU191D – April 2009 – Revised March 2013
Random-Number Generator
This chapter provides more information about the random-number generator and its usage.
Topic
14.1
14.2
14.3
...........................................................................................................................
Page
Introduction .................................................................................................... 144
Random-Number-Generator Operation ............................................................... 144
Random-Number-Generator Registers ............................................................... 145
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Random-Number Generator
143
Introduction
www.ti.com
14.1 Introduction
The random-number generator has the following features.
• Generates pseudorandom bytes which can be read by the CPU or used directly by the command
strobe processor (see Section 23.14).
• Calculates CRC16 of bytes that are written to RNDH.
• Seeded by value written to RNDL.
The random-number generator is a 16-bit linear-feedback shift register (LFSR) with polynomial X16 + X15 +
X2 + 1 (i.e., CRC16). It uses different levels of unrolling depending on the operation it performs. The basic
version (no unrolling) is shown in Figure 14-1.
The random-number generator is turned off when ADCCON1.RCTRL = 11.
15
in_bit
+
14
13
12
11
10
9
8
7
6
5
4
3
2
+
1
0
+
M0105-01
Figure 14-1. Basic Structure of the Random-Number Generator
14.2 Random-Number-Generator Operation
The operation of the random-number generator is controlled by the ADCCON1.RCTRL bits (see also
Section 12.2.10). The current value of the 16-bit shift register in the LFSR can be read from the RNDH and
RNDL registers.
14.2.1 Pseudorandom Sequence Generation
The default operation (ADCCON1.RCTRL is 00) is to clock the LFSR once (13× unrolling; where clocking
with 13× unrolling means performing an operation equivalent to doing 13 shifts with feedback) each time
the command strobe processor (Section 23.14) reads the random value. This leads to the availability of a
fresh pseudorandom byte from the LSB end of the LFSR.
Another way to update the LFSR is to set ADCCON1.RCTRL to 01. This clocks the LFSR once (13×
unrolling), and the ADCCON1.RCTRL bits are automatically cleared when the operation has completed.
14.2.2 Seeding
The LFSR can be seeded by writing to the RNDL register twice. Each time the RNDL register is written, the
8 LSBs of the LFSR are copied to the 8 MSBs and the 8 LSBs are replaced with the new data byte that
was written to RNDL.
For the CC253x, when a random value is required, the LFSR should be seeded by writing RNDL with
random bits from the IF_ADC in the RF receive path. To use this seeding method, the radio must first be
powered on. The radio should be placed in the infinite RX state to avoid possible sync detect in the RX
state. The random bits from the IF_ADC are read from the least significant bit position of the RF register
RFRND. These bits should be concatenated over time to form the bytes needed for the random-numbergenerator seed. See Section 23.12 for a description of the randomness of these numbers. Note that this
cannot be done while the radio is in use for normal tasks.
Note that a seed value of 0x0000 or 0x8003 always leads to an unchanged value in the LFSR after
clocking, as no values are pushed in via in_bit (see Figure 14-1); hence, neither of these seed values
should not be used for random-number generation.
14.2.3 CRC16
The LFSR can also be used to calculate the CRC value of a sequence of bytes. Writing to the RNDH
register triggers a CRC calculation. The new byte is processed from the MSB end and an 8× unrolling is
used, so that a new byte can be written to RNDH every clock cycle.
144
Random-Number Generator
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Random-Number-Generator Registers
www.ti.com
Note that the LFSR must be properly seeded by writing to RNDL before the CRC calculations start.
Usually, the seed value for CRC calculations should be 0x0000 or 0xFFFF.
14.3 Random-Number-Generator Registers
This section describes the random-number-generator registers.
RNDL (0xBC) – Random-Number-Generator Data, Low Byte
Bit
7:0
Name
RNDL[7:0]
Reset
0xFF
R/W
R/W
Description
Random value/seed or CRC result, low byte
When used for random-number generation, writing to this register twice seeds the randomnumber generator. Writing to this register copies the 8 LSBs of the LFSR to the 8 MSBs and
replaces the 8 LSBs with the data value written.
The value returned when reading from this register is the 8 LSBs of the LFSR.
When used for random-number generation, reading this register returns the 8 LSBs of the
random number. When used for CRC calculations, reading this register returns the 8 LSBs of the
CRC result.
RNDH (0xBD) – Random-Number-Generator Data, High Byte
Bit
7:0
Name
RNDH[7:0]
Reset
0xFF
R/W
R/W
Description
Random value or CRC result/input data, high byte
When written, a CRC16 calculation is triggered, and the data value written is processed starting
with the MSB.
The value returned when reading from this register is the 8 MSBs of the LFSR.
When used for random-number generation, reading this register returns the 8 MSBs of the
random number. When used for CRC calculations, reading this register returns the 8 MSBs of the
CRC result.
ADCCON1 (0xB4) – ADC Control 1 (see also Section 12.2.10)
Bit
7:4
Name
–
Reset
0011
R/W
-
3:2
RCTRL[1:0]
00
R/W
1:0
–
11
R/W
Description
For CC2533, these bits are reserved. For the other devices, see the ADCCON1 (0xB4) –
ADC Control 1 description in Section 12.2.10.
Controls the 16-bit random-number generator (Chapter 14). When 01 is written, the setting
automatically returns to 00 when the operation has completed.
00:
Normal operation. (13× unrolling)
01:
Clock the LFSR once (13× unrolling)
10:
Reserved
11:
Stopped. Random-number generator is turned off.
Reserved. Always set to 11
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Random-Number Generator
145
Chapter 15
SWRU191D – April 2009 – Revised March 2013
AES Coprocessor
The Advanced Encryption Standard (AES) coprocessor allows encryption/decryption to be performed with
minimal CPU usage.
The coprocessor has the following features:
• Supports all security suites in IEEE 802.15.4
• ECB, CBC, CFB, OFB, CTR, and CBC-MAC modes
• Hardware support for CCM mode
• 128-bit key and IV/nonce
• DMA transfer trigger capability
Topic
...........................................................................................................................
15.1
15.2
15.3
15.4
15.5
15.6
15.7
15.8
15.9
15.10
146
AES Operation ................................................................................................
Key and IV ......................................................................................................
Padding of Input Data ......................................................................................
Interface to CPU ..............................................................................................
Modes of Operation .........................................................................................
CBC-MAC .......................................................................................................
CCM Mode ......................................................................................................
AES Interrupts .................................................................................................
AES DMA Triggers ...........................................................................................
AES Registers ................................................................................................
AES Coprocessor
Page
147
147
147
147
147
147
148
150
150
150
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
AES Operation
www.ti.com
15.1 AES Operation
To
•
•
•
encrypt a message, the following procedure must be followed (ECB, CBC):
Load key
Load initialization vector (IV)
Download and upload data for encryption/decryption.
The AES coprocessor works on blocks of 128 bits. A block of data is loaded into the coprocessor,
encryption is performed, and the result must be read out before the next block can be processed. Before
each block is loaded , a dedicated start command must be sent to the coprocessor.
15.2 Key and IV
Before a key or IV/nonce load starts, an appropriate load key or IV/nonce command must be issued to the
coprocessor. When loading the IV, it is important also to set the correct mode.
A key load or IV load operation aborts any processing that could be running. The key, once loaded, stays
valid until a key reload takes place.
The IV must be downloaded before the beginning of each message (not each block).
Both the key and IV values are cleared by a reset of the device and when PM2 or PM3 is entered.
15.3 Padding of Input Data
The AES coprocessor works on blocks of 128 bits. If the last block contains less than 128 bits, it must be
padded with zeros when written to the coprocessor.
15.4 Interface to CPU
The CPU communicates with the coprocessor using three SFR registers:
• ENCCS, encryption control and status register
• ENCDI, encryption input register
• ENCDO, encryption output register
Read/write to the status register is done directly by the CPU, whereas access to the input/output registers
should be performed using direct memory access (DMA).
When using DMA with the AES coprosessor, two DMA channels must be used, one for input data and one
for output data. The DMA channels must be initialized before a start command is written to ENCCS. Writing
a start command generates a DMA trigger, and the transfer is started. After each block is processed, an
interrupt is generated. The interrupt is used to issue a new start command to ENCCS.
15.5 Modes of Operation
When using CFB, OFB, or CTR mode, the 128-bit blocks are divided into four 32-bit blocks. The 32 bits
are loaded into the AES coprocessor, and the resulting 32 bits are read out. This continues until all 128
bits have been encrypted. The only time one must consider this is if data is loaded/read directly using the
CPU. When using DMA, this is handled automatically by the DMA triggers generated by the AES
coprocessor; thus, DMA is preferred.
Both encryption and decryption are performed similarly.
The CBC-MAC mode is a variant of the CBC mode. See Section 15.6 for an explanation.
CCM is a combination of CBC-MAC and CTR. Parts of the CCM must therefore be done in software. The
following section gives a short explanation of the necessary steps to be done.
15.6 CBC-MAC
When performing CBC-MAC encryption, data is downloaded to the coprocessor in CBC-MAC mode one
block at a time, except for the last block. Before the last block is loaded, the mode is changed to CBC.
The last block is downloaded and the block uploaded is the message MAC.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
AES Coprocessor
147
CCM Mode
www.ti.com
CBC-MAC decryption is similar to encryption. The message MAC uploaded must be compared with the
MAC to be verified.
15.7 CCM Mode
To encrypt a message in CCM mode, the following sequence can be conducted (key is already loaded):
Message Authentication Phase
This phase takes place during the following steps 1–6.
1. The software loads the IV with zeros.
2. The software creates block B0. The layout of block B0 is shown in Figure 15-1.
Name
B0
0
Flag
Byte
Name
1
2
3
4
Designation
First Block for Authentication in CCM Mode
5
6
7
8
9
10
Nonce
11
12
13
14
15
L_M
Figure 15-1. Message Authentication Phase Block B0
There is no restriction on the nonce value. L_M is the message length in bytes.
For 802.15.4, nonce is 13 bytes and L_M is 2 bytes.
The content of the authentication flag byte is described in Figure 15-2.
L is set to 6 in this example. So, L – 1 is set to 5. M and A_Data can be set to any value.
Name
FLAG/B0
7
Reserved
0
Bit
Name
Value
3.
4.
5.
6.
7.
148
6
A_Data
x
Designation
Authentication Flag Field for CCM mode
5
4
3
(M – 2)/2
x
x
x
2
1
1
L–1
0
0
1
Figure 15-2. Authentication Flag Byte
If some additional authentication data (denoted a, following) is needed (that is, A_Data = 1), the
software creates the A_Data length field, called L(a) by:
• (a) If l(a) = 0, (that is, A_Data = 0), then L(a) is the empty string. Note that l(a) is the length of a in
octets.
• (b) If 0 < l(a) < 216 – 28, then L(a) is the 2-octet encoding of l(a).
The additional authentication data is appended to the A_Data length field L(a). The additional
authentication blocks are padded with zeros until the last additional authentication block is full. There is
no restriction on the length of a.
AUTH-DATA = L(a) + Authentication Data + (zero padding)
The last block of the message is padded with zeros until full (that is, if its length is not an integral
multiple of 128 bits).
The software concatenates block B0, the additional authentication blocks if any, and the message;
Input message = B0 + AUTH-DATA + Message + (zero padding of message)
Once the input message authentication by CBC-MAC is finished, the software leaves the uploaded
buffer contents unchanged (M = 16), or keeps only the higher-M bytes of the buffer unchanged, while
setting the lower bits to 0 (M != 16).
The result is called T.
Message Encryption
The software creates the key stream block A0. Note that L = 6, with the current example of the CTR
generation. The content is shown in Figure 15-3.
Note that when encrypting authentication data T to generate U in OFB mode, the CTR value must be
zero. When encrypting message blocks using CTR mode, the CTR value must be any value but zero.
The content of the encryption-flag byte is described in Figure 15-4.
AES Coprocessor
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CCM Mode
www.ti.com
Byte
Name
Name
A0
0
Flag
1
2
3
4
Designation
First CTR Value for CCM Mode
5
6
7
8
Nonce
9
10
11
12
13
14
15
CTR
Figure 15-3. Message Encryption Phase Block
Bit
Name
Value
Name
FLAG/A0
7
Designation
Encryption Flag Field for CCM Mode
5
4
3
—
0
0
0
6
Reserved
0
0
2
1
1
L–1
0
0
1
Figure 15-4. Encryption Flag Byte
8. The software loads A0 by selecting a Load IV/nonce command. To do so, it sets the mode to CFB or
OFB at the same time it selects the Load IV/nonce command.
9. The software calls a CFB or an OFB encryption on the authenticated data T. The uploaded buffer
contents stay unchanged (M = 16), or only its first M bytes stay unchanged, the others being set to 0
(M – 16). The result is U, which is used later.
10. The software calls a CTR-mode encryption immediately on the still-padded message blocks. It must
reload the IV when the CTR value is any value but zero.
11. The encrypted authentication data U is appended to the encrypted message. This gives the final
result, c.
Result C = encrypted message(m) + U
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
AES Coprocessor
149
AES Interrupts
www.ti.com
Message Decryption
CCM Mode Decryption
In the coprocessor, the automatic generation of CTR works on 32 bits; therefore, the maximum length of a
message is 128 × 232 bits, that is 236 bytes, which can be written in a 6-bit word. So, the value L is set to 6.
To decrypt a CCM-mode processed message, the following sequence can be conducted (key is already
loaded).
Message Parsing Phase
1. The software parses the message by separating the M rightmost octets, namely U, and the other
octets, namely string C.
2. C is padded with zeros until it can fill an integral number of 128-bit blocks.
3. U is padded with zeros until it can fill a 128-bit block.
4. The software creates the key stream block A0. It is done the same way as for CCM encryption.
5. The software loads A0 by selecting a Load IV/nonce command. To do so, it sets the mode to CFB or
OFB at the same time as it selects the IV load.
6. The software calls a CFB or an OFB encryption on the encrypted authenticated data U. The uploaded
buffer contents stay unchanged (M = 16), or only its first M bytes stay unchanged, the others being set
to 0 (M != 16). The result is T.
7. The software calls a CTR-mode decryption immediately on the encrypted message blocks C.
Reloading the IV/CTR is not necessary.
Reference Authentication Tag Generation
This phase is identical to the authentication phase of CCM encryption. The only difference is that the
result is named MACTag (instead of T).
Message Authentication Checking Phase
The software compares T with MACTag.
15.8 AES Interrupts
The AES interrupt, ENC, is produced when encryption or decryption of a block is completed. The interrupt
enable bit is IEN0.ENCIE, and the interrupt flag is S0CON.ENCIF.
15.9 AES DMA Triggers
Two DMA triggers are associated with the AES coprocessor. These are ENC_DW, which is active when
input data must be downloaded to the ENCDI register, and ENC_UP, which is active when output data
must be uploaded from the ENCDO register.
The ENCDI and ENCDO registers should be set as destination and source locations for DMA channels
used to transfer data to or from the AES coprocessor.
15.10 AES Registers
The AES coprocessor registers have the layout shown in this section.
The registers return to their reset value when the chip enters PM2 or PM3.
150
AES Coprocessor
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
AES Registers
www.ti.com
ENCCS (0xB3) – Encryption Control and Status
Bit
7
6:4
Name
–
MODE[2:0]
Reset
0
000
R/W
R0
R/W
3
RDY
1
R
2:1
CMD[1:0]
0
R/W
Description
Reserved, always read as 0
Encryption/decryption mode
000: CBC
001: CFB
010: OFB
011: CTR
100: ECB
101: CBC MAC
110: Reserved
111: Reserved
Encryption/decryption ready status
0:
Encryption/decryption in progress
1:
Encryption/decryption is completed.
Command to be performed when a 1 is written to ST
R/W1
H0
00:
Encrypt block
01:
Decrypt block
10:
Load key
11:
Load IV/nonce
Start processing command set by CMD. Must be issued for each command or 128-bit block of data.
Cleared by hardware.
0
ST
0
ENCDI (0xB1) – Encryption Input Data
Bit
7:0
Name
DIN[7:0]
Reset
0x00
R/W
R/W
Description
Encryption input data
ENCDO (0xB2) – Encryption Output Data
Bit
7:0
Name
DOUT[7:0]
Reset
0x00
R/W
R/W
Description
Encryption output data
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
AES Coprocessor
151
Chapter 16
SWRU191D – April 2009 – Revised March 2013
Watchdog Timer
The Watchdog Timer (WDT) is intended as a recovery method in situations where the CPU may be
subjected to a software upset. The WDT resets the system when software fails to clear the WDT within
the selected time interval. The watchdog can be used in applications that are subject to electrical noise,
power glitches, electrostatic discharge, etc., or where high reliability is required. If the watchdog function is
not needed in an application, it is possible to configure the Watchdog Timer to be used as an interval timer
that can be used to generate interrupts at selected time intervals.
The features of the Watchdog Timer are as follows:
• Four selectable timer intervals
• Watchdog mode
• Timer mode
• Interrupt request generation in timer mode
The WDT is configured as either a Watchdog Timer or as a timer for general-purpose use. The operation
of the WDT module is controlled by the WDCTL register. The Watchdog Timer consists of a 15-bit counter
clocked by the 32-kHz clock source. Note that the contents of the 15-bit counter are not user-accessible.
The contents of the 15-bit counter are retained during all power modes, and the Watchdog Timer
continues counting when entering active mode again.
Topic
16.1
16.2
16.3
152
...........................................................................................................................
Page
Watchdog Mode .............................................................................................. 153
Timer Mode ..................................................................................................... 153
Watchdog Timer Register ................................................................................. 153
Watchdog Timer
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Watchdog Mode
www.ti.com
16.1 Watchdog Mode
The WDT is disabled after a system reset. To start the WDT in watchdog mode, the WDCTL.MODE[1:0]
bits must be set to 10. The Watchdog Timer counter then starts incrementing from 0. When the timer is
enabled in watchdog mode, it is not possible to disable the timer. Therefore, writing 00 or 01 to
WDCTL.MODE[1:0] has no effect if the WDT is already operating in Watchdog mode.
The WDT operates with a Watchdog Timer clock frequency of 32.768 kHz (when the 32-kHz XOSC is
used). This clock frequency gives time-out periods equal to 1.9 ms, 15.625 ms, 0.25 s, and 1 s,
corresponding to the count value settings 64, 512, 8192, and 32,768, respectively.
If the counter reaches the selected timer interval value, the Watchdog Timer generates a reset signal for
the system. If a watchdog clear sequence is performed before the counter reaches the selected timer
interval value, the counter is reset to 0 and continues incrementing its value. The watchdog clear
sequence consists of writing 0xA to WDCTL.CLR[3:0], followed by writing 0x5 to the same register bits
within one watchdog clock period. If this complete sequence is not performed before the end of the
watchdog period, the Watchdog Timer generates a reset signal for the system.
When the WDT has been enabled in watchdog mode, it is not possible to change the mode by writing to
the WDCTL.MODE[1:0] bits, and the timer interval value cannot be changed.
In watchdog mode, the WDT does not produce interrupt requests.
16.2 Timer Mode
To start the WDT in timer mode, the WDCTL.MODE[1:0] bits must be set to 11. The timer is started and
the counter starts incrementing from 0. When the counter reaches the selected interval value, the timer
produces an interrupt request (IRCON2.WDTIF/IEN2.WDTIE).
In timer mode, it is possible to clear the timer contents by writing a 1 to WDCTL.CLR[0]. When the timer
is cleared, the content of the counter is set to 0. Writing 00 to WDCTL.MODE[1:0] stops the timer and
clears it to 0.
The timer interval is set by the WDCTL.INT[1:0] bits. The interval cannot be changed during timer
operation, and should be set when the timer is started. In timer mode, a reset is not produced when the
timer interval has been reached.
Note that if the watchdog mode is selected, the timer mode cannot be selected before the chip is reset.
16.3 Watchdog Timer Register
This section describes the register, WDCTL, for the Watchdog Timer.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Watchdog Timer
153
Watchdog Timer Register
www.ti.com
WDCTL (0xC9) – Watchdog Timer Control
Bit
Name
7:4 CLR[3:0]
Reset
0000
R/W
R0/W
3:2
MODE[1:0] 00
R/W
1:0
INT[1:0]
00
R/W
154
Watchdog Timer
Description
Clear timer. In watchdog mode, when 0xA followed by 0x5 is written to these bits, the timer is
cleared (i.e. loaded with 0). Note that the timer is only cleared when 0x5 is written within one
watchdog clock period after 0xA was written. Writing these bits when the Watchdog Timer is IDLE
has no effect. When operating in timer mode, the timer can be cleared to 0x0000 (but not stopped)
by writing 1 to CLR[0] (the other 3 bits are don't care).
Mode select. These bits are used to start the WDT in watchdog mode or timer mode. Setting these
bits to IDLE stops the timer when in timer mode. Note: to switch to watchdog mode when operating
in timer mode, first stop the WDT - then start the WDT in Watchdog mode. When operating in
Watchdog mode, writing these bits has no effect.
00: IDLE
01: Reserved
10: Watchdog mode
11: Timer mode
Timer interval select. These bits select the timer interval, which is defined as a given number of 32kHz oscillator periods. Note that the interval can only be changed when the WDT is IDLE, so the
interval must be set at the same time as the timer is started.
00: Clock period × 32,768 (~1 s) when running the 32-kHz XOSC
01: Clock period × 8192 (~0.25 s)
10: Clock period × 512 (~15.625 ms)
11: Clock period × 64 (~1.9 ms)
For CC253x and CC2540, when clock division is enabled through CLKCONCMD.CLKSPD, the
length of the watchdog timer interval is reduced by a factor equal to the current oscillator clock
frequency divided by the set clock speed. E.g., if 32-MHz crystal is selected and clock speed is set
to 4 MHz, then the watchdog timeout is reduced by a factor o 32 MHz/4 MHz = 8. If the watchdog
interval set by WDCTL.INT was 1 s, nominally it is 1/8 s with this clock division factor. For CC2541,
the watchdog timer interval is independent of the clock division rate.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 17
SWRU191D – April 2009 – Revised March 2013
USART
USART 0 and USART 1 are serial communications interfaces that can be operated separately in either
asynchronous UART mode or in synchronous SPI mode. The two USARTs have identical function, and
are assigned to separate I/O pins. See Section 7.6 for I/O configuration.
Topic
17.1
17.2
17.3
17.4
17.5
17.6
17.7
17.8
...........................................................................................................................
UART Mode .....................................................................................................
SPI Mode ........................................................................................................
SSN Slave-Select Pin .......................................................................................
Baud-Rate Generation ......................................................................................
USART Flushing ..............................................................................................
USART Interrupts ............................................................................................
USART DMA Triggers .......................................................................................
USART Registers .............................................................................................
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USART
Page
156
157
158
158
159
159
159
159
155
UART Mode
www.ti.com
17.1 UART Mode
For asynchronous serial interfaces, the UART mode is provided. In the UART mode, the interface uses a
two-wire or four-wire interface consisting of the pins RXD and TXD, and optionally RTS and CTS. The
UART mode of operation includes the following features:
• 8 or 9 payload bits
• Odd, even, or no parity
• Configurable start- and stop-bit levels
• Configurable LSB- or MSB-first transfer
• Independent receive and transmit interrupts
• Independent receive and transmit DMA triggers
• Parity and framing error status
The UART mode provides full-duplex asynchronous transfers, and the synchronization of bits in the
receiver does not interfere with the transmit function. A UART byte transfer consists of a start bit, eight
data bits, an optional ninth data or parity bit, and one or two stop bits. Note that the data transferred is
referred to as a byte, although the data can actually consist of eight or nine bits.
The UART operation is controlled by the USART control and status registers, UxCSR, and the UART
control registers, UxUCR, where x is the USART number, 0 or 1.
The UART mode is selected when UxCSR.MODE is set to 1.
17.1.1 UART Transmit
A UART transmission is initiated when the USART receive/transmit data buffers, UxDBUF, are written. The
byte is transmitted on the TXDx output pins. The UxDBUF registers are double-buffered.
The UxCSR.ACTIVE bit goes high when the byte transmission starts and low when it ends. When the
transmission ends, the UxCSR.TX_BYTE bit is set to 1. An interrupt request is generated when the
UxDBUF register is ready to accept new transmit data. This happens immediately after the transmission
has been started; hence, a new data byte value can be loaded into the data buffer while the byte is being
transmitted.
17.1.2 UART Receive
Data reception on the UART is initiated when a 1 is written to the UxCSR.RE bit. The UART then searches
for a valid start bit on the RXDx input pin and sets the UxCSR.ACTIVE bit high. When a valid start bit has
been detected, the received byte is shifted into the receive register. The UxCSR.RX_BYTE bit is set and a
receive interrupt is generated when the operation has completed. At the same time, UxCSR.ACTIVE goes
low.
The received data byte is available through the UxDBUF register. When UxDBUF is read, UxCSR.RX_BYTE
is cleared by hardware.
NOTE:
When the application has read UxDBUF, it is important that it does not clear
UxCSR.RX_BYTE. Clearing UxCSR.RX_BYTE implicitly makes the UART believe that the
UART RX shift register is empty, even though it might hold pending data (typically due to
back-to-back transmission). Consequently, the UART asserts (TTL low) the RT/RTS line,
which allows flow into the UART, leading to potential overflow. Hence, the
UxCSR.RX_BYTE flag integrates closely with the automatic RT/RTS function and must
therefore be controlled solely by the SoC UART itself. Otherwise, the application could
typically experience that the RT/RTS line remains asserted (TTL low), even though a backto-back transmission clearly suggests it ought to intermittently pause the flow.
17.1.3 UART Hardware Flow Control
Hardware flow control is enabled when the UxUCR.FLOW bit is set to 1. The RTS output is driven low when
the receive register is empty and reception is enabled. Transmission of a byte does not occur before the
CTS input goes low.
156
USART
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
SPI Mode
www.ti.com
17.1.4 UART Character Format
If the BIT9 and PARITY bits in register UxUCR are set high, parity generation and detection is enabled.
The parity is computed and transmitted as the ninth bit, and during reception, the parity is computed and
compared to the received ninth bit. If there is a parity error, the UxCSR.ERR bit is set high. This bit is
cleared when UxCSR is read.
The number of stop bits to be transmitted is set to one or two bits, as determined by the register bit
UxUCR.SPB. The receiver always checks for one stop bit. If the first stop bit received during reception is
not at the expected stop bit level, a framing error is signaled by setting register bit UxCSR.FE high.
UxCSR.FE is cleared when UxCSR is read. The receiver checks both stop bits when UxUCR.SPB is set.
Note that the RX interrupt is set when the first stop bit is checked OK. If second stop bit is not OK, there is
a delay in setting the framing error bit, UxCSR.FE. This delay is baud-rate dependent (bit duration).
17.2 SPI Mode
This section describes the SPI mode of operation for synchronous communication. In SPI mode, the
USART communicates with an external system through a three-wire or four-wire interface. The interface
consists of the pins MOSI, MISO, SCK, and SS_N. See Section 7.6 for a description of how the USART
pins are assigned to the I/O pins.
The SPI mode includes the following features:
• Three-wire (master) and four-wire SPI interface
• Master and slave modes
• Configurable SCK polarity and phase
• Configurable LSB- or MSB-first transfer
The SPI mode is selected when UxCSR.MODE is set to 0.
In SPI mode, the USART can be configured to operate either as a SPI master or as a SPI slave by writing
the UxCSR.SLAVE bit.
17.2.1 SPI Master Operation
A SPI byte transfer in master mode is initiated when the UxDBUF register is written. The USART generates
the SCK serial clock using the baud-rate generator (see Section 17.4) and shifts the provided byte from
the transmit register onto the MOSI output. At the same time, the receive register shifts in the received
byte from the MISO input pin.
The UxCSR.ACTIVE bit goes high when the transfer starts and low when the transfer ends. When the
transfer ends, the UxCSR.TX_BYTE bit is set to 1.
The polarity and clock phase of the serial clock SCK is selected by UxGCR.CPOL and UxGCR.CPHA. The
order of the byte transfer is selected by the UxGCR.ORDER bit.
At the end of the transfer, the received data byte is available for reading from the UxDBUF. A receive
interrupt is generated when this new data is ready in the UxDBUF USART receive/transmit data register.
A transmit interrupt is generated when the unit is ready to accept another data byte for transmission.
Because UxDBUF is double-buffered, this happens just after the transmission has been initiated. Note that
data should not be written to UxDBUF until UxCSR.TX_BYTE is 1. For DMA transfers, this is handled
automatically. For back-to-back transmits using DMA, the UxGCR.CPHA bit must be set to zero; if not,
transmitted bytes can become corrupted. For systems requiring setting of UxGCR.CPHA, polling
UxCSR.TX_BYTE is needed.
Also, note the difference between transmit interrupt and receive interrupt, as the former arrives
approximately eight bit-periods prior to the latter.
SPI master-mode operation as described previously is a three-wire interface. No select input is used to
enable the master. If the external slave requires a slave-select signal, this can be implemented through
software using a general-purpose I/O pin.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USART
157
SSN Slave-Select Pin
www.ti.com
17.2.2 SPI Slave Operation
A SPI byte transfer in slave mode is controlled by the external system. The data on the MOSI input is
shifted into the receive register controlled by the serial clock, SCK, which is an input in slave mode. At the
same time, the byte in the transmit register is shifted out onto the MISO output.
The UxCSR.ACTIVE bit goes high when the transfer starts and low when the transfer ends. Then the
UxCSR.RX_BYTE bit is set and a receive interrupt is generated.
The expected polarity and clock phase of SCK is selected by UxGCR.CPOL and UxGCR.CPHA. The
expected order of the byte transfer is selected by the UxGCR.ORDER bit.
At the end of the transfer, the received data byte is available for reading from UxDBUF.
The transmit interrupt is generated at the start of the operation.
17.3 SSN Slave-Select Pin
When the USART is operating in SPI mode, configured as a SPI slave, a four-wire interface is used with
the slave-select (SSN) pin as an input to the SPI. When SSN is low, the SPI slave is active, receives data
on the MOSI input, and outputs data on the MISO output. When SSN is high, the SPI slave is inactive and
does not receive data. The MISO output is in the high-impedance state when SSN is high. Also note that
the release of SSN (SSN going high) must be aligned to the end of the byte received or sent. If released
during a byte, the next received byte is not received properly, as information about the previous byte is
present in the SPI system. A USART flush can be used to remove this information.
In SPI master mode, the SSN pin is not used. When the USART operates as a SPI master and a slaveselect signal is required by an external SPI slave device, then a general-purpose I/O pin should be used to
implement the slave-select signal function in software.
17.4 Baud-Rate Generation
An internal baud-rate generator sets the UART baud rate when operating in UART mode and the SPI
master clock frequency when operating in SPI mode.
The UxBAUD.BAUD_M[7:0] and UxGCR.BAUD_E[4:0] registers define the baud rate used for UART
transfers and the rate of the serial clock for SPI transfers. The baud rate is given by the following
equation:
Baud Rate =
(256 + BAUD _ M) ´ 2BAUD _ E
228
´f
(3)
where f is the system clock frequency, 16 MHz for the RCOSC or 32 MHz for the XOSC.
The register values required for standard baud rates are shown in Table 17-1 for a typical system clock
set to 32 MHz. The table also gives the difference in actual baud rate to standard baud rate value as a
percentage error.
The maximum baud rate for the UART mode is f/16 when BAUD_E is 16 and BAUD_M is 0, and where f is
the system clock frequency.
See the device data sheet for the maximum baud rate in SPI mode.
Note that the baud rate must be set through the UxBAUD and UxGCR registers before any other UART or
SPI operations take place. If the baud rate is changed while in UART mode, it may take up to one bit
period of the old baud rate before the change takes effect.
Table 17-1. Commonly Used Baud-Rate Settings for 32 MHz System Clock
Baud Rate (bps)
158
USART
UxBAUD.BAUD_M
UxGCR.BAUD_E
Error (%)
2400
59
6
0.14
4800
59
7
0.14
9600
59
8
0.14
14,400
216
8
0.03
19,200
59
9
0.14
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USART Flushing
www.ti.com
Table 17-1. Commonly Used Baud-Rate Settings for 32 MHz System
Clock (continued)
Baud Rate (bps)
UxBAUD.BAUD_M
UxGCR.BAUD_E
Error (%)
28,800
216
9
0.03
38,400
59
10
0.14
57,600
216
10
0.03
76,800
59
11
0.14
115,200
216
11
0.03
230,400
216
12
0.03
17.5 USART Flushing
The current operation can be aborted by setting the UxUCR.FLUSH register bit. This event stops the
current operation and clears all data buffers. It should be noted that when setting the flush bit in the middle
of a TX/RX bit, the flushing does not take place until this bit has ended (buffers are cleared immediately,
but timers keeping knowledge of bit duration are not). Thus, using the flush bit should either be aligned
with USART interrupts or use a wait time of one bit duration at the current baud rate before updated data
or configuration can be received by the USART.
17.6 USART Interrupts
Each USART has two interrupts. These are the RX complete interrupt (URXx) and the TX interrupt
(UTXx). The TX interrupt is triggered when transmission starts and the data buffer is offloaded.
The USART interrupt enable bits are found in the IEN0 and IEN2 registers. The interrupt flags are located
in the TCON and IRCON2 registers. See Section 2.5 for details of these registers. The interrupt enables
and flags are summarized as follows.
Interrupt enables:
• USART0 RX: IEN0.URX0IE
• USART1 RX: IEN0.URX1IE
• USART0 TX: IEN2.UTX0IE
• USART1 TX: IEN2.UTX1IE
Interrupt flags:
• USART0 RX: TCON.URX0IF
• USART1 RX: TCON.URX1IF
• USART0 TX: IRCON2.UTX0IF
• USART1 TX: IRCON2.UTX1IF
17.7 USART DMA Triggers
There are two DMA triggers associated with each USART. The DMA triggers are activated by RX
complete and TX complete events, i.e., the same events as the USART interrupt requests. A DMA
channel can be configured using a USART receive/transmit buffer, UxDBUF, as source or destination
address.
See Table 8-1 for an overview of the DMA triggers.
17.8 USART Registers
The registers for the USART are described in this section. For each USART there are five registers
consisting of the following (x refers to the USART number, i.e., 0 or 1):
• UxCSR, USART x control and status
• UxUCR, USART x UART control
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USART
159
USART Registers
•
•
•
www.ti.com
UxGCR, USART x generic control
UxDBUF, USART x receive/transmit data buffer
UxBAUD, USART x baud-rate control
U0CSR (0x86) – USART 0 Control and Status
Bit
7
Name
MODE
Reset
0
R/W
R/W
6
RE
0
R/W
5
SLAVE
0
R/W
4
FE
0
R/W0
3
2
ERR
RX_BYTE
0
0
R/W0
R/W0
1
TX_BYTE
0
R/W0
0
ACTIVE
0
R
160
USART
Description
USART mode select
0:
SPI mode
1:
UART mode
UART receiver enable. Note: Do not enable receive before UART is fully configured.
0:
Receiver disabled
1:
Receiver enabled
SPI master or slave mode select
0:
SPI master
1:
SPI slave
UART framing error status. This bit is automatically cleared on a read of the U0CSR register or bits
in the U0CSR register.
0:
No framing error detected
1:
Byte received with incorrect stop-bit level
UART parity error status. This bit is automatically cleared on a read of the U0CSR register or bits in
the U0CSR register.
0:
No parity error detected
1:
Byte received with parity error
Receive byte status. UART mode and SPI slave mode. This bit is automatically cleared when
reading U0DBUF; clearing this bit by writing 0 to it effectively discards the data in U0DBUF.
0:
No byte received
1:
Received byte ready
Transmit byte status. UART mode and SPI master mode
0:
Byte not transmitted
1:
Last byte written to data-buffer register has been transmitted
USART transmit/receive active status. In SPI slave mode, this bit equals slave select.
0:
USART idle
1:
USART busy in transmit or receive mode
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USART Registers
www.ti.com
U0UCR (0xC4) – USART 0 UART Control
Bit
7
6
Name
FLUSH
FLOW
Reset
0
0
R/W
R0/W1
R/W
5
D9
0
R/W
4
3
BIT9
PARITY
0
0:
Odd parity
1:
Even parity
Set this bit to 1 in order to enable the parity bit tranfer (as 9th bit). The content of this 9th bit is given
by D9, if parity is enabled by PARITY.
R/W
0
0:
8-bit transfer
1:
9-bit transfer
UART parity enable. One must set BIT9 in addition to setting this bit for parity to be calculated.
R/W
2
SPB
0
R/W
1
STOP
1
R/W
0
START
0
R/W
Description
Flush unit. When set, this event stops the current operation and returns the unit to the idle state.
UART hardware flow enable. Selects use of hardware flow control with RTS and CTS pins
0:
Flow control disabled
1:
Flow control enabled
If parity is enabled (see PARITY, bit 3 in this register), then this bit sets the parity level as follows:
0:
1:
UART
0:
1:
UART
0:
1:
UART
0:
1:
Parity disabled
Parity enabled
number of stop bits. Selects the number of stop bits to transmit
1 stop bit
2 stop bits
stop-bit level must be different from the start-bit level
Low stop bit
High stop bit
start-bit level. Ensure that the polarity of the start bit is opposite the level of the idle line.
Low start bit
High start bit
U0GCR (0xC5) – USART 0 Generic Control
Bit
7
Name
CPOL
Reset
0
R/W
R/W
6
CPHA
0
R/W
5
ORDER
0
R/W
4:0
BAUD_E[4:0]
0 0000
R/W
Description
SPI clock polarity
0:
Negative clock polarity
1:
Positive clock polarity
SPI clock phase
Data is output on MOSI when SCK goes from CPOL inverted to CPOL, and data input
0:
is sampled on MISO when SCK goes from CPOL to CPOL inverted.
1:
Data is output on MOSI when SCK goes from CPOL to CPOL inverted, and data input
is sampled on MISO when SCK goes from CPOL inverted to CPOL.
Bit order for transfers
0:
LSB first
1:
MSB first
Baud rate exponent value. BAUD_E along with BAUD_M determines the UART baud rate and
the SPI master SCK clock frequency.
U0DBUF (0xC1) – USART 0 Receive/Transmit Data Buffer
Bit
7:0
Name
DATA[7:0]
Reset
0x00
R/W
R/W
Description
USART receive and transmit data. When writing this register, the data written is written to the
internal transmit-data register. When reading this register, the data from the internal read-data
register is read.
U0BAUD (0xC2) – USART 0 Baud-Rate Control
Bit
7:0
Name
Reset
BAUD_M[7:0] 0x00
R/W
R/W
Description
Baud-rate mantissa value. BAUD_E along with BAUD_M decides the UART baud rate and the
SPI master SCK clock frequency.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USART
161
USART Registers
www.ti.com
U1CSR (0xF8) – USART 1 Control and Status
Bit
7
Name
MODE
Reset
0
R/W
R/W
6
RE
0
R/W
5
SLAVE
0
R/W
4
FE
0
R/W0
3
2
ERR
RX_BYTE
0
0
R/W0
R/W0
1
TX_BYTE
0
R/W0
0
ACTIVE
0
R
Description
USART mode select
0:
SPI mode
1:
UART mode
UART receiver enable. Note: Do not enable receive before UART is fully configured.
0:
Receiver disabled
1:
Receiver enabled
SPI master- or slave-mode select
0:
SPI master
1:
SPI slave
UART framing error status. This bit is automatically cleared on a read of the U1CSR register or bits
in the U1CSR register.
0:
No framing error detected
1:
Byte received with incorrect stop-bit level
UART parity error status. This bit is automatically cleared on a read of the U1CSR register or bits in
the U1CSR register.
0:
No parity error detected
1:
Byte received with parity error
Receive byte status. UART mode and SPI slave mode. This bit is automatically cleared when
reading U1DBUF; clearing this bit by writing 0 to it effectively discards the data in U1DBUF.
0:
No byte received
1:
Received byte ready
Transmit byte status. UART mode and SPI master mode
0:
Byte not transmitted
1:
Last byte written to data buffer register has been transmitted
USART transmit/receive active status. In SPI slave mode, this bit equals slave select.
0:
USART idle
1:
USART busy in transmit or receive mode
U1UCR (0xFB) – USART 1 UART Control
Bit
7
6
Name
FLUSH
FLOW
Reset
0
0
R/W
R0/W1
R/W
5
D9
0
R/W
4
3
BIT9
PARITY
0
0
R/W
R/W
2
SPB
0
R/W
1
STOP
1
R/W
0
START
0
R/W
162
USART
Description
Flush unit. When set, this event stops the current operation and returns the unit to the idle state.
UART hardware flow enable. Selects use of hardware flow control with RTS and CTS pins
0:
Flow control disabled
1:
Flow control enabled
If parity is enabled (see PARITY, bit 3 in this register), then this bit sets the parity level as follows.
0:
Odd parity
1:
Even parity
Set this bit to 1 in order to enable the parity bit tranfer (as 9th bit). The content of this 9th bit is given
by D9, if parity is enabled by PARITY.
0:
8-bit transfer
1:
9-bit transfer
UART parity enable. One must set BIT9 in addition to setting this bit for parity to be calculated.
0:
1:
UART
0:
1:
UART
0:
1:
UART
0:
1:
Parity disabled
Parity enabled
number of stop bits. Selects the number of stop bits to transmit
1 stop bit
2 stop bits
stop-bit level must be different from start-bit level.
Low stop bit
High stop bit
start-bit level. Ensure that the polarity of the start bit is opposite the level of the idle line.
Low start bit
High start bit
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USART Registers
www.ti.com
U1GCR (0xFC) – USART 1 Generic Control
Bit
7
Name
CPOL
Reset
0
R/W
R/W
6
CPHA
0
R/W
5
ORDER
0
R/W
4:0
BAUD_E[4:0]
0 0000
R/W
Description
SPI clock polarity
0:
Negative clock polarity
1:
Positive clock polarity
SPI clock phase
0:
Data is output on MOSI when SCK goes from CPOL inverted to CPOL, and data input
is sampled on MISO when SCK goes from CPOL to CPOL inverted.
1:
Data is output on MOSI when SCK goes from CPOL to CPOL inverted, and data input
is sampled on MISO when SCK goes from CPOL inverted to CPOL.
Bit order for transfers
0:
LSB first
1:
MSB first
Baud rate exponent value. BAUD_E along with BAUD_M determines the UART baud rate
and the SPI master SCK clock frequency.
U1DBUF (0xF9) – USART 1 Receive/Transmit Data Buffer
Bit
7:0
Name
DATA[7:0]
Reset
0x00
R/W
R/W
Description
USART receive and transmit data. When writing this register, the data written is written to the
internal transmit-data register. When reading this register, the data from the internal read-data
register is read.
U1BAUD (0xFA) – USART 1 Baud-Rate Control
Bit
7:0
Name
Reset
BAUD_M[7:0] 0x00
R/W
R/W
Description
Baud rate mantissa value. BAUD_E along with BAUD_M determines the UART baud rate and
the SPI master SCK clock frequency.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USART
163
Chapter 18
SWRU191D – April 2009 – Revised March 2013
Operational Amplifier
The operational amplifier (in the CC2530, CC2531, and CC2540) has the following features:
• Low offset
• Ideal for use in combination with the onboard ADC in sensor applications
Topic
18.1
18.2
18.3
18.4
164
...........................................................................................................................
Description .....................................................................................................
Calibration ......................................................................................................
Clock Source ..................................................................................................
Registers ........................................................................................................
Operational Amplifier
Page
165
165
165
165
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Description
www.ti.com
18.1 Description
The operational amplifier is connected to the I/O pins as follows:
• The positive input pin is connected to P0_0.
• The negative input pin is connected to P0_1.
• The output is connected to P0_2.
The pins used by the operational amplifier must be configured as analog pins, by setting bits APCFG[2:0]
to 1. The OPAMPC.EN bit is used to enable/disable the operational amplifier. When power mode 2/3 is
entered, the operational amplifier is shut down automatically and must be restarted when entering PM0
again.
18.2 Calibration
The operational amplifier must be calibrated. A calibration is started by writing 1 to OPAMPC.CAL. During
calibration, OPAMPS.CAL_BUSY is 1. A new calibration is not accepted before OPAMPS.CAL_BUSY goes
low. Every time after enabling the operational amplifier, calibration must be performed.
18.3 Clock Source
The operational amplifier uses a divided version of the system clock. The division factor depends on which
clock source is used, HSOSC or XOSC. While the operational amplifier is enabled, the clock source
should not be changed.
18.4 Registers
This section describes the registers for the operational amplifier.
A
OPAMPMC (CC2530/CC2531: 0x61A6. CC2540: 0x61AD) – Operational Amplifier Mode Control
Bit
7:2
1:0
Name
–
MODE
Reset
0000 00
00
R/W
R/W
R/W
Description
Reserved. Always write 0000 00.
Operational amplifier mode
00 and 01: Non-chop mode – Higher offset (~500 µV), but no chopper ripple.
Use in conjunction with Mode 10 if offset cancellation is required. Offset for
these two modes is the opposite of the offset seen in Mode 10.
10: Non-chop mode – Higher offset (~500 µV), but no chopper ripple. Use in
conjunction with Mode 00 or Mode 01 to double sample and correct for the
offset by averaging the two samples.
11: Chop mode – Very low offset (~ 50 µV), and very low noise (1/f noise
shifted to 1 MHz due to chopping), and 1 MHz ripple
OPAMPC (0x62C0) – Operational Amplifier Control
Bit
7:2
1
0
Name
–
CAL
EN
Reset
0000 00
0
0
R/W
R0
W1/R0
R/W
Description
Reserved
Start calibration. Calibration only starts if OPAMPC.EN is 1.
Operational amplifier enable
OPAMPS (0x62C1) – Operational Amplifier Status
Bit
7:1
0
Name
–
CAL_BUSY
Reset
0000 000
0
R/W
R0
R
Description
Reserved
Calibration in progress
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Operational Amplifier
165
Chapter 19
SWRU191D – April 2009 – Revised March 2013
Analog Comparator
The analog comparator (in the CC2530, CC2531, CC240 and CC2541) has the following features:
• Low-power operation
• Wake-up source
Topic
19.1
19.2
166
...........................................................................................................................
Page
Description ..................................................................................................... 167
Register .......................................................................................................... 167
Analog Comparator
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Description
www.ti.com
19.1 Description
The analog comparator is connected to the I/O pins as follows:
• The positive input pin is connected to P0_5.
• The negative input pin is connected to P0_4.
• The output can be read from CMPCTL.OUTPUT.
The comparator pins must be configured as analog pins by setting bits APCFG[5:4] to 1. The
CMPCTL.EN bit is used to enable/disable the comparator. The output from the comparator is connected
internally to the edge detector that controls P0IFG[5]. This makes it possible to associate an I/O interrupt
with a rising/falling edge on the comparator output. When enabled, the comparator remains active while in
power mode 2 or 3. Thus, it is possible to wake up from power mode 2/3 on a rising or falling edge on the
comparator output.
ENB
P0_4
(Pad)
Pad I/O Driver
CMPCTL.EN
–
EN
+
Analog
Comparator
1
Edge Detector
for P0_5
ENB
P0_5
(Pad)
0
Pad I/O Driver
S0385-01
Figure 19-1. Analog Comparator
19.2
Register
This section describes the registers for the analog comparator.
A
CMPCTL (0x62D0) – Analog Comparator Control and Status
Bit
7:2
1
0
Name
–
EN
OUTPUT
Reset
0000 00
0
0
R/W
R0
R/W
R
Description
Reserved
Comparator enable
Comparator output
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Analog Comparator
167
Chapter 20
SWRU191D – April 2009 – Revised March 2013
I2C
The I2C module (in the CC2533 and CC2541) provides an interface between the device and I2Ccompatible devices connected by the two-wire I2C serial bus. External components attached to the I2C bus
serially transmit and/or receive serial data to/from the I2C module through the two-wire I2C interface.
The I2C module features include:
• Compliance with the I2C specification v2.1 (published by Philips Semiconductor)
• 7-bit device addressing modes
• General call
• START/RESTART/STOP
• Multi-master transmitter/receiver mode
• Slave receiver/transmitter mode
• Standard mode up to 100-kbps and fast mode up to 400-kbps support
Figure 20-1 shows the block diagram of the I2C module.
On the CC2533 and CC2541, the I2C module is connected to pins 2 and 3 on the chip and uses the P2
interrupt to the CPU. Pins 2 and 3 can alternatively be controlled as two GPIO pins if they are not used by
the I2C module.
The I2C pins cannot be used to wake the device from PM2/3. To wake up on activity on the I2C, the I2C
bus has to be connected to a normal GPIO in parallel.
Topic
20.1
20.2
168
I2C
...........................................................................................................................
Page
Operation ........................................................................................................ 169
I2C Registers ................................................................................................... 178
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Operation
www.ti.com
Address register (I2CADDR)
Address comparator
XDATA Bus
Data shift register (I2CDATA)
ACK
SDA
Arbitration and synchronization logic
System
clock
Serial clock generator
Control register (I2CCFG)
SCL
Control
logic
Interrupt
P2
interrupt
Status register (I2CSTAT)
Figure 20-1. Block Diagram of the I2C Module
20.1 Operation
The I2C module supports any slave or master I2C-compatible device. Figure 20-2 shows an example of an
I2C bus. Each I2C device is recognized by a unique address and can operate as either a transmitter or a
receiver. A device connected to the I2C bus can be considered as the master or the slave when
performing data transfers. A master initiates a data transfer and generates the clock signal, SCL. Any
device addressed by a master is considered a slave.
I2C data is communicated using the serial data (SDA) pin and the serial clock (SCL) pin. Both SDA and
SCL are bidirectional and must be connected to a positive supply voltage using a pullup resistor.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I2C
169
Operation
www.ti.com
VCC
CC253x
CC2541
Device A
Serial Data (SDA)
Serial Clock (SCL)
Device C
Device B
Figure 20-2. I2C Bus Connection Diagram
20.1.1 I2C Initialization and Reset
The I2C module is enabled by setting the I2CCFG.ENS1 bit. It is then in the not-addressed slave state.
The I2C configuration and state is not retained in power modes PM2 and PM3. It must be reconfigured
after coming out of sleep mode.
The I2C module is not reset when disabled, and retains its internal state until the next time I2CCFG.ENS1
is set.
20.1.2 I2C Serial Data
One clock pulse is generated by the master device for each data bit transferred. The I2C module operates
with byte data. Data is transferred MSB first as shown in Figure 20-3.
The first byte after a START condition consists of a 7-bit slave address and the R/W bit. When R/W = 0,
the master transmits data to a slave. When R/W = 1, the master receives data from a slave. The ACK bit
is sent from the receiver after each byte on the ninth SCL clock.
SDA
MSB
Acknowledgement
Signal From Receiver
Acknowledgement
Signal From Receiver
SCL
1
START
Condition (S)
2
7
8
R/W
9
ACK
1
2
8
9
ACK
STOP
Condition (P)
Figure 20-3. I2C Module Data Transfer
START and STOP conditions are generated by the master and are shown in Figure 20-3. A START
condition is a high-to-low transition on the SDA line while SCL is high. A STOP condition is a low-to-high
transition on the SDA line while SCL is high.
Data on SDA must be stable during the high period of SCL (see Figure 20-4). The state of SDA can only
change when SCL is low, otherwise a START or STOP condition is generated.
170
I2C
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Operation
www.ti.com
Data Line
Stable Data
SDA
SCL
Change of Data Allowed
Figure 20-4. Bit Transfer on I2C Bus
20.1.3 I2C Addressing Modes
The I2C module supports 7-bit addressing mode.
20.1.3.1 7-Bit Addressing
In the 7-bit addressing format (see Figure 20-5), the first byte is the 7-bit slave address and the R/W bit.
The ACK bit is sent from the receiver after each byte.
1
S
1
1
R/W
ACK
7
Slave Address
1
8
Data
1
8
ACK
Data
1
ACK P
Figure 20-5. I2C Module 7-Bit Addressing Format
20.1.3.2 Repeated Start Conditions
The direction of data flow on SDA can be changed by the master, without first stopping a transfer, by
issuing a repeated START condition. This is called a RESTART. After a RESTART is issued, the slave
address is again sent out with the new data direction specified by the R/W bit. The RESTART condition is
shown in Figure 20-6.
1
7
1
S
Slave Address
1
R/W ACK
1
8
1
1
Data
ACK
S
Any
Number
1
7
Slave Address
1
R/W ACK
1
8
1
1
Data
ACK
P
Any Number
Figure 20-6. I2C Module Addressing Format With Repeated START Condition
20.1.4 I2C Module Operating Modes
The I2C module can operate in master transmitter, master receiver, slave transmitter, or slave receiver
mode. The modes are discussed in the following sections.
20.1.4.1 Slave Mode
Initially, the I2C module is configured in receiver mode by setting the I2CCFG.ENS1 bit to receive the I2C
address. Afterwards, transmit and receive operations are controlled automatically, depending on the R/W
bit received, together with the slave address.
The I2C slave address is programmed with the I2CADDR.ADDR bits. The value of the I2CADDR.GC bit
determines whether the slave responds to a general call.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I2C
171
Operation
www.ti.com
When a START condition is detected on the bus, the I2C module receives the transmitted address and
compares it against its own address stored in I2CADDR.ADDR. If the compare is successful, an interrupt is
generated and the I2CCFG.SI bit is set. The same is done for a general call address match if the
I2CADDR.GC bit is set.
20.1.4.1.1 I2C Slave Transmitter Mode
Slave transmitter mode is entered when the slave address transmitted by the master is identical to this
device's own address with a set R/W bit. The slave transmitter shifts the serial data out on SDA with the
clock pulses that are generated by the master device. The slave device does not generate the clock, but it
does hold SCL low while intervention of the CPU is required after a byte has been transmitted.
If the master requests data from the slave, the I2C module is automatically configured as a transmitter, and
I2CCFG.SI is set. The SCL line is held low until the first data to be sent is written into the data buffer
I2CDATA. Then the address is acknowledged and the data is transmitted. After the data is acknowledged
by the master, the bus is stalled during the acknowledge cycle by holding SCL low until new data is written
into I2CDATA. If the master sends a NACK the I2C module returns to the not-addressed slave state.
Table 20-1 provides more details regarding the slave transmitter operation.
Table 20-1. Slave Transmitter Mode
Status
Code
(Value of
I2CSTAT)
0xA8
0xB0
0xB8
0xC0
172 I2C
Application Software Response
Status of the
I2C
To/From I2CDATA
Next Action Taken by I2C Hardware
To I2CCFG
STA
STO
Load data byte
X
0
0
0
Last data byte is transmitted and ACK is received.
or
load data byte
X
0
0
1
Data byte is transmitted; ACK is received.
Arbitration lost Load data byte
in SLA+R/W as
or
master; own
load data byte
SLA+R has
been received;
ACK has been
returned.
X
0
0
0
Last data byte is transmitted and ACK is received.
X
0
0
1
Data byte is transmitted; ACK is received.
Data byte has
been
transmitted;
ACK has been
received.
Load data byte
X
0
0
0
Last data byte is transmitted and ACK is received
or
load data byte
X
0
0
1
Data byte is transmitted; ACK is received.
Data byte has
been
transmitted;
not-ACK has
been received.
No action
0
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general call address
or
no action
0
0
0
1
Switched to not-addressed SLV mode; own SLA or
general call address is recognized.
or
no action
1
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general call address;
START condition is transmitted when the bus
becomes free.
or
no action
1
0
0
1
Switched to not-addressed SLV mode; own SLA or
general-call address is recognized; START
condition is transmitted when the bus becomes
free.
Own SLA+R
has been
received; ACK
has been
returned.
SI
AA
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Operation
www.ti.com
Table 20-1. Slave Transmitter Mode (continued)
Status
Code
(Value of
I2CSTAT)
0xC8
Application Software Response
Status of the
I2C
Last data byte
has been
transmitted;
ACK has been
received.
To/From I2CDATA
Next Action Taken by I2C Hardware
To I2CCFG
STA
STO
No action
0
0
0
SI
0
AA
Switched to not-addressed SLV mode; no
recognition of own SLA or general call address
or
no action
0
0
0
1
Switched to not-addressed SLV mode; own SLA or
general call address is recognized.
or
no action
1
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general call address;
START condition is transmitted when the bus
becomes free.
or
no action
1
0
0
1
Switched to not-addressed SLV mode; own SLA or
general call address is recognized; START
condition is transmitted when the bus becomes
free.
20.1.4.1.2 I2C Slave Receiver Mode
Slave receiver mode is entered when the slave address transmitted by the master is identical to its own
address and a cleared R/W bit is received. In slave receiver mode, serial data bits received on SDA are
shifted in with the clock pulses that are generated by the master device. The slave device does not
generate the clock, but it can hold SCL low if intervention of the CPU is required after a byte has been
received.
If the slave interrupt is triggered from the master, the I2C module is automatically configured as a receiver
and I2CCFG.SI is set. After the first data byte is received, the interrupt flag I2CCFG.SI is set again. The
I2C module automatically acknowledges the received data.
While the I2CCFG.SI flag is set, the bus is stalled by holding SCL low.
When the master generates a STOP condition, the I2CCFG.STO flag is set.
If the master generates a repeated START condition, the I2C state machine returns to its address
reception state.
Table 20-2 provides more details regarding slave receiver operation.
Table 20-2. Slave Receiver Mode
Status
Code
(Value of
I2CSTAT)
0x60
0x68
0x70
Application Software Response
Status of the
I2C
To/From I2CDATA
Next Action Taken by I2C Hardware
To I2CCFG
STA
STO
No action
X
0
0
0
Data byte is received and not-ACK is returned
or
no action
X
0
0
1
Data byte is received and ACK is returned.
Arbitration lost No action
in SLA+R/W as
or
master; own
no action
SLA+W has
been received,
ACK returned.
X
0
0
0
Data byte is received and not-ACK is returned.
X
0
0
1
Data byte is received and ACK is returned.
General-call
address (0x00)
has been
received; ACK
has been
returned
No action
X
0
0
0
Data byte is received and not-ACK is returned.
or
no action
X
0
0
1
Data byte is received and ACK is returned.
Own SLA+W
has been
received; ACK
has been
returned.
SI
AA
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I2C 173
Operation
www.ti.com
Table 20-2. Slave Receiver Mode (continued)
Status
Code
(Value of
I2CSTAT)
0x78
0x80
0x88
0x90
0x98
0xA0
174 I2C
Application Software Response
Status of the
I2C
To/From I2CDATA
Next Action Taken by I2C Hardware
To I2CCFG
STA
STO
Arbitration lost No action
in SLA+R/W as
or
master;
no action
general-call
address has
been received,
ACK returned.
X
0
0
0
Data byte is received and not-ACK is returned.
X
0
0
1
Data byte is received and ACK is returned.
Previously
Read data byte
addressed with
or
own SLV
read data byte
address; DATA
has been
received, ACK
returned.
X
0
0
0
Data byte is received and not-ACK is returned.
X
0
0
1
Data byte is received and ACK is returned.
Previously
addressed with
own SLA;
DATA byte has
been received,
not-ACK
returned.
Read data byte
0
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general-call address
or
read data byte
0
0
0
1
Switched to not-addressed SLV mode; own SLA or
general call address is recognized.
or
read data byte
1
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general-call address;
START condition is transmitted when the bus
becomes free.
or
read data byte
1
0
0
1
Switched to not-addressed SLV mode; own SLA or
general-call address is recognized; START
condition is transmitted when the bus becomes
free.
Previously
Read data byte
addressed with
or
general-call
read data byte
address; DATA
has been
received, ACK
returned.
X
0
0
0
Data byte is received and not-ACK is returned.
X
0
0
1
Data byteis received and ACK is returned.
Previously
addressed with
own SLA;
DATA byte has
been received,
not-ACK
returned.
Read data byte
0
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general-call address
or
read data byte
0
0
0
1
Switched to not-addressed SLV mode; own SLA or
general-call address is recognized.
or
read data byte
1
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general-call address;
START condition is transmitted when the bus
becomes free.
or
read data byte
1
0
0
1
Switched to not-addressed SLV mode; own SLA or
general-call address is recognized; START
condition is transmitted when the bus becomes
free.
No action
0
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general-call address
or
no action
0
0
0
1
Switched to not-addressed SLV mode; own SLA or
general-call address is recognized.
or
no action
1
0
0
0
Switched to not-addressed SLV mode; no
recognition of own SLA or general-call address;
START condition is transmitted when the bus
becomes free.
or
no action
1
0
0
1
Switched to not-addressed SLV mode; own SLA or
general-call address is recognized; START
condition is transmitted when the bus becomes
free.
A STOP
condition or
repeated
START
condition has
been received
while still
addressed as
SLV/REC or
SLV/TRX.
SI
AA
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Operation
www.ti.com
20.1.4.2 Master Mode
The I2C module is configured as an I2C master by setting the I2CCFG.ENS1 and I2CCFG.STA bits. When
the master is part of a multi-master system, its own address must be programmed into the
I2CADDR.ADDR register. The value of the I2CADDR.GC bit determines whether the I2C module responds
to a general call.
20.1.4.2.1 I2C Master Transmitter Mode
To enable master transmit mode, set the I2CCFG.ENS1 and I2CCFG.STA bits. The I2C module then
waits until the I2C bus is free. When the I2C bus is free, it generates a START condition, sends the slave
address, and transfers a transmit direction bit. It then generates an interrupt, and the first byte of data can
be written to the I2CDATA register. The I2C core sends I2CDATA content if arbitration is not lost, and then
generates another interrupt. The I2CSTAT register contains a value of 0x18 or 0x20, depending on the
received ACK bit (see Table 20-3). If a not-ACK is received from the slave, the master must react with
either a repeated START condition or a STOP condition. Setting I2CCFG.STA during transmission causes
a repeated START condition to be transmitted. Setting I2CCFG.STO during transmission causes a STOP
condition to be transmitted and the I2CCFG.STO bit to be reset.
Table 20-3 provides more details regarding the master transmitter operation.
Table 20-3. Master Transmitter Mode
Status
Code
(Value of
I2CSTAT)
Application Software Response
Status of the
I2C
To/From I2CDATA
Next Action Taken by I2C Hardware
To I2CCFG
STA
STO
SI
AA
0x08
A START
condition has
been
transmitted.
Load SLA+W
X
0
0
X
SLA+W is transmitted.
ACK is received.
0x10
A repeated
START
condition has
been
transmitted.
Load SLA+W
X
0
0
X
As for START condition (0x08)
or
load SLA+R
X
0
0
X
SLA+W is transmitted; I2C is switched to MST/REC
mode.
SLA+W has
been
transmitted;
ACK has been
received.
Load data byte
0
0
0
X
Data byte is transmitted; ACK is received.
or
no action
1
0
0
X
Repeated START is transmitted.
or
no action
0
1
0
X
STOP condition is transmitted; STO flag is reset.
or
no action
1
1
0
X
STOP condition followed by a START condition is
transmitted; STO flag is reset.
Load data byte
0
0
0
X
Data byte is transmitted; ACK is received.
or
no action
1
0
0
X
Repeated START is transmitted.
or
no action
0
1
0
X
STOP condition is transmitted; STO flag is reset.
or
no action
1
1
0
X
STOP condition followed by a START condition is
transmitted; STO flag is reset.
Load data byte
0
0
0
X
Data byte is transmitted; ACK is received.
or
no action
1
0
0
X
Repeated START is transmitted.
or
no action
0
1
0
X
STOP condition is transmitted; STO flag is reset.
or
no action
1
1
0
X
STOP condition followed by a START condition is
transmitted; STO flag is reset.
0x18
0x20
0x28
SLA+W has
been
transmitted;
not-ACK has
been received.
Data byte is
transmitted;
ACK is
received.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I2C 175
Operation
www.ti.com
Table 20-3. Master Transmitter Mode (continued)
Status
Code
(Value of
I2CSTAT)
0x30
0x38
Application Software Response
Status of the
I2C
Data byte in
I2CDATA has
been
transmitted.
Arbitration lost
in SLA+R/W or
data bytes
To/From I2CDATA
Next Action Taken by I2C Hardware
To I2CCFG
STA
STO
Data byte
0
0
0
SI
X
AA
Data byte is transmitted; ACK is received.
or
no action
1
0
0
X
Repeated START is transmitted.
or
no action
0
1
0
X
STOP condition is transmitted; STO flag is reset.
or
no action
1
1
0
X
STOP condition followed by a START condition is
transmitted; STO flag is reset.
No action
0
0
0
X
I2C bus is released; not-addressed slave is entered.
or
no action
1
0
0
X
A START condition is transmitted when the bus
becomes free.
20.1.4.2.2 I2C Master Receiver Mode
To enable master receive mode, set the I2CCFG.ENS1 and the I2CCFG.STA bits. The I2C module then
waits until the I2C bus is free. When the I2C bus is free, it generates a START condition, sends the slave
address, and transfers a receive direction bit. It then generates an interrupt, and the first byte is received.
Table 20-4 provides more details regarding the master receiver operation.
Table 20-4. Master Receiver Mode
Status
Code
(Value of
I2CSTAT)
To/From I2CDATA
Next Action Taken by I2C Hardware
To I2CCFG
STA
STO
0x08
A START
condition has
been
transmitted.
Load SLA+R
X
0
0
X
SLA+R is transmitted.
ACK is received.
0x10
A repeated
START
condition has
been
transmitted.
Load SLA+R
X
0
0
X
As above
or
load SLA+W
X
0
0
X
SLA+W is transmitted; I2C is switched to MST/TRX
mode.
Arbitration lost
in not-ACK bit.
No action
0
0
0
X
I2C bus is released; I2C enters slave mode.
or
no action
1
0
0
X
A start condition is transmitted when the bus
becomes free.
SLA+R has
been
transmitted;
ACK has been
received.
No action
0
0
0
0
Data byte is received; not-ACK is returned.
or
no action
0
0
0
1
Data byte is received; ACK is returned.
SLA+R has
been
transmitted;
not-ACK has
been received.
No action
1
0
0
X
Repeated START condition is transmitted.
or
no action
0
1
0
X
STOP condition is transmitted; STO flag is reset.
or
no action
1
1
0
X
STOP condition followed by a START condition is
transmitted; STO flag is reset.
Data byte has
been received;
ACK has been
returned.
Read data byte
0
0
0
0
Data byte is received; not-ACK is returned.
or
read data byte
0
0
0
1
Data byte is received; ACK is returned
Data byte has
been received;
not-ACK has
been returned.
Read data byte
1
0
0
X
Repeated START condition is transmitted.
or
read data byte
0
1
0
X
STOP condition is transmitted; STO flag is reset.
or
read data byte
1
1
0
X
STOP condition followed by a START condition is
transmitted; STO flag is reset.
0x38
0x40
0x48
0x50
0x58
176 I2C
Application Software Response
Status of the
I2C
SI
AA
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Operation
www.ti.com
20.1.4.3 Arbitration
If two or more master transmitters simultaneously start a transmission on the bus, an arbitration procedure
is invoked. Figure 20-7 shows the arbitration procedure between two devices. The arbitration procedure
uses the data presented on SDA by the competing transmitters. The first master transmitter that generates
a logic high is overruled by the opposing master generating a logic low. The arbitration procedure gives
priority to the device that transmits the serial data stream with the lowest binary value. The master
transmitter that lost arbitration switches to the slave receiver mode. If two or more devices send identical
first bytes, arbitration continues on the subsequent bytes.
Bus Line
SCL
Device #1 Lost Arbitration
and Switches Off
n
Data From
Device #1
1
0
Data From
Device #2
0
0
1
1
0
Bus Line
SDA
1
0
0
1
1
1
Figure 20-7. Arbitration Procedure Between Two Master Transmitters
20.1.5 I2C Clock Generation and Synchronization
The I2C clock SCL is provided by the master on the I2C bus. When the I2C module is in master mode, the
serial clock generator generates the SCL clock from the system clock. The serial clock generator is
switched off when the I2C module is in slave mode.
The frequency of the SCL is determined by the system clock frequency, and the division factor given by
the I2CCFG.CRx bits. Example frequencies for a 32-MHz system clock are given in the I2CCFG register
description.
During the arbitration procedure, the clocks from the different masters must be synchronized. A device
that first generates a low period on SCL overrules the other devices, forcing them to start their own low
periods. SCL is then held low by the device with the longest low period. The other devices must wait for
SCL to be released before starting their high periods. Figure 20-8 shows the clock synchronization. This
allows a slow slave to slow down a fast master.
Wait
State
Start HIGH
Period
SCL From
Device #1
SCL From
Device #2
Bus Line
SCL
Figure 20-8. Synchronization of Two I2C Clock Generators During Arbitration
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I2C
177
I2C Registers
www.ti.com
20.1.6 Bus Error
When an incorrect format of a frame is detected, a bus error condition is entered. The cause is that a
START or STOP condition was detected during transfer of an address, data, or an acknowledge bit. When
a bus error condition is entered, an interrupt is requested. The core leaves the bus error state when the
I2CCFG.STO flag is set and the interrupt request is cleared. It goes into the slave mode and the
I2CCFG.STO flag is automatically reset. The SDA and SCL lines are released (the STOP condition is not
transmitted).
Table 20-5. Miscellaneous States
Status
Code
(Value of
I2CSTAT.
STAC)
0x00
Application Software Response
Status of the
I2C
Bus error
during MST or
selected slave
modes
To I2CCFG
To/From I2CDATA
No action
STA
STO
0
1
SI
0
Next Action Taken by I2C Hardware
AA
X
Only the internal hardware is affected in the MST or
addressed SLV modes. In all cases, the bus is
released and I2C is switched to the not-addressed
SLV mode. The Sto flag is reset.
20.1.7 I2C Interrupt
The I2C module has an interrupt line to the CPU to signal that it requires service. The I2C module uses
interrupt #6, which is also shared with Port 2 inputs; hence, the interrupt routine must also handle Port 2
interrupts if they are enabled.
For an interrupt request to be generated, IEN2.P2IE must be set to 1. When an interrupt request has
been generated, the CPU starts executing the ISR if there are no higher-priority interrupts pending.
An interrupt is generated from the I2C module when one of the 26 out of 27 possible I2C component states
is entered. The only state that does not cause an interrupt to be generated is state 0xF8, which indicates
that no relevant state information is available. The corresponding I2CCFG.SI flag must be cleared by
software at the end of the ISR.
20.1.8 I2C Pins
The SCL and SDA pins of the I2C module are connected to pins 2 and 3, respectively, on the
CC2533/CC2541. These pins are pulled up during reset to avoid floating pins. After reset, they are
controlled by the I2C module and use an internal pullup resistor of 20 kΩ to hold bus signals high. If these
pins are not to be used for I2C, they can be used as GPIO by setting the I2CWC.OVR bit. In this mode,
pins 2 and 3 can be set up as outputs, as inputs with optional pullup, or as 4-mA drive-strength outputs
like the other GPIO pads on the device by using the configuration bits in I2CWC. Their values are read or
controlled using the I2CIO register. These pins cannot be configured to generate GPIO interrupts.
20.2 I2C Registers
This section describes all I2C registers used for control and status of the I2C module.
The registers return to their reset values when the chip enters PM2 or PM3.
178
I2C
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I2C Registers
www.ti.com
I2CCFG (0x6230) – I2C Control
Bit
7
6
Name
CR2
ENS1
Reset
0
0
R/W
R/W
R/W
5
4
STA
STO
0
0
R/W
R/W1
3
2
SI
AA
0
0
R/W0
R/W
1
0
CR1
CR0
0
0
R/W
R/W
Description
Clock rate bit 2
Enable bit.
0:
I2C module disabled.
SCL and SDA are set to high impedance inputs. The inputs are ignored by the I2C
module.
Note that setting ENS1 = 0 disables the I2C module but does not reset its state.
1:
I2C module enabled.
START flag. When set, HW detects when I2C is free and generates a START condition.
STOP flag. When set and in master mode, a STOP condition is transmitted on the I2C bus.
HW is cleared when transmit has completed successfully.
Interrupt flag
Assert acknowledge flag for the I2C module.
When set (AA = 1), an acknowledge is returned when:
● Slave address is recognized
● General call is recognized, when the I2C module is enabled
● Data byte received while in master/slave receive mode
When not set (AA = 0), an acknowledge is returned when:
● Data byte is received while in master/slave receive mode
Clock rate bit 1
Clock rate bit 0
Table 20-6. Clock Rates Defined at 32 MHz
CR2
CR1
CR0
Bit Frequency
(kHz)
Clock Divided by
0
0
0
123
256
0
0
1
144
244
0
1
0
165
192
0
1
1
197
160
1
0
0
33
960
1
0
1
267
120
1
1
0
533
60
1
1
1
Reserved
N/A
I2CSTAT (0x6231) – I2C Status
Bit
7:3
Name
STAC
Reset
1111 1
R/W
R
Description
Status code. Contains the state of the I2C core. 27 states are defined: 0 to 25 and 31.
Interrupt is only requested when in states 0 to 25.
The value 0xF8 indicates that there is no relevant state information available and that
I2CCFG.SI = 0.
2:0
–
000
R0
Reserved
I2CDATA (0x6232) – I2C Data
Bit
7:0
Name
SD
Reset
R/W
0000 00 R/W
00
Description
Serial data in/out (MSB is bit 7, LSB is bit 0). Contains data byte to be transmitted or byte
which has just been received. Can be read or written while not in the process of shifting a
byte. The register is not shadowed or double buffered, so it should only be accessed upon an
interrupt.
I2CADDR (0x6233) – I2C Own Slave Address
Bit
7:1
Name
ADDR
0
GC
Reset
R/W
0000 00 R/W
0
0
R/W
Description
Own slave address
General-call address acknowledge. If set, the general-call address is recognized.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
I2C
179
I2C Registers
www.ti.com
I2CWC (0x6234) – Wrapper control
Bit
7
Name
OVR
Reset
0
R/W
R/W
6:4
3
2
1
0
–
SCLPUE
SDAPUE
SCLOE
SDAOE
000
1
1
0
0
R0
R/W
R/W
R/W
R/W
Description
Override enable:
0: I2C functionality (ignore other bits in this register)
1: GPIO functionality.Reserved
SCL pin pullup enable
SDA pin pullup enable
SCL pin output enable
SDA pin output enable
I2CIO (0x6235) – GPIO
Bit
7:2
1
Name
–
SCLD
Reset
000
0
R/W
R0
R/W
0
SDAD
0
R/W
180
I2C
Description
Reserved
SCL data value
When I2CWC.SCLOE is set, reading SCLD reads the output register, not the pin. When
I2CWC.SCLOE is cleared, reading SCLD reads the pin. Writing SCLD writes to the output
register.
SDA data value
When I2CWC.SDAOE is set, reading SDAD reads the output register, not the pin. When
I2CWC.SDAOE is cleared, reading SDAD reads the pin. Writing SDAD writes to the output
register.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 21
SWRU191D – April 2009 – Revised March 2013
USB Controller
This section focuses on describing the functionality of the USB controller (in the CC2531/CC2540 only),
and it is assumed that the reader has a good understanding of USB and is familiar with the terms and
concepts used. See the Universal Serial Bus Specification for details ([8], Appendix C).
Standard USB nomenclature is used regarding IN and OUT. I.e., IN is always into the host (PC) and OUT
is out of the host.
Topic
...........................................................................................................................
21.1
21.2
21.3
21.4
21.5
21.6
21.7
21.8
21.9
21.10
21.11
21.12
USB Introduction .............................................................................................
USB Enable .....................................................................................................
48-MHz USB PLL .............................................................................................
USB Interrupts ................................................................................................
Endpoint 0 ......................................................................................................
Endpoint-0 Interrupts .......................................................................................
Endpoints 1–5 .................................................................................................
DMA ...............................................................................................................
USB Reset ......................................................................................................
Suspend and Resume .....................................................................................
Remote Wake-Up ............................................................................................
USB Registers ................................................................................................
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Controller
Page
182
182
182
183
183
183
185
189
189
189
190
190
181
USB Introduction
www.ti.com
21.1 USB Introduction
The USB controller monitors the USB for relevant activity and handles packet transfers.
Appropriate response to USB interrupts and loading/unloading of packets into/from endpoint FIFOs is the
responsibility of the firmware. The firmware must be able to reply correctly to all standard requests from
the USB host and work according to the protocol implemented in the driver on the PC.
The USB controller has the following features:
• Full-speed operation (up to 12 Mbps)
• Five endpoints (in addition to endpoint 0) that can be used as IN, OUT, or IN/OUT and can be
configured as bulk/interrupt or isochronous.
• 1 KB SRAM FIFO available for storing USB packets
• Endpoints supporting packet sizes from 8–512 bytes
• Support for double buffering of USB packets
Figure 21-1 shows a block diagram of the USB controller. The USB PHY is the physical interface with
input and output drivers. The USB SIE is the serial-interface engine, which controls the packet transfer
to/from the endpoints. The USB controller is connected to the rest of the system through the memory
arbiter.
USB Controller
EP0
EP1
DP
EP2
USB PHY
USB SIE
Memory
Arbiter
EP3
DM
EP4
EP5
1 KB
SRAM
(FIFOs)
B0305-01
Figure 21-1. USB Controller Block Diagram
21.2 USB Enable
The USB is enabled by setting USBCTRL.USB_EN to 1. Setting USBCTRL.USB_EN to 0 resets the USB
controller.
21.3 48-MHz USB PLL
The 48-MHz internal USB PLL must be powered up and stable for the USB controller to operate correctly.
It is important that the crystal oscillator is selected as souce and is stable before the USB PLL is enabled.
The USB PLL is enabled by setting the USBCTRL.PLL_EN bit and waiting for the USBCTRL.PLL_LOCKED
status flag to go high. When the PLL has locked, it is safe to use the USB controller.
Note: The PLL must be disabled before exiting active mode and re-enabled after entering active mode.
182
USB Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Interrupts
www.ti.com
21.4 USB Interrupts
There are three interrupt flag registers with associated interrupt-enable mask registers.
Table 21-1. USB Interrupt Flags Interrupt-Enable Mask Registers
Interrupt Flag
Description
Associated Interrupt
Enable Mask Register
USBCIF
Contains flags for common USB interrupts
USBCIE
USBIIF
Contains interrupt flags for endpoint 0 and all the IN endpoints
USBIIE
USBOIF
Contains interrupt flags for all OUT endpoints
USBOIE
Note: All interrupts except SOF and suspend are initially enabled after reset.
The USB controller uses interrupt #6 for USB interrupts. This interrupt number is shared with Port 2 inputs;
hence, the interrupt routine must also handle Port 2 interrupts if they are enabled. For an interrupt request
to be generated, IEN2.P2IE must be set to 1 together with the desired interrupt enable bits from the
USBCIE, USBIIE, and USBOIE registers. When an interrupt request has been generated, the CPU starts
executing the ISR if there are no higher-priority interrupts pending. The interrupt routine should read all the
interrupt flag registers and take action depending on the status of the flags. The interrupt flag registers are
cleared when they are read, and the status of the individual interrupt flags should therefore be saved in
memory (typically in a local variable on the stack) to allow them to be accessed multiple times.
At the end of the ISR, after the interrupt flags have been read, the interrupt flags should be cleared to
allow for new USB and P2 interrupts to be detected. The Port 2 interrupt status flags in the P2IFG register
should be cleared prior to clearing IRCON2.P2IF.
When waking up from suspend (typically in PM1), the USB D+ interrupt flag, P2IFG.DPIF, is set. The D+
interrupt flag indicates that there has been a falling edge on the D+ USB data pin. This is a resume event.
21.5 Endpoint 0
Endpoint 0 (EP0) is a bidirectional control endpoint, and during the enumeration phase all communication
is performed across this endpoint. Before the USBADDR register has been set to a value other than 0, the
USB controller is only able to communicate through endpoint 0. Setting the USBADDR register to a value
between 1 and 127 brings the USB function out of the default state in the enumeration phase and into the
address state. All configured endpoints are then available for the application.
The EP0 FIFO is only used as either IN or OUT, and double buffering is not provided for endpoint 0. The
maximum packet size for endpoint 0 is fixed at 32 bytes.
Endpoint 0 is controlled through the USBCS0 register by setting the USBINDEX register to 0. The USBCNT0
register contains the number of bytes received.
21.6 Endpoint-0 Interrupts
The following events may generate an EP0 interrupt request:
• A data packet has been received (USBCS0.OUTPKT_RDY = 1)
•
•
•
•
A data packet that was loaded into the EP0 FIFO has been sent to the USB host.
(USBCS0.INPKT_RDY should be set to 1 when a new packet is ready to be transferred. This bit is
cleared by hardware when the data packet has been sent.)
An IN transaction has been completed (the interrupt is generated during the status stage of the
transaction).
A STALL has been sent (USBCS0.SENT_STALL =
1)
A control transfer ends due to a premature end-of-control transfer (USBCS0.SETUP_END = 1)
Any of these events causes USBIIF.EP0IF to be asserted, regardless of the status of the EP0 interrupt
mask bit USBIIE.EP0IE. If the EP0 interrupt mask bit is set to 1, the CPU interrupt flag IRCON2.P2IF is
also asserted. An interrupt request is only generated if IEN2.P2IE and USBIIE.EP0IE are both set to 1.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Controller
183
Endpoint-0 Interrupts
www.ti.com
21.6.1 Error Conditions
When a protocol error occurs, the USB controller sends a STALL handshake. The USBCS0.SENT_STALL
bit is asserted, and an interrupt request is generated if the endpoint-0 interrupt is properly enabled. A
protocol error can be any of the following:
• An OUT token is received after USBCS0.DATA_END has been set to complete the OUT data stage (the
host tries to send more data than expected).
• An IN token is received after USBCS0.DATA_END has been set to complete the IN data stage (the host
tries to receive more data than expected).
• The USB host tries to send a packet that exceeds the maximum packet size during the OUT data
stage.
• The size of the DATA1 packet received during the status stage is not 0.
The firmware can also terminate the current transaction by setting the USBCS0.SEND_STALL bit to 1. The
USB controller then sends a STALL handshake in response to the next request from the USB host.
If an EP0 interrupt is caused by the assertion of the USBCS0.SENT_STALL bit, this bit should be deasserted, and firmware should consider the transfer as aborted (and consequently free the memory
buffers, etc.).
If EP0 receives an unexpected token during the data stage, the USBCS0.SETUP_END bit is asserted, and
an EP0 interrupt is generated (if enabled properly). EP0 then switches to the IDLE state. Firmware should
then set the USBCS0.CLR_SETUP_END bit to 1 and abort the current transfer. If USBCS0.OUTPKT_RDY is
asserted, this indicates that another setup packet has been received that firmware should process.
21.6.2 SETUP Transactions (IDLE State)
The control transfer consists of two or three stages of transactions (setup – data – status or setup –
status). The first transaction is a setup transaction. A successful setup transaction comprises three
sequential packets (a token packet, a data packet, and a handshake packet), where the data field
(payload) of the data packet is exactly 8 bytes long and is referred to as the setup packet. In the setup
stage of a control transfer, EP0 is in the IDLE state. The USB controller rejects the data packet if the
setup packet is not 8 bytes. Also, the USB controller examines the contents of the setup packet to
determine whether or not there is a data stage in the control transfer. If there is a data stage, EP0
switches state to TX (IN transaction) or RX (OUT transaction) when the USBCS0.CLR_OUTPKT_RDY bit is
set to 1 (if USBCS0.DATA_END = 0).
When a packet is received, the USBCS0.OUTPKT_RDY bit is asserted and an interrupt request is
generated (EP0 interrupt) if the interrupt has been enabled. Firmware should perform the following when a
setup packet has been received:
1. Unload the setup packet from the EP0 FIFO
2. Examine the contents and perform the appropriate operations
3. Set the USBCS0.CLR_OUTPKT_RDY bit to 1. This denotes the end of the setup stage. If the control
transfer has no data stage, the USBCS0.DATA_END bit must also be set. If there is no data stage, the
USB controller stays in the IDLE state.
21.6.3 IN Transactions (TX State)
If the control transfer requires data to be sent to the host, the setup stage is followed by one or more IN
transactions in the data stage. In this case, the USB controller is in the TX state and only accepts IN
tokens. A successful IN transaction comprises two or three sequential packets (a token packet, a data
packet, and a handshake packet (1)). If more than 32 bytes (maximum packet size) is to be sent, the data
must be split into a number of 32-byte packets followed by a residual packet. If the number of bytes to
send is a multiple of 32, the residual packet is a zero-length data packet, because a packet size less than
32 bytes denotes the end of the transfer.
(1)
184
For isochronous transfers there would not be a handshake packet from the host.
USB Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Endpoints 1–5
www.ti.com
Firmware should load the EP0 FIFO with the first data packet and set the USBCS0.INPKT_RDY bit as
soon as possible after the USBCS0.CLR_OUTPKT_RDY bit has been set. The USBCS0.INPKT_RDY is
cleared and an EP0 interrupt is generated when the data packet has been sent. Firmware might then load
more data packets as necessary. An EP0 interrupt is generated for each packet sent. Firmware must set
USBCS0.DATA_END in addition to USBCS0.INPKT_RDY when the last data packet has been loaded. This
starts the status stage of the control transfer.
EP0 switches to the IDLE state when the status stage has completed. The status stage may fail if the
USBCS0.SEND_STALL bit is set to 1. The USBCS0.SENT_STALL bit is then asserted, and an EP0
interrupt is generated.
If USBCS0.INPKT_RDY is not set when receiving an IN token, the USB controller replies with a NAK to
indicate that the endpoint is working, but temporarily has no data to send.
21.6.4 OUT Transactions (RX State)
If the control transfer requires data to be received from the host, the setup stage is followed by one or
more OUT transactions in the data stage. In this case, the USB controller is in the RX state and only
accepts OUT tokens. A successful OUT transaction comprises two or three sequential packets (a token
packet, a data packet, and a handshake packet (2)). If more than 32 bytes (maximum packet size) is to be
received, the data must be split into a number of 32-byte packets followed by a residual packet. If the
number of bytes to receive is a multiple of 32, the residual packet is a zero-length data packet, because a
data packet with payload less than 32 bytes denotes the end of the transfer.
The USBCS0.OUTPKT_RDY bit is set and an EP0 interrupt is generated when a data packet has been
received. The firmware should set USBCS0.CLR_OUTPKT_RDY when the data packet has been unloaded
from the EP0 FIFO. When the last data packet has been received (packet size less than 32 bytes)
firmware should also set the USBCS0.DATA_END bit. This starts the status stage of the control transfer.
The size of the data packet is kept in the USBCNT0 registers. Note that this value is only valid when
USBCS0.OUTPKT_RDY = 1.
EP0 switches to the IDLE state when the status stage has completed. The status stage may fail if the
DATA1 packet received is not a zero-length data packet or if the USBCS0.SEND_STALL bit is set to 1. The
USBCS0.SENT_STALL bit then is asserted and an EP0 interrupt is generated.
21.7 Endpoints 1–5
Each endpoint can be used as an IN only, an OUT only, or IN/OUT. For an IN/OUT endpoint, there are
basically two endpoints, an IN endpoint and an OUT endpoint associated with the endpoint number.
Configuration and control of IN endpoints is performed through the USBCSIL and USBCSIH registers. The
USBCSOL and USBCSOH registers are used to configure and control OUT endpoints. Each IN and OUT
endpoint can be configured as either an isochronous (USBCSIH.ISO = 1 and/or USBCSOH.ISO = 1) or
bulk/interrupt (USBCSIH.ISO = 0 and/or USBCSOH.ISO = 0) endpoint. Bulk and interrupt endpoints are
handled identically by the USB controller but have different properties from a firmware perspective.
The USBINDEX register must have the value of the endpoint number before the indexed endpoint registers
are accessed.
21.7.1 FIFO Management
Each endpoint has a certain number of FIFO memory bytes available for incoming and outgoing data
packets. Table 21-2 shows the FIFO size for endpoints 1–5. The firmware is responsible for setting the
USBMAXI and USBMAXO registers correctly for each endpoint to prevent data from being overwritten.
When both the IN and OUT endpoints of an endpoint number do not use double buffering, the sum of
USBMAXI and USBMAXO must not exceed the FIFO size for the endpoint. Figure 21-2 a) shows how the IN
and OUT FIFO memory for an endpoint is organized with single buffering. The IN FIFO grows down from
the top of the endpoint memory region, whereas the OUT FIFO grows up from the bottom of the endpoint
memory region.
(2)
For isochronous transfers, there is no handshake packet from the device.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Controller
185
Endpoints 1–5
www.ti.com
When the IN or OUT endpoint of an endpoint number uses double buffering, the sum of USBMAXI and
USBMAXO must not exceed half the FIFO size for the endpoint. Figure 21-2 b) illustrates the IN and OUT
FIFO memory for an endpoint that uses double buffering. Notice that the second OUT buffer starts from
the middle of the memory region and grows upwards. The second IN buffer also starts from the middle of
the memory region but grows downwards.
To configure an endpoint as IN-only, set USBMAXO to 0, and to configure an endpoint as OUT-only, set
USBMAXI to 0.
For unused endpoints, both USBMAXO and USBMAXI should be set to 0.
Table 21-2. FIFO Sizes for EP 1–5
EP Number
FIFO Size (in Bytes)
1
32
2
64
3
128
4
256
5
512
0
0
IN FIFO
(Buffer 1)
IN FIFO
USBMAXI-1
USBMAXO-1
USBMAXI-1
OUT FIFO
(Buffer 2)
0
0
IN FIFO
(Buffer 2)
USBMAXI -1
USBMAXO-1
USBMAXO-1
OUT FIFO
(Buffer 1)
OUT FIFO
0
0
a) Single Buffering
b) Double Buffering
M0106-02
Figure 21-2. IN/OUT FIFOs
21.7.2 Double Buffering
To enable faster transfer and reduce the need for retransmissions, double buffering can be used. This
allows two packets to be buffered in the FIFO in each direction. This is highly recommended for
isochronous endpoints, which are expected to transfer one data packet every USB frame without any
retransmission. For an isochronous endpoint, one data packet is sent/received every USB frame.
However, the data packet may be sent/received at any time during the USB frame period, and there is a
chance that two data packets may be sent/received at a few-microseconds interval. For isochronous
endpoints, an incoming packet is lost if there is no buffer available, and a zero-length data packet is sent if
there is no data packet ready for transmission when the USB host requests data. Double buffering is not
as critical for bulk and interrupt endpoints as it is for isochronous endpoints, because packets are not lost.
Double buffering, however, may improve the effective data rate for bulk endpoints.
To enable double buffering for an IN endpoint, USBCSIH.IN_DBL_BUF must be set to 1. To enable
double buffering for an OUT endpoint, set USBCSOH.OUT_DBL_BUF to 1.
186
USB Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Endpoints 1–5
www.ti.com
21.7.3 FIFO Access
The endpoint FIFOs are accessed by reading and writing to the registers USBF0–USBF6. Writing to a
register causes the byte written to be inserted into the IN FIFO. Reading a register causes the next byte in
the OUT FIFO to be extracted and the value of this byte to be returned.
When a data packet has been written to an IN FIFO, the USBCSIL.INPKT_RDY bit must be set to 1. If
double buffering is enabled, the USBCSIL.INPKT_RDY bit is cleared immediately after it has been written,
and another data packet can be loaded. This does not generate an IN endpoint interrupt, because an
interrupt is only generated when a packet has been sent. When double buffering is used, firmware should
check the status of the USBCSIL.PKT_PRESENT bit before writing to the IN FIFO. If this bit is 0, two data
packets can be written. Double-buffered isochronous endpoints should only load two packets the first time
the IN FIFO is loaded. After that, one packet is loaded for every USB frame. To send a zero-length data
packet, USBCSIL.INPKT_RDY should be set to 1 without loading a data packet into the IN FIFO.
A data packet can be read from the OUT FIFO when the USBCSOL.OUTPKT_RDY bit is 1. An interrupt is
generated when this occurs, if enabled. The size of the data packet is kept in the USBCNTH:USBCNTL
registers. Note that this value is only valid when USBCSOL.OUTPKT_RDY = 1. When the data packet has
been read from the OUT FIFO, the USBCSOL.OUTPKT_RDY bit must be cleared. If double buffering is
enabled, there may be two data packets in the FIFO. If another data packet is ready when the
USBCSOL.OUTPKT_RDY bit is cleared, the USBCSOL.OUTPKT_RDY bit is asserted immediately, and an
interrupt is generated (if enabled) to signal that a new data packet has been received. The
USBCSOL.FIFO_FULL bit is set when there are two data packets in the OUT FIFO.
The AutoClear feature is supported for OUT endpoints. When enabled, the USBCSOL.OUTPKT_RDY bit is
cleared automatically when USBMAXO bytes have been read from the OUT FIFO. The AutoClear feature is
enabled by setting USBCSOH.AUTOCLEAR = 1. The AutoClear feature can be used to reduce the time the
data packet occupies the OUT FIFO buffer and is typically used for bulk endpoints.
A complementary AutoSet feature is supported for IN endpoints. When enabled, the
USBCSIL.INPKT_RDY bit is set automatically when USBMAXI bytes have been written to the IN FIFO.
The AutoSet feature is enabled by setting USBCSIH.AUTOSET =
1. The AutoSet feature can reduce the overall time it takes to send a data packet and is
typically used for bulk endpoints.
21.7.4 Endpoint 1–5 Interupts
The following events may generate an IN EPx interrupt request (x indicates the endpoint number):
• A data packet that was loaded into the IN FIFO has been sent to the USB host.
(USBCSIL.INPKT_RDY should be set to 1 when a new packet is ready to be transferred. This bit is
cleared by hardware when the data packet has been sent.)
• A STALL has been sent (USBCSIL.SENT_STALL
= 1). Only bulk/interrupt endpoints can be stalled
• The IN FIFO is flushed due to the USBCSIH.FLUSH_PACKET bit being set to 1.
Any of these events causes USBIIF.INEPxIF to be asserted, regardless of the status of the IN EPx
interrupt mask bit USBIIE.INEPxIE. If the IN EPx interrupt mask bit is set to 1, the CPU interrupt flag
IRCON2.P2IF is also asserted. An interrupt request is only generated if IEN2.P2IE and
USBIIE.INEPxIE are both set to 1. The x in the register name refers to the endpoint number, 1–5)
The following events may generate an OUT EPx interrupt request:
• A data packet has been received (USBCSOL.OUTPKT_RDY = 1).
• A STALL has been sent (USBCSIL.SENT_STALL
= 1). Only bulk/interrupt endpoints can be stalled.
Any of these events causes USBOIF.OUTEPxIF to be asserted, regardless of the status of the OUT EPx
interrupt mask bit USBOIE.OUTEPxIE. If the OUT EPx interrupt mask bit is set to 1, the CPU interrupt flag
IRCON2.P2IF is also asserted. An interrupt request is only generated if IEN2.P2IE and
USBOIE.OUTEPxIE are both set to 1.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Controller
187
Endpoints 1–5
www.ti.com
21.7.5 Bulk/Interrupt IN Endpoint
Interrupt IN transfers occur at regular intervals, whereas bulk IN transfers use available bandwidth not
allocated to isochronous, interrupt, or control transfers.
Interrupt IN endpoints may set the USBCSIH.FORCE_DATA_TOG bit. When this bit is set, the data toggle
bit is continuously toggled, regardless of whether an ACK was received or not. This feature is typically
used by interrupt IN endpoints that are used to communicate rate feedback for isochronous endpoints.
A bulk/interrupt IN endpoint can be stalled by setting the USBCSIL.SEND_STALL bit to 1. When the
endpoint is stalled, the USB controller responds with a STALL handshake to IN tokens. The
USBCSIL.SENT_STALL bit is then set, and an interrupt is generated, if enabled.
A bulk transfer longer than the maximum packet size is performed by splitting the transfer into a number of
data packets of maximum size followed by a smaller data packet containing the remaining bytes. If the
transfer length is a multiple of the maximum packet size, a zero-length data packet is sent last. This
means that a packet with a size less than the maximum packet size denotes the end of the transfer. The
AutoSet feature can be useful in this case, because many data packets are of maximum size.
21.7.6 Isochronous IN Endpoint
An isochronous IN endpoint is used to transfer periodic data from the USB controller to the host (one data
packet every USB frame).
If there is no data packet loaded in the IN FIFO when the USB host requests data, the USB controller
sends a zero-length data packet, and the USBCSIL.UNDERRUN bit is asserted.
Double buffering requires that a data packet is loaded into the IN FIFO during the frame preceding the
frame where it should be sent. If the first data packet is loaded before an IN token is received, the data
packet is sent during the same frame as it was loaded and hence violates the double-buffering strategy.
Thus, when double buffering is used, the USBPOW.ISO_WAIT_SOF bit should be set to 1 to avoid this.
Setting this bit ensures that a loaded data packet is not sent until the next SOF token has been received.
The AutoSet feature typically is not used for isochronous endpoints, because the packet size increases or
decreases from frame to frame.
21.7.7 Bulk/Interrupt OUT Endpoint
Interrupt OUT transfers occur at regular intervals, whereas bulk OUT transfers use available bandwidth
not allocated to isochronous, interrupt, or control transfers.
A bulk/interrupt OUT endpoint can be stalled by setting the USBCSOL.SEND_STALL bit to 1. When the
endpoint is stalled, the USB controller responds with a STALL handshake when the host is done sending
the data packet. The data packet is discarded and is not placed in the OUT FIFO. The USB controller
asserts the USBCSOL.SENT_STALL bit when the STALL handshake is sent and generates an interrupt
request if the OUT endpoint interrupt is enabled.
As the AutoSet feature is useful for bulk IN endpoints, the AutoClear feature is useful for OUT endpoints,
because many packets are of maximum size.
21.7.8 Isochronous OUT Endpoint
An isochronous OUT endpoint is used to transfer periodic data from the host to the USB controller (one
data packet every USB frame).
If there is no buffer available when a data packet is being received, the USBCSOL.OVERRUN bit is asserted
and the packet data is lost. Firmware can reduce the chance for this to happen by using double buffering
and using DMA to unload data packets effectively.
An isochronous data packet in the OUT FIFO may have bit errors. The hardware detects this condition
and sets USBCSOL.DATA_ERROR. Firmware should therefore always check this bit when unloading a data
packet.
The AutoClear feature typically is not used for isochronous endpoints, because the packet size increases
or decreases from frame to frame.
188
USB Controller
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
DMA
www.ti.com
21.8 DMA
DMA should be used to fill the IN endpoint FIFOs and empty the OUT endpoint FIFOs. Using DMA
improves the read/write performance significantly compared to using the CPU. It is therefore highly
recommended to use DMA unless timing is not critical or only a few bytes are to be transferred.
There are no DMA triggers for the USB controller, meaning that DMA transfers must be triggered by
firmware.
Byte-size transfer should be used.
21.9 USB Reset
When reset signaling is detected on the bus, the USBCIF.RSTIF flag is asserted. If USBCIE.RSTIE is
enabled, IRCON2.P2IF is also asserted, and an interrupt request is generated if IEN2.P2IE = 1. The
firmware should take appropriate action when a USB reset occurs. A USB reset should place the device in
the default state, where it only responds to address 0 (the default address). One or more resets normally
take place during the enumeration phase, immediately after the USB cable is connected.
The following actions are performed by the USB controller when a USB reset occurs:
• USBADDR is set to 0.
• USBINDEX is set to 0.
•
•
All endpoint FIFOs are flushed.
USBMAXI, USBCS0, USBCSIL, USBCSIH, USBMAXO, USBCSOL, USBCSOH, USBCNT0, USBCNTL, and
USBCNTH are cleared.
•
•
All interrupts, except SOF and suspend, are enabled.
An interrupt request is generated (if IEN2.P2IE = 1 and USBCIE.RSTIE =
1).
Firmware should close all pipes and wait for a new enumeration phase when USB reset is detected.
21.10 Suspend and Resume
The USB controller asserts USBCIF.SUSPENDIF and enters suspend mode when the USB has been
continuously idle for 3 ms, provided that USBPOW.SUSPEND_EN =
1. IRCON2.P2IF is asserted if USBCIE.SUSPENDIE is enabled, and an interrupt request is
generated if IEN2.P2IE = 1.
While in suspend mode, only limited current can be sourced from the USB. See the USB 2.0 Specification
[3] for details about this. To be able to meet the suspend-current requirement, the device should be taken
down to PM1 when suspend is detected. The device should not enter PM2 or PM3, because this resets
the USB controller. Before entering PM1, the 48-MHz USB PLL must be turned off. This is done by setting
USBCTRL.PLL_EN to 0 and waiting for USBCTRL.PLL_LOCKED to be cleared.
Any valid nonidle signaling on the USB causes USBCIF.RESUMEIF to be asserted and an interrupt
request to be generated, and wakes up the system if the USB resume interrupt is enabled.
When the system wakes up (enters active mode) from suspend, no USB registers except USBCTRL can
be accessed before the 48-MHz USB PLL has been activated. This is done by setting USBCTRL.PLL_EN
to 1 and waiting until USBCTRL.PLL_LOCKED is set.
A USB reset also wakes up the system from suspend. A USB resume interrupt request is generated if the
interrupt is enabled, but the USBCIF.RSTIF interrupt flag is set instead of the USBCIF.RESUMEIF
interrupt flag.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Controller
189
Remote Wake-Up
www.ti.com
21.11 Remote Wake-Up
The USB controller can resume from suspend by signaling resume to the USB hub. Resume is performed
by setting USBPOW.RESUME to 1 for approximately 10 ms. According to the USB 2.0 Specification [3],
the resume signaling must be present for at least 1 ms and no more than 15 ms. It is, however,
recommended to keep the resume signaling for approximately 10 ms. Notice that support for remote
wakeup must be declared in the USB descriptor, and that the USB host must grant the device the privilege
to perform remote wakeup (through a SET_FEATURE request).
21.12 USB Registers
This section describes all USB registers used for control and status for the USB. The USB registers reside
in XDATA memory space in the region 0x6200–0x622B. These registers can be divided into three groups:
The common USB registers, the indexed endpoint registers, and the endpoint FIFO registers. The indexed
endpoint registers represent the currently selected endpoint. The USBINDEX register is used to select the
endpoint.
The registers return to their reset values and the FIFOs are cleared when the chip enters PM2 or PM3.
USBADDR (0x6200) – Function Address
Bit
7
Name
UPDATE
Reset
0
R/W
R
6:0
USBADDR[6:0]
000 000 R/W
0
Description
This bit is set when the USBADDR register is written and cleared when the address becomes
effective.
Device address
USBPOW (0x6201) – Power/Control Register
Bit
7
Name
ISO_WAIT_SOF
Reset
0
R/W
R/W
6:4
3
2
–
RST
RESUME
000
0
0
R0
R
R/W
1
SUSPEND
0
R
0
SUSPEND_EN
0
R/W
Description
When this bit is set to 1, the USB controller sends zero-length data packets from the time
INPKT_RDY is asserted and until the first SOF token has been received. This only applies to
isochronous endpoints.
Reserved
During reset signaling, this bit is set to 1.
Drives resume signaling for remote wakeup. According to the USB Specification, the duration
of driving resume must be at least 1 ms and no more than 15 ms. It is recommended to keep
this bit set for approximately 10 ms.
Suspend mode entered. This bit is only used when SUSPEND_EN = 1. Reading the USBCIF
register or asserting RESUME clears this bit.
Suspend enable. When this bit is set to 1, suspend mode is entered when the USB has been
idle for 3 ms.
USBIIF (0x6202) – IN Endpoints and EP0 Interrupt Flags
Bit
7:6
5
4
3
2
1
0
190
Name
–
INEP5IF
INEP4IF
INEP3IF
INEP2IF
INEP1IF
EP0IF
USB Controller
Reset
00
0
0
0
0
0
0
R/W
R0
R, H0
R, H0
R, H0
R, H0
R, H0
R, H0
Description
Reserved
Interrupt flag
Interrupt flag
Interrupt flag
Interrupt flag
Interrupt flag
Interrupt flag
for
for
for
for
for
for
IN endpoint 5. Cleared by hardware when read
IN endpoint 4. Cleared by hardware when read
IN endpoint 3. Cleared by hardware when read
IN endpoint 2. Cleared by hardware when read
IN endpoint 1. Cleared by hardware when read
endpoint 0. Cleared by hardware when read
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Registers
www.ti.com
USBOIF (0x6204) – OUT-Endpoint Interrupt Flags
Bit
7:6
5
4
3
2
1
0
Name
–
OUTEP5IF
OUTEP4IF
OUTEP3IF
OUTEP2IF
OUTEP1IF
Reset
–
0
0
0
0
0
–
–
R/W
R0
R, H0
R, H0
R, H0
R, H0
R, H0
R0
Description
Reserved
Interrupt flag
Interrupt flag
Interrupt flag
Interrupt flag
Interrupt flag
Reserved
for
for
for
for
for
OUT
OUT
OUT
OUT
OUT
endpoint
endpoint
endpoint
endpoint
endpoint
5. Cleared
4. Cleared
3. Cleared
2. Cleared
1. Cleared
by
by
by
by
by
hardware when
hardware when
hardware when
hardware when
hardware when
read
read
read
read
read
USBCIF (0x6206) – Common USB Interrupt Flags
Bit
7:4
3
2
1
0
Name
–
SOFIF
RSTIF
RESUMEIF
SUSPENDIF
Reset
–
0
0
0
0
R/W
R0
R, H0
R, H0
R, H0
R, H0
Description
Reserved
Start-of-frame interrupt flag. Cleared by hardware when read
Reset interrupt flag. Cleared by hardware when read
Resume interrupt flag. Cleared by hardware when read
Suspend interrupt flag. Cleared by hardware when read
USBIIE (0x6207) – IN Endpoints and EP0 Interrupt-Enable Mask
Bit
7:6
5
Name
INEP5IE
Reset
00
1
R/W
R/W
R/W
4
INEP4IE
1
R/W
3
INEP3IE
1
R/W
2
INEP2IE
1
R/W
1
INEP1IE
1
R/W
0
EP0IE
1
R/W
Description
Reserved. Always write 00
IN endpoint-5 interrupt enable
0:
Interrupt disbled
1:
Interrupt enabled
IN endpoint- 4 interrupt enable
0:
Interrupt disbled
1:
Interrupt enabled
IN endpoint-3 interrupt enable
0:
Interrupt disbled
1:
Interrupt enabled
IN endpoint-2 interrupt enable
0:
Interrupt disbled
1:
Interrupt enabled
IN endpoint-1 interrupt enable
0:
Interrupt disbled
1:
Interrupt enabled
Endpoint-0 interrupt enable
0:
Interrupt disbled
1:
Interrupt enabled
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Controller
191
USB Registers
www.ti.com
USBOIE (0x6209) – Out Endpoints Interrupt Enable Mask
Bit
7:6
5
Name
OUTEP5IE
Reset
00
1
R/W
R/W
R/W
4
OUTEP4IE
1
R/W
3
OUTEP3IE
1
R/W
2
OUTEP2IE
1
R/W
1
OUTEP1IE
1
R/W
0
–
1
R0
Description
Reserved. Always write 00
OUT endpoint 5 interrupt enable
0: Interrupt disbled
1: Interrupt enabled
OUT endpoint 4 interrupt enable
0: Interrupt disbled
1: Interrupt enabled
OUT endpoint 3 interrupt enable
0: Interrupt disbled
1: Interrupt enabled
OUT endpoint 2 interrupt enable
0: Interrupt disbled
1: Interrupt enabled
OUT endpoint 1 interrupt enable
0: Interrupt disbled
1: Interrupt enabled
Reserved
USBCIE (0x620B) – Common USB Interrupt-Enable Mask
Bit
7:4
3
Name
SOFIE
Reset
–
0
R/W
R0
R/W
2
RSTIE
1
R/W
1
RESUMEIE
1
R/W
0
SUSPENDIE
0
R/W
Description
Reserved
Start-of-frame interrupt enable
0: Interrupt disbled
1: Interrupt enabled
Reset interrupt enable
0: Interrupt disbled
1: Interrupt enabled
Resume interrupt enable
0: Interrupt disbled
1: Interrupt enabled
Suspend interrupt enable
0: Interrupt disbled
1: Interrupt enabled
USBFRML (0x620C) – Current Frame Number (Low Byte)
Bit
7:0
Name
FRAME[7:0]
Reset
0x00
R/W
R
Description
Low byte of 11-bit frame number
USBFRMH (0x620D) – Current Frame Number (High Byte)
Bit
7:3
2:0
Name
–
FRAME[10:8]
Reset
–
000
R/W
R0
R
Description
Reserved
3 MSBs of 11-bit frame number
USBINDEX (0x620E) – Current-Endpoint Index Register
Bit
7:4
3:0
Name
–
USBINDEX[3:0]
192
USB Controller
Reset
–
0000
R/W
R0
R/W
Description
Reserved
Endpoint selected. Must be set to a value in the range 0–5
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Registers
www.ti.com
USBCTRL (0x620F) – USB Control Register
Bit
7
6:3
2
1
Name
PLL_LOCKED
–
–
PLL_EN
Reset
0
–
0
0
R/W
R
R0
R/W
R/W
0
USB_EN
0
R/W
Description
PLL locked status
Reserved
Reserved. Always write 0
48-MHz USB PLL enable. When this bit is set, the 48-MHz PLL is started. However, the USB
must not be accessed before the PLL has locked, i.e. ,PLL_LOCKED is 1. This bit can only be set
when USB_EN = 1.
Note: The PLL must be disabled before exiting active mode and re-enabled after entering active
mode.
USB enable. The USB controller is reset when writing 0 to this bit.
USBMAXI (0x6210) – Max. Packet Size for IN Endpoint{1–5}
Bit
7:0
Name
USBMAXI[7:0]
Reset
0x00
R/W
R/W
Description
Maximum packet size, in units of 8 bytes, for IN endpoint selected by USBINDEX register. The
value of this register should correspond to the wMaxPacketSize field in the standard endpoint
descriptor for the endpoint. This register must not be set to a value greater than the available
FIFO memory for the endpoint.
USBCS0 (0x6211) – EP0 Control and Status (USBINDEX = 0)
Bit
7
Name
CLR_SETUP_END
Reset
0
6
CLR_OUTPKT_RDY
0
5
SEND_STALL
0
4
SETUP_END
0
3
DATA_END
0
2
SENT_STALL
0
1
INPKT_RDY
0
0
OUTPKT_RDY
0
R/W
Description
R/W H0 Set this bit to 1 to de-assert the SETUP_END bit of this register. This bit is cleared
automatically.
R/W H0 Set this bit to 1 to de-assert the OUTPKT_RDY bit of this register. This bit is cleared
automatically.
R/W H0 Set this bit to 1 to terminate the current transaction. The USB controller sends the STALL
handshake and this bit is de-asserted.
R
This bit is set if the control transfer ends due to a premature end-of-control transfer. The
FIFO is flushed and an interrupt request (EP0) is generated if the interrupt is enabled.
Setting CLR_SETUP_END = 1 de-asserts this bit.
R/W H0 This bit is used to signal the end of a data transfer and must be asserted in the following
three situations:
1: When the last data packet has been loaded and USBCS0.INPKT_RDY is set to 1
2: When the last data packet has been unloaded and USBCS0.CLR_OUTPKT_RDY is set
to 1
3: When USBCS0.INPKT_RDY has been asserted without having loaded the FIFO (for
sending a zero-length data packet).
The USB controller clears this bit automatically.
R/W H1 This bit is set when a STALL handshake has been sent. An interrupt request (EP0) is
generated if the interrupt is enabled. This bit must be cleared from firmware.
R/W H0 Set this bit when a data packet has been loaded into the EP0 FIFO to notify the USB
controller that a new data packet is ready to be transferred. When the data packet has been
sent, this bit is cleared, and an interrupt request (EP0) is generated if the interrupt is
enabled.
R
Data packet received. This bit is set when an incoming data packet has been placed in the
OUT FIFO. An interrupt request (EP0) is generated if the interrupt is enabled. Set
CLR_OUTPKT_RDY = 1 to de-assert this bit.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Controller
193
USB Registers
www.ti.com
USBCSIL (0x6211) – IN EP{1–5} Control and Status, Low
Bit
7
6
Name
–
CLR_DATA_TOG
Reset
–
0
5
SENT_STALL
0
R/W
R0
R/W
H0
R/W
4
SEND_STALL
0
R/W
3
FLUSH_PACKET
0
R/W
H0
2
UNDERRUN
0
R/W
1
0
PKT_PRESENT
INPKT_RDY
0
0
R
R/W
H0
Description
Reserved
Setting this bit resets the data toggle to 0. Thus, setting this bit forces the next data packet
to be a DATA0 packet. This bit is automatically cleared.
This bit is set when a STALL handshake has been sent. The FIFO is flushed and the
INPKT_RDY bit in this register is de-asserted. An interrupt request (IN EP{1–5}) is
generated if the interrupt is enabled. This bit must be cleared from firmware.
Set this bit to 1 to make the USB controller reply with a STALL handshake when receiving
IN tokens. Firmware must clear this bit to end the STALL condition. It is not possible to stall
an isochronous endpoint; thus, this bit only has an effect if the IN endpoint is configured as
bulk/interrupt.
Set to 1 to flush next packet that is ready to transfer from the IN FIFO. The INPKT_RDY bit
in this register is cleared. If there are two packets in the IN FIFO due to double buffering,
this bit must be set twice to completely flush the IN FIFO. This bit is automatically cleared.
In isochronous mode, this bit is set if an IN token is received when
INPKT_RDY =
0, and a zero-length data packet is transmitted in response to the
IN token. In bulk/interrupt mode, this bit is set when a NAK is returned in response to an IN
token. Firmware should clear this bit.
This bit is 1 when there is at least one packet in the IN FIFO.
Set this bit when a data packet has been loaded into the IN FIFO to notify the USB
controller that a new data packet is ready to be transferred. When the data packet has
been sent, this bit is cleared, and an interrupt request (IN EP{1–5}) is generated if the
interrupt is enabled.
USBCSIH (0x6212) – IN EP{1–5} Control and Status, High
Bit
7
Name
AUTOSET
Reset
0
R/W
R/W
6
ISO
0
R/W
5:4
3
FORCE_DATA_TOG
10
0
R/W
R/W
2:1
0
IN_DBL_BUF
–
0
R0
R/W
Description
When this bit is 1, the USBCSIL.INPKT_RDY bit is automatically asserted when a data
packet of maximum size (specified by USBMAXI) has been loaded into the IN FIFO.
Selects IN endpoint type
0:
Bulk/interrupt
1:
Isochronous
Reserved. Always write 10
Setting this bit forces the IN endpoint data toggle to switch and the data packet to be
flushed from the IN FIFO, even though an ACK was received. This feature can be useful
when reporting rate feedback for isochronous endpoints.
Reserved
Double buffering enable (IN FIFO)
0:
Double buffering disabled
1:
Double buffering enabled
USBMAXO (0x6213) – Max. Packet Size for OUT EP{1–5}
Bit
7:0
194
Name
USBMAXO[7:0]
USB Controller
Reset
0x00
R/W
R/W
Description
Maximum packet size, in units of 8 bytes, for OUT endpoint selected by USBINDEX register.
The value of this register should correspond to the wMaxPacketSize field in the standard
endpoint descriptor for the endpoint. This register must not be set to a value greater than the
available FIFO memory for the endpoint.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Registers
www.ti.com
USBCSOL (0x6214) – OUT EP{1–5} Control and Status, Low
Bit
7
Name
CLR_DATA_TOG
Reset
0
6
SENT_STALL
0
R/W
R/W
H0
R/W
5
SEND_STALL
0
R/W
4
FLUSH_PACKET
0
R/W
H0
3
DATA_ERROR
0
R
2
OVERRUN
0
R/W
1
FIFO_FULL
0
R
0
OUTPKT_RDY
0
R/W
Description
Setting this bit resets the data toggle to 0. Thus, setting this bit forces the next data
packet to be a DATA0 packet. This bit is automatically cleared.
This bit is set when a STALL handshake has been sent. An interrupt request (OUT
EP{1–5}) is generated if the interrupt is enabled. This bit must be cleared from firmware.
Set this bit to 1 to make the USB controller reply with a STALL handshake when
receiving OUT tokens. Firmware must clear this bit to end the STALL condition. It is not
possible to stall an isochronous endpoint; thus, this bit only has an effect if the IN
endpoint is configured as bulk/interrupt.
Set to 1 to flush the next packet that is to be read from the OUT FIFO. The
OUTPKT_RDY bit in this register is cleared. If there are two packets in the OUT FIFO
due to double buffering, this bit must be set twice to completely flush the OUT FIFO. This
bit is automatically cleared after a write to 1.
This bit is set if there is a CRC or bit-stuff error in the packet received. Cleared when
OUTPKT_RDY is cleared. This bit is only valid if the OUT endpoint is isochronous.
This bit is set when an OUT packet cannot be loaded into the OUT FIFO. Firmware
should clear this bit. This bit is only valid in isochronous mode.
This bit is asserted when no more packets can be loaded into the OUT FIFO because it is
full.
This bit is set when a packet has been received and is ready to be read from the OUT
FIFO. An interrupt request (OUT EP{1–5}) is generated if the interrupt is enabled. This bit
should be cleared when the packet has been unloaded from the FIFO.
USBCSOH (0x6215) – OUT EP{1–5} Control and Status, High
Bit
7
Name
AUTOCLEAR
Reset
0
R/W
R/W
6
ISO
0
R/W
OUT_DBL_BUF
00
–
0
R/W
R0
R/W
5:4
3:1
0
Description
When this bit is set to 1, the USBCSOL.OUTPKT_RDY bit is automatically cleared when a
data packet of maximum size (specified by USBMAXO) has been unloaded to the OUT
FIFO.
Selects OUT endpoint type
0
Bulk/interrupt
1
Isochronous
Reserved. Always write 00
Reserved
Double buffering enable (OUT FIFO)
0
Double buffering disabled
1
Double buffering enabled
USBCNT0 (0x6216) – Number of Received Bytes in EP0 FIFO (USBINDEX = 0)
Bit
7:6
5:0
Name
–
USBCNT0[5:0]
Reset
–
00 0000
R/W
R0
R
Description
Reserved
Number of received bytes into EP 0 FIFO. Only valid when OUTPKT_RDY is asserted
USBCNTL (0x6216) – Number of Bytes in EP{1–5} OUT FIFO, Low
Bit
7:0
Name
USBCNT[7:0]
Reset
0x00
R/W
R
Description
8 lsbs of number of received bytes in OUT FIFO selected by USBINDEX register. Only
valid when USBCSOL.OUTPKT_RDY is asserted.
USBCNTH (0x6217) – Number of Bytes in EP{1–5} OUT FIFO, High
Bit
7:3
2:0
Name
–
USBCNT[10:8]
Reset
–
000
R/W
R0
R
Description
Reserved
3 msbs of number of received bytes in OUT FIFO selected by USBINDEX register. Only
valid when USBCSOL.OUTPKT_RDY is set
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
USB Controller
195
USB Registers
www.ti.com
USBF0 (0x6220) – Endpoint-0 FIFO
Bit
7:0
Name
USBF0[7:0]
Reset
0x00
R/W
R/W
Description
Endpoint 0 FIFO. Reading this register unloads one byte from the EP0 FIFO. Writing to
this register loads one byte into the EP0 FIFO.
Note: The FIFO memory for EP0 is used for both incoming and outgoing data packets.
R/W
R/W
Description
Endpoint 1 FIFO register. Reading this register unloads one byte from the EP1 OUT
FIFO. Writing to this register loads one byte into the EP1 IN FIFO.
R/W
R/W
Description
Endpoint 2 FIFO register. Reading this register unloads one byte from the EP2 OUT
FIFO. Writing to this register loads one byte into the EP2 IN FIFO.
R/W
R/W
Description
Endpoint 3 FIFO register. Reading this register unloads one byte from the EP3 OUT
FIFO. Writing to this register loads one byte into the EP3 IN FIFO.
R/W
R/W
Description
Endpoint 4 FIFO register. Reading this register unloads one byte from the EP4 OUT
FIFO. Writing to this register loads one byte into the EP4 IN FIFO.
R/W
R/W
Description
Endpoint 5 FIFO register. Reading this register unloads one byte from the EP5 OUT
FIFO. Writing to this register loads one byte into the EP5 IN FIFO.
USBF1 (0x6222) – Endpoint-1 FIFO
Bit
7:0
Name
USBF1[7:0]
Reset
0x00
USBF2 (0x6224) – Endpoint-2 FIFO
Bit
7:0
Name
USBF2[7:0]
Reset
0x00
USBF3 (0x6226) – Endpoint-3 FIFO
Bit
7:0
Name
USBF3[7:0]
Reset
0x00
USBF4 (0x6228) – Endpoint-4 FIFO
Bit
7:0
Name
USBF4[7:0]
Reset
0x00
USBF5 (0x622A) – Endpoint-5 FIFO
Bit
7:0
196
Name
USBF5[7:0]
USB Controller
Reset
0x00
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 22
SWRU191D – April 2009 – Revised March 2013
Timer 2 (MAC Timer)
Timer 2 is mainly used to provide timing for 802.15.4 CSMA-CA algorithms and for general timekeeping in
the 802.15.4 MAC layer on CC253x devices, for timekeeping in the BLE link layer on CC2540/41, and for
general radio timekeeping when running the radio in proprietary mode on CC2541. Timer 2 must not be
used by the application on the CC2540/41 when the BLE stack is running. When Timer 2 is used together
with the Sleep Timer, the timing function is provided even when the system enters low-power modes PM1
and PM2. The timer runs at a speed according to the system clock. If Timer 2 is to be used with the Sleep
Timer, the system clock source must be the 32-MHz crystal whenever Timer 2 is running, and an external
32-kHz XOSC should be used for accurate results.
The main features of Timer 2 are the following:
• 16-bit timer up-counter providing, for example, a symbol/frame period of 16 μs/320 μs
• Adjustable period with accuracy of 31.25 ns
• 2 × 16-bit timer compare function
• 24-bit overflow count
• 2 × 24-bit overflow compare function
• Start-of-frame-delimiter capture function
• Timer start/stop synchronous with 32-kHz clock and timekeeping maintained by Sleep Timer.
• Interrupts generated on compare and overflow
• DMA trigger capability
• Possible to adjust timer value while counting by introducing delay
Topic
22.1
22.2
22.3
22.4
22.5
...........................................................................................................................
Timer Operation ..............................................................................................
Interrupts ........................................................................................................
Event Outputs (DMA Trigger and Radio Events) ..................................................
Timer Start/Stop Synchronization ......................................................................
Timer 2 Registers ............................................................................................
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 (MAC Timer)
Page
198
199
200
200
202
197
Timer Operation
www.ti.com
22.1 Timer Operation
This section describes the operation of the timer.
22.1.1 General
After a reset, the timer is in the timer IDLE mode, where it is stopped. The timer starts running when
T2CTRL.RUN is set to 1. The timer then enters the timer RUN mode. Either the entry is immediate, or it is
performed synchronously with the 32-kHz clock. See Section 22.4 for a description of the synchronous
start and stop mode.
Once the timer is running in RUN mode, it can be stopped by writing a 0 to T2CTRL.RUN. The timer then
enters the timer IDLE mode. The stopping of the timer is performed either immediately or synchronously
with the 32-kHz clock.
22.1.2 Up Counter
Timer 2 contains a 16-bit timer, which increments on each clock cycle. The counter value can be read
from registers T2M1:T2M0 with register T2MSEL.T2MSEL set to 000. Note that the register content in
T2M1 is latched when T2M0 is read, meaning that T2M0 must always be read first.
When the timer is idle, the counter can be modified by writing to registers T2M1:T2M0 with register
T2MSEL.T2MSEL set to 000. T2M0 must be written first.
22.1.3 Timer Overflow
At the same time as the timer counts to a value that is equal to the set timer period, a timer overflow
occurs. When the timer overflow occurs, the timer is set to 0x0000. If the overflow interrupt mask bit
T2IRQM.TIMER2_PERM is 1, an interrupt request is generated. The interrupt flag bit
T2IRQF.TIMER2_PERF is set to 1, regardless of the interrupt mask value.
22.1.4 Timer Delta Increment
The timer period may be adjusted once during a timer period by writing a timer delta value. When the
timer is running and a timer delta value is written to multiplexed registers T2M1:T2M0 with
T2MSEL.T2MSEL set to 000, the 16-bit timer halts at its current value and a delta counter starts counting.
The T2M0 register must be written before T2M1. The delta counter starts counting from the delta value
written, down to zero. Once the delta counter reaches zero, the 16-bit timer starts counting again.
The delta counter decrements at the same rate as the timer. When the delta counter has reached zero, it
does not start counting again until a delta value is written once again. In this way, a timer period may be
increased by the delta value in order to make adjustments to the timer overflow events over time.
22.1.5 Timer Compare
A timer compare occurs at the same time as the timer counts to a value that is equal to one of the 16-bit
compare values set. When a timer compare occurs, the interrupt flag T2IRQF.TIMER2_COMPARE1F or
T2IRQF.TIMER2_COMPARE2F is set to 1, depending of which compare value is reached. An interrupt
request is also generated if the corresponding interrupt mask in T2IRQM.TIMER2_COMPARE1M or
T2IRQM.TIMER2_COMPARE2M is set to 1.
22.1.6 Overflow Count
At each timer overflow, the 24-bit overflow counter is incremented by 1. The overflow counter value is read
through registers T2MOVF2:T2MOVF1:T2MOVF0 with register T2MSEL.T2MOVEFSEL set to 000. The
registers are latched as in the following description.
If one wants a unique timestamp, where both timer and overflow counter are latched at the same time, do
the following: Read T2M0 with T2MSEL.T2MSEL set to 000 and T2CTRL.LATCH_MODE set to 1. This
returns the low byte of the timer value, and also latches the high byte of the timer and the entire overflow
counter, so the rest of the timestamp is ready to be read.
198
Timer 2 (MAC Timer)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Interrupts
www.ti.com
If one wants to read just the overflow counter without reading timer first, read T2MOVF0 with
T2MSEL.T2MOVFSEL set to 000 and T2CTRL.LATCH_MODE set to 0. This returns the low byte of the
overflow counter, and latches the two most-significant bytes of the overflow counter so the values are
ready to be read.
22.1.7 Overflow-Count Update
The overflow count value can be updated by writing to registers T2MOVF2:T2MOVF1:T2MOVF0 with
T2MSEL.T2MOVFSEL set to 000. Always write the least-significant byte first, and always write all three
bytes. The write takes effect once the high byte is written.
22.1.8 Overflow-Count Overflow
At the same time as the overflow counter counts to a value that is equal to the overflow period setting, an
overflow period event occurs. When the period event occurs, the overflow counter is set to 0x00 0000. If
the overflow interrupt mask bit T2IRQM.TIMER2_OVF_PERM is 1, an interrupt request is generated. The
interrupt flag bit T2IRQF.TIMER2_OVF_PERF is set to 1, regardless of the interrupt mask value.
22.1.9 Overflow-Count Compare
Two compare values may be set for the overflow counter. The compare values are set by writing to
T2MOVF2:T2MOVF1:T2MOVF0 with register T2MSEL.T2MOVFSEL set to 011 or 100. At the same time as
the overflow counter counts to a value equal to one of the overflow count compare values, an overflow
count compare event occurs. If the corresponding overflow compare interrupt mask bit
T2IRQM.TIMER2_OVF_COMPARE1M or T2IRQM.TIMER2_OVF_COMPARE2M is 1, an interrupt request is
generated. The interrupt flags bit T2IRQF.TIMER2_OVF_COMPARE1F and
T2IRQF.TIMER2_OVF_COMPARE2F are set to 1, regardless of the interrupt mask value.
22.1.10 Capture Input
Timer 2 has a timer capture function, which captures the time when the start-of-frame delimiter (SFD)
status in the radio goes high.
When the capture event occurs, the current timer value is captured in the capture register. The capture
value can be read from registers T2M1:T2M0 if register T2MSEL.T2MSEL is set to 001. The value of the
overflow count is also captured at the time of the capture event and can be read from registers
T2MOVF2:T2MOVF1:T2MOVF0 if T2MSEL.T2MOVFSEL is set to 001.
22.1.11 Long Compare (CC2541 Only)
In the CC2541, two compare values may be set for the combination of the 16-bit timer and the overflow
counter. The compare values are a combination of either timer compare 1 and overflow compare 1, or
timer compare 2 and overflow compare 2. These combinations are known as the long compare 1 and long
compare 2 values, respectively. At the same time as the combination of the 16-bit timer and the 24-bit
overflow counter counts to a value equal to one of the long compare values, a long compare event occurs.
If the corresponding overflow compare interrupt mask bit T2IRQM.TIMER2_LONG_COMPARE1M or
T2IRQM.TIMER2_LONG_COMPARE2M is 1, an interrupt request is generated. The corresponding interrupt
flag bit T2IRQF.TIMER2_LONG_COMPARE1F or T2IRQF.TIMER2_LONG_COMPARE2F is set to 1,
regardless of the interrupt mask value.
22.2 Interrupts
The timer has six (eight on CC2541) individually maskable interrupt sources. These are the following:
• Timer overflow
• Timer compare 1
• Timer compare 2
• Overflow-count overflow
• Overflow-count compare 1
• Overflow-count compare 2
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 (MAC Timer)
199
Event Outputs (DMA Trigger and Radio Events)
•
•
www.ti.com
Long compare 1 (CC2541 only)
Long compare 2 (CC2541 only)
The interrupt flags are given in the T2IRQF registers. The interrupt flag bits are set only by hardware and
can be cleared only by writing to the SFR register.
Each interrupt source can be masked by its corresponding mask bit in the T2IRQM register. An interrupt is
generated when the corresponding mask bit is set; otherwise, the interrupt is not generated. The interrupt
flag bit is set, however, regardless of the state of the interrupt mask bit.
22.3 Event Outputs (DMA Trigger and Radio Events)
Timer 2 has two event outputs, T2_EVENT1 and T2_EVENT2. These can be used as DMA triggers, as
inputs to the radio, for conditions in conditional instructions in the CSP on CC253x, for use by the BLE
stack on CC2540/41, or for timing TX or RX in CC2541 when running the radio in proprietary mode. The
event outputs can be configured individually to any of the following events:
• Timer overflow
• Timer compare 1
• Timer compare 2
• Overflow-count overflow
• Overflow-count compare 1
• Overflow-count compare 2
• Long compare 1 (CC2541 only)
• Long compare 2 (CC2541 only)
The DMA triggers are configured using T2EVTCFG.TIMER2_EVENT1_CFG and
T2EVTCFG.TIMER2_EVENT2_CFG.
22.4 Timer Start/Stop Synchronization
This section describes the synchronized timer start and stop.
22.4.1 General
The timer can be started and stopped synchronously with the 32-kHz clock rising edge. Note that this
event is derived from a 32-kHz clock signal, but is synchronous with the 32-MHz system clock and thus
has a period approximately equal to that of the 32-kHz clock period. Syncronous starting and stopping
must not be attempted unless both the 32-kHz clock and 32-MHz XOSC are running and stable.
At the time of a synchronous start, the timer is reloaded with new calculated values for the timer and
overflow count such that it appears that the timer has not been stopped.
22.4.2 Timer Synchronous Stop
After the timer has started running, i.e., entered timer RUN mode, it is stopped synchronously by writing 0
to T2CTRL.RUN when T2CTRL.SYNC is 1. After T2CTRL.RUN has been set to 0, the timer continues
running until the 32-kHz clock rising edge is sampled as 1. When this occurs, the timer is stopped, the
current Sleep Timer value is stored, and T2CTRL.STATE goes from 1 to 0.
200
Timer 2 (MAC Timer)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer Start/Stop Synchronization
www.ti.com
22.4.3 Timer Synchronous Start
When the timer is in the IDLE mode, it is started synchronously by writing 1 to T2CTRL.RUN when
T2CTRL.SYNC is 1. After T2CTRL.RUN has been set to 1, the timer remains in the IDLE mode until the
32-kHz clock rising edge is detected. When this occurs, the timer first calculates new values for the 16-bit
timer value and for the 24-bit timer overflow count, based on the current and stored Sleep Timer values
and the current 16-bit timer values. The new Timer 2 and overflow count values are loaded into the timer,
and the timer enters the RUN mode. T2CTRL.STATE = 1 indicates that the module is running. This
synchronous start process takes 86 clock cycles from the time when the 32-kHz clock rising edge is
sampled high. The synchronous start-and-stop function requires that the system clock frequency is
selected to be 32 MHz. If the 16-MHz clock is selected, an offset is added to the new calculated value.
If a synchronous start is done without a previous synchronuous stop, the timer is loaded with
unpredictable values. To avoid this, do the first start of the timer asynchronously, then enable
synchronous mode for subsequent stops and starts.
The method for calculating the new Timer 2 value and overflow-count value is given as follows. Because
the Timer 2 and Sleep Timer clocks are asynchronous with a noninteger clock ratio, there is an error of
maximum ±1 in the calculated timer value compared to the ideal timer value, not taking clock inaccuracies
into account.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 (MAC Timer)
201
Timer 2 Registers
www.ti.com
Calculation of New Timer Value and Overflow Count Value
Nc = Current Sleep Timer value
NST = Stored Sleep Timer value
Kck = Clock ratio = 976.5625(1)
stw = Sleep Timer width = 24
PT = Timer 2 period
POVF = Overflow period
OST = Stored overflow-count value
OTICK = Overflow tics while sleeping
tST = Stored timer value
TOH = Overhead = 86
Nt = Nc – NST
Nt ≤ 0 → Nd = 2stw + Nt; Nt > 0 → Nd = Nt
C = Nd × Kck + TST + TOH (rounded to nearest integer value)
T = C mod PT
Timer2Value = T
OTICK =
(C - T )
PT
O = (OTICK + OST) mod POVF
Timer2OverflowCount = O
(1)
Clock ratio of Timer 2 clock frequency (32 MHz) and Sleep Timer clock frequency (32 kHz)
For a given Timer 2 period value, PT, there is a maximum duration between Timer 2 synchronous stop and
start for which the timer value is correctly updated after starting. The maximum value is given in terms of
the number of Sleep Timer clock periods, i.e., 32-kHz clock periods, tST(max).
t ST(max) £
(224 - 1) ´ PT + TOH
K ck
22.5 Timer 2 Registers
The SFR registers associated with Timer 2 are listed in this section. These registers are the following:
•
•
•
•
•
•
•
•
•
•
202
T2MSEL
T2M1
T2M0
T2MOVF2
T2MOVF1
T2MOVF0
T2IRQF
T2IRQM
T2EVTCFG
T2CTRL
Timer 2 (MAC Timer)
– Timer 2
– Timer 2
– Timer 2
– Timer 2
– Timer 2
– Timer 2
– Timer 2
– Timer 2
– Timer 2
– Timer 2
multiplexed register control
multiplexed count high
multiplexed count low
multiplexed overflow count 2
multiplexed overflow count 1
multiplexed overflow count 0
interrupt flags
interrupt masks
event output configuration
configuration
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 Registers
www.ti.com
Timer 2 has several multiplexed registers. This is to be able to fit all the registers into the limited SFR
address space. The internal registers listed in Table 22-1 can be accessed indirectly through T2M0, T2M1,
T2MOVF0, T2MOVF1, and T2MOVF2.
Table 22-1. Internal Registers
Register Name
Reset
R/W
Function
t2tim[15:0]
0x0000
R/W
Holds the 16-bit upcounter
t2_cap[15:0]
0x0000
R
Holds the last captured value of the upcounter
t2_per[15:0]
0x0000
R/W
Holds the period of the upcounter
t2_cmp1[15:0]
0x0000
R/W
Holds compare value 1 for the upcounter
t2_cmp2[15:0]
0x0000
R/W
Holds compare value 2 for the upcounter
t2ovf[23:0]
0x00 0000
R/W
Holds the 24-bit overflow counter
t2ovf_cap[23:0]
0x00 0000
R
Holds the last captured value of the overflow counter
t2ovf_per[23:0]
0x00 0000
R/W
Holds the period of the overflow counter
t2ovf_cmp1[23:0]
0x00 0000
R/W
Holds compare value 1 for the overflow counter
t2ovf_cmp2[23:0]
0x00 0000
R/W
Holds compare value 2 for the overflow counter
The registers listed in the remainder of this section are directly accessible in the SFR address space.
T2MSEL (0xC3) – Timer 2 Multiplex Select
Bit
No.
7:0
6:4
3
2:0
Name
–
T2MOVFSEL
–
T2MSEL
Reset
R/W
0
0
R0
R/W
0
0
R0
R/W
Function
Reserved. Read as 0
The value of this register selects the internal registers that are modified or read
when accessing T2MOVF0, T2MOVF1, and T2MOVF2.
000: t2ovf (overflow counter)
001: t2ovf_cap (overflow capture)
010: t2ovf_per (overflow period)
011: t2ovf_cmp1 (overflow compare 1)
100: t2ovf_cmp2 (overflow compare 2)
101 to 111: Reserved
Reserved. Read as 0
The value of this register selects the internal registers that are modified or read
when accessing T2M0 and T2M1.
000: t2tim (timer count value)
001: t2_cap (timer capture)
010: t2_per (timer period)
011: t2_cmp1 (timer compare 1)
100: t2_cmp2 (timer compare 2)
101 to 111: Reserved
T2M0 (0xA2) – Timer 2 Multiplexed Register 0
Bit
No.
7:0
Name
T2M0
Reset
R/W
0
R/W
Function
Indirectly returns/modifies bits [7:0] of an internal register depending on the
T2MSEL.T2MSEL value.
When reading the T2M0 register with T2MSEL.T2MSEL set to 000 and
T2CTRL.LATCH_MODE set to 0, the timer (t2tim) value is latched.
When reading the T2M0 register with T2MSEL.T2MSEL set to 000 and
T2CTRL.LATCH_MODE set to 1, the timer (t2tim) and overflow counter (t2ovf)
values are latched.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 (MAC Timer)
203
Timer 2 Registers
www.ti.com
T2M1 (0xA3) – Timer 2 Multiplexed Register 1
Bit
No.
7:0
Name
Reset
R/W
0
T2M1
R/W
Function
Indirectly returns/modifies bits [15:8] of an internal register, depending on
T2MSEL.T2MSEL value.
When reading the T2M0 register with T2MSEL.T2MSEL set to 000, the timer
(t2tim) value is latched.
Reading this register with T2MSEL.T2MSEL set to 000 returns the latched value
of t2tim[15:8].
T2MOVF0 (0xA4) – Timer 2 Multiplexed Overflow Register 0
Bit
No.
7:0
Name
Reset
R/W
0
T2MOVF0
R/W
Function
Indirectly returns/modifies bits [7:0] of an internal register, depending on the
T2MSEL.T2MOVFSEL value.
When reading the T2MOVF0 register with T2MSEL.T2MOVFSEL set to 000 and
T2CTRL.LATCH_MODE set to 0, the overflow counter value (t2ovf) is latched.
When reading the T2M0 register with T2MSEL.T2MOVFSEL set to 000 and
T2CTRL.LATCH_MODE set to 1, the overflow counter value (t2ovf) is latched.
T2MOVF1 (0xA5 – Timer 2 Multiplexed Overflow Register 2
Bit
No.
7:0
Name
Reset
R/W
0
T2MOVF1
R/W
Function
Indirectly returns/modifies bits [15:8] of an internal register, depending on the
T2MSEL.T2MSEL value.
Reading this register with T2MSEL.T2MOVFSEL set to 000 returns the latched
value of t2ovf[15:8].
T2MOVF2 (0xA6) – Timer 2 Multiplexed Overflow Register 2
Bit
No.
7:0
Name
Reset
R/W
0
T2MOVF2
R/W
Function
Indirectly returns/modifies bits [23:16] of an internal register, depending on the
T2MSEL.T2MOVFSEL value.
Reading this register with T2MSEL.T2MOVFSEL set to 000 returns the latched
value of t2ovf[23:16].
T2IRQF (0xA1) – Timer 2 Interrupt Flags (CC253x/CC2540)
Bit
No.
7:6
5
4
Name
–
TIMER2_OVF_COMPARE2F
TIMER2_OVF_COMPARE1F
3
Reset
R/W
Function
0
0
0
R0
R/W0
R/W0
TIMER2_OVF_PERF
0
R/W0
2
1
0
TIMER2_COMPARE2F
TIMER2_COMPARE1F
TIMER2_PERF
0
0
0
R/W0
R/W0
R/W0
Reserved. Read as 0
Set when the Timer 2 overflow counter counts to the value set at t2ovf_cmp2
Set when the Timer 2 overflow counter counts to the value set at Timer 2
t2ovf_cmp1
Set when the Timer 2 overflow counter would have counted to a value equal to
t2ovf_per, but instead wraps to 0
Set when the Timer 2 counter counts to the value set at t2_cmp2
Set when the Timer 2 counter counts to the value set at t2_cmp1
Set when the Timer 2 counter would have counted to a value equal to t2_per,
but instead wraps to 0.
204
Timer 2 (MAC Timer)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 Registers
www.ti.com
T2IRQF (0xA1) – Timer 2 Interrupt Flags (CC2541 Only)
Bit
No.
7
Name
TIMER2_LONG_COMPARE2F
6
Reset
R/W
Function
0
R/W0
TIMER2_LONG_COMPARE1F
0
R/W0
5
4
TIMER2_OVF_COMPARE2F
TIMER2_OVF_COMPARE1F
0
0
R/W0
R/W0
3
TIMER2_OVF_PERF
0
R/W0
2
1
0
TIMER2_COMPARE2F
TIMER2_COMPARE1F
TIMER2_PERF
0
0
0
R/W0
R/W0
R/W0
Set when the Timer 2 overflow counter is equal to t2ovf_cmp2 and the timer
counts to the value set at t2_cmp2
Set when the Timer 2 overflow counter is equal to t2ovf_cmp1 and the timer
counts to the value set at t2_cmp1
Set when the Timer 2 overflow counter counts to the value set at t2ovf_cmp2
Set when the Timer 2 overflow counter counts to the value set at Timer 2
t2ovf_cmp1
Set when the Timer 2 overflow counter would have counted to a value equal to
t2ovf_per, but instead wraps to 0
Set when the Timer 2 counter counts to the value set at t2_cmp2
Set when the Timer 2 counter counts to the value set at t2_cmp1
Set when the Timer 2 counter would have counted to a value equal to t2_per,
but instead wraps to 0.
T2IRQM (0xA7) – Timer 2 Interrupt Mask (CC253x/CC2540)
Bit
No.
7:6
5
4
3
2
1
0
Name
–
TIMER2_OVF_COMPARE2M
TIMER2_OVF_COMPARE1M
TIMER2_OVF_PERM
TIMER2_COMPARE2M
TIMER2_COMPARE1M
TIMER2_PERM
Reset
0
0
0
0
0
0
0
R/W
R0
R/W
R/W
R/W
R/W
R/W
R/W
Function
Reserved. Read as 0
Enables the TIMER2_OVF_COMPARE2 interrupt
Enables the TIMER2_OVF_COMPARE1 interrupt
Enables the TIMER2_OVF_PER interrupt
Enables the TIMER2_COMPARE2 interrupt
Enables the TIMER2_COMPARE1 interrupt
Enables the TIMER2_PER interrupt
T2IRQM (0xA7) – Timer 2 Interrupt Mask (CC2541 Only)
Bit
No.
7
6
5
4
3
2
1
0
Name
TIMER2_LONG_COMARE2M
TIMER2_LONG_COMARE1M
TIMER2_OVF_COMPARE2M
TIMER2_OVF_COMPARE1M
TIMER2_OVF_PERM
TIMER2_COMPARE2M
TIMER2_COMPARE1M
TIMER2_PERM
Reset
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Function
Enables the
Enables the
Enables the
Enables the
Enables the
Enables the
Enables the
Enables the
TIMER2_LONG_COMPARE2F interrupt
TIMER2_LONG_COMPARE1F interrupt
TIMER2_OVF_COMPARE2 interrupt
TIMER2_OVF_COMPARE1 interrupt
TIMER2_OVF_PER interrupt
TIMER2_COMPARE2 interrupt
TIMER2_COMPARE1 interrupt
TIMER2_PER interrupt
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 (MAC Timer)
205
Timer 2 Registers
www.ti.com
T2CTRL (0x94) – Timer 2 Control Register
Bit
No.
7:4
3
Name
–
LATCH_MODE
Reset
R/W
0
0
R0
R/W
Function
Reserved. Read as 0
Reading T2M0 with T2MSEL.T2MSEL = 000 latches the high byte of the timer,
making it ready to be read from T2M1. Reading T2MOVF0 with T2MSEL.T2MOVFSEL
= 000 latches the two most-significant bytes of the overflow counter, making it
possible to read these from T2MOVF1 and T2MOVF2.
Reading T2M0 with T2MSEL.T2MSEL = 000 latches the high byte of the timer and
the entire overflow counter at once, making it possible to read the values from T2M1,
T2MOVF0, T2MOVF1, and T2MOVF2.
0:
1:
2
STATE
0
R
1
SYNC
1
R/W
0
RUN
0
R/W
State of Timer 2
0:
Timer idle
1:
Timer running
0:
Starting and stopping of timer is immediate, i.e., synchronous with clk_rf_32m.
1:
Starting and stopping of timer happens at the first positive edge of the 32-kHz clock.
Read Section 22.4 for more details regarding timer start and stop.
Write 1 to start timer, write 0 to stop timer. When read, it returns the last written
value.
T2EVTCFG (0x9C) – Timer 2 Event Configuration (CC253x/CC2540)
Bit
No.
Name
Reset
R/W
7
6:4
–
TIMER2_EVENT2_CFG
0
0
R0
R/W
3
2:0
–
TIMER2_EVENT1_CFG
0
0
R0
R/W
206
Timer 2 (MAC Timer)
Function
Reserved. Read as 0
Selects the event that triggers a T2_EVENT2 pulse
000: t2_per_event
001: t2_cmp1_event
010: t2_cmp2_event
011: t2ovf_per_event
100: t2ovf_cmp1_event
101: t2ovf_cmp2_event
110: Reserved
111: No event
Reserved. Read as 0
Selects the event that triggers a T2_EVENT1 pulse
000: t2_per_event
001: t2_cmp1_event
010: t2_cmp2_event
011: t2ovf_per_event
100: t2ovf_cmp1_event
101: t2ovf_cmp2_event
110: Reserved
111: No event
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 Registers
www.ti.com
T2EVTCFG (0x9C) – Timer 2 Event Configuration (CC2541)
Bit
No.
7:4
Name
Reset
R/W
TIMER2_EVENT2_CFG
0
R/W
3:0
TIMER2_EVENT1_CFG
0
R/W
Function
Selects the event that triggers a T2_EVENT2 pulse
0000: t2_per_event
0001: t2_cmp1_event
0010: t2_cmp2_event
0011: t2ovf_per_event
0100: t2ovf_cmp1_event
0101: t2ovf_cmp2_event
0110: Reserved
0111: No event
1000: t2ovf_long_cmp1_event
1001: t2ovf_long_cmp2_event
1010–1110: Reserved
1111: No event
Selects the event that triggers a T2_EVENT1 pulse
0000: t2_per_event
0001: t2_cmp1_event
0010: t2_cmp2_event
0011: t2ovf_per_event
0100: t2ovf_cmp1_event
0101: t2ovf_cmp2_event
0110: Reserved
0111: No event
1000: t2ovf_long_cmp1_event
1001: t2ovf_long_cmp2_event
1010–1110: Reserved
1111: No event
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Timer 2 (MAC Timer)
207
Chapter 23
SWRU191D – April 2009 – Revised March 2013
CC253x Radio
The RF Core controls the analog radio modules. In addition, it provides an interface between the MCU
and the radio which makes it possible to issue commands, read status, and automate and sequence radio
events.
Topic
...........................................................................................................................
23.1
23.2
23.3
23.4
23.5
23.6
23.7
23.8
23.9
23.10
23.11
23.12
23.13
23.14
23.15
208
RF Core ..........................................................................................................
FIFO Access ...................................................................................................
DMA ...............................................................................................................
Memory Map ...................................................................................................
Frequency and Channel Programming ...............................................................
IEEE 802.15.4-2006 Modulation Format ...............................................................
IEEE 802.15.4-2006 Frame Format ......................................................................
Transmit Mode ................................................................................................
Receive Mode ..................................................................................................
RXFIFO Access ..............................................................................................
Radio Control State Machine ...........................................................................
Random-Number Generation ...........................................................................
Packet Sniffing and Radio Test Output Signals ..................................................
Command Strobe/CSMA-CA Processor .............................................................
Registers .......................................................................................................
CC253x Radio
Page
209
213
213
213
215
215
217
218
222
232
234
236
237
238
255
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RF Core
www.ti.com
23.1 RF Core
The RF Core controls the analog radio modules. In addition, it provides an interface between the MCU
and the radio which makes it possible to issue commands, read status, and automate and sequence radio
events.
The FSM submodule controls the RF transceiver state, the transmitter and receiver FIFOs, and most of
the dynamically controlled analog signals, such as power up/down of analog modules. The FSM is used to
provide the correct sequencing of events (such as performing an FS calibration before enabling the
receiver/transmitter). Also, it provides step-by-step processing of incoming frames from the demodulator:
reading the frame length, counting the number of bytes received, checking the FCS, and finally, optionally
handling automatic transmission of ACK frames after successful frame reception. It performs similar tasks
in TX, including performing an optional CCA before transmission and automatically going to RX after the
end of transmission to receive an ACK frame. Finally, the FSM controls the transfer of data between
modulator/demodulator and the TXFIFO/RXFIFO in RAM.
The modulator transforms raw data into I/Q signals to the transmitter DAC. This is done in compliance
with the IEEE 802.15.4 standard.
The demodulator is responsible for retrieving the over-the-air data from the received signal.
The amplitude information from the demodulator is used by the automatic gain control (AGC). The AGC
adjusts the gain of the analog LNA so that the signal level within the receiver is approximately constant.
The frame filtering and source matching supports the FSM in the RF Core by performing all operations
needed in order to do frame filtering and source address matching, as defined by IEEE 802.15.4.
The frequency synthesizer (FS) generates the carrier wave for the RF signal.
The command strobe processor (CSP) processes all commands issued by the CPU. It also has a short
program memory of 24 bytes, making it possible to automate CSMA-CA algorithms.
The radio RAM holds a FIFO for transmit data (TXFIFO) and a FIFO for receive data (RXFIFO). Both
FIFOs are 128 bytes long. In addition, the RAM holds parameters for frame filtering and source matching,
and for which 128 bytes are reserved.
Timer 2 (MAC Timer) is used for timing of radio events and to capture time stamps of incoming packets.
This timer keeps counting even in power modes PM1 and PM2.
23.1.1 Interrupts
The radio is associated with two interrupt vectors on the CPU. These are the RFERR interrupt (interrupt
0) and the RF interrupt (interrupt 12) with the following functions.
• RFERR: Error situations in the radio are signaled using this interrupt.
• RF: Interrupts coming from normal operation are signaled using this interrupt.
The RF interrupt vector combines the interrupts in RFIF. Note that these RF interrupts are rising-edge
triggered. Thus, an interrupt is generated when, e.g., the SFD status flag goes from 0 to 1. The RFIF
interrupt flags are described in Section 23.1.2.
23.1.2 Interrupt Registers
Two of the main interrupt control SFR registers are used to enable the RF and RFERR interrupts. These
are the following:
• RFERR: IEN0.RFERRIE
• RF: IEN2.RFIE
Two main interrupt flag SFR registers hold the RF and RFERR interrupt flags. These are the following:
• RFERR: TCON.RFERRIF
• RF: S1CON.RFIF
The two interrupts generated by the RF Core are a combination of several sources within the RF Core.
Each of the individual sources has its own enable and interrupt flags in the RF Core. Flags can be found
in RFIRQF0, RFIRQF1, and RFIERRF. Interrupt masks can be found in RFIRQM0, RFIRQM1, and RFERRM.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
209
RF Core
www.ti.com
The interrupt-enable bits in the mask registers are used to enable individual interrupt sources for the two
RF interrupts. Note that masking an interrupt source does not affect the updating of the status in the flag
registers.
Due to the use of individual interrupt masks in the RF Core, the interrupts coming from RF Core have twolayered masking, and care must be taken when processing these interrupts. The procedure is described
as follows.
To clear an interrupt from the RF Core, one must clear two flags, both the flag set in RF Core and the one
set in S1CON or TCON (depending on which interrupt is triggered). If a flag is cleared in the RF Core and
there are other unmasked flags standing, another interrupt is generated.
RFIRQF0 (0xE9) RF Interrupt Flags
Bit
7
Name
RXMASKZERO
Reset
0
R/W
R/W0
6
RXPKTDONE
0
R/W0
5
FRAME_ACCEPTED
0
R/W0
4
SRC_MATCH_FOUND
0
R/W0
3
SRC_MATCH_DONE
0
R/W0
2
FIFOP
0
R/W0
1
SFD
0
R/W0
0
ACT_UNUSED
0
R/W0
210
CC253x Radio
Description
The RXENABLE register has gone from a nonzero state to an all-zero state.
0: No interrupt pending
1: Interrupt pending
A complete frame has been received.
0: No interrupt pending
1: Interrupt pending
Frame has passed frame filtering.
0: No interrupt pending
1: Interrupt pending
Source match found
0: No interrupt pending
1: Interrupt pending
Source matching complete
0: No interrupt pending
1: Interrupt pending
The number of bytes in the RXFIFO is above the threshold. Also raised when
a complete frame has been received, and when a packet has been read out
completely and there are more complete packets available.
0: No interrupt pending
1: Interrupt pending
SFD has been received or transmitted.
0: No interrupt pending
1: Interrupt pending
Reserved
0: No interrupt pending
1: Interrupt pending
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RF Core
www.ti.com
RFIRQF1 (0x91) RF Interrupt Flags
Bit
7:6
5
Name
–
CSP_WAIT
Reset
00
0
R/W
R0
R/W0
4
CSP_STOP
0
R/W0
3
CSP_MANINT
0
R/W0
2
RFIDLE
0
R/W0
1
TXDONE
0
R/W0
0
TXACKDONE
0
R/W0
Description
Reserved. Read as 0
Execution continued after a wait instruction in CSP.
0: No interrupt pending
1: Interrupt pending
CSP has stopped program execution.
0: No interrupt pending
1: Interrupt pending
Manual interrupt generated from CSP.
0: No interrupt pending
1: Interrupt pending
Radio state machine has entered the idle state.
0: No interrupt pending
1: Interrupt pending
A complete frame has been transmitted.
0: No interrupt pending
1: Interrupt pending
An acknowledgement frame has been completely transmitted.
0: No interrupt pending
1: Interrupt pending
RFERRF (0xBF) – RF Error Interrupt Flags
Bit
7
6
Name
–
STROBEERR
Reset
0
0
R/W
R0
R/W0
5
TXUNDERF
0
R/W0
4
TXOVERF
0
R/W0
3
RXUNDERF
0
R/W0
2
RXOVERF
0
R/W0
1
RXABO
0
R/W0
0
NLOCK
0
R/W0
Description
Reserved. Read as 0
A command strobe was issued at a time it could not be processed. Triggered if
trying to disable radio when already disabled, or when trying to do a SACK,
SACKPEND, or SNACK command when not in active RX.
0: No interrupt pending
1: Interrupt pending
TXFIFO underflowed
0: No interrupt pending
1: Interrupt pending
TXFIFO overflowed
0: No interrupt pending
1: Interrupt pending
RXFIFO underflowed
0: No interrupt pending
1: Interrupt pending
RXFIFO overflowed
0: No interrupt pending
1: Interrupt pending
Reception of a frame was aborted.
0: No interrupt pending
1: Interrupt pending
Frequency synthesizer failed to achieve lock after timeout, or lock is lost during
reception. Receiver must be restarted to clear this error situation.
0: No interrupt pending
1: Interrupt pending
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
211
RF Core
www.ti.com
RFIRQM0 (0x61A3) – RF Interrupt Masks
Bit
7
Name
RXMASKZERO
Reset
0
R/W
R/W
6
RXPKTDONE
0
R/W
5
FRAME_ACCEPTED
0
R/W
4
SRC_MATCH_FOUND
0
R/W
3
SRC_MATCH_DONE
0
R/W
2
FIFOP
0
R/W
1
SFD
0
R/W
0
ACT_UNUSED
0
R/W
Description
The RXENABLE register has gone from a nonzero state to an all-zero state.
0: Interrupt disabled
1: Interrupt enabled
A complete frame has been received.
0: Interrupt disabled
1: Interrupt enabled
Frame has passed frame filtering.
0: Interrupt disabled
1: Interrupt enabled
Source match found
0: Interrupt disabled
1: Interrupt enabled
Source matching complete
0: Interrupt disabled
1: Interrupt enabled
The number of bytes in the RXFIFO is above the threshold. Also raised when
a complete frame has been received, or when a complete packet has been
read out and there are more complete packets available.
0: Interrupt disabled
1: Interrupt enabled
SFD has been received or transmitted.
0: Interrupt disabled
1: Interrupt enabled
Reserved
0: Interrupt disabled
1: Interrupt enabled
RFIRQM1 (0x61A4 ) – RF Interrupt Masks
Bit
7:6
5
Name
–
CSP_WAIT
Reset
00
0
R/W
R0
R/W
4
CSP_STOP
0
R/W
3
CSP_MANINT
0
R/W
2
RFIDLE
0
R/W
1
TXDONE
0
R/W
0
TXACKDONE
0
R/W
212
CC253x Radio
Description
Reserved. Read as 0
Execution continued after a wait instruction in CSP.
0: Interrupt disabled
1: Interrupt enabled
CSP has stopped program execution.
0: Interrupt disabled
1: Interrupt enabled
Manual interrupt generated from CSP.
0: Interrupt disabled
1: Interrupt enabled
Radio state machine has entered the idle state.
0: Interrupt disabled
1: Interrupt enabled
A complete frame has been transmitted.
0: Interrupt disabled
1: Interrupt enabled
An acknowledgement frame has been completely transmitted.
0: Interrupt disabled
1: Interrupt enabled
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
FIFO Access
www.ti.com
RFERRM (0x61A5) – RF Error Interrupt Masks
Bit
7
7:6
Name
–
STROBEERR
Reset
0
0
R/W
R0
R/W
5
TXUNDERF
0
R/W
4
TXOVERF
0
R/W
3
RXUNDERF
0
R/W
2
RXOVERF
0
R/W
1
RXABO
0
R/W
0
NLOCK
0
R/W
Description
Reserved. Read as 0
A command strobe was issued at a time it could not be processed. Triggered if
trying to disable radio when already disabled, or when trying to do a SACK,
SACKPEND, or SNACK command when not in active RX.
0: Interrupt disabled
1: Interrupt enabled
TXFIFO underflowed
0: Interrupt disabled
1: Interrupt enabled
TXFIFO overflowed
0: Interrupt disabled
1: Interrupt enabled
RXFIFO underflowed
0: Interrupt disabled
1: Interrupt enabled
RXFIFO overflowed
0: Interrupt disabled
1: Interrupt enabled
Reception of a frame was aborted.
0: Interrupt disabled
1: Interrupt enabled
Frequency synthesizer failed to achieve lock after timeout, or lock is lost during
reception. Receiver must be restarted to clear this error situation.
0: Interrupt disabled
1: Interrupt enabled
23.2 FIFO Access
The TXFIFO and RXFIFO may be accessed though the SFR register RFD (0xD9). Data is written to the
TXFIFO when writing to the RFD register. Data is read from the RXFIFO when the RFD register is read.
The XREG registers RXFIFOCNT and TXFIFOCNT provide information on the amount of data in the
FIFOs. The FIFO contents can be cleared by issuing SFLUSHRX and SFLUSHTX.
RFD (0xD9) – RF Data
Bit
7:0
Name
RFD[7:0]
Reset
0x00
R/W
R/W
Description
Data written to the register is written to the TXFIFO. When reading this
register, data from the RXFIFO is read.
23.3 DMA
It is possible to use direct memory access (DMA) to move data between memory and the radio. The DMA
controller is described in Chapter 8. See this section for a detailed description on how to set up and use
DMA transfers.
To support the DMA controller, there is one DMA trigger associated with the radio, the RADIO DMA trigger
(DMA trigger 19). The RADIO DMA trigger is activated by two events. The first event to cause a RADIO
DMA trigger is when the first data is present in the RXFIFO, i.e. when the RXFIFO goes from the empty
state to a nonempty state. The second event that causes a RADIO DMA trigger is when data is read from
the RXFIFO (through RFD) and there is still more data available in the RXFIFO.
23.4 Memory Map
The RF Core contains 384 bytes of physical RAM located at addresses 0x6000 to 0x0617F. The
configuration and status registers of the RF Core are located at addresses from 0x6180 to 0x61EF.
Configuration registers, RXFIFO, and TXFIFO are all preserved during sleep modes.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
213
Memory Map
www.ti.com
23.4.1 RXFIFO
The RXFIFO memory area is located at addresses 0x6000 to 0x607F and is thus 128 bytes. Although this
memory area is intended for the RXFIFO, it is not protected in any way, so it is still accessible in the
XREG memory space. Normally, only the designated instructions should be used to manipulate the
contents of the RXFIFO. The RXFIFO can contain more than one frame at a time.
23.4.2 TXFIFO
The TXFIFO memory area is located at addresses 0x6080 to 0x60FF and is thus 128 bytes. Although this
memory area is intended for the TXFIFO, it is not protected in any way, so it is still accessible in the
XREG memory space. Normally, only the designated instructions should be used to manipulate the
contents of the TXFIFO. The TXFIFO can only contain one frame at a time.
23.4.3 Frame-Filtering and Source-Matching Memory Map
The frame-filtering and source-address-matching functions use a 128 byte block of the RF Core RAM to
store local-address information and source-matching configuration and results; this is located in the area
0x6100 to 0x617F. This memory space is described in Table 23-1. Values that do not fill an entire
byte/word are in the least-significant part of the byte/word. Note that the values in these registers are
unknown after reset. However, the values are retained during power modes.
Table 23-1. Frame Filtering and Source Matching Memory Map
ADDRESS
REGISTER/VARIABLE
0x6176–0x617F
Temporary storage
ENDIAN
DESCRIPTION
RESERVED
Memory space used for temporary storage of variables
LOCAL ADDRESS INFORMATION
0x6174–0x6175
SHORT_ADDR
LE
The short address used during destination address filtering
0x6172–0x6173
PAN_ID
LE
The PAN ID used during destination address filtering
0x616A–0x6171
EXT_ADD
LE
The IEEE extended address used during destination address filtering
SOURCE ADDRESS MATCHING CONTROL
8 MSBs of the 24-bit mask that enables/disables automatic pending
for each of the 24 short addresses
0x6169
SRCSHORTPENDEN2
0x6168
SRCSHORTPENDEN1
0x6167
SRCSHORTPENDEN0
8 LSBs of the 24-bit mask that enables/disables automatic pending
for each of the 24 short addresses
0x6166
SRCEXTPENDEN2
8 MSBs of the 24-bit mask that enables/disables automatic pending
for each of the 12 extended addresses. Entry n is mapped to
SRCEXTPENDEN[2n]. All SRCEXTPENDEN[2n + 1] bits are don't
care.
0x6165
SRCEXTPENDEN1
0x6164
SRCEXTPENDEN0
8 LSBs of the 24-bit mask that enables/disables automatic pending
for each of the 12 extended addresses. Entry n is mapped to
SRCEXTPENDEN[2n]. All SRCEXTPENDEN[2n + 1] bits are don't
care.
SOURCE ADDRESS MATCHING RESULT
0x6163
SRCRESINDEX
The bit index of the least-significant 1 in SRCRESMASK, or 0x3F when
there is no source match. On a match, bit 5 is 0 when the match is
on a short address and 1 when it is on an extended address. On a
match, bit 6 is 1 when the conditions for automatic pending bit in
acknowledgment have been met (see the description of
SRCMATCH.AUTOPEND). The bit gives no indication of whether or not
the acknowledgment actually is transmitted, and does not take the
PENDING_OR register bit and the SACK/SACKPEND/SNACK strobes
into account.
0x6162
SRCRESMASK2
24-bit mask that indicates source address match for each individual
entry in the source address table
0x6161
SRCRESMASK1
Short address matching. When there is a match on entry panid_n +
short_n, bit n is set in SRCRESMASK.
214 CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Frequency and Channel Programming
www.ti.com
Table 23-1. Frame Filtering and Source Matching Memory Map (continued)
ADDRESS
0x6160
REGISTER/VARIABLE
ENDIAN
DESCRIPTION
Extended address matching. When there is a match on entry ext_n,
bits 2n and 2n + 1 are set in SRCRESMASK.
SRCRESMASK0
SOURCE ADDRESS TABLE
0x615E–0x615F
short_23
0x615C–0x615D
panid_23
0x615A–0x615B
short_22
0x6158–0x6159
panid_22
...
...
LE
ext_11
LE
LE
LE
Two individual short-address entries (combination of 16-bit PAN ID
and 16-bit short address) or one extended address entry
...
...
LE
Two individual short address entries (combination of 16-bit PAN ID
and 16-bit short address) or one extended address entry
LE
Two individual short address entries (combination of 16-bit PAN ID
and 16-bit short address) or one extended address entry
LE
...
...
0x610E–0x610F
short_03
0x610C–0x610D
panid_03
LE
0x610A–0x610B
short_02
0x6108–0x6109
panid_02
LE
0x6106–0x6107
short_01
LE
0x6104–0x6105
panid_01
0x6102–0x6103
short_00
0x6100–0x6101
panid_00
ext_01
ext_00
LE
LE
LE
LE
LE
23.5 Frequency and Channel Programming
The carrier frequency is set by programming the 7-bit frequency word located in FREQCTRL.FREQ[6:0].
Changes take effect after the next recalibration. Carrier frequencies in the range from 2394 MHz to 2507
MHz are supported. The carrier frequency fC, in MHz, is given by fC = (2394 + FREQCTRL.FREQ[6:0])
MHz, and is programmable in 1-MHz steps.
IEEE 802.15.4-2006 specifies 16 channels within the 2.4-GHz band. They are numbered 11 through 26
and are 5 MHz apart. The RF frequency of channel k is given by Equation 4.
fc = 2405 + 5(k - 11) éëMHz ùû k Î éë11, 26 ùû
(4)
For operation in channel k, the FREQCTRL.FREQ register should therefore be set to
FREQCTRL.FREQ = 11 + 5 (k – 11).
23.6 IEEE 802.15.4-2006 Modulation Format
This section is meant as an introduction to the 2.4-GHz direct-sequence spread-spectrum (DSSS) RF
modulation format defined in IEEE 802.15.4-2006. For a complete description, see the standard document
[1].
The modulation and spreading functions are illustrated at the block level in Figure 23-1. Each byte is
divided into two symbols, 4 bits each. The least-significant symbol is transmitted first. For multibyte fields,
the least-significant byte is transmitted first, except for security-related fields, where the most-significant
byte is transmitted first.
Each symbol is mapped to one out of 16 pseudorandom sequences, 32 chips each. The symbol-to-chip
mapping is shown in Table 23-2. The chip sequence is then transmitted at 2 Mchips/s, with the leastsignificant chip (C0) transmitted first for each symbol. The transmitted bit stream and the chip sequences
are observable on GPIO pins P1[0:5]. See Chapter 7 for details on how to configure the GPIO to do this.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
215
IEEE 802.15.4-2006 Modulation Format
www.ti.com
62. 5 ksymbol/s
250 kbps
2 Mchips/s
1 Mchips/s
I
Transmitted
Bit-Stream
(LSB First)
Bit-toSymbol
Symbolto-Chip
O-QPSK
Modulator
Q
Modulated
Signal
( to DACs)
1 Mchips/s
B0306-01
Figure 23-1. Modulation
Table 23-2. IEEE 802.15.4-2006 Symbol-to-Chip Mapping
Symbol
Chip Sequence (C0, C1, C2, … , C31)
0
11011001110000110101001000101110
1
11101101100111000011010100100010
2
00101110110110011100001101010010
3
00100010111011011001110000110101
4
01010010001011101101100111000011
5
00110101001000101110110110011100
6
11000011010100100010111011011001
7
10011100001101010010001011101101
8
10001100100101100000011101111011
9
10111000110010010110000001110111
10
01111011100011001001011000000111
11
01110111101110001100100101100000
12
00000111011110111000110010010110
13
01100000011101111011100011001001
14
10010110000001110111101110001100
15
11001001011000000111011110111000
The modulation format is offset – quadrature phase shift keying (O-QPSK) with half-sine chip shaping.
This is equivalent to MSK modulation. Each chip is shaped as a half-sine, transmitted alternately in the I
and Q channels with one-half chip-period offset. This is illustrated for the zero-symbol in Figure 23-2.
tC
I-Phase
Q-Phase
0
1
1
0
1
1
0
0
1
1
1
0
0
0
1
0
1
0
1
0
0
1
0
1
0
1
0
1
0
1
1
0
2tC
M0107-01
Figure 23-2. I/Q Phases When Transmitting a Zero-Symbol Chip Sequence, tC = 0.5 μs
216
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
IEEE 802.15.4-2006 Frame Format
www.ti.com
23.7 IEEE 802.15.4-2006 Frame Format
This section gives a brief summary of the IEEE 802.15.4 frame format [1]. The radio has built-in support
for processing of parts of the frame. This is described in the following sections.
Figure 23-3 shows a schematic view of the IEEE 802.15.4 frame format. Similar figures describing specific
frame formats (data frames, beacon frames, acknowledgment frames, and MAC command frames) are
included in the standard document [1].
Bytes:
2
Frame
Control Field
(FCF)
MAC
Layer
Bytes:
PHY
Layer
4
1
Preamble
Sequence
Frame
Length
1
Start-of-Frame
Delimiter
(SFD)
Synchronization Header
(SHR)
1
0 to 20
2
Frame Check
Sequence
(FCS)
MAC Footer
(MFR)
n
Data
Address
Sequence
Information
Number
MAC Header (MHR)
Frame Payload
MAC Payload
5 + (0 to 20) + n
MAC Protocol
Data Unit
(MPDU)
PHY Service Data Unit
(PSDU)
PHY Header
(PHR)
11 + (0 to 20) + n
PHY Protocol Data Unit
(PPDU)
M0108-01
Figure 23-3. Schematic View of the IEEE 802.15.4 Frame Format [1]
23.7.1 PHY Layer
Synchronization Header
The synchronization header (SHR) consists of the preamble sequence followed by the start-of-frame
delimiter (SFD). In the IEEE 802.15.4 specification [1], the preamble sequence is defined to be 4 bytes of
0x00. The SFD is one byte with value 0xA7.
PHY Header
The PHY header consists only of the frame-length field. The frame-length field defines the number of
bytes in the MPDU. Note that the value of the frame-length field does not include the frame-length field
itself. It does, however, include the frame-check sequence (FCS), even if this is inserted automatically by
the hardware.
The frame-length field is 7 bits long and has a maximum value of 127. The most-significant bit in the
frame-length field is reserved, and should always be set to zero.
PHY Service Data Unit
The PHY service data unit (PSDU) contains the MAC protocol data unit (MPDU). It is the MAC layer's
function to generate/interpret the MPDU, and the radio has built-in support for processing of some of the
MPDU subfields.
23.7.2 MAC Layer
The FCF, data sequence number, and address information follow the frame-length field as shown in
Figure 23-3. Together with the MAC data payload and frame check sequence, they form the MPDU. The
format of the FCF is shown in Figure 23-4. For full details, see the IEEE 802.15.4 specification [1].
Bits: 0–2
Frame type
3
Security
enabled
4
Frame
pending
5
Acknowledge
request
6
Intra PAN
7–9
Reserved
10–11
Destination
addressing
mode
12–13
Reserved
14–15
Source
addressing
mode
Figure 23-4. Format of the Frame Control Field (FCF)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
217
Transmit Mode
www.ti.com
Frame-Check Sequence
A 2-byte frame-check sequence (FCS) follows the last MAC payload byte as shown in Figure 23-3. The
FCS is calculated over the MPDU, i.e., the frame-length field is not part of the FCS.
The FCS polynomial defined in [1] is
G(s) = x16 + x12 + x5 + 1
The radio supports automatic calculation/verification of the FCS. See Section 23.8.10 for details.
23.8 Transmit Mode
This section describes how to control the transmitter, how to control the integrated frame processing, and
how to use the TXFIFO.
23.8.1 TX Control
The radio has many built-in features for frame processing and status reporting. Note that the radio
provides features that make it easy to have precise control of the timing of outgoing frames. This is very
important in an IEEE 802.15.4/ ZigBee® system, because there are strict timing requirements to such
systems.
Frame transmission is started by the following actions:
• The STXON command strobe
– The SAMPLED_CCA signal is not updated.
•
The STXONCCA command strobe, provided that the CCA signal is high.
– Aborts ongoing transmission/reception and forces a TX calibration followed by transmission.
– The SAMPLED_CCA signal is updated.
Clear channel assessment is described in detail in Section 23.8.12.
Frame transmission is aborted by the following command actions:
• The SRXON command strobe
– Aborts ongoing transmission and forces an RX calibration
• The SRFOFF command strobe
– Aborts ongoing transmission/reception and forces the FSM to the IDLE state.
• The STXON command strobe
– Aborts ongoing transmission and forces an RX calibration
To enable the receiver after transmission with STXON, the FRMCTRL1.SET_RXENMASK_ON_TX bit should
be set. This sets bit 6 in RXENABLE when STXON is executed. When transmitting with STXONCCA, the
receiver is on before the transmission and is turned back on afterwards (unless the RXENABLE registers
have been cleared in the meantime).
23.8.2 TX State Timing
Transmission of preamble begins 192 μs after the STXON or STXONCCA command strobe. This is
referred to as TX turnaround time in [1]. There is an equal delay when returning to receive mode.
When returning to idle or receive mode, there is a 2-μs delay while the modulator ramps down the signals
to the DACs. The down ramping happens automatically after the complete MPDU (as defined by the
length byte) has been transmitted or if TX underflow occurs. This affects:
• The SFD signal, which is stretched by 2 μs.
• The radio FSM transition to the IDLE state, which is delayed by 2 μs.
23.8.3 TXFIFO Access
The TXFIFO can hold 128 bytes and only one frame at a time. The frame can be buffered before or after
the TX command strobe is executed, as long as it does not generate TX underflow (see the error
conditions listed in Section 23.8.5).
218
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Transmit Mode
www.ti.com
Figure 23-5 illustrates what must be written to the TXFIFO (marked blue). Additional bytes are ignored,
unless TX overflow occurs (see the error conditions listed in Section 23.8.5).
AUTOCRC = 0 LEN
LEN –2 Bytes
AUTOCRC = 1 LEN
LEN –2 Bytes
FCS
(2 Bytes)
–
(Ignored)
–
(Ignored)
M0109-01
Figure 23-5. Frame Data Written to the TXFIFO
There are two ways to write to the TXFIFO.
• Write to the RFD register.
•
Frame buffering always begins at the start of the TXFIFO memory. By enabling the
FRMCTRL1.IGNORE_TX_UNDERF bit, it is possible to write directly into the RAM area in the radio
memory, which holds the TXFIFO. Note that it is recommended to use the RFD register for writing data
to the TXFIFO.
The number of bytes in the TXFIFO is stored in the TXFIFOCNT register.
The TXFIFO can be emptied manually with the SFLUSHTX command strobe. TX underflow occurs if the
FIFO is emptied during transmission.
23.8.4 Retransmission
In order to support simple retransmission of frames, the radio does not delete the TXFIFO contents as
they are transmitted. After a frame has been successfully transmitted, the FIFO contents are left
unchanged. To retransmit the same frame, simply restart TX by issuing an STXON or STXONCCA
command strobe. Note that a retransmission of a packet is only possible if the packet has been completely
transmitted; i.e., a packet cannot be aborted and then be retransmitted.
If a different frame is to be transmitted, issue an ISFLUSHTX strobe and then write the new frame to the
TXFIFO.
23.8.5 Error Conditions
There are two error conditions associated with the TXFIFO:
• Overflow happens when the TXFIFO is full and another byte write is attempted.
• Underflow happens when the TXFIFO is empty and the radio attempts to fetch another byte for
transmission.
TX overflow is indicated by the TX_OVERFLOW interrupt flag being set. When this error occurs, the
writing is aborted, i.e., the data byte that caused the overflow is lost. The error condition must be cleared
with the SFLUSHTX strobe.
TX underflow is indicated by the TX_UNDERFLOW interrupt flag being set. When this error occurs, the
ongoing transmission is aborted. The error condition must be cleared with the SFLUSHTX strobe.
The TX_UNDERFLOW exception can be disabled by setting the FRMCTRL1.IGNORE_TX_UNDERF bit. In
this case, the radio continues transmitting the bytes that happen to be in the TXFIFO memory, until the
number of bytes given by the first byte (i.e., the length byte) has been transmitted.
23.8.6 TX Flow Diagram
Figure 23-6 summarizes the previous sections in a flow diagram:
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
219
Transmit Mode
www.ti.com
No CSMA-CA
Unslotted CSMA-CA
Data buffering
Slotted CSMA-CA
SSAMPLECCA
Yes
(SAMPLED_CCA = 1)
STXON
Success?
No
(SAMPLED_CCA = 0)
STXONCCA
Yes
(SAMPLED_CCA = 1)
Write a frame to the
TX buffer using:
- TXBUF
- TXBUFCP
- Memory access
- A combination of
these methods
This can be done
before, after, or in
parallel with the TX
strobe.
No
(SAMPLED_CCA = 0)
TX started?
TX buffer overfilled
TX completes?
No
TX is aborted by
SRXON,
STXON or SRFOFF
Why?
Yes
TX_FRM_DONE
TX_UNDERFLOW
Frame transmitted successfully
Error condition
TX_OVERFLOW
Incomplete or no frame transmission
TIME
Error condition
(left side of the flow
diagram should be
ignored because the
TX buffer is corrupted.)
Between two transmissions, there can be multiple other activities such as frame reception, RX FIFO access, and acknowledgment transmission (using SACK,
SACKPEND, or AUTOACK), or idle periods (random backoffs). This has no side effects on the state of the TX buffer.
The placement of the SFLUSHTX strobe in the diagram shows the latest point in time where this strobe can be executed. If fewer special cases is desired, it is
always possible to use the SFLUSHTX strobe and then load or reload TXBUF with the next frame to be transmitted.
Next time...
Next time...
To retransmit or
transmit a
different frame...
To retransmit the
current frame...
To transmit a
different frame...
Restart from the
top of the diagram
Restart from the
top of the diagram
Do not write
anything to the TX
buffer
Write the new
frame to the TX
buffer
(before, after, or in
parallel with the
TX strobe)
SFLUSHTX
Restart from the
top of the diagram
To retransmit or
transmit a
different frame...
To (re)transmit
what is
currently in
the TX buffer...
To transmit a
different frame...
Restart from the
top of the diagram
SFLUSHTX
SFLUSHTX
Restart from the
top of the diagram
Restart from the
top of the diagram
Write the new
frame to the TX
buffer
before, after, or in
parallel with the
TX strobe)
Write the next
frame to the TX
buffer
before, after, or in
parallel with the
TX strobe)
If anything is
written to the TX
buffer, it is
appended to the
current data.
Write the next
frame to the TX
buffer
(before, after, or in
parallel with the
TX strobe)
F0035-01
Figure 23-6. TX Flow
220
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Transmit Mode
www.ti.com
23.8.7 Transmitted Frame Processing
The radio performs the following frame generation tasks for TX frames:
Received Frame
Preamble
SFD
LEN
(1)
MAC Payload
MHR
FCS
(2)
(3)
M0110-01
(1)
Generation and automatic transmission of the PHY layer synchronization header, which consists of the
preamble and the SFD
(2)
Transmission of the number of bytes specified by the frame-length field
(3)
Calculation of and automatic transmission of the FCS (can be disabled)
Figure 23-7. Single Transmitted Frame
The recommended usage is to write the frame-length field followed by the MAC header and MAC payload
to the TXFIFO and let the radio handle the rest. Note that the frame-length field must include the two FCS
bytes, even though the radio adds these automatically.
23.8.8 Synchronization Header
Synchronization Header
Preamble
1 Symbol
IEEE 802.15.4
0
0
SFD
1 Byte
0
0
0
0
0
0
7
A
2 (PREAMBLE_LENGTH + 2) Zero Symbols
CC2530
M0111-02
Figure 23-8. Transmitted Synchronization Header
The radio has programmable preamble length. The default value is compliant with [1], and changing the
value makes the system noncompliant to IEEE 802.15.4.
The preamble sequence length is set by MDMCTRL0.PREAMBLE_LENGTH. Figure 23-8 shows how the
synchronization header relates to the IEEE 802.15.4 specification.
When the required number of preamble bytes has been transmitted, the radio automatically transmits the
1-byte SFD. The SFD is fixed, and it is not possible to change this value from software.
23.8.9 Frame-Length Field
When the SFD has been transmitted, the modulator starts to read data from the TXFIFO. It expects to find
the frame-length field followed by the MAC header and MAC payload. The frame-length field is used to
determine how many bytes are to be transmitted.
Note that the minimum frame length is 3 bytes when AUTOCRC = 1 and 1 byte when AUTOCRC = 0.
23.8.10 Frame Check Sequence
When the FRMCTRL0.AUTOCRC control bit is set, the FCS field is automatically generated and appended
to the transmitted frame at the position defined by the frame-length field. The FCS is not written to the
TXFIFO, but stored in a separate 16-bit register. It is recommended always to have AUTOCRC enabled,
except possibly for debug purposes. If FRMCTRL0.AUTOCRC = 0, then the modulator expects to find the
FCS in the TXFIFO, so software must generate the FCS and write it to the TXFIFO along with the rest of
the MPDU.
The hardware implementation of the FCS calculator is shown in Figure 23-9. See [1] for further details.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
221
Receive Mode
Data Input
(LSB First)
www.ti.com
+
+
+
r0
r1
r2
r4
r3
r5
r6
r7
r8
r9
r10
r11
r12
r13
r14
r15
B0307-01
Figure 23-9. FCS Hardware Implementation
23.8.11 Interrupts
The SFD interrupt is raised when the SFD field of the frame has been transmitted. At the end of the frame,
the TX_FRM_DONE interrupt is raised when the complete frame has been successfully transmitted.
Note that there is a second SFD signal available on GPIO (through radio observation mux) that should not
be confused with the SFD interrupt.
23.8.12 Clear-Channel Assessment
The clear-channel assessment (CCA) status signal indicates whether the channel is available for
transmission or not. The CCA function is used to implement the CSMA-CA functionality specified in the
IEEE 802.15.4 specification [1]. The CCA signal is valid when the receiver has been enabled for at least
eight symbol periods. The RSSI_VALID status signal can be used to verify this.
The CCA is based on the RSSI value and a programmable threshold. The exact behavior is configurable
in the CCACTRL0 and CCACTRL1 registers.
There are two variations of the CCA signal, one that is updated at every new RSSI sample and one that is
only updated on SSAMPLECCA/ISAMPLECCA and STXONCCA/ISTXONCCA command strobes. They
are both available in the FSMSTAT1 register.
Note that the CCA signal is updated four clock cycles (system clock) after the RSSI_VALID signal has
been set.
23.8.13 Output Power Programming
The RF output power is controlled by the 7-bit value in the TXPOWER register. The device data sheet
(Appendix C) shows typical output power and current consumption for recommended settings when the
center frequency is set to 2.440 GHz. Note that the recommended settings are only a small subset of all
the possible register settings.
23.8.14 Tips and Tricks
•
•
Note that there is no requirement to have the complete frame in the TXFIFO before starting a
transmission. Bytes may be added to the TXFIFO during transmission.
It is possible to transmit non-IEEE 802.15.4 compliant frames by setting
MDMTEST1.MODULATION_MODE = 1.
23.9 Receive Mode
This section describes how to control the receiver, control the integrated RX frame processing, and how to
use the RXFIFO.
23.9.1 RX Control
The receiver is turned on and off with the SRXON and SRFOFF command strobes, and with the
RXENABLE registers. The command strobes provide a hard on/off mechanism, whereas RXENABLE
manipulation provides a soft on/off mechanism.
The receiver is turned on by the following actions:
• The SRXON strobe:
222
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Receive Mode
www.ti.com
– Sets RXENABLE[7]
•
•
– Aborts ongoing transmission/reception by forcing a transition to RX calibration.
The STXON strobe, when FRMCTRL1.SET_RXENMASK_ON_TX is enabled:
– Sets RXENABLE[6]
– The receiver is enabled after transmission completes.
Setting RXENABLE != 0x00 by writing to RXENMASKOR:
– Does not abort ongoing transmission/reception.
The receiver is turned off by the following actions:
• The SRFOFF strobe:
– Clears RXENABLE[7:0]
•
– Aborts ongoing transmission/reception by forcing the transition to IDLE mode.
Setting RXENABLE = 0x00 by writing to RXENMASKAND
– Does not abort ongoing transmission/reception. Once the ongoing transmission/reception is
finished, the radio returns to the IDLE state.
There are several ways to manipulate the RXENABLE registers:
• The SRXMASKBITSET and SRXMASKBITCLR strobes (affecting RXENABLE[5])
• The SRXON, SRFOFF and STXON strobes, including the FRMCTRL1.SET_RXMASK_ON_TX setting
23.9.2 RX State Timing
The receiver is ready 192 μs after RX has been enabled by one of the methods described in
Section 23.9.1. This is referred to as RX turnaround time in [1].
When returning to receive mode after frame reception, there is by default an interval of 192 μs where SFD
detection is disabled. This interval can be disabled by clearing FSMCTRL.RX2RX_TIME_OFF.
23.9.3 Received Frame Processing
The radio integrates critical portions of the RX requirements in IEEE 802.15.4-2003 and -2006 in
hardware. This reduces the CPU interruption rate, simplifies the software that handles frame reception,
and provides the results with minimum latency.
During reception of a single frame, the following frame-processing steps are performed:
Received Frame
Preamble
SFD
(1)
LEN
MHR
(2)
Transmitted Acknowledgment Frame
MAC Payload
(3)
FCS
Preamble
(4)
SFD
LEN
MHR
FCS
(5)
M0110-02
(1)
Detection and removal of the received PHY synchronization header (preamble and SFD), and reception of
the number of bytes specified by the frame-length field.
(2)
Frame filtering as specified by [1], section 7.5.6.2, third filtering level.
(3)
Matching of the source address against a table containing up to 24 short addresses or 12 extended IEEE
addresses. The source address table is stored in the radio RAM.
(4)
Automatic FCS checking, and attaching this result and other status values (RSSI, correlation and source
match result) to received frames.
(5)
Automatic acknowledgment transmission with correct timing, and correct setting of the frame pending bit,
based on the results from source address matching and FCS checking.
Figure 23-10. Single Received Frame and Transmitted Acknowledgement Frame
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
223
Receive Mode
www.ti.com
23.9.4 Synchronization Header and Frame-Length Fields
Frame reception starts with detection of a start-of-frame delimiter (SFD), followed by the length byte,
which determines when the reception is complete. The SFD signal, which can be output on GPIO, can be
used to capture the start of received frames:
Received Frame
Preamble
SFD
MPDU (LEN[6:0] Bytes)
LEN
SFD (Accepted Frame)
SFD (Rejected Frame)
Frame Rejected
T0319-01
Figure 23-11. SFD Signal Timing
Preamble and SFD are not written to the RXFIFO.
The radio uses a correlator to detect the SFD. The correlation threshold value in MDMCTRL1.CORR_THR
determines how closely the received SFD must match an ideal SFD. The threshold must be adjusted with
care:
• If set too high, the radio misses many actual SFDs, effectively reducing the receiver sensitivity.
• If set too low, the radio detects many false SFDs. Although this does not reduce the receiver
sensitivity, the effect is similar, because false frames might overlap with the SFDs of actual frames. It
also increases the risk of receiving false frames with correct FCS.
In addition to SFD detection, it is also possible to require a number of valid preamble symbols (also above
the correlation threshold) prior to SFD detection. See the register descriptions of MDMCTRL0 and
MDMCTRL1 for available options and recommended settings.
23.9.5 Frame Filtering
The frame filtering function rejects nonintended frames as specified by [1], section 7.5.6.2, third filtering
level. In addition, it provides filtering on:
• The eight different frame types (see the FRMFILT1 register)
•
The reserved bits in the frame control field (FCF)
The function is controlled by:
• The FRMFILT0 and FRMFILT1 registers
•
The PAN_ID, SHORT_ADDR and EXT_ADDR values in RAM
Filtering Algorithm
The FRMFILT0.FRM_FILTER_EN bit controls whether frame filtering is applied or not. When disabled, the
radio accepts all received frames. When enabled (which is the default setting), the radio only accepts
frames that fulfill all of the following requirements:
• The length byte must be equal to or higher than the minimum frame length, which is derived from the
source- and destination-address mode and PAN ID compression subfields of the FCF.
• The reserved FCF bits [9:7] ANDed together with FRMFILT0.FCF_RESERVED_BITMASK must equal
000b.
• The value of the frame version subfield of the FCF cannot be higher than
FRMFILT0.MAX_FRAME_VERSION.
•
•
224
The source and destination address modes cannot be reserved values (1).
Destination address:
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Receive Mode
www.ti.com
•
– If a destination PAN ID is included in the frame, it must match PAN_ID or must be the broadcast
PAN identifier (0xFFFF).
– If a short destination address is included in the frame, it must match either SHORT_ADDR or the
broadcast address (0xFFFF).
– If an extended destination address is included in the frame, it must match EXT_ADDR.
Frame type:
– Beacon frames (0) are only accepted when:
• FRMFILT1.ACCEPT_FT0_BEACON = 1
• Length byte ≥ 9
• The destination address mode is 0 (no destination address).
• The source address mode is 2 or 3 (i.e., a source address is included).
• The source PAN ID matches PAN_ID, or PAN_ID equals 0xFFFF.
– Data (1) frames are only accepted when:
• FRMFILT1.ACCEPT_FT1_DATA = 1
Length byte ≥ 9
A destination address and/or source address is included in the frame. If no destination address
is included in the frame, the FRMFILT0.PAN_COORDINATOR bit must be set, and the source
PAN ID must equal PAN_ID.
– Acknowledgment (2) frames are only accepted when:
• FRMFILT1.ACCEPT_FT2_ACK = 1
•
•
• Length byte = 5
– MAC command (3) frames are only accepted when:
• FRMFILT1.ACCEPT_FT3_MAC_CMD = 1
Length byte ≥ 9
A destination address and/or source address is included in the frame. If no destination address
is included in the frame, the FRMFILT0.PAN_COORDINATOR bit must be set, and the source
PAN ID must equal PAN_ID for the frame to be accepted.
– Reserved frame types (4, 5, 6, and 7) are only accepted when
• FRMFILT1.ACCEPT_FT4TO7_RESERVED = 1 (default is 0)
•
•
•
Length byte ≥ 9
The following operations are performed before the filtering begins, with no effect on the frame data stored
in the RXFIFO:
• Bit 7 of the length byte is masked out (don't care).
• If FRMFILT1.MODIFY_FT_FILTER is not equal to zero, the MSB of the frame type subfield of the FCF
is either inverted or forced to 0 or 1.
If a frame is rejected, the radio only starts searching for a new frame after the rejected frame has been
completely received (as defined by the frame-length field) to avoid detecting false SFDs within the frame.
Note that a rejected frame can generate RX overflow if it occurs before the frame is rejected.
Interrupts
When frame filtering is enabled and the filtering algorithm accepts a received frame, an
RX_FRM_ACCEPTED interrupt is generated. It is not generated if frame filtering is disabled or
RX_OVERFLOW or RX_FRM_ABORTED is generated before the filtering result is known.
Figure 23-12 illustrates the three different scenarios (not including the overflow and abort-error conditions).
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
225
Receive Mode
www.ti.com
SFD
LEN
FCF + SEQ + Destination +
Source PAN ID
Remainder of Received Frame
Filtering is Enabled, Frame Rejected
SFD
Interrupt
Frame
Rejected
SFD Search
Resumed
Filtering is Enabled, Frame Accepted
FIFOP interrupt occurs during this interval
(Depending on FIFOPCTRL Value)
SFD
Interrupt
RX_FRM_ACCEPTED
Interrupt
RX_FRM_DONE
Interrupt
Filtering is Disabled
FIFOP interrupt occurs during this interval
(Depending on FIFOPCTRL Value)
SFD
Interrupt
RX_FRM_DONE
Interrupt
M0112-01
Figure 23-12. Filtering Scenarios (Exceptions Generated During Reception)
The FSMSTAT1.SFD register bit goes high when a start-of-frame delimiter is completely received and
remains high until either the last byte in MPDU is received or the received frame has failed to pass
address recognition and been rejected.
Tips and Tricks
The following register settings must be configured correctly:
• FRMFILT0.PAN_COORDINATOR must be set if the device is a PAN coordinator, and cleared if not.
• FRMFILT0.MAX_FRAME_VERSION must correspond to the supported version(s) of the IEEE 802.15.4
standard.
• The local address information must be loaded into RAM.
To avoid completely the receiving of frames during energy-detection scanning, set FRMCTRL0.RX_MODE =
11b and then (re)start RX. This disables symbol search and thereby prevents SFD detection.
To resume normal RX mode, set FRMCTRL0.RX_MODE = 00b and (re)start RX.
During operation in a busy IEEE 802.15.4 environment, the radio receives large numbers of nonintended
acknowledgment frames. To block reception of these frames effectively, use the
FRMFILT1.ACCEPT_FT2_ACK bit to control when acknowledgment frames should be received:
• Set FRMFILT1.ACCEPT_FT2_ACK after successfully starting a transmission with acknowledgment
request, and clear the bit again after the acknowledgment frame has been received or the time-out has
226
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Receive Mode
www.ti.com
•
been reached.
Keep the bit cleared otherwise.
It is not necessary to turn off the receiver while changing the values of the FRMFILT0/1 registers and the
local address information stored in RAM. However, if the changes take place between reception of the
SFD byte and the source PAN ID (i.e., between the SFD and RX_FRM_ACCEPTED exceptions), the
modified values must be considered as don't care for that particular frame (the radio uses either the old or
the new value).
Note that it is possible to make the radio ignore all IEEE 802.15.4 incoming frames by setting
MDMTEST1.MODULATION_MODE = 1.
23.9.6 Source Address Matching
The radio supports matching of the source address in received frames against a table stored in the onchip memory. The table is 96 bytes long, and hence it can contain up to:
• 24 short addresses (2 byte pan id + 2 byte short address).
• 12 IEEE extended addresses (8 bytes each).
Source address matching is only performed when frame filtering is also enabled and the received frame
has been accepted. The function is controlled by:
• The SRCMATCH, SRCSHORTEN0, SRCSHORTEN1, SRCSHORTEN2, SRCEXTEN0, SRCEXTEN1, and
SRCEXTEN2 registers
•
The source address table in RAM
Applications
Automatic acknowledgment transmission with correct setting of the frame-pending bit: When using indirect
frame transmission, the devices send data requests to poll frames stored on the coordinator. To indicate
whether it actually has a frame stored for the device, the coordinator must set or clear the frame-pending
bit in the returned acknowledgment frame. On most 8- and 16-bit MCUs, however, there is not enough
time to determine this, and so the coordinator ends up setting the pending bit regardless of whether there
are pending frames for the device (as required by IEEE 802.15.4 [1]). This is wasteful in terms of power
consumption, because the polling device must keep its receiver enabled for a considerable period of time,
even if there are no frames for it. By loading the destination addresses in the indirect frame queue into the
source address table and enabling the AUTOPEND function, the radio sets the pending bit in outgoing
acknowledgment frames automatically. This way, the operation is no longer timing-critical, as the effort
done by the microcontroller is when adding or removing frames in the indirect frame queue and updating
the source address table accordingly.
Security material look-up: To reduce the time needed to process secured frames, the source address
table can be set up so the entries match the table of security keys on the CPU. A second level of masking
on the table entries allows this application to be combined with automatic setting of the pending bit in
acknowledgment frames.
Other applications: The two previous applications are the main targets for the source-address matching
function. However, for proprietary protocols that only rely on the basic IEEE 802.15.4 frame format, there
are several other useful applications. For instance, it is possible to create firewall functionality where only
a specified set of nodes is to be acknowledged.
The Source Address Table
The source address table begins at address 0x6100 in RAM. The space is shared between short and
extended addresses, and the SRCSHORTEN0/1/2 and SRCEXTEN0/1/2 registers are used to control
which entries are enabled. All values in the table are little-endian (as in the received frames).
• A short address entry starts with the 16-bit PAN ID followed by the 16-bit short address. These entries
are stored at address 0x6100 + (4 × n), where n is a number between 0 and 23.
• An extended address entry consists only of the 64-bit IEEE extended address. These entries are
stored at address 0x6100 + (8 × n), where n is a number between 0 and 11.
Address Enable Registers
Software is responsible for allocating table entries and for making sure that active short and extended
address entries do not overlap. There are separate enable bits for short and extended addresses:
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
227
Receive Mode
•
•
www.ti.com
Short address entries are enabled in the SRCSHORTEN0, SRCSHORTEN1, and SRCSHORTEN2 registers.
Register bit n corresponds to short address entry n.
Extended address entries are enabled in the SRCEXTEN0, SRCEXTEN1, and SRCEXTEN2 registers. In
this case, register bit 2n corresponds to extended address entry n. This mapping is convenient when
creating a combined bit vector (of short and extended enable bits) to find unused entries. Moreover,
when read, register bit 2n + 1 always has the same value as register bit 2n, because an extended
address occupies the same memory as two short-address entries.
Matching Algorithm
The SRCMATCH.SRC_MATCH_EN bit controls whether source address matching is enabled or not. When
enabled (which is the default setting) and a frame passes the filtering algorithm, the radio applies one of
the algorithms outlined in Figure 23-15, depending on which type of source address is present.
The result is reported in two different forms:
• A 24-bit vector called SRCRESMASK contains a 1 for each enabled short entry with a match, or two
1s for each enabled extended entry with a match (the bit mapping is the same as for the addressenable registers on read access).
• A 7-bit value called SRCRESINDEX:
– When no source address is present in the received frame, or there is no match on the received
source address:
• Bits 6:0: 011 1111
– If there is a match on the received source address:
• Bits 4:0: The index of the first entry (i.e., the one with the lowest index number) with a match,
0–23 for short addresses or 0–11 for extended addresses.
• Bit 5: 0 if the match is on a short address, 1 if the match is on an extended address
• Bit 6: The result of the AUTOPEND function
Short Source Address (Mode 2)
Extended Source Address (Mode 3)
The received source PAN ID is called srcPanid. The received short
address is called srcShort.
The received extended address is called srcExt.
SRCRESMASK = 0x000000;
SRCRESINDEX = 0x3F;
for (n = 0; n < 24; n++) {
bitVector = 0x000001 << n;
if (SRCSHORTEN & bitVector) {
if ((panid[n] == srcPanid) &&
(short[n] == srcShort)) {
SRCRESMASK |= bitVector;
if (SRCRESINDEX == 0x3F) {
SRCRESINDEX = n;
}
}
}
}
SRCRESMASK = 0x000000;
SRCRESINDEX = 0x3F;
for (n = 0; n < 12; n++) {
bitVector = 0x000003 << (2*n);
if (SRCEXTEN & bitVector) {
if (ext[n] == srxExt) {
SRCRESMASK |= bitVector;
if (SRCRESINDEX == 0x3F) {
SRCRESINDEX = n | 0x20;
}
}
}
}
Figure 23-13. Matching Algorithm for Short and Extended Addresses
SRCRESMASK and SRCRESINDEX are written to RF Core memory as soon as the result is available.
SRCRESINDEX is also appended to received frames if the FRMCTRL0.AUTOCRC and
FRMCTRL0.APPEND_DATA_MODE bits have been set. The value then replaces the 7-bit correlation value
of the 16-bit status word.
Interrupts
When source address matching is enabled and the matching algorithm completes, the
SRC_MATCH_DONE interrupt flag is set, regardless of the result. If a match is found, the
SRC_MATCH_FOUND flag is also set immediately before SRC_MATCH_DONE.
Figure 23-14 illustrates the timing of the setting of flags:
228
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Receive Mode
www.ti.com
When There Is No Source Address:
SFD
LEN
Last
Byte
-----
FCF + SEQ + Destination
SRC_MATCH_DONE
Interrupt
SFD
Interrupt
RX_FRM_ACCEPTED
Interrupt
RX_FRM_DONE
Interrupt
When There Is a Source Address:
SFD
LEN
FCF + SEQ + Destination +
Source PAN ID
Source
Address
Last
Byte
-----
SRC_MATCH_FOUND interrupt
may occur during this interval
SRC_MATCH_DONE interrupt
occurs during this interval
SFD
Interrupt
RX_FRM_ACCEPTED
Interrupt
RX_FRM_DONE
Interrupt
M0113-01
Figure 23-14. Interrupts Generated by Source Address Matching
Tips and Tricks
• The source address table can be modified safely during frame reception. If one address replaces
another while the receiver is active, the corresponding enable bit should be turned off during the
modification. This prevents the RF Core from using a combination of old and new values, because it
only considers entries that are enabled throughout the whole source matching process.
The following measures can be taken to avoid the next received frame overwriting the results from source
address matching:
• Use the appended SRCRESINDEX result instead of the value written to RAM (this is the
recommended approach).
• Read the results from RAM before RX_FRM_ACCEPTED occurs in the next received frame. For the
shortest frame type, this happens after the sequence number, so the total available time (absolute
worst-case with a small safety margin) becomes:
16 μs (required preamble) + 32 μs (SFD) + 128 μs (4 bytes) = 176 μs
• To increase the available time, clear the FSMCTRL.RX2RX_TIME_OFF bit. This adds another 192 μs,
for a total of 368 μs. This also reduces the risk of RX overflow.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
229
Receive Mode
www.ti.com
23.9.7 Frame-Check Sequence
In receive mode, the FCS is verified by hardware if FRMCTRL0.AUTOCRC is enabled. The user is normally
only interested in the correctness of the FCS, not the FCS sequence itself. The FCS sequence itself is
therefore not written to the RXFIFO during receive. Instead, when FRMCTRL0.AUTOCRC is set, the two
FCS bytes are replaced by other more-useful values. The values that are substituted for the FCS
sequence are configurable in the FRMCTRL0 register.
FRMCTRL0 Settings
Data in RXFIFO
Length Byte
AUTOCRC = 0
MPDU
AUTOCRC = 1 and
APPEND_DATA_MODE = 0
7
AUTOCRC = 1 and
APPEND_DATA_MODE = 1
7
MPDU2
MPDU1
n
6
5
3
4
2
7
0
1
5
4
3
2
1
0
7
5
6
CRC
OK
RSSI (Signed 2s Complement)
6
MPDUn–2
••••
4
3
FCS2
2
1
0
Correlation Value (Unsigned)
6
5
CRC
OK
RSSI (Signed 2s Complement)
FCS1
4
3
2
1
0
SRCRESINDEX
M0114-01
Figure 23-15. Data in RXFIFO for Different Settings
Field Descriptions:
• The RSSI value is measured over the first eight symbols following the SFD.
• The CRC_OK bit indicates whether the FCS is correct (1) or incorrect (0). When incorrect, software is
responsible for discarding the frame.
• The correlation value is the average correlation value over the first eight symbols following the SFD.
• SRCRESINDEX is the same value that is written to RAM after completion of source address matching.
Calculation of the LQI value used by IEEE 802.15.4 is described in Section 23.10.4.
23.9.8 Acknowledgement Transmission
The radio includes hardware support for acknowledgment transmission after successful frame reception
(i.e., the FCS of the received frame must be correct). Figure 23-16 shows the format of the
acknowledgment frame.
Bytes:
4
1
Preamble
Sequence
Start-of-Frame
Delimiter
(SFD)
Synchronization Header
(SHR)
1
2
Frame
Length
Frame
Control Field
(FCF)
PHY Header
(PHR)
1
2
Data
Sequence
Number
Frame Check
Sequence
(FCS)
MAC Header (MHR)
MAC Footer
(MFR)
M0115-01
Figure 23-16. Acknowledge Frame Format
There are three variable fields in the generated acknowledgment frame:
• The pending bit, which may be controlled with command strobes and the AUTOPEND feature (in the
FCF field)
• The data sequence number (DSN), which is taken automatically from the last received frame
• The FCS, which is given implicitly
230
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Receive Mode
www.ti.com
There are three different sources for setting the pending bit in an ACK frame (i.e., the SACKPEND strobe,
the PENDING_OR register bit, and the AUTOPEND feature). The pending bit is set if one or more of these
sources are set.
Transmission Timing
Acknowledgment frames can only be transmitted immediately after frame reception. The transmission
timing is controlled by the FSMCTRL.SLOTTED_ACK bit.
12 Symbol Periods = 192 ms
Unslotted ACK (0)
Preamble
RX Frame
SFD
Preamble
SFD
ACK Frame
Preamble
SFD
ACK Frame
n Backoff Periods = n ´ 320 ms
Slotted ACK (1)
Preamble
RX Frame
SFD
12–31 Symbol Periods
T0320-01
Figure 23-17. Acknowledgment Timing
The IEEE 802.15.4 requires unslotted mode in nonbeacon-enabled PANs, and slotted mode for beaconenabled PANs.
Manual Control
The SACK, SACKPEND, and SNACK command strobes can only be issued during frame reception. If the
strobes are issued at any other time, they have no effect but generating a STROBE_ERROR interrupt.
Preamble
RX Frame (Rejected or Accepted)
SFD
STROBE_ ERROR
Valid Strobe Interval
STROBE_ ERROR
T0321-01
Figure 23-18. Command Strobe Timing
The command strobes may be issued several times during reception; however, only the last strobe has an
effect:
• No strobe / SNACK / incorrect FCS: No acknowledgment transmission
• SACK: Acknowledgment transmission with the frame pending-bit cleared
• SACKPEND: Acknowledgment transmission with the frame pending-bit set
Automatic Control (AUTOACK)
When FRMFILT0.FRM_FILTER_EN and FRMCTRL0.AUTOACK are both enabled, the radio determines
automatically whether or not to transmit acknowledgment frames:
• The RX frame must be accepted by frame filtering (indicated by the RX_FRM_ACCEPTED exception).
• The acknowledgment request bit must be set in the RX frame.
• The RX frame must not be a beacon or an acknowledgment frame.
• The FCS of the RX frame must be correct.
Automatic acknowledgments can be overridden by the SACK, SACKPEND, and SNACK command
strobes. For instance, if the microcontroller is low on memory resources and cannot store a received
frame, the SNACK strobe can be issued during reception and prevent acknowledging the discarded frame.
By default, the AUTOACK feature never sets the frame-pending bit in the acknowledgment frames. Apart
from manual override with command strobes, there are two options:
• Automatic control, using the AUTOPEND feature
• Manual control, using the FRMCTRL1.PENDING_OR bit
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
231
RXFIFO Access
www.ti.com
Automatic Setting of the Frame Pending Field (AUTOPEND)
When the SRCMATCH.AUTOPEND bit is set, the result from source address matching determines the value
of the frame-pending field. On reception of a frame, the frame-pending field in the (possibly) returned
acknowledgment is set, given that:
• FRMFILT0.FRAME_FILTER_EN is set.
• SRCMATCH.SRC_MATCH_EN is set.
• SRCMATCH.AUTOPEND is set.
• The received frame matches the current SRCMATCH.PEND_DATAREQ_ONLY setting.
•
The received source address matches at least one source-match table entry, which is enabled in both
SRCSHORTEN and SRCSHORTPENDEN, or SRCEXTEN and SRCEXTPENDEN.
If the source-matching table runs full, the FRMCTRL1.PENDING_OR bit may be used to override the
AUTOPEND feature and temporarily acknowledge all frames with the frame-pending field set.
23.10 RXFIFO Access
The RXFIFO can hold one or more received frames, provided that the total number of bytes is 128 or less.
There are two ways to determine the number of bytes in the RXFIFO:
• Reading the RXFIFOCNT register
• Using the FIFOP and FIFO signals in combination with the FIFOPCTRL.FIFOPTHR setting
The RXFIFO is accessed through the RFD register.
The data in the RXFIFO can also be accessed by accessing the radio RAM directly. The FIFO pointers
are readable in RXFIRST_PTR, RXLAST_PTR, and RXP1_PTR. This can be useful if one wants to access
quickly a certain byte within a frame without having to read out the entire frame first. Note that when using
this direct accessing, no FIFO pointers are updated.
The ISFLUSHRX command strobe resets the RXFIFO, resetting all FIFO pointers and clearing all
counters, status signals, and sticky error conditions. However, if the receiver is actively receiving a frame
when the FIFO is flushed, the RFERRF.ABO flag is asserted.
The SFLUSHRX command strobe resets the RXFIFO, removing all received frames and clearing all
counters, status signals, and sticky-error conditions.
23.10.1 Using the FIFO and FIFOP
The FIFO and FIFOP signals are useful when reading out received frames in small portions while the
frame is received:
• FSMSTAT1.FIFO goes high when one or more bytes are in the RXFIFO, but low when RX overflow
has occurred.
• The FSMSTAT1.FIFOP signal goes high when:
– The number of valid bytes in the RXFIFO exceeds the FIFOP threshold value programmed into
FIFOPCTRL. When frame filtering is enabled, the bytes in the frame header are not considered
valid until the frame has been accepted.
– The last byte of a new frame is received, even if the FIFOP threshold is not exceeded. If so, FIFOP
goes back to low at the next RXFIFO read access.
232
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RXFIFO Access
www.ti.com
Received Frame
Preamble
SFD
LEN
MPDU (LEN[6:0] Bytes)
Accepted Frame
FSMSTAT1:SFD
FSMSTAT1:FIFO
FSMSTAT1:FIFOP
(Low Threshold)
Rejected Frame
FSMSTAT1:FIFOP
(High Threshold)
FSMSTAT1:SFD
FSMSTAT1:FIFO
FSMSTAT1:FIFOP
First Byte
Received
Frame
Filtering
Complete
Last Byte
Received
T0322-01
Figure 23-19. Behavior of FIFO and FIFOP Signals
When using the FIFOP as an interrupt source for the microcontroller, the FIFOP threshold should be
adjusted by the interrupt service routine to prepare for the next interrupt. When preparing for the last
interrupt for a frame, the threshold should match the number of bytes remaining.
23.10.2 Error Conditions
There are two error conditions associated with the RXFIFO:
• Overflow, in which case the RXFIFO is full when another byte is received
• Underflow, in which case software attempts to read a byte from an empty RXFIFO
RX overflow is indicated by the RFERRF.RXOVERF flag being set and by the signal values
FSMSTAT1.FIFO = 0 and FSMSTAT1.FIFOP = 1. When the error occurs, frame reception is halted. The
frames currently stored in the RXFIFO may be read out before the condition is cleared with the
ISFLUSHRX strobe. Note that rejected frames can generate RX overflow if the condition occurs before the
frame is rejected.
RX underflow is indicated by the RFERRF.RXUNDERF flag being set. RX underflow is a serious error
condition that should not occur in error-free software, and the RXUNDERF event should only be used for
debugging or in a watchdog function. Note that the RXUNDERF error is not generated when the read
operation occurs simultaneously with the reception of a new byte.
23.10.3 RSSI
The radio has a built-in received signal-strength indication (RSSI), which calculates an 8-bit signed digital
value that can be read from a register or automatically appended to received frames. The RSSI value is
the result of averaging the received power over eight symbol periods (128 μs) as specified by IEEE
802.15.4 [1].
The RSSI value is a 2s-complement signed number on a logarithmic scale with 1-dB steps.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
233
Radio Control State Machine
www.ti.com
The status bit RSSI_VALID should be checked before reading the RSSI value register. RSSI_VALID
indicates that the RSSI value in the register is in fact valid, which means that the receiver has been
enabled for at least eight symbol periods.
To find the actual signal power P at the RF pins with reasonable accuracy, an offset must be added to the
RSSI value.
P = RSSI – OFFSET [dBm]
For example, with an offset of 73 dB, reading an RSSI value of –10 from the RSSI register means that the
RF input power is approximately –83 dBm. For the correct offset value to use, see the data sheet
(Appendix C).
There are two ways the radio can update the RSSI register after it has first become valid. If
FRMCTRL0.ENERGY_SCAN = 0 (default), the RSSI register contains the latest value available, but if this bit
is set to 1, a peak search is performed, and the RSSI register contains the largest value since the energy
scan was enabled.
23.10.4 Link Quality Indication
The link quality indication (LQI) is a measurement of the strength and/or quality of the received frame as
defined by the IEEE 802.15.4 standard [1]. The LQI value is required by the IEEE 802.15.4 standard [1] to
be limited to the range 0 through 255, with at least eight unique values. The radio does not provide an LQI
value directly, but reports several measurements that can be used by the microcontroller to calculate an
LQI value.
The RSSI value can be used by the MAC software to calculate the LQI value. This approach has the
disadvantage that, e.g., a narrowband interferer inside the channel bandwidth can increase the RSSI and
thus the LQI value, although the true link quality actually decreases. The radio therefore also provides an
average correlation value for each incoming frame, based on the first eight symbols following the SFD.
This unsigned 7-bit value can be looked on as a measurement of the chip error rate, although the radio
does not do chip decision.
As described in Section 23.9.7, the average correlation value for the first eight symbols is appended to
each received frame, together with the RSSI and CRC OK/not OK, when FRMCTRL0.AUTOCRC is set. A
correlation value of ~110 indicates a maximum-quality frame, whereas a value of ~50 is typically the
lowest-quality frame detectable by the radio.
Software must convert the correlation value to the range 0–255 as defined by [1], for instance by
calculating:
LQI = (CORR – a)b
limited to the range 0–255, where a and b are found empirically based on PER measurements as a
function of the correlation value.
A combination of RSSI and correlation values may also be used to generate the LQI value.
23.11 Radio Control State Machine
The FSM module is responsible for maintaining the TXFIFO and RXFIFO pointers, control of analog
dynamic signals such as power up/power down, control of the data flow within the RF Core, generation of
automatic acknowledgement frames, and control of all analog RF calibration.
234
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RX overflow
17
Overflow
SFD wait
3–6
ACK delay
55
Slotted ACK
RX
7–13
SFD detected
Frame completed and
no ack scheduled
RX/RX wait
14
Timeout 192 ms
or
rx2rx_time_off = 1
SFLUSHRX
Timeout
192 ms
RXFIFO
reset
16
ACK
calibration
48
Timeout 190 ms
SRXON or SFLUSHRX
rxenable ! = 0
Timeout x ms
(depending on length byte
of the received frame)
Unslotted ACK
Frame not for me
RX
calibration
2
rxenable = 0
Timeout
192 ms
TX/RX transit
40
any
RX state
idle
0
all states
ACK
49–54
rxenmask! = 0
rxenable! = 0
TX underflow
56
STXONCCA andcca = 1
SRFOFF and
tx_active = 0
r xenable = 0
Timeout 2 ms
Underflow
STXON
all TX
and
ACK
states
SRFOFF or
SRXON
TX shutdown
26, 57
TX final
39
Frame sent
TX
34–38
Timeout
192 ms
TX
calibration
32
F0036-01
www.ti.com
Radio Control State Machine
Figure 23-20. Main FSM
CC253x Radio
235
Random-Number Generation
www.ti.com
Table 23-3 shows the mapping from FSM state to the number which can be read from the FSMSTAT0
register. Note that although it is possible to read the state of the FSM, this information should not be used
to control the program flow in the application software. The states may change very quickly (every 32-MHz
clock cycle), and an 8-MHz SPI is not able to capture all the activities.
Table 23-3. FSM State Mapping
State Name
State Number, Decimal
Number, Hex
TX_ACTIV
E
RX_ACTIVE
Idle
0
0x00
0
0
RX calibration
SFD wait
2
0x02
0
1
3–6
0x03–0x06
0
1
RX
7–13
0x07–0x0D
0
1
RX/RX wait
14
0x0E
0
1
RXFIFO reset
16
0x10
0
1
RX overflow
17
0x11
0
0
TX calibration
32
0x20
1
0
TX
34–38
0x22–0x26
1
0
TX final
39
0x27
1
0
TX/RX transit
40
0x28
1
0
ACK calibration
48
0x30
1
0
ACK
49–54
0x31–0x36
1
0
ACK delay
55
0x37
1
0
TX underflow
56
0x38
1
0
TX shutdown
26, 57
0x1A, 0x39
1
0
23.12 Random-Number Generation
The RF Core can generate random bits. The chip should be in RX when generation of random bits is
required. One must also make sure that the chip has been in RX long enough for the transients to have
died out. A convenient way to do this is to wait for the RSSI-valid signal to go high.
Single random bits from either the I or Q channel can be read from the register RFRND.
Randomness tests show good results for this module. However, a slight dc component exists. In a simple
test where the RFRND.IRND register was read a number of times and the data grouped into bytes, about
20 million bytes were read. When interpreted as unsigned integers between 0 and 255, the mean value
was 127.6518, which indicates that there is a dc component.
The FFT of the first 214 bytes is shown in Figure 23-21. Note that the dc component is clearly visible. A
histogram (32 bins) of the 20 million values is shown in Figure 23-22.
236
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Packet Sniffing and Radio Test Output Signals
www.ti.com
650
645
–10
640
–20
635
–30
630
Count
PSD – Power Spectral Density – Power/Bin
0
–40
625
620
–50
615
–60
610
–70
–80
605
600
–3
–2
–1
0
1
2
0
3
50
100
150
200
250
Bin Number
f – Frequency – rad
G002
G001
Figure 23-21. FFT of the Random Bytes
Figure 23-22. Histogram of 20 Million Bytes Generated
With the RANDOM Instruction
For the first 20 million individual bits, the probability of a one is P(1) = 0.500602 and P(0) = 1 – P(1) =
0.499398.
Note that to fully qualify the random generator as true random, much more elaborate tests are required.
There are software packages available on the internet that may be useful in this respect [8], [9].
23.13 Packet Sniffing and Radio Test Output Signals
Packet sniffing is a nonintrusive way of observing data that is either being transmitted or received. The
packet sniffer outputs a clock and a data signal, which should be sampled on the rising edges of the clock.
The two packet sniffer signals are observable as GPIO outputs. For accurate time stamping, the SFD
signal should also be output.
Because the radio has a data rate of 250 kbps, the packet sniffer clock frequency is 250 kHz. The data is
output serially, with the MSB of each byte first, which is opposite of the actual RF transmission, but more
convenient when processing the data. It is possible to use a SPI slave to receive the data stream.
When sniffing frames in TX mode, the data that is read from the TXFIFO by the modulator is the same
data that is output by the packet sniffer. However, if automatic CRC generation is enabled, the packet
sniffer does NOT output these 2 bytes. Instead, it replaces the CRC bytes with 0x8080. This value can
never occur as the last two bytes of a received frame (when automatic CRC checking is enabled), and
thus it provides a way for the receiver of the sniffed data to separate frames that were transmitted and
frames that were received.
When sniffing frames in RX mode, the data that is written to the RXFIFO by the demodulator is the same
data that is output by the packet sniffer. In other words, the last two bytes are either the received CRC
value or the CRC OK/RSSI/correlation/SRCRESINDEX value that may automatically replace the CRC
value, depending on configuration settings.
To set up the packet sniffer signals or some of the other RF Core observation outputs (in total maximum
3; rfc_obs_sig0, rfc_obs_sig1, and rfc_obs_sig2), the user must perform the following steps:
Step1: Determine which signal (rfc_obs_sig) to output on which GPIO pin (P1[0:5]). This is done using
the OBSSELx control registers (OBSSEL0–OBSSEL5) that control the observation output to pins P1[0:5]
(overriding the standard GPIO behavior for those pins).
Step2: Set the RFC_OBS_CTRL control registers (RFC_OBS_CTRL0–RFC_OBS_CTRL2) to select the
correct signals (rfc_obs_sig); e.g., for packet sniffing one needs the rfc_sniff_data for the packet sniffer
data signal and rfc_sniff_clk for the corresponding clock signal.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
237
Command Strobe/CSMA-CA Processor
www.ti.com
Step3: For packet sniffing, the packet sniffer module must be enabled in the MDMTEST1 register.
23.14 Command Strobe/CSMA-CA Processor
The command strobe/CSMA-CA processor (CSP) provides the control interface between the CPU and the
radio.
The CSP interfaces with the CPU through the SFR register RFST and the XREG registers CSPX, CSPY,
CSPZ, CSPT, CSPSTAT, CSPCTRL, and CSPPROG<n> (where n is in the range 0 to 23). The CSP produces
interrupt requests to the CPU. In addition, the CSP interfaces with the MAC Timer by observing MAC
Timer events.
The CSP allows the CPU to issue command strobes to the radio, thus controlling the operation of the
radio.
The CSP has two modes of operation, which are described as follows.
• Immediate command strobe execution
• Program execution
Immediate command strobes are written as Immediate Command Strobe instructions to the CSP, which
are issued instantly to the radio module. The Immediate Command Strobe instructions are also used to
control the CSP. The Immediate Command Strobe instructions are described in Section 23.14.8.
Program execution mode means that the CSP executes a sequence of instructions, comprising a short
user-defined program, from a program memory or instruction memory. The available instructions are from
a set of 20 instructions. The instruction set is defined in Section 23.14.8. The required program is first
loaded into the CSP by the CPU, and then the CPU instructs the CSP to start executing the program.
The program execution mode, together with the MAC Timer, allows the CSP to automate CSMA-CA
algorithms and thus act as a coprocessor for the CPU.
The operation of the CSP is described in detail in the following sections. The command strobes and other
instructions supported by the CSP are given in Section 23.14.9.
RFST (0xE1) – RF CSMA-CA/Strobe Processor
Bit
7:0
Name
INSTR[7:0]
Reset
0xD0
R/W
R/W
Description
Data written to this register is written to the CSP instruction memory. Reading this register
returns the CSP instruction currently being executed.
23.14.1 Instruction Memory
The CSP executes single-byte program instructions which are read from a 24-byte instruction memory.
Writes to the instruction memory are sequential, written through SFR register RFST. An instruction write
pointer is maintained within the CSP to hold the location within the instruction memory where the next
instruction written to RFST is to be stored. For debugging purposes, the program currently loaded into the
CSP can be read from the XREG registers CSPPROG<n>. Following a reset, the write pointer is reset to
location 0. During each RFST register write, the write pointer is incremented by 1 until the end of memory
is reached, at which time the write pointer stops incrementing. The first instruction written to RFST is
stored in location 0, the location where program execution starts. Thus, a complete 24-instruction program
is written to the instruction memory by writing each instruction in the desired order to the RFST register.
The write pointer can be reset to 0 by writing the immediate command strobe instruction ISSTOP. In
addition, the write pointer is reset to 0 when the command strobe SSTOP is executed in a program.
Following a reset, the instruction memory is filled with SNOP (No Operation) instructions (opcode value
0xC0). The immediate strobe ISCLEAR clears the instruction memory, filling it with SNOP instructions.
While the CSP is executing a program, there must be no attempts to write instructions to the instruction
memory by writing to RFST. Failure to observe this rule can lead to incorrect program execution and
corrupt instruction memory contents. However, Immediate Command Strobe instructions may be written to
RFST (see Section 23.14.3).
238
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Command Strobe/CSMA-CA Processor
www.ti.com
23.14.2 Data Registers
The CSP has three data registers, CSPT, CSPX, CSPY, and CSPZ, which are read/write accessible for the
CPU as XREG registers. These registers are read or modified by some instructions, thus allowing the
CPU to set parameters to be used by a CSP program, or allowing the CPU to read CSP program status.
The CSPT data register is not modified by any instruction. The CSPT data register is used to set a MAC
Timer overflow-compare value. Once program execution has started on the CSP, the content of this
register is decremented by 1 each time the MAC Timer overflows. When CSPT reaches zero, program
execution is halted and the interrupt IRQ_CSP_STOP is asserted. The CSPT register is not decremented if
the CPU writes 0xFF to this register.
NOTE: If the CSPT register compare function is not used, this register must be set to 0xFF before
the program execution is started.
23.14.3 Program Execution
After the instruction memory has been filled, program execution is started by writing the immediate
command strobe instruction ISSTART to the RFST register. Program execution continues until the
instruction at the last location has been executed, the CSPT data register content is zero, an SSTOP
instruction has been executed, an immediate ISSTOP instruction is written to RFST, or a SKIP instruction
returns a location beyond the last location in the instruction memory. The CSP runs at the set system
clock frequency, which must be set to 32 MHz for correct radio operation.
Immediate command strobe instructions may be written to RFST while a program is being executed. In this
case, the immediate instruction is executed before the instruction in the instruction memory, which is
executed once the immediate instruction has been completed.
During program execution, reading RFST returns the current instruction being executed. An exception to
this is the execution of immediate command strobes, during which RFST returns 0xD0.
23.14.4 Interrupt Requests
The CSP has three interrupt flags which can produce the RF interrupt vector. These are the following:
• IRQ_CSP_STOP: asserted when the processor has executed the last instruction in memory or when
the processor stops due to an SSTOP or ISSTOP instruction or the CSPT register being equal to zero
•
•
IRQ_CSP_WT: asserted when the processor continues executing the next instruction after a WAIT W
or WAITX instruction
IRQ_CSP_INT: asserted when the processor executes an INT instruction
23.14.5 Random Number Instruction
There is a delay in the update of the random number used by the RANDXY instruction. Therefore, if the
instruction RANDXY, which uses this value, is issued immediately after a previous RANDXY instruction,
the random value read may be the same in both cases.
23.14.6 Running CSP Programs
The basic flow for loading and running a program on the CSP is shown in Figure 23-23. When program
execution stops at the end of the program, the current program remains in program memory so that the
same program can be run again by starting execution once again with the ISSTART command. To clear
the program contents, use the ISCLEAR instruction.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
239
Command Strobe/CSMA-CA Processor
www.ti.com
Write instruction to
RFST
No
All instructions
written?
Yes
Set up CSPT, CSPX,
CSPY, CSPZ, and
CSPCTRL registers
Start execution by
writing ISSTART to
RFST
Clear program by
writing ISCLEAR
to RFST
SSTOP instruction,
end of program, or
writing ISTOP to
RFST stops program
No
Rerun last
program?
Yes
F0037-01
Figure 23-23. Running a CSP Program
23.14.7 Registers
CSPROG<N> (N Ranging From 0 to 23) (0x61C0 + N) – CSP Program
Bit
7:0
Name
CSP_INSTR
Reset
0xD0
R/W
R
Description
Byte N of the CSP program memory
CSPCTRL (0x61E0) – CSP Control Bit
Bit
7:1
0
240
Name
—
MCU_CTRL
CC253x Radio
Reset
0000 000
0
R/W
R0
R/W
Description
Reserved. Read as 0
CSP MCU control input
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Command Strobe/CSMA-CA Processor
www.ti.com
CSPSTAT (0x61E1) – CSP Status Register
Bit
7:6
5
Name
—
CSP_RUNNING
Reset
00
0
R/W
R0
R
4:0
CSP_PC
0 0000
R
Description
Reserved. Read as 0
1: CSP is running.
0: CSP is idle.
CSP program counter
CSPX (0x61E2) – CSP X Register
Bit
7:0
Name
CSPX
Reset
0x00
R/W
R/W
Description
CSP X data register. Used by CSP instructions WAITX, RANDXY, INCX, DECX,
and conditional instructions
R/W
R/W
Description
CSP Y data register. Used by CSP instructions RANDXY, INCY, DECY, and
conditional instructions
R/W
R/W
Description
CSP Z data register. Used by CSP instructions INCZ, DECZ, and conditional
instructions
R/W
R/W
Description
CSP T data register. Content is decremented each time the MAC Timer
overflows while the CSP program is running. The SCP program stops when
decremented to 0. Setting CSPT = 0xFF prevents the register from being
decremented.
CSPY (0x61E3) – CSP Y Register
Bit
7:0
Name
CSPY
Reset
0x00
CSPZ (0x61E4) – CSP Z Register
Bit
7:0
Name
CSPZ
Reset
0x00
CSPT (0x61E5) – CSP T Register
Bit
7:0
Name
CSPT
Reset
0xFF
23.14.8 Instruction Set Summary
This section gives an overview of the instruction set. This is intended as a summary and definition of
instruction opcodes. See Section 23.14.9 for a description of each instruction. Each instruction consists of
one byte, which is written to the RFST register to be stored in the instruction memory.
The Immediate Strobe instructions (ISxxx) are not used in a program. When these instructions are written
to the RFST register, they are executed immediately. If the CSP is already executing a program, the
current instruction is delayed until the immediate strobe instruction has completed.
For undefined opcodes, the behavior of the CSP is defined as a no-operation strobe command (SNOP).
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
241
Command Strobe/CSMA-CA Processor
www.ti.com
Table 23-4. Instruction Set Summary
Mnemonic
7
SKIP <C>, <S>
0
6
5
S2 S1
4
3
2
1
S0
N
C2
C1
0
Description
C0 Skip S instructions on condition C. When condition (C XOR N) is true,
skip the next S instructions, else execute the next instruction. If S = 0,
re-execute the conditional jump (i.e., busy loop until condition is
false). Skipping past the last instruction in the command buffer results
in an implicit STOP command. The conditions are:
C = 0 CCA true
C = 1 Synchronization word received and still receiving packet or
synchronization word transmitted and still transmitting packet
(SFD found, not yet frame end)
C = 2 MCU control bit is 1.
C = 3 Reserved
C = 4 Register X = 0
C = 5 Register Y = 0
C = 6 Register Z = 0
C = 7 RSSI_VALID = 1
WAIT <W>
1
0
0
RPT <C>
1
0
1
W4 W3 W2 W1 W0 Wait for MAC Timer to overflow W times. Waits until the MAC Timer
has overflowed W times (W = 0 waits 32 times), then continues
execution. Generates an IRQ_CSP_WAIT interrupt request when
execution continues.
0
N
C2
C1
C0 Repeat loop while condition C. If condition C is true, go to the
instruction following the last LABEL instruction (address in loop-start
register); if the condition is false or no LABEL instruction has been
executed, go to the next instruction.
Note condition C is as defined for SKIP, defined previously in this
table. It is not possible to have a RPT instruction placed at index 23 of
the command buffer.
WEVENT1
1
0
1
1
1
0
0
0
Wait for mact_event1 to go high, and then continue execution.
WEVENT2
1
0
1
1
1
0
0
1
Wait for mact_event2 to go high, and then continue execution.
INT
1
0
1
1
1
0
1
0
Generate an IRQ_CSP_MANINT. Issues an IRQ_CSP_MANINT
interrupt request.
LABEL
1
0
1
1
1
0
1
1
Set the next instruction as the start of a repeat loop. Enters the
address of the next instruction into the loop-start register.
WAITX
1
0
1
1
1
1
0
0
Wait for MAC Timer to overflow [X] times, where [X] is the value of
register X. Each time a MAC Timer overflow is detected, X is
decremented. Execution continues as soon as X = 0. (If X = 0 when
instruction is run, no wait is performed and execution continues
directly). An IRQ_CSP_WAIT interrupt request is generated when
execution continues.
RANDXY
1
0
1
1
1
1
0
1
Load the [Y] LSBs of register X with random value.
SETCMP1
1
0
1
1
1
1
1
0
Set the output csp_mact_setcmp1 high. This sets the compare value
of the MAC Timer to the current timer value.
INCX
1
1
0
0
0
0
0
0
Increment register X.
INCY
1
1
0
0
0
0
0
1
Increment register Y.
INCZ
1
1
0
0
0
0
1
0
Increment register Z.
DECX
1
1
0
0
0
0
1
1
Decrement register X.
DECY
1
1
0
0
0
1
0
0
Decrement register Y.
DECZ
1
1
0
0
0
1
0
1
Decrement register Z.
INCMAXY <M>
1
1
0
0
1
Sxxx
1
1
0
1
S3
M2 M1 M0 Register Y ≤ min(Y + 1, M). Increment Y, but not beyond M.
S2
S1
S0 Execute command strobe S. Send command strobe S to FFCTRL. Up
to 32 command strobes are supported. In addition to the regular
command strobes, two additional command strobes that only apply to
the command strobe processor are supported:
SNOP: Do nothing.
SSTOP: Stops the command strobe processor execution and
invalidates any set label. An IRQ_CSP_STOP interrupt request is
issued.
242 CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Command Strobe/CSMA-CA Processor
www.ti.com
Table 23-4. Instruction Set Summary (continued)
Mnemonic
ISxxx
7
6
5
4
3
2
1
1
1
1
0
S3
S2
S1
0
Description
S0 Execute command strobe S immediately. Send command strobe S to
FFCTRL immediately, bypassing the instructions in the command
buffer. If the current buffer instruction is a strobe, it is delayed. In
addition to the regular command strobes, two additional command
strobes that only apply to the command strobe processor are
supported:
ISSTART: The command strobe processor starts execution at the first
instruction in the command buffer. Do not issue an ISSTART
instruction if the CSP is already running.
ISSTOP: Stops the command strobe processor execution and
invalidates any set label. An IRQ_CSP_STOP interrupt request is
issued.
ISCLEAR
1
1
1
1
1
1
1
1
Clear the CSP program. Reset PC.
23.14.9 Instruction Set Definition
There are 20 basic instruction types. Furthermore, the command-strobe and immediate-strobe instructions
can each be divided into 16 subinstructions, giving an effective number of 42 different instructions. The
following subsections describe each instruction in detail.
Note: the following definitions are used in this section
PC = CSP program counter
X = RF register CSPX
Y = RF register CSPY
Z = RF register CSPZ
T = RF register CSPT
23.14.9.1 DECZ
Function:
Description:
Operation:
Decrement Z
The Z register is decremented by 1. An original value of 0x00 underflows to 0xFF.
Z=Z–1
Opcode: 0xC5
7
1
6
1
5
0
4
0
3
0
2
1
1
0
0
1
23.14.9.2 DECY
Function:
Description:
Operation:
Decrement Y
The Y register is decremented by 1. An original value of 0x00 underflows to 0xFF.
Y=Y–1
Opcode: 0xC4
7
1
6
1
5
0
4
0
3
0
2
1
1
0
0
0
23.14.9.3 DECX
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
243
Command Strobe/CSMA-CA Processor
Function:
Description:
Operation:
www.ti.com
Decrement X
The X register is decremented by 1. An original value of 0x00 underflows to 0xFF.
X=X–1
Opcode: 0xC3
7
1
6
1
5
0
4
0
3
0
2
0
1
1
0
1
23.14.9.4 INCZ
Function:
Description:
Operation:
Increment Z
The X register is incremented by 1. An original value of 0xFF overflows to 0x00.
Z=Z+1
Opcode: 0xC2
7
1
6
1
5
0
4
0
3
0
2
0
1
1
0
0
23.14.9.5 INCY
Function:
Description:
Operation:
Increment Y
The Y register is incremented by 1. An original value of 0xFF overflows to 0x00.
Y=Y+1
Opcode: 0xC1
7
1
6
1
5
0
4
0
3
0
2
0
1
0
0
1
23.14.9.6 INCX
Function:
Description:
Operation:
Increment X
The X register is incremented by 1. An original value of 0xFF overflows to 0x00.
X=X+1
Opcode: 0xC0
7
1
6
1
5
0
4
0
3
0
2
0
1
0
0
0
23.14.9.7 INCMAXY
Function:
Description:
Operation:
244
CC253x Radio
Increment Y not greater than M.
The Y register is incremented by 1 if the result is less than M; otherwise, Y register is
loaded with value M.
Y = min(Y + 1, M)
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Command Strobe/CSMA-CA Processor
www.ti.com
Opcode: 0xC8 | M (M = 0–7)
7
1
6
1
5
0
4
0
3
1
2
1
M
0
23.14.9.8 RANDXY
Function:
Description:
Operation:
Load random value into X.
The [Y] LSBs of the X register are loaded with a random value. Note that if a second
RANDXY instruction is issued immediately (within 13 clock cycles) after the first, the
same random value is used in both cases. If Y equals zero or is greater than 7, then an
8-bit random value is loaded into X.
X[(Y – 1):0]: = RNG_DOUT[(Y – 1):0], X[7:Y]: = 0
Opcode: 0xBD
7
1
6
0
5
1
4
1
3
1
2
1
1
0
0
1
23.14.9.9 INT
Function:
Description:
Operation:
Interrupt
The interrupt IRQ_CSP_INT is asserted when this instruction is executed.
IRQ_CSP_INT = 1
Opcode: 0xBA
7
1
6
0
5
1
4
1
3
1
2
0
1
1
0
0
23.14.9.10 WAITX
Function:
Description:
Operation:
Wait for X MAC Timer overflows
Wait for MAC Timer to overflow [X] times, where [X] is the value of register X. Each time
a MAC Timer overflow is detected, the value in register X is decremented. Program
execution continues as soon as X = 0. (If X = 0 when instruction is run, no wait is
performed and execution continues directly.) An IRQ_CSP_WAIT interrupt request is
generated when execution continues. Note: The difference compared to WAIT W is that
W is a fixed value, whereas X is a register value (which could potentially be changed,
such that the number of overflows actually does not correspond to the value of X at the
time WAITX instruction is run).
X = X – 1 when MAC Timer overflow = true
PC = PC while X > 0
PC = PC + 1 when X = 0
Opcode: 0xBC
7
1
6
0
5
1
4
1
3
1
2
1
1
0
0
0
23.14.9.11 SETCMP1
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
245
Command Strobe/CSMA-CA Processor
Function:
Description:
Operation:
www.ti.com
Set the compare value of the MAC Timer to the current timer value.
Set the compare value of the MAC Timer to the current timer value.
Csp_mact_setcmp1 = 1
Opcode: 0xBE
7
1
6
0
5
1
4
1
3
1
2
1
1
1
0
0
23.14.9.12 WAIT W
Function:
Description:
Operation:
Wait for W MAC Timer overflows
Wait until MAC Timer overflows a number of times equal to the value of W. If W = 0, the
instruction waits for 32 overflows. Program execution continues with the next instruction,
and the interrupt flag IRQ_CSP_WT is asserted when the wait condition is true.
PC = PC while number of MAC Timer overflows < W
PC = PC + 1 when number of MAC Timer overflows = W
Opcode: 0x80 | W (W = 0–31)
7
1
6
0
5
0
4
3
2
W
1
0
23.14.9.13 WEVENT1
Function:
Description:
Operation:
Wait until MAC Timer event 1
Wait until next MAC Timer event. Program execution continues with the next instruction
when the wait condition is true.
PC = PC while MAC Timer compare = false
PC = PC + 1 when MAC Timer compare = true
Opcode: 0xB8
7
1
6
0
5
1
4
1
3
1
2
0
1
0
0
0
23.14.9.14 WEVENT2
Function:
Description:
Operation:
Wait until MAC Timer event 2
Wait until next MAC Timer event. Program execution continues with the next instruction
when the wait condition is true.
PC = PC while MAC Timer compare = false
PC = PC + 1 when MAC Timer compare = true
Opcode: 0xB9
7
1
246
6
0
5
1
CC253x Radio
4
1
3
1
2
0
1
0
0
1
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Command Strobe/CSMA-CA Processor
www.ti.com
23.14.9.15 LABEL
Function:
Description:
Operation:
Set loop label
Sets next instruction as start of loop. If the current instruction is the last instruction in the
instruction memory, then the current PC is set as start of loop. If several label
instructions are executed, the last label executed is the active label. Earlier labels are
removed, which means that only one level of loops is supported.
LABEL: = PC + 1
Opcode: 0xBB
7
1
6
0
5
1
4
1
3
1
2
0
1
1
0
1
23.14.9.16 RPT C
Function:
Description:
Condition
Code C
000
001
010
011
100
101
110
111
Conditional repeat
If condition C is true, then jump to the instruction defined by the last LABEL instruction,
i.e., jump to start of loop. If the condition is false or if a LABEL instruction has not been
executed, then execution continues from next instruction. The condition C may be
negated by setting N = 1 and is described in the following table.
Description
Function
CCA is true
Synchronization word
received and still receiving
packet or synchronization
word transmitted and still
transmitting packet
CPU control true
Reserved
Register X = 0
Register Y = 0
Register Z = 0
RSSI is valid
CCA = 1
SFD = 1
CSPCTRL.CPU_CTRL = 1
X=0
Y=0
Z=0
RSSI_VALID = 1
Operation: PC = LABEL when (C XOR N) = true
PC = PC + 1 when (C XOR N) = false or LABEL = not set
Opcode: 0xA0 | N | C (N = 0, 8; C = 0–7)
7
1
6
0
5
1
4
0
3
N
2
1
C
0
23.14.9.17 SKIP S, C
Function:
Description:
Conditional skip instruction
Skip S instructions on condition C (where condition C may be negated; N = 1). When
condition (C xor N) is true, skip the next S instructions, else execute the next instruction.
If S = 0, re-execute the conditional jump (i.e., busy loop until condition is false). Skipping
past the last instruction in the command buffer results in an implicit STOP command.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
247
Command Strobe/CSMA-CA Processor
Condition
Code C
000
001
010
011
100
101
110
111
www.ti.com
Description
Function
CCA is true
Synchronization word
received and still receiving
packet or synchronization
word transmitted and still
transmitting packet
CPU control true
Reserved
Register X = 0
Register Y = 0
Register Z = 0
RSSI is valid
CCA = 1
SFD = 1
CSPCTRL.CPU_CTRL = 1
X=0
Y=0
Z=0
RSSI_VALID = 1
Operation: PC = PC + S + 1 when (C XOR N) = true
PC = PC + 1 when (C XOR N) = false
Opcode: 0x00 | S | N | C
7
0
6
5
S
4
3
N
2
1
C
0
23.14.9.18 STOP
Function:
Description:
Operation:
Stop program execution
The SSTOP instruction stops the CSP program execution.
Stop execution
Opcode: 0xD2
7
1
6
1
5
0
4
1
3
0
2
0
1
1
0
0
23.14.9.19 SNOP
Function:
Description:
Operation:
No operation
Operation continues at the next instruction.
PC = PC + 1
Opcode: 0xD0
7
1
6
1
5
0
4
1
3
0
2
0
1
0
0
0
23.14.9.20 SRXON
Function:
Description:
Operation:
248
CC253x Radio
Enable and calibrate frequency synthesizer for RX
The SRXON instruction asserts the output FFCTL_SRXON_STRB to enable and
calibrate the frequency synthesizer for RX. The instruction waits for the radio to
acknowledge the command before executing the next instruction.
SRXON
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Command Strobe/CSMA-CA Processor
www.ti.com
Opcode: 0xD3
7
1
6
1
5
0
4
1
3
0
2
0
1
1
0
1
23.14.9.21 STXON
Function:
Description:
Operation:
Enable TX after calibration
The STXON instruction enables TX after calibration. The instruction waits for the radio to
acknowledge the command before executing the next instruction. Sets a bit in RXENABLE if
SET_RXENMASK_ON_TX is set
STXON
Opcode: 0xD9
7
1
6
1
5
0
4
1
3
1
2
0
1
0
0
1
23.14.9.22 STXONCCA
Function:
Description:
Operation:
Enable calibration and TX if CCA indicates a clear channel
The STXONCCA instruction enables TX after calibration if CCA indicates a clear
channel.
STXONCCA
Opcode: 0xDA
7
1
6
1
5
0
4
1
3
1
2
0
1
1
0
0
23.14.9.23 SSAMPLECCA
Function:
Description:
Operation:
Sample the current CCA value to SAMPLED_CCA
The current CCA value is written to SAMPLED_CCA in XREG.
SSAMPLECCA
Opcode: 0xDB
7
1
6
1
5
0
4
1
3
1
2
0
1
1
0
1
23.14.9.24 SRFOFF
Function:
Description:
Operation:
Disable RX/TX and frequency synthesizer.
The SRFOFF instruction disables RX/TX and the frequency synthesizer.
SRFOFF
Opcode: 0xDF
7
1
6
1
5
0
4
1
3
1
2
1
1
1
0
1
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
249
Command Strobe/CSMA-CA Processor
www.ti.com
23.14.9.25 SFLUSHRX
Function:
Description:
Operation:
Flush RXFIFO buffer and reset demodulator
The SFLUSHRX instruction flushes the RXFIFO buffer and resets the demodulator. The
instruction waits for the radio to acknowledge the command before executing the next
instruction.
SFLUSHRX
Opcode: 0xDD
7
1
6
1
5
0
4
1
3
1
2
1
1
0
0
1
23.14.9.26 SFLUSHTX
Function:
Description:
Operation:
Flush TXFIFO buffer
The SFLUSHTX instruction flushes the TXFIFO buffer. The instruction waits for the radio
to acknowledge the command before executing the next instruction.
SFLUSHTX
Opcode: 0xDE
7
1
6
1
5
0
4
1
3
1
2
1
1
1
0
0
23.14.9.27 SACK
Function:
Description:
Operation:
Send acknowledge frame with pending field cleared
The SACK instruction sends an acknowledge frame. The instruction waits for the radio to
acknowledge the command before executing the next instruction.
SACK
Opcode: 0xD6
7
1
6
1
5
0
4
1
3
0
2
1
1
1
0
0
23.14.9.28 SACKPEND
Function:
Description:
Operation:
Send acknowledge frame with the pending field set
The SACKPEND instruction sends an acknowledge frame with the pending field set. The
instruction waits for the radio to acknowledge the command before executing the next
instruction.
SACKPEND
Opcode: 0xD7
7
1
250
6
1
5
0
CC253x Radio
4
1
3
0
2
1
1
1
0
1
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Command Strobe/CSMA-CA Processor
www.ti.com
23.14.9.29 SNACK
Function:
Description:
Operation:
Abort sending of acknowledge frame
The SACKPEND instruction aborts sending acknowldedge to the frame currently being
received.
SNACK
Opcode: 0xD8
7
1
6
1
5
0
4
1
3
1
2
0
1
0
0
0
23.14.9.30 SRXMASKBITSET
Function:
Description:
Operation:
Set bit in RXENABLE
The SRXMASKBITSET instruction sets bit 5 in the
SRXMASKBITSET
RXENABLE
register.
Opcode: 0xD4
7
1
6
1
5
0
4
1
3
0
2
1
1
0
0
0
23.14.9.31 SRXMASKBITCLR
Function:
Description:
Operation:
Clear bit in RXENABLE
The SRXMASKBITCLR instruction clears bit 5 in the
SRXMASKBITCLR
RXENABLE
register.
Opcode: 0xD5
7
1
6
1
5
0
4
1
3
0
2
1
1
0
0
1
23.14.9.32 ISSTOP
Function:
Description:
Operation:
Stop program execution
The ISSTOP instruction stops the CSP program execution and the IRQ_CSP_STOP
interrupt flag is asserted.
Stop execution
Opcode: 0xE2
7
1
6
1
5
1
4
0
3
0
2
0
1
1
0
0
23.14.9.33 ISSTART
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
251
Command Strobe/CSMA-CA Processor
Function:
Description:
Operation:
www.ti.com
Start program execution
The ISSTART instruction starts the CSP program execution from first instruction written
to instruction memory. Do not issue an ISSTART instruction if CSP is already running.
PC := 0, start execution
Opcode: 0xE1
7
1
6
1
5
1
4
0
3
0
2
0
1
0
0
1
23.14.9.34 ISRXON
Function:
Description:
Operation:
Enable and calibrate frequency synthesizer for RX
The ISRXON instruction immediately enables and calibrates the frequency synthesizer
for RX.
SRXON
Opcode: 0xE3
7
1
6
1
5
1
4
0
3
0
2
0
1
1
0
1
23.14.9.35 ISRXMASKBITSET
Function:
Description:
Operation:
Set bit in RXENABLE
The ISRXMASKBITSET instruction immediately sets bit 5 in the
SRXMASKBITSET
RXENABLE
register.
Opcode: 0xE4
7
1
6
1
5
1
4
0
3
0
2
1
1
0
0
0
23.14.9.36 ISRXMASKBITCLR
Function:
Description:
Operation:
Clear bit in RXENABLE
The ISRXMASKBITCLR instruction immediately clears bit 5 in the
SRXMASKBITCLR
RXENABLE
register.
Opcode: 0xE5
7
1
6
1
5
1
4
0
3
0
2
1
1
0
0
1
23.14.9.37 ISTXON
Function:
Description:
Operation:
252
CC253x Radio
Enable TX after calibration
The ISTXON instruction immediately enables TX after calibration. The instruction waits
for the radio to acknowledge the command before executing the next instruction.
STXON_STRB
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Command Strobe/CSMA-CA Processor
www.ti.com
Opcode: 0xE9
7
1
6
1
5
1
4
0
3
1
2
0
1
0
0
1
23.14.9.38 ISTXONCCA
Function:
Description:
Operation:
Enable calibration and TX if CCA indicates a clear channel
The ISTXONCCA instruction immediately enables TX after calibration if CCA indicates a
clear channel.
STXONCCA
Opcode: 0xEA
7
1
6
1
5
1
4
0
3
1
2
0
1
1
0
0
23.14.9.39 ISSAMPLECCA
Function:
Description:
Operation:
Sample the current CCA value to SAMPLED_CCA
The current CCA value is immediately written to
SSAMPLECCA
SAMPLED_CCA
in XREG.
Opcode: 0xEB
7
1
6
1
5
1
4
0
3
1
2
0
1
1
0
1
23.14.9.40 ISRFOFF
Function:
Description:
Operation:
Disable RX/TX and frequency synthesizer.
The ISRFOFF instruction immediately disables RX/TX and the frequency synthesizer.
FFCTL_SRFOFF_STRB = 1
Opcode: 0xEF
7
1
6
1
5
1
4
0
3
1
2
1
1
1
0
1
23.14.9.41 ISFLUSHRX
Function:
Description:
Operation:
Flush RXFIFO buffer and reset demodulator
The ISFLUSHRX instruction immediately flushes the RXFIFO buffer and resets the
demodulator.
SFLUSHRX
Opcode: 0xED
7
1
6
1
5
1
4
0
3
1
2
1
1
0
0
1
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
253
Command Strobe/CSMA-CA Processor
www.ti.com
23.14.9.42 ISFLUSHTX
Function:
Description:
Operation:
Flush TXFIFO buffer
The ISFLUSHTX instruction immediately flushes the TXFIFO buffer.
SFLUSHTX
Opcode: 0xEE
7
1
6
1
5
1
4
0
3
1
2
1
1
1
0
0
23.14.9.43 ISACK
Function:
Description:
Operation:
Send acknowledge frame with the pending field cleared
The ISACK instruction immediately sends an acknowledge frame.
SACK
Opcode: 0xE6
7
1
6
1
5
1
4
0
3
0
2
1
1
1
0
0
23.14.9.44 ISACKPEND
Function:
Description:
Operation:
Send acknowledge frame with the pending field set
The ISACKPEND instruction immediately sends an acknowledge frame with the pending
field set. The instruction waits for the radio to receive and interpret the command before
executing the next instruction.
SACKPEND
Opcode: 0xE7
7
1
6
1
5
1
4
0
3
0
2
1
1
1
0
1
23.14.9.45 ISNACK
Function:
Description:
Operation:
Abort sending of acknowledge frame
The ISNACK instruction immediately prevents sending of an acknowledge frame to the
currently received frame.
SNACK
Opcode: 0xE8
7
1
6
1
5
1
4
0
3
1
2
0
1
0
0
0
23.14.9.46 ISCLEAR
254
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
Function:
Description:
Operation:
Clear CSP program memory, reset program counter
The ISCLEAR clears the program memory, resets the program counter, and aborts any
running program. No stop interrupt is generated. The LABEL pointer is cleared. The
ISCLEAR instruction must be issued twice to reset the program counter.
PC := 0, clear program memory
Opcode: 0xFF
7
1
6
1
5
1
4
1
3
1
2
1
1
1
0
1
23.15 Registers
Table 23-5. Register Overview
Address (Hex)
+ 0x000
+ 0x001
+ 0x002
+ 0x003
0x6180
FRMFILT0
FRMFILT1
SRCMATCH
SRCSHORTEN0
0x6184
SRCSHORTEN1
SRCSHORTEN2
SRCEXTEN0
SRCEXTEN1
0x6188
SRCEXTEN2
FRMCTRL0
FRMCTRL1
RXENABLE
0x618C
RXMASKSET
RXMASKCLR
FREQTUNE
FREQCTRL
0x6190
TXPOWER
TXCTRL
FSMSTAT0
FSMSTAT1
0x6194
FIFOPCTRL
FSMCTRL
CCACTRL0
CCACTRL1
0x6198
RSSI
RSSISTAT
RXFIRST
RXFIFOCNT
0x619C
TXFIFOCNT
RXFIRST_PTR
RXLAST_PTR
RXP1_PTR
TXFIRST_PTR
TXLAST_PTR
RFIRQM0
0x61A0
0x61A4
RFIRQM1
RFERRM
MONMUX
RFRND
0x61A8
MDMCTRL0
MDMCTRL1
FREQEST
RXCTRL
0x61AC
FSCTRL
FSCAL1
FSCAL2
0x61B0
FSCAL3
AGCCTRL0
AGCCTRL1
AGCCTRL2
0x61B4
AGCCTRL3
ADCTEST0
ADCTEST1
ADCTEST2
0x61B8
MDMTEST0
MDMTEST1
DACTEST0
DACTEST1
0x61BC
DACTEST2
ATEST
PTEST0
PTEST1
0x61C0
CSPPROG0
CSPPROG1
CSPPROG2
CSPPROG3
0x61C4
CSPPROG4
CSPPROG5
CSPPROG6
CSPPROG7
0x61C8
CSPPROG8
CSPPROG9
CSPPROG10
CSPPROG11
0x61CC
CSPPROG12
CSPPROG13
CSPPROG14
CSPPROG15
0x61D0
CSPPROG16
CSPPROG17
CSPPROG18
CSPPROG19
0x61D4
CSPPROG20
CSPPROG21
CSPPROG22
CSPPROG23
0x61E0
CSPCTRL
CSPSTAT
CSPX
CSPY
0x61E4
CSPZ
CSPT
RFC_OBS_CTRL1
RFC_OBS_CTRL2
0x61D8
0x61DC
0x61E8
0x61EC
RFC_OBS_CTRL0
0x61F0
0x61F4
0x61F8
TXFILTCFG
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio 255
Registers
www.ti.com
23.15.1 Register Settings Update
This section contains a summary of the register settings that must be updated from their default value to
have optimal performance.
The following settings should be set for both RX and TX. Although not all settings are necessary for both
RX and TX, it is recommended for simplicity (allowing one set of settings to be written at the initialization
of the code).
Table 23-6. Registers That Require Update From Their Default Value
Register Name
New Value (Hex)
Description
AGCCTRL1
0x15
Adjusts AGC target value.
TXFILTCFG
0x09
Sets TX anti-aliasing filter to appropriate bandwidth.
FSCAL1
0x00
Recommended setting for lowest spurious emission
23.15.2 Register Access Modes
The Mode column in Table 23-7 shows what kind of accesses are allowed for each bit. The Description
column gives the meaning of the different alternatives.
Table 23-7. Register-Bit Access Modes
Mode
Description
R
Read
W
Write
R0
Read constant zero
R1
Read constant one
W1
Only possible to write one
W0
Only possible to write zero
R*
The value read is not the actual register value, but rather the value seen by the module. This is typically used where a
configuration value may be generated automatically (through calibration, dynamic control etc.) or manually overridden with
a register value. An example structure is shown for the AGCCTRL2 register in Figure 23-24.
read_data
write_ data
AGCCTRL2
Register
LNA_ CURRENT_OE
rf_ input
LNA
1
AGC
Module
0
B0308-01
Figure 23-24. Example Hardware Structure for the R* Register Access Mode
256
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
23.15.3 Register Descriptions
FRMFILT0 (0x6180) – Frame Filtering
Bit
No.
7
6:4
Name
Reset
R/W
Description
–
FCF_RESERVED_MASK[2:0]
0
000
R/W
R/W
3:2
MAX_FRAME_VERSION[1:0]
11
R/W
1
PAN_COORDINATOR
0
R/W
0
FRAME_FILTER_EN
1
R/W
Reserved. Always write 0
Used for filtering on the reserved part of the frame control field (FCF).
FCF_RESERVED_MASK[2:0] is ANDed with FCF[9:7]. If the result is
nonzero and frame filtering is enabled, the frame is rejected.
Used for filtering on the frame version field of the frame control field (FCF).
If FCF[13:12] (the frame version subfield) is higher than
MAX_FRAME_VERSION[1:0] and frame filtering is enabled, the frame is
rejected.
Should be set high when the device is a PAN coordinator, to accept
frames with no destination address (as specified in section 7.5.6.2 of IEEE
802.15.4(b))
0: Device is not PAN coordinator.
1: Device is PAN coordinator.
Enables frame filtering
When this bit is set, the radio performs frame filtering as specified in
section 7.5.6.2 of IEEE 802.15.4(b), third filtering level. FRMFILT0[6:1]
and FRMFILT1[7:1], together with the local address information, define
the behavior of the filtering algorithm.
0: Frame filtering off. (FRMFILT0[6:1], FRMFILT1[7:1] and
SRCMATCH[2:0] are don't care.)
1: Frame filtering on.
FRMFILT1 (0x6181) – Frame Filtering
Bit
No.
7
Name
Reset
R/W
Description
ACCEPT_FT_4TO7_RESERVED
0
R/W
6
ACCEPT_FT_3_MAC_CMD
1
R/W
5
ACCEPT_FT_2_ACK
1
R/W
4
ACCEPT_FT_1_DATA
1
R/W
3
ACCEPT_FT_0_BEACON
1
R/W
2:1
MODIFY_FT_FILTER[1:0]
00
R/W
0
–
0
R/W
Defines whether reserved frames are accepted or not. Reserved frames
have frame type = 100, 101, 110, or 111.
0: Reject
1: Accept
Defines whether MAC command frames are accepted or not. MAC
command frames have frame type = 011.
0: Reject
1: Accept
Defines whether acknowledgment frames are accepted or not.
Acknowledgement frames have frame type = 010.
0: Reject
1: Accept
Defines whether data frames are accepted or not. Data frames have
frame type = 001.
0: Reject
1: Accept
Defines whether beacon frames are accepted or not. Beacon frames
have frame type = 000.
0: Reject
1: Accept
These bits are used to modify the frame-type field of a received frame
before frame-type filtering is performed. The modification does not
influence the frame that is written to the RXFIFO.
00: Leave as it is
01: Invert MSB
10: Set MSB to 0
11: Set MSB to 1
Reserved. Always write 0
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
257
Registers
www.ti.com
SRCMATCH (0x6182) – Source Address Matching and Pending Bits
Bit
No.
7:3
2
Name
Reset
R/W
Description
–
PEND_DATAREQ_ONLY
0000 0
1
R/W
R/W
1
AUTOPEND
1
R/W
Reserved. Always write 0
When this bit is set, the AUTOPEND function also requires that the
received frame is a DATA REQUEST MAC command frame.
Automatic acknowledgment pending flag enable.
On reception of a frame, the pending bit in the (possibly) returned
acknowledgment is set automatically, given that:
– FRMFILT0.FRAME_FILTER_EN is set.
– SRCMATCH.SRC_MATCH_EN is set.
– SRCMATCH.AUTOPEND is set.
– The received frame matches the current
SRCMATCH.PEND_DATAREQ_ONLY setting.
0
SRC_MATCH_EN
1
R/W
– The received source address matches at least one source-match
table entry, which is enabled in both SHORT_ADDR_EN and
SHORT_PEND_EN or EXT_ADDR_EN and EXT_PEND_EN.
Note: Details for SHORT_PEND_EN and EXT_PEND_EN is found in the
memory map description (Section 23.4).
Source address matching enable (This bit is don't care if
FRMFILT0.FRAME_FILTER_EN = 0.)
SRCSHORTEN0 (0x6183) – Short Address Matching
Bit
No.
7:0
Name
Reset
R/W
Description
SHORT_ADDR_EN[7:0]
0x00
R/W
The 7:0 part of the 24-bit word SHORT_ADDR_EN that enables/disables
source address matching for each of the 24 short address table entries.
Optional safety feature: To ensure that an entry in the source-matching
table is not used while it is being updated, set the corresponding
SHORT_ADDR_EN bit to 0 while updating.
SRCSHORTEN1 (0x6184) – Short Address Matching
Bit
No.
7:0
Name
Reset
R/W
Description
SHORT_ADDR_EN[15:8]
0x00
R/W
The 15:8 part of the 24-bit word SHORT_ADDR_EN. See previoius
description of SRCSHORTEN0.
SRCSHORTEN2 (0x6185) – Short Address Matching
Bit
No.
7:0
Name
Reset
R/W
Description
SHORT_ADDR_EN[23:16]
0x00
R/W
The 23:16 part of the 24-bit word SHORT_ADDR_EN. See previous
description of SRCSHORTEN0.
SRCEXTEN0 (0x6186) – Extended Address Matching
Bit
No.
7:0
Name
Reset
R/W
Description
EXT_ADDR_EN[7:0]
0x00
R/W
The 7:0 part of the 24-bit word EXT_ADDR_EN that enables/disables
source address matching for each of the 12 extended address table
entries.
Write access: Extended address enable for table entry n (0 to 7) is
mapped to EXT_ADDR_EN[2n]. All EXT_ADDR_EN[2n + 1] bits are
read-only and don't care when written to.
Read access: Extended address enable for table entry n (0 to 7) is
mapped to both EXT_ADDR_EN[2n] and EXT_ADDR_EN[2n + 1].
To ensure that an entry in the source matching table is not used while it
is being updated, set the corresponding EXT_ADDR_EN bit to 0 while
updating.
258
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
SRCEXTEN1 (0x6187) – Extended Address Matching
Bit
No.
7:0
Name
Reset
R/W
Description
EXT_ADDR_EN[15:8]
0x00
R/W
The 15:8 part of the 24-bit word EXT_ADDR_EN. See previous description
of SRCEXTEN0.
SRCEXTEN2 (0x6188) – Extended Address Matching
Bit
No.
7:0
Name
Reset
R/W
Description
EXT_ADDR_EN[23:16]
0x00
R/W
The 23:16 part of the 24-bit word EXT_ADDR_EN. See previous
description of SRCEXTEN0.
Name
Reset
R/W
Description
APPEND_DATA_MODE
0
R/W
When AUTOCRC = 0: Don't care
When AUTOCRC = 1:
FRMCTRL0 (0x6189) – Frame Handling
Bit
No.
7
0:
1:
6
1
AUTOCRC
R/W
RSSI + The CRC_OK bit and the 7-bit correlation value are
appended at the end of each received frame
RSSI + The CRC_OK bit and the 7-bit SRCRESINDEX are
appended at the end of each received frame. See Table 23-1 for
details.
In TX
1:
A CRC-16 (ITU-T) is generated in hardware and appended to the
transmitted frame. There is no need to write the last 2 bytes to
TXBUF.
0:
No CRC-16 is appended to the frame. The last 2 bytes of the frame
must be generated manually and written to TXBUF (if not,
TX_UNDERFLOW occurs).
In RX
1:
The CRC-16 is checked in hardware, and replaced in the RXFIFO
by a 16-bit status word which contains a CRC OK bit. The status
word is controllable through APPEND_DATA_MODE.
0:
5
AUTOACK
0
R/W
4
ENERGY_SCAN
0
R/W
3:2
RX_MODE[1:0]
00
R/W
1:0
TX_MODE[1:0]
00
R/W
The last two bytes of the frame (CRC-16 field) are stored in the
RXFIFO. The CRC check (if any) must be done manually.
Note that this setting does not influence acknowledgment transmission
(including AUTOACK).
Defines whether the radio automatically transmits acknowledge frames or
not. When autoack is enabled, all frames that are accepted by address
filtering, have the acknowledge request flag set, and have a valid CRC
are automatically acknowledged 12 symbol periods after being received.
0:
Autoack disabled
1:
Autoack enabled
Defines whether the RSSI register contains the most-recent signal
strength or the peak signal strength since the energy scan was enabled.
0:
Most-recent signal strength
1:
Peak signal strength
Set RX modes
00: Normal operation, use RXFIFO.
01: Reserved
10: RXFIFO looping ignores overflow in RXFIFO; infinite reception.
11: Same as normal operation except that symbol search is disabled.
Can be used for RSSI or CCA measurements when it is not desired
to find symbol.
Set test modes for TX
00: Normal operation, transmit TXFIFO
01: Reserved. Should not be used
10: TXFIFO looping ignores underflow in TXFIFO and reads cyclically;
infinite transmission.
11: Send pseudorandom data from CRC, infinite transmission.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
259
Registers
www.ti.com
FRMCTRL1 (0x618A) – Frame Handling
Bit
No.
7:3
2
Name
Reset
R/W
Description
–
PENDING_OR
0000 0
0
R0
R/W
1
IGNORE_TX_UNDERF
0
R/W
0
SET_RXENMASK_ON_TX
1
R/W
Read as zero
Defines whether the pending data bit in outgoing acknowledgment frames
is always set to 1 or controlled by the main FSM and the address filtering.
0:
Pending data bit is controlled by main FSM and address filtering.
1:
Pending data bit is always 1.
Defines whether TX underflow should be ignored or not.
0:
Normal TX operation. TX underflow is detected and TX is aborted if
underflow occurs.
1:
Ignore TX underflow. Transmit the number of bytes given by the
frame-length field.
Defines whether STXON sets bit 6 in the RXENABLE register or leaves it
unchanged.
Does not affect RXENABLE.
0:
Sets bit 6 in RXENABLE. Used for backwards compatibility with the
1:
CC2420.
Name
Reset
R/W
Description
RXENMASK[7:0]
0x00
R
RXENABLE enables the receiver. A nonzero value in this register causes
the main FSM to enable the receiver when in idle, after transmission, and
after acknowledgement transmission.
The following strobes can modify RXENMASK:
SRXON: Sets bit 7 in RXENMASK
STXON: Sets bit 6 in RXENMASK if SET_RXENMASK_ON_TX = 1
SRFOFF: Clears all bits in RXENMASK
SRXMASKBITSET: Sets bit 5 in RXENMASK
SRXMASKBITCLR: Clears bit 5 in RXENMASK
RXENABLE can be modified directly by the CPU by accessing registers
RXMASKSET and RXMASKCLR.
RXENABLE (0x618B) – RX Enabling
Bit
No.
7:0
There might be conflicts between the CSP and CPU operations if both try
to modify RXENMASK simultaneously. To handle the case of simultaneous
access to RXENMASK, the following rules apply:
– If two sources are not in conflict (they modify different parts of the
register), both their requests to modify RXENMASK are processed.
– If both try to modify the mask simultaneously, bus-write operations to
RXMASKSET and RXMASKCLR have priority over the CSP. It is strongly
recommended to avoid this situation.
RXMASKSET (0x618C) – RX Enabling
Bit
No.
7:0
Name
Reset
R/W
Description
RXENMASKSET[7:0]
0x00
R0/W
When written, the written data is ORed with RXENMASK and stored in
RXENMASK.
RXMASKCLR (0x618D) – RX Disabling
Bit
No.
7:0
Name
Reset
R/W
Description
RXENMASKCLR[7:0]
0x00
R0/W
When written, the written data is inverted and ANDed with RXENMASK and
stored in RXENMASK.
For example, if a 1 is written to one or more bit positions in this register,
the corresponding bits are cleared in RXENMASK.
260
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
RFIRQM0 (0x61A3) – RF Interrupt Masks
Bit
No.
7:0
Name
Reset
R/W
Description
RFIRQM[7:0]
0x00
R/W
Bit mask masking out interrupt sources
Bit position
7: RXMASKZERO
6: RXPKTDONE
5: FRAME_ACCEPTED
4: SRC_MATCH_FOUND
3: SRC_MATCH_DONE
2: FIFOP
1: SFD
0: ACT_UNUSED
RFIRQM1 (0x61A4) – RF Interrupt Masks
Bit
No.
7:0
Name
Reset
R/W
Description
RFIRQM[14:8]
0x00
R/W
Bit mask masking out interrupt sources
Bit position
7: Reserved
6: Reserved
5: CSP_WAIT
4: CSP_STOP
3: CSP_MANINT
2: RF_IDLE
1: TXDONE
0: TXACKDONE
RFERRM (0x61A5) – RF Error Interrupt Mask
Bit
No.
7:0
Name
Reset
R/W
Description
RFERRM[7:0]
0x00
R/W
Bit mask masking out interrupt sources
Bit position
7: Reserved
6: STROBE_ERR
5: TXUNDERF
4: TXOVERF
3: RXUNDERF
2: RXOVERF
1: RXABO
0: NLOCK
FREQCTRL (0x618F) – Controls the RF Frequency
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
FREQ[6:0]
0
0x0B
(2405 MHz)
R0
R/W
Read as zero
Frequency control word
ƒRF = ƒLO = (2394 + FREQ[6:0]) MHz
The frequency word in FREQ[6:0] is an offset value from 2394. The device
supports the frequency range from 2394 MHz to 2507 MHz. The usable settings
for FREQ[6:0] are consequently 0 to 113. Settings outside this range (114–127)
give a frequency of 2507 MHz.
IEEE 802.15.4-2006 specifies a frequency range from 2405 MHz to 2480 MHz
with 16 channels 5 MHz apart. The channels are numbered 11 through 26. For an
IEEE 802.15.4-2006 compliant system, the only valid settings are thus
FREQ[6:0] = 11 + 5 (channel number – 11).
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
261
Registers
www.ti.com
FREQTUNE (0x618E) – Crystal Oscillator Frequency Tuning
Bit
No.
7:4
3:0
Reset
Name
–
XOSC32M_TUNE[3:0]
0x0
0xF
R/W
Description
R0
R/W
Read as zero
Tune crystal oscillator
The default setting, 1111, leaves the XOSC not tuned. Changing the
setting from default switches in extra capacitance to the oscillator,
effectively lowering the XOSC frequency. Hence, a higher setting gives
a higher frequency.
R/W
Description
R/W
PA power control.
NOTE: Before going to TX, this value should be updated. Please
consult the device's data sheet (Appendix C) for recommended values;
see also Section 23.8.13.
TXPOWER (0x6190) – Controls the Output Power
Bit
No.
7:0
Reset
Name
PA_POWER
0xF5
[7:0]
TXCTRL (0x6191) – Controls the TX Settings
Bit
No.
7
6:4
3:2
1:0
Name
Reset
R/W
Description
–
DAC_CURR[2:0]
DAC_DC[1:0]
TXMIX_CURRENT[1:0]
0
110
10
01
R0
R/W
R/W
R/W
Reserved
Change the current in the DAC.
Adjusts the dc level to the TX mixer.
Transmit mixers core current: current increases with increasing setting.
FSMSTAT0 (0x6192) – Radio Status Register
Bit
No.
7
6
5:0
262
Name
Reset
R/W
Description
Reserved
Frequency synthesizer calibration status
0:
Calibration done or not started
1:
Calibration in progress
Gives the current state of the FIFO and frame-control (FFCTRL) finite-state
machine.
–
CAL_RUNNING
0
0
R
R
FSM_FFCTRL_STATE[5:0]
–
R
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
FSMSTAT1 (0x6193) – Radio Status Register
Bit
No.
7
6
Name
Reset
R/W
Description
FIFO
FIFOP
0
0
R
R
FIFO is high whenever there is data in the RXFIFO. Low during RXFIFO overflow
FIFOP is set high when there are more than FIFOP_THR bytes of data in the
RXFIFO that have passed frame filtering.
FIFOP is set high when there is at least one complete frame in the RXFIFO.
FIFOP is set low again when a byte is read from the RXFIFO and this leaves
FIFOP_THR bytes in the FIFO.
FIFOP is high during RXFIFO overflow.
5
SFD
0
R
4
CCA
0
R
In TX
0:
When a complete frame with SFD has been sent or no SFD has been sent
1:
SFD has been sent.
In RX
0:
When a complete frame has been received or no SFD has been received
1:
SFD has been received.
Clear-channel assessment. Dependent on CCA_MODE settings. Seethe following
description of CCACTRL1.
3
SAMPLED_CCA
0
R
2
1
0
LOCK_STATUS
TX_ACTIVE
RX_ACTIVE
0
0
0
R
R
R
Contains a sampled value of the CCA. The value is updated whenever a
SSAMPLECCA or STXONCCA strobe is issued.
1 when PLL is in lock, otherwise 0.
Status signal, active when FFCTRL is in one of the transmit states
Status signal, active when FFCTRL is in one of the receive states
FIFOPCTRL (0x6194) – FIFOP Threshold
Bit
No.
7
6:0
Name
Reset
R/W
–
FIFOP_THR[6:0]
0
R0
100 0000 R/W
Description
Read as zero
Threshold used when generating FIFOP signal
FSMCTRL (0x6195) – FSM Options
Bit
No.
7:2
1
Name
Reset
R/W
Description
–
SLOTTED_ACK
0000 00
0
R0
R/W
0
RX2RX_TIME_OFF
1
R/W
Read as zero
Controls timing of transmission of acknowledge frames
0:
The acknowledge frame is sent 12 symbol periods after the end
of the received frame which requests the aknowledge.
1:
The acknowledge frame is sent at the first backoff-slot boundary
more than 12 symbol periods after the end of the received frame
which requests the aknowledge.
Defines whether or not a 12-symbol time-out should be used after frame
reception has ended.
0:
No time-out
1:
12-symbol-period time-out
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
263
Registers
www.ti.com
CCACTRL0 (0x6196) – CCA Threshold
Bit
No.
7:0
Name
Reset
R/W
Description
CCA_THR[7:0]
0xE0
R/W
Clear-channel-assessment threshold value, signed 2s-complement
number for comparison with the RSSI.
The unit is 1 dB, offset is about 76 dBm. The CCA signal goes high
when the received signal is below this value. The CCA signal is
available on the CCA pin and in the FSMSTAT1 register.
Note that the value should never be set lower than CCA_HYST – 128 in
order to avoid erroneous behavior of the CCA signal.
NOTE: The reset value translates to an input level of approximately
–32 – 76 = –108 dBm, which is well below the sensitivity limit. That
means the CCA signal never indicates a clear channel.
This register should be updated to 0xF8, which translates to an input
level of about –8 – 76 = –84 dBm.
CCACTRL1 (0x6197) – Other CCA Options
Bit
No.
7:5
4:3
Name
Reset
R/W
Description
–
CCA_MODE[1:0]
000
11
R0
R/W
Read as zero
00:
CCA always set to 1
CCA = 1 when RSSI < CCA_THR –
01:
0 when RSSI ≥ CCA_THR
CCA_HYST; CCA =
10:
11:
2:0
CCA_HYST[2:0]
010
CCA = 1 when not receiving a frame, else CCA = 0
CCA = 1 when RSSI < CCA_THR –
CCA_HYST and not
receiving a frame; CCA = 0 when RSSI ≥ CCA_THR or when
receiving a frame
Sets the level of CCA hysteresis. Unsigned values given in dB
R/W
RSSI (0x6198) – RSSI Status Register
Bit
No.
7:0
Name
Reset
R/W
Description
RSSI_VAL[7:0]
0x80
R
RSSI estimate on a logarithmic scale, signed number in 2s complement
Unit is 1 dB. The offset to use in order to convert the register value into
the real RSSI value can be found in the device's data sheet
(Appendix C). The RSSI value is averaged over 8 symbol periods. The
RSSI_VALID status bit should be checked before reading RSSI_VAL
the first time.
The reset value of –128 also indicates that the RSSI value is invalid.
RSSISTAT (0x6199) – RSSI Valid Status Register
Bit
No.
7:1
0
Name
Reset
R/W
Description
–
RSSI_VALID
0000 000
0
R0
R
Read as zero
RSSI value is valid. Occurs eight symbol periods after entering RX
RXFIRST (0x619A) – First Byte in RXFIFO
Bit
No.
7:0
Name
Reset
R/W
Description
DATA[7:0]
0x00
R
First byte of the RXFIFO.
Note: Reading this register does not modify the contents of the FIFO.
RXFIFOCNT (0x619B) – Number of Bytes in RXFIFO
Bit
No.
7:0
264
Name
Reset
R/W
Description
RXFIFOCNT[7:0]
0x00
R
Number of bytes in the RXFIFO. Unsigned integer.
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
TXFIFOCNT (0x619C) – Number of Bytes in TXFIFO
Bit
No.
7:0
Name
Reset
R/W
Description
TXFIFOCNT[7:0]
0x00
R
Number of bytes in the TXFIFO. Unsigned integer
R/W
Description
RXFIRST_PTR (0x619D) – RXFIFO Pointer
Bit
No.
7
6:0
Name
Reset
–
RXFIRST_PTR[6:0]
0
R
000 0000 R
Reserved
RAM address offset of the first byte in the RXFIFO
RXLAST_PTR (0x619E) – RXFIFO Pointer
Bit
No.
7
6:0
Name
Reset
R/W
–
RXLAST_PTR[6:0]
0
R
000 0000 R
Description
Reserved
RAM address offset of the last byte +1 byte in the RXFIFO
RXP1_PTR (0x619F) – RXFIFO Pointer
Bit
No.
7:0
Name
Reset
R/W
Description
RXP1_PTR[7:0]
0x00
R
RAM address offset of the first byte of the first frame in the RXFIFO
TXFIRST_PTR (0x61A1) – TXFIFO Pointer
Bit
No.
7:0
Name
Reset
R/W
Description
TXFIRST_PTR[7:0]
0x00
R
RAM address offset of the next byte to be transmitted from the TXFIFO
TXLAST_PTR (0x61A2) – TXFIFO Pointer
Bit
No.
7:0
Name
Reset
R/W
Description
TXLAST_PTR[7:0]
0x00
R
RAM address offset of the last byte +1 byte of the TXFIFO
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
265
Registers
www.ti.com
MDMCTRL0 (0x61A8) – Controls Modem
Bit
No.
7:6
Name
Reset
R/W
Description
DEM_NUM_ZEROS[1:0]
10
R/W
Sets how many zero symbols must be detected before the sync word when
searching for sync. Note that only one is required to have a correlation value
above the correlation threshold set in the MDMCTRL1 register.
5
DEMOD_AVG_MODE
0
R/W
4:1
PREAMBLE_LENGTH [3:0]
0010
R/W
0
TX_FILTER
1
R/W
00:
Reserved
01:
1 zero symbol
10:
2 zero symbols
11:
3 zero symbols
Defines the behavior or the frequency offset averaging filter.
0:
Lock average level after preamble match. Restart frequency offset
calibration when searching for the next frame.
1:
Continuously update average level.
The number of preamble bytes (two zero-symbols) to be sent in TX mode prior
to the SFD, encoded in steps of 2 symbols (1 byte). The reset value of 2 is
compliant with IEEE 802.15.4.
0000:
2 leading-zero bytes
0001:
3 leading-zero bytes
0010:
4 leading-zero bytes
…
…
1111:
17 leading-zero bytes
Defines the kind of TX filter that is used. The normal TX filter is as defined by
the IEEE802.15.4 standard. Extra filtering may be applied in order to lower the
out-of-band emissions.
0:
Normal TX filtering
1:
Enable extra filtering
MDMCTRL1 (0x61A9) – Controls Modem
Bit
No.
7:6
5
Name
Reset
R/W
Description
–
CORR_THR_SFD
00
0
R0
R/W
4:0
CORR_THR[4:0]
0x14
R/W
Read as zero
Defines requirements for SFD detection:
0:
The correlation value of one of the zero symbols of the preamble must
be above the correlation threshold.
1:
The correlation value of one zero symbol of the preamble and both
symbols in the SFD must be above the correlation threshold.
Demodulator correlator threshold value, required before SFD search.
Threshold value adjusts how the receiver synchronizes to data from the radio. If
the threshold is set too low, sync can more easily be found on noise. If set too
high, the sensitivity is reduced, but sync is not likely to be found on noise.
In combination with DEM_NUM_ZEROS, the system can be tuned so sensitivity
is high with less sync found on noise.
FREQEST (0x61AA) – Estimated RF Frequency Offset
Bit
No.
7:0
266
Name
Reset
R/W
Description
FREQEST[7:0]
0x00
R
Signed 2s-complement value. Contains an estimate of the frequency offset
between carrier and the receiver LO. The offset frequency is
FREQEST × 7800 Hz. DEM_AVG_MODE controls when this estimate is updated.
If DEM_AVG_MODE = 0, it is updated until sync is found. Then the frequency
offset estimate is frozen until the end of the received frame. If DEM_AVG_MODE
= 1, it is updated as long as the demodulator is enabled. To calculate the
correct value, one must use an offset (FREQEST_offset), which can be found
in the device's data sheet (Appendix C). Real FREQEST value = FREQEST –
FREQEST_offset.
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
RXCTRL (0x61AB) – Tune Receive Section
Bit
No.
7:6
5:4
Name
Reset
R/W
Description
–
GBIAS_LNA2_REF[1:0]
00
11
R0
R/W
3:2
GBIAS_LNA_REF[1:0]
11
R/W
1:0
MIX_CURRENT[1:0]
11
R/W
Reserved
Adjusts front-end LNA2/mixer PTAT current output (from M = 3 to M = 6),
default: M = 5
Adjusts front-end LNA PTAT current output (from M = 3 to M = 6), default: M =
5
Control of the receiver mixers output current. The current increases with
increasing setting.
FSCTRL (0x61AC) – Tune Frequency Synthesizer
Bit
No.
7:6
5:4
3:2
1:0
Name
Reset
R/W
Description
PRE_CURRENT
[1:0]
LODIV_BUF_CURRENT_TX [1:0]
LODIV_BUF_CURRENT_RX [1:0]
LODIV_CURRENT [1:0]
01
01
10
10
R/W
R/W
R/W
R/W
Prescaler current setting
Adjusts current in mixer and PA buffers. Used when TX_ACTIVE = 1
Adjusts current in mixer and PA buffers. Used when TX_ACTIVE = 0
Adjusts divider currents, except mixer and PA buffers.
FSCAL1 (0x61AE) – Tune Frequency Calibration
Bit
No.
7:2
1:0
Name
Reset
R/W
VCO_CURR[1:0]
0010 10 R/W0
11
R/W
Description
Reserved
Defines current in VCO core. Sets the multiplier between calibrated current and
VCO current. For the best value to use, see Table 23-6 in Section 23.15.1.
FSCAL2 (0x61AF) – Tune Frequency Calibration
Bit
No.
7
6
5:0
Name
Reset
R/W
Description
–
VCO_CAPARR_OE
VCO_CAPARR[5:0]
0
0
10 0000
R0
R/W
R*/W
Reserved. Read as 0
Override the calibration result with the value from VCO_CAPARR[5:0].
VCO capacitor array setting. Programmed during calibration. Override value when
VCO_CAPARR_OE = 1.
FSCAL3 (0x61B0) – Tune Frequency Calibration
Bit
No.
7
6
5:2
1:0
Name
Reset
R/W
Description
–
VCO_DAC_EN_OV
VCO_VC_DAC
[3:0]
VCO_CAPARR_CAL_CTRL[1:0]
0
0
1010
10
R0
R/W
R/W
R/W
Reserved. Read as 0
Enables the VCO DAC when 1
Bit vector for programming varactor control voltage from VC DAC.
Calibration accuracy setting for the capacitor array part of the calibration
00: 80 XOSC periods
01: 100 XOSC periods
10: 125 XOSC periods
11: 250 XOSC periods
AGCCTRL0 (0x61B1) – AGC Dynamic Range Control
Bit
No.
7
6
Name
Reset
R/W
Description
–
AGC_DR_XTND_EN
0
1
R0
R/W
5:0
AGC_DR_XTND_THR[5:0]
01 1111
R/W
Reserved. Read as 0
0: The AGC performs no adjustment of attenuation in the AAF.
1: The AGC adjusts the gain in the AAF to achieve extra dynamic range for
the receiver.
If the measured error between the AGC reference magnitude and the actual
magnitude in dB is larger than this threshold, the extra attenuation is enabled in
the front end. This threshold should be set higher than 0x0C.
This feature is enabled by AGC_DR_XTND_EN.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
267
Registers
www.ti.com
AGCCTRL1 (0x61B2) – AGC Reference Level
Bit
No.
7:6
5:0
Name
Reset
R/W
Description
–
AGC_REF[5:0]
00
01 0001
R0
R/W
Reserved. Read as 0
Target value for the AGC control loop, given in 1-dB steps. For the best value
to use see Table 23-6 in Section 23.15.1.
AGCCTRL2 (0x61B3) – AGC Gain Override
Bit
No.
7:6
Name
Reset
R/W
Description
LNA1_CURRENT[1:0]
00
R*/W
5:3
LNA2_CURRENT[2:0]
000
R*/W
2:1
LNA3_CURRENT[1:0]
00
R*/W
0
LNA_CURRENT_OE
0
R/W
Overrride value for LNA 1. Only used when LNA_CURRENT_OE = 1. When read,
this register returns the current applied gain setting.
00: 0-dB gain (reference level)
01: 3-dB gain
10: Reserved
11: 6-dB gain
Overrride value for LNA 2. Only used when LNA_CURRENT_OE = 1. When read,
this register returns the current applied gain setting.
000: 0-dB gain (reference level)
001: 3-dB gain
010: 6-dB gain
011: 9-dB gain
100: 12-dB gain
101: 15-dB gain
110: 18-dB gain
111: 21-dB gain
Overrride value for LNA 3. Only used when LNA_CURRENT_OE = 1. When read,
this register returns the current applied gain setting.
00: 0-dB gain (reference level)
01: 3-dB gain
10: 6-dB gain
11: 9-dB gain
Write 1 to override the AGC LNA current setting with the values above
(LNA1_CURRENT, LNA2_CURRENT, and LNA3_CURRENT).
AGCCTRL3 (0x61B4) – AGC Control
Bit
No.
7
6:5
Name
Reset
R/W
Description
–
AGC_SETTLE_WAIT[1:0]
0
01
R0
R/W
4:3
AGC_WIN_SIZE[1:0]
01
R/W
2:1
AAF_RP[1:0]
11
R*/W
0
AAF_RP_OE
0
R/W
Reserved. Read as 0
Time for AGC to wait for analog gain to settle after a gain change. During this
period, the energy measurement in the AGC is paused.
00: 15 periods
01: 20 periods
10: 25 periods
11: 30 periods
Window size for the accumulate and dump function in the AGC
00: 16 samples
01: 32 samples
10: 64 samples
11: 128 samples
Overrides the AGC control signals to AAF when AAF_RP_OE = 1. When read, it
returns the applied signal to the AAF.
00: 9-dB attenuation in AAF
01: 6-dB attenuation in AAF
10: 3-dB attenuation in AAF
11: 0-dB attenuation in AAF (reference level)
Write 1 to override the AGC AAF control signals with the values stored in AAF_RP.
268
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
ADCTEST0 (0x61B5) – ADC Tuning
Bit
No.
7:6
5:4
3:1
0
Name
Reset
R/W
Description
ADC_VREF_ADJ[1:0]
ADC_QUANT_ADJ[1:0]
ADC_GM_ADJ[2:0]
ADC_DAC2_EN
00
01
000
0
R/W
R/W
R/W
R/W
Quantizer threshold control for test/debug
Quantizer threshold control for test/debug
Gm control for test/debug
Enables DAC2 for enhanced ADC stability
Name
Reset
R/W
Description
ADC_TEST_CTRL[3:0]
ADC_C2_ADJ[1:0]
ADC_C3_ADJ[1:0]
0000
11
10
R/W
R/W
R/W
ADC test mode selector
Used to adjust capacitor values in ADC
Used to adjust capacitor values in ADC
ADCTEST1 (0x61B6) – ADC Tuning
Bit
No.
7:4
3:2
1:0
ADCTEST2 (0x61B7) – ADC Tuning
Bit
No.
7
6:5
Name
Reset
R/W
Description
–
ADC_TEST_MODE
0
00
R0
R/W
4:3
2:1
0
AAF_RS[1:0]
ADC_FF_ADJ[1:0]
ADC_DAC_ROT
00
01
1
R/W
R/W
R/W
Reserved. Read as 0
Test mode to enable output of ADC data from demodulator. When enabled, raw
ADC data is clocked out on the GPIO pins.
00:
Test mode disabled
01:
Data from both I and Q ADCs is output, data rate 76 MHz
10:
Data from I ADC is output. Two and two ADC samples grouped, data rate
38 MHz
11:
Data from Q ADC is output. Two and two ADC samples grouped, data rate
38 MHz
Controls series resistance of AAF
Adjust feedforward
Control of DAC DWA scheme
0: DWA (scrambling) disabled
1: DWA enabled
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
269
Registers
www.ti.com
MDMTEST0 (0x61B8) – Test Register for Modem
Bit
No.
7:4
Name
Reset
R/W
Description
TX_TONE[3:0]
0111
R/W
3:2
DC_WIN_SIZE[1:0]
01
R/W
1:0
DC_BLOCK_MODE[1:0]
01
R/W
Enables the possibility to transmit a baseband tone by picking samples from the sine
tables with a controllable phase step between the samples. The step size is
controlled by TX_TONE. If MDMTEST1.MOD_IF is 0, the tone is superpositioned on
the modulated data, effectively giving modulation with an IF. If MDMTEST1.MOD_IF is
1, only the tone is transmitted.
0000: –6 MHz
0001: –4 MHz
0010: –3 MHz
0011: –2 MHz
0100: –1 MHz
0101: –500 kHz
0110: –4 kHz
0111: 0
1000: 4 kHz
1001: 500 kHz
1010: 1 MHz
1011: 2 MHz
1100: 3 MHz
1101: 4 MHz
1110: 6 MHz
Others: Reserved
Controls the number of samples to be accumulated between each dump of the
accumulate-and-dump filter used in dc removal.
00:
32 samples
01:
64 samples
10:
128 samples
11:
256 samples
Selects the mode of operation:
00:
The input signal to the dc blocker is passed on to the output without any
attempt to remove dc.
01:
Enable dc cancellation. Normal operation
10:
Freeze estimates of dc when sync is found. Start estimating dc again when
searching for the next frame
11:
Reserved
MDMTEST1 (0x61B9) – Test Register for Modem
Bit
No.
7:5
4
Name
Reset
R/W
Description
–
MOD_IF
000
0
R0
R/W
Reserved. Read as 0
0: Modulation is performed at an IF set by MDMTEST0.TX_TONE.
1: A tone is transmitted with frequency set by MDMTEST0.TX_TONE.
3
RAMP_AMP
1
R/W
2
RFC_SNIFF_EN
0
R/W
1
MODULATION_MODE
0
R/W
0
RESERVED
0
R/W
1: Enable ramping of DAC output amplitude during startup and finish.
0: Disable ramping of DAC output amplitude
0:
Packet sniffer module disabled
1:
Packet sniffer module enabled. The received and transmitted data can be
observed on GPIO pins.
Set one of two RF modulation modes for RX/TX
0:
IEEE 802.15.4 compliant mode
1:
Reversed phase, non-IEEE compliant
Reserved. Do not write.
270
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
DACTEST0 (0x61BA) – DAC Override Value
Bit
No.
7
6:0
Name
Reset
R/W
–
DAC_Q_O[6:0]
0
R0
000 0000 R/W
Description
Reserved. Read as 0
Q-branch DAC override value when DAC_SRC = 001
If DAC_SRC is set to be ADC data, CORDIC magnitude, channel filtered data, then
DAC_Q_O controls the part of the word in question that actually is muxed to the
DAC as described in the following list.
00 0110 ≥ bits 6:0
00 0111 ≥ bits 7:1
00 1000 ≥ bits 8:2
and so on
If an invalid setting is chosen, then the DAC outputs only zeros (minimum value).
DACTEST1 (0x61BB) – DAC Override Value
Bit
No.
7
6:0
Name
Reset
R/W
–
DAC_I_O[6:0]
0
R0
000 0000 R/W
Description
Reserved. Read as 0
I-branch DAC override value when DAC_SRC = 001
If DAC_SRC is set to be ADC data, CORDIC magnitude, channel filtered data, then
DAC_I_O controls the part of the word in question that actually is muxed to the
DAC as described in the following list.
00 0110 ≥ bits 6:0
00 0111 ≥ bits 7:1
00 1000 ≥ bits 8:2
And so on
If an invalid setting is chosen, then the DAC outputs only zeros (minimum value),
DACTEST2 (0x61BC) – DAC Test Setting
Bit
No.
7:3
2:0
Name
Reset
R/W
Description
–
DAC_SRC[2:0]
0010 1
000
R0
R/W
Reserved
The TX DAC's data source is selected by DAC_SRC according to:
000:
001:
010:
011:
100:
101:
111:
Normal operation (from modulator).
The DAC_I_O and DAC_Q_O override values
ADC data after decimation, magnitude-controlled by DAC_I_O and DAC_Q_O
I/Q after decimation, channel and dc filtering, magnitude-controlled by DAC_I_O and
DAC_Q_O
Cordic magnitude output and front-end gain is output, magnitude-controlled by
DAC_I_O and DAC_Q_O
RSSI I output on the I DAC
Reserved
ATEST (0x61BD) – Analog Test Control
Bit
No.
7:6
5:0
Name
Reset
R/W
–
ATEST_CTRL[5:0]
00
R0
00 0000 R/W
Description
Reserved. Read as 0
Controls the analog test mode:
00 0000: Disabled
00 0001: Enables the temperature sensor (see also the TR0 register description in
Section 12.2.10).
00 0010 : Enables the temperature sensor in the CC2533 (see also the TR0
register description in Section 12.2.10)
Other values reserved.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
271
Registers
www.ti.com
RFRND (0x61A7) – Random Data
Bit
No.
7:2
1
0
Name
Reset
R/W
–
QRND
IRND
0000 00 R0
0
R0
0
R0
Description
Reserved. Read as 0
Random bit from the Q channel of the receiver
Random bit from the I channel of the receiver
PTEST0 (0x61BE) – Override Power-Down Register
Bit
No.
7
6
5
4
3:2
Name
Reset
R/W
Description
PRE_PD
CHP_PD
ADC_PD
DAC_PD
LNA_PD[1:0]
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
Prescaler power-down signal when PD_OVERRIDE = 1
Charge-pump power-down signal when PD_OVERRIDE = 1
Analog-to-digital converter power-down signal when PD_OVERRIDE = 1
Digital-to-analog converter power-down signal when PD_OVERRIDE = 1
1
0
TXMIX_PD
AAF_PD
0
0
R/W
R/W
Low-noise amplifier power-down signal. Defines LNA/mixer PD modes.
00:
Power up
01:
LNA off, mixer/regulator on
10:
LNA/mixer off, regulator on
11:
PD
When PD_OVERRIDE = 1
Transmit mixer power-down signal when PD_OVERRIDE = 1
Antialiasing filter power-down signal when PD_OVERRIDE = 1
PTEST1 (0x61BF) – Override Power-Down Register
Bit
No.
7:4
3
Name
Reset
R/W
Description
–
PD_OVERRIDE
0000
0
R0
R/W
2
1
0
PA_PD
VCO_PD
LODIV_PD
0
0
0
R/W
R/W
R/W
Reserved. Read as 0
Override enabling/disabling of various modules. For debug and testing only. It is
impossible to override hard-coded BIAS_PD[1:0] depenency.
Power amplifier power-down signal when PD_OVERRIDE = 1
Voltage-controlled oscillator power-down signal when PD_OVERRIDE = 1
LO power-down signal when PD_OVERRIDE = 1
272
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
RFC_OBS_CTRL0 (0x61EB) – RF Observation Mux Control
Bit
No.
7
6
5:0
Name
Reset
R/W
–
RFC_OBS_POL0
RFC_OBS_MUX0
0
R0
0
R/W
00 0000 R/W
Description
Reserved. Read as 0
The signal chosen by RFC_OBS_MUX0 is XORed with this bit.
Controls which observable signal from RF Core is to be muxed out to rfc_obs_sigs[0].
00 0000: 0 – Constant value
00 0001: 1 – Constant value
00 1000: rfc_sniff_data – Data from packet sniffer. Sample data on rising edges of
sniff_clk.
00 1001: rfc_sniff_clk – 250kHz clock for packet sniffer data.
00 1100: rssi_valid – Pin is high when the RSSI value has been updated at least once
since RX was started. Cleared when leaving RX.
00 1101: demod_cca – Clear channel assessment. See FSMSTAT1 register for details
on how to configure the behavior of this signal.
00 1110: sampled_cca – A sampled version of the CCA bit from demodulator. The
value is updated whenever a SSAMPLECCA or STXONCCA strobe is issued.
00 1111: sfd_sync – Pin is high when a SFD has been received or transmitted.
Cleared when leaving RX/TX respectively. Not to be confused with the SFD exception.
01 0000: tx_active – Indicates that FFCTRL is in one of the TX states. Active-high.
Note: This signal might have glitches, because it has no output flip-flop and is based
on the current state register of the FFCTRL FSM.
01 0001: rx_active – Indicates that FFCTRL is in one of the RX states. Active-high.
Note: This signal might have glitches, because it has no output flip-flop and is based
on the current state register of the FFCTRL FSM.
01 0010: ffctrl_fifo – Pin is high when one or more bytes are in the RXFIFO. Low
during RXFIFO overflow.
01 0011: ffctrl_fifop – Pin is high when the number of bytes in the RXFIFO exceeds the
programmable threshold or at least one complete frame is in the RXFIFO. Also high
during RXFIFO overflow. Not to be confused with the FIFOP exception.
01 0100: packet_done – A complete frame has been received. I.e., the number of
bytes set by the frame-length field has been received.
01 0110: rfc_xor_rand_i_q – XOR between I and Q random outputs. Updated at 8
MHz.
01 0111: rfc_rand_q – Random data output from the Q channel of the receiver.
Updated at 8 MHz.
01 1000: rfc_rand_i – Random data output from the I channel of the receiver. Updated
at 8 MHz
01 1001: lock_status – 1 when PLL is in lock, otherwise 0
10 1000: pa_pd – Power amplifier power-down signal
10 1010: lna_pd – LNA power-down signal
Others: Reserved
RFC_OBS_CTRL1 (0x61EC) – RF Observation Mux Control
Bit
No.
7
6
5:0
Name
Reset
R/W
–
RFC_OBS_POL1
RFC_OBS_MUX1
0
R0
0
R/W
00 0000 R/W
Description
Reserved. Read as 0
The signal chosen by RFC_OBS_MUX1 is XORed with this bit.
Controls which observable signal from RF Core is to be muxed out to rfc_obs_sigs[1].
See description of RFC_OBS_CTRL0 for details.
RFC_OBS_CTRL2 (0x61ED) – RF Observation Mux Control
Bit
No.
7
6
5:0
Name
Reset
R/W
–
RFC_OBS_POL2
RFC_OBS_MUX2
0
R0
0
R/W
00 0000 R/W
Description
Reserved. Read as 0
The signal chosen by RFC_OBS_MUX2 is XORed with this bit.
Controls which observable signal from RF Core is to be muxed out to rfc_obs_sigs[2].
See description of RFC_OBS_CTRL0 for details.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC253x Radio
273
Registers
www.ti.com
TXFILTCFG (0x61FA) – TX Filter Configuration
Bit
No.
7:4
3:0
Name
Reset
R/W
Description
–
FC
0
0xF
R0
R/W
Reserved
Sets TX anti-aliasing filter to appropriate bandwidth. Reduces spurious emissions
close to signal. For the best value to use, see Table 23-6 in Section 23.15.1.
IVCTRL (0x6265) – Analog control register (CC2533 only)
Bit
No.
7:6
5:4
Name
Reset
R/W
Description
–
DAC_CURR_CTRL
00
01
R0
R/W
3
LODIV_BIAS_CTRL
0
R/W
2
1:0
TXMIX_DC_CTRL
PA_BIAS_CTRL
0
11
R/W
R/W
Reserved. Always read as 0.
Controls bias current to DAC
00: 100% IVREF, 0% IREF bias
01: 60% IVREF, 40% IREF bias
10: 40% IVREF, 60% IREF bias
11: 0% IVREF, 100% IREF bias
Controls bias current to LODIV
1: PTAT bias
0: IVREF bias
Controls dc bias in TXMIX
Controls bias current to PA
00: IREF bias
01: IREF and IVREF bias
10: PTAT bias
11: Increased PTAT slope bias
274
CC253x Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 24
SWRU191D – April 2009 – Revised March 2013
CC2540/CC2541 Bluetooth low energy Radio
The CC2540 and CC2541 provide a Bluetooth low energy compliant radio transceiver. On the CC2540/41,
radio operation is controlled by the Bluetooth low energy stack. The application is not allowed to access
the radio directly. The application interacts with the radio by sending API commands to the stack. The TI
BLE stack with documentation is available at www.ti.com/blestack. The CC2541 may also be run in
proprietary mode; see Chapter 25 for a description of the operation in that case.
Topic
24.1
...........................................................................................................................
Registers
Page
........................................................................................................ 276
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
CC2540/CC2541 Bluetooth low energy Radio
Copyright © 2009–2013, Texas Instruments Incorporated
275
Registers
www.ti.com
24.1 Registers
The following status registers are available to the user:
RFSTAT (0x618D) RF Core Status
Bit
7
6:5
Name
MOD_UNDERFLOW
DEM_STATUS
Reset
0
00
R/W
R/W0
R
4
SFD
0
R
3
CAL_RUNNING
0
R
2
1
0
LOCK_STATUS
TX_ACTIVE
RX_ACTIVE
0
0
0
R
R
R
Description
Modulator has underflowed. Must be cleared by software
Demodulator status
00: Idle
01: Active
10: Finishing
11: Error
High when the access address has been sent in TX or when sync has been
obtained in RX
Frequency synth calibration status
0: Calibration done or not started
1: Calibration in progress
1 when PLL is in lock, otherwise 0
Status signal, active when radio is in transmit state
Status signal, active when radio is in receive state
RFC_OBS_CTRL0 (0x61AE) RF Observation Mux Control 0
Bit
7
6
5:0
Name
–
RFC_OBS_POL0
RFC_OBS_MUX0
Reset
0
0
00 0000
R/W
R0
R/W
R/W
Description
Reserved. Read as 0
The signal chosen by RFC_OBS_MUX0 is XORed with this bit.
Controls which observable signal from rf_core is to be muxed out to
rfc_obs_sigs(0).
00 0000: 0 – Constant value
00 0001: 1 – Constant value
00 1001: TX active
00 1010: RX_active
11 0000: High from when receiver has found access address until packet is
finished, low otherwise
11 0001: High from the access address has been transmitted until end of
packet, low otherwise
Other values reserved
RFC_OBS_CTRL1 (0x61AF) RF Observation Mux Control 1
Bit
7
6
5:0
276
Name
–
RFC_OBS_POL1
RFC_OBS_MUX1
Reset
0
0
00 0000
R/W
R0
R/W
R/W
CC2540/CC2541 Bluetooth low energy Radio
Description
Reserved. Read as 0
The signal chosen by RFC_OBS_MUX1 is XORed with this bit.
Controls which observable signal from rf_core is to be muxed out to
rfc_obs_sigs(1).
00 0000: 0 – Constant value
00 0001: 1 – Constant value
00 1001: TX active
00 1010: RX_active
11 0000: High from when receiver has found access address until packet is
finished, low otherwise
11 0001: High from the access address has been transmitted until end of
packet, low otherwise
Other values reserved
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
RFC_OBS_CTRL2 (0x61B0) RF Observation Mux Control 2
Bit
7
6
Name
–
RFC_OBS_POL2
Reset
0
0
R/W
R0
R/W
Description
Reserved. Read as 0
5:0
RFC_OBS_MUX2
00 0000
R/W
Controls which observable signal from rf_core is to be muxed out to
rfc_obs_sigs(2).
00 0000: 0 – Constant value
00 0001: 1 – Constant value
00 1001: TX active
00 1010: RX_active
11 0000: High from when receiver has found access address until packet is
finished, low otherwise
11 0001: High from the access address has been transmitted until end of
packet, low otherwise
Other values reserved
R/W
R0
R/W
Description
Reserved. Read as 0
Controls the analog test mode:
00 0000: Disabled
00 0001: Enables the temperature sensor (see also the TR0 register
description in Section 12.2.10).
Other values reserved.
The signal chosen by RFC_OBS_MUX2 is XORed with this bit.
ATEST (0x61A9) – Analog Test Control
Bit
7:6
5:0
Name
–
ATEST_CTRL[5:0]
Reset
00
00 0000
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
CC2540/CC2541 Bluetooth low energy Radio
Copyright © 2009–2013, Texas Instruments Incorporated
277
Chapter 25
SWRU191D – April 2009 – Revised March 2013
CC2541 Proprietary Mode Radio
In proprietary mode, the CC2541 radio supports data rates up to 2 Mbps, and has extensive baseband
automation, including auto-acknowledgment and address decoding. The RF Core controls the analog
radio module and the RF transceiver state. In addition, it provides an interface between the MCU and the
radio which makes it possible to issue commands, read status, and automate and sequence radio events.
It has 1 KB of dedicated RAM, which holds the 128-byte transmit and receive FIFO.
This chapter describes the proprietary mode operation of the CC2541 devices and features in the LLE
program. For Bluetooth low energy operation, see Chapter 24.
Topic
...........................................................................................................................
25.1
25.2
25.3
25.4
25.5
25.6
25.7
25.8
25.9
25.10
25.11
25.12
278
RF Core ..........................................................................................................
Interrupts ........................................................................................................
RF Core Data Memory ......................................................................................
Bit-Stream Processor .......................................................................................
Frequency and Channel Programming ...............................................................
Modulation Formats .........................................................................................
Receiver .........................................................................................................
Packet Format .................................................................................................
Link Layer Engine ............................................................................................
Random Number Generation ...........................................................................
Packet Sniffing ...............................................................................................
Registers .......................................................................................................
CC2541 Proprietary Mode Radio
Page
279
279
280
291
296
296
296
297
301
317
318
319
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RF Core
www.ti.com
25.1 RF Core
The RF core contains several submodules that support and control the analog radio modules. In addition,
it provides an interface between the MCU and the radio which makes it possible to issue commands, read
status, and automate and sequence radio events.
The link-layer engine (LLE) controls the RF transceiver state and most of the dynamically controlled
analog signals such as power up / down of analog modules. The LLE is used to provide the correct
sequencing of events (such as performing an FS calibration before enabling the receiver). It handles
packet assembly and decoding, including automatic length field handling, address insertion and filtering
and CRC generation and checking.
The radio data RAM holds a FIFO for transmit data (Tx FIFO) and a FIFO for receive data (Rx FIFO).
Both FIFOs are 128 bytes long and have hardware control of pointers when data is entered and removed
from the FIFOs. In addition, the RAM contains six segments of 128 bytes, one of which is used for
communication with the LLE.
The bit-stream processor is used for whitening and de-whitening transferred signals and CRC
generation and check.
The modulator transforms raw data into I/Q signals to the transmitter DAC.
The demodulator is responsible for retrieving the over-the-air data from the received signal.
The frequency synthesizer (FS) generates the carrier wave for the RF signal.
25.2 Interrupts
The radio is associated with two interrupt vectors on the CPU. These are the RFERR interrupt (interrupt
0) and the RF interrupt (interrupt 12) with the following functions.
• RFERR: Error situations in the radio are signaled using this interrupt.
• RF: Interrupts coming from normal operation are signaled using this interrupt.
The RF interrupt vector combines the interrupts in RFIF. Note that these RF interrupts are rising-edge
triggered. Thus, an interrupt is generated when, for example, the TASKDONE status flag in the RFIRQF1
register goes from 0 to 1. The RFIF interrupt flags are described in Section 25.2.1.
25.2.1 Interrupt Registers
Two main interrupt-control SFR registers are used to enable the RF and RFERR interrupts. These are the
following:
• RFERR: IEN0.RFERRIE
• RF: IEN2.RFIE
Two main interrupt-flag SFR registers hold the RF and RFERR interrupt flags. These are the following:
• RFERR: TCON.RFERRIF
• RF:S1CON.RFIF
The two interrupts generated from the RF core are a combination of several sources within the RF core.
Each of the individual sources has its own enable and interrupt flags in RF core. Flags can be found in
RFIRQF0, RFIRQF1, and RFERRF. Interrupt enable masks can be found in RFIRQM0, RFIRQM1, and
RFERRM.
The interrupt enable bits in the mask registers are used to enable individual interrupt sources. Note that
masking an interrupt source does not affect the updating of the corresponding status in the flag registers.
Due to the use of individual interrupt masks in the RF core, the interrupts coming from the RF core have
two-layered masking, and care must be taken when processing these interrupts. The procedure is
described as follows.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
279
RF Core Data Memory
www.ti.com
To clear an interrupt from the RF core one needs to clear two flags, both the flag set in the RF core and
the one set in the main interrupt flag SFR registers, S1CON or TCON (depending on which interrupt is
triggered). If a flag is cleared in the RF core and there are other unmasked flags standing, the main
interrupt flag is set. Exiting the interrupt service routine with the main interrupt flag set causes the interrupt
service routine to be executed again.
TIP: For proper handling of interrupts in ISRs, the following is advised:
• At the start of the ISR, read and store the RF core flags
• Process the interrupts
• Clear the main interrupt flag
• Clear the processed RF core flags. It is important that this is done in a single operation.
25.3 RF Core Data Memory
The radio core has 1024 bytes of data RAM divided into eight pages of 128 bytes each. The pages are to
be used as shown in Table 25-1.
Table 25-1. Radio RAM Pages
Page Number
Assignment
0
RAM-based registers
1
For Rx with auto ACK: ACK payload FIFO for addresses 2 and 3
2
For Rx with auto ACK: ACK payload FIFO for addresses 4 and 5
3
For Rx with auto ACK: ACK payload FIFO for addresses 6 and 7
4
Free for MCU use
5
Additional RAM-based registers/Reserved for LLE
6
Rx FIFO
7
Tx FIFO/for Rx with auto ACK: ACK payload FIFO for addresses
0 and 1
The active memory page is selected in register RFRAMCFG.PRE. The selected page is accessible at
XDATA addresses 0x6000–0x607F. The Rx FIFO page (page 6) is also accessible at XDATA addresses
0x6080–0x60FF. The Tx FIFO page (page 7) is also accessible at XDATA addresses 0x6100–0x617F.
A page is used for transferring parameters to the LLE, see Section 25.3.3.
There is no hardware protection to prevent the MCU from overwriting memory used by the LLE and the
FIFO. Thus the MCU should never write to page 5 (except for special dedicated registers). The MCU
should write to pages 0, 1, 2, 3, and 7 only as specified in this chapter. Writes to the FIFO pages should
only be done in ways compatible with FIFO operation, except for accessing the Tx FIFO page while
running an Rx task with auto ACK.
Pages 0, 1, 6, and 7 have retention in all power modes, whereas the contents of pages 2–5 are lost in
PM2 and PM3.
Radio core hardware registers are located at XDATA addresses 0x6180–0x61F7. Figure 25-1 shows the
mapping of radio memory to MCU XDATA memory space.
280
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RF Core Data Memory
www.ti.com
0x61F7
Radio XREG
0x6180
0x617F
Page7
Tx FIFO
Page 6
Page 5
Page 4
0x6100
0x60FF
.
.
.
Rx FIFO
Page 3
Page 2
Page 1
0x6080
0x607F
.
.
.
Radio RAM
0x6000
Page 0
M0219-01
Figure 25-1. Mapping of Radio Memory to MCU XDATA Memory Space
25.3.1 FIFOs
The FIFOs are used for transporting data between the MCU and the radio. The FIFOs have hardware
support for read and write pointer increment with circular buffering, overflow and underflow detection, and
flushing of last entry or the entire FIFO.
The Rx and Tx FIFOs are fundamentally two similar modules. Each FIFO has four pointers: the write
pointer (WP), the read pointer (RP), the start-of-packet write pointer (SWP), and the start-of-packet read
pointer (SRP). WP and RP give the index in the FIFO where the next byte is to be written and read,
respectively. SWP is used to indicate the start of the current packet being written, and SRP is used to
indicate the start of the current packet being read. The use of the pointers is indicated in Figure 25-2.
127
Last Packet
Available Data
SWP
RP
n -Packets
First Packet
SRP
Free Space
WP
0
M0220-01
Figure 25-2. FIFO Pointers
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
281
RF Core Data Memory
www.ti.com
The Tx FIFO and Rx FIFO may be accessed though the SFR register RFD (0xD9). Data is written to the
Tx FIFO when writing to the RFD register. Data is read from the Rx FIFO when the RFD register is read. In
addition, there are separate read and write registers for each FIFO (RFRXFRD, RFRXFWR, RFTXFRD,
RFTXFWR).
The Rx FIFO or Tx FIFO can be cleared by issuing CMD_RXFIFO_RESET or CMD_TXFIFO_RESET (see
Section 25.3.1.2), respectively. The contents of both FIFOs can be cleared by issuing
CMD_FIFO_RESET.
Four operations are defined to handle the four pointers:
• Deallocate is setting SRP equal to RP. This should be done when the treatment of a packet that has
been read from the FIFO is finished.
• Retry is setting RP equal to SRP. This is done to re-read a packet that has been read from the FIFO
previously.
• Discard is setting WP equal to SWP. This is done to remove a packet that had been written to the
FIFO.
• Commit is setting SWP equal to WP. This is done to confirm the writing of a packet to the FIFO and
making it available to be read out.
Using the register RFFCFG, it is possible to set up auto-commit and auto-deallocate for each of the FIFOs.
If auto-commit is enabled, SWP is set equal to WP each time a byte is written to the FIFO. If autodeallocate is enabled, SRP is set equal to RP each time a byte is read from the FIFO. By default, autocommit is enabled for the Tx FIFO and auto-deallocate is enabled for the Rx FIFO. This is also the
recommended setting. However, if packets that exceed the FIFO size are to be supported, auto-commit
must be enabled for the Rx FIFO and auto-deallocate for the Tx FIFO; see Section 25.8.1 and
Section 25.8.2 for details. If auto-commit is disabled for the Tx FIFO, the MCU must issue a commit
command after writing a packet to the Tx FIFO, and if auto-deallocate is disabled for the Rx FIFO, the
MCU must issue a deallocate command after reading a packet from the Rx FIFO.
25.3.1.1 FIFO Status and Interrupts
The XREG registers RFRXFLEN and RFTXFLEN provide information on the amount of data in the FIFOs.
This is the number of bytes between SRP and WP, i.e., the number of bytes that is not free space in
Figure 25-2. The register RFFSTATUS contains status bits for each of the FIFOs. FIFO empty is defined as
the length being 0, and FIFO full is defined as the length being 128. The amount of data between RP and
SWP is known as available data, and there is a status bit in the RFFSTATUS register telling whether there
is available data for each of the FIFOs.
An attempt to write to a full FIFO results in a FIFO overflow. The data written is then ignored and the
RXOVERF or TXOVERF flag is set in the RFERRF register, causing an RFERR interrupt. An attempt to read
from a FIFO when no data is available results in a FIFO underflow. The value read is then zero, and the
RXUNDERF or TXUNDERF flag is set in the RFERRF register, causing an RFERR interrupt.
Registers RFTXFTHRS and RFRXFTHRS are used to set threshold points for the Tx and Rx FIFOs,
respectively. Each FIFO has one status flag and two interrupt flags; when the amount of data in the FIFO
crosses the threshold, an interrupt flag is set. The FIFO status flags are available in RFFSTATUS, and the
interrupt flags are available in RFIRQF0.
When the amount of data in the FIFO is above the threshold, i.e. RFxXFLEN is greater than or equal to
RFxXFTHRS, the status bit xXDTHEX of RFFSTATUS is 1, otherwise it is 0.
When data is written to the FIFO causing the FIFO threshold to be crossed, i.e., xXDTHEX going from 0 to
1, the corresponding interrupt flag is set.
When data is read from the FIFO causing the FIFO threshold to be crossed, i.e., xXDTHEX going from 1 to
0, the corresponding interrupt flag is set.
25.3.1.2 Command Register
The command register RFST can be used for sending commands to the FIFO. Commands in the range
0x80–0xFF are commands to the FIFO. Other commands are commands to the LLE; see Section 25.9.1.
282
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RF Core Data Memory
www.ti.com
The supported FIFO commands are listed in Table 25-2. A command in the range of 0x80–0xFF that does
not match any of the listed commands is ignored.
Table 25-2. Commands to FIFO via RFST Register
Number
Command Name
Description
0x81
CMD_RXFIFO_RESET
Reset (empty) Rx FIFO. Set RFRXF* := 0
0x82
CMD_RXFIFO_DEALLOC
Deallocate Rx FIFO. This sets RFRXFSRP := RFRXFRP.
0x83
CMD_RXFIFO_RETRY
Retry Rx FIFO. This sets RFRXFRP := RFRXFSRP
0x84
CMD_RXFIFO_DISCARD
Discard Rx FIFO. This sets RFRXFWP := RFRXFSWP
0x85
CMD_RXFIFO_COMMIT
Commit Rx FIFO. This sets RFRXFSWP := RFRXFWP
0x91
CMD_TXFIFO_RESET
Reset (empty) Tx FIFO. Set RFRXF* := 0
0x92
CMD_TXFIFO_DEALLOC
Deallocate Tx FIFO. This sets RFTXFSRP := RFTXFRP.
0x93
CMD_TXFIFO_RETRY
Retry Tx FIFO. This sets RFTXFRP := RFTXFSRP
0x94
CMD_TXFIFO_DISCARD
Discard Tx FIFO. This sets RFTXFWP := RFTXFSWP
0x95
CMD_TXFIFO_COMMIT
Commit Tx FIFO. This sets RFTXFSWP := RFTXFWP
0xF1
CMD_FIFO_RESET
Reset both FIFOs
0xF2
CMD_FIFO_DEALLOC
Deallocate both FIFOs
0xF3
CMD_FIFO_RETRY
Retry both FIFOs
0xF4
CMD_FIFO_DISCARD
Discard both FIFOs
0xF5
CMD_FIFO_COMMIT
Commit both FIFOs
25.3.1.3 FIFO Pointer Operations
The FIFO pointers can be accessed directly through registers RFFRXFWP, RFFRXFRP, RFFRXFSWP,
RFFRXFSRP, RFFTXFWP, RFFTXFRP, RFFTXFSWP, and RFFTXFSRP.
Because the placement of the pointers may be the same for an empty and a full FIFO, there are internal
states distinguishing between these situations. This means that while any value can be written to the
pointer registers, certain rules must be observed for the FIFO to function reliably after the pointer write.
Any writes to a pointer must be considered to move that pointer up. Hence, writing N to a pointer already
holding N is considered equivalent to moving that pointer up 128 places, writing N-1 is equivalent to
moving the pointer up 127 places, and so on.
The pointers must maintain a specific ordering: (Going from lowest position to highest) SRP, RP, SWP,
WP.
A lower pointer may be moved up to but not past a higher pointer, whereas the highest pointer (WP) may
be moved down to, but not past the lower.
25.3.1.4 Cooperation With LLE
The LLE performs FIFO operations as part of its operation. In order to avoid conflicts between the LLE
and the MCU, access to FIFO registers should be done according to Table 25-3. Read accesses can
always be made, except for the data-read registers, which causes the read pointers to be modified. If the
MCU reads a register, one must take into account that the value may change at any time due to accesses
from the LLE. The reset FIFO commands should only be run by the MCU between LLE tasks. They are
marked with an asterisk in Table 25-3.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
283
RF Core Data Memory
www.ti.com
Table 25-3. Access to FIFO Registers
Register
Read Access
RFD
Write Access
MCU
MCU
Depends on command:
0x81 : Reset Rx FIFO: MCU*
0x82 : Deallocate Rx FIFO: MCU
0x83 : Retry Rx FIFO: MCU
0x84: Discard Rx FIFO: LLE
0x85: Commit Rx FIFO: LLE
0x91 : Reset Tx FIFO: MCU*
0x92 : Deallocate Tx FIFO: LLE (1)
0x93 : Retry Tx FIFO: LLE (1)
0x94: Discard Tx FIFO: MCU
0x95: Commit Tx FIFO: MCU
0xF1: Reset both fifos: MCU*
0xF2: Deallocate both fifos: none
0xF3: Retry both fifos: none
0xF4: Discard both fifos: none
0xF5: Commit both fifos: none
RFST (FIFO commands)
N/A
RFFDMA0
Both
MCU
RFFDMA1
Both
MCU
RFFSTATUS
Both
N/A
RFFCFG
Both
MCU
RFRXFLEN
Both
N/A
RFRXFTHRS
Both
MCU
RFRXFWR
N/A
LLE
RFRXFRD
MCU
N/A
RFRXFWP
Both
LLE
RFRXFRP
Both
MCU
RFRXFSWP
Both
LLE
RFRXFSRP
Both
MCU
RFTXFLEN
Both
N/A
RFTXFTHRS
Both
MCU
RFTXFWR
N/A
MCU
RFTXFRD
LLE
N/A
RFTXFWP
Both
MCU
RFTXFRP
Both
LLE
RFTXFSWP
Both
MCU
RFTXFSRP
Both
LLE
(1)
MCU if PRF_ADDR_ENTRYn.CONF.RETRY is 1
25.3.2 DMA
It is possible to use direct memory access (DMA) to move data between memory and the radio. See
Chapter 8 for a detailed description on how to set up and use DMA transfers.
There are two DMA triggers associated with the radio: the RADIO DMA triggers 0 and 1 (DMA triggers 19
and 11).
The radio DMA trigger source is selected in registers RFFDMA0 and RFFDMA1. See the register
descriptions in Section 25.12 for details.
284
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RF Core Data Memory
www.ti.com
25.3.3 RAM-Based Registers
A list of the memory entries of the general radio RAM area used for parameter transfer is shown in
Table 25-5. All these registers are in page 0 of the radio RAM. Each memory entry is considered a RAMbased register and has a name. Numeric values that are two bytes long are represented in little-endian
format.
The radio RAM registers have no defined reset value and must therefore be initialized by the MCU.
The registers SEMAPHORE0 and SEMAPHORE1 can be used to verify data integrity. These registers are
changed to 0 when they are read. If a semaphore register is read and the value was 1, the semaphore
has been successfully taken, and subsequent reads of the register return 0 until the semaphore is
released. If a semaphore register is read as 0, the semaphore was not free. A semaphore can be released
by writing 1 to the semaphore register; this should only be done if the semaphore has previously been
taken by the MCU. The LLE takes SEMAPHORE0 when a task starts and SEMAPHORE1 when the radio has
been set up. Both semaphores are released by the LLE at the end of the task. SEMAPHORE2 is not used
by the LLE. If the LLE is not granted the semaphore, it generates an error. If SEMAPHORE0 and
SEMAPHORE1 are taken by the MCU before registers protected by these semaphores are modified by the
MCU, data integrity is ensured, and an error occurs if the LLE is accidentally started while such an access
is going on.
Where bit numbering is used, bit 0 is the LSB and bit 7 is the MSB. Multi-byte fields are little-endian.
The detailed breakdown of the address entries ADDR_ENTRY0–ADDR_ENTRY7 is shown in Table 25-4 or
Table 25-6, depending on the operational mode.
The Prot columns of Table 25-4, Table 25-5, and Table 25-6 list the type of protection for each entry:
Sem0: Entries protected by SEMAPHORE0. Should only be written by the MCU while the LLE does not
have SEMAPHORE0. Is not modified by the LLE.
Sem1: Entries protected by SEMAPHORE1. Should only be written by the MCU while the LLE does not
have SEMAPHORE1. Is not modified by the LLE.
Sem1/R: Entries containing state variables and accumulative counters that are updated by the LLE. They
may be read by the MCU after a receive or transmit interrupt to see how many packets have been
received or transmitted. The MCU must take into account that at the time these values are read, some of
them may have been updated for the next interrupt and some not. When the LLE does not have
SEMAPHORE1, the MCU may write to them to initialize. The counters are not initialized by the LLE.
None: No semaphore protection; special rules apply for access.
Table 25-4. RAM-Based Registers
Name
PRF_CHAN
Addr
0x6000
Prot
Sem0
Description
Bits 0–6: FREQ
Frequency to use.
0: 2379 MHz
... 1 MHz steps
116: 2495 MHz
117–126: Reserved
127: The LLE does not program frequency; it is to
be set up by the MCU through the FREQCTRL
and MDMTEST1 registers.
Bit 7: SYNTH_ON
0: Turn off synthesizer when task is done.
1: Leave synthesizer running after task is done.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio 285
RF Core Data Memory
www.ti.com
Table 25-4. RAM-Based Registers (continued)
Name
Addr
Prot
Description
Configuration of task control
Bits 0–1: MODE (operation mode)
00: Basic mode, fixed length
01: Basic mode, variable length
10: Auto mode, 9-bit header
11: Auto mode, 10-bit header
Bit 2: REPEAT (repeated operation)
0: Single operation
1: Repeated operation
Bit 3: START_CONF (start configuration)
0: Start each receive/transmit immediately
1: Start each receive/transmit on Timer 2 event 1
PRF_TASK_CONF
0x6001
Sem0
Bits 4–5: STOP_CONF (stop configuration)
00: No stop based on Timer 2.
01: End task after current packet is done on Timer
2 event 2 (end immediately in sync search or wait)
10: Stop transmit/receive immediately on Timer 2
event 2
11: End task on Timer 2 event 2 in first sync
search or clear channel assessment. No stop after
first sync search or clear channel assessment.
Bit 6: TX_ON_CC_CONF
0: Listen until RSSI drops below given level, then
start Tx.
1: End task if RSSI is above given level
Bit 7: REPEAT_CONF
For TX_ON_CC with REPEAT = 1:
0: Listen again on repeated operation and
retransmissions
1: Listen only before the first transmission, then
transmit every time
For RX with REPEAT = 1:
0: Recalibrate the synthesizer before listening for
new packets
1: Recalibrate the synthesizer only when the task
starts
286 CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RF Core Data Memory
www.ti.com
Table 25-4. RAM-Based Registers (continued)
Name
Addr
Prot
Description
Configure FIFO use
Bit 0: AUTOFLUSH_IGN
Keep received packets with unexpected sequence
number in the Rx FIFO.
0: Keep
1: Auto-flush
Bit 1: AUTOFLUSH_CRC
Keep received packets with CRC error in the Rx
FIFO.
0: Keep
1: Auto-flush
Bit 2: AUTOFLUSH_EMPTY
Keep packets with no payload in the Rx FIFO.
0: Keep
1: Auto-flush
Bit 3: RX_STATUS_CONF
Rx FIFO channel information
0: Do not append RSSI and RES
1: Append RSSI and RES
PRF_FIFO_CONF
0x6002
Sem1
Bits 4–5: RX_ADDR_CONF
Rx FIFO address and config byte configuration
00: Do not include address or config byte in Rx
FIFO
01: Include received address in Rx FIFO (1-byte
addresses only), but no config byte
10: Include config byte in Rx FIFO, but no address
byte
11: Include received address (1-byte addresses
only) and config byte in Rx FIFO
Bits 6–7: TX_ADDR_CONF
Tx FIFO address and config byte configuration
00: No address or config byte; read address from
PRF_ADDR_ENTRY0
01: Include address byte in Tx FIFO, no config byte
10: Include config byte and use address index in
that byte to find address from
PRF_ADDR_ENTRYn
11: Read address from Tx FIFO followed by config
byte (where address information is ignored). Not
allowed for PRF_TASK_CONF.MODE = 00 or
01.
Packet configuration
Bit 0: ADDR_LEN. Number of address bytes (0 or
1).
PRF_PKT_CONF
0x6003
Sem0
Bit 1: AGC_EN
0: Do not use AGC
1: Use AGC (Section 25.9.2.1)
Bit 2: START_TONE
0: Ordinary transmission
1: Override extra preamble bytes with tone and
reduce synthesizer calibration time accordingly
(Section 25.9.2.2)
Bits 3-7: Reserved, always write 0.
PRF_CRC_LEN
0x6004
Sem1
Number of CRC bytes. Permitted values: 0–4
PRF_RSSI_LIMIT
0x6005
Sem1
For transmit on clear channel. Start a transmit task
by listening to the channel; start transmitting if the
RSSI drops below the level (signed) given in this
register.
PRF_RSSI_COUNT
0x6006–0x6007
Sem1
For transmit on clear channel. Number of additional
RSSI measurements that must be below the RSSI
limit before transmission takes place.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio 287
RF Core Data Memory
www.ti.com
Table 25-4. RAM-Based Registers (continued)
Name
Addr
Prot
Description
PRF_CRC_INIT
0x6008–0x600B
Sem1
Initialization value for CRC. For less than a 4-byte
CRC, the first bytes shall be 0 and the last bytes
the desired value.
PRF_W_INIT
0x600C
Sem1
Byte to write to register BSP_W before a packet;
initializes the PN7 whitener if that is used. If PN9
whitener is used, bit 7 should be 1.
PRF_RETRANS_CNT
0x600D
Sem1
Maximum number of retransmissions in automatic
retransmit
PRF_TX_DELAY
0x600E–0x600F
Sem1
Time from end of transmission to new transmission
of different payload, given in units of 62.5 ns
PRF_RETRANS_DELAY
0x6010–0x6011
Sem1
Time from end of transmission to retransmission in
auto retransmit mode, given in units of 62.5 ns
PRF_SEARCH_TIME
0x6012–0x6013
Sem1
Time to perform search before giving up or
retransmitting, given in 31.25-ns units. 0: Never
give up. Must be at least 256 if not 0.
PRF_RX_TX_TIME
0x6014–0x6015
Sem1
Time to add to Rx-Tx turnaround time in Rx with
auto ACK, given in 31.25-ns units
PRF_TX_RX_TIME
0x6016–0x6017
Sem1
Time to add to Tx-Rx turnaround time in Tx with
auto retransmission, given in 31.25-ns units
PRF_ADDR_ENTRY0
0x6018–0x6023
Address structure for address number 0. See
Table 25-5 and Table 25-6 for details.
PRF_ADDR_ENTRY1
0x6024–0x602F
Address structure for address number 1. See
Table 25-5 and Table 25-6 for details.
PRF_ADDR_ENTRY2
0x6030–0x603B
Address structure for address number 2. See
Table 25-5 and Table 25-6 for details.
PRF_ADDR_ENTRY3
0x603C–0x6047
Address structure for address number 3. See
Table 25-5 and Table 25-6 for details.
PRF_ADDR_ENTRY4
0x6048–0x6053
Address structure for address number 4. See
Table 25-5 and Table 25-6 for details.
PRF_ADDR_ENTRY5
0x6054–0x605F
Address structure for address number 5. See
Table 25-5 and Table 25-6 for details.
PRF_ADDR_ENTRY6
0x6060–0x606B
Address structure for address number 6. See
Table 25-5 and Table 25-6 for details.
PRF_ADDR_ENTRY7
0x606C–0x6077
Address structure for address number 7. See
Table 25-5 and Table 25-6 for details.
PRF_N_TX
0x6078
Sem1/R
Total number of packets transmitted
PRF_LAST_RSSI
0x6079
Sem1/R
RSSI of last received packet
PRF_LAST_DCOFF
0x607A–0x607D
Sem1/R
DC offset of last received packet
288 CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
RF Core Data Memory
www.ti.com
Table 25-4. RAM-Based Registers (continued)
Name
Addr
Prot
Description
Configure radio hardware
Bits 0–1: RXCAP
00: Do not capture on Rx packets
01: Capture start of every Rx packet
10: Capture end of every Rx packet
11: Capture start of first Rx packet only
Bits 2–3: TXCAP
00: Do not capture on Tx packets
01: Capture start of every Tx packet
10: Capture end of every Tx packet
11: Capture start of first Tx packet only
PRF_RADIO_CONF
0x607E
Sem0
Bits 4–5: TXIF: Tx IF configuration (for 2 Mbps
only)
00: Zero IF
01: ±1 MHz IF
10: ±2 MHz IF
11: ±3 MHz IF
Bit 6: DCOFF: Special dc offset handling
0: Standard dc offset
1: Use special dc offset routine measuring dc offset
right after Rx start
Bit 7: DCWB: Write back dc offset estimate to
override registers
0: Do not write back
1: Write back after each received packet with CRC
OK
PRF_ENDCAUSE
0x607F
None
Reason why LLE ended task
Table 25-5. Address Structure for Auto Mode
Name
Index
Prot
Description
Bit 0: ENA0 (Enable for primary sync word – Rx task only)
0: Disable address entry for primary sync word
1: Enable address entry for primary sync word
Bit 1: ENA1 (Enable for secondary sync word – Rx task only)
0: Disable address entry for secondary sync word
1: Enable address entry for secondary sync word
Bit 2: REUSE (Allow reuse of transmitted packet)
0: LLE deallocates packet after it has been acknowledged
1: LLE does not deallocate packet after it has been acknowledged
(this is up to the MCU)
CONF
0x00
Sem1
Bit 3: AA (Enable auto acknowledgement/auto retranmsmission)
0: Disable auto ack (Rx) or auto retransmission (Tx) for this address
1: Enable auto ack (Rx) or auto retransmission (Tx) for this address
Bit 4: VARLEN (variable length support)
0: Use fixed length given by RXLENGTH in receiver when receiving
packets or ACKs
1: Use variable length up to RXLENGTH in receiver when receiving
packets or ACKs
Bit 5: FIXEDSEQ (fixed sequence number – Tx task only)
0: Insert sequence number from SEQSTAT.SEQ
1: Read sequence number from Tx FIFO
Bit 6: TXLEN
0: Insert packet length in header when transmitting
1: Used fixed-length word when transmitting
Note: Must not be set to 1 unless the peer uses fixed length
RXLENGTH
0x01
Sem1
Maximum length of received packet (0–127)
ADDRESS
0x02
Sem1
Address of packet
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio 289
RF Core Data Memory
www.ti.com
Table 25-5. Address Structure for Auto Mode (continued)
Name
Index
Prot
Description
Bit 0: VALID (Rx task only)
0: The status is not valid. Any packet is viewed as new. On
successful reception of a packet, the LLE sets this bit.
1: The status is valid. Only packets with a sequence number and
CRC different from the previous one are accepted.
Bits 1–2: SEQ (sequence number). For Rx, the sequence number of
the last successfully received packet. For Tx, the sequence number
of the next or current packet to be transmitted
SEQSTAT
0x03
Sem1/R
Bits 3–4: ACKSEQ (ACK sequence number – Rx task only) For Rx
with auto ACK, the sequence number of the next or current ACK to
be transmitted
Bit 5: ACK_PAYLOAD_SENT (Rx task only)
0: The last received packet was not acknowledged with payload.
1: The last received packet was acknowledged with payload.
Bit 6: NEXTACK (next ACK buffer to use – Rx task only)
0: Use ACK buffer 0.
1: Use ACK buffer 1.
ACKLENGTH0
0x04
None
For Rx with auto ACK: Length of payload to be transmitted from
buffer 0. When 0, the buffer is free. After the payload has been
transmitted and a packet with a new sequence number is received,
the value is set to 0 by the LLE. The MCU only writes to the register
when it is zero; the LLE only writes it to zero when it is non-zero.
None
For Rx with auto ACK: Length of payload to be transmitted from
buffer 1. When 0, the buffer is free. After the payload has been
transmitted and a packet with a new sequence number is received,
the value is set to 0 by the LLE. The MCU only writes to the register
when it is zero; the LLE only writes it to zero when it is non-zero.
ACKLENGTH1
0x05
CRCVAL
0x06–0x07
Sem1/R
CRC value (last two bytes if more than 2 CRC bytes) of last
successfully received packet
N_TXDONE
0x08
Sem1/R
Number of packets transmitted. For auto retransmission, only
acknowledged packets with new sequence number are counted. For
auto ACK, only packets with new payload are counted when the
payload has been confirmed.
N_RXIGNORED
0x09
Sem1/R
Number of retransmitted packets received with CRC OK
N_RXOK
0x0A
Sem1/R
Number of new packets received with CRC OK or ACK packets
without payload received
N_RXNOK
0x0B
Sem1/R
Number of packets received with CRC error
Table 25-6. Address Structure for Basic Mode
Name
Index
Prot
Description
Bit 0: ENA0 (enable for primary sync word – Rx task only)
0: Disable address entry for primary sync word
1: Enable address entry for primary sync word
CONF
0x00
Sem1
Bit 1: ENA1 (enable for secondary sync word – Rx task only)
0: Disable address entry for secondary sync word
1: Enable address entry for secondary sync word
Bit 2: REUSE (allow reuse of transmitted packet)
0: LLE deallocates packet after it has been transmitted
1: LLE does not deallocate packet after it has been transmitted (this is
up to the MCU)
RXLENGTH
ADDRESS
0x01
Sem1
Maximum length of received packet (0–255)
0x02
Sem1
Address of packet
0x03–0x09
Reserved
N_RXOK
0x0A
Sem1/R
Number of packets received with CRC OK
N_RXNOK
0x0B
Sem1/R
Number of packets received with CRC error
290 CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Bit-Stream Processor
www.ti.com
25.3.4 Variables in RAM Page 5
Some additional RAM registers are placed in page 5 of the RFCORE RAM. These variables have the
prefix PRFX and are listed in Table 25-7. The addresses overlap other RAM registers, and to access them
page 5 must be selected using the RFRAMCFG register; see Section 25.3. Some of the registers have a
reset value. This value is written by the LLE shortly after it has been taken out of reset by
LLECTRL.LLE_EN being set to 1. If the MCU must modify these registers, the modification must be done
each time the LLE is reset. After taking the LLE out of reset, the MCU may modify the registers after
LLASTAT.LLE_IDLE has gone high.
Table 25-7. RAM-Based Registers in RAM Page 5 (1)
Name
Addr
Prot
Reset Val
Description
Last frequency offset estimate, read from the
FREQEST register at the end of receiving each
packet
PRFX_LAST_FREQEST
0x6006
Sem1/R
–
PRFX_RSSI_LIM_LOWER
0x6008
Sem1
0x20
Lower RSSI limit for use in AGC algorithm
PRFX_RSSI_LIM_UPPER
0x6009
Sem1
0x3C
Upper RSSI limit for use in AGC algorithm
PRFX_RSSI_DIFF
0x600A
Sem1
0x14
Difference between high and low RSSI gain
PRFX_LNAGAIN_SAT
0x600B
Sem1
0x4A
0x600C–0x600D
Sem1
0x064A
LNAGAIN setting to use while close to
saturation
Duration of tone in start of packet if
PRFX_TONE_DURATION
PRF_PKT_CONF.START_TONE = 1,
given in 31.25-ns units
PRFX_TONE_OFFSET
(1)
0x600E–0x600F
Sem0
0x0600
Time to subtract from Tx synthesizer calibration
time if PRF_PKT_CONF.START_TONE =
1, given in 31.25-ns units
Note that the LLE is reset when the device enters PM2 or PM3. This means that the PRFX registers must be re-initialized after
coming up from one of these power modes.
The parts of RAM page 5 that are not listed in Table 25-7 are reserved for use by the LLE and should not
be written by the MCU.
25.4 Bit-Stream Processor
The bit-stream processor (BSP) supports automatic insertion of CRC and detection of CRC error with a
programmable polynomial of 8, 16, 24, or 32 bits.
The bit-stream processor also supports whitening and de-whitening. The whitening sequences
supported are a PN7 sequence and a PN 9 sequence compatible with CC2500/CC2510.
The bit-stream processor is used by the LLE to do the whitening and CRC generation and checking. This
operation is based on the configuration set up by the MCU. The BSP can also be run in a coprocessor
mode to calculate whitened sequences and CRCs. This must only be done while the LLE is not running.
25.4.1 Whitening
The BSP supports two whiteners, a PN7 and a PN9 whitener. The register BSP_MODE is used to enable or
disable each whitener. When no whitener is enabled, it outputs zero. The whitener sequence is XORed
with the transmitted or received signal.
It is possible to enable both whiteners. This is useful, e.g., in conjunction with the test command
CMD_TX_TEST ( #IMPLIED) to transmit a white test signal.
25.4.1.1 PN7 Whitening
The PN7 whitener is shown in Figure 25-3. It has a 7-bit whitening shift register w used for calculating the
PN sequence given by the polynomial x 7 + x 4 + 1. The output is the same as the shift register feedback.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
291
Bit-Stream Processor
www.ti.com
The w register must be initialized by writing w into register BSP_W.W before starting receiving or
transmitting a packet. Doing this sets w 6 to BSP_W[0], w 5 to BSP_W[1] and so on up to w 1 to
BSP_W[5]; w 0 is set to 1.
When running normal receive or transmit tasks, writing to BSP_W is done by the LLE, which writes the
value in PRF_W_INIT to this register, but for test commands and co-processor mode, the BSP_W register
must be written by the MCU.
The PN7 whitener is enabled by the bit W_PN7_EN of the BSP_MODE register.
PN7 Whitening
w0
x
w1
w2
0
w3
w4
x
w5
4
w6
x
Output
7
BSP_MODE.W_PN7_EN
B0466-01
Figure 25-3. PN7 Whitening
25.4.2 CC2500-Compatible PN9 Whitening
The CC2500-compatible PN9 whitener is shown in Figure 25-4. It has a 9-bit whitening shift register s and
an 8-bit output register b. It produces a whitening sequence compatible with CC2500, CC2510 and other
TI devices. These devices use the polynomial x9 + x4 + 1. The whitening sequence is produced one byte
at a time, and the byte is bit-reversed before being XORed with a received or transmitted byte. Before
starting reception or transmission of a packet, the s and b registers must be initialized to all ones by
writing a 1 to register BSP_W.W_PN9_RESET. As for the PN7 whitener, this is done by the LLE for normal
receive and transmit tasks, provided that bit 7 of PRF_W_INIT is 1.
In Figure 25-4, the dashed arrows going from the s blocks to the b blocks denote a copy that takes place
after the whitening of one byte is done. This means that the first byte is whitened by the 8 bits that are in
the b register after initialization (all ones). As this byte is being whitened, the s register is updated. After
the first byte is whitened, the value of the s register is copied into the b register and used for whitening the
second byte.
The CC2500-compatible whitener is enabled by bit W_PN9_EN of the BSP_MODE register.
292
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Bit-Stream Processor
www.ti.com
PN9 Whitening
s8
BSP_MODE.W_PN9_EN
s7
s6
s5
s4
s3
s2
s1
s0
b0
b1
b2
b3
b4
b5
b6
b7
Output
B0467-01
Figure 25-4. CC2500-Compatible Whitening
25.4.3 CRC
A block diagram showing the operation of the CRC module is given in Figure 25-5. The CRC sub-module
has two registers:
• A 32-bit data shift register d
• A 32-bit register p for holding the polynomial
The p register defines the shift register used for calculating CRC. There is a feedback tap in the locations
where the corresponding bit of p is set to 1. The module input is XORed by the output of the shift register,
and this becomes the feedback of the shift register.
The current value of the data shift register d is the CRC value. Prior to the start of CRC calculation, the d
and p registers should be initialized by writing d to registers BSP_D[0–3] and p to registers
BSP_P[0–3]. The BSP_P[0–3] registers only must be set once, whereas the BSP_D[0–3] registers
should be set again for each packet. In normal transmit and receive modes, this is handled by the LLE,
which writes the value of PRF_CRC_INIT[0–3] to BSP_D[0–3]. At the end of CRC calculation, the value
of the register is serially shifted out on the output. When performing CRC checking, all the BSP_D[0–3]
registers should be 0 for the CRC to be OK after the received CRC has been fed through the shift register.
If whitening is enabled, calculated CRC bytes are whitened before transmission, and received CRC bytes
are de-whitened before CRC checking.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
293
Bit-Stream Processor
www.ti.com
CRC
x
x
0
p2
p1
p0
x
1
x
2
x
30
d2
d1
d0
p30
32
p31
x
31
d30
d31
Input
Output
B0468-01
Figure 25-5. CRC Module
A 32-bit CRC polynomial can be described by the equation x 32+ a 31x 31+ … + a 1x 1+ 1, where all an are 0
or 1. To represent this, each P[n] bit in the BSP_P0–BSP_P3 registers should be set to an , and P[0]
should be set to 1. To reduce the size of the polynomial to k, set the bits P[33 – k:0] to 0 and
P[32 – k] to 1. In this case, the initialization value must have zeros at D[33 – k:0]. In practice, only
polynomials of order 8, 16, 24, and 32 are supported, as the number of CRC bits produced in the
transmitter and checked in the receiver is always a multiple of 8. The number of CRC bytes produced in
normal transmit tasks is given by RAM register PRF_CRC_LEN.
This is summarized in Table 25-8 for the four CRC polynomial orders supported. In the BSP_Px column,
the numbers are binary, with the most significant bit at the left. In the PRF_CRC_INIT column, an X
indicates the initialization value to use (each X does not have to be the same). Some examples are shown
in Table 25-9.
Table 25-8. Register Settings for Different CRCs
Order
8
16
Polynomial
PRF_CRC_LEN
1
2
8
7
BSP_Px
1
x + a7x + … + a1 x + 1
x16 + a15x15 + … + a1x1 + 1
BSP_P0 = 0000 0000
BSP_P1 = 0000 0000
BSP_P2 = 0000 0000
BSP_P3 = a7 a6 a5 a4 a3 a2 a11
PRF_CRC_INIT[0] = 0
PRF_CRC_INIT[1] = 0
PRF_CRC_INIT[2] = 0
PRF_CRC_INIT[3] = X
BSP_P0 = 0000 0000
BSP_P1 = 0000 0000
BSP_P2 = a7 a6 a5 a4 a3 a2 a11
BSP_P3 = a15 a14 a13 a12 a11 a10
PRF_CRC_INIT[0] = 0
PRF_CRC_INIT[1] = 0
PRF_CRC_INIT[2] = X
PRF_CRC_INIT[3] = X
a9 a8
24
3
24
23
1
x + a23 x + … + a1 x + 1
PRF_CRC_INIT
BSP_P0 = 0000 0000
BSP_P1 = a7 a6 a5 a4 a3 a2 a11
BSP_P2 = a15 a14 a13 a12 a11 a10
a9 a8
BSP_P3 = a23 a22 a21 a20 a19 a18
PRF_CRC_INIT[0] = 0
PRF_CRC_INIT[1] = X
PRF_CRC_INIT[2] = X
PRF_CRC_INIT[3] = X
a17 a16
BSP_P0 = a7 a6 a5 a4 a3 a2 a11
BSP_P1 = a15 a14 a13 a12 a11 a10
32
4
x32 + a31 x31 + … + a1 x1 + 1
a9 a8
BSP_P2 = a23 a22 a21 a20 a19 a18
a17 a16
BSP_P3 = a31 a30 a29 a28 a27 a26
PRF_CRC_INIT[0] = X
PRF_CRC_INIT[1] = X
PRF_CRC_INIT[2] = X
PRF_CRC_INIT[3] = X
a25 a24
294 CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Bit-Stream Processor
www.ti.com
Table 25-9. Register Settings for Some Commonly Used CRCs, Assuming Initialization With All 1s
Order
8
8
16
16
24
32
CRC
PRF_CRC_LEN
1
BSP_Px
CRC-8-ATM x8 + x2 + x + 1
PRF_CRC_INIT
BSP_P0 = 0x00
BSP_P1 = 0x00
BSP_P2 = 0x00
BSP_P3 = 0x07
PRF_CRC_INIT[0] = 0x00
PRF_CRC_INIT[1] = 0x00
PRF_CRC_INIT[2] = 0x00
PRF_CRC_INIT[3] = 0xFF
PRF_CRC_INIT[0] = 0x00
PRF_CRC_INIT[1] = 0x00
PRF_CRC_INIT[2] = 0x00
PRF_CRC_INIT[3] = 0xFF
1
CRC-8 x + x + x + x + x + 1
BSP_P0 = 0x00
BSP_P1 = 0x00
BSP_P2 = 0x00
BSP_P3 = 0xD3
2
CRC-16 (used in CC2500) x16 + x15 +
x2 + 1
BSP_P0 = 0x00
BSP_P1 = 0x00
BSP_P2 = 0x05
BSP_P3 = 0x80
PRF_CRC_INIT[0] = 0x00
PRF_CRC_INIT[1] = 0x00
PRF_CRC_INIT[2] = 0xFF
PRF_CRC_INIT[3] = 0xFF
PRF_CRC_INIT[0] = 0x00
PRF_CRC_INIT[1] = 0x00
PRF_CRC_INIT[2] = 0xFF
PRF_CRC_INIT[3] = 0xFF
8
7
6
4
2
2
CRC-16-CCITT x + x + x + 1
BSP_P0 = 0x00
BSP_P1 = 0x00
BSP_P2 = 0x21
BSP_P3 = 0x10
3
CRC-24 x24 + x22 + x20 + x19 + x18 + x16
+ x14 + x13 + x11 + x10 + x8 + x7 + x6 + x3
+x+1
BSP_P0 = 0x00
BSP_P1 = 0xCB
BSP_P2 = 0x6D
BSP_P3 = 0x5D
PRF_CRC_INIT[0] = 0x00
PRF_CRC_INIT[1] = 0xFF
PRF_CRC_INIT[2] = 0xFF
PRF_CRC_INIT[3] = 0xFF
4
CRC-32-IEEE 802.3 x32 + x26 + x23 +
x22 + x16 + x12 + x11 + x10 + x8 + x7 + x5
+ x4 + x2 + x + 1
BSP_P0 = 0xB7
BSP_P1 = 0x1D
BSP_P2 = 0xC1
BSP_P3 = 0x04
PRF_CRC_INIT[0] = 0xFF
PRF_CRC_INIT[1] = 0xFF
PRF_CRC_INIT[2] = 0xFF
PRF_CRC_INIT[3] = 0xFF
16
12
5
25.4.4 Coprocessor Mode
The coprocessor mode is used to run the BSP as a stand-alone and not part of the signal path. It must not
be used while the LLE is running. Coprocessor mode is selected by setting BSP_MODE.CP_MODE to 01 or
11. In these modes, one byte to be processed is written to the BSP_DATA register, and the result of
processing this byte can later be read back from the same register. When BSP_MODE.CP_MODE is 01, the
coprocessor is in receive mode, where the whitener is applied before the CRC. When
BSP_MODE.CP_MODE is 11, the coprocessor is in transmit mode, where the whitener is applied after the
CRC.
To apply the BSP operations to a byte, write it to the BSP_DATA register. When this register is written to,
the BSP_MODE.CP_BUSY bit goes high.
If CP_MODE.CP_END is 0, the first bit provided is the LSB and the last bit is the MSB. If CP_MODE.CP_END
is 1, the first bit provided is the MSB and the last bit is the LSB.
When BSP_MODE.CP_BUSY goes low, the processed data can be read from the BSP_DATA register. If one
or both whiteners are enabled, this byte is whitened or de-whitened. Otherwise, it is the same as the byte
written, except if the CRC is being read as described in the following text.
To read out a CRC in transmit mode, set BSP_MODE.CP_READOUT to 1. A zero must be written to the
BSP_DATA register, and when BSP_MODE.CP_BUSY goes low, a CRC byte can be read from BSP_DATA.
This should be repeated for each CRC byte. If whitening is enabled, the read back CRC bytes are
whitened.
The BSP must not be set in coprocessor mode while the LLE is processing a packet.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
295
Frequency and Channel Programming
www.ti.com
25.5 Frequency and Channel Programming
For normal transmit and receive tasks, the carrier frequency is set by using the PRF_CHAN.FREQ register.
The carrier frequency is 2379 + n MHz, where n is the value of this register, and n can be from 0 to 116.
This gives a frequency range from 2379 MHz to 2495 MHz. Note that this frequency range extends
beyond the ISM band.
If PRF_CHAN.FREQ is set to 127, and for the Rx and Tx test commands, the frequency must be
programmed directly in hardware registers. In this case, the synthesizer frequency is set by programming
the 7-bit frequency word located in FREQCTRL.FREQ[6:0]. The synthesizer frequency is given by
2379 + n MHz, where n is the value of FREQCTRL.FREQ[6:0], and is programmable in 1-MHz steps.
The device supports synthesizer frequencies in the range from 2379 MHz to 2495 MHz. The usable
settings for FREQ[6:0] is consequently 0 to 116.
In Rx, the system operates on a low intermediate frequency (IF) of 1 MHz for data rates up to 1 Mbps, and
on a zero IF for 2 Mbps. In Tx, the system supports operating on low IF or zero IF. The IF to be used for
Tx can be programmed in the register MDMTEST1.TX_TONE. The receiver may operate on a positive or
negative IF when the data rate is 1 Mbps and lower; this is controlled with MDMTEST1.RX_IF.
When the symbol rate is 1 Mbps or lower and the LLE programs the frequency, it uses a ±1 MHz IF on Tx.
For both Rx and Tx, a negative IF is used when PRF_CHAN.FREQ < 62, and a positive IF is used when
PRF_CHAN.FREQ ≥ 62
When the symbol rate is 2 Mbps and the LLE programs the frequency, it uses an IF on Tx as specified in
PRF_RADIO_CONF.TXIF. This IF may be zero, or ±1 MHz, ±2 MHz, or ±3 MHz. The recommended
setting is ±1 MHz. A negative IF is used when PRF_CHAN.FREQ < 62, and a positive IF is used when
PRF_CHAN.FREQ ≥ 62.
For all data rates, the setting of MDMCTRL1.PHASE_INVERT is taken into account by the LLE when finding
the setting for MDMTEST1.TX_TONE. The FREQCTRL register is programmed corresponding to the
programmed IF in order to operate on the channel specified by PRF_CHAN.FREQ.
25.6 Modulation Formats
The CC2541 supports GFSK and MSK modulation formats. For GFSK modulation, the deviation can be
set to 160 kHz or 250 kHz (320 kHz or 500 kHz for 2 Mbps). The data rate can be set to 250 kbps, 500
kbps, 1 Mbps, or 2 Mbps. The desired modulation scheme is set in the MDMCTRL0.MODULATION register.
Not all combinations of modulation format, data rate and deviation are supported. Table 25-10 gives an
overview of supported combinations.
Table 25-10. Supported Modulation Formats, Data Rates, and Deviations
Modulation Format
Data Rate
Deviation
MDMCTRL0.MODULATION
GFSK
2 Mbps
500 kHz
0011
GFSK
2 Mbps
320 kHz
0111
GFSK
1 Mbps
250 kHz
0010
GFSK
1 Mbps
160 kHz
0110
GFSK
250 kbps
160 kHz
0100
MSK
500 kbps
–
1001
MSK
250 kbps
–
1000
25.7 Receiver
When the receiver is started, it searches for the preamble and the sync word. These are used for
frequency offset compensation and bit and byte synchronization. The sync word can be programmed to be
from 16 to 32 bits.
296
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Packet Format
www.ti.com
Checking the sync word is done in a two-stage process. First, a correlation value is calculated. If this
correlation is above a programmable threshold, a data decision of the received sync word is done. It can
be programmed in MDMCTRL3.SYNC_MODE whether this data decision is to be ignored, no bit errors are to
be accepted, or one bit error is to be accepted. The correlation threshold value is programmed in
MDMCTRL1.CORR_THR. This threshold value should depend on the sync word length. As a rule of thumb,
a value of 0.25 times the number of bits (rounded down) can be used.
For the bit synchronization to work well, some guidelines should be followed for the sync word. It should
have enough transitions, but not long runs of 10 1010... or other short, repeated patterns. Generally, a
longer sync word gives better performance.
The CC2541 devices have support for two independent sync words. The primary and secondary sync
words are specified in two sets of registers. The secondary sync word can be enabled by the
SW_CONF.DUAL_RX bit, and if enabled, the received signal is correlated against both sync words. If the
correlation with one of the sync words is above the threshold, data decision is done against that sync
word.
While the receiver is running, a received signal strength indicator (RSSI) is updated. The RSSI is available
some time after the receiver is started, regardless of whether sync is found. It can be read from the RSSI
register, which is 0x80 when no RSSI is available. The value given is in the range 0 to approximately 64,
with a change of 1 corresponding to a 1-dB change. The offset from a true dBm value depends on the
receiver mode and can be found in the device data sheet. For high received signal levels, the reported
RSSI saturates at one of the highest possible reported values. The accuracy and update time of the RSSI
can be traded off using MDMTEST0.RSSI_ACC. The RSSI can be calculated over a window of 5.33 µs or
21.3 µs, and 1, 2, or 4 such windows can be averaged to give the result. Using a longer average time
gives higher accuracy, but it takes longer before a result is ready, and doing the average over a longer
time means that the result may be wrong for short packets. An average of n windows of length t RSSIshould
only be used for packets lasting longer than (n + 1) tRSSI (including preamble, sync word, and CRC).
The receiver must run dc offset estimation and removal. The dc offset estimation mode can be controlled
with MDMTEST0.DC_BLOCK_MODE. For data rates of 1 Mbps and lower, where the receiver runs on a low
IF, it is recommended to use the default setting for this register (continuous estimation). For 2 Mbps,
where the receiver runs on zero IF, delayed dc offset estimation should normally be used. This causes the
dc offset estimation to be done in front of the packet. The delay can be controlled through
MDMTEST0.DC_BLOCK_LENGTH and MDMTEST1.DC_DELAY. The recommendation is to set
MDMTEST0.DC_BLOCK_LENGTH to 11 (128 samples) and MDMTEST1.DC_DELAY to 00 (5 delays), which
allows for up to approximately 105 µs of energy in front of the packet payload, including the preamble and
sync word. As an alternative for 2 Mbps, dc offset estimation can be turned off, and a previously found
value can be used, written into the DC_I_L, DC_I_H, DC_Q_L, and DC_Q_H registers. Values can be
found in advance, but differ for each frequency. For auto acknowledgments and other packets that are
received at a known time, the LLE can perform a special dc offset algorithm as described in
Section 25.9.2.
25.8 Packet Format
The packet format is configurable. There are two operation modes for radio packet control, basic mode
and auto mode. Of these, only auto mode supports automatic acknowledgment and retransmissions. The
LLE-controlled part of the packet format is also different for the two modes. In basic mode, there is an
optional length field followed by an optional address of 1 byte, as shown in Figure 25-6. In auto mode,
there is a 9-bit or 10-bit header field containing length and sequence number information. This format is
shown in Figure 25-7. The figures show the packet formats with their configurability. The fields with a
header in gray are controlled directly by the modem and are used in the acquisition of received packets.
The fields with header in white are controlled by the LLE.
Preamble
Sync word
1–16 bytes
16–32 bits
Handled by modem
Length
Address
Payload
CRC
0–1 byte
0–1 byte
0–255 bytes
0–4 bytes
R0009-01
Figure 25-6. Air Interface Packet Format for Basic Mode
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
297
Packet Format
www.ti.com
Preamble
Sync word
1–16 bytes
16–32 bits
Address
Header
Payload
CRC
0–1 byte
9–10 bits
0–127 bytes
0–4 bytes
Handled by modem
R0010-02
When using a 9 bit header, the payload length is limited to the range 0-63 bytes. Note that the LLE must be reset
when the device enters PM2 or PM3. This means that the PRFX registers must be re-initialized after the LLE has
been re-enabled after coming up from one of these power modes.
Figure 25-7. Air Interface Packet Format for Auto Mode
The preamble is a sequence of 1010 1010 or 0101 0101. It can be from 1 to 16 bytes. The type of
preamble and the number of bytes can be set up in the MDMCTRL2 register.
The Sync word field is a synchronization word that can have any length from 16 to 32 bits. The length is
programmed in the SW_CONF.SW_LEN register. The sync word itself is programmed in the SW0, SW1, SW2,
and SW3 registers for the primary sync word, and SW4, SW5, SW6, and SW7 for the secondary sync word.
The bit ordering of the sync word is set up with MDMCTRL2.SW_BIT_ORDER. If SW_BIT_ORDER is 0, the
LSB of SW0 (SW4) is transmitted first and the MSB of SW3 (SW7) is transmitted last. If SW_BIT_ORDER
is 1, the MSB of SW3 (SW7) is transmitted first and the LSB of SW0 (SW4) is transmitted last. The first bit
transmitted is always the same regardless of the sync word length; the unused bits for sync word length of
less than 32 bits are the ones that would have been transmitted last.
The optional length byte in basic mode (see Figure 25-6) is present if PRF_TASK_CONF.MODE = 01. It
indicates the number of address and payload bytes following the length byte. If the length field is not
present, the length is fixed as described in Section 25.9.2.
The optional address is 1 byte if present; the length is configured with the PRF_PKT_CONF.ADDR_LEN
register. In the transmitter, the address can be used for identification or to direct the message to a
particular receiver, and in the receiver, the address can be used to filter out messages from unknown or
unwanted transmitters and to distinguish between messages from different transmitters. See
Section 25.9.2 for details on how the address is used. Note that for the packet format in Figure 25-7 or if a
length field is not used, the address field immediately follows the sync word, and can thus be seen as an
extension of it.
The 9-bit or 10-bit header shown in Figure 25-7 is shown in more detail in Figure 25-8 and Figure 25-9.
This field consists of a 6-bit or 7-bit length followed by a 2-bit sequence number and a flag called
NO_ACK (NOA in Figure 25-8 and Figure 25-9) to inform that acknowledgment of the packet is not
expected. If the configuration is to use a fixed length, the value of the length field is ignored in the
receiver. It can be configured always to set the length field to 110011 in the transmitter for fixed-length
packets.
Length
Bit 8
Bit 7
Bit 6
SEQ
Bit 5
Bit 4
Bit 3
Bit 2
NOA
Bit 1
LSB
R0011-01
Figure 25-8. Bits of 9-Bit Header
Length
Bit 9
Bit 8
Bit 7
Bit 6
SEQ
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
NOA
LSB
R0012-01
Figure 25-9. Bits of 10-Bit Header
The payload can be from zero to 255 bytes in basic mode, but the sum of the number of address and
payload bytes must not exceed 255. In auto mode, the payload can be from 0 to 63 bytes with a 9-bit
header or 0 to 127 bytes with a 10-bit header. The maximum packet length can be limited, see
Section 25.9.2.3.1 and Section 25.9.2.3.2
298
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Packet Format
www.ti.com
The bit ordering when transmitting the length, address, payload, and CRC bytes is set up with the
ENDIANNESS bit of the FRMCTRL0 register; if 0, the LSB of each byte is transmitted first and if 1, the MSB
is transmitted first. Normally, FRMCTRL0.ENDIANNESS and MDMCTRL2.SW_BIT_ORDER should have the
same value. Note that for correct operation in auto mode, FRMCTRL0.ENDIANNESS must be set to 1 so
that MSB is transmitted first.
The CRC field contains 0 to 4 bytes and is used to check the packet for errors if present. See
Section 25.4.3 on how to set up the CRC generation and checking.
25.8.1 Rx FIFO Packet Organization
Length
Address
config
1 Byte
0–1 Bytes
0–1 Bytes
Address Index
SW
Bit 1
Bit 3
LSB
Bit 2
Unused NOA
Bit 4
Bit 5
Status
Payload
(Length – (Length of Address + config)) Bytes
SEQ
Bit 6
MSB
Address Index
SW
Bit 1
Bit 3
LSB
Bit 2
RSSI
RES
1 Byte
1 Byte
Unused
Bit 4
Bit 5
IGN
CRC
Bit 6
MSB
R0013-02
Figure 25-10. Structure of Packets in the Rx FIFO
The structure of a packet in the Rx FIFO is shown in Figure 25-10. All packets start with a length byte,
regardless of whether a length byte is present on the air. The length is the number of bytes in the address,
config, and payload fields following the length byte, and it may be modified compared to the length
received on the air or configured as fixed-length. If packets are longer than what can fit in the FIFO,
packets must be read from the FIFO while reception takes place, either by DMA or directly by the MCU.
The auto-flush options in PRF_FIFO_CONF cannot be used in this case, and auto-commit and autodeallocate must be enabled for the Rx FIFO in RFFCFG.
The address byte is placed after the length byte and is present if configured in
PRF_FIFO_CONF.RX_ADDR_CONF. The address is written in the FIFO as it was received on the air.
The config byte following the length byte and address byte is present if configured in
PRF_FIFO_CONF.RX_ADDR_CONF. In this case, the index n to the PRF_ADDR_ENTRYn containing the
received address is present in bits 0–2, and bit 3 is 0 if the primary sync word was received and 1 if the
secondary sync word was received. In auto mode, the 3 MSBs of the config byte are set to the 3 LSBs of
the received header.
The payload is as received on the air. In case of an empty packet, there is no payload.
The status field consists of 2 bytes appended to the FIFO entry if configured in
PRF_FIFO_CONF.RX_STATUS_CONF. The presence of a status field is not reflected in the value of the
length byte, so if a status field is present, the MCU must read 2 extra bytes. It is possible to configure this
even using DMA with automatic length extraction. The status bytes are:
• RSSI is the received signal-strength indication from the demodulator.
• RES contains information on the address and CRC result.
– The 3 LSBs contain the address index as in the config byte.
– Bit 3 is 0 if the primary sync word was received and 1 if the secondary sync word was received.
– IGN is 1 for packets that may be ignored by the MCU due to repeated sequence number and 0
otherwise.
– CRC is 1 if there was a CRC error and 0 otherwise.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
299
Packet Format
www.ti.com
25.8.2 Tx FIFO Packet Organization
Length
Address
config
Payload
1 Byte
0–1 Bytes
0–1 Bytes
(Length – (Length of Address + config)) Bytes
Unused
Address Index
LSB
Bit 1
Bit 2
Bit 3
Bit 4
SEQ
NOA
Bit 5
Bit 6
MSB
R0014-02
Figure 25-11. Structure of Packets in the Tx FIFO
The structure of a packet in the Tx FIFO is shown in Figure 25-11. All packets start with a length byte,
regardless of whether a length byte is present on the air. The length is the number of bytes in the address,
config, and payload fields following the length byte, and it may be modified before being transmitted on the
air. If a fixed length is used, it is up to the MCU to ensure that the length is correct given the fixed length
expected by the receiver. If packets are longer than what can fit in the FIFO, packets must be written to
the FIFO while transmission takes place, either by DMA or directly by the MCU. Auto-commit and autodeallocate must then be enabled for the Tx FIFO in RFFCFG.
The address byte is placed after the length byte and is present if configured in
PRF_FIFO_CONF.TX_ADDR_CONF. If it is included, the address is transmitted on the air as it is read from
the FIFO. If it is not included, but a config byte is included, the three LSBs of the config byte tell the index
n of PRF_ADDR_ENTRYn from which the address is inserted. If neither an address nor a config byte is
included, the address is inserted from PRF_ADDR_ENTRY0.ADDRESS.
The config byte following the length byte and optional address byte is present if configured in
PRF_FIFO_CONF.TX_ADDR_CONF. This byte contains an address index which is used to determine the
address if no address byte is included as explained previously. If an address byte is included, the address
index is used to determine which address entry to read the configuration from, but the ADDRESS field in
that address entry is ignored. In auto mode, the NO_ACK bit (LSB) of the transmitted header is set to bit 5
of the config byte. If PRF_ADDR_ENTRYn.CONF.FIXEDSEQ, where n is the index of the address used, is
1, the SEQ field of the transmitted header is taken from the SEQ field (bits 6–7) of the config byte;
otherwise, the sequence number on the air is inserted from PRF_ADDR_ENTRYn.SEQSTAT.SEQ. If the
config byte is not included, the NO_ACK bit is always sent as 0 and
PRF_ADDR_ENTRYn.CONF.FIXEDSEQ should be 0 (otherwise the SEQ field always remains 0).The
payload is transmitted as present in the FIFO.
25.8.3 Tx Buffers for ACK Payload
The hardware Tx FIFO is not used for ACK payload in Rx tasks in auto mode. Instead, an
acknowledgment packet for each address can be placed in one of two dedicated buffers for that address.
These two buffers constitute a FIFO capable of holding two packets. The buffers for the first two
addresses are placed in the RAM page normally used for the hardware Tx FIFO. These four buffers can
either be accessed from the Tx FIFO space at 0x6100 or by selecting page 7 through RFRAMCFG, but the
Tx FIFO registers should not be used. The other twelve buffers must be addressed from the configurable
radio memory bank through the RFRAMCFG register. The mapping of each buffer is shown in Table 25-11.
Table 25-11. Segments for Holding ACK Payload for Each Address Entry
Address Entry Number
Buffer Number
Setting of RFRAMCFG
Start Address
0
0
7 or X
0x6000 or 0x6100
0
1
7 or X
0x6020 or 0x6120
1
0
7 or X
0x6040 or 0x6140
1
1
7 or X
0x6060 or 0x6160
2
0
1
0x6000
300 CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Link Layer Engine
www.ti.com
Table 25-11. Segments for Holding ACK Payload for Each Address Entry (continued)
Address Entry Number
Buffer Number
Setting of RFRAMCFG
Start Address
2
1
1
0x6020
3
0
1
0x6040
3
1
1
0x6060
4
0
2
0x6000
4
1
2
0x6020
5
0
2
0x6040
5
1
2
0x6060
6
0
3
0x6000
6
1
3
0x6020
7
0
3
0x6040
7
1
3
0x6060
The status of buffer k for address n is contained in the PRF_ADDR_ENTRYn.ACKLENGTHk register. If the
value is 0, the buffer is free.
In order to enter a payload for address n, the MCU must follow the following procedure:
1. Read PRF_ADDR_ENTRYn.ACKLENGTH0 and PRF_ADDR_ENTRYn.ACKLENGTH1. Call the values
len_0 and len_1, respectively.
2. Read PRF_ADDR_ENTRYn.SEQSTAT.NEXTACK and call this value k. Let m be NOT k (i.e., 1 – k).
3. Check if len_k is 0. If so, write the payload to buffer k for address entry n (see Table 25-11), then write
the payload length to PRF_ADDR_ENTRYn.ACKLENGTHk. End the procedure.
4. Otherwise, check whether len_ m is 0. If so, write the payload to buffer m for address entry n (see
Table 25-11), then write the payload length to PRF_ADDR_ENTRYn.ACKLENGTHm. End the procedure.
5. Otherwise, no ACK payload buffer for that address is free, and no payload can be entered at this time.
The ACK payload length can be 1–32. When a buffer becomes free, the LLE writes the
PRF_ADDR_ENTRYn.ACKLENGTHk to 0 and raises a TXDONE interrupt.
A buffer contains only the payload to be transmitted. The length is given by
PRF_ADDR_ENTRYn.ACKLENGTHk, and the address and sequence number are as described in
Section 25.9.2.3.2.
In order to flush the buffers for address n, issue the command CMD_FLUSH_ACK n (see Table 25-12).
This causes the LLE to write PRF_ADDR_ENTRYn.ACKLENGTH0 and PRF_ADDR_ENTRYn.ACKLENGTH1
to 0 and clear PRF_ADDR_ENTRYn.SEQSTAT.ACK_PAYLOAD_SENT. If no task is running, the LLE takes
SEMAPHORE1; if it fails, it does not write to PRF_ADDR_ENTRYn.SEQSTAT.ACK_PAYLOAD_SENT. If the
transmission of an acknowledgment with payload had started on that address, flushing happens after the
transmission is finished. After the flushing is done, the LLE raises a TXFLUSHED interrupt.
25.9 Link Layer Engine
The link layer engine controls radio operation. It is started by setting the LLECTRL.LLE_EN bit to 1. The
LLE must be started before the radio can be operated.
The LLE can be reset by clearing and setting LLECTRL.LLE_EN. The LLE should not be reset while the
radio is active. The MCU should not enter PM1, PM2, or PM3 while the LLE is running a task. Before
entering PM2 or PM3, LLECTRL.LLE_EN must be set to 0, otherwise the behavior of the RF core may be
unpredictable after waking up. The mode of the LLE is selected with LLECTRL.LLE_MODE_SEL. For the
proprietary-mode operation described in this chapter, this field must be written to 00. For BLE operation,
this field is 01; that value should only be written by the TI BLE stack. In order to switch modes, the LLE
must be reset; writing to LLECTRL.LLE_MODE_SEL while LLECTRL.LLE_EN is 1 has no effect.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
301
Link Layer Engine
www.ti.com
25.9.1 Command Register
The command register RFST can be used for sending commands to the LLE and the FIFOs. Commands
in the range 0x80–0xFF are commands to the FIFOs, see Section 25.3.1. Other commands are
commands to the LLE.
The commands are listed in Table 25-12. There are commands for starting receive and transmit modes. In
addition, there is a command CMD_SHUTDOWN to stop the radio operation and end the task directly.
The commands CMD_SEND_EVENT1 and CMD_SEND_EVENT2 do the same action as receiving a
Timer 2 event 1 or event 2.
If an unknown command is entered, the LLE responds by generating an LLEERR interrupt. If a task is
running, it stops.
When sending a command to the LLE, the RFST register retains its value until the LLE has received the
command (but not necessarily executed it) and then is set to 0. Commands should not be sent to the LLE
unless RFST is 0. FIFO commands may be sent at any time.
Table 25-12. Commands From MCU to LL Engine via RFST Register
Number
Command Name
Description
0x01
CMD_SHUTDOWN
Stop operation immediately
0x02
CMD_DEMOD_TEST
Start demodulator without sync search
0x03
CMD_RX_TEST
Start demodulator and sync search
0x04
CMD_TX_TEST
Start transmitter and transmit zeros
0x05
CMD_TX_FIFO_TEST
Start transmitter and transmit from Tx FIFO
0x06
CMD_PING
Respond with a PINGRSP interrupt
0x08
CMD_RX
Start receive operation
0x09
CMD_TX
Start transmit operation
0x0A
CMD_TX_ON_CC
Start transmit operation on clear channel
0x0B
CMD_STOP
Gracefully stop radio task
0x21
CMD_SEND_EVENT1
Do the same action as if Timer 2 event 1 was observed
0x22
CMD_SEND_EVENT2
Do the same action as if Timer 2 event 2 was observed
0x30
CMD_FLUSH_ACK0
Flush the ACK payload buffers for address 0
0x31
CMD_FLUSH_ACK1
Flush the ACK payload buffers for address 1
0x32
CMD_FLUSH_ACK2
Flush the ACK payload buffers for address 2
0x33
CMD_FLUSH_ACK3
Flush the ACK payload buffers for address 3
0x34
CMD_FLUSH_ACK4
Flush the ACK payload buffers for address 4
0x35
CMD_FLUSH_ACK5
Flush the ACK payload buffers for address 5
0x36
CMD_FLUSH_ACK6
Flush the ACK payload buffers for address 6
0x37
CMD_FLUSH_ACK7
Flush the ACK payload buffers for address 7
25.9.2 Radio Tasks
Before starting a task, radio registers should be set up with the desired packet format, and the desired
input sensitivity and output power should be programmed. Furthermore, the sync word in use must be
programmed in the SW0, SW1, SW2, and SW3 registers. If a secondary sync word is used, it must be
programmed in the SW4, SW5, SW6, and SW7 registers. The RAM registers must be programmed to
configure the task. The way the task runs depends on the PRF_TASK_CONF register. The operation mode
is set up by the MODE bits of this register. A value of 00 or 01 gives basic mode and thus disables auto
ACK or auto retransmission. A value of 10 or 11 gives auto mode where auto acknowledgment or auto
retransmission can be enabled per the address in PRF_ADDR_ENTRYn.CONF.AA.
All tasks start with a start-of-task command from the MCU. The LLE takes SEMAPHORE0 at this time; if the
semaphore is not available, the task ends with an error. Depending on the configuration in
PRF_TASK_CONF.START_CONF, the LLE either starts the task immediately or waits for a Timer 2 event 1
before starting. Note that a Timer 2 event 1 may be pending from before the LLE starts waiting; in that
case, the task starts immediately. To clear a pending Timer 2 event 1, reset the LLE. To prevent
302
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Link Layer Engine
www.ti.com
unwanted events from reaching the LLE, Timer 2 event 1 can be disabled in the T2EVTCFG register; see
Chapter 22. The frequency word is programmed according to the setting of PRF_CHAN.FREQ, except if it
is 127, in which case no frequency programming is done and any value written by the MCU is retained.
When using auto mode on 2 Mbps, the frequency must be programmed through the PRF_CHAN.FREQ
register. Then the LLE will change the IF frequency automatically (for 2 Mbps, the recommended settings
use different IF for transmission and reception) when changing from receive operation to transmit
operation (for sending an acknowledgement packet) and vice versa. The LLE starts configuring the
transmitter or receiver, depending on the type of task. After the transmitter or receiver has been set up,
the LLE takes SEMAPHORE1 to gain access to the remaining RAM-based registers, read the parameters,
and start transmission or reception.
Programming of frequency is done as described in Section 25.5. For symbol rates of 1 Mbps and lower,
Rx and Tx are done on the same synthesizer frequency, whereas for a symbol rate of 2 Mbps, the
synthesizer frequency changes between Rx and Tx. This change is done without a recalibration of the
synthesizer.
At the end of a packet, the LLE reads the RSSI register and writes the value to the PRF_LAST_RSSI
register and, if so configured, to the RSSI byte of the Rx FIFO. This read is done after the next-to-last byte
has been obtained from the demodulator. Note that for a bit rate of 2 Mbps and for sync words shorter
than 32 bits, MDMCTRL3.RSSI_MODE should be set to 11 to ensure a correct reading. Before turning off
the demodulator, the LLE reads the dc offset from the DC_I_L, DC_I_H, DC_Q_L, and DC_Q_H registers
and writes the result to PRF_LAST_DCOFF (in the byte order listed for the register read). The LLE also
reads the frequency offset from the FREQEST register and writes the result to PRFX_LAST_FREQEST (see
Table 25-7).
If PRF_RADIO_CONF.DCOFF is 1, the LLE runs a procedure that estimates the dc offset right after
receiver startup. This mode is suitable for packets that are known to be received at a certain time, such as
acknowledgment packets. In this mode, the LLE starts the receiver with normal dc cancellation mode and
forces the LNA gain to minimum. After a short time, the LLE reads out the value of the dc offset estimate,
writes it into the override registers, and selects manual override mode for dc offset estimation. It sets the
LNA gain back to the programmed value and after a waiting time to allow the LNA to stabilize, starts sync
search. The time to start Rx with this mode is the same as for ordinary start of Rx.
If PRF_RADIO_CONF.DCWB is 1, the LLE writes the dc offset estimate read out at the end of the packet
into the dc offset override register, provided that the received packet did not have a CRC error. This is
suited for the delayed dc offset mode, where the override value for dc offset is used before a delayed dc
offset is available.
Some of the RAM registers are checked by the LLE to verify that their values are permitted. This applies
to PRF_CHAN.FREQ, PRF_FIFO_CONF.TX_ADDR_CONF, and PRF_CRC_LEN. If any of these registers
have values that are not permitted, the task ends with an error.
A CMD_SHUTDOWN command, undefined command, or any command starting a new task, ends the task
immediately. If a packet was being transmitted or received, an RXTXABO interrupt to the MCU is raised.
This means that to avoid unwanted abort of commands, the CPU should wait for a TASKDONE interrupt
or check that LLESTAT.LLE_IDLE is 1 before starting another command.
If a CMD_STOP command is received, the task ends after the current reception or transmission is done.
Timer 2 event 2 can be configured to end a task: If PRF_TASK_CONF.STOP_CONF is 01, Timer 2 event 2
behaves as a CMD_STOP, and if PRF_TASK_CONF.STOP_CONF is 10, Timer 2 event 2 behaves as a
CMD_SHUTDOWN. Setting PRF_TASK_CONF.STOP_CONF to 00 disables Timer 2 event 2 as a stop
event. With the 11 setting, Timer 2 event 2 only applies to sync search or listen right after a CMD_RX or
CMD_TX_ON_CC (this setting is not meaningful for a CMD_TX task) or a start by Timer 2 event 1. This is
explained in later subsections.
Timer 2 may capture the time of a packet based on the setting in PRF_RADIO_CONF. The fields TXCAP
and RXCAP decide how capture is configured for Tx and Rx, respectively; see Table 25-13. The captured
value can be read from the registers T2M0, T2M1, T2MOVF0, T2MOVF1, and T2MOVF2 when t2_cap and
t2ovf_cap are selected using the T2MSELregister; see Chapter 22.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
303
Link Layer Engine
www.ti.com
Table 25-13. Timer 2 Capture Settings
TXCAP
Description
00
Capture of transmitted packets off
01
Capture the start (after the sync word) of every transmitted packet
10
Capture the end of every transmitted packet
11
Capture the start of the first transmitted packet, i.e., capture of transmitted packets is turned off
after a packet has been transmitted.
RXCAP
Description
00
Capture of received packets off
01
Capture the start (after the sync word) of every received packet
10
Capture the end of every received packet
11
Capture the start of the first received packet, i.e., capture of received packets is turned off after
a packet has been fully received.
When capture is done at the beginning of a packet, the time captured is the time right after the sync word
has been received or transmitted. Setting TXCAP or RXCAP to 11 enables capture at the start of a packet,
but the capture is turned off after a packet has been transmitted or fully received in a task, so it is the start
of the first packet in the task that is captured. The MCU should normally only read the captured value after
a task is done; otherwise, the captured value may be overwritten with a new value. The user must take
into account that a timer value may be captured on a received packet that does not match the address or
that has a length which is not permitted, and that is thus not reported. It is possible to turn on capture for
both received and transmitted packets in the same task. If so, it is up to the user to determine if the
captured value was from a received or transmitted packet.
When a task is finished, the LLE writes an end-of-task cause in PRF_ENDCAUSE, frees the semaphores,
raises a TASKDONE interrupt, and halts its operation. The possible values of PRF_ENDCAUSE are listed in
Table 25-14.
If PRF_CHAN.SYNTH_ON is 1, the synthesizer is not turned off after the task ends. This can be used to
start a new task immediately on the same channel and get faster start of Rx or Tx. To do so, the next task
should be started with PRF_CHAN.FREQ set to 127. Note that the synthesizer should not be allowed to run
for a long time after a task has ended, as this causes excessive power consumption. The synthesizer can
be stopped by sending a CMD_SHUTDOWN command.
Table 25-14. End-of-Task Causes
Number
Name
Description
Normal task ending
0
TASK_ENDOK
Task ended normally
1
TASK_RXTIMEOUT
Timer 2 event 2 or CMD_STOP observed while waiting for Rx
sync
2
TASK_NOSYNC
Sync was not obtained in the specified time
3
TASK_NOCC
TX_ON_CC ended because channel was not clear
4
TASK_MAXRT
Task ended because maximum number of retransmissions was
reached
5
TASK_STOP
Task ended after transmission or reception by Timer 2 event 2
or CMD_STOP while transmitting or receiving or with ACK or
retransmission in progress
6
TASK_ABORT
Task aborted by command
255
TASKERR_INTERNAL
Internal program error
254
TASKERR_CMD
Unknown command
253
TASKERR_SEM
Unable to obtain semaphore
252
TASKERR_PAR
Unpermitted parameter
251
TASKERR_TXFIFO
Tx FIFO without available data when not permitted
250
TASKERR_RXFIFO
Overfull Rx FIFO in Tx task
MCU interface error
304 CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Link Layer Engine
www.ti.com
Table 25-14. End-of-Task Causes (continued)
Number
249
Name
TASKERR_MODUNF
Description
Modulator underflow observed
25.9.2.1 AGC Algorithm
If PRF_PKT_CONF.AGC_EN is 1, an automatic gain control (AGC) algorithm is run while the receiver is
looking for sync. The AGC algorithm switches between two different front-end gain settings in the
LNAGAIN register. It is recommended to use AGC when running on 2 Mbps to improve the saturation
performance.
Parameters for control of the AGC algorithm are found in page 5 of the radio RAM; see Table 25-7. This
table lists reset values that the LLE sets for these parameters.
The LLE polls the RSSI value at every update and compares it to the values of PRFX_RSSI_LIM_LOWER
and PRFX_RSSI_LIM_UPPER. If the observed RSSI is below PRFX_RSSI_LIM_LOWER, the LNA gain is
set to the high gain setting. If the observed RSSI is above PRFX_RSSI_LIM_UPPER, the LNA gain is set
to the low gain setting. If the observed RSSI is between these limits, the LNA gain is not changed.
The high gain to use is the value found in the LNAGAIN register when the task is started. The low gain to
use is the value found in the PRFX_LNAGAIN_SAT RAM register.
The PRFX_RSSI_LIM_LOWER and PRFX_RSSI_LIM_UPPER values must differ in order to account for the
difference that is observed from the RSSI register when the LNA gain is changed and to have hysteresis
to avoid too-frequent gain changes.
When sync is obtained on the receiver, the AGC algorithm stops updating the LNA gain, which remains at
the value last set. When the receiver is switched off, the LNAGAIN register is set back to the value it had
when the task started, i.e., the high gain setting.
When the gain is reduced during the reception of a packet, the value found in the PRF_LAST_RSSI
register and (if configured) in the RSSI byte of the Rx FIFO is updated to reflect this. This update is done
by adding the value of the register PRFX_RSSI_DIFF to the value found in the RSSI register.
PRFX_RSSI_DIFF should therefore contain the difference between the RSSI offset for the two LNA gain
settings in use, available from the device data sheet. Note that the hardware RSSI register is not updated
this way.
For the AGC algorithm to operate correctly, it requires some signal, having the same power as the packet,
transmitted in the band in front of the packet. That signal can be extra preamble bytes or tone. The length
required for this signal depends on the RSSI accuracy setting in MDMTEST0.RSSI_ACC, see Section 25.7.
An average of n windows of length t RSSI requires the extra signal to last at least (n + 1) tRSSI. Extra
preamble bytes can be set up using MDMCTRL2.NUM_PREAM_BYTES. Note that the extra signal required
comes in addition to the 1 preamble byte always used in a packet. When adding extra preamble bytes,
this must be accounted for in PRF_TX_DELAY, PRF_RETRANS_DELAY and PRF_RX_TX_TIME. The Rx
requires n × tRSSI extra time to start when using the AGC. In the dc offset estimation, the extra signal must
be accounted for when setting the delay.
25.9.2.2 Tone in Front of Packet
In order to get the transmission format to resemble that of other vendors, a tone may be sent in front of
the preamble. This tone can be used by the AGC algorithm on the receiver side. If
PRF_PKT_CONF.START_TONE is 1, such a tone is transmitted as a replacement of the first preamble
bytes. This means that this feature must only be used in combination with increasing the number of
preamble bytes. The tone lasts for a time given by the RAM register PRFX_TONE_DURATION. In order to
get a smooth transition from tone to preamble, it is recommended to set PRFX_TONE_DURATION as given
in Table 25-15.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
305
Link Layer Engine
www.ti.com
Tone transmission is allowed to coincide with the synthesizer stabilizing (this may, however, cause the
start of the tone to have larger frequency variations than the packet). For this reason, when
PRF_PKT_CONF.START_TONE is 1, the synthesizer startup time is reduced by the value of the register
PRFX_TONE_OFFSET. This should normally correspond to the time of the extra preambles, but it must not
be larger than 4096 (corresponding to 128 µs). PRFX_TONE_OFFSET can thus be used to compensate for
the extra time added by the extra preamble bytes used for tone generation. However, the duration of the
extra preamble bytes configured must be accounted for in PRF_TX_DELAY, PRF_RETRANS_DELAY, and
PRF_RX_TX_TIME.
The default values of PRFX_TONE_DURATION and PRFX_TONE_OFFSET correspond to 48 µs and are
tuned for using 12 extra preamble bytes (13 in total) on 2 Mbps. When using the reset values,
MDMCTRL2.NUM_PREAM_BYTES should thus be set to 0x0C.
If PRFX_TONE_DURATION is set too large compared to the number of preamble bytes configured, the
modulator underflows. If this happens, the task ends with TASKERR_MODUNF as end cause.
Table 25-15. Recommended RAM Register Settings for Start Tone
Data Rate
2 Mbps
1 Mbps
500 kbps
250 kbps
PRFX_TONE_DURATION
MDMCTRL2.NUM_PREAM_BYTES × 0x80 +
0x4A
PRFX_TONE_OFFSET
MDMCTRL2.NUM_PREAM_BYTES × 0x80
MDMCTRL2.NUM_PREAM_BYTES × 0x100 + MDMCTRL2.NUM_PREAM_BYTES × 0x100
0x52
MDMCTRL2.NUM_PREAM_BYTES × 0x200 +
0x62
MDMCTRL2.NUM_PREAM_BYTES × 0x400 +
0x82
min(MDMCTRL2.NUM_PREAM_BYTES × 0x200,
0x1000)
min(MDMCTRL2.NUM_PREAM_BYTES × 0x400,
0x1000)
25.9.2.3 Receive Task
When a CMD_RX command is received, the LLE configures the radio on the channel given by
PRF_CHAN.FREQ and starts listening for a sync word.
The LLE can set up an internal time-out for the sync search in the PRF_SEARCH_TIME register. If this
register is non-zero and no sync has been obtained in the number of 32 MHz cycles given by this register,
the task ends with a TASK_NOSYNC end cause. Note that the value of this register must be large enough
to have time for the duration of the sync word and one preamble byte, in addition to some margin, in order
to get sync. The task can also be set up to end on Timer 2 event 2, based on
PRF_TASK_CONF.STOP_CONF. If this bit field is 11, the Timer 2 event 2 time-out applies only during the
first sync search after a CMD_RX command has been issued if PRF_TASK_CONF.START_CONF is 0. In
this case, the time-out in PRF_SEARCH_TIME does not apply to the first sync search, but it still applies to
subsequent sync searches in the same task. If PRF_TASK_CONF.STOP_CONF is 11 and
PRF_TASK_CONF.START_CONF is 1, the time-out applies to every sync search and PRF_SEARCH_TIME
never applies, but the Timer 2 event 2 timeout does not apply after sync is obtained or while waiting for
Timer 2 event 1 to restart listening. If sync is obtained, the LLE starts reading the packet.
If sync is found on a packet, the time of sync is captured by the Timer 2 capture function (see
Section 22.1.10).
25.9.2.3.1 Basic Mode
This section describes the receive operation if PRF_TASK_CONF.MODE is 00 or 01.
If PRF_TASK_CONF.MODE is 01, the length byte is read first. It gives the number of bytes between the
length byte and the CRC, including the address. If the length is too small to contain the address, the
reception of the packet is stopped and the device goes back to sync search (regardless of the setting in
PRF_TASK_CONF.REPEAT).
306
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Link Layer Engine
www.ti.com
Next, the address byte is read if PRF_PKT_CONF.ADDR_LEN is 1. It is compared against
PRF_ADDR_ENTRYn.ADDRESS for the values of n where the entry is enabled for the received sync word.
If there is a matching entry, this entry is used when receiving the packet; otherwise, reception is stopped
and the device goes back to sync search. If PRF_PKT_CONF.ADDR_LEN is 0, the first entry that is enabled
for the received sync word is used. If PRF_TASK_CONF.MODE is 00, the packet length is then read from
PRF_ADDR_ENTRYn.RXLENGTH. This length includes the address, so it must be greater than or equal to
the number of address bytes. If PRF_TASK_CONF.MODE is 01, the received length byte is compared
against PRF_ADDR_ENTRYn.RXLENGTH. If it is greater than that value, reception is stopped and the
device goes back to sync search.
If reception is stopped due to address mismatch or invalid length, the time-out given by
PRF_SEARCH_TIME or Timer 2 event 2 still applies. If the first packet of the task is being received and
PRF_TASK_CONF.STOP_CONF is 11, the next packet still counts as the first packet.
If a CRC field is present, it is checked using the polynomial configured in the BSP and the initialization
value from PRF_CRC_INIT. The result of the CRC check is written in the MSB of the RES byte in the
status field if a status field is configured. If the CRC is not correct and PRF_FIFO_CONF.AUTOFLUSH_CRC
is 1, the LLE sends a discard Rx FIFO command to remove the packet from the Rx FIFO.
A packet where the length is equal to the address size contains no payload. Such a packet is known as an
empty packet. If PRF_FIFO_CONF.AUTOFLUSH_EMPTY is 1 and an empty packet is received, the LLE
sends a discard Rx FIFO command to remove the empty packet from the Rx FIFO.
Note that if the CRC length is 1 byte, the CRC check is not correct for empty packets if fixed length is
configured or no address bytes are used.
If the Rx FIFO becomes full while receiving a packet, the packet is discarded from the FIFO and no more
bytes are stored in the Rx FIFO, but the packet is received to its end. After that, it is checked whether the
packet would be discarded from the Rx FIFO anyway due to the setting of PRF_FIFO_CONF. If so, the
task proceeds as normally. Otherwise, an RXFIFOFULL error interrupt is raised in lieu of the normal
interrupt for received packets.
After receiving a packet, the LLE raises an interrupt to the MCU. Depending on CRC result and whether
the packet was empty, the interrupts are generated as shown in Table 25-16, provided an RXFIFOFULL
interrupt is not raised as described previously. The table also shows which of the counters among the
RAM registers are to be updated.
Table 25-16. Interrupt and Counter Operation for Received Messages
CRC Result
Payload Length > Address
Length
Counter Incremented
Interrupt Raised
OK
No
PRF_ADDR_ENTRYn.N_RXOK
RXEMPTY
OK
Yes
PRF_ADDR_ENTRYn.N_RXOK
RXOK
NOK
No
PRF_ADDR_ENTRYn.N_RXNOK
RXNOK
NOK
Yes
PRF_ADDR_ENTRYn.N_RXNOK
RXNOK
An address entry should not be modified while the receiver is running. In order to modify, stop the
receiver, modify the entry or entries, and restart the receiver.
25.9.2.3.2 Auto Mode
This section describes the receive operation if PRF_TASK_CONF.MODE is 10 or 11.
If PRF_PKT_CONF.ADDR_LEN is 1, the address byte is compared against PRF_ADDR_ENTRYn.ADDRESS,
where n ranges from 0 to 7. It is compared against PRF_ADDR_ENTRYn.ADDRESS for the values of n
where the entry is enabled for the received sync word. If there is a matching entry, this entry is used when
receiving the packet, otherwise reception is stopped and the device goes back to sync search. If
PRF_PKT_CONF.ADDR_LEN is 0, the first entry that is enabled for the received sync word is used.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
307
Link Layer Engine
www.ti.com
Next, the 9-bit or 10-bit header is read. If PRF_ADDR_ENTRYn.CONF.VARLEN is 1, the length is fetched
from the header and compared against PRF_ADDR_ENRYn.RXLENGTH. If it is greater than that value,
reception is stopped and the device goes back to sync search. If PRF_ADDR_ENTRYn.CONF.VARLEN is 0,
the length field in the received header is ignored and the packet length is read from
PRF_ADDR_ENTRYn.RXLENGTH. In both cases, the length is the number of bytes after the header and
before the CRC. The length must be less than or equal to 63 for a 9-bit header and 127 for a 10-bit
header. When a 10-bit header is used, the MCU must ensure that an entire packet can fit in the Rx FIFO
for auto ACK to be possible. This limits the maximum packet size based on the settings in
PRF_FIFO_CONF.
If reception is stopped due to address mismatch or invalid length, the time-out given by
PRF_SEARCH_TIME or Timer 2 event 2 still applies. If the first packet of the task is being received and
PRF_TASK_CONF.STOP_CONF is 11, this still counts as the first packet.
If a CRC field is present, it is checked using the polynomial configured in the BSP and the initialization
value from PRF_CRC_INIT. The result of the CRC is written in the MSB of the RES byte in the status field
if a status field is configured. If the CRC is not correct and PRF_FIFO_CONF.AUTOFLUSH_CRC is 1, the
LLE sends a discard Rx FIFO command to remove the packet from the Rx FIFO.
If the CRC is correct, the sequence number is checked against the sequence number stored in
PRF_ADDR_ENTRYn.SEQSTAT.SEQ. If the sequence numbers are equal and
PRF_ADDR_ENTRYn.SEQSTAT.VALID is 1, the two last received CRC bytes are compared against the 2
bytes in PRF_ADDR_ENTRYn.LASTCRC. If they are equal, the packet is determined to be a retransmission
which can be ignored. If the CRC is 1 byte only, the received CRC byte is compared to
PRF_ADDR_ENTRYn.LASTCRC[0] only, and if there is no CRC, the comparison is always viewed as
equal. If the packet was a retransmission, the IGN bit of the RES byte in the status field is set if a status
field is configured. After reception of a packet with CRC OK and which fit in the Rx FIFO,
PRF_ADDR_ENTRYn.SEQSTAT.VALID is set to 1, PRF_ADDR_ENTRYn.SEQSTAT.SEQ is set to the
sequence number of the header of the received packet, and PRF_ADDR_ENTRYn.LASTCRC is set to the
value of the last two received CRC bytes.
If the Rx FIFO becomes full while receiving a packet, the packet is discarded from the FIFO and no more
bytes are stored in the Rx FIFO, but the packet is received to its end. After that, it is checked whether the
packet would be discarded from the Rx FIFO anyway due to the setting of PRF_FIFO_CONF. If so, the
task proceeds as normally. Otherwise, an RXFIFOFULL error interrupt is raised, and no acknowledgment
is transmitted. The sequence number is not updated so that a retransmission of the packet is not ignored.
If the received packet was not a retransmission and PRF_ADDR_ENTRYn.SEQSTAT.ACK_PAYLOAD_SENT
is 1, the packet is seen as a confirmation of the last transmitted acknowledgment payload. If so,
PRF_ADDR_ENTRYn.SEQSTAT.ACK_PAYLOAD_SENT is set to 0, a TXDONE interrupt is raised, and the
PRF_ADDR_ENTRYn.NTXDONE counter is incremented. PRF_ADDR_ENTRYn.ACKLENGTHk is set to 0 for
the k found in PRF_ADDR_ENTRYn.SEQSTAT.NEXTACK, and PRF_ADDR_ENTRYn.SEQSTAT.NEXTACK is
inverted.
After receiving a packet, the LLE raises an interrupt to the MCU. Depending on the CRC result, the
payload length, and whether the received packet is a retransmission to be ignored, the interrupts are
generated as shown in Table 25-17. The table also shows which of the counters among the RAM registers
are to be updated.
Table 25-17. Interrupt and Counter Operation for Received Messages
308
CRC Result
Ignore
Length
OK
No
>0
PRF_ADDR_ENTRYn.N_RXOK
RXOK
OK
No
=0
PRF_ADDR_ENTRYn.N_RXOK
RXEMPTY
OK
Yes
X
PRF_ADDR_ENTRYn.N_RXIGNORED
RXIGNORED
NOK
X
X
PRF_ADDR_ENTRYn.N_RXNOK
RXNOK
CC2541 Proprietary Mode Radio
Counter Incremented
Interrupt Raised
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Link Layer Engine
www.ti.com
After reception of a packet, the next action is determined as follows:
• If the CRC of the received packet was not correct, the treatment of the packet is finished and the next
action is as described in Section 25.9.2.3.3.
• If PRF_ADDR_ENTRYn.CONF.AA is 0, the treatment of the packet is finished and the next action is as
described in Section 25.9.2.3.3.
• If the NO_ACK bit of the received header is 1 and the CRC was correct, the treatment of the packet is
finished and the next action is as described in Section 25.9.2.3.3.
• If the packet did not fit in the Rx FIFO and was not otherwise to be discarded, the treatment of the
packet is finished and the next action is as described in Section 25.9.2.3.3.
• Otherwise, an acknowledgment is transmitted as described in the following text.
After receiving a packet where the CRC is correct and where an acknowledgment is supposed to be sent,
the transmitter is configured. The transmission starts at a time given by the PRF_RX_TX_TIME register.
Synthesizer recalibration is performed only if there is time. The LLE checks
PRF_ADDR_ENTRYn.SEQSTAT.NEXTACK to find k. If PRF_ADDR_ENTRYn.ACKLENGTHk is nonzero,
payload is included the packet. In this case, PRF_ADDR_ENTRYn.SEQSTAT.ACK_PAYLOAD_SENT is set
to 1 by the LLE. The transmitted packet has the same sync word and address as the received packet. If
PRF_ADDR_ENTRYn.CONF.TXLEN is 0, the length field in the header is set equal to
PRF_ADDR_ENTRYn.ACKLENGTHk. If PRF_ADDR_ENTRYn.CONF.TXLEN is 1, the length field is set to 11
0011 for a 9-bit header and to 011 0011 for a 10-bit header. Note that a value of 0 for
PRF_ADDR_ENTRYn.CONF.TXLEN may be used regardless of the VARLEN setting in the peer device, as
the length field is ignored for fixed length. A value of 1 must only be used if the peer is configured to use
fixed length for the ACK payload, and should only be used with ACKs without payload. The sequence
number is set to the value of PRF_ADDR_ENTRYn.SEQSTAT.ACKSEQ, and NO_ACK is set to 0. If there is
payload, it is read from the buffer as described in Section 25.8.3.
After the acknowledgement has been transmitted, PRF_ADDR_ENTRYn.SEQSTAT.ACKSEQ is incremented
modulo 4, the PRF_N_TX counter is incremented, and the next action is as described in
Section 25.9.2.3.3.
25.9.2.3.3 Continuation and Ending of Receive Tasks
When a task ends, a TASKDONE interrupt is raised and an end cause is then available in PRF_ENDCAUSE.
After a packet has been received and potentially an acknowledgment has been transmitted, the next
action depends on PRF_TASK_CONF.REPEAT. If this value is 0, the task ends. In this case, the
PRF_ENDCAUSE register is set to TASK_ENDOK.
If PRF_TASK_CONF.REPEAT is 1, reception restarts. If PRF_TASK_CONF.START_CONF is 1, the LLE
behaves as if the task was started again, with the LLE waiting for Timer 2 event 1; then starting to listen. If
PRF_TASK_CONF.START_CONF is 0, the receiver restarts as soon as possible, as starting a new task
(except for the behavior of Timer 2 event 2 if PRF_TASK_CONF.STOP_CONF is 11). In both cases,
synthesizer recalibration is done if PRF_TASK_CONF.REPEAT_CONF is 0, otherwise not. Skipping
synthesizer recalibration reduces the time before listening is restarted.
If a CMD_SHUTDOWN or a command starting a new task is observed while the task is running, it ends
immediately with TASK_ABORT as the end cause. If the receiver or transmitter was running, an RXTXABO
interrupt is also raised.
If CMD_STOP is received while in sync search, the task ends immediately with TASK_RXTIMEOUT as
the end cause. If CMD_STOP is received while receiving or while transmitting an ACK or in the transition
between those, the task ends with TASK_STOP as the end cause after the packet is fully received and (if
ACK is to be sent) the ACK is sent. If CMD_STOP is received while waiting for Timer 2 event 1 to restart
reception, the task ends immediately with TASK_STOP as the end cause
If Timer 2 event 2 (either from Timer 2 or from CMD_SEND_EVENT2) is observed during the task, the
behavior depends on PRF_TASK_CONF.STOP_CONF:
•
•
•
00: Nothing happens
01: Behaves as if a CMD_STOP was received
10: Behaves as if a CMD_SHUTDOWN was received
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
309
Link Layer Engine
•
www.ti.com
11: If received while in sync search for the first packet after the task was started, or if
PRF_TASK_CONF.START_CONF is 1 while in sync search for any packet, the task ends immediately
with TASK_RXTIMEOUT as the end cause. Otherwise, nothing happens.
In addition, the task can end due to an internal time-out as described in the beginning of Section 25.9.2.1,
or it can end due to an error condition. The full list of possible end causes is summarized in Table 25-18.
Table 25-18. End-of-Receive Tasks
Condition
End-of-Task Cause
Received packet (and potentially sent ACK) with
PRF_TASK_CONF.REPEAT = 0
TASK_ENDOK
Received packet (and potentially sent ACK) with
PRF_TASK_CONF.REPEAT = 1 after having
observed CMD_STOP or Timer 2 event 2 with
PRF_TASK_CONF.STOP_CONF = 01 and
PRF_TASK_CONF.REPEAT = 1
TASK_STOP
Comment
While in sync search, observed CMD_STOP or Timer 2
TASK_RXTIMEOUT
event 2 with PRF_TASK_CONF.STOP_CONF = 01
Observed Timer 2 event 2 while in sync search of the
first packet with PRF_TASK_CONF.STOP_CONF
= 11
Did not get sync in the time specified by
PRF_SEARCH_TIME
TASK_RXTIMEOUT
TASK_NOSYNC
Received command for starting new task or
CMD_SHUTDOWN or observed Timer 2 event 2 with
PRF_TASK_CONF.STOP_CONF = 10
TASK_ABORT
If transmitter was running or receiver was
running and had obtained sync, an
RXTXABO interrupt is also raised.
Received unknown command
TASKERR_CMD
Semaphore is not free when expected
TASKERR_SEM
Task ends without any radio operation.
LLEERR interrupt is also raised.
Unpermitted value of RAM register
TASKERR_PAR
LLEERR interrupt is also raised.
TASKERR_PAR
LLEERR interrupt is also raised.
TASKERR_PAR
LLEERR interrupt is also raised.
LLEERR interrupt is also raised. If
transmitter was running or receiver was
running and had obtained sync, an
RXTXABO interrupt is also raised.
For PRF_TASK_CONF.MODE = 00:
PRF_ADDR_ENTRYn.RXLENGTH corresponding
to the received address is smaller than address length
For auto mode:
PRF_ADDR_ENTRYn.ACKLENGTGHm for the
ACK payload to be transmitted exceeded 32
25.9.2.4 Transmit Task
When a CMD_TX command is received, the LLE configures the radio on the channel given by
PRF_CHAN.FREQ and starts transmitting the packet from the Tx FIFO.
If the Tx FIFO has no available data, the task ends with TASKERR_TXFIFO as the end cause. Otherwise,
the number of bytes given by the length byte in the Tx FIFO is read from the Tx FIFO and transmitted or
otherwise handled as described in following sections. No check of data availability is done after the length
byte is read, so if the FIFO contains fewer bytes than indicated in the length field, a Tx FIFO underflow
interrupt is raised by the FIFO hardware.
25.9.2.4.1 Basic Mode
This section describes the transmit operation if PRF_TASK_CONF.MODE is 00 or 01.
If PRF_TASK_CONF.MODE is 01, the length field is calculated from the length field in the FIFO and
transmitted. It is up to the MCU to ensure that the calculated length field does not exceed 255. If
PRF_TASK_CONF.MODE is 00, no length field is transmitted.
310
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Link Layer Engine
www.ti.com
If an address is configured, it is found based on the setting in PRF_FIFO_CONF.TX_ADDR_CONF. It can
be set to take the address from PRF_ADDR_ENTRY0, to read it from the Tx FIFO (which for the transmitter
is equivalent to not having an address configured), or to read an index n from the config byte in the FIFO
and read the address from PRF_ADDR_ENTRYn.ADDRESS. The values of ENA0 and ENA1 in
PRF_ADDR_ENTRYn.CONF are ignored for the transmitter; the primary sync word is always transmitted.
The payload (if any) is transmitted as given in the FIFO.
If configured, a CRC with the number of bytes given by PRF_CRC_LEN is transmitted at the end.
When a packet has been transmitted, the LLE sends a deallocate Tx FIFO command if
PRF_ADDR_ENTRYn.REUSE is 0. Otherwise, the MCU must issue either a deallocate Tx FIFO (to send a
new packet) or a retry Tx FIFO (to reuse) before sending again. The PRF_N_TX counter is incremented. A
TXDONE interrupt to the MCU is raised when the packet has been completely read out of the Tx FIFO by
the LLE. Note that due to modulator delay, CRC transmission and ramp-down, this will happen before the
packet transmission is finished. The next action is as given in Section 25.9.2.4.3.
25.9.2.4.2 Auto Mode
This section describes the transmit operation if PRF_TASK_CONF.MODE is 10 or 11.
When a 10-bit header is used, the MCU must ensure that an entire packet can fit in the Tx FIFO for auto
retransmission to be possible. This limits the maximum packet size based on the settings in
PRF_FIFO_CONF.
If an address is configured, it is the first byte transmitted. It is found based on the setting in
PRF_FIFO_CONF.TX_ADDR_CONF. It can be set to take the address from PRF_ADDR_ENTRY0, to read it
from the Tx FIFO, or to read an index n from the config byte in the FIFO and read the address from
PRF_ADDR_ENTRYn.ADDRESS. In other cases, n is always be assumed to be 0 in the following text. The
values of ENA0 and ENA1 in PRF_ADDR_ENTRYn.CONF are ignored for the transmitter; the primary sync
word is always transmitted.
The 9-bit or 10-bit header is transmitted next. If PRF_ADDR_ENTRYn.CONF.TXLEN is 0, the length field is
set to the number of payload bytes after the header, which is calculated from the length byte in the Tx
FIFO. If PRF_ADDR_ENTRYn.CONF.TXLEN is 1, the length field is set to 11 0011 for a 9-bit header and to
011 0011 for a 10-bit header. Note that a value of 0 for PRF_ADDR_ENTRYn.CONF.TXLEN may be used
regardless of the VARLEN setting in the receiver, as a receiver configured to use fixed length ignores the
length field. A value of 1 must only be used if the receiver is configured to use fixed length. The NO_ACK
bit transmitted is set according to bit 5 of the config byte read from the Tx FIFO if present, otherwise to 0.
If PRF_ADDR_ENTRYn.CONF.FIXEDSEQ is 1, the SEQ bits transmitted are set equal to bits 6 and 7 of the
config byte read from the FIFO. Otherwise, the SEQ bits are set to the value of
PRF_ADDR_ENTRYn.SEQSTAT.SEQ.
The payload (if any) is transmitted as given in the FIFO.
If configured, a CRC with the number of bytes given by PRF_CRC_LEN is transmitted at the end.
When a packet has been transmitted, the N_TX counter is incremented.
After transmission of a packet, the action depends on PRF_ADDR_ENTRYn.CONF.AA and the NO_ACK bit
in the transmitted header. If PRF_ADDR_ENTRYn.CONF.AA = 0 or NO_ACK = 1, no acknowledgment is
expected, and the action is as if a valid acknowledgment had been received.
If PRF_ADDR_ENTRYn.CONF.AA is 1 and the transmitted NO_ACK bit was 0, the LLE configures Rx to
listen for an acknowledgment. To listen for acknowledgment, the receiver is configured at a time given by
the PRF_TX_RX_TIME register. Synthesizer recalibration is performed only if there is time. The unit looks
for sync. The sync search times out at the time given by PRF_SEARCH_TIME. If sync is found, the packet
is received into the Rx FIFO. If PRF_PKT_CONF.ADDR_LEN is 1, the address byte is compared against
PRF_ADDR_ENTRYn.ADDRESS for the n that was used in transmission. If not matching, reception is
stopped.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
311
Link Layer Engine
www.ti.com
Next, the 9-bit or 10-bit header is read. If PRF_ADDR_ENTRYn.CONF.VARLEN is 1, the length is fetched
from the header and compared against PRF_ADDR_ENTRYn.RXLENGTH. The maximum allowed value of
this register is 32. If the received length is greater than PRF_ADDR_ENTRYn.RXLENGTH, reception is
stopped and the device goes back to sync search. If PRF_ADDR_ENTRYn.CONF.VARLEN is 0, the length
field in the received header is ignored and the packet length is read from PRF_ADDR_ENTRYn.RXLENGTH,
which should normally be 0 in this case. The length is the number of bytes after the header and before the
CRC.
If a CRC field is present, it is checked using the polynomial configured in the BSP and the initialization
value from PRF_CRC_INIT. The result of the CRC is written in the MSB of the RES byte in the status field
if a status field is configured. If the CRC is not correct and PRF_FIFO_CONF.AUTOFLUSH_CRC is set, the
LLE sends a discard Rx FIFO command to remove the packet from the Rx FIFO.
The received sequence number is written to the config byte of the Rx FIFO if configured, but is otherwise
ignored.
If the Rx FIFO goes full while receiving an acknowledgment packet, the packet is discarded from the FIFO
and no more bytes are stored in the Rx FIFO, but the packet is received to its end. After that, it is checked
to see whether the packet would be discarded from the Rx FIFO anyway due to the setting of
PRF_FIFO_CONF. If so, the task proceeds as normally. Otherwise, the task ends after the packet is
received and an RXFIFOFULL error interrupt is raised. In this case, the treatment of the packet is as if the
acknowledgment were not successfully received. This means that the next time a transmit task is started,
the packet is retransmitted so that the receiver retransmits the ACK payload.
After receiving an acknowledgment, the LLE raises an interrupt to the MCU. Depending on the CRC
result, the payload length, and whether the received packet had the same sequence number as the
transmitted one, the interrupts are generated as shown in Table 25-19. It also shows which of the counters
among the RAM registers that are to be updated.
Table 25-19. Interrupt and Counter Operation for Received ACK Packets
CRC Result
Length
Counter Incremented
Interrupt Raised
OK
>0
PRF_ADDR_ENTRYn.N_RXOK
OK
=0
PRF_ADDR_ENTRYn.N_RXIGNORED
NOK
X
RXOK
PRF_ADDR_ENTRYn.N_RXNOK
RXEMPTY
RXNOK
If an acknowledgment was not received (because no sync was obtained in time, the address did not
match, the sequence number was wrong, the CRC check failed, or the ACK did not fit in the Rx FIFO and
was not otherwise to be discarded) the LLE sends a retry Tx FIFO command. If the number of
retransmissions already performed (not including the original transmission) is equal to
PRF_RETRANS_CNT, the task ends. Otherwise, the packet is retransmitted. The time from the end of the
previous transmission to the start of the retransmission is given in units of 62.5 ns by
PRF_RETRANS_DELAY.
If the received packet was a valid acknowledgment, or if a packet was completely read out of the Tx FIFO
and no acknowledgment was expected, the LLE sends a deallocate Tx FIFO command if
PRF_ADDR_ENTRYn.REUSE is 0. Otherwise, the MCU must issue either a deallocate Tx FIFO (to send a
new packet) or a retry Tx FIFO (to reuse) before sending again. The PRF_ADDR_ENTRYn.NTXDONE
counter is incremented. A TXDONE interrupt to the MCU is raised. If
PRF_ADDR_ENTRYn.CONF.FIXEDSEQ = 0, PRF_ADDR_ENTRYn.SEQSTAT.SEQ is incremented by 1
modulo 4. The next action is as given in Section 25.9.2.3.3.
If the task ends because of a maximum number of retransmissions, a retry Tx FIFO command is sent
before the task ends, and PRF_ADDR_ENTRYn.SEQSTAT.SEQ is not incremented. This means that by
default, the packet retransmission is attempted in the next task. If this is not desired, the packet must be
removed from the FIFO. This can be done either by issuing a CMD_TXFIFO_RESET (this also removes
any subsequent packets in the Tx FIFO), by reading out the packet using the RFTXFRD register and
issuing a CMD_TX_FIFO_DEALLOC command, or by Tx FIFO pointer manipulation (Section 25.3.1.3).
PRF_ADDR_ENTRYn.SEQSTAT.SEQ should then be incremented by one. These operations should only
take place between tasks (i.e., while the LLE does not have SEMAPHORE1).
312
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Link Layer Engine
www.ti.com
25.9.2.4.3 Continuation and Ending of Transmit Tasks
When a task ends, a TASKDONE interrupt is raised and an end cause is then available in PRF_ENDCAUSE.
After a packet has been transmitted and potentially a valid acknowledgment has been received, the next
action depends on PRF_TASK_CONF.REPEAT. If this value is 0, the task ends. In this case, the
PRF_ENDCAUSE register is set to TASK_ENDOK.
If PRF_TASK_CONF.REPEAT is 1, the Tx FIFO status is checked. If the Tx FIFO has no available data, the
task ends with TASK_ENDOK as the end cause. Otherwise, transmission restarts. If
PRF_TASK_CONF.START_CONF is 1, it behaves as if the task was started again with the LLE waiting for
Timer 2 event 1, then performing a synthesizer calibration and starting to transmit. If
PRF_TASK_CONF.START_CONF is 0, the transmitter restarts PRF_TX_DELAY after the end of the
previously transmitted packet, with synthesizer recalibration only if there is enough time, but in other
respects as starting a new task. The PRF_TX_DELAY register gives the wait time in units of 62.5 ns. If the
value is too small to fulfill, the transmission starts as soon as possible.
If a CMD_SHUTDOWN or a command starting a new task is observed while the task is running, it ends
immediately with TASK_ABORT as the end cause. If the transmitter or receiver was running, an RXTXABO
interrupt is also raised.
If CMD_STOP is received while transmitting a packet, waiting for transmission of another packet, waiting
for an ACK, receiving an ACK, or in the transition between those, the task ends with TASK_STOP as the
end cause after the packet is fully transmitted and (if ACK is expected) the ACK is received or given up. If
CMD_STOP is received while waiting for Timer 2 event 1 to restart reception, the task ends immediately
with TASK_STOP as the end cause.
If Timer 2 event 2 (either from Timer 2 or from CMD_SEND_EVENT2) is observed during the task, the
behavior depends on PRF_TASK_CONF.STOP_CONF.
•
•
•
•
00:
01:
10:
11:
Nothing happens.
Behaves as if a CMD_STOP was received
Behaves as if a CMD_SHUTDOWN was received
Nothing happens.
In addition, the task can end for reasons described earlier, or it can end due to an error condition. The full
list of possible end causes is summarized in Table 25-20.
Table 25-20. End-of-Transmit Tasks
Condition
End-of-Task Cause
Transmitted packet (and potentially received ACK)
with PRF_TASK_CONF.REPEAT = 0
TASK_ENDOK
Transmitted packet (and potentially received ACK)
and observed Tx FIFO with no available data
TASK_ENDOK
Comment
Transmitted packet (and potentially received ACK)
after having observed CMD_STOP or Timer 2 event 2 TASK_STOP
with PRF_TASK_CONF.STOP_CONF = 01
Did not get valid acknowledgment after having
retransmitted the number of times given by
TASK_MAXRT
PRF_RETRANS_CNT
LLEERR and RXTXABO interrupts are
Observed empty Tx FIFO when packet transmission
is supposed to start, or Tx FIFO is in an invalid state
TASKERR_TXFIFO
Rx FIFO went overfull while receiving an ACK that
was not otherwise to be discarded
TASKERR_RXFIFO
RXFIFOFULL interrupt is also raised.
Received command for starting new task or
CMD_SHUTDOWN or observed Timer 2 event 2 with
PRF_TASK_CONF.STOP_CONF = 10
TASK_ABORT
If transmitter was running, an RXTXABO
interrupt is also raised.
also raised.
LLEERR interrupt is also raised. If
Received unknown command
TASKERR_CMD
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
transmitter or receiver was running and
had obtained sync, an RXTXABO
interrupt is also raised.
CC2541 Proprietary Mode Radio 313
Link Layer Engine
www.ti.com
Table 25-20. End-of-Transmit Tasks (continued)
Condition
End-of-Task Cause
Comment
Semaphore is not free when expected
TASKERR_SEM
Task ends without any radio operation.
LLEERR interrupt is also raised.
Invalid value of RAM register
TASKERR_PAR
LLEERR interrupt is also raised.
Length field to be transmitted exceeded maximum
allowed value (255 in basic mode, 127 or 63 in auto
mode with variable length)
TASKERR_PAR
LLEERR interrupt is also raised.
25.9.2.5 Transmit on Clear-Channel Task
When a CMD_TX_ON_CC command is received, the LLE configures the receiver on the channel given by
PRF_CHAN.FREQ, but sync search is not enabled. The LLE polls the RSSI register every 5.33 µs and
compares it to the value of PRF_RSSI_LIMIT. If a valid RSSI below the value of PRF_RSSI_LIMIT is
observed more than PRF_RSSI_COUNT times in a row, the system starts transmitting. From there, the
operation is as a normal transmit task, see Section 25.9.2.4, except for the operation after a packet has
been transmitted and potentially acknowledged, which is described in Section 25.9.2.5.1.
25.9.2.5.1 Continuation and Ending of Transmit on Clear-Channel Tasks
If PRF_TASK_CONF.TX_ON_CC_CONF is 1, the task ends if a valid RSSI value is not below the limit. If
PRF_TASK_CONF.TX_ON_CC_CONF is 0, the device keeps listening until an RSSI below the given value
is found.
If PRF_TASK_CONF.STOP_CONF is 11, Timer 2 event 2 may give a time-out while listening for a clear
channel the first time, but not after the first transmission has been started.
If PRF_TASK_CONF.TX_ON_CC_CONF is 0, the clear-channel assessment must not be used as the only
medium access control scheme in a multiuser environment, as this may cause all the devices to start
transmission at the same time.
If retransmission is enabled, the LLE listens for acknowledgment and retransmits if needed as for normal
Tx tasks. However, if PRF_TASK_CONF.REPEAT_CONF is 0 after applying the retransmit delay, the device
returns to listening, performing the same operation as when the task started, before possibly
retransmitting.
If PRF_TASK_CONF.REPEAT is 0, the task ends after transmission as for normal Tx tasks.
If PRF_TASK_CONF.REPEAT is 1, the Tx FIFO status is checked. If the Tx FIFO has no available data, the
task ends with TASK_ENDOK as the end cause. Otherwise, transmission restarts. If
PRF_TASK_CONF.REPEAT_CONF is 0, the task returns to listening, whereas if it is 1, the task restarts as if
it were a standard transmit task. If PRF_TASK_CONF.START_CONF is 1, it behaves as if the task was
started again with the LLE waiting for Timer 2 event 1, then performing a synthesizer calibration and
starting to listen or transmit. If PRF_TASK_CONF.START_CONF is 0, the listening or transmission restarts
PRF_TX_DELAY after the end of the previously transmitted packet, with synthesizer recalibration only if
there is enough time, but in other respects as starting a new task.
If Timer 2 event 2 (either from Timer 2 or from CMD_SEND_EVENT2) is observed during the task, the
behavior depends on PRF_TASK_CONF.STOP_CONF.
•
•
•
•
314
00: Nothing happens.
01: If received while transmitting a packet or waiting for or receiving an ACK or in the transition
between those, the task ends with TASK_STOP as the end cause after the packet is fully transmitted
and (if ACK is expected) the ACK is received or given up. If received while waiting for Timer 2 event 1
to restart reception, the task ends immediately with TASK_STOP as the end cause.
10: Behaves as if a CMD_SHUTDOWN was received
11: If received while listening for RSSI below the level before the first packet after the task was started,
or if PRF_TASK_CONF.START_CONF is 1 while listening before any packet, the task ends immediately
with TASK_NOCC as the end cause. Otherwise, nothing happens.
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Link Layer Engine
www.ti.com
The task can end for all the same reasons as a normal transmit task summarized in Table 25-20. In
addition it can end for the reasons listed in Table 25-21.
Table 25-21. Additional Reasons for End-of-Transmit on Clear-Channel Tasks
Condition
End-of-Task Cause
Observed valid RSSI above the threshold with
PRF_TASK_CONF.TX_ON_CC_CONF = 1
TASK_NOCC
Observed CMD_STOP or Timer 2 event 2 with
PRF_TASK_CONF.STOP_CONF = 01 or 11 while listening for
RSSI below the threshold
TASK_NOCC
Comment
25.9.2.6 Timing
The timing in tasks is given the registers PRF_TX_DELAY, PRF_RETRANS_DELAY, PRF_SEARCH_TIME,
PRF_RX_TX_TIME, and PRF_TX_RX_TIME. The first two of these registers are multiplied by 2 and then
represent the number of 32-MHz samples, while the rest directly give the number of 32-MHz samples.
PRF_TX_DELAY gives the time from the end of the previous transmission in the task to the start of the
next transmission. Some examples of these delays are shown in Figure 25-12 and Figure 25-13 for Rx
and Tx tasks, respectively. The time from the end of a received packet to the beginning of a transmitted
packet is 130 µs in an Rx task with auto ACK.
When sync search takes place, either for receiving a normal packet or for receiving ACK, a time-out can
be set up for when to give up the search. This time-out, given in 32-MHz cycles, is set up in the
PRF_SEARCH_TIME register. Setting this register to 0 disables the time-out. In case of a time-out, the task
ends for a normal sync search, or a packet is retransmitted in case of an ACK sync search.
Receive task, PRF_TASK_CONF.MODE = 0X, .START_CONF = 0, .REPEAT = 1:
tSynth
tRx
Sync Search
tSearch
tSynth
tSynth
tRx
Receive Packet
tRx
Sync Search
Receive Packet
Sync Search
CMD_RX
Give up sync search,
end task
Receive task, PRF_TASK_CONF.MODE = 1X, .START_CONF = 0, .REPEAT = 1:
tSynth
tRx-Tx
tRx
Sync Search
tSynth
tTx
Receive Packet
tRx
Transmit ACK
Sync Search
Receive Packet
CMD_RX
Receive task, PRF_TASK_CONF.MODE = 1X, .START_CONF = 0, .REPEAT = 1:
tSynth
tRx-Tx
tRx
Sync Search
Receive Packet
tSearch
tSynth
tTx
tRx
Transmit ACK
CMD_RX
Sync Search
Give up sync search,
end task
T0536-01
NOTE: The time given by PRF_SEARCH_TIME is denoted tSearch and the time given by PRF_RX_TX is denoted
tRx-Tx. The setup and wait time for the synthesizer, receiver, and transmitter are denoted tSynth, tTx, and tRx, respectively.
Figure 25-12. Timing of Packets in Rx Tasks
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
315
Link Layer Engine
www.ti.com
For auto retransmit tasks, the time PRF_RETRANS_DELAY is the time from the end of a transmission to the
retransmission of the packet in case an ACK is not found or there is a CRC error; see Figure 25-13. The
values of PRF_SEARCH_TIME and the maximum packet length in PRF_PAYLOAD_LEN should be set such
that this time can always be achieved. If it is not possible to achieve the retransmission time, the packet is
retransmitted as early as possible.
Transmit task, PRF_TASK_CONF.MODE = 0X, .START_CONF = 0, .REPEAT = 1:
tSynth
tTx_Delay
tTx
tTx_Delay
tTx
tTx
Transmit Packet 0
Transmit Packet 2
Transmit Packet 1
CMD_TX
Transmit task, PRF_TASK_CONF.MODE = 0X, .START_CONF = 0, .REPEAT = 1:
tTx_Delay
tTx-Rx
tSynth
tTx
tRx
Transmit Packet 0
tTx
Sync Search
Receive ACK
Transmit Packet 1
CMD_TX
Transmit task, PRF_TASK_CONF.MODE = 1X, .START_CONF = 0, .REPEAT = 1,
no ACK found first time:
tRetrans
tSynth
tSearch
tTx-Rx
tTx
tTx-Rx
tTx
tRx
Transmit Packet 0
Sync Search
tRx
Sync
Search
Retransmit Packet 0
Receive
ACK 0
CMD_TX
Transmit task, PRF_TASK_CONF.MODE = 1X, .START_CONF = 0, .REPEAT = 1,
CRC error on ACK:
tRetrans
tSynth
tSearch
tTx-Rx
tTx
Transmit Packet 0
tTx-Rx
tTx
tRx
Sync
Search
Receive
ACK 0
tRx
Retransmit Packet 0
Sync
Search
Receive
ACK 0
CMD_TX
T0537-01
NOTE: The time given by PRF_TX_DELAY is denoted t Tx_Delay, the time given by PRF_SEARCH_TIME is
denoted tSearch, the time given by PRF_RETRANS_DELAY is denoted tRetrans, and the time given by PRF_TX_RX
is denoted tTx-Rx. The setup and wait times for the synthesizer, receiver, and transmitter are denoted tSynth, tTx, and tRx,
respectively.
Figure 25-13. Timing of Packets in Tx Tasks
316
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Random Number Generation
www.ti.com
25.9.3 RF Test Commands
In Table 25-12, there are listed some commands for modem test purposes. No registers are programmed
by the LLE based on these commands, so all registers must be set up by the MCU. This includes the
frequency word, which is otherwise written by the LLE; see Section 25.5 on how to program this. These
commands take no parameters, and do not cause the LLE to create any interrupts or to write any end
cause.
When receiving CMD_DEMOD_TEST, the LLE starts the receiver, but does not start sync search. The
receiver runs until a CMD_SHUTDOWN command (or a command starting another task) is sent.
When receiving CMD_RX_TEST, the LLE starts the receiver and starts sync search. Any received data is
discarded. The receiver runs until a CMD_SHUTDOWN command (or a command starting another task) is
sent.
When receiving a CMD_TX_TEST, the LLE starts the transmitter and starts sending all zeros. The Tx test
command is normally combined with configuration of the modem to send a tone or by the BSP to send a
whitening sequence. The transmitter runs until a CMD_SHUTDOWN command (or a command starting
another task) is sent. In order to send random modulated data for test purposes, it is recommended to set
BSP_MODE to 0x03 to enable both whiteners.
In order to send a continuous wave (CW), MDMCTRL0.TX_IF can be set to 1 before the CMD_TX_TEST
command is issued. In this case, the radio outputs a tone with the frequency given in
MDMTEST1.TX_TONE. In most cases, a tone at the synthesizer frequency is wanted (e.g., to measure
phase noise), in which case MDMTEST.TX_TONE should be set to 0x0A. The frequency synthesizer must
be programmed to the carrier frequency with no offset in this case; see Section 25.5.
When receiving a CMD_TX_FIFO_TEST, the LLE starts the transmitter and starts sending from the Tx
FIFO; otherwise, the command is as CMD_TX_TEST. The MCU must feed data into the Tx FIFO to avoid
underflow, and the Tx FIFO must be set up with auto commit and auto deallocate.
When receiving a CMD_PING command, the LLE responds with a PINGRSP. This can be used for
checking that the LLE code is running.
25.10 Random Number Generation
The CC2541 has a hardware pseudo-random register, as explained in Section 14.1. The RF core register
bank provides a second interface to this register. Reading the RFPSRND register is equivalent to reading
RNDL, then writing 01 to ADCCON1.RCTL.
For seeding the pseudo-random number generator and for tasks where higher entropy of the random
numbers is needed, the radio can be used as a true-random generator. The register RFRND provides
access to the least-significant bits of the radio ADC, which is random when noise is received. In order to
get values on this register, the receiver must be turned on. The value in RFRND is updated every 0.17 µs,
so the sampling of that register must be slower than that in order to get a new value with every sample.
To get true random numbers, the following procedure can be followed:
1. Program FREQCTRL to a channel that is not likely to contain a narrow-band signal. A frequency outside
the ISM band, such as a setting of 0, is recommended.
2. Program LNAGAIN to 0 to have minimum reception of a signal on the air.
3. Start the receiver in test mode by issuing a CMD_DEMOD_TEST command.
4. Wait until ADC data are ready. This can be seen by the RFRND register having a value different from 0.
5. Read the number of values needed from RFRND. Make sure that there is at least 0.17 µs between each
read (i.e., at least 6 cycles if running on 32 MHz). For instance, to seed the pseudo-random generator,
two values are needed.
6. Shut down the receiver by issuing a CMD_SHUTDOWN command
The values read from the RFRND register do not have a perfectly uniform distribution. In order to improve
this, several random numbers can be combined to produce one random number. One way of doing this is
to use the pseudo-random generator in CRC mode and combine 8 numbers into one. An example of how
this can be done is given in the C code below:
// Store LNA gain setting and set minimum LNA gain
lnagain_stored = LNAGAIN;
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
317
Packet Sniffing
www.ti.com
LNAGAIN = 0x00;
// Set lowest possible frequency to avoid signals in ISM band
FREQCTRL = 0x00;
// Enable radio in Rx without sync search
while (RFST != 0);
RFST = CMD_DEMOD_TEST;
// Wait for modem to be running
while (RFRND == 0);
// Seed RNG
RNDL = RFRND;
RNDL = RFRND;
for (j=0; j<ARRAY_SZ; j++) {
// Read 8 random bytes into CRC generation
RNDH = RFRND;
RNDH = RFRND;
RNDH = RFRND;
RNDH = RFRND;
RNDH = RFRND;
RNDH = RFRND;
RNDH = RFRND;
RNDH = RFRND;
// Read out LSB of CRC state
rndarray[j] = RNDL;
}
// Shut down radio
while (RFST != 0);
RFST = CMD_SHUTDOWN;
// Restore LNA gain setting
LNAGAIN = lnagain_stored;
25.11 Packet Sniffing
Packet sniffing is a nonintrusive way of observing data that is either being transmitted or received. The
packet sniffer outputs a clock and a data signal, which should be sampled on the rising edges of the clock.
The two packet-sniffer signals are observable as GPIO outputs. For accurate time stamping, the SFD
signal should also be output. The packet sniffer does not work for the 2 Mbps data rate.
The packet-sniffer mode is selected in register MDMCTRL3.RFC_SNIFF_CTRL; see Table 25-22 for a
description of the different modes of operation.
Table 25-22. Packet-Sniffer Modes of Operation
MDMCTRL3.RFC_SNIFF_CTRL
Description
00
Packet sniffer disabled
01
Data output from BSP. Tx data, including CRC, is whitened if the whitener is
enabled. Rx data, including CRC, is always de-whitened.
10
Data output from modulator. Only Tx data before whitening is output. Any CRC
bytes are 0.
11
Data output from the demodulator. Only Rx data before de-whitening is output.
The packet-sniffer clock frequency is equal to the RF data rate. The data is output serially, in the
received/transmitted order. It is possible to use a SPI slave to receive the data stream.
When a complete packet has been transferred, the packet sniffer appends a status byte that tells which
value of the FREQCTRL register was used and if it was a Tx or Rx packet (bit 0 is high if it was a Tx
packet). The appended byte is formatted as follows (first transmitted bit to the left):
318
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
FREQ(6)
FREQ(5)
FREQ(4)
FREQ(3)
FREQ(2)
FREQ(1)
FREQ(0)
TX
R0015-01
Figure 25-14. Complete Appended Packet
This allows for the external receiver to differentiate between Rx and Tx packets.
To set up the packet-sniffer signals or some of the other RF core observation outputs (in total maximum 3;
rfc_obs_sig0, rfc_obs_sig1, and rfc_obs_sig2), the user must perform the following steps:
Step1: Determine which signal (RFC_OBS_CTRL[0–2]) to output on which GPIO pin (P1[0:5]). This is
done using the OBSSELx control registers (OBSSEL0–OBSSEL5) that control the observation output to the
pins P1[0:5] (overriding the standard GPIO behavior for those pins).
Step2: Set the (RFC_OBS_CTRL[0–2]) control registers to select the correct signals (rfc_obs_sig); e.g.,
for packet sniffing one needs the rfc_sniff_data for the packet-sniffer data signal.and rfc_sniff_clk for the
corresponding clock signal.
Step3: Enable the packet-sniffer module in the MDMCTRL3 register.
25.12 Registers
25.12.1 Register Overview
25.12.1.1 SFR Registers
1 - RFIRQF0 (0xE9) RF interrupt flags
2 - RFIRQF1 (0x91) RF interrupt flags
3 - RFERRF (0xBF) RF error interrupt flags
4 - RFD (0xD9) RF data
5 - RFST (0x6189) LLE and FIFO commands
25.12.1.2 XREG Registers
Table 25-23. XREG Register Overview
Address (Hex)
+ 0x0000
+ 0x001
+ 0x002
+ 0x003
0x6180
FRMCTRL0
RFIRQM0
RFIRQM1
RFERRM
0x6184
FREQCTRL
FREQTUNE
TXPOWER
TXCTRL
0x6188
LLESTAT
SEMAPHORE0
SEMAPHORE1
0x618C
SEMAPHORE2
RFSTAT
RSSI
RFPSRND
0x6190
MDMCTRL0
MDMCTRL1
MDMCTRL2
MDMCTRL3
0x6194
SW_CONF
SW0
SW1
SW2
0x6198
SW3
FREQEST
RXCTRL
FSCTRL
LNAGAIN
AAFGAIN
ADCTEST0
0x61A4
MDMTEST0
MDMTEST1
0x61A8
ATEST
0x619C
0x61A0
RFC_OBS_CTRL0
0x61AC
0x61B0
RFC_OBS_CTRL2
0x61BC
TXFILTCFG
RFC_OBS_CTRL1
LLECTRL
RFRND
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio 319
Registers
www.ti.com
Table 25-23. XREG Register Overview (continued)
Address (Hex)
+ 0x0000
+ 0x001
+ 0x002
+ 0x003
0x61C0
RFRAMCFG
0x61C4
RFFDMA1
RFFSTATUS
RFFCFG
RFFDMA0
0x61C8
RFRXFLEN
RFRXFTHRS
RFRXFWR
RFRXFRD
0x61CC
RFRXFWP
RFRXFRP
RFRXFSWP
RFRXFSRP
0x61D0
RFTXFLEN
RFTXFTHRS
RFTXFWR
RFTXFRD
0x61D4
RFTXFWP
RFTXFRP
RFTXFSWP
RFTXFSRP
0x61E0
BSP_P0
BSP_P1
BSP_P2
BSP_P3
0x61E4
BSP_D0
BSP_D1
BSP_D2
BSP_D3
0x61E8
BSP_W
BSP_MODE
BSP_DATA
0x61F8
SW4
SW5
SW6
SW7
0x61FC
DC_I_L
DC_I_H
DC_Q_L
DC_Q_H
25.12.2 Register Settings Update
This section contains a summary of the register settings that should be updated from their default value to
have optimal performance. For some of the registers, the setting depends on the required gain in the
receiver chain for bit rates of 1 Mbps and lower. For 2 Mbps, other values are needed, and different
values should be used for Rx and Tx tasks. Note that registers that are listed in only one of the following
two tables should have their reset value in the other case.
Table 25-24. Registers That Should Be Updated From Their Default Value,
Bit Rates 1 Mbps and Lower
Address (Hex)
Standard Gain:
New Value (Hex) (1)
High Gain:
New Value (Hex)
FRMCTRL0
6180
43
43
Amplitude weight in frequency offset
compensation (assuming sync word
included in CRC and MSB first)
MDMCTRL1
6191
48
48
Sync word correlation threshold (32-bit
sync word)
MDMCTRL2
6192
C0
C0
Use inverse of preamble for frequency
offset estimation (assuming MSB first)
MDMCTRL3
6193
63
63
Set RSSI mode to peak detect after
sync
RXCTRL
619A
33
3F
Receiver currents
FSCTRL
619B
55
5A
Prescaler and mixer currents
LNAGAIN
61A0
3A
7F
LNA gain
TXFILTCFG
61BC
03
03
Sets Tx anti-aliasing filter to appropriate
bandwidth
TXPOWER
6186
E1
E1
Tx power (0 dBm)
TXCTRL
6187
19
19
DAC current
IVCTRL
6265
1B
1B
PA, mixer, and DAC bias
Register Name
(1)
320
Description
Not all modulation types are characterized for the standard gain setting; see the CC2541 2.4-GHz Bluetooth low energy and
Proprietary System-on-Chip data sheet (SWRS110).
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
Table 25-25. Registers That Should Be Updated From Their Default Value, Bit Rate 2 Mbps
Register Name
Address (Hex)
Rx Tasks
Tx Tasks
Description
FRMCTRL0
6180
43
43
Amplitude weight in frequency offset
compensation (assuming sync word is
included in CRC and MSB first)
MDMCTRL1
6191
48
48
Sync word correlation threshold (32-bit sync
word)
MDMCTRL2
6192
CC
CC
Use inverse of preamble for frequency offset
estimation (assuming MSB first); set extra
preamble bytes
MDMCTRL3
6193
63
63
Set RSSI mode to peak detect after sync
RXCTRL
619A
29
29
Receiver currents
FSCTRL
619B
5A
5A
Prescaler and mixer currents
ADCTEST0
61A2
66
66
Reduce ADC gain
MDMTEST0
61A5
0F
01
Select dc offset compensation method;
change RSSI averaging
TXFILTCFG
61BC
03
03
Transmit filter bandwidth
PRF_PKT_CONF
6003
06
06
Enable AGC and start tone
PRF_RADIO_CONF
607E
90
D0
Set 1-MHz Tx IF and dc write-back; for Tx
tasks also special dc offset compensation
TXPOWER
6186
E1
E1
Tx power (0 dBm)
TXCTRL
6187
19
19
DAC current
IVCTRL
6265
1B
1B
PA, mixer, and DAC bias
The values for FRMCTRL0, MDMCTRL2, and PRF_PKT_CONF may require further adjustment based on the
frame format, and the correlation threshold in MDMCTRL1 should be adjusted according to the sync word
length, see Section 25.7.
In addition to these modifications, registers must be set in order to set up the modulation format, packet
handling, etc., as explained throughout this chapter.
25.12.3 SFR Register Descriptions
RFIRQF0 (0xE9) – RF Interrupt Flags
Bit
No.
7:4
3
Name
Reset
R/W
Description
–
RXTHSHUP
0000
0
R0
R/W0
2
TXTHSHUP
0
R/W0
1
RXTHSHDN
0
R/W0
0
TXTHSHDN
0
R/W0
Reserved
Rx FIFO goes above its upper threshold.
0: No interrupt pending
1: Interrupt pending
Tx FIFO goes above its upper threshold.
0: No interrupt pending
1: Interrupt pending
Rx FIFO goes below its lower threshold.
0: No interrupt pending
1: Interrupt pending
Tx FIFO goes below its lower threshold.
0: No interrupt pending
1: Interrupt pending
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
321
Registers
www.ti.com
RFIRQF1 (0x91) – RF Interrupt Flags
Bit
No.
7
Name
Reset
R/W
Description
PINGRSP
0
R/W0
6
TASKDONE
0
R/W0
5
TXDONE
0
R/W0
4
RXEMPTY
0
R/W0
3
RXIGNORED
0
R/W0
2
RXNOK
0
R/W0
1
TXFLUSHED
0
R/W0
0
RXOK
0
R/W0
When receiving a CMD_PING command, the LLE responds with a
PINGRSP. This can be used for checking that the LLE is running.
0: No interrupt pending
1: Interrupt pending
Tx FIFO packet completed
0: No interrupt pending
1: Interrupt pending
Tx FIFO packet completed
0: No interrupt pending
1: Interrupt pending
Empty packet received
0: No interrupt pending
1: Interrupt pending
Packet received with unexpected sequence number
0: No interrupt pending
1: Interrupt pending
Packet received with CRC error
0: No interrupt pending
1: Interrupt pending
Tx ACK buffer flushed
0: No interrupt pending
1: Interrupt pending
Packet received correctly
0: No interrupt pending
1: Interrupt pending
RFERRF (0xBF) – RF Error Interrupt Flags
Bit
No.
7
6
Name
Reset
R/W
Description
–
RXFIFOFULL
0
0
R/W0
R/W0
5
LLEERR
0
R/W0
4
RXTXABO
0
R/W0
3
RXOVERF
0
R/W0
2
TXOVERF
0
R/W0
1
RXUNDERF
0
R/W0
0
TXUNDERF
0
R/W0
Reserved
Rx FIFO is full when trying to store received data
0: No interrupt pending
1: Interrupt pending
LLE command or parameter error
0: No interrupt pending
1: Interrupt pending
Receive or transmit operation aborted
0: No interrupt pending
1: Interrupt pending
Rx FIFO overflow
0: No interrupt pending
1: Interrupt pending
Tx FIFO overflow
0: No interrupt pending
1: Interrupt pending
Rx FIFO underflow
0: No interrupt pending
1: Interrupt pending
Tx FIFO underflow
0: No interrupt pending
1: Interrupt pending
Name
Reset
R/W
Description
RFD[7:0]
0x00
R/W
Data written to the register is written to the Tx FIFO. When reading this
register, data from the Rx FIFO is read.
RFD (0xD9) – RF data
Bit
No.
7:0
322
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
RFST (0xE1) – LLE and FIFO commands
Bit
No.
7:0
Name
Reset
R/W
Description
RFST[7:0]
0x00
R/WH0
Commands to radio are written to this register. The register is cleared (set
to 0x00) when the radio is ready for a new command.
25.12.3.1 XREG Register Descriptions
FRMCTRL0 (0x6180) – Frame Control
Bit
No.
7:5
Name
Reset
R/W
Description
FOC_MAGN_CONT[2:0]
000
R/W
4:2
1
–
SW_CRC_MODE
000
0
R/W
R/W
0
ENDIANNESS
0
R/W
Controls how signal amplitude is weighted into the frequency offset
compensation scheme.
000: Magnitude has no effect
001 to 111: Low-to-high weighting of the magnitude
Reserved always write 0.
0: The sync word is not included in the crc calculation.
1: The sync word is included in the crc calculation. Only to be used with
whitening disabled.
0: The data goes LSB-first over the air.
1: The data goes MSB-first over the air.
Name
Reset
R/W
Description
–
RXTHSHUP
0000
0
R0
R/W
2
TXTHSHUP
0
R/W
1
RXTHSHDN
0
R/W
0
TXTHSHDN
0
R/W
Reserved
Rx FIFO goes above its upper threshold.
0: Interrupt disabled
1: Interrupt enabled
Tx FIFO goes above its upper threshold.
0: Interrupt disabled
1: Interrupt enabled
Rx FIFO goes below its lower threshold.
0: Interrupt disabled
1: Interrupt enabled
Tx FIFO goes below its lower threshold.
0: Interrupt disabled
1: Interrupt enabled
RFIRQM0 (0x6181) – RF Interrupt Masks
Bit
No.
7:4
3
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
323
Registers
www.ti.com
RFIRQM1 (0x6182) – RF Interrupt Masks
Bit
No.
7
Name
Reset
R/W
Description
PINGRSP
0
R/W
6
TASKDONE
0
R/W
5
TXDONE
0
R/W
4
RXEMPTY
0
R/W
3
RXIGNORED
0
R/W
2
RXNOK
0
R/W
1
TXFLUSHED
0
R/W
0
RXOK
0
R/W
When receiving a CMD_PING command, the LLE responds with a
PINGRSP. This can be used for checking that the LLE is running.
0: Interrupt disabled
1: Interrupt enabled
Tx FIFO packet completed
0: Interrupt disabled
1: Interrupt enabled
TX FIFO packet completed
0: Interrupt disabled
1: Interrupt enabled
Empty packet received
0: Interrupt disabled
1: Interrupt enabled
Packet received with unexpected sequence number
0: Interrupt disabled
1: Interrupt enabled
Packet received with CRC error
0: Interrupt disabled
1: Interrupt enabled
Tx ACK buffer flushed
0: Interrupt disabled
1: Interrupt enabled
Packet received correctly
0: Interrupt disabled
1: Interrupt enabled
RFERRM (0x6183) – RF Error Interrupt Masks
Bit
No.
7
6
Name
Reset
R/W
Description
–
RXFIFOFULL
0
0
R/W
R/W
5
LLEERR
0
R/W
4
RXTXABO
0
R/W
3
RXOVERF
0
R/W
2
TXOVERF
0
R/W
1
RXUNDERF
0
R/W
0
TXUNDERF
0
R/W
Reserved
RX FIFO is full when trying to store received data.
0: Interrupt disabled
1: Interrupt enabled
LLE command or parameter error
0: Interrupt disabled
1: Interrupt enabled
Receive or transmit operation aborted
0: Interrupt disabled
1: Interrupt enabled
Rx FIFO overflow
0: Interrupt disabled
1: Interrupt enabled
Tx FIFO overflow
0: Interrupt disabled
1: Interrupt enabled
Rx FIFO underflow
0: Interrupt disabled
1: Interrupt enabled
Tx FIFO underflow
0: Interrupt disabled
1: Interrupt enabled
324
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
FREQCTRL (0x6184) – Synthesizer Frequency Control
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
FREQ[6:0]
0
0x16
R0
R/W
Reserved. Read as zero
Frequency control word. Controls frequency of local oscillator. See
Section 25.5 for the relation between the LO frequency and the RF
frequency.
ƒLO = (2379 + FREQ[6:0]) MHz
The frequency word in FREQ[6:0] is an offset value from 2379 MHz. The
device supports the frequency range from 2379 MHz to 2495 MHz. The
usable settings for FREQ[6:0] are consequently 0 to 116. Settings outside
this range (117–127) give a frequency of 2495 MHz.
FREQTUNE (0x6185) – Crystal Oscillator Frequency Tuning
Bit
No.
7:4
3:0
Name
Reset
R/W
Description
–
XOSC32M_TUNE[3:0]
0x0
0xF
R0
R/W
Reserved. Read as zero
Tune crystal oscillator. The default setting of 1111 leaves the XOSC not
tuned. Changing setting from default switches in extra capacitance to the
oscillator, effectively lowering the XOSC frequency. Hence, a higher
setting gives a higher frequency.
TXPOWER (0x6186) – Output Power Control
Bit
No.
7:0
Name
Reset
R/W
Description
PA_POWER[7:0]
0xF5
R/W
PA power control. NOTE: This value should be updated. See the device
data sheet (Appendix C) for recommended values.
Name
Reset
R/W
Description
–
–
DAC_CURR[1:0]
DAC_DC[1:0]
TXMIX_CURRENT[1:0]
0
1
10
10
01
R0
R/W
R/W
R/W
R/W
Reserved
Reserved
Change the current in the DAC to change the current steps
Adjusts the dc level to the Tx mixer.
Transmit mixers core current: Current increases with increasing setting.
Name
Reset
R/W
Description
–
AGC_LOWGAIN
WAIT_T2E1
000
0
0
R0
R
R
2
LLE_IDLE
0
R
1
SYNC_SEARCH
0
R
0
VCO_ON
0
R
Reserved
1 if the AGC algorithm has reduced the front-end gain; 0 otherwise
Indication on the LLE waiting for Timer 2 event 1 to start a task
0: Not waiting for Timer 2 event 1
1: Command processed, event 1 not yet received
Link-layer engine idle
0: The LLE is busy processing or finishing a command, or in reset.
1: The LLE is idle waiting for a command to start a new task.
Rx search for sync
0: The modem is not ready to receive a packet
1: The modem is in search for a sync word or is receiving a packet
VCO on
0: The VCO is powered down, so the next Rx or Tx operation must start
and calibrate the synthesizer before transmitting or receiving
1: The VCO is powered up. If the LLE is idle, it means the next task starts
quickly if frequency programming is disabled (PRF_CHAN.FREQ =
127)
TXCTRL (0x6187) – Tx Settings
Bit
No.
7
6
5:4
3:2
1:0
LLESTAT (0x6188) – LLE Status
Bit
No.
7:5
4
3
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
325
Registers
www.ti.com
SEMAPHORE0 (0x618A) – Semaphore for Accessing RF Data Memory
Bit
No.
7:1
0
Name
Reset
R/W
Description
–
0000
000
1
R0
Reserved, read as 0
R/W1
When SEMAPHORE = 1 and SEMAPHORE0 is read, SEMAPHORE is
set to 0. SEMAPHORE can only be set to 1 by a reset or by writing 1 to
it.
SEMAPHORE
SEMAPHORE1 (0x618B) – Semaphore for Accessing RF Data Memory
Bit
No.
7:1
Name
Reset
R/W
Description
–
0000
000
1
R0
Reserved, read as 0
R/W1
When SEMAPHORE = 1 and SEMAPHORE1 is read, SEMAPHORE is
set to 0. SEMAPHORE can only be set to 1 by a reset or by writing 1 to
it.
Name
Reset
R/W
Description
–
0000
000
1
R0
Reserved, read as 0
R/W1
When SEMAPHORE = 1 and SEMAPHORE2 is read, SEMAPHORE is
set to 0. SEMAPHORE can only be set to 1 by a reset or by writing 1 to
it.
Name
Reset
R/W
Description
MOD_UNDERFLOW
DEM_STATUS
0
00
R/W0
R
4
SFD
0
R
3
CAL_RUNNING
0
R
2
1
0
LOCK_STATUS
TX_ACTIVE
RX_ACTIVE
0
0
0
R
R
R
Modulator has underflowed. Must be cleared by software
Demodulator status
00: Idle
01: Active
10: Finishing
11: Error
High when the sync word has been sent in TX or when sync has been
obtained in RX
Frequency synthesizer calibration status.
0: Calibration done or not started
1: Calibration in progress.
1 when PLL is in lock; 0 otherwise
Status signal, active when the LLE is in one of the transmit states
Status signal, active when the LLE is in one of the receive states
0
SEMAPHORE
SEMAPHORE2 (0x618C) – Semaphore
Bit
No.
7:1
0
SEMAPHORE
RFSTAT (0x618D) – RF Core Status
Bit
No.
7
6:5
RSSI (0x618E) – Received Signal Strength Indicator
Bit
No.
7:0
Name
Reset
R/W
Description
RSSI_VAL[7:0]
0x80
R
RSSI estimate on a logarithmic scale. Unit is 1 dB; offset depends on the
gain of the Rx chain, including external components; see the device data
sheet. The reset value of 0x80 also indicates that the RSSI value is
invalid/measurement not yet complete.
RFPSRND (0x618F) – Pseudorandom Number Generator
Bit
No.
7:0
326
Name
Reset
R/W
Description
RNG_DOUT[7:0]
0x00
R
The value read from the pseudorandom number generator, see
Chapter 14. Reading this register generates causes the shift register to be
updated with 13 times rollout.
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
MDMCTRL0 (0x6190) – Modem Configuration
Bit
No.
7:6
Name
Reset
R/W
Description
FOC_DECAY[1:0]
00
R/W
5
TX_IF
0
R/W
4:1
MODULATION[3:0]
0010
R/W
0
PHASE_INVERT
0
R/W
Controls decay ratio of frequency offset compensation mechanism. Value
by which to increment preamble cost function at each decay
00: 8
01: 16
10: 32
11: 64
0: Modulation is done at an IF set by rfr_tx_tone.
1: Modulator outputs tone set by rfr_tx_tone.
Modulation scheme
0010: GFSK 250-kHz deviation, 1-Mbps data rate
0011: GFSK 500-kHz deviation, 2-Mbps data rate
0100: GFSK 160-kHz deviation, 250-kbps data rate
0110: GFSK 160-kHz deviation, 1-Mbps data rate
0111: GFSK 320-kHz deviation, 2-Mbps data rate
1000: MSK, 250-kbps data rate
1001: MSK, 500-kbps data rate
Others: Reserved
Set one of two RF modulation modes for Rx / Tx
0: Normal (binary 0 represented with negative frequency deviation, binary
1 represented with positive frequency deviation)
1: Inverted phase (binary 0 represented with positive frequency deviation,
binary 1 represented with negative frequency deviation)
MDMCTRL1 (0x6191) – Modem Configuration
Bit
No.
7:6
5
4:0
Name
Reset
R/W
Description
FOC_MODE
01
R/W
–
CORR_THR[4:0]
0
0x0F
R0
R/W
Frequency-offset average filter behavior
00: No frequency-offset compensation done
01: Freeze frequency-offset estimate after sync
10: Continuously estimate and remove frequency offset
11: Freeze the frequency-offset estimate after sync, use double decay rate
Reserved
Demodulator correlator threshold value, used in sync search. Optimal
threshold value depends on SW_CONF.SW_LEN. CORR_THR adjusts
how the receiver synchronizes to data from the radio. If threshold is set
too low, sync can more easily be found on noise. If set too high, the
sensitivity is reduced but sync is not likely to be found on noise.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
327
Registers
www.ti.com
MDMCTRL2 (0x6192) – Modem Configuration
Bit
No.
7
Name
Reset
R/W
Description
SW_BIT_ORDER
0
R/W
6
DEM_PREAM_MODE
0
R/W
5:4
PREAM_SEL[1:0]
00
R/W
3:0
NUM_PREAM_BYTES[3:0]
0000
R/W
0: The sync word is transmitted LSB to MSB (from SYNC_WORD[0] to
SYNC_WORD[31]), and in receive the correlator expects this bit
ordering.
1: The sync word is transmitted MSB to LSB (from SYNC_WORD[31]
to SYNC_WORD[0]), and in receive the correlator expects this bit
ordering.
Use PREAM_SEL[1:0] or 1s complement of PREAM_SEL[1:0] for
frequency offset estimation.
00: Select preamble based on first bit of sync word; last bit of preamble is
inverse of first bit of sync word.
01: Select preamble based on first bit of sync word; last bit of preamble is
same as first bit of sync word.
10: Use preamble 0101 0101
11: Use preamble 1010 1010
The number of preamble bytes to be sent in Tx mode prior to the sync
word
0000: 1 leading preamble byte
0001: 2 leading preamble bytes
0010: 3 leading preamble bytes
0011: 4 leading preamble bytes
…
1111: 16 leading preamble bytes
MDMCTRL3 (0x6193) – Modem Configuration
Bit
No.
7:6
Name
Reset
R/W
Description
SYNC_MODE[1:0]
01
R/W
5
RAMP_AMP
1
R/W
4:3
RFC_SNIFF_CTRL[1:0]
00
R/W
2
1:0
–
RSSI_MODE[1:0]
0
00
R0
R/W
00: Correlation above threshold is sufficient as sync criterion.
01: Correlation value above threshold and data decision on all symbols of
sync word is used as sync criterion.
10: Correlation value above threshold and data decision on all symbols of
sync word is used as sync criterion. Accept one bit error in sync word
11: Reserved
1: Enable ramping of DAC output amplitude during startup and finish.
0: Disable ramping of DAC output amplitude.
Enable / disable rfc_sniff.
00: Sniffer disabled
01: Output data out of the BSP
10: Output data out of the modulator before the BSP
11: Output data out of the demodulator before the BSP
Reserved. Read as 0.
Controls mode of RSSI
00 : Continuous mode
01 : Freeze estimate at sync
10 : Peak detect
11 : Continuous before sync, peak detect after sync
328
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
SW_CONF (0x6194) – Sync Word Configuration
Bit
No.
7
Name
Reset
R/W
Description
DUAL_RX
0
R/W
6
5
–
SW_RX
0
0
R/W
R
0: Only search for primary SW
1: Search for both primary and secondary SW
Reserved. Always write 0.
0: Primary SW received
1: Secondary SW received
Valid only when RFSTAT.SFD is 1
4:0
SW_LEN[4:0]
0 0000
R/W
Determines how many of the bits in SYNC_WORD are to be used. This
allows for arbitrary sync word lengths.
0 0000: 32-bit SW
0 0001 to 0 11111: Reserved
1 0000: 16-bit SW
1 0001: 17-bit SW
1 0010: 18-bit SW
1 0011: 19-bit SW
…
1 1111: 31-bit SW
SW0 (0x6195) – Primary Sync Word Byte 0
Bit
No.
7:0
Name
Reset
R/W
Description
SYNC_WORD[7:0]
0x00
R/W
Contains bits 7:0 of the primary synchronization word
SW1 (0x6196) – Primary Sync Word Byte 1
Bit
No.
7:0
Name
Reset
R/W
Description
SYNC_WORD[15:8]
0x00
R/W
Contains bits 15:8 of the primary synchronization word
SW2 (0x6197) – Primary Sync Word Byte 2
Bit
No.
7:0
Name
Reset
R/W
Description
SYNC_WORD[23:16]
0x00
R/W
Contains bits 23:16 of the primary synchronization word
SW3 (0x6198) – Primary Sync Word Byte 3
Bit
No.
7:0
Name
Reset
R/W
Description
SYNC_WORD[31:24]
0x00
R/W
Contains bits 31:24 of the primary synchronization word
SW4 (0x61F8) – Secondary Sync Word Byte 0
Bit
No.
7:0
Name
Reset
R/W
Description
SYNC_WORD2[7:0]
0x00
R/W
Contains bits 7:0 of the secondary synchronization word
SW5 (0x61F9) – Secondary Sync Word Byte 1
Bit
No.
7:0
Name
Reset
R/W
Description
SYNC_WORD2[15:8]
0x00
R/W
Contains bits 15:8 of the secondary synchronization word
SW6 (0x61FA) – Secondary Sync Word Byte 2
Bit
No.
7:0
Name
Reset
R/W
Description
SYNC_WORD2[23:16]
0x00
R/W
Contains bits 23:16 of the secondary synchronization word
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
329
Registers
www.ti.com
SW7 (0x61FB) – Secondary Sync Word Byte 3
Bit
No.
7:0
Name
Reset
R/W
Description
SYNC_WORD2[31:24]
0x00
R/W
Contains bits 31:24 of the secondary synchronization word
FREQEST (0x6199) – Estimated RF Frequency Offset
Bit
No.
7:0
Name
Reset
R/W
Description
FREQEST[7:0]
0x00
R
Signed value. Contains an estimate of the frequency offset between
carrier and the receiver frequency. FOC_MODE controls when this
estimate is updated.
RXCTRL (0x619A) – Receive Section Tuning
Bit
No.
7:6
5:4
Name
Reset
R/W
Description
–
GBIAS_LNA2_REF[1:0]
00
10
R0
R/W
3:2
GBIAS_LNA_REF[1:0]
10
R/W
1:0
MIX_CURRENT[1:0]
01
R/W
Reserved
Adjusts front-end LNA2/mixer PTAT current output (M =
GBIAS_LNA2_REF[1:0] + 3), default: M = 5.
Adjusts front-end LNA PTAT current output (M = GBIAS_LNA_REF[1:0]
+ 3), default: M = 5.
Control of the receiver mixers output current. The current increases with
increasing setting.
FSCTRL (0x619B) – Frequency Synthesizer Tuning
Bit
No.
7:6
5:4
Name
Reset
R/W
Description
PRE_CURRENT [1:0]
LODIV_BUF_CURRENT_TX
[1:0]
LODIV_BUF_CURRENT_RX
[1:0]
LODIV_CURRENT [1:0]
01
01
R/W
R/W
01
R/W
01
R/W
Prescaler current setting
Adjusts current in mixer and PA buffers (lodiv_buf_current). Used when
lle_tx_active = 1
Adjusts current in mixer and PA buffers (lodiv_buf_current). Used when
lle_tx_active = 0
Adjusts divider currents, except mixer and PA buffers.
Name
Reset
R/W
Description
–
LNA1_CURRENT[1:0]
0
11
R0
R/W
4:2
LNA2_CURRENT[2:0]
111
R/W
1:0
LNA3_CURRENT[1:0]
11
R/W
Reserved, read as 0
Gain setting LNA1
00: 0-dB gain (reference level)
01: 3-dB gain
10: Reserved
11: 6-dB gain
Gain setting LNA2
000: 0-dB gain (reference level)
001: 3-dB gain
010: 6-dB gain
011: 9-dB gain
100: 12-dB gain
101: 15-dB gain
110: 18-dB gain
111: 21-dB gain
Gain setting LNA3
00: 0-dB gain (reference level)
01: 3-dB gain
10: 6-dB gain
11: 9-dB gain
3:2
1:0
LNAGAIN (0x61A0) – LNA Gain Setting
Bit
No.
7
6:5
330
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
AAFGAIN (0x61A1) – AAF Gain Setting
Bit
No.
7:2
1:0
Name
Reset
R/W
Description
–
AAF_GAIN[1:0]
0000 00 R0
11
R/W
Reserved. Read as zero
Controls attenuation in AAF
00: 9-dB attenuation in AAF
01: 6-dB attenuation in AAF
10: 3-dB attenuation in AAF
11: 0-dB attenuation in AAF (reference level)
Name
Reset
R/W
Description
ADC_ADJ[7:0]
0x10
R/W
Adjust ADC gain
ADCTEST0 (0x61A2) – ADC Tuning
Bit
No.
7:0
MDMTEST0 (0x61A5) – Modem Configuration
Bit
No.
7:5
Name
Reset
R/W
Description
RSSI_ACC[2:0]
101
R/W
4
3:2
–
DC_BLOCK_LENGTH[1:0]
0
00
R/W
R/W
1:0
DC_BLOCK_MODE[1:0]
01
R/W
RSSI accuracy
000: 5.33-µs average window
001: Mean of two 5.33-µs average windows
010: Reserved
011: Mean of four 5.33-µs average windows
100: 21.3-µs average window
101: Mean of two 21.3-µs average windows
110: Reserved
111: Mean of four 21.3-µs average windows
Reserved, always write 0.
Controls the number of samples to be accumulated between each dump of
the accumulate-and-dump filter used in dc removal.
00: 16 samples
01: 32 samples
10: 64 samples
11: 128 samples
Selects the mode of operation:
00 : Manual override mode
01 : Enable dc cancellation. Normal operation
10 : Freeze estimates of dc when sync is found. Start estimating dc again
when searching for the next frame.
11 : Delayed dc offset estimate used. Delay set by
MDMTEST1.DC_DELAY. Until the first estimate is ready, the manual
override value is used.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
331
Registers
www.ti.com
MDMTEST1 (0x61A6) – Modem Configuration
Bit
No.
7:6
Name
Reset
R/W
Description
DC_DELAY
00
R/W
Controls delay of dc estimate delayed dc block mode. Delay unit is set by
MDMTEST0.DC_BLOCK_LENGTH
5
RX_IF
0
R/W
4:0
TX_TONE[4:0]
0 0000
R/W
Name
Reset
R/W
–
ATEST_CTRL[5:0]
00
R0
00 0000 R/W
00: 5 delays
01: 6 delays
10: 7 delays
11: 8 delays
Controls mixer frequency in demodulator (not 2 Mbps)
0: 1 MHz
1: –1 MHz
For 2 Mbps, always write 0. The receiver then operates at zero IF.
Controls baseband frequency of transmission
Note: If MDMCTRL0.PHASE_INVERT is 1, the sign of the frequency
is inverted
0: –8 MHz
1: –6 MHz
2: –4 MHz
3: –3 MHz
4: –2 MHz
5: –1 MHz
6: –500 kHz
7: –250 kHz
8: –125 kHz
9: –4 kHz
10: 0 Hz
11: 4 kHz
12: 125 kHz
13: 250 kHz
14: 500 kHz
15: 1 MHz
16: 2 MHz
17: 3 MHz
18: 4 MHz
19: 6 MHz
20: 8 MHz
ATEST (0x61A9) – Analog Test Control
Bit
No.
7:6
5:0
332
CC2541 Proprietary Mode Radio
Description
Reserved. Read as zero
Controls the analog test mode:
00 0000: Disabled
00 0001: Enables the temperature sensor (see also the TR0 register
description in #IMPLIED).Other values reserved.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
RFC_OBS_CTRL0 (0x61AE) – RF Observation Mux Control 0
Bit
No.
7
6
5:0
Name
Reset
R/W
–
RFC_OBS_POL0
RFC_OBS_MUX0
0
R0
0
R/W
00 0000 R/W
Description
Reserved. Read as 0
The signal chosen by RFC_OBS_MUX0 is XORed with this bit
Controls which observable signal from rf_core is to be muxed out to
rfc_obs_sigs(0); see Section 7.9.
00 0111: rfc_sniff_data – Data from packet sniffer, see Section 25.11
00 1000: rfc_sniff_clk – Clock for packet sniffer data, see Section 25.11
00 1001: tx_active
00 1010: rx_active
00 1011: vco_on – VCO on
Low: The VCO is powered down, so the next Rx or Tx operation must start
and calibrate the synthesizer before transmitting or receiving
High: The VCO is powered up. If the LLE is idle, it means the next task
starts quickly if frequency programming is disabled (PRF_CHAN.FREQ
= 127)
00 1100: sync_search – Rx search for sync
Low: The modem is not ready to receive a packet
High: The modem is in search for a sync word or receiving a packet
00 1101: lle_idle – Link-layer engine idle
Low: The LLE is busy processing or finishing a command, or in reset
High: The LLE is idle waiting for a command to start a new task
00 1110: wait_t2e1 – Indication on the LLE waiting for Timer 2 event 1 to
start a task
Low: Not waiting for Timer 2 event 1
High: Command processed, event 1 not yet received
00 1111: agc_lowgain – High if the AGC algorithm has reduced the frontend gain; low otherwise
01 0011: fsc_lock – High when PLL is in lock; low otherwise
01 1011: pa_pd - Power amplifier power-down signal
10 1100: lnamix_pd - Low-noise amplifier power-down signal
11 0000: dem_sync_found - High when demodulator has detected a sync
word. Stays high until end of packet.
11 0001: mod_sync_sent - High when modulator has sent a sync word.
Stays high until end of packet.
Others: Reserved
RFC_OBS_CTRL1 (0x61AF) – RF Observation Mux Control 1
Bit
No.
7
6
5:0
Name
Reset
R/W
–
RFC_OBS_POL1
RFC_OBS_MUX1
0
R0
0
R/W
00 0000 R/W
Description
Reseved. Read as 0
The signal chosen by RFC_OBS_MUX1 is XORed with this bit.
Controls which observable signal from rf_core is to be muxed out to
rfc_obs_sigs(1). See description of RFC_OBS_CTRL0.
RFC_OBS_CTRL2 (0x61B0) – RF Observation Mux Control 2
Bit
No.
7
6
5:0
Name
Reset
R/W
–
RFC_OBS_POL2
RFC_OBS_MUX2
0
R0
0
R/W
00 0000 R/W
Description
Reseved. Read as 0
The signal chosen by RFC_OBS_MUX2 is XORed with this bit.
Controls which observable signal from rf_core is to be muxed out to
rfc_obs_sigs(2). See description of RFC_OBS_CTRL0.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
333
Registers
www.ti.com
LLECTRL (0x61B1) – LLE Control
Bit
No.
7:3
2:1
0
Name
Reset
R/W
Description
–
LLE_MODE_SEL
0
00
R0
R/W
LLE_EN
0
R/W
Reseved. Read as 0
LLE mode. Changing this field has no effect unless LLE_EN is changed
from 0 to 1.
00: Proprietary mode (described in this chapter)
01: BLE mode (only for use by the BLE stack)
Others: Reserved
Must be set to 0 before entering PM2 or PM3, otherwise the behavior of
the RF core after waking up may be unpredictable.
0: LLE held in reset
1: LLE enabled
TXFILTCFG (0x61BC) – TX Filter Configuration
Bit
No.
7:4
3:2
1:0
Name
Reset
R/W
Description
–
–
FC
0000
11
11
R0
R/W
R/W
Reserved
Reserved
Sets TX anti-aliasing filter to appropriate bandwidth. Reduces spurious
emissions close to signal. For the best value to use, see Table 25-24 and
Table 25-25.
Name
Reset
R/W
Description
RND
0x00
R
Random bits, provided analog part is in random number generation mode
(receiver running without sync)
RFRND (0x61BF) – Random Data
Bit
No.
7:0
RFRAMCFG (0x61C0) – Radio RAM Configuration
Bit
No.
7:3
2:0
334
Name
Reset
R/W
Description
–
PRE
0000 1
000
R
R/W
Reserved
Selects active memory page for RF core data memory
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
RFFDMA0, (0x61C3) – Radio DMA Trigger 0 Control
Bit
No.
7:5
4:0
Name
Reset
R/W
Description
–
DMA0
000
0x00
R
R/W
Reserved
Generate a pulse on radio DMA trigger 0 (DMA trigger 19) when:
0x00: Never
0x01: A byte is read from Rx FIFO and more bytes remain or when a byte
arrives in Rx FIFO and it was previously empty.
0x02: A byte is written to Rx FIFO and there is available space left or
when there becomes available space when the Rx FIFO was full.
0x03: Rx FIFO is empty.
0x04: Rx FIFO is full.
0x05: Rx FIFO length equals RFRXFTHRS after a write to Rx FIFO.
0x06: Rx FIFO is read when its size equals RFRXFTHRS.
0x07: Rx FIFO is reset (see Table 25-2).
0x08: Rx FIFO is deallocated (see Table 25-2).
0x09: Rx FIFO is retried (see Table 25-2).
0x0A: Rx FIFO is discarded (see Table 25-2).
0x0B: Rx FIFO is committed (see Table 25-2).
0x0C–0x0F: Reserved (never)
0x10: Never
0x11: A byte is read from Tx FIFO and more bytes remain or when a byte
arrives in Tx FIFO and it was previously empty.
0x12: A byte is written to Tx FIFO and there is available space left or
when there becomes available space when the Tx FIFO was full.
0x13: Tx FIFO is empty.
0x14: Tx FIFO is full.
0x15: Tx FIFO length equals RFTXFTHRS after a write to Tx FIFO.
0x16: Tx FIFO is read when its size equals RFTXFTHRS.
0x17: Tx FIFO is reset (see Table 25-2).
0x18: Tx FIFO is deallocated (see Table 25-2).
0x19: Tx FIFO is retried (see Table 25-2).
0x1A: Tx FIFO is discarded (see Table 25-2).
0x1B: Tx FIFO is committed (see Table 25-2).
0x1C–0x1F: Reserved (never)
RFFDMA1, (0x61C4) – Radio DMA Trigger 1 Control
Bit
No.
7:5
4:0
Name
Reset
R/W
Description
–
DMA1
0x0
0x0
R
R/W
Reserved
Condition for generating a pulse on radio DMA trigger 1 (DMA trigger 11).
See RFFDMA0 for the list of conditions.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
335
Registers
www.ti.com
RFFSTATUS (0x61C5) – FIFO Status
Bit
No.
7
Name
Reset
R/W
Description
TXAVAIL
0
R
6
TXFEMPTY
1
R
5
TXDTHEX
1
R
0: No readable data in Tx FIFO
1: Readable data present in Tx FIFO
0: Data present in Tx FIFO
1: Tx FIFO is empty
0: There is less data in Tx FIFO than the threshold amount given by
RFTXFTHRS.
1: There is more than or equal amount of data in Tx FIFO than the
threshold amount given by RFTXFTHRS
4
TXFFULL
0
R
3
RXAVAIL
0
R
2
RXFEMPTY
1
R
1
RXDTHEX
1
R
0
RXFFULL
0
R
0: Rx FIFO has available space
1: Rx FIFO is full
Name
Reset
R/W
Description
–
TXAUTOCOMMIT
0
1
R
R/W
Reserved
0: Commit Tx FIFO only on command 0x95
1: Always set RFTXSWP = RFTXWP
4
TXFAUTODEALLOC
0
R/W
0: Deallocate Tx FIFO only on command 0x92
1: Always set RFTXFSRP = RFTXFRP.
3:2
1
–
RXAUTOCOMMIT
0
0
R
R/W
Reserved
0: Commit Rx FIFO only on command 0x85
1: Always set RFRXSWP = RFRXWP
0
RXFAUTODEALLOC
1
R/W
0: Deallocate Rx FIFO only on command 0x82
1: Always set RFRXFSRP = RFRXFRP.
Name
Reset
R/W
Description
D
0x00
R
Amount of data present in Rx FIFO
0: Tx FIFO has available space
1: Tx FIFO is full
0: No readable data in Rx FIFO
1: Readable data present in Rx FIFO
0: Data present in Rx FIFO
1: Rx FIFO is empty
0: There is less data in Rx FIFO than the threshold amount given by
RFRXFTHRS.
1: There is more than or equal amount of data in Rx FIFO than the
threshold amount given by RFRXFTHRS
RFFCFG (0x61C6) – FIFO Configuration
Bit
No.
7:6
5
RFRXFLEN (0x61C8) – Rx FIFO Length
Bit
No.
7:0
RFRXFTHRS (0x61C9) – Rx FIFO Threshold
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Threshold value for Rx FIFO
RFRXFWR (0x61CA) – Rx FIFO Write Register
Bit
No.
7:0
336
Name
Reset
R/W
Description
D
0x00
W
Data written to this register is written to the Rx FIFO address at offset
RFRXFWP from the start of the Rx FIFO area (see Figure 25-1).
RFRXFWP (and RFRXFSWP if RFFCFG.RXAUTODEALLOC = 1) is
incremented by 1 modulo 0x80 unless the write fails.
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
RFRXFRD (0x61CB) – Rx FIFO Read Register
Bit
No.
7:0
Name
Reset
R/W
Description
D
0x00
R
When this register is read, the data in Rx FIFO address offset RFRXFRP
from the start of the Rx FIFO area is returned (see Figure 25-1).
RFRXFRP (and RFRXFSRP if RFFCFG.RXAUTOCOMMIT = 1) is
incremented by 1 modulo 0x80 unless the read fails.
RFRXFWP (0x61CC) – Rx FIFO Write Pointer
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Rx FIFO write pointer. This is the offset into Rx FIFO the next write
operation writes to.
RFRXFRP (0x61CD) – Rx FIFO Read Pointer
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Rx FIFO read pointer. This is the offset into Rx FIFO the next read
operation reads from.
RFRXFSWP (0x61CE) – Rx FIFO Start-of-Frame Write Pointer
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Rx FIFO start of written package. This is the point to which the write
pointer can be reset if a discard command is issued.
RFRXFSRP (0x61CF) – Rx FIFO Start-of-Frame Read Pointer
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Rx FIFO start of read package. This is the start of the allocated part of the
Rx FIFO.
Name
Reset
R/W
Description
D
0x00
R
Amount of data present in Tx FIFO
RFTXFLEN (0x61D0) – Tx FIFO Length
Bit
No.
7:0
RFTXFTHRS (0x61D1) – Tx FIFO Threshold
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Threshold value for Tx FIFO
RFTXFWR (0x61D2) – Tx FIFO Write Register
Bit
No.
7:0
Name
Reset
R/W
Description
D
0x00
W
Data written to this register is written to the Tx FIFO address at offset
RFTXFWP from the start of the Tx FIFO area (see Figure 25-1) is
returned. RFTXFWP (and RFTXFSWP if
RFFCFG.TXAUTODEALLOC = 1) is incremented by 1 modulo 0x80
unless the write fails.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
337
Registers
www.ti.com
RFTXFRD (0x61D3) – Tx FIFO Read Register
Bit
No.
7:0
Name
Reset
R/W
Description
D
0x00
R
When this register is read, the data in Tx FIFO address offset RFTXFRP
from the start of the Tx FIFO area is returned (see Figure 25-1).
RFTXFRP (and RFTXFSRP if RFFCFG.TXAUTOCOMMIT = 1) is
incremented by 1 modulo 0x80 unless the read fails.
RFTXFWP (0x61D4) – Tx FIFO Write Pointer
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Tx FIFO write pointer. This is the offset into Tx FIFO the next write
operation writes to.
RFTXFRP (0x61D5) – Tx FIFO Read Pointer
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Tx FIFO read pointer. This is the offset into Tx FIFO the next read
operation reads from.
RFTXFSWP (0x61D6) – Tx FIFO Start-of-Frame Write Pointer
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
000
0000
R
R/W
Reserved
Tx FIFO start of written package. This is the point to which the write
pointer can be reset if a discard command is issued.
RFTXFSRP (0x61D7) – Tx FIFO Start-of-Frame Read Pointer
Bit
No.
7
6:0
Name
Reset
R/W
Description
–
D
0
0x00
R
R/W
Reserved
Tx FIFO start-of-read package. This is the start of the allocated part of the
Tx FIFO.
BSP_P0 (0x61E0) – CRC Polynomial Byte 0
Bit
No.
7:0
Name
Reset
R/W
Description
P[7:0]
0x00
R/W
Bits 7:0 of p register in CRC sub-module
BSP_P1 (0x61E1) – CRC Polynomial Byte 1
Bit
No.
7:0
Name
Reset
R/W
Description
P[15:8]
0x5B
R/W
Bits 15:8 of p register in CRC sub-module
BSP_P2 (0x61E2) – CRC Polynomial Byte 2
Bit
No.
7:0
Name
Reset
R/W
Description
P[23:16]
0x06
R/W
Bits 23:16 of p register in CRC sub-module
BSP_P3 (0x61E3) – CRC Polynomial Byte 3
Bit
No.
7:0
338
Name
Reset
R/W
Description
P[31:24]
0x00
R/W
Bits 31:24 of p register in CRC sub-module
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Registers
www.ti.com
BSP_D0 (0x61E4) – CRC Value Byte 0
Bit
No.
7:0
Name
Reset
R/W
Description
D[7:0]
0x00
R/W
Bits 7:0 of d register in CRC sub-module
Name
Reset
R/W
Description
D[15:8]
0x5B
R/W
Bits 15:8 of d register in CRC sub-module
Name
Reset
R/W
Description
D[23:16]
0x06
R/W
Bits 23:16 of d register in CRC sub-module
Name
Reset
R/W
Description
D[31:24]
0x00
R/W
Bits 31:24 of d register in CRC sub-module
Name
Reset
R/W
Description
W_PN9_RESET
0
R0
When a 1 is written to this bit, the CC2500-compatible whitener is reset,
and all bits in the s and b registers are set to 1.
W[6:0]
110
0101
R/W
Write: Writes all whitening registers. w6 is set to BSP_W[0], w5 is set to
BSP_W[1] and so on up to w1 is set to BSP_W[5]. w0 is set to 1.
Read: Reads back w register. BSP_W[0] is set to w6, BSP_W[1] is
set to w5 and so on up to BSP_W[6] is set to w0.
BSP_D1 (0x61E5) – CRC Value Byte 1
Bit
No.
7:0
BSP_D2 (0x61E6) – CRC Value Byte 2
Bit
No.
7:0
BSP_D3 (0x61E7) – CRC Value Byte 3
Bit
No.
7:0
BSP_W (0x61E8) – Whitener Value
Bit
No.
7
6:0
BSP_MODE (0x61E9) – Bit Stream Processor Configuration
Bit
No.
7
6
Name
Reset
R/W
Description
–
CP_BUSY
0
0
R0
R
Reserved. Read as zero
Coprocessor mode busy. Goes to 1 after a byte has been written to
BSP_DATA. Goes to 0 when a byte is ready to be read back from
5
4
CP_READOUT
CP_END
0
0
R/W
R/W
3:2
CP_MODE[1:0]
00
R/W
1
0
W_PN9_EN
W_PN7_EN
0
1
R/W
R/W
BSP_DATA
Coprocessor mode readout
Endianness of data in coprosessor mode.
0: LSB processed first
1: MSB processed first
Coprocessor mode
00: Coprocessor disabled
01: Coprocessor receive mode
10: Reserved
11: Coprocessor transmit mode
Enable CC2500-compatible PN9 whitener
Enable PN7 whitener
BSP_DATA (0x61EA) – Bit Stream Processor Coprocessor Data
Bit
No.
7:0
Name
Reset
R/W
Description
BSP_DATA[7:0]
0x00
R/W
When BSP_MODE.CP_BUSY = 0:
Write: Provide byte to be processed in coprocessor mode
Read: Read processed byte
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
CC2541 Proprietary Mode Radio
339
Registers
www.ti.com
DC_I_L (0x61FC) – In-Phase DC Offset Estimate, Low Byte
Bit
No.
7:0
Name
Reset
R/W
Description
DC_I[7:0]
0x00
R*/W
When running dc estimation, this register reflects the 8 LSBs of the dc
estimate in the I channel. When manual dc override is selected, the
override value is written to this register.
DC_I_H (0x61FD) – In-Phase DC Offset Estimate, High Byte
Bit
No.
7:0
Name
Reset
R/W
Description
DC_I[15:8]
0x00
R*/W
When running dc estimation, this register reflects the 8 MSBs of the dc
estimate in the I channel. When manual dc override is selected, the
override value is written to this register.
DC_Q_L (0x61FE) – Quadrature-Phase DC Offset Estimate Low Byte
Bit
No.
7:0
Name
Reset
R/W
Description
DC_Q[7:0]
0x00
R*/W
When running dc estimation. this register reflects the 8 LSBs of the dc
estimate in the Q channel. When manual dc override is selected, the
override value is written to this register.
DC_Q_H (0x61FF) – Quadrature-Phase DC Offset Estimate High Byte
Bit
No.
7:0
Name
Reset
R/W
Description
DC_Q[15:8]
0x00
R*/W
When running dc estimation, this register reflects the 8 MSBs of the dc
estimate in the Q channel. When manual dc override is selected, the
override value is written to this register.
IVCTRL (0x6265) – Analog Control Register
Bit
No.
7:6
5:4
Name
Reset
R/W
Description
–
TX_MIX_LOAD
00
01
R0
R/W
3
LODIV_BIAS_CTRL
0
R/W
2
1:0
TXMIX_DC_CTRL
PA_BIAS_CTRL
0
11
R/W
R/W
Reserved
Controls load capacitor in Tx mixer
00: Minimum load
... (Intermediate loads)
11: Maximum load
Controls bias current to LODIV
0: IVREF bias
1: PTAT bias
Controls dc bias in TXMIX
Controls bias current to PA
00: IREF bias
01: IREF and IVREF bias
10: PTAT bias
11: Increased PTAT slope bias
340
CC2541 Proprietary Mode Radio
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Chapter 26
SWRU191D – April 2009 – Revised March 2013
Voltage Regulator
The digital voltage regulator is used to power the digital core. The output of this regulator is available on
the DCOUPL pin and requires capacitive decoupling to function properly (see, e.g., the CC2530 reference
design).
The voltage regulator is disabled in power modes PM2 and PM3 (see Chapter 4). When the voltage
regulator is disabled, register and RAM contents are retained while the unregulated 2 V to 3.6 V power
supply is present
NOTE:
The voltage regulator should not be used to provide power to external circuits.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Voltage Regulator
341
Chapter 27
SWRU191D – April 2009 – Revised March 2013
Available Software
This chapter presents the various available software solutions relevant to the CC253x/CC2540/CC2541
family. They are all available free of charge on the TI Web site at www.ti.com/lprf when used with TI LPRF
devices.
As shown in Table 0-1 in the Preface, the members of the CC253x/CC2540/CC2541 family have different
flash/RAM sizes; hence, they are not equally well suited for the different software offerings in the sections
below. For example, a user designing a ZigBee device should use the CC2530F256 as the Z-Stack™
requires in most cases more than 128 KB of flash and needs the 8-KB RAM.
Topic
27.1
27.2
27.3
27.4
27.5
27.6
342
...........................................................................................................................
SmartRF™ Software for Evaluation (www.ti.com/smartrfstudio) ............................
RemoTI™ Network Protocol (www.ti.com/remoti) ................................................
SimpliciTI™ Network Protocol (www.ti.com/simpliciti) .........................................
TIMAC Software (www.ti.com/timac) ..................................................................
Z-Stack™ Software (www.ti.com/z-stack) ............................................................
BLE Stack Software .........................................................................................
Available Software
Page
343
343
344
344
345
345
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
SmartRF™ Software for Evaluation (www.ti.com/smartrfstudio)
www.ti.com
27.1
SmartRF™ Software for Evaluation (www.ti.com/smartrfstudio)
Texas Instruments’ SmartRF Studio can be used for radio performance and functionality evaluation and is
great for exploring and gaining knowledge about the RF-IC products. This software helps the designers of
radio systems to evaluate the RF-ICs easily at an early stage in the design process. It is especially useful
for generation of the configuration data and for finding optimized external component values.
SmartRF Studio software runs on Microsoft™ Windows™ 98, Windows 2000, Windows XP, Windows
Vista (32 bit) and Windows 7 (32 bit). SmartRF Studio software can be downloaded from the Texas
Instruments Web page: www.ti.com/smartrfstudio (http://www.ti.com/litv/zip/swrc046m).
Features
• Link tests. Send and receive packets between nodes
• Packet error-rate (PER) tests
• Communication with evaluation boards through the USB port or the parallel port
• Up to eight USB devices are supported on a single computer.
• Normal view with preferred register settings
• Register view with possibilities to read and write each individual register. Each register given with
detailed information
• Save/Open configuration data from file
• Save/Load register settings from file
• Export/Import register values from text file
• Exports register settings into a C-compatible software structure
27.2
RemoTI™ Network Protocol (www.ti.com/remoti)
Most existing remote controls use infrared technology to communicate commands to consumer electronics
devices. However, radio frequency (RF) remote controls enable non-line-of-sight operation and provide
more advanced features based on bidirectional RF communication.
ZigBee Radio Frequency for Consumer Electronics (RF4CE) is the result of a recent agreement between
the ZigBee Alliance and the RF4CE Consortium (http://www.zigbee.org/rf4ce) and has been designed to
be deployed in a wide range of remotely-controlled audio/visual consumer electronics products, such as
TVs and set-top boxes. ZigBee RF4CE promises you:
• Richer communication and increased reliability
• Enhanced features and flexibility
• Interoperability
• No line-of-sight barrier
The RemoTI network protocol is Texas Instruments’ implementation of the ZigBee RF4CE standard. It is a
complete solution offering hardware and software support for TI’s low-power RF product portfolio. With the
RemoTI network protocol we provide:
• The industry-leading RF4CE-compliant stack featuring the interoperable CERC profile support, a
simple API, easy-to-understand sample application code, full development kits and reference designs,
and much more.
• Operation on our best-in-class IEEE 802.15.4 compliant System-on-Chip, the CC2530, with excellent
RF co-existence and RF performance. The four flexible power modes include the lowest-currentconsumption power-down mode for long battery in life low-duty-cycle applications.
• Extensive worldwide support and tools to ensure that development of ZigBee RF4CE-based products
is simple, fast, and can be completed at minimal cost.
• The RemoTI network protocol is a Golden Unit platform; i.e., it is used for testing other
implementations of the ZigBee RF4CE standard for standard compliance.
For more information on TI’s RemoTI network protocol, see the Texas Instruments RemoTI network
protocol Web site www.ti.com/remoti.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Available Software
343
SimpliciTI™ Network Protocol (www.ti.com/simpliciti)
27.3
www.ti.com
SimpliciTI™ Network Protocol (www.ti.com/simpliciti)
The SimpliciTI network protocol is a low-power RF protocol (for sub-1 GHz, 2.4 GHz, and IEEE 802.15.4
RF ICs) targeting simple, small RF networks. This open-source software is an excellent start for building a
network with battery-operated devices using a TI low-power RF System-on-Chip (SoC). The SimpliciTI
network protocol was designed for easy implementation and deployment out-of-the-box on several TI RF
platforms. It provides several sample applications.
Key Applications
• Alarm and security: occupancy sensors, light sensors, carbon monoxide sensors, glass-breakage
detectors
• Smoke detectors
• Automatic meter reading: gas meters, water meters, e-meters
• Active RFID applications
Key Features
• Low power: A TI-proprietary low-power network protocol
• Flexible:
– Direct device-to-device communication
– Simple star with access point for store and forward to end device
– Range extenders to increase range to four hops
• Simple: uses a five-command API
• Low data rate and low duty cycle
• Ease of use
For more information about the SimpliciTI network protocol, see the Texas Instruments SimpliciTI network
protocol Web site www.ti.com/simpliciti.
27.4 TIMAC Software (www.ti.com/timac)
TIMAC software is an IEEE 802.15.4 medium-access-control software stack for TI’s IEEE 802.15.4
transceivers and System-on-Chips.
You can use TIMAC when you:
• Need a wireless point-to-point or point-to-multipoint solution; e.g., multiple sensors reporting directly to
a master
• Need a standardized wireless protocol
• Have battery-powered and/or mains-powered nodes
• Need support for acknowledgement and retransmission
• Have low data-rate requirements (around 100-kbps effective data rate)
Features
• Support for IEEE 802.15.4 standard
• Support for beacon-enabled and non-beaconing systems
• Multiple platforms
• Easy application development
The TIMAC software stack is certified to be compliant with the IEEE 802.15.4 standard. TIMAC software is
distributed as object code free of charge. There are no royalties for using TIMAC software.
For more information about TIMAC software, see the Texas Instruments TIMAC Web site
www.ti.com/timac.
344
Available Software
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Z-Stack™ Software (www.ti.com/z-stack)
www.ti.com
27.5
Z-Stack™ Software (www.ti.com/z-stack)
The Z-Stack software is TI’s ZigBee-compliant protocol stack for a growing portfolio of IEEE 802.15.4
products and platforms. The Z-Stack software stack is compliant with the ZigBee-2007 specification,
supporting both the ZigBee and ZigBee PRO features sets. The Z-Stack software includes implementation
of two ZigBee application profiles – SmartEnergy and HomeAutomation. Other application profiles can
easily be implemented by the user.
Z-Stack software notables include:
• A fully compliant ZigBee and ZigBee PRO feature set
• A range of sample applications including support for the ZigBee Smart Energy and ZigBee Home
Automation profiles
• Over-the-air download and serial boot loader support
• Can be used together with the RF front ends, CC2590 and CC2591, which support 10 dBm and 20
dBm output power, respectively, and improved receive sensitivity.
The Z-Stack software has been awarded the ZigBee Alliance's golden-unit status for both the ZigBee and
ZigBee PRO stack profiles and is used by ZigBee developers worldwide.
Z-Stack software is well suited for:
• Smart energy (AMI)
• Home automation
• Commercial building automation
• Medical, assisted living, or personal health and hospital care
• Monitoring and control applications
• Wireless sensor networks
• Alarm and security
• Asset tracking
• Applications that require interoperability
For more information about Z-Stack software, see the Texas Instruments Z-Stack software Web site
www.ti.com/z-stack.
27.6 BLE Stack Software
TI’s single-mode Bluetooth low energy stack has been certified according to the Bluetooth 4.0 low energy
specification. Key features:
• Supports all BLE roles
• Range of example applications
• Multi-role capabilities
For more information about TI's BLE stack software, visit Texas Instruments Bluetooth low energy stack
software Web site at www.ti.com/blestack.
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Available Software
345
Appendix A
SWRU191D – April 2009 – Revised March 2013
Abbreviations
Abbreviations used in this user's guide:
AAF
Anti-aliasing filter
ACK
Acknowledge
ADC
Analog-to-digital converter
AES
Advanced Encryption Standard
AGC
Automatic gain control
ARIB
Association of Radio Industries and Businesses
BCD
Binary-coded decimal
BER
Bit error rate
BLE
Bluetooth low-energy
BOD
Brownout detector
BOM
Bill of materials
BSP
Bit-stream process
CBC
Cipher block chaining
CBC-MAC
Cipher block chaining message authentication code
CCA
Clear channel assessment
CCM
Counter mode + CBC-MAC
CFB
Cipher feedback
CFR
Code of Federal Regulations
CMRR
Common-mode rejection ratio
CPU
Central processing unit
CRC
Cyclic redundancy check
CSMA-CA
Carrier sense multiple access with collision avoidance
CSP
CSMA/CA strobe processor
CTR
Counter mode (encryption)
CW
Continuous wave
DAC
Digital-to-analog converter
DC
Direct current
DMA
Direct memory access
DSM
Delta-sigma modulator
DSSS
Direct-sequence spread spectrum
ECB
Electronic code book (encryption)
EM
Evaluation module
ENOB
Effective number of bits
ETSI
European Telecommunications Standards Institute
EVM
Error vector magnitude
FCC
Federal Communications Commission
FCF
Frame control field
FCS
Frame check sequence
FFCTRL
FIFO and frame control
346 Abbreviations
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Appendix A
www.ti.com
FIFO
First in, first out
FS
Full scale
GPIO
General-purpose input/output
HF
High frequency
HSSD
High-speed serial data
I/O
Input/output
I/Q
In-phase/quadrature-phase
IEEE
Institute of Electrical and Electronics Engineers
IF
Intermediate frequency
IOC
I/O controller
IRQ
Interrupt request
IR
Infrared
ISM
Industrial, scientific and medical
ITU-T
International Telecommunication Union – Telecommunication
IV
Initialization vector
KB
1024 bytes
kbps
Kilobits per second
LFSR
Linear feedback shift register
LLE
Link-layer engine
LNA
Low-noise amplifier
LO
Local oscillator
LQI
Link quality indication
LSB
Least-significant bit/byte
MAC
Medium access control
MAC
Message authentication code
MCU
Microcontroller unit
MFR
MAC footer
MHR
MAC header
MIC
Message integrity code
MISO
Master in, slave out
MOSI
Master out, slave in
MPDU
MAC protocol data unit
MSB
Most-significant bit/byte
MSDU
MAC service data unit
MUX
Multiplexer
NA
Not applicable/available
NC
Not connected
OFB
Output feedback (encryption)
O-QPSK
Offset – quadrature phase-shift keying
PA
Power amplifier
PC
Program counter
PCB
Printed circuit board
PER
Packet error rate
PHR
PHY header
PHY
Physical layer
PLL
Phase-locked loop
PM1, PM2,
PM3
Power mode 1, 2, and 3
PMC
Power management controller
PN7, PN9
7-bit or 9-bit pseudo-random sequence
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Abbreviations 347
Appendix A
348
www.ti.com
POR
Power-on reset
PSDU
PHY service data unit
PWM
Pulse-width modulator
RAM
Random access memory
RBW
Resolution bandwidth
RC
Resistor-capacitor
RCOSC
RC oscillator
RF
Radio frequency
RSSI
Receive signal strength indicator
RTC
Real-time clock
RX
Receive
SCK
Serial clock
SFD
Start of frame delimiter
SFR
Special function register
SHR
Synchronization header
SINAD
Signal-to-noise and distortion ratio
SPI
Serial peripheral interface
SRAM
Static random-access memory
ST
Sleep timer
T/R
Tape and reel
T/R
Transmit/receive
THD
Total harmonic distortion
TI
Texas Instruments
TX
Transmit
UART
Universal asynchronous receiver/transmitter
USART
Universal synchronous/asynchronous receiver/transmitter
VCO
Voltage-controlled oscillator
VGA
Variable-gain amplifier
WDT
Watchdog timer
XOSC
Crystal oscillator
Abbreviations
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Appendix B
SWRU191D – April 2009 – Revised March 2013
Additional Information
Texas Instruments offers a wide selection of cost-effective, low-power RF solutions for proprietary and
standard-based wireless applications for use in industrial and consumer applications. Our selection
includes RF transceivers, RF transmitters, RF front ends and System-on-Chips as well as various software
solutions for the sub-1 and 2.4-GHz frequency bands.
In addition, Texas Instruments provides a large selection of support collateral such as development tools,
technical documentation, reference designs, application expertise, customer support, third-party and
university programs.
The Low-Power RF E2E Online Community provides you with technical support forums, videos and blogs,
and the chance to interact with fellow engineers from all over the world.
With a broad selection of product solutions, end application possibilities, and the range of technical
support, Texas Instruments offers the broadest low-power RF portfolio. We make RF easy!
The following subsections point to where to find more information.
Topic
B.1
B.2
B.3
B.4
...........................................................................................................................
Texas Instruments Low-Power RF Web Site ........................................................
Low-Power RF Online Community .....................................................................
Texas Instruments Low-Power RF Developer Network .........................................
Low-Power RF eNewsletter ...............................................................................
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Additional Information
Page
350
350
350
350
349
Texas Instruments Low-Power RF Web Site
B.1
www.ti.com
Texas Instruments Low-Power RF Web Site
Texas Instruments’ Low-Power RF Web site has all our latest products, application and design notes, FAQ
section, news and events updates, and much more. Just go to www.ti.com/lprf.
B.2
Low-Power RF Online Community
•
•
•
Forums, videos, and blogs
RF design help
E2E interaction - Posting one's own and reading other users' questions
Join us today at www.ti.com/lprf-forum
B.3
Texas Instruments Low-Power RF Developer Network
Texas Instruments has launched an extensive network of low-power RF development partners to help
customers speed up their application development. The network consists of recommended companies, RF
consultants, and independent design houses that provide a series of hardware module products and
design services, including:
• RF circuit, low-power RF and ZigBee design services
• Low-power RF and ZigBee module solutions and development tools
• RF certification services and RF circuit manufacturing
Need help with modules, engineering services or development tools?
Search the Low-Power RF Developer Network to find a suitable partner! www.ti.com/lprfnetwork
B.4
Low-Power RF eNewsletter
The Low-Power RF eNewsletter keeps you up to date on new products, news releases, developers’ news,
and other news and events associated with low-power RF products from TI. The Low-Power RF
eNewsletter articles include links to get more online information.
Sign up today on www.ti.com/lprfnewsletter.
350
Additional Information
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
Appendix C
SWRU191D – April 2009 – Revised March 2013
References
References and other useful material:
1. IEEE Std. 802.15.4-2006: Wireless Medium Access Control (MAC) and Physical Layer (PHY)
specifications for Low Rate Wireless Personal Area Networks (LR-WPANs)
http://standards.ieee.org/getieee802/download/802.15.4-2006.pdf
2. CC2530 Data Sheet (SWRS081)
3. CC2531 Data Sheet (SWRS086)
4. CC2533 Data Sheet (SWRS087)
5. CC2540 Data Sheet (SWRS084)
6. CC2541 Data Sheet (SWRS110)
7. Bluetooth® Core Technical Specification document, version 4.0
https://www.bluetooth.org/technical/specifications/adopted.htm
8. Universal Serial Bus Revision 2.0 specification http://www.usb.org/developers/docs/usb_20_101111.zip
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
References
351
Revision History
www.ti.com
Revision History
Changes from C Revision (January 2012) to D Revision ............................................................................................... Page
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Deleted is intended for use in a laboratory test environment only. It fromFCC Warning ...................................... 15
Deleted You can also see the TI Knowledgebase for Analog & Mixed-Signal. ................................................. 15
Deleted OP-AMP connection to 12-bit (delta-Sigma) ADC ........................................................................ 18
Deleted OP-AMP connection to 12-bit (delta-Sigma) ADC ........................................................................ 19
Changed location of USB block, disconnecting it from the I/O controller and adding its connection to USB PHY ........ 19
Changed word wrapping in code. ..................................................................................................... 41
Changed TIMIF.OVFIM to TIMIF.T1OVFIM in Interrupt Overview image ....................................................... 43
Changed first and second paragraph of section 3.4.1 ............................................................................. 57
Added second row in table beginning with 126:FLASH_PAGES-1 in the first column ........................................ 57
Added sentence to first full paragraph in section 4.4.3 ............................................................................ 67
Added "This bit does not give an indication of the stability of the 32 kHz XOSC." to bit 7 description ...................... 68
Added The RESET_N input is always configured as an input with pull-up. ..................................................... 70
Added sentence to section 7.4, third paragraph .................................................................................... 80
Changed destination of pointer from "Yes" branch of "Repetitive Transfer Mode?" decision ................................ 94
Changed TIMIF.OVFIM to TIMIF.T1OVFIM ....................................................................................... 113
Changed TIMIF.OVFIM to TIMIF.T1OVFIM ....................................................................................... 114
Changed OVFIM to T1OVFIM ....................................................................................................... 127
Added paragraph before section 12.2.2 ............................................................................................ 134
Added phrase 'on the chip' to second full paragraph of Chapter 20 ............................................................ 168
Added paragraph after second full paragraph of Chapter 20 .................................................................... 168
Changed I2CSTAT.STAC to I2CSTAT in first column of Table 20-1 ........................................................... 172
Changed I2CSTAT.STAC to I2CSTAT in first column of Table 20-2 ........................................................... 173
Changed I2CSTAT.STAC to I2CSTAT in first column of Table 20-3 ........................................................... 175
Changed I2CSTAT.STAC to I2CSTAT in first column of Table 20-4 ........................................................... 176
Changed 01 to 11 in FSCAL reg. table under Reset on 1:0 bits row ........................................................... 267
Changed information to the description column for bits 5:0 in ATEST 0x61BD register table .............................. 271
Added R/W column in RFRND 0x61A7 register table bits 1 and 0 to R0 ...................................................... 272
Changed third sentence in second full paragraph of section 25.2 .............................................................. 279
Added note to Figure 25-7 ........................................................................................................... 297
Added sentence after third sentence of second paragraph of section 25.9. .................................................. 301
Added sentence immediately before table 25-12. ................................................................................. 302
Added sentence to second paragraph of section 25.9.2. ........................................................................ 303
Added sentence to eighth paragraph of section 25.9.2. .......................................................................... 303
Changed second to last sentence in last paragraph of section 25.9.2.4.1 and fixed section cross reference to 25.9.2.4.3 at
the end of the paragraph. ............................................................................................................ 311
Changed "transmitted" to "completely read out of the Tx FIFO" in second to last paragraph of section 25.9.2.4.2. ..... 312
Deleted "PRF_SEARCH_TIME and" from first sentence in section 25.9.2.6. ................................................. 315
Added sentence to end of first paragraph in section 25.11. ..................................................................... 318
Changed "Packet transmitted" to "Tx FIFO packet completed" in RFIRQF1 register table, bit 5 ........................... 322
Changed "Task ended" to "Tx FIFO packet completed" in RFIRQM1 register table, bit 6 .................................. 324
Changed "Packet transmitted" to "Tx FIFO packet completed" in RFIRQM1 register table, bit 5 .......................... 324
Added "Must be set to 0 before entering PM2 or PM3, otherwise the behavior of the RF core after waking up may be
unpredictable." to Bit 0 description of LLECTRL register table. ................................................................. 334
Deleted "The value is reset to 00 when coming up from PM2 or PM3." from the description column for Bit 2:1 of the
LLECTRL register table. .............................................................................................................. 334
NOTE: Page numbers for previous revisions may differ from page numbers in the current version.
352
Revision History
SWRU191D – April 2009 – Revised March 2013
Submit Documentation Feedback
Copyright © 2009–2013, Texas Instruments Incorporated
IMPORTANT NOTICE
Texas Instruments Incorporated and its subsidiaries (TI) reserve the right to make corrections, enhancements, improvements and other
changes to its semiconductor products and services per JESD46, latest issue, and to discontinue any product or service per JESD48, latest
issue. Buyers should obtain the latest relevant information before placing orders and should verify that such information is current and
complete. All semiconductor products (also referred to herein as “components”) are sold subject to TI’s terms and conditions of sale
supplied at the time of order acknowledgment.
TI warrants performance of its components to the specifications applicable at the time of sale, in accordance with the warranty in TI’s terms
and conditions of sale of semiconductor products. Testing and other quality control techniques are used to the extent TI deems necessary
to support this warranty. Except where mandated by applicable law, testing of all parameters of each component is not necessarily
performed.
TI assumes no liability for applications assistance or the design of Buyers’ products. Buyers are responsible for their products and
applications using TI components. To minimize the risks associated with Buyers’ products and applications, Buyers should provide
adequate design and operating safeguards.
TI does not warrant or represent that any license, either express or implied, is granted under any patent right, copyright, mask work right, or
other intellectual property right relating to any combination, machine, or process in which TI components or services are used. Information
published by TI regarding third-party products or services does not constitute a license to use such products or services or a warranty or
endorsement thereof. Use of such information may require a license from a third party under the patents or other intellectual property of the
third party, or a license from TI under the patents or other intellectual property of TI.
Reproduction of significant portions of TI information in TI data books or data sheets is permissible only if reproduction is without alteration
and is accompanied by all associated warranties, conditions, limitations, and notices. TI is not responsible or liable for such altered
documentation. Information of third parties may be subject to additional restrictions.
Resale of TI components or services with statements different from or beyond the parameters stated by TI for that component or service
voids all express and any implied warranties for the associated TI component or service and is an unfair and deceptive business practice.
TI is not responsible or liable for any such statements.
Buyer acknowledges and agrees that it is solely responsible for compliance with all legal, regulatory and safety-related requirements
concerning its products, and any use of TI components in its applications, notwithstanding any applications-related information or support
that may be provided by TI. Buyer represents and agrees that it has all the necessary expertise to create and implement safeguards which
anticipate dangerous consequences of failures, monitor failures and their consequences, lessen the likelihood of failures that might cause
harm and take appropriate remedial actions. Buyer will fully indemnify TI and its representatives against any damages arising out of the use
of any TI components in safety-critical applications.
In some cases, TI components may be promoted specifically to facilitate safety-related applications. With such components, TI’s goal is to
help enable customers to design and create their own end-product solutions that meet applicable functional safety standards and
requirements. Nonetheless, such components are subject to these terms.
No TI components are authorized for use in FDA Class III (or similar life-critical medical equipment) unless authorized officers of the parties
have executed a special agreement specifically governing such use.
Only those TI components which TI has specifically designated as military grade or “enhanced plastic” are designed and intended for use in
military/aerospace applications or environments. Buyer acknowledges and agrees that any military or aerospace use of TI components
which have not been so designated is solely at the Buyer's risk, and that Buyer is solely responsible for compliance with all legal and
regulatory requirements in connection with such use.
TI has specifically designated certain components as meeting ISO/TS16949 requirements, mainly for automotive use. In any case of use of
non-designated products, TI will not be responsible for any failure to meet ISO/TS16949.
Products
Applications
Audio
www.ti.com/audio
Automotive and Transportation
www.ti.com/automotive
Amplifiers
amplifier.ti.com
Communications and Telecom
www.ti.com/communications
Data Converters
dataconverter.ti.com
Computers and Peripherals
www.ti.com/computers
DLP® Products
www.dlp.com
Consumer Electronics
www.ti.com/consumer-apps
DSP
dsp.ti.com
Energy and Lighting
www.ti.com/energy
Clocks and Timers
www.ti.com/clocks
Industrial
www.ti.com/industrial
Interface
interface.ti.com
Medical
www.ti.com/medical
Logic
logic.ti.com
Security
www.ti.com/security
Power Mgmt
power.ti.com
Space, Avionics and Defense
www.ti.com/space-avionics-defense
Microcontrollers
microcontroller.ti.com
Video and Imaging
www.ti.com/video
RFID
www.ti-rfid.com
OMAP Applications Processors
www.ti.com/omap
TI E2E Community
e2e.ti.com
Wireless Connectivity
www.ti.com/wirelessconnectivity
Mailing Address: Texas Instruments, Post Office Box 655303, Dallas, Texas 75265
Copyright © 2013, Texas Instruments Incorporated