Renesas HD6417021 Superhâ ¢ risc engine Datasheet

To all our customers
Regarding the change of names mentioned in the document, such as Hitachi
Electric and Hitachi XX, to Renesas Technology Corp.
The semiconductor operations of Mitsubishi Electric and Hitachi were transferred to Renesas
Technology Corporation on April 1st 2003. These operations include microcomputer, logic, analog
and discrete devices, and memory chips other than DRAMs (flash memory, SRAMs etc.)
Accordingly, although Hitachi, Hitachi, Ltd., Hitachi Semiconductors, and other Hitachi brand
names are mentioned in the document, these names have in fact all been changed to Renesas
Technology Corp. Thank you for your understanding. Except for our corporate trademark, logo and
corporate statement, no changes whatsoever have been made to the contents of the document, and
these changes do not constitute any alteration to the contents of the document itself.
Renesas Technology Home Page: http://www.renesas.com
Renesas Technology Corp.
Customer Support Dept.
April 1, 2003
Cautions
Keep safety first in your circuit designs!
1.
Renesas Technology Corporation puts the maximum effort into making semiconductor products better and more reliable, but
there is always the possibility that trouble may occur with them. Trouble with semiconductors may lead to personal injury, fire
or property damage.
Remember to give due consideration to safety when making your circuit designs, with appropriate measures such as (i)
placement of substitutive, auxiliary circuits, (ii) use of nonflammable material or (iii) prevention against any malfunction or
mishap.
Notes regarding these materials
1.
These materials are intended as a reference to assist our customers in the selection of the Renesas Technology Corporation
product best suited to the customer's application; they do not convey any license under any intellectual property rights, or any
other rights, belonging to Renesas Technology Corporation or a third party.
2.
Renesas Technology Corporation assumes no responsibility for any damage, or infringement of any third-party's rights,
originating in the use of any product data, diagrams, charts, programs, algorithms, or circuit application examples contained in
these materials.
3.
All information contained in these materials, including product data, diagrams, charts, programs and algorithms represents
information on products at the time of publication of these materials, and are subject to change by Renesas Technology
Corporation without notice due to product improvements or other reasons. It is therefore recommended that customers contact
Renesas Technology Corporation or an authorized Renesas Technology Corporation product distributor for the latest product
information before purchasing a product listed herein.
The information described here may contain technical inaccuracies or typographical errors.
Renesas Technology Corporation assumes no responsibility for any damage, liability, or other loss rising from these
inaccuracies or errors.
Please also pay attention to information published by Renesas Technology Corporation by various means, including the
Renesas Technology Corporation Semiconductor home page (http://www.renesas.com).
4.
When using any or all of the information contained in these materials, including product data, diagrams, charts, programs, and
algorithms, please be sure to evaluate all information as a total system before making a final decision on the applicability of
the information and products. Renesas Technology Corporation assumes no responsibility for any damage, liability or other
loss resulting from the information contained herein.
5.
Renesas Technology Corporation semiconductors are not designed or manufactured for use in a device or system that is used
under circumstances in which human life is potentially at stake. Please contact Renesas Technology Corporation or an
authorized Renesas Technology Corporation product distributor when considering the use of a product contained herein for
any specific purposes, such as apparatus or systems for transportation, vehicular, medical, aerospace, nuclear, or undersea
repeater use.
6.
The prior written approval of Renesas Technology Corporation is necessary to reprint or reproduce in whole or in part these
materials.
7.
If these products or technologies are subject to the Japanese export control restrictions, they must be exported under a license
from the Japanese government and cannot be imported into a country other than the approved destination.
Any diversion or reexport contrary to the export control laws and regulations of Japan and/or the country of destination is
prohibited.
8.
Please contact Renesas Technology Corporation for further details on these materials or the products contained therein.
SuperH™ RISC engine
SH7020 and SH7021
HD6437020, HD6477021,
HD6437021, HD6417021
Hardware Manual
ADE-602-074B
Rev. 3.0
3/7/03
Hitachi, Ltd.
Cautions
1. Hitachi neither warrants nor grants licenses of any rights of Hitachi’s or any third party’s
patent, copyright, trademark, or other intellectual property rights for information contained in
this document. Hitachi bears no responsibility for problems that may arise with third party’s
rights, including intellectual property rights, in connection with use of the information
contained in this document.
2. Products and product specifications may be subject to change without notice. Confirm that you
have received the latest product standards or specifications before final design, purchase or
use.
3. Hitachi makes every attempt to ensure that its products are of high quality and reliability.
However, contact Hitachi’s sales office before using the product in an application that
demands especially high quality and reliability or where its failure or malfunction may directly
threaten human life or cause risk of bodily injury, such as aerospace, aeronautics, nuclear
power, combustion control, transportation, traffic, safety equipment or medical equipment for
life support.
4. Design your application so that the product is used within the ranges guaranteed by Hitachi
particularly for maximum rating, operating supply voltage range, heat radiation characteristics,
installation conditions and other characteristics. Hitachi bears no responsibility for failure or
damage when used beyond the guaranteed ranges. Even within the guaranteed ranges,
consider normally foreseeable failure rates or failure modes in semiconductor devices and
employ systemic measures such as fail-safes, so that the equipment incorporating Hitachi
product does not cause bodily injury, fire or other consequential damage due to operation of
the Hitachi product.
5. This product is not designed to be radiation resistant.
6. No one is permitted to reproduce or duplicate, in any form, the whole or part of this document
without written approval from Hitachi.
7. Contact Hitachi’s sales office for any questions regarding this document or Hitachi
semiconductor products.
Introduction
The SH7020 and SH7021 are part of a new generation of reduced instruction-set computer-type
(RISC) microcomputers that integrate RISC-type CPUs and the peripheral functions required for
system configuration onto a single chip to achieve high-performance operations processing. They
can operate in a power-down state, which is an essential feature for portable equipment.
The SH7020 and SH7021 CPUs have RISC-type instruction sets. Basic instructions can be
executed in a single clock cycle, which strikingly improves instruction execution speed. The
SH7020 and SH7021 include peripheral functions such as large-capacity ROM (PROM or masked
ROM), RAM, a direct memory access controller (DMAC), timers, a serial communication
interface (SCI), an interrupt controller (INTC), and I/O ports. These on-chip elements enable
users to construct systems with the fewest possible components. External memory access support
functions enable direct connection to SRAM and DRAM. without the use of glue logics.
This Hardware Manual describes in detail the hardware functions of the SH7020 and SH7021. For
information on the instructions, please refer to the Programming Manual.
Related Manuals
SH7000 Series Instructions
"SH-1/SH-2/SH-DSP Programming Manual"
For development support tools, contact your Hitachi sales office.
Organization of This Manual
Table 1 describes how this manual is organized. Figure 1 shows the relationships between the
Sections within this manual.
Table 1 Manual Organization
Category
Section Title
Abbreviation
Overview
1.
Overview
—
Features, internal block diagram, pin layout,
pin functions
CPU
2.
CPU
CPU
Register configuration, data structure.
instruction features, instruction types,
instruction lists
Operating
Modes
3.
Operating
Modes
—
MCU mode, PROM mode
Internal
Modules
4.
Exception
Processing
—
Resets, address errors, interrupts, trap
instructions, illegal instructions
5.
Interrupt
Controller
INTC
NMI interrupts, user break interrupts, IRQ
interrupts, on-chip module interrupts
6.
User Break
Controller
UBC
Break address and break bus cycles selection
Clock
7.
Clock Pulse
Generator
CPG
Crystal pulse generator, duty correction circuit
Buses
8.
Bus State
Controller
BSC
Division of memory space, DRAM interface,
refresh, wait state control, parity control
9.
Direct Memory
Access
Controller
DMAC
Auto request, external request, on-chip
peripheral module request, cycle steal mode,
burst mode
10. 16-Bit
IntegratedTimer Pulse
Unit
ITU
Waveform output mode, input capture
function, counter clear function, buffer
operation, PWM mode, complementary PWM
mode, reset synchronized mode, synchronized
operation, phase counting mode, compare
match output mode
11. Programmable
Timing Pattern
Controller
TPC
Compare match output triggers, non-overlap
operation
12. Watchdog
Timer
WDT
Watchdog timer mode, interval timer mode
13. Serial
Communication Interface
SCI
Asynchronous mode, clock synchronous
mode, multiprocessor communication function
Timers
Data
Processing
Contents
Table 1 Manual Organization (cont)
Abbreviation
Contents
14. Pin Function
Controller
PFC
Pin function selection
15. Parallel I/O
Ports
I/O
I/O port
16. ROM
ROM
On-chip ROM
17. RAM
RAM
On-chip RAM
Power-Down
States
18. Power-Down
States
—
Sleep mode, standby mode
Electrical
Characteristics
19. Electrical
Characteristics
—
Absolute maximum ratings, AC characteristics,
DC characteristics, operation timing
Category
Section Title
Pins
Memory
1. Overview
3. Operating modes
2. CPU
On-chip modules
4. Exception processing
5. Interrupt controller (INTC)
6. User break controller (UBC)
7. Clock pulse generator (CPG)
Timers
Buses
8. Bus state controller (BC)
10. 16-bit integrated-timer
pulse unit (ITU)
9. Direct memory access
controller (DMAC)
11. Programmable timing
pattern controller (TPC)
12. Watchdog timer (WDT)
Memory
Data processing
16. ROM
17. RAM
13. Serial communication
interface (SCI)
Pins
14. Pin function
controller (PFC)
15. Parallel I/O ports
18. Power-down states
19. Electrical characteristics
Manual Organization Scheme
Addresses of On-Chip Peripheral Module Registers
The on-chip peripheral module registers are located in the on-chip peripheral module space (area
5: H'5000000–H'5FFFFFF), but since the actual register space is only 512 bytes, address bits
A23–A9 are ignored. 32k shadow areas in 512 byte units that contain exactly the same contents as
the actual registers are thus provided in the on-chip peripheral module space.
In this manual, register addresses are specified as though the on-chip peripheral module registers
were in the 512 bytes H'5FFFE00–H'5FFFFFF. Only the values of the A27–A24 and A8–A0 bits
are valid; the A23–A9 bits are ignored. When H'5000000–H'50001FF is accessed, for example,
the result will be the same as when H'5FFFE00–H'5FFFFFF is accessed. For more details, see
Section 8.3.5, Area Description: Area 5.
Free Addresses in the On-chip Peripheral Module Space (Area 5)
Avoid reading/writing from/to the free addresses without registers in the on-chip peripheral
module space (area 5: H'5000000-H'5FFFFFF).
Contents
Section 1 Overview ............................................................................................................
1.1
1.2
1.3
SH Microcomputer Features..............................................................................................
Block Diagram...................................................................................................................
Pin Descriptions.................................................................................................................
1.3.1 Pin Arrangement ..................................................................................................
1.3.2 Pin Functions........................................................................................................
1.3.3 Pin Layout by Mode .............................................................................................
1
1
7
8
8
9
13
Section 2 CPU ...................................................................................................................... 15
2.1
2.2
2.3
2.4
2.5
Register Configuration ......................................................................................................
2.1.1 General Registers (Rn) .........................................................................................
2.1.2 Control Registers..................................................................................................
2.1.3 System Registers ..................................................................................................
2.1.4 Initial Values of Registers ....................................................................................
Data Formats......................................................................................................................
2.2.1 Data Format in Registers......................................................................................
2.2.2 Data Format in Memory .......................................................................................
2.2.3 Immediate Data Format........................................................................................
Instruction Features ..............................................................................................................
2.3.1 RISC-Type Instruction Set ...................................................................................
2.3.2 Addressing Modes................................................................................................
2.3.3 Instruction Formats ..............................................................................................
Instruction Set....................................................................................................................
2.4.1 Instruction Set by Classification ..........................................................................
2.4.2 Operation Code Map ............................................................................................
CPU State ..........................................................................................................................
2.5.1 State Transitions ...................................................................................................
2.5.2 Power-Down State................................................................................................
15
15
15
16
17
17
17
18
18
19
19
22
25
28
28
39
41
41
43
Section 3 Operating Modes .............................................................................................. 45
3.1
3.2
Types of Operating Modes and Their Selection................................................................
Operating Mode Descriptions............................................................................................
3.2.1 Mode 0 (MCU Mode 0)........................................................................................
3.2.2 Mode 1 (MCU Mode 1)........................................................................................
3.2.3 Mode 2 (MCU Mode 2)........................................................................................
3.2.4 Mode 7 (PROM Mode) ........................................................................................
45
45
45
45
45
45
Section 4 Exception Processing ..................................................................................... 47
4.1
Overview............................................................................................................................
47
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.1.1 Exception Processing Types and Priorities ..........................................................
4.1.2 Exception Processing Operation ..........................................................................
4.1.3 Exception Process Vector Table ..........................................................................
Reset ..................................................................................................................................
4.2.1 Reset Types ..........................................................................................................
4.2.2 Power-On Reset....................................................................................................
4.2.3 Manual Reset........................................................................................................
Address Errors ...................................................................................................................
4.3.1 Address Error Sources..........................................................................................
4.3.2 Address Error Exception Processing....................................................................
Interrupts............................................................................................................................
4.4.1 Interrupt Sources ..................................................................................................
4.4.2 Interrupt Priority Rankings...................................................................................
4.4.3 Interrupt Exception Processing ............................................................................
Instruction Exceptions .......................................................................................................
4.5.1 Types of Instruction Exceptions...........................................................................
4.5.2 Trap Instruction ....................................................................................................
4.5.3 Illegal Slot Instruction ..........................................................................................
4.5.4 General Illegal Instructions ..................................................................................
Cases in Which Exceptions Are Not Accepted .................................................................
4.6.1 Immediately after Delayed Branch Instructions...................................................
4.6.2 Immediately after Interrupt-Disabling Instructions..............................................
Stack Status after Exception Processing............................................................................
Notes..................................................................................................................................
4.8.1 Value of the Stack Pointer (SP)............................................................................
4.8.2 Value of the Vector Base Register (VBR) ...........................................................
4.8.3 Address Errors that Are Caused by Stacking During Address Error
Exception Processing............................................................................................
47
49
49
51
51
51
52
52
52
53
54
54
54
55
55
55
55
56
56
57
57
57
58
59
59
59
59
Section 5 Interrupt Controller (INTC) .......................................................................... 61
5.1
5.2
5.3
Overview............................................................................................................................ 61
5.1.1 Features ................................................................................................................ 61
5.1.2 Block Diagram...................................................................................................... 61
5.1.3 Pin Configuration ................................................................................................. 63
5.1.4 Registers ............................................................................................................... 63
Interrupt Sources................................................................................................................ 63
5.2.1 NMI Interrupts...................................................................................................... 64
5.2.2 User Break Interrupt ............................................................................................. 64
5.2.3 IRQ Interrupts ...................................................................................................... 64
5.2.4 On-Chip Interrupts................................................................................................ 64
5.2.5 Interrupt Exception Vectors and Priority Rankings ............................................. 65
Register Descriptions......................................................................................................... 68
5.3.1 Interrupt Priority Registers A–E (IPRA–IPRE) ................................................... 68
5.4
5.5
5.5
5.3.2 Interrupt Control Register (ICR) .......................................................................... 69
Interrupt Operation ............................................................................................................ 70
5.4.1 Interrupt Sequence................................................................................................ 70
5.4.2 Stack after Interrupt Exception Processing .......................................................... 72
Interrupt Response Time.................................................................................................... 73
Usage Notes ....................................................................................................................... 74
Section 6 User Break Controller (UBC) ...................................................................... 75
6.1
6.2
6.3
6.4
6.5
Overview............................................................................................................................
6.1.1 Features ................................................................................................................
6.1.2 Block Diagram......................................................................................................
6.1.3 Register Configuration .........................................................................................
Register Descriptions.........................................................................................................
6.2.1 Break Address Registers (BAR) ..........................................................................
6.2.2 Break Address Mask Register (BAMR)...............................................................
6.2.3 Break Bus Cycle Register (BBR) .........................................................................
Operation ...........................................................................................................................
6.3.1 Flow of the User Break Operation........................................................................
6.3.2 Break on Instruction Fetch Cycles to On-Chip Memory......................................
6.3.3 Program Counter (PC) Value Saved in User Break Interrupt Exception
Processing.............................................................................................................
Setting User Break Conditions ..........................................................................................
Notes..................................................................................................................................
6.5.1 On-Chip Memory Instruction Fetch .....................................................................
6.5.2 Instruction Fetch at Branches ...............................................................................
6.5.3 Instruction Fetch Break ........................................................................................
75
75
75
76
77
77
78
79
81
81
84
84
84
86
86
86
87
Section 7 Clock Pulse Generator (CPG) ..................................................................... 89
7.1
7.2
7.3
Overview............................................................................................................................ 89
Clock Source...................................................................................................................... 89
7.2.1 Connecting a Crystal Resonator ........................................................................... 89
7.2.2 External Clock Input ............................................................................................ 90
Usage Notes ....................................................................................................................... 91
Section 8 Bus State Controller (BSC) ........................................................................... 93
8.1
8.2
Overview............................................................................................................................ 93
8.1.1 Features ................................................................................................................ 93
8.1.2 Block Diagram...................................................................................................... 93
8.1.3 Pin Configuration ................................................................................................. 95
8.1.4 Register Configuration ......................................................................................... 95
8.1.5 Overview of Areas................................................................................................ 96
Register Descriptions......................................................................................................... 97
8.2.1 Bus Control Register (BCR) ................................................................................ 97
8.2.2 Wait State Control Register 1 (WCR1)................................................................
8.2.3 Wait State Control Register 2 (WCR2)................................................................
8.2.4 Wait State Control Register 3 (WCR3)................................................................
8.2.5 DRAM Area Control Register (DCR)..................................................................
8.2.6 Refresh Control Register (RCR) ..........................................................................
8.2.7 Refresh Timer Control/Status Register (RTCSR) ................................................
8.2.8 Refresh Timer Counter (RTCNT) ........................................................................
8.2.9 Refresh Time Constant Register (RTCOR)..........................................................
8.2.10 Parity Control Register (PCR)..............................................................................
8.2.11 Notes on Register Access .....................................................................................
8.3 Address Space Subdivision................................................................................................
8.3.1 Address Spaces and Areas....................................................................................
8.3.2 Bus Width.............................................................................................................
8.3.3 Chip Select Signals (CS0–CS7)............................................................................
8.3.4 Shadows................................................................................................................
8.3.5 Area Description ..................................................................................................
8.4 Accessing External Memory Space ...................................................................................
8.4.1 Basic Timing ........................................................................................................
8.4.2 Wait State Control ................................................................................................
8.4.3 Byte Access Control .............................................................................................
8.5 DRAM Interface Operation ...............................................................................................
8.5.1 DRAM Adress Multiplexing ...............................................................................
8.5.2 Basic Timing ........................................................................................................
8.5.3 Wait State Control ................................................................................................
8.5.4 Byte Access Control .............................................................................................
8.5.5 DRAM Burst Mode ..............................................................................................
8.5.6 Refresh Control ....................................................................................................
8.6 Address/Data Multiplexed I/O Space Access....................................................................
8.6.1 Basic Timing ........................................................................................................
8.6.2 Wait State Control ................................................................................................
8.6.3 Byte Access Control .............................................................................................
8.7 Parity Check and Generation .............................................................................................
8.8 Warp Mode........................................................................................................................
8.9 Wait State Control .............................................................................................................
8.10 Bus Arbitration ..................................................................................................................
8.10.1 The Operation of Bus Arbitration ........................................................................
8.10.2 BACK Operation..................................................................................................
8.11 Usage Notes .......................................................................................................................
8.11.1 Usage Notes on Manual Reset..............................................................................
8.11.2 Usage Notes on Parity Data Pins DPH and DPL .................................................
8.11.3 Maximum Number of States from BREQ Input to Bus Release .........................
98
101
103
104
107
108
110
112
112
114
115
115
117
117
118
120
128
128
129
133
134
134
136
138
140
142
148
151
152
153
153
154
154
155
157
158
159
161
161
164
164
Section 9 Direct Memory Access Controller (DMAC) ......................................... 169
9.1
9.2
9.3
9.4
9.5
Overview............................................................................................................................
9.1.1 Features ................................................................................................................
9.1.2 Block Diagram......................................................................................................
9.1.3 Pin Configuration .................................................................................................
9.1.4 Register Configuration .........................................................................................
Register Descriptions.........................................................................................................
9.2.1 DMA Source Address Registers 0–3 (SAR0–SAR3) ..........................................
9.2.2 DMA Destination Address Registers 0-3 (DAR0–DAR3) ..................................
9.2.3 DMA Transfer Count Registers 0–3 (TCR0–TCR3) ...........................................
9.2.4 DMA Channel Control Registers 0–3 (CHCR0–CHCR3)...................................
9.2.5 DMA Operation Register (DMAOR)...................................................................
Operation ...........................................................................................................................
9.3.1 DMA Transfer Flow .............................................................................................
9.3.2 DMA Transfer Requests.......................................................................................
9.3.3 Channel Priority....................................................................................................
9.3.4 DMA Transfer Types ...........................................................................................
9.3.5 Number of Bus Cycle States and DREQ Pin Sample Timing..............................
9.3.6 DMA Transfer Ending Conditions .......................................................................
Examples of Use................................................................................................................
9.4.1 DMA Transfer between On-Chip RAM and a Memory-Mapped
External Device ....................................................................................................
9.4.2 Example of DMA Transfer between On-Chip SCI and External Memory..........
Cautions .............................................................................................................................
169
169
170
172
173
174
174
174
175
175
180
182
182
184
186
191
198
205
206
206
207
208
Section 10 16-Bit Integrated-Timer Pulse Unit (ITU) ............................................ 213
10.1 Overview............................................................................................................................
10.1.1 Features ................................................................................................................
10.1.2 Block Diagram......................................................................................................
10.1.3 Input/Output Pins..................................................................................................
10.1.4 Register Configuration .........................................................................................
10.2 ITU Register Descriptions .................................................................................................
10.2.1 Timer Start Register (TSTR)................................................................................
10.2.2 Timer Synchro Register (TSNC)..........................................................................
10.2.3 Timer Mode Register (TMDR) ............................................................................
10.2.4 Timer Function Control Register (TFCR)............................................................
10.2.5 Timer Output Control Register (TOCR) ..............................................................
10.2.6 Timer Counters (TCNT).......................................................................................
10.2.7 General Registers A and B (GRA and GRB) .......................................................
10.2.8 Buffer Registers A and B (BRA, BRB)................................................................
10.2.9 Timer Control Register (TCR) .............................................................................
10.2.10 Timer I/O Control Register (TIOR) .....................................................................
10.2.11 Timer Status Register (TSR) ................................................................................
10.2.12 Timer Interrupt Enable Register (TIER) ..............................................................
213
213
216
221
222
224
224
226
227
230
231
232
233
234
235
237
239
240
10.3 CPU Interface ....................................................................................................................
10.3.1 16-Bit Accessible Registers..................................................................................
10.3.2 8-Bit Accessible Registers....................................................................................
10.4 Description of Operation ...................................................................................................
10.4.1 Overview ..............................................................................................................
10.4.2 Basic Functions ....................................................................................................
10.4.3 Synchronizing Mode ............................................................................................
10.4.4 PWM Mode ..........................................................................................................
10.4.5 Reset-Synchronized PWM Mode .........................................................................
10.4.6 Complementary PWM Mode................................................................................
10.4.7 Phase Counting Mode ..........................................................................................
10.4.8 Buffer Mode .........................................................................................................
10.4.9 ITU Output Timing ..............................................................................................
10.5 Interrupts............................................................................................................................
10.5.1 Timing of Setting Status Flags .............................................................................
10.5.2 Clear Timing of Status Flags................................................................................
10.5.3 Interrupt Sources and Activating the DMAC.......................................................
10.6 Notes and Precautions........................................................................................................
10.6.1 Contention between TCNT Write and Clear........................................................
10.6.2 Contention between TCNT Word Write and Increment ......................................
10.6.3 Contention between TCNT Byte Write and Increment........................................
10.6.4 Contention between GR Write and Compare Match............................................
10.6.5 Contention between TCNT Write and Overflow/Underflow ...............................
10.6.6 Contention between General Register Read and Input Capture ...........................
10.6.7 Contention Between Counter Clearing by Input Capture and Counter
Increment..............................................................................................................
10.6.8 Contention between General Register Write and Input Capture..........................
10.6.9 Note on Waveform Cycle Setting ........................................................................
10.6.10 Contention Between BR Write and Input Capture ...............................................
10.6.11 Note on Writing in the Synchronizing Mode .......................................................
10.6.12 Note on Setting Reset-synchronized PWM Mode/Complementary
PWM Mode ..........................................................................................................
10.6.13 Clearing the Complementary PWM Mode...........................................................
10.6.14 ITU Operating Modes ..........................................................................................
241
241
243
244
244
245
256
258
262
264
272
274
280
281
281
283
284
285
285
286
287
288
289
290
291
292
292
293
294
294
295
295
Section 11 Programmable Timing Pattern Controller (TPC) ............................... 303
11.1 Overview............................................................................................................................
11.1.1 Features ................................................................................................................
11.1.2 Block Diagram......................................................................................................
11.1.3 Input/Output Pins..................................................................................................
11.1.4 Registers ...............................................................................................................
11.2 Register Descriptions.........................................................................................................
11.2.1 Port B Control Registers 1 and 2 (PBCR1, PCBR2)............................................
303
303
304
305
306
306
306
11.2.2 Port B Data Register (PBDR)...............................................................................
11.2.3 Next Data Register A (NDRA) ............................................................................
11.2.4 Next Data Register B (NDRB) .............................................................................
11.2.5 Next Data Enable Register A (NDERA)..............................................................
11.2.6 Next Data Enable Register B (NDERB) ..............................................................
11.2.7 TPC Output Control Register (TPCR) .................................................................
11.2.8 TPC Output Mode Register (TPMR) ...................................................................
11.3 Operation ...........................................................................................................................
11.3.1 Overview ..............................................................................................................
11.3.2 Output Timing ......................................................................................................
11.3.3 Examples of Use of Ordinary TPC Output ..........................................................
11.3.4 TPC Output Non-Overlap Operation....................................................................
11.3.5 TPC Output by Input Capture ..............................................................................
11.4 Usage Notes .......................................................................................................................
11.4.1 Non-Overlap Operation........................................................................................
307
308
310
311
312
313
314
316
316
317
317
320
324
325
325
Section 12 Watchdog Timer (WDT) ............................................................................ 327
12.1 Overview............................................................................................................................
12.1.1 Features ................................................................................................................
12.1.2 Block Diagram......................................................................................................
12.1.3 Pin Configuration .................................................................................................
12.1.4 Register Configuration .........................................................................................
12.2 Register Descriptions.........................................................................................................
12.2.1 Timer Counter (TCNT) ........................................................................................
12.2.2 Timer Control/Status Register (TCSR) ................................................................
12.2.3 Reset Control/Status Register (RSTCSR) ............................................................
12.2.4 Register Access ....................................................................................................
12.3 Operation ...........................................................................................................................
12.3.1 Operation in the Watchdog Timer Mode..............................................................
12.3.2 Operation in the Interval Timer Mode..................................................................
12.3.3 Operation in the Standby Mode............................................................................
12.3.4 Timing of Setting the Overflow Flag (OVF)........................................................
12.3.5 Timing of Setting the Watchdog Timer Overflow Flag (WOVF)........................
12.4 Usage Notes .......................................................................................................................
12.4.1 TCNT Write and Count Up Contention ...............................................................
12.4.2 Changing CKS2-CKS0 Bit Values ......................................................................
12.4.3 Changing Watchdog Timer/Interval Timer Modes ..............................................
12.4.4 System Reset With WDTOVF .............................................................................
12.4.5 Internal Reset With the Watchdog Timer ............................................................
327
327
328
328
329
329
329
330
331
333
334
334
336
336
337
337
338
338
338
338
339
339
Section 13 Serial Communication Interface (SCI) .................................................. 341
13.1 Overview............................................................................................................................ 341
13.1.1 Features ................................................................................................................ 341
13.2
13.3
13.4
13.5
13.1.2 Block Diagram......................................................................................................
13.1.3 Input/Output Pins..................................................................................................
13.1.4 Register Configuration .........................................................................................
Register Descriptions.........................................................................................................
13.2.1 Receive Shift Register ..........................................................................................
13.2.2 Receive Data Register ..........................................................................................
13.2.3 Transmit Shift Register ........................................................................................
13.2.4 Transmit Data Register.........................................................................................
13.2.5 Serial Mode Register ............................................................................................
13.2.6 Serial Control Register .........................................................................................
13.2.7 Serial Status Register............................................................................................
13.2.8 Bit Rate Register (BRR).......................................................................................
Operation ...........................................................................................................................
13.3.1 Overview ..............................................................................................................
13.3.2 Operation in Asynchronous Mode........................................................................
13.3.3 Multiprocessor Communication ...........................................................................
13.3.4 Clocked Synchronous Operation..........................................................................
SCI Interrupt Sources and the DMAC...............................................................................
Usage Notes .......................................................................................................................
342
343
343
344
344
344
344
345
345
347
351
355
363
363
366
376
384
394
394
Section 14 Pin Function Controller (PFC) ................................................................... 399
14.1 Overview............................................................................................................................ 399
14.2 Register Configuration ...................................................................................................... 401
14.3 Register Descriptions......................................................................................................... 401
14.3.1 Port A I/O Register (PAIOR) ............................................................................... 401
14.3.2 Port A Control Registers (PACR1 and PACR2) .................................................. 402
14.3.3 Port B I/O Register (PBIOR)................................................................................ 407
14.3.4 Port B Control Registers (PBCR1 and PBCR2)................................................... 408
14.3.5 Column Address Strobe Pin Control Register (CASCR) ..................................... 413
Section 15 Parallel I/O Ports ............................................................................................ 415
15.1 Overview............................................................................................................................ 415
15.2 Port A................................................................................................................................. 415
15.2.1 Register Configuration ......................................................................................... 415
15.2.2 Port A Data Register (PADR) .............................................................................. 416
15.3 Port B ................................................................................................................................. 417
15.3.1 Register Configuration ......................................................................................... 417
15.3.2 Port B Data Register (PBDR)............................................................................... 418
Section 16 ROM .................................................................................................................. 419
16.1 Overview............................................................................................................................ 419
16.2 PROM Mode...................................................................................................................... 421
16.2.1 Setting the PROM Mode ...................................................................................... 421
16.2.2 Socket Adapter Pin Correspondence and Memory Map .....................................
16. 3 PROM Programming .........................................................................................................
16.3.1 Selecting the Programming Mode........................................................................
16.3.2 Write/Verify and Electrical Characteristics..........................................................
16.3.3 Points to Note About Writing...............................................................................
16.3.4 Reliability After Writing ......................................................................................
421
423
423
424
428
429
Section 17 RAM .................................................................................................................. 431
17.1 Overview............................................................................................................................ 431
17.2 Operation ........................................................................................................................... 431
Section 18 Power-Down States ...................................................................................... 433
18.1 Overview............................................................................................................................
18.1.1 Power-Down Modes.............................................................................................
18.1.2 Register.................................................................................................................
18.2 Standby Control Register (SBYCR)..................................................................................
18.3 Sleep Mode........................................................................................................................
18.3.1 Transition to the Sleep Mode ...............................................................................
18.3.2 Canceling the Sleep Mode....................................................................................
18.4 S tandby Mode ....................................................................................................................
18.4.1 Transition to the Standby Mode ...........................................................................
18.4.2 Canceling the Standby Mode................................................................................
18.4.3 Standby Mode Application...................................................................................
433
433
434
434
435
435
435
436
436
438
439
Section 19 Electrical Characteristics ............................................................................ 441
19.1 Absolute Maximum Ratings..............................................................................................
19.2 DC Characteristics .............................................................................................................
19.3 AC Characteristics .............................................................................................................
19.3.1 Clock Timing........................................................................................................
19.3.2 Control Signal Timing..........................................................................................
19.3.3 Bus Timing ...........................................................................................................
19.3.4 DMAC Timing .....................................................................................................
19.3.5 16-bit Integrated Timer Pulse Unit Timing..........................................................
19.3.6 Programmable Timing Pattern Controller and I/O Port Timing ..........................
19.3.7 Watchdog Timer Timing ......................................................................................
19.3.8 Serial Communications Interface Timing ............................................................
19.3.9 AC Characteristics Measurement Conditions ......................................................
19.4 Usage Note ........................................................................................................................
441
442
449
449
451
454
490
491
493
494
495
497
498
Appendix A On-Chip Peripheral Module Registers ................................................ 499
A.1
A.2
List of Registers................................................................................................................. 499
Register tables.................................................................................................................... 509
A.2.1 Serial Mode Register (SMR)................................................................................ 509
A.2.2
A.2.3
A.2.4
A.2.5
A.2.6
A.2.7
A.2.8
A.2.9
A.2.10
A.2.11
A.2.12
A.2.13
A.2.14
A.2.15
A.2.16
A.2.17
A.2.18
A.2.19
A.2.20
A.2.21
A.2.22
A.2.23
A.2.24
A.2.25
A.2.26
A.2.27
A.2.28
A.2.29
A.2.30
A.2.31
A.2.32
A.2.33
A.2.34
A.2.35
A.2.36
A.2.37
A.2.38
A.2.39
A.2.40
A.2.41
A.2.42
A.2.43
A.2.44
Bit Rate Register (BRR).......................................................................................
Serial Control Register (SCR)..............................................................................
Transmit Data Register (TDR) .............................................................................
Serial Status Register (SSR).................................................................................
Receive Data Register (RDR) ..............................................................................
Timer Start Register (TSTR)................................................................................
Timer Synchronization Register (TSNC).............................................................
Timer Mode Register (TMDR) ............................................................................
Timer Function Control Register (TFCR)............................................................
Timer Control Registers 0–4 (TCR0–TCR4) .......................................................
Timer I/O Control Registers 0–4 (TIO0–TIO4)...................................................
Timer Interrupt Enable Registers 0–4 (TIER0–TIER4).......................................
Timer Status Registers 0–4 (TSR0–TSR4) ..........................................................
Timer Counter 0–4 (TCNT0–TCNT4).................................................................
General Registers A0–4 (GRA0–GRA4) .............................................................
General Registers B0–4 (GRB0–GRB4)..............................................................
Buffer Registers A3, A4 (BRA3, BRA4).............................................................
Buffer registers B3, B4 (BRB3, BRB4) ...............................................................
Timer Output Control Register (TOCR) ..............................................................
DMA Source Address Registers 0–3 (SAR0–SAR3) ..........................................
DMA Destination Address Registers 0–3 (DAR0–DAR3)..................................
DMA Transfer Count Registers 0–3 (TCR0–TCR3) ...........................................
DMA Channel Control Registers 0–3 (CHCR0–CHCR3)...................................
DMA Operation Registers (DMAOR) .................................................................
Interrupt Priority Setting Register A (IPRA)........................................................
Interrupt Priority Setting Register B (IPRB)........................................................
Interrupt Priority Setting Register C (IPRC)........................................................
Interrupt Priority Setting Register D (IPRD)........................................................
Interrupt Priority Setting Register E (IPRE) ........................................................
Interrupt Control Register (ICR) ..........................................................................
Break Address Register H (BARH) .....................................................................
Break Address Register L (BARL) ......................................................................
Break Address Mask Register H (BAMRH)........................................................
Break Address Mask Register L (BAMRL).........................................................
Break Bus Cycle Register (BBR) .........................................................................
Bus Control Register (BCR) ................................................................................
Wait State Control Register 1 (WCR1)................................................................
Wait State Control Register 2 (WCR2)................................................................
Wait State Control Register 3 (WCR3)................................................................
DRAM Area Control Register (DCR)..................................................................
Parity Control Register (PCR)..............................................................................
Refresh Control Register (RCR) ..........................................................................
Refresh Timer Control/Status Register (RSTCR) ................................................
510
510
512
512
514
515
515
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
535
536
537
538
539
540
541
542
543
544
545
546
548
549
550
552
553
555
556
557
A.2.45
A.2.46
A.2.47
A.2.48
A.2.49
A.2.50
A.2.51
A.2.52
A.2.53
A.2.54
A.2.55
A.2.56
A.2.57
A.2.58
A.2.59
A.2.60
A.2.61
A.2.62
A.2.63
A.2.64
A.3
Refresh Timer Counter Register (RTCNT)..........................................................
Refresh Timer Constant Register (RTCOR) ........................................................
Timer Control/Status Register (TCSR) ................................................................
Timer Counter (TCNT) ........................................................................................
Reset Control/Status Register (RSTCSR) ............................................................
Standby Control Register (SBYCR) ....................................................................
Port A Data Register (PADR) ..............................................................................
Port B Data Register (PBDR)...............................................................................
Port A I/O Register (PAIOR) ...............................................................................
Port B Data Register (PBIOR) .............................................................................
Port A Control Register 1 (PACR1).....................................................................
Port A Control Register 2 (PACR2).....................................................................
Port B Control Register 1 (PBCR1) .....................................................................
Port B Control Register 2 (PBCR2) .....................................................................
Column Address Strobe Pin Control Register (CASCR) .....................................
TPC Output Mode Register (TPMR) ...................................................................
TPC Output Control Register (TPCR) .................................................................
Next Data Enable Register A (NDERA)..............................................................
Next Data Enable Register B (NDERB) ..............................................................
Next Data Register A (NDRA) (When the output triggers of TPC output
groups 0 and 1 are the same) ................................................................................
A.2.65 Next Data Register A (NDRA) (When the output triggers of TPC output
groups 0 and 1 are the same) ................................................................................
A.2.66 Next Data Register A (NDRA) (When the output triggers of TPC output
groups 0 and 1 are different) ................................................................................
A.2.67 Next Data Register A (NDRA) (When the output triggers of TPC output
groups 0 and 1 are different) ................................................................................
A.2.68 Next Data Register B (NDRB) (When the output triggers of TPC output
groups 2 and 3 are the same) ................................................................................
A.2.69 Next Data Register B (NDRB) (When the output triggers of TPC output
groups 2 and 3 are the same) ................................................................................
A.2.70 Next Data Register B (NDRB) (When the output triggers of TPC output
groups 2 and 3 are different) ................................................................................
A.2.71 Next Data Register B (NDRB) (When the output triggers of TPC output
groups 2 and 3 are different) ................................................................................
Register Status in Reset and Power-Down States..............................................................
558
559
559
561
561
562
563
564
565
566
567
569
571
573
575
576
577
579
579
580
581
581
582
582
583
584
584
585
Appendix B Pin States ....................................................................................................... 588
Appendix C External Dimensions ................................................................................. 594
Section 1 Overview
1.1
SuperH Microcomputer Features
The SuperH microcomputer (SH7000 series) is a new generation reduced instruction set computer
(RISC) in which a Hitachi-original CPU and the peripheral functions required for system
configuration are integrated onto a single chip.
The CPU has a RISC-type instruction set. Most instructions can be executed in one clock cycle,
which strikingly improves instruction execution speed. In addition, the CPU has a 32-bit internal
architecture for enhanced data-processing ability. As a result, the CPU enables high-performance
systems to be constructed with advanced functionality at low cost, even in applications such as
realtime control that require very high speeds, an impossibility with conventional microcomputers.
The SH microcomputer includes peripheral functions such as large-capacity ROM, RAM, a direct
memory access controller (DMAC), timers, a serial communication interface (SCI), an interrupt
controller (INTC), and I/O ports. External memory access support functions enable direct
connection to SRAM and DRAM. These features can drastically reduce system cost.
For on-chip ROM, masked ROM or electrically programmable ROM (PROM) can be selected.
The PROM version can be programmed by users with a general-purpose EPROM programmer.
Table 1.1 lists the features of the SH microcomputers (SH7020 and SH7021).
RENESAS 1
Table 1.1
Features of the SH7020 and SH7021 Microcomputers
Feature
Description
CPU
Original Hitachi architecture
32-bit internal data paths
General-register machine:
• Sixteen 32-bit general registers
• Three 32-bit control registers
• Four 32-bit system registers
RISC-type instruction set:
• Instruction length: 16-bit fixed length for improved code
efficiency
• Load-store architecture (basic arithmetic and logic operations
are executed between registers)
• Delayed unconditional branch instructions reduce pipeline
disruption
• Instruction set optimized for C language
Instruction execution time: one instruction/cycle (50 ns/instruction
at 20-MHz operation)
Address space: 4 Gbytes available on the architecture
On-chip multiplier: multiplication operations (16 bits × 16 bits →
32 bits) executed in 1–3 cycles, and multiplication/accumulation
operations (16 bits × 16 bits + 42 bits → 42 bits) executed in 2–3
cycles
Five-stage pipeline
Operating modes
Operating modes:
• On-chip ROMless mode
• On-chip ROM mode
Processing states:
• Power-on reset state
• Manual reset state
• Exception processing state
• Program execution state
• Power-down state
• Bus-released state
Power-down states:
• Sleep mode
• Software standby mode
2 RENESAS
Table 1.1
Features of the SH7020 and SH7021 Microcomputers (cont)
Feature
Description
Interrupt controller (INTC)
Nine external interrupt pins (NMI, IRQ0–IRQ7)
Thirty internal interrupt sources
Sixteen programmable priority levels
User break controller (UBC)
Generates an interrupt when the CPU or DMAC generates a bus
cycle with specified conditions
Simplifies configuration of a self-debugger
Clock pulse generator (CPG)
On-chip clock pulse generator (maximum operating frequency:
20 MHz):
• 20-MHz pulses can be generated from a 20-MHz crystal with a
duty cycle correcting circuit
Bus state controller (BSC)
Supports external memory access:
• Sixteen-bit external data bus
Address space divided into eight areas with the following preset
features:
• Bus size (8 or 16 bits)
• Number of wait cycles can be defined by user.
• Type of area (external memory area, DRAM area, etc.)
—
Simplifies connection to ROM, SRAM, DRAM, and
peripheral I/O
• When the DRAM area is accessed:
—
RAS and CAS signals for DRAM are output
—
Tp cycles can be generated to assure RAS precharge
time
—
Address multiplexing is supported internally, so DRAM
can be connected directly
• Chip select signals (CS0 to CS7) are output for each area
DRAM refresh function:
• Programmable refresh interval
• Supports CAS-before-RAS refresh and self-refresh modes
DRAM burst access function:
• Supports high-speed access modes for DRAM
Wait cycles can be inserted by an external WAIT signal
One-stage write buffer improves the system performance
Data bus parity can be generated and checked
RENESAS 3
Table 1.1 Features of the SH7032 and SH7034 Microcomputers (cont)
Feature
Description
Direct memory access
controller (DMAC)
(4 channels)
Permits DMA transfer between the following modules:
• External memory
• External I/O
• On-chip memory
• Peripheral on-chip modules (except DMAC)
DMA transfer can be requested from external pins, on-chip SCI, onchip timers, and on-chip A/D converter
Cycle-steal mode or burst mode
Channel priority level is selectable
Channels 0 and 1: dual or single address transfer mode is
selectable; external request sources are supported; Channels 2 and
3: dual address transfer mode, internal request sources only
16-bit integrated-timer
pulse unit (ITU)
Ten types of waveforms can be output
Input pulse width and cycle can be measured
PWM mode: pulse output with 0–100% duty cycle (maximum
resolution: 50 ns)
Complementary PWM mode: can output a maximum of three pairs
of non-overlapping PWM waveforms
Phase counting mode: can count up or down according to the phase
of an external two-phase clock
Timing pattern controller
(TPC)
Maximum 16-bit output (4 bits × 4 channels) can be output
Non-overlap intervals can be established between pairs of
waveforms
Timing-source timer is selectable
Watchdog timer (WDT)
(1 channel)
Can be used as watchdog timer or interval timer
Timer overflow can generate an internal reset, external signal, or
interrupt
Power-on reset or manual reset can be selected as the internal reset
Serial communication
Asynchronous or clocked synchronous mode is selectable
interface (SCI) (2 channels) Can transmit and receive simultaneously (full duplex)
On-chip baud rate generator in each channel
Multiprocessor communication function
4 RENESAS
Table 1.1
Features of the SH7032 and SH7034 Microcomputers (cont)
Feature
Description
I/O ports
Total of 40 I/O lines (32 input/output lines, 8 input-only lines):
• Port A: 16 input/output lines (input or output can be selected for
each bit)
• Port B: 16 input/output lines (input or output can be selected for
each bit)
On-chip memory
SH7020: 16-kbyte masked ROM, and 1-kbyte RAM
SH7021: 32-kbyte electrically programmable ROM or masked Rom,
and 1-kbyte RAM
32-bit data can be accessed in one clock cycle
RENESAS 5
ROMless
ROM
masked
PROM
6 RENESAS
SH7020
masked
SH7021
ROM
On-Chip
ROM
HD6437021X
HD6437021XI
HD6437021VX
HD6437021VXI
HD6477021X
HD6477021XI
HD6477021VX
HD6477021VXI
HD6437020X
HD6437020XI
HD6437020VX
HD6437020VXI
HD6417020SX20I
HD6417020SVX12I
-20 to +75 °C
-40 to +85 °C
-20 to +75 °C
-40 to +85 °C
-20 to +75 °C
-40 to +85 °C
-20 to +75 °C
-40 to +85 °C
-20 to +75 °C
-40 to +85 °C
-20 to +75 °C
-40 to +85 °C
-40 to +85 °C
-40 to +85 °C
2 to 20MHz
2 to 16.6MHz
2 to 20MHz
2 to 12.5MHz
3.3V
2 to 12.5MHz
2 to 16.6MHz
2 to 20MHz
2 to 12.5MHz
2 to 12.5MHz
2 to 16.6MHz
2 to 20MHz
Model
Operating
temperature
Operating
Frequency
5.0V
3.3V
5.0V
3.3V
5.0V
3.3V
5.0V
Operating
Voltage
Product Line
Product
Number
Table 1.2
HD6417020VX12I
HD6417020X20I
HD6437020VTEI
HD6437020VTE
HD6437020TEI
HD6437020TE
HD6477021VTEI
HD6477021VTE
HD6477021TEI
HD6477021TE
HD6437021VTEI
HD6437021VTE
HD6437021TEI
HD6437021TE
Marking
Model No.
(TFP-100B)
plastic TQFP
100-pin
Package
Address
RES(Vpp)*2
WDTOVF
MD2
MD1
MD0
NMI
CK
EXTAL
XTAL
VCC
VCC
VCC
VCC
VCC
VCC
VSS
VSS
VSS
VSS
VSS
VSS
VSS
VSS
VSS
VSS
AVref
AVCC
AVSS
PROM or masked ROM*1
Serial communication interface
(2 channels)
Programmable
timing pattern
controller
16-bit
integrated-timer
pulse unit
Watchdog timer
PB15/TP15/IRQ7
PB14/TP14/IRQ6
PB13/TP13/IRQ5/SCK1
PB12/TP12/IRQ4/SCK0
PB11/TP11/TxD1
PB10/TP10/RxD1
PB9/TP9/TxD0
PB8/TP8/RxD0
PB7/TP7/TOCXB4/TCLKD
PB6/TP6/TOCXA4/TCLKC
PB5/TP5/TIOCB4
PB4/TP4/TIOCA4
PB3/TP3/TIOCB3
PB2/TP2/TIOCA3
PB1/TP1/TIOCB2
PB0/TP0/TIOCA2
: Peripheral data bus (16 bits)
: Internal address bus (24 bits)
;;;;;
;;;;; : Internal upper data bus (16 bits)
;;;;;
;;;;; : Internal lower data bus (16 bits)
AD15
AD14
AD13
AD12
AD11
AD10
AD9
AD8
AD7
AD6
AD5
AD4
AD3
AD2
AD1
AD0
Bus state controller
Port B
: Peripheral address bus (24 bits)
A15
A14
A13
A12
A11
A10
A9
A8
A7
A6
A5
A4
A3
A2
A1
A0 (HBS)
RAM*1
;;;;
;;
;;
;;
;;
;;;;
;;;;
;;
;;
;;
;;
;;
;;
;;
;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;;
;;;;
;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
Direct
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
memory
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;
;;;;;
CPU
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;
;;;;;
access
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;
;;;; controller
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
User
Interrupt
break
controller controller
Address
CS3/CASL
CS2
CS1/CASH
CS0
A21
A20
A19
A18
A17
A16
Port A
Data/address
PA15/IRQ3/DREQ1
PA14/IRQ2/DACK1
PA13/IRQ1/DREQ0/TCLKB
PA12/IRQ0/DACK0/TCLKA
PA11/DPH/TIOCB1
PA10/DPL/TIOCA1
PA9/AH/IRQOUT/ADTRG
PA8/BREQ
PA7/BACK
PA6/RD
PA5/WRH (LBS)
PA4/WRL (WR)
PA3/CS7/WAIT
PA2/CS6/TIOCB0
PA1/CS5/RAS
PA0/CS4/TIOCA0
Block Diagram
Clock pulse
generator
1.2
Notes: *1. SH7020: 16-kbyte masked ROM and 1-kbyte RAM.
SH7021: 32-kbyte PROM or Masked ROM and 1-kbyte RAM.
*2. Vpp: SH7021 (PROM version)
Figure 1.1 Block Diagram
RENESAS 7
Pin Descriptions
1.3.1
Pin Arrangement
TFP-100B
(Top view)
75
74
73
72
71
70
69
68
67
66
65
64
63
62
61
60
59
58
57
56
55
54
53
52
51
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
A5
A6
A7
A8
A9
A10
VSS
A11
A12
A13
A14
A15
VCC
A16
A17
VSS
A18
A19
A20
A21
CS0
CS1/CASH
CS2
CS3/CASL
VSS
AD0
AD1
AD2
VSS
AD3
AD4
AD5
AD6
AD7
AD8
AD9
AD10
VCC
AD11
VSS
AD12
AD13
AD14
AD15
A0(HBS)
A1
A2
A3
VSS
A4
100
99
98
97
96
95
94
93
92
91
90
89
88
87
86
85
84
83
82
81
80
79
78
77
76
PB15/TP15/IRQ7
PB14/TP14/IRQ6
PB13/TP13/IRQ5/SCK1
PB12/TP12/IRQ4/SCK0
PB11/TP11/TxD1
PB10/TP10/RxD1
PB9/TP9/TxD0
PB8/TP8/RxD0
VSS
PB7/TP7/TOCXB4/TCLKD
PB6/TP6/TOCXA4/TCLKC
PB5/TP5/TIOCB4
VCC
PB4/TP4/TIOCA4
PB3/TP3/TIOCB3
PB2/TP2/TIOCA3
PB1/TP1/TIOCB2
PB0/TP0/TIOCA2
VSS
VSS
VCC
MD2
MD1
MD0
RES(Vpp)*
1.3
Notes: Vpp: SH7021 (PROM version)
Figure 1.2 Pin Arrangement
8 RENESAS
WDTOVF
NMI
VCC
XTAL
EXTAL
VSS
CK
PA15/IRQ3/DREQ1
PA14/IRQ2/DACK1
PA13/IRQ1/DREQ0/TCLKB
PA12/IRQ0/DACK0/TCLKA
PA11/DPH/TICOB1
VCC
PA10/DPL/TIOCA1
PA9/AH/IRQOUT
PA8/BREQ
VSS
PA7/BACK
PA6/RD
PA5/WRH(LBS)
PA4/WRL(WR)
PA3/CA7/WAIT
PA2/CS6/TIOCB0
PA1/CS5/RAS
PA0/CS4/TIOCA0
1.3.2
Pin Functions
Table 1.3 describes the pin functions.
Table 1.3
Pin Functions
Type
Symbol
Pin No.
I/O
Name and Function
Power
VCC
13, 38,
63, 73,
80, 88
I
Power: Connected to the power supply. Connect all
VCC pins to the system power supply . The chip will not
operate if any VCC pin is left unconnected.
VSS
4, 15, 24,
32, 41,
50, 59,
70, 81,
82, 92
I
Ground: Connected to ground. Connect all V SS pins to
the system ground. The chip will not operate if any VSS
pin is left unconnected.
VPP
76*
I
RES pin in the MCU mode. Apply +12.5V when
programming the PROM in the SH7021 (PROM
version).
EXTAL
71
I
Crystal/external clock: Connected to a crystal resonator
or external clock input having the same frequency as
the system clock (CK).
XTAL
72
I
Crystal: Connected to a crystal resonator with the same
frequency as the system clock (CK). If an external clock
is input at the EXTAL pin, leave XTAL open.
CK
69
O
System clock: Supplies the system clock (CK) to
peripheral devices.
RES
76
I
Reset: Low input causes a power-on reset if NMI is
high, or a manual reset if NMI is low.
WDTOVF
75
O
Watchdog timer overflow: Overflow output signal from
the watchdog timer.
BREQ
60
I
Bus request: Driven low by an external device to
request the bus ownership.
BACK
58
O
Bus request acknowledge: Indicates that bus ownership
has been granted to an external device. By receiving
the BACK signal, a device that has sent a BREQ signal
can confirm that it has been granted the bus.
Clock
System
control
Note: Pin 76 is RES in the SH7020, SH7021 (Masked ROM version) and Vpp in the SH7021
(PROM version).
RENESAS 9
Table 1.3
Type
Pin Functions (cont)
Symbol
Operating MD2,
mode
MD1,
control
MD0
Pin No.
I/O
Name and Function
79–77
I
Mode select: Selects the operating mode. Do not
change these inputs while the chip is operating. The
following table lists the possible operating modes and
their corresponding MD2–MD0 values.
Operating
MD2 MD1 MD0 Mode
Interrupts
0
0
0
0
0
1
0
1
0
0
1
1
1
0
0
1
0
1
1
1
0
1
1
1
MCU mode
On-chip
ROM
Bus
Size in
Area 0
Disabled
8 bits
16 bits
Enabled* 1
(Reserved)
PROM
mode* 2
NMI
74
I
Nonmaskable interrupt: Nonmaskable interrupt request
signal. The rising or falling edge can be selected for
signal detection.
IRQ0–
IRQ7
65–68,
I
Interrupt request 0–7: Maskable interrupt request
signals. Level input or edge-triggered input can be
selected.
IRQOUT
61
O
Slave interrupt request output: Indicates occurrence of
an interrupt while the bus is released.
Address
bus
A21–A0
45–42, 40, O
39, 37–33,
31–25,
23–20
Address bus: Outputs addresses.
Data bus
AD15–
AD0
19–16, 14, I/O
12-5, 3–1
Data bus: 16-bit bidirectional data bus that is
multiplexed with the lower 16 bits of the address bus.
DPH
64
I/O
Upper data bus parity: Parity data for D15–D8.
DPL
62
I/O
Lower data bus parity: Parity data for D7–D0.
WAIT
54
I
Wait: Requests the insertion of wait states (T W ) into
the bus cycle when the external address space is
accessed.
Bus
control
97–100
Notes : 1.Use prohibited in the SH7020 Romless version.
2.Can only be used in the SH7021 ZTAT version.
10 RENESAS
Table 1.3
Pin Functions (cont)
Type
Symbol
Pin No.
I/O
Name and Function
Bus
control
RAS
52
O
Row address strobe: DRAM row-address strobe-timing
signal.
(cont)
CASH
47
O
Column address strobe high: DRAM column-address
strobe-timing signal outputs low level to access the
upper eight data bits.
CASL
49
O
Column address strobe low: DRAM column-address
strobe-timing signal outputs low level to access the
lower eight data bits.
RD
57
O
Read: Indicates reading of data from an external device.
WRH
56
O
Upper write: Indicates write access to the upper eight
bits of an external device.
WRL
55
O
Lower write: Indicates write access to the lower eight
bits of an external device.
CS0–CS7
46–49,
51–54
O
Chip select 0–7: Chip select signals for accessing
external memory and devices.
AH
61
O
Address hold: Address hold timing signal for a device
using a multiplexed address/data bus.
O
Upper/lower byte strobe: Upper and lower byte strobe
signals. (Also used as WRH and A0.)
HBS, LBS 20, 56
DMAC
16-bit
integratedtimer pulse
unit (ITU)
WR
55
O
Write: Brought low during write access. (Also used as
WRL.)
DREQ0,
DREQ1
66, 68
I
DMA transfer request (channels 0 and 1): Input pins for
external DMA transfer requests.
DACK0,
DACK1
65, 67
O
DMA transfer acknowledge (channels 0 and 1):
Indicates that DMA transfer is acknowledged.
TIOCA0,
TIOCB0
51, 53
I/O
ITU input capture/output compare (channel 0): Input
capture or output compare pins.
TIOCA1,
TIOCB1
62, 64
I/O
ITU input capture/output compare (channel 1): Input
capture or output compare pins.
TIOCA2,
TIOCB2
83, 84
I/O
ITU input capture/output compare (channel 2): Input
capture or output compare pins.
TIOCA3,
TIOCB3
85, 86
I/O
ITU input capture/output compare (channel 3): Input
capture or output compare pins.
TIOCA4,
TIOCB4
87, 89
I/O
ITU input capture/output compare (channel 4): Input
capture or output compare pins.
RENESAS 11
Table 1.3
Pin Functions (cont)
Type
Symbol
Pin No.
I/O
Name and Function
16-bit
integratedtimer pulse
unit (ITU)
TOCXA4,
TOCXB4
90, 91
O
ITU output compare (channel 4): Output compare pins.
TCLKA–
TCLKD
65, 66,
90, 91
I
ITU timer clock input: External clock input pins for ITU
counters.
Timing
pattern
controller
(TPC)
TP15–
TP0
100–93,
91–89,
87–83
O
Timing pattern output 15–0: Timing pattern output pins.
Serial
communication
interface
(SCI)
TxD0,
TxD1
94, 96
O
Transmit data (channels 0 and 1): Transmit data output
pins for SCI0 and SCI1.
RxD0,
RxD1
93, 95
I
Receive data (channels 0 and 1): Receive data input
pins for SCI0 and SCI1.
SCK0,
SCK1
97, 98
I/O
Serial clock (channels 0 and 1): Clock input/output pins
for SCI0 and SCI1.
PA15–
PA0
68–64,
62–60,
58-51
I/O
Port A: 16-bit input/output pins. Input or output can be
selected individually for each bit.
PB15–
PB0
100–93,
91–89,
87–83
I/O
Port B: 16-bit input/output pins. Input or output can be
selected individually for each bit.
I/O ports
12 RENESAS
1.3.3
Pin Layout by Mode
Table 1.4 shows pin layout by mode
Table 1.4 Pin Layout by Mode
MCU Mode
PROM Mode
(SH7021PROM version)
Pin
No.
MCU Mode
PROM Mode
(SH7021PROM version)
1
AD0
AD0
29
A8
A8
2
AD1
AD1
30
A9
OE
3
AD2
AD2
31
A10
A10
4
Vss
Vss
32
VSS
VSS
5
AD3
AD3
33
A11
A11
6
AD4
AD4
34
A12
A12
7
AD5
AD5
35
A13
A13
8
AD6
AD6
36
A14
A14
9
AD7
AD7
37
A15
A15
10
AD8
NC
38
VCC
VCC
11
AD9
NC
39
A16
A16
12
AD10
NC
40
A17
VCC
13
VCC
VCC
41
VSS
VSS
14
AD11
NC
42
A18
VCC
15
VSS
VSS
43
A19
NC
16
AD12
NC
44
A20
NC
17
AD13
NC
45
A21
NC
18
AD14
NC
46
CS0
NC
19
AD15
NC
47
CS1/CASH
NC
20
A0(HBS)
A0
48
CS2
NC
21
A1
A1
49
CS3/CASL
NC
22
A2
A2
50
VSS
VSS
23
A3
A3
51
PA0/CS4/TIOCA0
NC
24
VSS
VSS
52
PA1/CS5/RAS
NC
25
A4
A4
53
PA2/CS6/ TIOCB0
PGM
26
A5
A5
54
PA3/CS7/WAIT
CE
27
A6
A6
55
PA4/WRL(WR)
NC
28
A7
A7
56
PA5/WRH(LBS)
NC
Pin
No.
RENESAS 13
Table 1.3.3 Pin Layout by Mode (cont)
MCU Mode
PROM Mode
(SH7021PROM version)
Pin
No.
MCU Mode
PROM Mode
(SH7021PROM version)
57
PA6/RD
NC
80
VCC
VCC
58
PA7/BACK
NC
81
VSS
VSS
59
VSS
VSS
82
VSS
VSS
60
PA8/BREQ
NC
83
PB0/TP0/TIOCA2
NC
61
PA9/AH/IRQOUT
NC
84
PB1/TP1/TIOCB2
NC
62
PA10/DPL/TIOCA1
NC
85
PB2/TP2/TIOCA3
NC
63
VCC
VCC
86
PB3/TP3/TIOCB3
NC
64
PA11/DPH/TIOCB1
NC
87
PB4/TP4/TIOCA4
NC
65
PA12/IRQ0/DACK0/
NC
Pin
No.
TCLKA
66
88
VCC
VCC
89
PB5/TP5/TIOCB4
NC
PB6/TP6/TOCXA4/
TCLKC
NC
PB7/TP7/TOCXB4/
NC
PA13/IRQ1/DREQ0/
TCLKLB
NC
90
91
67
PA14/IRQ2/DACK1
NC
68
PA15/IRQ3/DREQ1
NC
69
CK
NC
92
VSS
VSS
70
VSS
VSS
93
PB8/TP8/RxD0
NC
71
EXTAL
NC
94
PB9/TP9/TxD0
NC
72
XTAL
NC
95
PB10/TP10/RxD1
NC
73
VCC
VCC
96
PB11/TP11/TXD1
NC
74
NMI
A9
97
PB12/TP12/IRQ4/
NC
TCLKD
SCK0
75
WDTOVF
NC
76
RES
Vpp
77
MD0
VCC
78
MD1
VCC
99
PB14/TP14/IRQ6
NC
79
MD2
VCC
100
PB15/TP15/IRQ7
NC
14 RENESAS
98
PB13/TP13/IRQ5/
NC
SCK1
Section 2 CPU
2.1
Register Configuration
The register set consists of sixteen 32-bit general registers, three 32-bit control registers, and four
32-bit system registers.
2.1.1
General Registers (Rn)
General registers Rn consist of sixteen 32-bit registers (R0–R15). General registers are used for
data processing and address calculation. Register R0 also functions as an index register. For some
instructions, the R0 register must be used. Register R15 functions as a stack pointer to save or
recover status registers (SR) and program counter (PC) during exception processing.
31
0
R0
R1
R2
R3
R4
R0 functions as an index register
in the indexed register addressing
mode and indirect indexed GBR
addressing mode. In some instructions, R0 functions as a source
register or a destination register.
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
R15, SP
R15 functions as a stack pointer (SP)
during exception processing.
Figure 2.1 General Registers (Rn)
2.1.2
Control Registers
Control registers consist of the 32-bit status register (SR), global base register (GBR), and vector
base register (VBR). The status register indicates processing states. The global base register
RENESAS15
functions as a base address for the indirect GBR addressing mode to transfer data to the registers
of peripheral on-chip modules. The vector base register functions as the base address of the
exception processing vector area including interrupts.
9 8 7 6 5 4 32 1 0
31
SR
M Q I3 I2 I1 I0
ST
SR: Status register
T bit: The MOVT, CMP, TAS, TST,
BT, BF, SETT, and CLRT instructions
use the T bit to indicate a true (1) or
false (0). The ADDV, ADDC, SUBV,
SUBC, DIV0U, DIV0S, DIV1, NEGC,
SHAR, SHAL, SHLR, SHLL, ROTR,
ROTL, ROTCR and ROTCL
instructions also use the T bit to indicate
carry/borrow or overflow/underflow
S bit: Used by the MAC instruction.
Reserved bits. These bits always read 0.
The write value should always be 0.
Bits I0–I3: Interrupt mask bits.
M and Q bits: Used by the DIV0U, DIV0S,
and DIV1 instructions.
31
GBR
31
VBR
Global base register (GBR):
0 Indicates the base address of the indirect
GBR addressing mode. The indirect GBR
addressing mode is used to transfer data
to the register areas peripheral on-chip
modules.
0 Vector base register (VBR):
Stores the base address of the exception
processing vector area.
Figure 2.2 Control Registers
2.1.3
System Registers
System registers consist of four 32-bit registers: multiply and accumulate registers high and low
(MACH and MACL), procedure register (PR), and program counter (PC). The multiply and
accumulate registers store the results of multiply and accumulate operations. The procedure
register stores the return address from the subroutine procedure. The program counter stores
program addresses to control the flow of the processing.
16 RENESAS
31
9
(sign extended)
0
MACH
MACL
0
31
PR
0
31
PC
Multiply and accumulate (MAC) registers
high and low (MACH, MACL): Store the
results of multiply and accumulate operations. MACH is sign-extended when read
because only the lowest 10 bits are valid.
Procedure register (PR): Stores a return
address from a subroutine procedure.
Program counter (PC): Indicates the
fourth byte (second instruction) after
the current instruction.
Figure 2.3 System Registers
2.1.4
Initial Values of Registers
Table 2.1 lists the values of the registers after reset.
Table 2.1
Initial Values of Registers
Classification
Register
Initial Value
General register
R0–R14
Undefined
R15 (SP)
Value of the stack pointer in the vector address table
SR
Bits I0-I3 are 1111(H'F), reserved bits are 0, and other
bits are undefined
GBR
Undefined
VBR
H'00000000
MACH, MACL, PR
Undefined
PC
Value of the program counter in the vector address
table
Control register
System register
2.2
Data Formats
2.2.1
Data Format in Registers
Register operands are always long words (32 bits). When the memory operand is only a byte (8
bits) or a word (16 bits), it is sign-extended into a long word when stored into a register (figure
2.4).
RENESAS17
31
0
Long word
Figure 2.4 Data Format in Registers
2.2.2
Data Format in Memory
Memory data formats are classified into bytes, words, and long words. Byte data can be accessed
from any address, but an address error will occur if you try to access word data starting from an
address other than 2n or long word data starting from an address other than 4n. In such cases, the
data accessed cannot be guaranteed. The hardware stack area, which is referred to by the hardware
stack pointer (SP, R15), uses only long word data starting from address 4n because this area stores
the program counter and status register (figure 2.5).
Address m + 1
Address m
23
31
Address m + 3
Address m + 2
7
15
0
7 Byte 0 7 Byte 0 7 Byte 0 7 Byte 0
Address 2n
15
Address 4n
31
Word
0 15
Word
Long word
0
0
Figure 2.5 Data Format in Memory
2.2.3
Immediate Data Format
Byte (8-bit) immediate data is located in the instruction code. Immediate data accessed by the
MOV, ADD, and CMP/EQ instructions is sign-extended and is handled in registers as long word
data. Immediate data accessed by the TST, AND, OR, and XOR instructions is zero-extended and
is handled as long word data. Consequently, AND instructions with immediate data always clear
the upper 24 bits of the destination register.
Word or long word immediate data is not located in the instruction code but rather is stored in a
memory table. The memory table is accessed by a immediate data transfer instruction (MOV)
using the PC relative addressing mode with displacement.
18 RENESAS
2.3
Instruction Features
2.3.1
RISC-Type Instruction Set
All instructions are RISC type. Their features are as follows:
16-Bit Fixed Length: Every instruction is 16 bits long, making program coding much more
efficient.
One Instruction/Cycle: Basic instructions can be executed in one cycle using the pipeline system.
One-cycle instructions are executed in 50 ns at 20 MHz.
Data Length: Long word is the standard data length for all operations. Memory can be accessed
in bytes, words, or long words. Byte or word data accessed from memory is sign-extended and
handled as long word data. Immediate data is sign-extended for arithmetic operations or zeroextended for logic operations (handled as long word data).
Table 2.2
Sign Extension of Word Data
CPU of SH7000 Series
Description
Conventional CPUs
MOV.W @(disp,PC),R1
ADD
R1,R0
......................
..
.DATA.W
H'1234
Data is sign-extended to 32 bits, and
R1 becomes H'00001234. It is next
operated upon by an ADD instruction.
ADD.W #H'1234, R0
Note: The address of the immediate data is accessed by @(disp, PC).
Load-Store Architecture: Basic operations are executed between registers. For operations that
involve memory, data is loaded to the registers and executed (load-store architecture). Instructions
such as AND that manipulate bits, however, are executed directly in memory.
Delayed Branch Instructions: Unconditional branch instructions are delayed. Pipeline disruption
during branching is reduced by first executing the instruction that follows the branch instruction,
and then branching. See the SH-1/SH-2 Programming Manual for details.
Table 2.3
Delayed Branch Instructions
CPU of SH7000 Series
Description
Conventional CPU
BRA
ADD
Executes an ADD before
branching to TRGET.
ADD.W
BRA
TRGET
R1, R0
R1, R0
TRGET
Multiplication/Accumulation Operation: The five-stage pipeline system and the on-chip
multiplier enable 16-bit × 16-bit → 32-bit multiplication operations to be executed in 1–3 cycles.
16-bit × 16-bit + 42-bit → 42-bit multiplication/accumulation operations can be executed in 2–3
RENESAS19
cycles.
T bit: T bit (in the status register) is set according to the result of a comparison, and in turn is the
condition (True/False) that determines if the program will branch. The T bit in the status register is
only changed by selected instructions, thus improving the processing speed.
Table 2.4
T bit
CPU of SH7000 Series
Description
Conventional CPU
CMP/GE R1, R0
BT
TRGET0
BF
TRGET1
T bit is set when R0 ≥ R1. The program
branches to TRGET0 when R0 ≥ R1
and to TRGET1 when R0<R1.
CMP.W
BGE
BLT
R1, R0
TRGET0
TRGET1
ADD
TST
BT
T bit is not changed by ADD. T bit is set
when R0=0. The program branches if
R0=0.
SUB.W
BEQ
#1, R0
TRGET
#–1, R0
R0, R0
TRGET
Immediate Data: Byte (8-bit) immediate data is located in the instruction code. Word or long
word immediate data is not located in instruction codes but is stored in a memory table. The
memory table is accessed by a immediate data transfer instruction (MOV) using the PC relative
addressing mode with displacement.
Table 2.5
Immediate Data Accessing
Classification
CPU of SH7000 Series
Conventional CPU
8-bit immediate
MOV
MOV.B
#H'12, R0
16-bit immediate
MOV.W @(disp,PC), R0
.........
.DATA.W
H'1234
MOV.W
#H'1234, R0
32-bit immediate
MOV.L @(disp,PC), R0
.........
.DATA.L
H'12345678
MOV.L
#H'12345678, R0
#H'12, R0
Note: The address of the immediate data is accessed by @(disp, PC).
Absolute Address: When data is accessed by absolute address, the value already in the absolute
address is placed in the memory table. By loading the immediate data when the instruction is
executed, that value is transferred to the register and the data is accessed in the indirect register
addressing mode.
20 RENESAS
Table 2.6
Absolute Address Accessing
Classification
CPU of SH7000 Series
Conventional CPU
Absolute address
MOV.L @(disp,PC), R1
MOV. B @R1, R0
.........
.DATA.L
H'12345678
MOV.B
@H'12345678, R0
Note: The address of the immediate data is accessed by @(disp, PC).
16/32-Bit Displacement: When data is accessed by 16-bit or 32-bit displacement, the pre-existing
displacement value is placed in the memory table. By loading the immediate data when the
instruction is executed, that value is transferred to the register and the data is accessed in the
indirect indexed register addressing mode.
Table 2.7
Accessing by Displacement
Classification
CPU of SH7000 Series
Conventional CPU
16-bit displacement
MOV.W @(disp, PC), R0
MOV.W @(R0, R1), R2
.........
.DATA.W
H'1234
MOV.W
@(H'1234, R1),
R2
Note: The address of the immediate data is accessed by @(disp, PC).
RENESAS21
2.3.2
Addressing Modes
Addressing modes and effective address calculation are described in table 2.8.
Table 2.8
Addressing Modes and Effective Addresses
Addressing Mnemonic
Mode
Expression
Effective Addresses Calculation
Equation
Direct
register
addressing
Rn
The effective address is register Rn. (The operand is
the contents of register Rn.)
—
Indirect
register
addressing
@Rn
The effective address is the content of register Rn.
Rn
Post-increment
indirect
register
addressing
@Rn +
Rn
Rn
The effective address is the content of register Rn. A
constant is added to the content of Rn after the
instruction is executed. 1 is added for a byte
operation, 2 for a word operation, and 4 for a long
word operation.
Rn
Rn
Rn + 1/2/4
@–Rn
Rn
1/2/4
22 RENESAS
Byte: Rn + 1
→ Rn
Long word:
Rn + 4 → Rn
The effective address is the value obtained by
subtracting a constant from Rn. 1 is subtracted for a
byte operation, 2 for a word operation, and 4 for a
long word operation.
Rn – 1/2/4
(After the
instruction is
executed)
Word: Rn + 2
→ Rn
+
1/2/4
Pre-decrement
indirect
register
addressing
Rn
–
Rn – 1/2/4
Byte: Rn – 1
→ Rn
Word: Rn – 2
→ Rn
Long word:
Rn – 4 → Rn
(Instruction
executed with
Rn after
calculation)
Table 2.8
Addressing Modes and Effective Addresses (cont)
Addressing Mnemonic
Mode
Expression
Indirect
register
addressing
with
displacement
@(disp:4,
Rn)
Effective Addresses Calculation
Equation
The effective address is Rn plus a 4-bit displacement
(disp). disp is zero-extended, and remains the same
for a byte operation, is doubled for a word operation,
and is quadrupled for a long word operation.
Byte: Rn +
disp
Rn
disp
(zero-extended)
Rn + disp × 1/2/4
+
Word: Rn +
disp × 2
Long word:
Rn + disp × 4
×
1/2/4
Indirect
indexed
register
addressing
@(R0, Rn)
The effective address is the Rn value plus R0.
Rn + R0
Rn
+
Rn + R0
R0
Indirect
GBR
addressing
with
displacement
@(disp:8,
GBR)
The effective address is the GBR value plus an 8-bit
displacement (disp). The value of disp is zeroextended, and remains the same for a byte
operation, is doubled for a word operation, and is
quadrupled for a long word operation.
GBR
disp
(zero-extended)
+
GBR
+ disp × 1/2/4
Byte: GBR +
disp
Word: GBR +
disp × 2
Long word:
GBR + disp ×
4
×
1/2/4
Indirect
indexed
GBR
addressing
@(R0,
GBR)
The effective address is the GBR value plus the R0.
GBR + R0
GBR
+
GBR + R0
R0
RENESAS23
Table 2.8
Addressing Modes and Effective Addresses (cont)
Addressing Mnemonic
Mode
Expression
PC relative
addressing
with displacement
@(disp:8,
PC)
Effective Addresses Calculation
Equation
The effective address is the PC value plus an 8-bit
displacement (disp). disp is zero-extended, and
remains the same for a byte operation, is doubled for
a word operation, and is quadrupled for a long word
operation. For a long word operation, the lowest two
bits of the PC are masked.
Word: PC +
disp × 2
Long word:
PC &
H'FFFFFFFC
+ disp × 4
PC
& *
H'FFFFFFFC
+
disp
(zero-extended)
PC + disp × 2
or
PC & H'FFFFFFFC
+ disp × 4
×
2/4
*: For long word
PC relative
addressing
disp:8
The effective address is the PC value sign-extended
with an 8-bit displacement (disp), doubled, and
added to the PC.
PC + disp × 2
PC
disp
(zero-extended)
+
PC + disp × 2
×
2
disp:12
The effective address is the PC value sign-extended
with a 12-bit displacement (disp), doubled, and
added to the PC.
PC
disp
(zero-extended)
+
×
2
24 RENESAS
PC + disp × 2
PC + disp × 2
Table 2.8
Addressing Modes and Effective Addresses (cont)
Addressing Mnemonic
Mode
Expression
Effective Addresses Calculation
Equation
Immediate
addressing
#imm:8
The 8-bit immediate data (imm) for the TST, AND,
OR, and XOR instructions are zero-extended.
—
#imm:8
The 8-bit immediate data (imm) for the MOV, ADD,
and CMP/EQ instructions are sign-extended.
—
#imm:8
Immediate data (imm) for the TRAPA instruction is
zero-extended and is quadrupled.
—
2.3.3
Instruction Formats
The instruction format refers to the source operand and the destination operand. The meaning of
the operand depends on the instruction code. Symbols are as follows.
Instruction code
xxxx
mmmm Source register
nnnn
iiii
dddd
Table 2.9
Destination register
Immediate data
Displacement
Instruction Formats
Instruction Formats
Source
Operand
Destination
Operand
Instruction
Example
0 format
—
—
NOP
—
nnnn: Direct
register
MOVT
Rn
Control register
or system
register
nnnn: Direct
register
STS
MACH,Rn
Control register
or system
register
nnnn: Indirect predecrement register
STC.L
15
0
xxxx
xxxx
xxxx
xxxx
n format
15
0
xxxx
nnnn
xxxx
xxxx
SR,@-Rn
RENESAS25
Table 2.9
Instruction Formats (cont)
Instruction Formats
m format
15
0
xxxx mmmm xxxx
xxxx
nm format
15
0
xxxx
nnnn mmmm xxxx
Source
Operand
Destination
Operand
Instruction
Example
mmmm: Direct
register
Control register or
system register
LDC
mmmm: Indirect
post-increment
register
Control register or
system register
LDC.L
@Rm+,SR
mmmm: Direct
register
—
JMP
mmmm: Direct
register
nnnn: Direct
register
ADD
Rm,Rn
mmmm: Direct
register
nnnn: Direct
register
MOV.L
Rm,@Rn
mmmm: Indirect MACH, MACL
post-increment
register (multiply/
accumulate)
Rm,SR
@Rm
MAC.W
@Rm+,@Rn+
nnnn: Indirect
post-increment
register (multiply/
accumulate)*
md format
15
0
xxxx
xxxx mmmm dddd
nd4 format
15
0
xxxx
xxxx
nnnn
mmmm: Indirect
post-increment
register
nnnn: Direct
register
MOV.L
@Rm+,Rn
mmmm: Direct
register
nnnn: Indirect predecrement register
MOV.L
Rn
mmmm: Direct
register
nnnn: Indirect
indexed register
MOV.L
Rm,@(R0,Rn)
mmmmdddd:
indirect register
with
displacement
R0 (Direct register) MOV.B
@(disp,Rm),R0
R0 (Direct
register)
nnnndddd: Indirect
register with
displacement
dddd
Note: In MAC instructions, nnnn is the source register.
26 RENESAS
Rm,@-
MOV.B
R0,@(disp,Rn)
Table 2.9
Instruction Formats (cont)
Instruction Formats
nmd format
15
0
xxxx
15
0
xxxx
dddd
dddd
d12 format
15
0
xxxx
dddd
dddd
15
0
nnnn
dddd
dddd
i format
15
0
xxxx
xxxx
iiii
nnnndddd: Indirect
register with
displacement
MOV.L
Rm,@(disp,Rn)
mmmmdddd:
Indirect register
with
displacement
nnnn: Direct
register
MOV.L
@(disp,Rm),Rn
dddddddd:
Indirect GBR
with
displacement
R0 (Direct register) MOV.L
@(disp,GBR),R0
R0(Direct
register)
dddddddd: Indirect
GBR with
displacement
dddddddd: PC
relative with
displacement
R0 (Direct register) MOVA
@(disp,PC),R0
dddddddd: PC
relative
—
BF
disp
dddddddddddd:
PC relative
—
BRA
disp
dddddddd: PC
relative with
displacement
nnnn: Direct
register
MOV.L
@(disp,PC),Rn
iiiiiiii: Immediate
Indirect indexed
GBR
AND.B
#imm,@(R0,GBR)
iiiiiiii: Immediate
R0 (Direct register) AND
#imm,R0
iiiiiiii: Immediate
—
TRAPA
#imm
iiiiiiii: Immediate
nnnn: Direct
register
ADD
#imm,Rn
15
0
nnnn
iiii
MOV.L
R0,@(disp,GBR)
iiii
ni format
xxxx
mmmm: Direct
register
Example
dddd
nd8 format
xxxx
Destination
Operand
nnnn mmmm dddd
d format
xxxx
Source
Operand
iiii
RENESAS27
2.4
Instruction Set
2.4.1
Instruction Set by Classification
Table 2.10 lists instructions by classification.
Table 2.10 Classification of Instructions
Classification
Data
transfer
Arithmetic
operations
Logic
operations
28 RENESAS
Types
Operation
Code
5
MOV
Data transfer, immediate data transfer,
peripheral module data transfer, structure
data transfer
MOVA
Effective address transfer
MOVT
T bit transfer
SWAP
Swap of upper and lower bytes
XTRCT
Extraction of the middle of registers
connected
ADD
Binary addition
ADDC
Binary addition with carry
ADDV
Binary addition with overflow check
17
6
Function
CMP/cond
Comparison
DIV1
Division
DIV0S
Initialization of signed division
DIV0U
Initialization of unsigned division
EXTS
Sign extension
EXTU
Zero extension
MAC
Multiplication and accumulation
MULS
Signed multiplication
MULU
Unsigned multiplication
NEG
Negation
NEGC
Negation with borrow
SUB
Binary subtraction
SUBC
Binary subtraction with carry
SUBV
Binary subtraction with underflow check
AND
Logical AND
NOT
Bit inversion
OR
Logical OR
TAS
Memory test and bit set
Number of
Instructions
39
28
14
Table 2.10 Classification of Instructions (cont)
Classification
Types
Operation
Code
Function
Number of
Instructions
Logic operations(cont)
6
TST
Logical AND and T bit set
14
XOR
Exclusive OR
Shift
10
ROTL
One-bit left rotation
ROTR
One-bit right rotation
ROTCL
One-bit left rotation with T bit
ROTCR
One-bit right rotation with T bit
SHAL
One-bit arithmetic left shift
SHAR
One-bit arithmetic right shift
SHLL
One-bit logical left shift
SHLLn
n-bit logical left shift
SHLR
One-bit logical right shift
SHLRn
n-bit logical right shift
BF
Conditional branch (T = 0)
BT
Conditional branch (T = 1)
BRA
Unconditional branch
BSR
Branch to subroutine procedure
JMP
Unconditional branch
JSR
Branch to subroutine procedure
RTS
Return from subroutine procedure
CLRT
T bit clear
CLRMAC
MAC register clear
LDC
Load to control register
LDS
Load to system register
NOP
No operation
RTE
Return from exception processing
SETT
T bit set
SLEEP
Shift into power-down mode
STC
Storing control register data
STS
Storing system register data
TRAPA
Trap exception processing
Branch
System
control
Total
7
11
56
14
7
31
133
RENESAS29
Instruction codes, operation, and execution states are listed in the following format in order by
classification.
Table 2.11 Instruction Code Format
Item
Format
Instruction
mnemonic
OP.Sz
Instruction
code
MSB ↔ LSB
mmmm: Source register
nnnn: Destination register
0000: R0
0001: R1
...........
1111: R15
iiii: Immediate data
dddd: Displacement
Operation
summary
→, ←
(xx)
M/Q/T
&
|
^
~
<<n, >>n
Direction of transfer
Memory operand
Flag bits in the SR
Logical AND of each bit
Logical OR of each bit
Exclusive OR of each bit
Logical NOT of each bit
n-bit shift
Execution
cycle
Explanation
SRC,DEST OP: Operation code
Sz: Size
SRC: Source
DEST: Destination
Rm: Source register
Rn: Destination register
imm: Immediate data
disp: Displacement*
Value when no wait states are inserted
Instruction execution cycles: The execution cycles shown in
the table are minimums. The actual number of cycles may
be increased:
1. When contention occurs between instruction fetches
and data access, or
2. When the destination register of the load instruction
(memory → register) and the register used by the next
instruction are the same.
T bit
Value of T bit after instruction is executed
—
No change
Note: Scaling (×1, ×2, ×4) is performed according to the instruction operand size. See
"SH-1/SH-2 Programming Manual" for details.
30 RENESAS
Table 2.12 Data Transfer Instructions
Execution
Cycles
T
bit
Instruction
Instruction Code
Operation
MOV
#imm,Rn
1110nnnniiiiiiii
#imm → Sign extension →
Rn
1
—
MOV.W
@(disp,PC),Rn 1001nnnndddddddd
(disp×2 + PC) → Sign
extension → Rn
1
—
MOV.L
@(disp,PC),Rn 1101nnnndddddddd
(disp×4 + PC) → Rn
1
—
MOV
Rm,Rn
0110nnnnmmmm0011
Rm → Rn
1
—
MOV.B
Rm,@Rn
0010nnnnmmmm0000
Rm → (Rn)
1
—
MOV.W
Rm,@Rn
0010nnnnmmmm0001
Rm → (Rn)
1
—
MOV.L
Rm,@Rn
0010nnnnmmmm0010
Rm → (Rn)
1
—
MOV.B
@Rm,Rn
0110nnnnmmmm0000
(Rm) → Sign extension →
Rn
1
—
MOV.W
@Rm,Rn
0110nnnnmmmm0001
(Rm) → Sign extension →
Rn
1
—
MOV.L
@Rm,Rn
0110nnnnmmmm0010
(Rm) → Rn
1
—
MOV.B
Rm,@–Rn
0010nnnnmmmm0100
Rn–1 → Rn, Rm → (Rn)
1
—
MOV.W
Rm,@–Rn
0010nnnnmmmm0101
Rn–2 → Rn, Rm → (Rn)
1
—
MOV.L
Rm,@–Rn
0010nnnnmmmm0110
Rn–4 → Rn, Rm → (Rn)
1
—
MOV.B
@Rm+,Rn
0110nnnnmmmm0100
(Rm) → Sign extension →
Rn,Rm + 1 → Rm
1
—
MOV.W
@Rm+,Rn
0110nnnnmmmm0101
(Rm) → Sign extension →
Rn,Rm + 2 → Rm
1
—
MOV.L
@Rm+,Rn
0110nnnnmmmm0110
(Rm) → Rn,Rm + 4 → Rm
1
—
MOV.B
R0,@(disp,Rn) 10000000nnnndddd
R0 → (disp + Rn)
1
—
MOV.W
R0,@(disp,Rn) 10000001nnnndddd
R0 → (disp×2 + Rn)
1
—
MOV.L
Rm,@(disp,Rn) 0001nnnnmmmmdddd
Rm → (disp×4 + Rn)
1
—
MOV.B
@(disp,Rm),R0 10000100mmmmdddd
(disp + Rm) → Sign
extension → R0
1
—
MOV.W
@(disp,Rm),R0 10000101mmmmdddd
(disp×2 + Rm) → Sign
extension → R0
1
—
MOV.L
@(disp,Rm),Rn 0101nnnnmmmmdddd
(disp×4 + Rm) → Rn
1
—
MOV.B
Rm,@(R0,Rn)
0000nnnnmmmm0100
Rm → (R0 + Rn)
1
—
MOV.W
Rm,@(R0,Rn)
0000nnnnmmmm0101
Rm → (R0 + Rn)
1
—
RENESAS31
Table 2.12 Data Transfer Instructions (cont)
Instruction
Instruction Code
Operation
Execution
Cycles
MOV.L
Rm,@(R0,Rn)
0000nnnnmmmm0110
Rm → (R0 + Rn)
1
—
MOV.B
@(R0,Rm),Rn
0000nnnnmmmm1100
(R0 + Rm) → Sign
extension → Rn
1
—
MOV.W
@(R0,Rm),Rn
0000nnnnmmmm1101
(R0 + Rm) → Sign
extension → Rn
1
—
MOV.L
@(R0,Rm),Rn
0000nnnnmmmm1110
(R0 + Rm) → Rn
1
—
MOV.B
R0,@(disp,GBR)
11000000dddddddd
R0 → (disp + GBR)
1
—
MOV.W
R0,@(disp,GBR)
11000001dddddddd
R0 → (disp×2 + GBR)
1
—
MOV.L
R0,@(disp,GBR)
11000010dddddddd
R0 → (disp×4 + GBR)
1
—
MOV.B
@(disp,GBR),R0
11000100dddddddd
(disp + GBR) → Sign
extension → R0
1
—
MOV.W
@(disp,GBR),R0
11000101dddddddd
(disp×2 + GBR) → Sign
extension → R0
1
—
MOV.L
@(disp,GBR),R0
11000110dddddddd
(disp×4 + GBR) → R0
1
—
MOVA
@(disp,PC),R0
11000111dddddddd
disp×4 + PC → R0
1
—
MOVT
Rn
0000nnnn00101001
T → Rn
1
—
SWAP.B Rm,Rn
0110nnnnmmmm1000
Rm → Swap the bottom
two bytes → Rn
1
—
SWAP.W Rm,Rn
0110nnnnmmmm1001
Rm → Swap two
consecutive words → Rn
1
—
XTRCT
0010nnnnmmmm1101
Center 32 bits of Rm and
Rn → Rn
1
—
Rm,Rn
32 RENESAS
T
bit
Table 2.13 Arithmetic Instructions
Instruction
Instruction Code
Operation
Execution
Cycles
T bit
ADD
Rm,Rn
0011nnnnmmmm1100
Rn + Rm → Rn
1
—
ADD
#imm,Rn
0111nnnniiiiiiii
Rn + imm → Rn
1
—
ADDC
Rm,Rn
0011nnnnmmmm1110
Rn + Rm + T → Rn,
Carry → T
1
Carry
ADDV
Rm,Rn
0011nnnnmmmm1111
Rn + Rm → Rn,
Overflow → T
1
Overflow
CMP/EQ #imm,R0
10001000iiiiiiii
If R0 = imm, 1 → T
1
Comparison
result
CMP/EQ Rm,Rn
0011nnnnmmmm0000
If Rn = Rm, 1 → T
1
Comparison
result
CMP/HS Rm,Rn
0011nnnnmmmm0010
If Rn≥Rm with
1
unsigned data, 1 → T
Comparison
result
CMP/GE Rm,Rn
0011nnnnmmmm0011
If Rn ≥ Rm with
signed data, 1 → T
1
Comparison
result
CMP/HI Rm,Rn
0011nnnnmmmm0110
If Rn > Rm with
1
unsigned data, 1 → T
Comparison
result
CMP/GT Rm,Rn
0011nnnnmmmm0111
If Rn > Rm with
signed data, 1 → T
1
Comparison
result
CMP/PZ Rn
0100nnnn00010001
If Rn ≥ 0, 1 → T
1
Comparison
result
CMP/PL Rn
0100nnnn00010101
If Rn > 0, 1 → T
1
Comparison
result
CMP/STR Rm,Rn
0010nnnnmmmm1100
If Rn and Rm have
an equivalent byte, 1
→T
1
Comparison
result
DIV1
Rm,Rn
0011nnnnmmmm0100
Single-step division
(Rn/Rm)
1
Calculation
result
DIV0S
Rm,Rn
0010nnnnmmmm0111
MSB of Rn → Q,
MSB of Rm → M,
M^Q→T
1
Calculation
result
DIV0U
0000000000011001
0 → M/Q/T
1
0
EXTS.B Rm,Rn
0110nnnnmmmm1110
A byte in Rm is signextended → Rn
1
—
RENESAS33
Table 2.13 Arithmetic Instructions (cont)
Execution
Cycles
Instruction
Instruction Code
Operation
T bit
EXTS.W Rm,Rn
0110nnnnmmmm1111
A word in Rm is sign- 1
extended → Rn
—
EXTU.B Rm,Rn
0110nnnnmmmm1100
A byte in Rm is zeroextended → Rn
1
—
EXTU.W Rm,Rn
0110nnnnmmmm1101
A word in Rm is zero- 1
extended → Rn
—
MAC.W
@Rm+,@Rn+
0100nnnnmmmm1111
Signed operation of
(Rn) × (Rm) + MAC
→ MAC
3/(2)*
—
MULS
Rm,Rn
0010nnnnmmmm1111
Signed operation of
Rn × Rm → MAC
1–3*
—
MULU
Rm,Rn
0010nnnnmmmm1110
Unsigned operation
of Rn × Rm → MAC
1–3*
—
NEG
Rm,Rn
0110nnnnmmmm1011
0–Rm → Rn
1
—
NEGC
Rm,Rn
0110nnnnmmmm1010
0–Rm–T → Rn,
Borrow → T
1
Borrow
SUB
Rm,Rn
0011nnnnmmmm1000
Rn–Rm → Rn
1
—
SUBC
Rm,Rn
0011nnnnmmmm1010
Rn–Rm–T → Rn,
Borrow → T
1
Borrow
SUBV
Rm,Rn
0011nnnnmmmm1011
Rn–Rm → Rn,
Underflow → T
1
Underflow
Note: The normal minimum number of execution cycles (The number in parenthesis in the
number of cycles when there is contension with preceding/following instructions).
34 RENESAS
Table 2.14 Logic Operation Instructions
Executio
n Cycles
T bit
0010nnnnmmmm1001 Rn & Rm → Rn
1
—
#imm,R0
11001001iiiiiiii R0 & imm → R0
1
—
AND.B
#imm,@(R0,GBR)
11001101iiiiiiii (R0 + GBR) & imm →
(R0 + GBR)
3
—
NOT
Rm,Rn
0110nnnnmmmm0111 ~Rm → Rn
1
—
OR
Rm,Rn
0010nnnnmmmm1011 Rn | Rm → Rn
1
—
OR
#imm,R0
11001011iiiiiiii R0 | imm → R0
1
—
OR.B
#imm,@(R0,GBR)
11001111iiiiiiii (R0 + GBR) | imm →
(R0 + GBR)
3
—
TAS.B
@Rn
0100nnnn00011011 If (Rn) is 0, 1 → T; 1 →
MSB of (Rn)
4
Test
result
TST
Rm,Rn
0010nnnnmmmm1000 Rn & Rm; if the result is
0, 1 → T
1
Test
result
TST
#imm,R0
11001000iiiiiiii R0 & imm; if the result
is 0, 1 → T
1
Test
result
TST.B
#imm,@(R0,GBR)
11001100iiiiiiii (R0 + GBR) & imm; if
the result is 0, 1 → T
3
Test
result
XOR
Rm,Rn
0010nnnnmmmm1010 Rn ^ Rm → Rn
1
—
XOR
#imm,R0
11001010iiiiiiii R0 ^ imm → R0
1
—
XOR.B
#imm,@(R0,GBR)
11001110iiiiiiii (R0 + GBR) ^ imm →
(R0 + GBR)
3
—
Instruction
Instruction Code
AND
Rm,Rn
AND
Operation
RENESAS35
Table 2.15 Shift Instructions
Instruction
Instruction Code
Operation
Execution Cycles
T bit
ROTL
Rn
0100nnnn00000100
T ← Rn ← MSB
1
MSB
ROTR
Rn
0100nnnn00000101
LSB → Rn → T
1
LSB
ROTCL
Rn
0100nnnn00100100
T ← Rn ← T
1
MSB
ROTCR
Rn
0100nnnn00100101
T → Rn → T
1
LSB
SHAL
Rn
0100nnnn00100000
T ← Rn ← 0
1
MSB
SHAR
Rn
0100nnnn00100001
MSB → Rn → T
1
LSB
SHLL
Rn
0100nnnn00000000
T ← Rn ← 0
1
MSB
SHLR
Rn
0100nnnn00000001
0 → Rn → T
1
LSB
SHLL2
Rn
0100nnnn00001000
Rn<<2 → Rn
1
—
SHLR2
Rn
0100nnnn00001001
Rn>>2 → Rn
1
—
SHLL8
Rn
0100nnnn00011000
Rn<<8 → Rn
1
—
SHLR8
Rn
0100nnnn00011001
Rn>>8 → Rn
1
—
SHLL16 Rn
0100nnnn00101000
Rn<<16 → Rn
1
—
SHLR16 Rn
0100nnnn00101001
Rn>>16 → Rn
1
—
Table 2.16 Branch Instructions
Operation
Executio
n Cycles T bit
Instruction
Instruction Code
BF
label
10001011dddddddd If T = 0, disp×2 + PC → PC; if T = 1, nop
3/1*
—
BT
label
10001001dddddddd If T = 1, disp×2 + PC → PC; if T = 0, nop
3/1*
—
BRA label
1010dddddddddddd Delayed branch, disp×2 + PC → PC
2
—
BSR label
1011dddddddddddd Delayed branch, PC → PR, disp×2 + PC
→ PC
2
—
JMP @Rm
0100mmmm00101011 Delayed branch, Rm → PC
2
—
JSR @Rm
0100mmmm00001011 Delayed branch, PC → PR, Rm → PC
2
—
RTS
0000000000001011 Delayed branch, PR → PC
2
—
Note: The execution state is three cycles when program branches, and one cycle when program
does not branch.
36 RENESAS
Table 2.17 System Control Instructions
Execution
Cycles
T bit
0000000000001000 0 → T
1
0
0000000000101000 0 → MACH, MACL
1
—
Instruction
Instruction Code
CLRT
CLRMAC
Operation
LDC
Rm,SR
0100mmmm00001110 Rm → SR
1
LSB
LDC
Rm,GBR
0100mmmm00011110 Rm → GBR
1
—
LDC
Rm,VBR
0100mmmm00101110 Rm → VBR
1
—
LDC.L
@Rm+,SR
0100mmmm00000111 (Rm) → SR, Rm + 4 → Rm
3
LSB
LDC.L
@Rm+,GBR
0100mmmm00010111 (Rm) → GBR, Rm + 4 → Rm 3
—
LDC.L
@Rm+,VBR
0100mmmm00100111 (Rm) → VBR, Rm + 4 → Rm
3
—
LDS
Rm,MACH
0100mmmm00001010 Rm → MACH
1
—
LDS
Rm,MACL
0100mmmm00011010 Rm → MACL
1
—
LDS
Rm,PR
0100mmmm00101010 Rm → PR
1
—
LDS.L
@Rm+,MACH
0100mmmm00000110 (Rm) → MACH, Rm + 4 →
Rm
1
—
LDS.L
@Rm+,MACL
0100mmmm00010110 (Rm) → MACL, Rm + 4 →
Rm
1
—
LDS.L
@Rm+,PR
0100mmmm00100110 (Rm) → PR, Rm + 4 → Rm
1
—
NOP
0000000000001001 No operation
1
—
RTE
0000000000101011 Delayed branch, stack area
→ PC/SR
4
—
SETT
0000000000011000 1 → T
1
1
SLEEP
0000000000011011 Sleep
3*
—
STC
SR,Rn
0000nnnn00000010 SR → Rn
1
—
STC
GBR,Rn
0000nnnn00010010 GBR → Rn
1
—
STC
VBR,Rn
0000nnnn00100010 VBR → Rn
1
—
STC.L
SR,@–Rn
0100nnnn00000011 Rn–4 → Rn, SR → (Rn)
2
—
STC.L
GBR,@–Rn
0100nnnn00010011 Rn–4 → Rn, GBR → (Rn)
2
—
STC.L
VBR,@–Rn
0100nnnn00100011 Rn–4 → Rn, VBR → (Rn)
2
—
STS
MACH,Rn
0000nnnn00001010 MACH → Rn
1
—
STS
MACL,Rn
0000nnnn00011010 MACL → Rn
1
—
STS
RR,Rn
0000nnnn00101010 PR → Rn
1
—
Note: The number of execution states before the chip enters the sleep state.
RENESAS37
Table 2.17 System Control Instructions (cont)
Execution
Cycles
T bit
0100nnnn00000010 Rn–4 → Rn, MACH → (Rn)
1
—
MACL,@–Rn
0100nnnn00010010 Rn–4 → Rn, MACL → (Rn)
1
—
STS.L
PR,@–Rn
0100nnnn00100010 Rn–4 → Rn, PR → (Rn)
1
—
TRAPA
#imm
11000011iiiiiiii PC/SR → stack area,
(imm×4+VBR) → PC
8
—
Instruction
Instruction Code
STS.L
MACH,@–Rn
STS.L
Operation
Note: Instruction execution cycles: The execution cycles shown in the table are minimums. The
actual number of cycles may be increased:
1. When contention occurs between instruction fetches and data access, or
2. When the destination register of the load instruction (memory → register) and the
register used by the next instruction are the same.
38 RENESAS
2.4.2
Operation Code Map
Table 2.18 is an operation code map.
Table 2.18 Operation Code Map
Instruction Code
MSB
Fx: 0000
LSBMD: 00
Fx: 0001
Fx: 0010
Fx: 0011–1111
MD: 01
MD: 10
MD: 11
0000 Rn
Fx
0000
0000 Rn
Fx
0001
0000 Rn
Fx
0010 STC
0000 Rn
Fx
0011
0000 Rn
Rm
01MD MOV.B RM,
@(R0,Rn)
MOV.W RM,
@(R0,Rn)
MOV.L RM,
@(R0,Rn)
0000 0000 Fx
1000 CLRT
SETT
CLRMAC
0000 0000 Fx
1001 NOP
DIVOU
0000 0000 Fx
1010
SR,Rn
STC
GBR,Rn
STC
VBR,Rn
0000 0000 Fx
1011 RTS
0000 Rn
Fx
1000
0000 Rn
Fx
1001
0000 Rn
Fx
1010 STS
0000 Rn
Rm
1011
0000 Rn
Rm
11MD MOV.B
@(R0,Rm),Rn
0001 Rn
Rm
disp
0010 Rn
Rm
00MD MOV.B Rm,@Rn
0010 Rn
Rm
01MD MOV.B Rm,@-Rn MOV.W Rm,@-Rn MOV.L Rm,@-Rn DIV0S Rm,Rn
0010 Rn
Rm
10MD TST
0010 Rn
Rm
11MD CMP/STR
Rm,Rn
0011 Rn
Rm
00MD CMP/EQ Rm,Rn
CMP/HS Rm,Rn
CMP/GE Rm,Rn
0011 Rn
Rm
01MD DIV1
Rm,Rn
CMP/HI Rm,Rn
CMP/GT Rm,Rn
0011 Rn
Rm
10MD SUB
Rm,Rn
SUBC
Rm,Rn
SUBV
Rm,Rn
0011 Rn
Rm
11MD ADD
Rm,Rn
ADDC
Rm,Rn
ADDV
Rm,Rn
0100 Rn
Fx
0000 SHLL
Rn
SHAL
Rn
0100 Rn
Fx
0001 SHLR
Rn
CMP/PZ Rn
SHAR
Rn
0100 Rn
Fx
0010 STS.L MACH,
@–Rn
STS.L MACL,
@–Rn
STS.L PR,
@–Rn
SLEEP
MACH,Rn STS
RTE
MACL,Rn STS
MOV.W
@(R0,Rm),Rn
PR,Rn
MOV.L
@(R0,Rm),Rn
MOV.L Rm,@(disp:4,Rn)
Rm,Rn
MOV.W Rm,@Rn
AND
Rm,Rn
XTRCT Rm,Rn
MOV.L Rm,@Rn
XOR
Rm,Rn
OR
Rm,Rn
MULU
Rm,Rn
MULS
Rm,Rn
RENESAS39
Table 2.18 Operation Code Map (cont)
Instruction Code
Fx: 0000
MSB
LSB MD: 00
Fx: 0001
Fx: 0010
Fx: 0011–1111
MD: 01
MD: 10
MD: 11
STC.L
GBR,@–Rn
STC.L
VBR,@–Rn
0100 Rn
Fx
0011 STC.L
SR,@–Rn
0100 Rn
Fx
0100 ROTL
Rn
0100 Rn
Fx
0101 ROTR
Rn
CMP/PL Rn
ROTCR Rn
0100 Rm
Fx
0110 LDS.L
@Rm+,MACH
LDS.L
@Rm+,MACL
LDS.L
@Rm+,PR
0100 Rm
Fx
0111 LDC.L
@Rm+,SR
LDC.L
@Rm+,GBR
LDC.L
@Rm+,VBR
0100 Rn
Fx
1000 SHLL2 Rn
SHLL8 Rn
SHLL16 Rn
0100 Rn
Fx
1001 SHLR2 Rn
SHLR8 Rn
SHLL16 Rn
0100 Rm
Fx
1010 LDS
0100 Rm/Rn Fx
1011 JSR
0100 Rm
1100
Fx
ROTCL Rn
Rm,MACH LDS
Rm,MACL LDS
@Rm
TAS.B @Rn
Rm,Sr
LDC
JMP
Rm,PR
@Rm
0100 Rm
Fx
1101
0100 Rn
Fx
1110 LDC
0100 Rn
Rm 1111 MAC.W @Rm+,@Rn+
0101 Rn
Rm disp
0110 Rn
Rm 00MD MOV.B @Rm,Rn MOV.W @Rm,Rn MOV.L @Rm,Rn MOV
Rm,Rn
0110 Rn
Rm 01MD MOV.B
NOT
Rm,Rn
NEG
Rm,Rn
Rm,GBR LDC
Rm,VBR
MOV.L @(disp:4,Rm),Rn
MOV.W
@Rm+,R
n
MOV.L
@Rm+,R
@Rm+,R
n
n
0110 Rn
Rm 10MD SWAP.B
@Rm+,Rn
SWAP.W
@Rm+,Rn
NEGC
0110 Rn
Rm 11MD EXTU.B Rm,Rn
EXTU.W Rm,Rn
EXTS.B Rm,Rn
0111 Rn
imm
1000 00MD Rn
disp
1000 01MD Rm disp
ADD
MOV.B R0,
MOV.W R0,
@(disp:4,Rn) @(disp:4,Rn)
MOV.B
@(disp:4,
Rm),R0
imm/disp CMP/EQ
#imm:8,R0
1000 11MD
imm/disp
disp
MOV.W R0,
@(disp:4,
Rn),R0
BT
disp:8
MOV.W @(disp:8,PC),Rn
1010
disp
BRA
disp:12
1011
disp
BSR
disp:12
40 RENESAS
EXTS.W Rm,Rn
#imm:8,Rn
1000 10MD
1001 Rn
Rm,Rn
BF
disp:8
Table 2.18 Operation Code Map (cont)
Instruction Code
MSB
Fx: 0000
LSB MD: 00
Fx: 0001
Fx: 0010
Fx: 0011–1111
MD: 01
MD: 10
MD: 11
1100 00MD
imm/disp
1100 01MD
disp
MOV.B
@(disp:8,
GBR),R0
MOV.W
@(disp:8,
GBR),R0
MOV.L
@(disp:8,
GBR),R0
MOVA
@(disp:8,
PC),R0
1100 10MD
imm
TST
#imm:8,R0
AND
#imm:8,R0
XOR
#imm:8,R0
OR
#imm:8,R0
1100 11MD
imm
TST.B
#imm:8,
@(R0,GBR)
AND.B
#imm:8,
@(R0,GBR)
XOR.B
#imm:8,
@(R0,GBR)
OR.B
#imm:8,
@(R0,GBR)
1101 Rn
disp
MOV.L @(disp:8,PC),Rn
1110 Rn
imm
MOV
MOV.B R0,@
MOV.W R0,@
MOV.L R0,@
TRAPA #imm:8
(disp:8,GBR) (disp:8,GBR) (disp:8,GBR)
1111
...
2.5
CPU State
2.5.1
State Transitions
#imm:8,Rn
The CPU has five processing states: reset, exception processing, bus release, program execution
and power-down. The transitions between the states are shown in figure 2.6. For more information
on the reset and exception processing states, see section 4, Exception Processing. For details on
the power-down states, see section 19, Power Down States.
RENESAS41
From any state when
RES = 0 and NMI = 1
From any state when
RES = 0 and NMI = 0
RES = 0, NMI = 0
Power-on reset state
Manual reset state
RES = 0, NMI = 1
When an interrupt source
or DMA address error occurs
RES = 1,
NMI = 0
RES = 1,
NMI = 1
Reset states
Exception processing state
Bus request
cleared
NMI interrupt
source occurs
Bus request
generated
Bus release state
Bus request
generated
Bus request
generated
Exception
processing
source occurs
Exception
processing
ends
Bus request
cleared
Program execution state
Bus request
cleared
Sleep mode
SBY bit cleared for
SLEEP instruction
SBY bit set
for SLEEP
instruction
Standby mode
Power-down state
Figure 2.6 Transitions Between Processing States
Reset State: In the reset state the CPU is reset. This occurs when the RES pin level goes low.
When the NMI pin is high, the result is a power-on reset; when it is low, a manual reset will
occur.When turning on the power, make sure to carry out a power-on reset.
42 RENESAS
On a power-on reset, all CPU internal states and on-chip peripheral module registers are
initialized. In a manual reset, all CPU internal states and on-chip peripheral module registers, with
the exception of the bus state controller (BSC) and pin function controller (PFC), are initialized.
On a manual reset, the BSC is not initialized, so the refresh operation will continue.
Exception Processing State: Exception processing is a transient state that occurs when the CPU’s
processing state flow is altered by exception processing sources such as resets or interrupts.
For a reset, the initial values of the program counter PC (execution start address) and stack pointer
SP are fetched from the exception processing vector table and stored; the CPU then branches to
the execution start address and execution of the program begins.
For an interrupt, the stack pointer (SP) is accessed and the program counter (PC) and status
register (SR) are saved to the stack area. The exception service routine start address is fetched
from the exception processing vector table; the CPU then branches to that address and the program
starts executing, thereby entering the program execution state.
Program Execution State: In the program execution state, the CPU sequentially executes the
program.
Power-Down State: In the power-down state, the CPU operation halts and power consumption
declines. The SLEEP instruction places the CPU in the power-down state. This state has two
modes: sleep mode and standby mode. This is described in more detail in section 2.5.1, PowerDown State.
Bus Release State: In the bus release state, the CPU releases rights to the bus to the device that
has requested them.
2.5.2
Power-Down State
In addition to the ordinary program execution states, the CPU also has a power-down state in
which CPU operation halts and power consumption is lowered. There are two power-down state
modes: sleep mode and standby mode.
Sleep Mode: When the standby bit SBY (in the standby control register SBYCR) is cleared to 0
and a SLEEP instruction executed, the CPU moves from program execution state to sleep mode.
In the sleep mode, the CPU halts and the contents of its internal registers and the data in on-chip
RAM are stored. The on-chip peripheral modules other than the CPU do not halt in the sleep
mode.
To return from sleep mode, use a reset, any interrupt, or a DMA address error; the CPU returns to
ordinary program execution state through the exception processing state.
RENESAS43
Software Standby Mode: To enter the standby mode, set the standby bit SBY (in the standby
control register SBYCR) to 1 and execute a SLEEP instruction. In standby mode, all CPU, on-chip
peripheral module and oscillator functions are halted. CPU internal register contents and on-chip
RAM data are held.
To return from standby mode, use a reset or an external NMI interrupt. For resets, the CPU returns
to ordinary program execution state through the exception processing state when placed in a reset
state during oscillator stabilization time. For NMI interrupts, the CPU returns to ordinary program
execution state through the exception processing state after the oscillator stabilization time has
elapsed. In this mode, power consumption drops markedly, since the oscillator stops.
Table 2.19 Power-Down State
State
On-chip
CPU
Peripheral RegiClock CPU Modules sters RAM I/O Ports
Mode
Conditions
Sleep
mode
Execute SLEEP Run
instruction with
SBY bit cleared
to 0 in SBYCR
Halt Run
Held
Held Held
Canceling
1. Interrupt
2. DMA address
error
3. Power-on
reset
4. Manual reset
Stand
by
mode
Execute SLEEP Halt
instruction with
SBY bit set to 1
in SBYCR
Halt Halt and
initialize*
Held
Note: Differs depending on the peripheral module and pin.
44 RENESAS
Held Held or
1. NMI
high-Z*
2. Power-on
(selectable)
reset
3. Manual reset
Section 3 Operating Modes
3.1
Types of Operating Modes and Their Selection
The SH7020 and SH7021 operate in one of four operating modes (modes 0, 1, 2, and 7). Modes 0
and 1 differ in the bus width of memory area 0. The mode is selected by the mode pins (MD2–
MD0) as indicated in table 3.1. Do not change the mode selection while the chip is operating.
Table 3.1
Operating Mode Selection
Pin Settings
Operating Mode
MD2
MD1
MD0
Mode Name
Bus Width of Area 0
Mode 0*
2
0
0
0
MCU mode 0
8 bits
Mode 1*
2
0
0
1
MCU mode 1
16 bits
Mode 2
0
1
0
MCU mode 2
On-chip ROM
Mode 7*
1
1
1
PROM mode
—
Notes : 1.SH7021 PROM version only
2.Only modes 0 and 1are available in the SH7020 ROMless version.
3.2
Operating Mode Descriptions
3.2.1
Mode 0 (MCU Mode 0)
In mode 0, memory area 0 has an eight-bit bus width. For the memory map, see section 8, Bus
State Controller.
3.2.2
Mode 1 (MCU Mode 1)
In mode 1, memory area 0 has a 16-bit bus width.
3.2.3
Mode 2 (MCU Mode 2)
In mode 2, memory area 0 is assigned to the on-chip ROM.
3.2.4
Mode 7 (PROM Mode)
Mode 7 is a PROM mode. In this mode, the EPROM can be programmed.For details,see section
16, ROM. Do not set to mode 7 unless the product is the SH7021(PROM version).
RENESAS 45
Section 4 Exception Processing
4.1
Overview
4.1.1
Exception Processing Types and Priorities
As figure 4.1 indicates, exception processing may be caused by a reset, address error, interrupt, or
instruction. Exception sources are prioritized as indicated in figure 4.1. If two or more exceptions
occur simultaneously, they are accepted and processed in the priority order shown.
RENESAS 47
Priority
Reset
• Power-on reset
• Manual reset
Address
error
• CPU address error
• DMA address error
• NMI
• User break
• IRQ
Interrupt
Exception
source
• On-chip module
Instruction
Notes: 1.
2.
High
• IRQ0–IRQ7
• Direct memory access
controller
• 16-bit integrated-timer
pulse unit
• Serial communications
interface
• Parity control unit
(part of the bus controller)
• Watchdog timer
• DRAM refresh control
unit (part of the bus
controller)
• Trap instruction
• TRAPA instruction
• General illegal
instruction
• Illegal slot
instruction
• Undefined code
• Undefined instruction
or instruction that
rewrites the PC*1
Low
placed directly after
a delayed branch
instruction*2
The instructions that rewrite the PC are JMP, JSR, BRA, BSR, RTS, RTE, BT, BF,
and TRAPA.
The delayed branch instructions are JMP, JSR. BRA. BSR, RTS, and RTE.
Figure 4.1 Exception Source Types and Priority
48 RENESAS
4.1.2
Exception Processing Operation
Exception sources are detected at the times indicated in table 4.1, whereupon processing starts.
Table 4.1
Exception Source Detection and Time of the Start of Processing
Exception Type
Source Detection and Time of the Start of Processing
Reset
Power-on
Low-to-high transition at pin RES when NMI is high
Manual
Low-to-high transition at pin RES when NMI is low
Address error
Detected when instruction is decoded and starts after the
instruction that was executing prior to this point is completed.
Interrupt
Detected when instruction is decoded and starts after the
instruction that was executing prior to this point is completed.
Instruction
Trap instruction Starts when a trap instruction (TRAPA) is executed.
General illegal
instruction
Starts when undefined code is decoded at a position other than
directly after a delayed branch instruction (a delay slot).
Illegal slot
instruction
Starts when undefined code or an instruction that rewrites the PC is
decoded directly after a delayed branch instruction (in a delay slot).
When exception processing begins, the CPU operates as follows:
Resets: The initial values of the program counter (PC) and stack pointer (SP) are read from the
exception vector table (the respective PC and SP values are H'00000000 and H'00000004 for a
power-on reset and H'00000008 and H'0000000C for a manual reset). For more information on the
exception vector table, see section 4.1.3, Exception Vector Table. Next, the vector base register
(VBR) is cleared to zero and interrupt mask bits (I3–I0) in the status register (SR) are set to 1111.
Program execution starts from the PC address read from the exception vector table.
Address Errors, Interrupts and Instructions: SR and PC are pushed onto the stack indicated in
R15. For interrupts, the interrupt priority level is written in the interrupt mask bits (I3–I0). For
address errors and instructions, bits I3–I0 are not affected. Next, the start address is fetched from
the exception vector table, and program execution starts from this address.
4.1.3
Exception Process Vector Table
Before exception processing can execute, the exception vector table must be set in memory. The
exception processing vector table holds the start addresses of exception service routines (the table
for reset exception processing stores initial PC and SP values). Different vector numbers and
vector table address offsets are assigned to different exception sources. The vector table addresses
are calculated from the corresponding vector numbers and vector address offsets. In exception
processing, the exception service routine start address is fetched from the exception vector table
indicated by this vector table address.
RENESAS 49
Table 4.2 lists vector numbers and vector table address offsets. Table 4.3 shows how to calculate
vector table addresses.
Table 4.2
Exception Process Vector Table
Vector
Number
Vector table Address Offset
PC
0
H'00000000–H'00000003
SP
1
H'00000004–H'00000007
PC
2
H'00000008–H'0000000B
SP
3
H'0000000C–H'0000000F
General illegal instruction
4
H'00000010–H'00000013
(Reserved for system use)
5
H'00000014–H'00000017
Illegal slot instruction
6
H'00000018–H'0000001B
(Reserved for system use)
7
H'0000001C–H'0000001F
8
H'00000020–H'00000023
CPU address error
9
H'00000024–H'00000027
DMA address error
10
H'00000028–H'0000002B
NMI
11
H'0000002C–H'0000002F
User
break
12
H'00000030–H'00000033
(Reserved for system use)
13–31
H'00000034–H'00000037 to H'0000007C–
H'0000007F
Trap instruction (user
vectors)
32–63
H'00000080–H'00000083 to H'000000FC–
H'000000FF
IRQ0
64
H'00000100–H'00000103
IRQ1
65
H'00000104–H'00000107
IRQ2
66
H'00000108–H'0000010B
IRQ3
67
H'0000010C–H'0000010F
IRQ4
68
H'00000110–H'00000113
IRQ5
69
H'00000114–H'00000117
IRQ6
70
H'00000118–H'0000011B
IRQ7
71
H'0000011C–H'0000011F
On-chip
modules*
72–255
H'00000120–H'00000123 to H'000003FC–
H'000003FF
Exception Source
Power-on reset
Manual reset
Interrupts
Interrupts
Note: See table 5.3, Interrupt Exception Processing Vectors and Rankings, in section 5, Interrupt
Controller, for details on vector numbers and vector table address offsets of individual onchip peripheral module interrupts.
50 RENESAS
Table 4.3
Calculation of Exception Vector table Addresses
Exception Source
Calculation of Vector table Addresses
Reset
(Vector table address) = (vector table address offset) =
(vector number) × 4
Address error, interrupt, instructions
(Vector table address) = VBR + (vector table address offset)
= VBR + (vector number) × 4
Note: VBR: Vector base register. For vector table address offsets and vector numbers, see table
4.2.
4.2
Reset
4.2.1
Reset Types
A reset is the highest-priority exception. There are two types of reset: power-on reset and manual
reset. As table 4.4 shows, a power-on reset initializes the internal state of the CPU and all registers
of the on-chip peripheral modules. A manual reset initializes the internal state of the CPU and all
registers of the on-chip peripheral modules except the bus state controller (BSC), pin function
controller (PFC) and I/O ports (I/O).
Table 4.4
Reset Types
Transition Conditions
Internal State
Reset
NMI
RES
CPU
On-Chip Peripheral Module
Power-on Reset
High
Low
Initialize
Initialize
Manual Reset
Low
Low
Initialize
Initialize all except BSC, PFC and I/O
4.2.2
Power-On Reset
When the NMI pin is high, a low input at the RES pin drives the chip into the power-on reset state.
The RES pin should be driven low while the clock pulse generator (CPG) is stopped (or while the
CPG is operating during the oscillation settling time) for at least 20 tcyc to assure that the LSI is
reset. A power-on reset initializes the internal state of the CPU and all registers of the on-chip
peripheral modules. For pin states in the power-on reset state, see appendix B, Pin States.
While the NMI pin remains high, if the RES pin is held low for a certain time then driven high in
the power-on state, power-on reset exception processing begins. The CPU then:
1. Reads the start address (initial PC value) from the exception vector table.
2. Reads the initial stack pointer value (SP) from the exception vector table.
RENESAS 51
3. Clears the vector base register (VBR) to H'00000000, and sets interrupt mask bits I3–I0 in the
status register (SR) to H'F (1111).
4. Loads the values read from the exception vector table into PC and SP and starts program
execution.
Further, make sure to carry out a power-on reset when turning on the power of the system.
4.2.3
Manual Reset
When the NMI pin is high, a low input at the RES pin drives the chip into the manual reset state.
To be assured of resetting the LSI, drive the RES pin low for at least 20 tcyc. A manual reset
initializes the internal state of the CPU and all registers of the on-chip peripheral modules except
the bus state controller, pin function controller and I/O ports. Since a manual reset does not affect
the bus state controller, the DRAM refresh control function operates even if the manual reset state
continues for a long time. When a manual reset is performed during the bus cycle, the manual
reset exception processing waits for the end of the bus cycle before beginning. The manual reset
thus cannot be used to abort the bus cycle. For the pin states during the manual reset state, see
appendix B, Pin States.
While the NMI pin remains low, if the RES pin is held low for a certain time then driven high in
the manual reset state, manual reset exception processing begins. The CPU carries out the same
operations as for a power-on reset.
4.3
Address Errors
4.3.1
Address Error Sources
Address errors occur during instruction fetches and data reading/writing as shown in table 4.5.
52 RENESAS
Table 4.5
Address Error Sources
Bus Cycle
Type
Bus Master
Instruction fetch CPU
Data read/write
CPU or DMAC
Operation
Address Error
Instruction fetch from even address
None (normal)
Instruction fetch from odd address
Address error
Instruction fetch from outside on-chip
peripheral module space
None (normal)
Instruction fetch from on-chip peripheral
module space
Address error
Access to word data from even address
None (normal)
Access to word data from odd address
Address error
Access to long word data aligned on long
word boundary
None (normal)
Access to long word data not aligned on
long word boundary
Address error
Access to word or byte data in on-chip
peripheral module space*
None (normal)
Access to long word data in 16-bit on-chip
peripheral module space*
None (normal)
Access to long word data in 8-bit on-chip
peripheral module space*
Address error
Note: See section 8, Bus State Controller, for details on the on-chip peripheral module space.
4.3.2
Address Error Exception Processing
When an address error occurs, address error exception processing starts after both the bus cycle
that caused the address error and the instructions that were being executed at that time have been
completed. The CPU then:
1. Pushes the SR onto the stack.
2. Pushes the program counter onto the stack. The PC value saved is the top address of the
instruction following the last instruction to be executed.
3. Fetches the exception service routine start address from the exception vector table for the
address error that occurred and starts program execution from that address. The branch that
occurs here is not a delayed branch.
RENESAS 53
4.4
Interrupts
4.4.1
Interrupt Sources
Table 4.6 lists the types of interrupt exception processing sources (NMI, user break, IRQ, on-chip
peripheral module).
Table 4.6
Interrupt Sources
Interrupt
Requesting Pin or Module
Number of Sources
NMI
NMI pin (external input)
1
User break
User break controller
1
IRQ
IRQ0–IRQ7 pin (external input)
8
On-chip
Direct Memory Access Controller
4
16-bit integrated-timer pulse unit
15
Serial communications interface
8
Watchdog timer
1
Bus state controller
2
Each interrupt source has a different vector number and vector address offset value. See table 5.3,
Interrupt Exception Vectors and Rankings, in section 5, Interrupt Controller, for details on vector
numbers and vector table address offsets.
4.4.2
Interrupt Priority Rankings
Interrupt sources are assigned priorities. When multiple interrupts occur at the same time, the
interrupt controller (INTC) ascertains their priorities and starts exception processing based on its
findings. Priorities from 16–0 can be assigned, with 0 the lowest level and 16 the highest. The
NMI has priority level 16 and cannot be masked. NMI is always accepted. The user break priority
level is 15. The IRQ and on-chip peripheral module interrupt priority levels can be set in interrupt
priority level registers A–E (IPRA–IPRE) as shown in table 4.7. Priority levels 0–15 can be set.
See section 5.3.1, Interrupt Priority Level Registers A–E (IPRA–IPRE), for details.
Table 4.7
Interrupt Priority Rankings
Type
Priority
Comments
NMI
16
Fixed and unmaskable
User break
15
Fixed
IRQ and on-chip peripheral modules 0–15
54 RENESAS
Set in interrupt priority level registers A–E
(IPRA–IPRE)
4.4.3
Interrupt Exception Processing
When an interrupt is generated, the INTC ascertains the interrupt rankings. NMI is always
accepted, but other interrupts are only accepted if their ranking is higher than the ranking set in the
interrupt mask bits (I3–I0) of the SR.
When an interrupt is accepted, interrupt exception processing begins. In the interrupt exception
processing sequence, the SR and PC are pushed onto the stack, and the priority level of the
accepted interrupt is copied to the interrupt mask level bits (I3–I0) in the SR. In NMI exception
processing, the priority ranking is 16 but the value 15 (H'F) is stored in I3–I0. The exception
service routine start address for the accepted interrupt is fetched from the exception vector table
and the program branches to that address and starts executing. For further information on
interrupts, see section 5.4, Interrupt Operation.
4.5
Instruction Exceptions
4.5.1
Types of Instruction Exceptions
Table 4.8 shows the three types of instruction that start exception processing (trap instructions,
illegal slot instructions, and general illegal instructions).
Table 4.8
Types of Instruction Exceptions
Type
Source Instruction
Comments
Trap instruction
TRAPA
—
Illegal slot
instruction
Undefined code or instruction
that rewrites the PC located
immediately after a delayed
branch instruction (delay slot)
Delayed branch instructions are: JMP, JSR,
BRA, BSR, RTS, RTE. Instructions that
rewrite the PC are: JMP, JSR, BRA, BSR,
RTS, RTE, BT, BF and TRAPA
General illegal
instructions
Undefined code in other than
delayed slot
—
4.5.2
Trap Instruction
Trap instruction exception processing is carried out when a trap instruction (TRAPA) is executed.
The CPU then:
1. Saves the status register by pushing register contents onto the stack.
2. Pushes the program counter value onto the stack. The PC value saved is the top address of the
next instruction after the TRAPA instruction.
3. Reads an exception processing service routine start address from the vector table
corresponding to a vector number specified in the TRAPA instruction, branches to that
address, and starts program execution. The branch is not a delayed branch.
RENESAS 55
4.5.3
Illegal Slot Instruction
An instruction located immediately after a delayed branch instruction is called an “instruction
placed in a delay slot.” If an undefined instruction is located in a delay slot, illegal slot instruction
exception processing begins executing when the undefined code is decoded. Illegal slot instruction
exception processing also begins when the instruction located in the delay slot is an instruction
that rewrites the program counter. In this case, exception processing begins when the instruction
that rewrites the PC is decoded. The CPU performs illegal slot exception processing as follows:
1. Saves the status register onto the stack.
2. Pushes the program counter value onto the stack. The PC value saved is the branch destination
address of the delayed branch instruction immediately before the instruction that contains the
undefined code or rewrites the PC.
3. Fetches an exception processing service routine start address from the vector table
corresponding to the exception that occurred, branches to that address and the program starts
executing. The branch is not a delayed branch.
4.5.4
General Illegal Instructions
If an undefined instruction located other than a delay slot (immediately after a delayed branch
instruction) is decoded, general illegal instruction exception processing is executed. The CPU
follows the same procedure as for illegal slot exception processing, except that the program
counter (PC) value pushed on the stack in general illegal instruction exception processing is the
top address of the illegal instruction with the undefined code.
56 RENESAS
4.6
Cases in Which Exceptions Are Not Accepted
In some cases, address errors and interrupts that directly follow a delayed branch instruction or
interrupt-disabled instruction are not accepted immediately. Table 4.9 lists these cases. When this
occurs, the exception is accepted when an instruction that can accept the exception is decoded.
Table 4.9
Cases in Which Exceptions Are Not Accepted
Exception Source
Case
Immediately after delayed branch
instruction*1
Immediately after interrupt-disabled
instruction*2
Address Error
Interrupt
X
X
O
X
X: Not accepted
O: Accepted
Notes: 1. Delayed branch instructions: JMP, JSR, BRA, BSR, RTS, RTE
2. Interrupt-disabled instructions: LDC, LDC.L, STC, STC.L, LDS, LDS.L, STS, STS.L
4.6.1
Immediately after Delayed Branch Instructions
Address errors and interrupts are not accepted when an instruction in a delay slot immediately
following a delayed branch instruction is decoded. The delayed branch instruction and the
instruction in the delay slot are therefore always executed one after the other. Exception
processing is never inserted between them.
4.6.2
Immediately after Interrupt-Disabling Instructions
Interrupts are not accepted when the instruction immediately following an interrupt-disabled
instruction is decoded. Address errors are accepted, however.
RENESAS 57
4.7
Stack Status after Exception Processing
Table 4.10 shows the stack after exception processing.
Table 4.10 Stack after Exception Processing
Type
Address
error
Stack Status
Type
Stack Status
Interrupt
SP
Address of
instruction
Upper 16 bits
after instruction that has
finished
executing
SP
Address of
instruction
Upper 16 bits
after instruction that
has finished
executing
Lower 16 bits
Lower 16 bits
SR
SR
Upper 16 bits
Lower 16 bits
Lower 16 bits
Trap
instruction
SP
Address of
instruction
Upper 16 bits
after TRAPA
instruction
Illegal
slot
instruction
Lower 16 bits
SR
Upper 16 bits
SP
Branch
destination
address of
delayed
branch
instuction
Lower 16 bits
Upper 16 bits
Lower 16 bits
Upper 16 bits
SR
Upper 16 bits
Lower 16 bits
General
illegal
instruction
SP
Start address of
illegal
instruction
Upper 16 bits
Lower 16 bits
SR
Upper 16 bits
Lower 16 bits
Note: Stack status is based on a bus width of 16 bits.
58 RENESAS
4.8
Notes
4.8.1
Value of the Stack Pointer (SP)
An address error occurs if the stack is accessed for exception processing when the value of the
stack pointer (SP) is not a multiple of four. Therefore, a multiple of four should always be stored
in SP.
4.8.2
Value of the Vector Base Register (VBR)
An address error occurs if the vector table is accessed for exception processing when the value of
the vector base register (VBR) is not a multiple of four. Therefore, VBR should always be set to a
multiple of four.
4.8.3
Address Errors that Are Caused by Stacking During Address Error Exception
Processing
When the stack pointer is not a multiple of four, address errors will occur in the exception
processing (interrupt, etc.) stacking. After the exception processing ends, the CPU will then shift
to address error exception processing. An address error will also occur during the address error
exception processing stacking, but the CPU is set up to ignore the address error so that it can avoid
an infinite series of address errors. This allows it to shift program control to the address error
exception service routine and process the error.
When an address error does occur in exception processing stacking, the stacking bus cycle (write)
is executed. In SR and PC stacking, four is subtracted from each of the SPs so the SP values are
not multiples of four after stacking either. Since the address value output during stacking is the SP
value, the address that produced the error is exactly what is output. In such cases, the stacked write
data will be undefined.
RENESAS 59
Section 5 Interrupt Controller (INTC)
5.1
Overview
The interrupt controller (INTC) determines the priority of interrupt sources and controls interrupt
requests to the CPU. INTC has registers for assigning priority levels to interrupt sources. These
registers handle interrupt requests according to user-established priorities.
5.1.1
Features
The interrupt controller has the following features:
•
•
•
16 settable priority levels: Five interrupt priority registers can set 16 levels of interrupt
priorities for IRQ and on-chip peripheral interrupt sources.
The INTC has an NMI input level T bit that indicates NMI pin status. By reading this bit with
the interrupt exception service routine, the pin status can be checked for use in a noise
canceller function.
The interrupt controller can notify external devices (via the IRQOUT pin) that an onchip
interrupt has been occured. In this way an external device can, for example, be informed if an
on-chip interrupt occurs while the chip is operating in a bus-released mode and the bus has
been requested.
5.1.2
Block Diagram
Figure 5.1 is a block diagram of the interrupt controller.
RENESAS 61
IRQOUT
NMI
IRQ0
IRQ1
IRQ2
IRQ3
Input
control
Priority
decision
logic
IRQ4
IRQ5
IRQ6
Comparator
Interrupt request
IRQ7
SR
UBC
DMAC
ITU
SCI
PRT
WDT
REF
(Interrupt request)
(Interrupt request)
(Interrupt request)
(Interrupt request)
(Interrupt request)
(Interrupt request)
(Interrupt request)
I3
I2
I1
I0
CPU
IPR
ICR
Module bus
Bus
interface
Internal bus
IPRA–IPRE
INTC
UBC: User break controller
DMAC: Direct memory access controller
ITU: 16-bit integrated-timer pulse unit
SCI: Serial communications interface
PRT: Parity control unit of BSC
WDT: Watchdog timer
REF: DRAM refresh control unit of BSC
ICR: Interrupt control register
IPRA–IPRE: Interrupt priority registers A–E
SR: Status register
Figure 5.1 Block Diagram of the Interrupt Controller
62 RENESAS
5.1.3
Pin Configuration
INTC pins are summarized in table 5.1.
Table 5.1
INTC Pin Configuration
Name
Abbr.
I/O
Function
Nonmaskable interrupt input
pin
NMI
I
Inputs a non-maskable interrupt request
signal
Interrupt request input pins
IRQ0–IRQ7
I
Inputs maskable interrupt request signals
Interrupt request output pin
IRQOUT
O
Outputs a signal indicating an interrupt
source has occurred.
5.1.4
Registers
The interrupt controller has six registers as listed in table 5.2. These registers are used for setting
interrupt priority levels and controlling the detection of external interrupt input signals.
Table 5.2
Interrupt Controller Register Configuration
Name
Abbr.
R/W
Address* 2
Initial Value
Bus width
Interrupt priority register A
IPRA
R/W
H'5FFFF84
H'0000
8, 16, 32
Interrupt priority register B
IPRB
R/W
H'5FFFF86
H'0000
8, 16, 32
Interrupt priority register C
IPRC
R/W
H'5FFFF88
H'0000
8, 16, 32
Interrupt priority register D
IPRD
R/W
H'5FFFF8A
H'0000
8, 16, 32
Interrupt priority register E
IPRE
R/W
H'5FFFF8C
H'0000
8, 16, 32
Interrupt control register
ICR
R/W
H'5FFFF8E
*1
8, 16, 32
Note:
5.2
1. H'8000 when pin NMI is high, H'0000 when pin NMI is low.
2. Only the values of bits A27–A24 and A8–A0 are valid; bits A23–A9 are ignored. For
details on the register addresses, see section 8.3.5, Description of Areas.
Interrupt Sources
There are four types of interrupt sources: NMI, user break, IRQ, and on-chip peripheral module
interrupts.
Interrupt rankings are expressed as priority levels (0–16), with 0 the lowest and 16 the highest. An
interrupt set to level 0 is masked.
RENESAS 63
5.2.1
NMI Interrupts
NMI is the highest-priority interrupt (level 16) and is always accepted. Input at the NMI pin is
edge-sensed. Either the rising or falling edge can be selected by setting the NMI edge select bit
(NMIE) in the interrupt control register (ICR). NMI interrupt exception processing sets the
interrupt mask level bits (I3–I0) in the status register (SR) to level 15.
5.2.2
User Break Interrupt
A user break interrupt occurs when a break condition is satisfied in the user break controller
(UBC). A user break interrupt has priority level 15. User break interrupt exception processing sets
the interrupt mask level bits (I3–I0) in the status register (SR) to level 15. For further details about
the user break interrupt, see section 6, User Break Controller.
5.2.3
IRQ Interrupts
IRQ interrupts are requested by input from pins IRQ0-IRQ7. IRQ sense select bits 0–7 (IRQ0S–
IRQ7S) in the interrupt control register (ICR) can select low-level sensing or falling-edge sensing
for each pin independently. Interrupt priority registers A and B (IPRA and IPRB) can select
priority levels from 0–15 for each pin. IRQ interrupt exception processing sets the interrupt mask
level bits (I3–I0) in the status register (SR) to the priority level value of the IRQ interrupt that was
accepted.
5.2.4
On-Chip Interrupts
On-chip interrupts are interrupts generated by the following 5 on-chip peripheral modules:
•
•
•
•
•
Direct memory access controller (DMAC)
16-bit integrated-timer pulse unit (ITU)
Serial communications interface (SCI)
Bus state controller (BSC)
Watchdog timer (WDT)
A different interrupt vector is assigned to each interrupt source, so the exception service routine
does not have to decide which interrupt has occurred. Priority levels 0–15 can be assigned to
individual on-chip peripheral module in interrupt priority registers C–E (IPRC–IPRE). On-chip
interrupt exception processing sets the interrupt mask level bits (I3–I0) in the status register (SR)
to the priority level value of the on-chip interrupt that was accepted.
64 RENESAS
5.2.5
Interrupt Exception Vectors and Priority Rankings
Table 5.3 lists the vector numbers, vector table address offsets, and interrupt priority order of the
interrupt sources.
Each interrupt source is allocated a different vector number and vector table address offset. The
vector table address is calculated from this vector number and address offset. In interrupt
exception processing, the exception service routine start address is fetched from the vector table
indicated by this vector table address. See table 4.3, Calculation of Exception Vector table
Addresses, in section 4, Exception Processing, for details on this calculation.
Arbitrary interrupt priority levels between 0 and 15 can be assigned to IRQ and on-chip peripheral
module interrupt sources by setting interrupt priority registers A–E (IPRA–IPRE) for each pin or
module. The interrupt sources for IPRC–IPRE, however, must be ranked in the order listed under
Priority Within Module in table 5.3 and cannot be changed. A reset assigns priority level 0 to IRQ
and on-chip peripheral module interrupts. If the same priority level is assigned to two or more
interrupt sources, and interrupts from those sources occur simultaneously, their priority order is
the default priority order indicated at the right in table 5.3.
RENESAS 65
Table 5.3
Interrupt Exception Vectors and Rankings
Interrupt Priority Order IPR (bit
Interrupt Source (initial value) numbers)
Priority VecWithin tor Address Offset in
Module No. Vector table
NMI
16
—
—
11
H'0000002C–H'0000002F High
User break
15
—
—
12
H'00000030–H'00000033
IRQ0
0–15 (0)
IPRA (15–12) —
64
H'00000100–H'00000103
IRQ1
0–15 (0)
IPRA (11–8)
—
65
H'00000104–H'00000107
IRQ2
0–15 (0)
IPRA (7–4)
—
66
H'00000108–H'0000010B
IRQ3
0–15 (0)
IPRA (3–0)
—
67
H'0000010C–H'0000010F
IRQ4
0–15 (0)
IPRB (15–12) —
68
H'00000110–H'00000113
IRQ5
0–15 (0)
IPRB (11–8)
—
69
H'00000114–H'00000117
IRQ6
0–15 (0)
IPRB (7–4)
—
70
H'00000118–H'0000011B
IRQ7
0–15 (0)
IPRB (3–0)
—
71
H'0000011C–H'0000011F
DMAC0 DEI0
0–15 (0)
IPRC (15–12) 3
72
H'00000120–H'00000123
2
73
H'00000124–H'00000127
1
74
H'00000128–H'0000012B
0
75
H'0000012C–H'0000012F
3
76
H'00000130–H'00000133
2
77
H'00000134–H'00000137
1
78
H'00000138–H'0000013B
0
79
H'0000013C–H'0000013F
3
80
H'00000140–H'00000143
IMIB0
2
81
H'00000144–H'00000147
OVI0
1
82
H'00000148–H'0000014B
Reserved
0
83
H'0000014C–H'0000014F
3
84
H'00000150–H'00000153
IMIB1
2
85
H'00000154–H'00000157
OVI1
1
86
H'00000158–H'0000015B
Reserved
0
87
H'0000015C–H'0000015F
IPRD (15–12) 3
88
H'00000160–H'00000163
IMIB2
2
89
H'00000164–H'00000167
OVI2
1
90
H'00000168–H'0000016B
Reserved
0
91
H'0000016C–H'0000016F
Reserved
DMAC1 DEI1
Reserved
DMAC2 DEI2
0–15 (0)
IPRC (11–8)
Reserved
DMAC3 DEI3
Reserved
ITU0
ITU1
ITU2
IMIA0
IMIA1
IMIA2
66 RENESAS
0–15 (0)
0–15 (0)
0–15 (0)
IPRC (7–4)
IPRC (3–0)
Default
Priority
Order
Table 5.3
Interrupt Exception Vectors and Rankings (cont)
Interrupt Priority Order IPR (bit
Interrupt Source (initial value) numbers)
Priority VecWithin tor Address Offset in
Module No. Vector table
ITU3
3
92
H'00000170–H'00000173
IMIB3
2
93
H'00000174–H'00000177 _
OVI3
1
94
H'00000178–H'0000017B
Reserved
0
95
H'0000017C–H'0000017F
3
96
H'00000180–H'00000183
IMIB4
2
97
H'00000184–H'00000187
OVI4
1
98
H'00000188–H'0000018B
Reserved
0
99
H'0000018C–H'0000018F
3
100 H'00000190–H'00000193
RxI0
2
101 H'00000194–H'00000197
TxI0
1
102 H'00000198–H'0000019B
TEI0
0
103 H'0000019C–H'0000019F
IPRE (15–12) 3
104 H'000001A0–H'000001A3
RxI1
2
105 H'000001A4–H'000001A7
TxI1
1
106 H'000001A8–H'000001AB
TEI1
0
107 H'000001AC–H'000001AF
3
108 H'000001B0–H'000001B3
Reserved
2
109 H'000001B4–H'000001B7
Reserved
1
110 H'000001B8–H'000001BB
Reserved
0
111 H'000001BC–H'000001BF
3
112 H'000001C0–H'000001C3
ITU4
SCI0
SCI1
PRT*1
IMIA3
IMIA4
ERI0
ERI1
PEI
0–15 (0)
0–15 (0)
0–15 (0)
0–15 (0)
0–15 (0)
IPRD (7–4)
IPRD (3–0)
IPRE (11–8)
WDT
ITI
REF*2
CMI
2
113 H'000001C4–H'000001C7
Reserved
1
114 H'000001C8–H'000001CB
Reserved
0
115 H'000001CC–H'000001CF
—
116 H'000001D0–H'000001D3
to
to
255 H'000003FC–H'000003FF Low
Reserved
0–15 (0)
IPRD (11–8)
—
IPRE (7–4)
Default
Priority
Order
—
Notes: 1. PRT: Parity control unit of bus state controller.
2. REF: DRAM refresh control unit of bus state controller.
RENESAS 67
5.3
Register Descriptions
5.3.1
Interrupt Priority Registers A–E (IPRA–IPRE)
The five registers from IPRA–IPRE are 16-bit read/write registers that assign priority levels from
0–15 to the IRQ and on-chip peripheral module interrupt sources. Interrupt request sources are
mapped onto IPRA–IPRE as shown in table 5.4.
Bit:
15
14
13
12
11
10
9
8
Bit name:
Initial value:
R/W:
Bit:
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
Table 5.4
Interrupt Request Sources and IPRA–IPRE
Register
Bits 15–12
Bits 11–8
Bits 7–4
Bits 3–0
IPRA
IRQ0
IRQ1
IRQ2
IRQ3
IPRB
IRQ4
IRQ5
IRQ6
IRQ7
IPRC
DMAC0, DMAC1
DMAC2, DMAC3
ITU0
ITU1
IPRD
ITU2
ITU3
ITU4
SCI1
PRT*1
IPRE
WDT,
SCI0
REF*2
(Reserved)*3
Notes: 1. PRT: Parity control unit of bus state controller. See section 8, Bus State Controller, for
details.
2. REF: DRAM refresh control unit of bus controller. See section 8, Bus State Controller,
for details.
3. When read, always 0. Always write 0 in reserved bits.
As indicated in table 5.4, four IRQ pins or four groups of on-chip peripheral modules are assigned
to each interrupt priority register. The priority levels for the four pins or groups can be set by
setting the corresponding 4-bit groups of bits 15–12, bits 11–8, bits 7–4, and bits 3–0 (of IPRA–
IPRE) with values in the range of H'0 (0000) to H'F (1111). Setting H'0 gives interrupt priority
level 0 (the lowest). Setting H'F gives level 15 (the highest). When two on-chip peripheral
modules are assigned to the same bits (DMAC0 and DMAC1, or DMAC2 and DMAC3, or the
watchdog timer and DRAM refresh control unit), those two modules have the same priority. A
reset initializes IPRA–IPRE to H'0000. They are not initialized by the standby mode.
68 RENESAS
5.3.2
Interrupt Control Register (ICR)
ICR is a 16-bit register that sets the input detection mode of the external interrupt input pins NMI
and IRQ0–IRQ7 and indicates the input signal level to the NMI pin. A reset initializes ICR but the
standby mode does not.
Bit:
15
14
13
12
11
10
9
8
NMIL
—
—
—
—
—
—
NMIE
Initial value:
*
0
0
0
0
0
0
0
R/W:
R
—
—
—
—
—
—
R/W
Bit:
7
6
5
4
3
2
1
0
IRQ0S
IRQ1S
IRQ2S
IRQ3S
IRQ4S
IRQ5S
IRQ6S
IRQ7S
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Bit name:
Initial value:
R/W:
Note: When NMI input is high: 1; when NMI input is low: 0
•
Bit 15 (NMI input level (NMIL)): NMIL sets the level of the signal input at the NMI pin.
NMIL cannot be modified. The NMI input level can be read to determine the NMI pin level.
Bit 15: NMIL
Description
0
NMI input level is low
1
NMI input level is high
•
Bits 14–9 (reserved): These bits always read as 0. The write value should always be 0.
•
Bit 8 (NMI edge select (NMIE)): NMIE selects whether the falling or rising edge of the
interrupt request signal to the NMI pin is sensed.
Bit 8: NMIE
Description
0
Interrupt is requested on falling edge of NMI input (initial value)
1
Interrupt is requested on rising edge of NMI input
Bits 7–0 (IRQ0–IRQ7 sense select (IRQ0S–IRQ7S)): IRQ0–IRQ7 select whether the falling edge or
low level of the IRQ inputs is sensed at the pins IRQ0–IRQ7.
Bits 7–0: IRQ0S–IRQ7S
Description
0
Interrupt is requested when IRQ input is low (initial value)
1
Interrupt is requested on falling edge of IRQ input
RENESAS 69
5.4
Interrupt Operation
5.4.1
Interrupt Sequence
The sequence of interrupt operations will be explained below. Figure 5.2 is a flowchart of the
operations up to acceptance of the interrupt.
1. The interrupt request sources send interrupt request signals to the interrupt controller.
2. The interrupt controller selects the highest-priority interrupt in the interrupt requests sent,
following the priority order indicated in table 5.3 and the levels set in interrupt priority
registers A–E (IPRA–IPRE). Lower priority interrupts are ignored*. If two interrupts with the
same priority level are requested simultaneously or if there are multiple interrupts occurring
within a single module, the interrupt with the highest default priority or priority within module
as indicated in table 5.3 is selected.
3. The interrupt controller compares the priority level of the selected interrupt request with the
interrupt mask level bits (I3–I0) in the CPU’s status register (SR). If the request priority level
is equal to or less than the interrupt mask level, the request is ignored. If the request priority
level is higher than the interrupt mask level, the interrupt controller accepts the request and
sends an interrupt request signal to the CPU.
4. When the interrupt controller accepts an interrupt request, it drives the pin IRQOUT low.
5. The CPU detects the interrupt request sent from the interrupt controller when it decodes the
next instruction to be executed. Instead of executing that instruction, the CPU starts interrupt
exception processing.
6. In interrupt exception processing, first SR and PC are pushed onto the stack.
7. The priority level of the accepted interrupt is copied to the interrupt mask level bits (I3–I0) in
the status register (SR).
8. When the accepted interrupt is level-sensed or from an on-chip peripheral module, The pin
IRQOUT returns to the high level. If the accepted interrupt is edge-sensed, the pin IRQOUT
returns to the high level when the instruction to be executed by the CPU in (5) is replaced by
the interrupt exception processing. If the interrupt controller has accepted another interrupt (of
a level higher than the current interrupt), however, the pin IRQOUT remains low.
9. The CPU accesses the exception vector table at the entry for the vector number of the accepted
interrupt, reads the start address of the exception service routine, branches to that address, and
starts executing the program there. This branch is not delayed.
Note: A request for an external interrupt (IRQ) designated as edge-detected is held pending once
only. An external interrupt designated as level-detected is held pending as long as the
interrupt request continues, but if the request is cleared before the CPU next accepts an
interrupt, the interrupt request is regarded as not having been made.
Interrupt requests from on-chip supporting modules are level requests. When the status
flag in a particular module is set, an interrupt is requested. For details, see the descriptions
of the individual modules. Note that the interrupt request will be continued unless an
70 RENESAS
operation described in "Clearing Conditions" is performed.
Program
execution state
No
Interrupt?
Yes
No
NMI?
Yes
User break?
Yes
No
Level 15
interrupt?
Yes
Yes
No
Level 14
interrupt?
I3 to I0 ≤
level 14?
No Yes
No
Yes
Level 1
interrupt?
I3 to I0 ≤
level 13?
Yes
No
IRQOUT low *1
No
Pushes SR onto stack
Yes
I3 to I0 =
level 0?
No
Pushes PC onto stack
Copies level of acceptance from I3 to I0
IRQOUT high *2
Reads exception
vector table
Branches to exception
service routine
I3 to I0 : Interrupt mask bits of status register
Notes : *1. IRQOUT is the same signal as the interrupt request signal to the CPU (Figure 5.1).
The pin IRQOUT return to the high level when the interrupt controller has accepted
the interrupt of a level higher than the I3 to I0 bits of the status register in the CPU.
*2. If the accepted interrupt is edge-sensed, the pin IRQOUT returns to the high level
when the instruction to be executed by the CPU is replaced by the interrupt
exception processing (before the status register is saved to the stack ).
If the interrupt controller has accepted another interrupt of a level higher than the
current interrupt. and has requested the interrupt to the CPU, however, the pin
IRQOUT remains low.
Figure 5.2 Flowchart of Interrupt Operation
RENESAS 71
5.4.2
Stack after Interrupt Exception Processing
Figure 5.3 shows the stack after interrupt exception processing.
Address
4n–8
PC*2
4n–2
SP*3
Lower 16 bits
4n–6
4n–4
Upper 16 bits
SR
Upper 16 bits
Lower 16 bits
4n
Notes: 1.
2.
3.
Bus width is 16 bits.
PC stores the top address of the next instruction (return instruction) after the
executed instruction.
The value of SP must always be a multiple of four.
Figure 5.3 Stack after Interrupt Exception Processing
72 RENESAS
5.5
Interrupt Response Time
Table 5.5 indicates the interrupt response time, which is the time from the occurrence of an
interrupt request until the interrupt exception processing starts and fetching of the first instruction
of the interrupt service routine begins. Figure 5.4 shows the pipeline when an IRQ interrupt is
accepted.
Table 5.5
Interrupt Response Time
Number of States
Item
Interrupt priority decision
and comparison with SR
mask bit
NMI or On-Chip
Interrupt
IRQ
2
3
Wait for completion of
X (≥ 0)
sequence currently being
executed by CPU
Notes
The longest sequence is the
interrupt or address error
exception processing
sequence: X = 4 + m1 + m2
+ m3 + m4. If an interruptmasking instruction follows,
however, the time may be
longer.
Time from interrupt
exception processing
(saving PC and SR and
fetching vector address)
until fetching of first
instruction of interrupt
service routine starts
5 + m1 + m2 + m3
Interrupt
response
Total
7 + m1 + m2 + m3
8 + m1 + m2 + m3
Minimum
10
11
0.50–0.55 µs at 20 MHz
Maximum
11 + 2(m1 + m2 +
m3) + m4
12 + 2(m1 + m2 +
m3) + m4
(m1 = m2 = m3 = m4) 0.90–
0.95 µs at 20 MHz
Notes: m1–m4 are the number of states needed for the following memory accesses:
m1: SR save cycle (long word write)
m2: PC save cycle (long word write)
m3: Vector address read cycle (long word read)
m4: Fetch top instruction of interrupt service routine
RENESAS 73
Interrupt accepted
5 + m1 + m2 + m3
3
m1 m2 1 m3 1
3
IRQ
Instruction (instruction replaced by
interrupt exception handling)
F D E E M M E M E E
Overrun fetch
Interrupt service routine—
first instruction
F
F D E
IRQOUT
(edge)
(level)
When m1 = m2 = m3, the interrupt response time is 11 cycles.
F (Instruction fetch)
D (Instruction decoding)
E (Instruction execution)
M (Memory access)
Instruction fetched from memory where program is stored.
The fetched instruction is decoded.
Data operations and address calculations are performed
according to the decoded results.
Data in memory is accessed.
Note: For the interrupt acceptance timing, see table 4.1, Exception Source Detection and
Exception Handling Start Timing, in section 4.1.2, Exception Handling Operation.
Figure 5.4 Example of Pipelining in IRQ Interrupt Acceptance
5.6
Usage Notes
When the following operations are performed in the order shown when a pin to which IRQ input is
assigned is designated as a general input pin by the pin function controller (PFC) and inputs a lowlevel signal, the IRQ falling edge is detected, and an interrupt request is detected, immediately
after the setting in (b) is performed:
•
•
An interrupt control register (ICR) setting is made so that an interrupt is detected at the falling
edge of IRQ.
…… (a)
The function of pins to which IRQ input is assigned is switched from general input to IRQ
input by a pin function controller (PFC) setting.
…… (b)
Therefore, when switching the pin function from general input pin to IRQ input, the pin function
controller (PFC) setting should be changed to IRQ input while the pin to which IRQ input is
assigned is high.
74 RENESAS
Section 6 User Break Controller (UBC)
6.1
Overview
The user break controller (UBC) simplifies the debugging of user programs. Break conditions are
set in the UBC and a user break interrupt request is sent to the CPU in response to the contents of
a CPU or DMAC bus cycle. This function can implement an effective self-monitoring debugger,
enabling a program to be debugged by itself without using a large in-circuit emulator.
6.1.1
•
•
•
Features
The following break conditions can be set:
 Address
 CPU cycle or DMA cycle
 Instruction fetch or data access
 Read or write
 Operand size (long word access, word access, or byte access)
When break conditions are met, a user break interrupt is generated. A user-created user break
interrupt exception routine can then be executed.
When a break is set to a CPU instruction fetch, the break occurs just before the fetched
instruction.
6.1.2
Block Diagram
Figure 6.1 is the block diagram of the user break controller.
RENESAS 75
BBR
BAMRH
BARH
BAMRL
BARL
Internal bus
Bus
interface
Module bus
Break condition comparator
User break
interrupt
generating
circuit
UBC
Interrupt request
Interrupt controller
BARH, BARL: Break address registers H and L
BAMRH, BAMRL: Break address mask registers H and L
BBR: Break bus cycle register
Figure 6.1 Block Diagram of the User Break Controller
6.1.3
Register Configuration
The user break controller has five registers as listed in table 6.1. These registers are used for
setting break conditions.
76 RENESAS
Table 6.1
User Break Controller Registers
Name
Abbr.
R/W
Address*
Initial
Value
Bus width
Break address register high
BARH
R/W
H'5FFFF90
H'0000
8, 16, 32
Break address register low
BARL
R/W
H'5FFFF92
H'0000
8, 16, 32
Break address mask register high
BAMRH
R/W
H'5FFFF94
H'0000
8, 16, 32
Break address mask register low
BAMRL
R/W
H'5FFFF96
H'0000
8, 16, 32
Break bus cycle register
BBR
R/W
H'5FFFF98
H'0000
8, 16, 32
Note: Only the values of bits A27–A24 and A8–A0 are valid; bits A23–A9 are ignored. For details
on the register addresses, see section 8.3.5, Description of Areas.
6.2
Register Descriptions
6.2.1
Break Address Registers (BAR)
There are two break address registers—break address register H (BARH) and break address
register L (BARL)—that together form a single group. Both are 16-bit read/write registers. BARH
stores the upper bits (bits 31–16) of the address of the break condition. BARL stores the lower bits
(bits 15–0) of the address of the break condition. A reset initializes both BARH and BARL to
H'0000. Neither is initialized in standby mode.
BARH: Break address register H.
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
•
15
14
13
12
11
10
9
8
BA31
BA30
BA29
BA28
BA27
BA26
BA25
BA24
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
BA23
BA22
BA21
BA20
BA19
BA18
BA17
BA16
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
BARH Bits 15–0 (break address 31–16 (BA31–BA16)): BA31–BA16 store the upper bit
values (bits 31–16) of the address of the break condition.
RENESAS 77
BARL: Break address register L.
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
•
15
14
13
12
11
10
9
8
BA15
BA14
BA13
BA12
BA11
BA10
BA9
BA8
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
BA7
BA6
BA5
BA4
BA3
BA2
BA1
BA0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
BARL Bits 15–0 (break address 15–0 (BA15–BA0)): BA15–BA0 store the lower bit values
(bits 15–0) of the address of the break condition.
6.2.2
Break Address Mask Register (BAMR)
The two break address mask registers—break address mask register H (BAMRH) and break
address mask register L (BARML)—together form a single group. Both are 16-bit read/write
registers. BAMRH determines which of the bits in the break address set in BARH are masked.
BAMRL determines which of the bits in the break address set in BARL are masked. A reset
initializes BAMRH and BARML to H'0000. They are not initialized in the standby mode.
BAMRH: Break address mask register H.
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
•
15
14
13
12
11
10
9
8
BAM31
BAM30
BAM29
BAM28
BAM27
BAM26
BAM25
BAM24
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
BAM23
BAM22
BAM21
BAM20
BAM19
BAM18
BAM17
BAM16
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
BAMRH bits 15–0 (break address mask 31–16 (BAM31–BAM16)): BAM31–BAM16 specify
whether bits BA31–BA16 of the break address set in BARH are masked or not.
78 RENESAS
BAMRL: Break address mask register L.
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
•
15
14
13
12
11
10
9
8
BAM15
BAM14
BAM13
BAM12
BAM11
BAM10
BAM9
BAM8
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
BAM7
BAM6
BAM5
BAM4
BAM3
BAM2
BAM1
BAM0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
BAMRL bits 15–0 (break address mask 15–0 (BAM15–BAM0)): BAM15–BAM0 specify
whether bits BA15–BA0 of the break address set in BARH are masked or not.
Bits 15–0: BAMn
Description
0
Break address bit BAn is included in the break condition (initial value)
1
Break address bit BAn is not included in the break condition
n = 31–0
6.2.3
Break Bus Cycle Register (BBR)
The break bus cycle register (BBR) is a 16-bit read/write register that selects the following four
break conditions:
•
•
•
•
CPU cycle or DMA cycle
Instruction fetch or data access
Read or write
Operand size (byte, word, long word).
A reset initializes BBR to H'0000. It is not initialized in the standby mode.
RENESAS 79
Bit:
15
14
13
12
11
10
9
8
Bit name:
—
—
—
—
—
—
—
—
Initial value:
0
0
0
0
0
0
0
0
R/W:
—
—
—
—
—
—
—
—
Bit:
7
6
5
4
3
2
1
0
CD1
CD0
ID1
ID0
RW1
RW0
SZ1
SZ0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
•
Bits 15–8 (reserved): These bits always read as 0. The write value should always be 0.
•
Bits 7 and 6 (CPU cycle/DMA cycle select (CD1 and CD0)): CD1 and CD0 select whether to
break on CPU and/or DMA bus cycles.
Bit 7: CD1
Bit 6: CD0
Description
0
0
No break interrupt occurs (initial value)
1
Break only on CPU cycles
0
Break only on DMA cycles
1
Break on both CPU and DMA cycles
1
•
Bits 5 and 4 (instruction fetch/data access select (ID1, ID0)): ID1, ID0 select whether to break
on instruction fetch and/or data access bus cycles.
Bit 5: ID1
Bit 4: ID0
Description
0
0
No break interrupt occurs (initial value)
1
Break only on instruction fetch cycles
0
Break only on data access cycles
1
Break on both instruction fetch and data access cycles
1
80 RENESAS
•
Bits 3 and 2 (read/write select (RW1, RW0)): RW1, RW0 select whether to break on read
and/or write access cycles.
Bit 3: RW1
Bit 2: RW0
Description
0
0
No break interrupt occurs (initial value)
1
Break only on read cycles
0
Break only on write cycles
1
Break on both read and write cycles
1
•
Bits 1 and 0 (operand size select (SZ1, SZ0)): SZ1, SZ0 select bus cycle operand size as a
break condition.
Bit 1: SZ1
Bit 0: SZ0
Description
0
0
Operand size is not a break condition (initial value)
1
Break on byte access
0
Break on word access
1
Break on long word access
1
Note: When setting to break on an instruction fetch, set the SZ0 bit to 0. All instructions will be
considered to be accessed as words (even those instructions in on-chip memory for which
two instructions can be fetched simultaneously in a single bus cycle). Instruction fetch is by
word access and CPU/DMAC data access is by the specified operand size. They are not
determined by the bus width of the space being accessed.
6.3
Operation
6.3.1
Flow of the User Break Operation
The flow from setting of break conditions to user break interrupt exception processing is described
below.
1. Break conditions are set in break address register (BAR), break address mask register
(BAMR), and the break bus cycle register (BBR). Set the break address in the BAR, the
address bits to be masked in the BAMR and the type of breaking bus cycle in the BBR. When
even one of the BBR groups (CPU cycle/DMA cycle select bits (CD1, CD0), instruction
fetch/data access select bits (ID1, ID0), read/write select bits (RW1, RW0)) is set to 00 (no
user break interrupt), there will be no user break even when all other conditions are consistent.
To use a user break interrupt, set conditions for all three pairs.
2. The UBC checks to see if the set conditions are satisfied, using the system shown in figure 6.2.
When the break conditions are satisfied, the UBC sends a user break interrupt request to the
interrupt controller.
RENESAS 81
3. When receiving the user break interrupt request, the interrupt controller checks its priority
level. The user break interrupt has priority level 15, so it is accepted only if the interrupt mask
level in bits I3–I0 in the status register (SR) is 14 or lower. When the I3–I0 bit level is 15, the
user break interrupt cannot be accepted but it is held pending until user break interrupt
exception processing is carried out. NMI exception processing sets I3–I0 to level 15, so a user
break cannot occur during the NMI service routine unless the NMI service routine itself begins
by reducing I3–I0 to level 14 or lower. Section 5, Interrupt Controller, described the handling
of priority levels in greater detail.
4. The INTC sends a request signal for a user break interrupt to the CPU. When the CPU receives
it, it starts user break interrupt exception processing. Section 5.4, Interrupt Operation, describes
interrupt exception processing in more detail.
82 RENESAS
BARH/BARL
BAMRH/BAMRL
32
32
Internal address
bits 31–0
32
32
CD1
CD0
ID1
ID0
32
CPU cycle
DMA cycle
Instruction fetch
User
break
interrupt
Data access
RW1
RW0
SZ1
SZ0
Read cycle
Write cycle
Byte size
Word size
Long word size
Figure 6.2 Break Condition Logic
RENESAS 83
6.3.2
Break on Instruction Fetch Cycles to On-Chip Memory
On-chip memory (on-chip ROM and RAM) is always accessed 32 bits each bus cycle. Two
instructionsare therefore fetched in a bus cycle from on-chip memory . Although only a single bus
cycle occurs for the two-instruction fetch, a break can be set on either instruction by placing the
corresponding address in the break address registers (BAR). In other words, to break the second of
the two instructions fetched, set its start address in the BAR. The break will then occur after the
first instruction executes.
6.3.3
Program Counter (PC) Value Saved in User Break Interrupt Exception Processing
Break on Instruction Fetch: The program counter (PC) value saved in user break interrupt
exception processing for an instruction fetch is the address set as the break condition. The user
break interrupt is generated before the fetched instruction is executed. If a break condition is set on
the fetch cycle of a delayed slot instruction immediately following a delayed branch instruction or
on the fetch cycle of an instruction that follows an interrupt-disabling instruction, however, the
user break interrupt is not accepted immediately, so the instruction is executed. The user break
interrupt is not accepted until immediately after that instruction. The PC value that will be saved is
the start address of the next instruction that is able to accept the interrupt.
Break on Data Access (CPU/DMAC): The program counter (PC) value is the top address of the
next instruction after the last executed instruction at the time when the user break exception
processing is activated. When data access (CPU/DMAC) is set as a break condition, the place
where the break will occur cannot be specified exactly. The break will occur at the instruction
fetched close to where the data access that is to receive the break occurs.
6.4
Setting User Break Conditions
CPU Instruction Fetch Bus Cycle:
•
Register settings: BARH = H'0000, BARL = H'0404, BBR = H'0054
Conditions set: Address = H'00000404, Bus cycle = CPU, instruction fetch, read (operand size
not included in conditions)
A user break interrupt will occur immediately before the instruction at address H'00000404. If
the instruction at address H'00000402 can accept an interrupt, the user break exception
processing will be executed after that instruction is executed. The instruction at H'00000404
will not be executed. The value saved to PC is H'00000404.
•
Register settings: BARH = H'0015, BARL = H'389C, BBR = H'0058
Conditions set: Address = H'0015389C, Bus cycle = CPU, instruction fetch, write (operand
size not included in conditions)
No user break interrupt occurs, because no instruction fetch cycle is ever a write cycle.
84 RENESAS
•
Register settings: BARH = H'0003, BARL = H'0147, BBR = H'0054
Conditions set: Address = H'00030147, Bus cycle = CPU, instruction fetch, read (operand size
not included in conditions)
No user break interrupt occurs, because instructions are always fetched from even addresses. If
the first fetched address after a branch is odd and a user break is set on this address, however,
user break exception processing will be carried out after address error exception processing.
CPU Data Access Bus Cycle:
•
Register settings: BARH = H'0012, BARL = H'3456, BBR = H'006A
Conditions set: Address = H'00123456, Bus cycle = CPU, data access, write, word
A user break interrupt occurs when word data is written to address H'00123456.
•
Register settings: BARH = H'00A8, BARL = H'0391, BBR = H'0066
Conditions set: Address = H'00A80391, Bus cycle = CPU, data access, read, word
No user break interrupt occurs, because word data access is always to an even address.
DMA Cycle:
•
Register setting: BARH = H'0076, BARL = H'BCDC, BBR = H'00A7
Conditions set: Address = H'0076BCDC, Bus cycle = DMA, data access, read, long word
A user break interrupt occurs when long word data is read from address H'0076BCDC.
•
Register setting: BARH = H'0023, BARL = H'45C8, BBR = H'0094
Conditions set: Address = H'002345C8, Bus cycle = DMA, instruction fetch, read (operand
size not included)
No user break interrupt occurs, because a DMA cycle includes no instruction fetch.
RENESAS 85
6.5
Notes
6.5.1
On-Chip Memory Instruction Fetch
Two instructions are simultaneously fetched from on-chip memory. If a break condition is set on
the second of these two instructions but the contents of the UBC break condition registers are
changed so as to alter the break condition immediately after the first of the two instructions is
fetched, a user break interrupt will still occur when the second instruction is fetched.
6.5.2
Instruction Fetch at Branches
When a conditional branch instruction or TRAPA instruction causes a branch, instructions are
fetched and executed as follows:
1. Conditional branch instruction, branch taken: BT, BF
Instruction fetch cycles: Conditional branch fetch → Next-instruction overrun fetch → Nextinstruction overrun fetch → Branch destination fetch
Instruction execution: Conditional branch instruction execution → Branch destination
instruction execution
2. TRAPA instruction, branch taken: TRAPA
Instruction fetch cycles: TRAPA instruction fetch → Next-instruction overrun fetch → Nextinstruction overrun fetch → Branch destination fetch
Instruction execution: TRAPA instruction execution → Branch destination instruction
execution
When a conditional branch instruction or TRAPA instruction causes a branch, the branch
destination will be fetched after the next instruction or the one after that does an overrun fetch.
When the next instruction or the one after that is set as a break condition, a branch will result in
the generation of a user break interrupt at the next instruction or the instruction after that, neither
of which instructions will be executed.
86 RENESAS
6.5.3
Instruction Fetch Break
If a break is attempted at the task A return destination instruction fetch, task B is activated before
the UBC interrupt by interrupt B generated during task A processing, and the UBC interrupt is
handled after the interrupt B exception handling.
(1) Cause
The SH7032/SH7034 chip operates as follows.
Interrupt B accepted
UBC interrupt accepted
Interrupt exception
handling
F D E E M M E M E E
Interrupt exception
handling
F
<Address>
F D E
<Description>
0x00011a0a Instruction replaced by interrupt
exception handling
Break
condition
E M M E M E E
f
0x00011a0c Overrun fetch
0xf000974 Task B first instruction fetch
(instruction replaced by interrupt
exception handling)
F
(0xf000978 Overrun fetch)
0x02000030 UBC first instruction fetch
Figure 6.3
UBC Operation
It actually takes at least two cycles for the UBC interrupt generated by the address 0x00011a0c
instruction fetch cycle to be sent to the interrupt controller and interrupt exception handling to
begin. However, as shown in figure 6.3, when the UBC interrupt is generated, previously
generated interrupt B initiated by task B is accepted first, and the UBC interrupt is accepted after
completion of the interrupt B exception handling.
(2) Remedy
There is no way of preventing this operation by hardware. A software solution, such as the use of
a flag, must be employed.
RENESAS 87
Section 7 Clock Pulse Generator (CPG)
7.1
Overview
The SuperH microcomputer has a built-in clock pulse generator (CPG) that supplies the LSI and
external devices with a clock pulse. The CPG makes the LSI run at the oscillation frequency of the
crystal resonator. The CPG consists of an oscillator and a duty cycle correcting circuit (figure 7.1).
The CPG can be made to generate a clock signal by connecting it to a crystal resonator or by
inputting an external clock. (The CPG is halted in standby mode.)
CPG
XTAL
Oscillator
Duty correcting
circuit
EXTAL
Internal
clock (φ)
CK
System clock
Figure 7.1 Block Diagram of the Clock Pulse Generator
7.2
Clock Source
Clock pulses can be supplied from a connected crystal resonator or an external clock.
7.2.1
Connecting a Crystal Resonator
Circuit Configuration: A crystal resonator can be connected as shown in figure 7.2. Use the
damping resistance Rd shown in table 7.1. Use an AT-cut parallel resonating crystal with a
frequency equal to the system clock (CK) frequency. Connect load capacitors (C L1 and CL2) as
shown in the figure. The clock pulse produced by the crystal resonator and internal pulse generator
is sent to the duty cycle correction circuit where its duty cycle is corrected. It is then supplied to
the LSI and to external devices.
RENESAS 89
CL1
EXTAL
CL1 = CL2 = IC–22 pF
Rd
CL2
XTAL
Figure 7.2 Connection of the Crystal Resonator (Example)
Table 7.1
Damping Resistance
Frequency [MHz]
2
4
8
12
16
20
Rd [Ω]
1k
500
200
0
0
0
Crystal Resonator: Figure 7.3 shows an equivalent circuit of the crystal resonator. Use a crystal
resonator with the characteristics listed in table 7.2.
L
CL
Rs
XTAL
EXTAL
C0
Figure 7.3 Crystal Resonator Equivalent Circuit
Table 7.2
Crystal Resonator Parameters
Frequency (MHz)
Parameter
2
4
8
12
16
20
Rs max [Ω]
500
120
80
60
50
40
Co max [pF]
7
7
7
7
7
7
Value to be determined (TBD)
7.2.2
External Clock Input
An external clock signal can be input at the EXTAL pin as shown in figure 7.6. The XTAL pin
should be left open. The frequency must be equal to the system clock (CK) frequency. The
specifications for the waveform of the external clock input are given below. Make the external
clock frequency the same as the system clock (CK).
90 RENESAS
Open
External clock input
XTAL
EXTAL
Figure 7.4 External Clock Input Method
tcyc
tEXH
tEXL
VIH
1/2 Vcc
VIL
tEXr
tEXf
Figure 7.5 Input Clock Waveform
Table 7.3
Input Clock Specifications
5 V Specifications
(fmax = 20 MHz)
3.3 V Specifications
(fmax = 12.5 MHz)
Units
t EXr/f (VIL–VIH)
Max = 5
Max = 10
ns
t EXH/L (1/2 VCC
standard)
Min = 10
Min = 20
ns
7.3
Usage Notes
Board design: When designing the board, place the crystal resonator and its load capacitors as
close as possible to the XTAL and EXTAL pins. Route no other signal lines near the XTAL and
EXTAL pin signal lines to prevent induction from interfering with correct oscillation. See figure
7.6.
RENESAS 91
No crossing
signal lines
CL1
XTAL
CL2
EXTAL
Figure 7.6 Precaution on Oscillator Circuit Board Design
Duty cycle correction circuit: Duty cycle corrections are conducted for an input clock over 5
MHz. Duty cycles may not be corrected if under 5 MHz, but AC characteristics for the high-level
pulse width (tCH) and low-level pulse width (tCL) of the clock are satisfied, and the LSI will
operate normally. Figure 7.7 shows the standard characteristics of a duty cycle correction.
This duty cycle correction circuit is not for correcting the input clock's transient fluctuations and
jutters.
Output duty
70
60
Input duty
70
60
50
40
30
50
40
30
1
2
5
10
20
(MHz)
Input frequency
Figure 7.7 Duty Cycle Correction Circuit Standard Characteristics
92 RENESAS
Section 8 Bus State Controller (BSC)
8.1
Overview
The bus state controller (BSC) divides address space and outputs control signals for all kinds of
memory and peripheral LSIs. BSC functions enable the LSI to link directly with DRAM, SRAM,
ROM, and peripheral LSIs without the use of external circuits, simplifying system design and
allowing high-speed data transfers in a compact system.
8.1.1
Features
The BSC has the following features.
•
•
•
•
•
•
Address space is divided into eight areas
 A maximum 4-Mbyte of linear address space for each of eight areas, 0–7 (area 1 can be up
to 16-Mbyte linear space when set for DRAM) (The space that can actually be used varies
with the type of memory connected)
 Bus width (8 bits or 16 bits) can be selected by access address
 On-chip ROM and RAM is accessed in one cycle (32 bits wide)
 Wait states can be inserted using the WAIT pin
 Wait state insertion can be controlled through software. Register settings can be used to
specify the insertion of 1–4 cycles for areas 0, 2, and 6 (long wait function)
 The type of memory connected can be specified for each area.
 Outputs control signals for accessing the memory and peripheral LSIs connected to the area
Direct interface to DRAM
 Multiplexes row/column addresses according to DRAM capacity
 Two types of byte access signals (dual-CAS system and dual-WE system)
 Supports burst operation (high-speed page mode)
 Supports CAS-before-RAS refresh and self-refresh
Access control for all memory and peripheral LSIs
 Address/data multiplex function
Parallel execution of external writes and the like with internal access (warp mode)
Supports parity check and generation for data bus
 Odd parity/even parity selectable
 Interrupt request generated for parity error (PEI interrupt request signal)
Refresh counter can be used as an 8-bit interval timer
 Interrupt request generated at compare match (CMI interrupt request signal)
8.1.2
Block Diagram
Figure 8.1 shows the block diagram of the bus state controller.
RENESAS 93
WCR1
Wait control
unit
WAIT
WCR2
WCR3
RD
BCR
WRH, WRL
DCR
RCR
CASH, CASL
DRAM
control
unit
RAS
CMI interrupt request
Module bus
Area control
unit
HBS, LBS
AH
CS7 to CS0
RTCSR
RTCNT
Comparator
DPH, DPL
Interrupt
controller
Peripheral bus
PEI interrupt request
Parity control
unit
RTCOR
PCR
WCR: Wait state control register
BCR: Bus control register
DCR: DRAM area control register
RCR: Refresh control register
RTCSR: Refresh timer control/status register
RTCNT: Refresh timer counter
RTCOR: Refresh time constant register
PCR: Parity control register
Figure 8.1 BSC Block Diagram
94 RENESAS
Internal bus
Bus
interface
8.1.3
Pin Configuration
Table 8.1 shows the BSC pin configuration.
Table 8.1
Pin Configuration
Name
Abbreviation
I/O Function
Chip select 7–0
CS7–CS0
O
Chip select signal that indicates the area being
accessed
Read
RD
O
Strobe signal that indicates the read cycle
High write
WRH
O
Strobe signal that indicates write cycle to upper 8 bits
Low write
WRL
O
Strobe signal that indicates write cycle to lower 8 bits
Write
WR*1
O
Strobe signal that indicates write cycle
High byte strobe
HBS* 2
O
Strobe signal that indicates access to upper 8 bits
Low byte strobe
LBS* 3
O
Strobe signal that indicates access to lower 8 bits
Row address strobe
RAS
O
DRAM row address strobe signal
High column
address strobe
CASH
O
Column address strobe signal for accessing the
upper 8 bits of the DRAM
Low column address CASL
strobe
O
Column address strobe signal for accessing the lower
8 bits of the DRAM
Address hold
AH
O
Signal for holding the address for address/data
multiplexing
Wait
WAIT
I
Wait state request signal
Address bus
A21–A0
O
Address output
Data bus
AD15–AD0
I/O Data I/O. During address/data multiplexing, address
output and data input/output.
Data bus parity high
DPH
I/O Parity data I/O for upper byte
Data bus parity low
DPL
I/O Parity data I/O for lower byte
Notes: 1. Doubles with the WRL pin. (Selected by the BAS bit of the BCR. See section 8.2.1, Bus
Control Register, for details.
2. Doubles with the A0 pin. (Selected by the BAS bit of the BCR. See section 8.2.1, Bus
Control Register, for details.
3. Doubles with the WRH pin. (Selected by the BAS bit of the BCR. See section 8.2.1, Bus
Control Register, for details.
8.1.4
Register Configuration
The BSC has ten registers (listed in table 8.2) which control space division, wait states, DRAM
interface, and parity check.
RENESAS 95
Table 8.2
Register Configuration
Name
Abbr.
R/W
Initial Value
Address* 1
Bus width
Bus control register
BCR
R/W
H'0000
H'5FFFFA0
8,16,32
Wait state control register 1
WCR1
R/W
H'FFFF
H'5FFFFA2
8,16,32
Wait state control register 2
WCR2
R/W
H'FFFF
H'5FFFFA4
8,16,32
Wait state control register 3
WCR3
R/W
H'F800
H'5FFFFA6
8,16,32
DRAM area control register
DCR
R/W
H'0000
H'5FFFFA8
8,16,32
Parity control register
PCR
R/W
H'0000
H'5FFFFAA 8,16,32
Refresh control register
RCR
R/W
H'0000
H'5FFFFAC 8,16,32* 2
Refresh timer control/status register
RTCSR
R/W
H'0000
H'5FFFFAE 8,16,32* 2
Refresh timer counter
RTCNT
R/W
H'0000
H'5FFFFB0
8,16,32* 2
Refresh time constant register
RTCOR
R/W
H'00FF
H'5FFFFB2
8,16,32* 2
Notes: 1. Only the values of bits A27–A24 and A8–A0 are valid; bits A23–A9 are ignored. For
details on the register addresses, see section 8.3.5, Description of Areas.
2. Write only with word transfer instructions. See section 8.2.11, Register Access, for
details on writing.
8.1.5
Overview of Areas
The SH microprocessors have 32-bit address spaces on the architecture, but the top 4 bits are
ignored. Table 8.3 outlines the division of space. As shown, the space is divided into areas 0–7 by
the value of the top addresses.
Each area is allocated a specific type of space. When the area is accessed, a strobe signal that
matches the type of area space is generated. This allocates peripheral LSIs and memory devices
according to the type of the area spaces and allows them to be directly linked to this LSI. Some
areas are of a fixed type based on their address while others can be selected in registers.
Area 0 can be used as an on-chip ROM space or external memory space. Area 1 can be used as
DRAM space or external memory space. DRAM space enables direct connection to DRAM and
outputs RAS, CAS and multiplexed addresses. Areas 2–4 can only be used as external memory
space. Area 5 can be used as on-chip peripheral module space or external memory space. Area 6
can be used as address/data multiplexed I/O space or external memory space. For address/data
multiplexed I/O space, an address and data are multiplexed and input/output from AD15–AD0
pins. Area 7 can be used as on-chip RAM space or external memory space.
The bus width of the data bus is basically switched between 8 bit and 16 bit by the value of
address bit A27. For the following areas, however, the bus width is determined by conditions other
than the A27 bit value.
96 RENESAS
•
•
•
•
•
On-chip ROM space in area 0: Always 32 bits
External memory space in area 0: 8 bits when MD0 pin is 0, 16 bits when the pin is 1
On-chip peripheral module space in area 5: 8 bits when the A8 address bit is 0, 16 bits when it
is 1
Area 6: If A27 = 0, area 6 is 8 bits when the A14 address bit is 0, 16 bits when A14 is 1
On-chip RAM space in area 7: Always 32 bits
See table 8.6 in section 8.3, Address Space Subdivision, for more information on how the space is
divided.
8.2
Register Descriptions
8.2.1
Bus Control Register (BCR)
The bus control register (BCR) is a 16-bit read/write register that selects the functions of areas and
status of bus cycles. It is initialized to H'0000 by a power-on reset, but is not initialized by a
manual reset or by the standby mode.
Bit:
15
14
13
12
11
10
9
8
IOE
WARP
RDDTY
BAS
—
—
—
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
—
—
—
Bit:
7
6
5
4
3
2
1
0
Bit name: DRAME
Initial value:
R/W:
•
Bit name:
—
—
—
—
—
—
—
—
Initial value:
0
0
0
0
0
0
0
0
R/W:
—
—
—
—
—
—
—
—
Bit 15 (DRAM enable bit (DRAME)): DRAME selects whether area 1 is used as an external
memory space or DRAM space. 0 sets it for external memory space and 1 sets it for DRAM
space. The setting of the DRAM area control register is valid only when this bit is set to 1.
Bit 15: DRAME
Description
0
Area 1 is external memory space (initial value)
1
Area 1 is a DRAM space
•
Bit 14 (multiplexed I/O enable bit (IOE)): IOE selects whether area 6 is used as external
memory space or an address/data multiplexed I/O area. 0 sets it for external memory space and
1 sets it for address/data multiplexed I/O space. With address/data multiplexed I/O space,
address and data are multiplexed and input/output is from AD15–AD0.
RENESAS 97
Bit 14: IOE
Description
0
Area 6 is external memory space (initial value)
1
Area 6 is an address/data multiplexed I/O area
•
Bit 13 (warp mode bit (WARP)): WARP selects warp or normal mode. 0 sets it for normal
mode and 1 sets it for warp mode. In warp mode, some external accesses are carried out in
parallel with internal access.
Bit 13: WARP
Description
0
Normal mode: External and internal accesses are not simultaneously
performed (initial value)
1
Warp mode: External and internal accesses are simultaneously
performed
•
Bit 12 (RD duty (RDDTY)): RDDTY selects 35% or 50% of the T1 state as the high-level duty
cycle ratio of signal RD. 0 sets it for 50%, 1 sets it for 35%. Only set to 1 when the operating
frequency is a minimum of 10 MHz.
Bit 12: RDDTY
Description
0
RD signal high-level duty cycle is 50% of T1 state (initial value)
1
RD signal high-level duty cycle is 35% of T1 state
•
Bit 11 (byte access select (BAS)): BAS selects whether byte access control signals are WRH,
WRL, and A0, or LBS, WR and HBS during word space accesses. When this bit is cleared to
0, WRH, WRL, and A0 signals are valid; when set to 1, LBS, WR, and HBS, signals are valid.
Bit 11: BAS
Description
0
WRH, WRL, and A0 enabled (initial value)
1
LBS, WR, and HBS enabled
•
Bits 10–0 (reserved): These bits always read as 0. The write value should always be 0.
8.2.2
Wait State Control Register 1 (WCR1)
Wait state control register 1 is a 16-bit read/write register that controls the number of states for
accessing each area and the whether wait states are used. WCR1 is initialized to H'FFFF by a
power-on reset. It is not initialized by a manual reset or by the standby mode.
98 RENESAS
Bit:
15
14
13
12
11
10
9
8
RW7
RW6
RW5
RW4
RW3
RW2
RW1
RW0
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
—
WW1
—
Initial value:
1
1
1
1
1
1
1
1
R/W:
—
—
—
—
—
—
R/W
—
Bit name:
Initial value:
R/W:
•
Bits 15–8 (wait state control during read (RW7–RW0)): RW7–RW0 determine the number of
states in read cycles for each area and whether or not to sample the signal input from the
WAIT pin. Bits RW7–RW0 correspond to areas 7–0, respectively. If a bit is cleared to 0, the
WAIT signal is not sampled during the read cycle for the corresponding area. If it is set to 1,
sampling takes place.
For the external memory spaces of areas 1, 3–5, and 7, read cycles are completed in one state
when the corresponding bits are cleared to 0. When they are set to 1, the number of wait states
is 2 plus the WAIT signal value. For the external memory space of areas 0, 2, and 6, read
cycles are completed in one state plus the number of long wait states (set in wait state
controller 3 (WCR3)) when the corresponding bits are cleared to 0. When they are set to 1, the
number of wait states is 1 plus the long wait state; when the WAIT signal is at low level as
well, a wait state is inserted.
The DRAM space (area 1) finishes the column address output cycle in one state (short pitch)
when the RW1 bit is 0, and in 2 states plus the WAIT signal value (long pitch) when RW1 is 1.
When RW1 is set to 1, the number of wait states selected in wait state insertion bits 1 and 0
(RLW0 and RLW1) for CAS-before-RAS (CBR) refresh of the refresh control register (RCR)
are inserted during the CBR refresh cycle, regardless of the status of the WAIT signal.
The read cycle of the address/data multiplexed I/O space (area 6) is 4 states plus the wait states
from the WAIT signal, regardless of the setting of the RW6 bit. The read cycle of the on-chip
peripheral module space (area 5) finishes in 3 states, regardless of the setting of the RW5 bit,
and the WAIT signal is not sampled. The read cycles of on-chip ROM (area 0) and on-chip
RAM (area 7) finish in 1 state, regardless of the settings of bits RW0 and RW7. The WAIT
signal is not sampled for either.
Table 8.3 summarizes read cycle state information.
RENESAS 99
Table 8.3
Read Cycle State Description
Read Cycle States
External Memory Space
WAIT
Bits 15–8: Pin Input External Memory
RW7–RW0 Signal
Space
0
Not
sampled
during
read
cycle*1
• Areas 1, 3–5,7: 1
state, fixed
1
Sampled
during
read
cycle
(initial
value)
Areas 1, 3–5, 7: 2
states + wait states
from WAIT
DRAM Space
Column address cycle: 1
Areas 0, 2, 6: 1 state state, fixed
(short pitch)
+ long wait state
Internal space
MultiPlexed
I/O
On-chip
Peripheral
Module
4 states 3 states,
fixed
+ wait
states
from
WAIT
On-chip
ROM and
RAM
1 state,
fixed
Column
address cycle:
2 states + wait
Areas 0, 2, 6: 1 state state from
WAIT (long
+ long wait state +
pitch)*2
wait state from
WAIT
Notes: 1. Sampled in the address/data multiplexed I/O space
2. During a CBR refresh, the WAIT signal is ignored and the wait state from the RLW1
and RLW0 bits of RCR is inserted.
•
Bits 7–2 (reserved): These bits always read as 1. The write value should always be 1.
•
Bit 1 (wait state control during write (WW1)): WW1 determines the number of states in write
cycles for the DRAM space (area 1) and whether or not to sample the WAIT signal. When the
DRAM enable bit (DRAME) of the BCR is set to 1 and area 1 is being used as DRAM space,
clearing WW1 to 0 makes the column address output cycle finish in 1 states (short pitch).
When WW1 is set to 1, it finishes in 2 states plus the wait states from the WAIT signal (long
pitch).
Note: Write 0 to WW1 only when area 1 is used as DRAM space (DRAME bit of BCR is 1).
Never write 0 to WW1 when area 1 is used as external memory space (DRAME is 0).
Bit 1: WW1
DRAM Space (DRAME = 1)
Area 1's External Memory
Space (DRAME = 0)
0
Column address cycle: 1 state (short pitch)
Setting inhibited
1
Column address cycle: 2 states + wait state
from WAIT (long pitch) (initial value)
2 states + wait state from WAIT
•
Bit 0 (reserved): This bit always reads 1. The write value should always be 1.
100 RENESAS
8.2.3
Wait State Control Register 2 (WCR2)
Wait state control register 2 is a 16-bit read/write register that controls the number of states for
accessing each area with a DMA single address mode transfer and whether wait states are used.
WCR2 is initialized to H'FFFF by a power-on reset. It is not initialized by a manual reset or by the
standby mode.
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
•
15
14
13
12
11
10
9
8
DRW7
DRW6
DRW5
DRW4
DRW3
DRW2
DRW1
DRW0
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
DWW7
DWW6
DWW5
DWW4
DWW3
DWW2
DWW1
DWW0
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bits 15–8 (wait state control during single-mode DMA transfer (DRW7–DRW0)): DRW7–
DRW0 determine the number of states in single-mode DMA memory read cycles for each area
and whether or not to sample the WAIT signal. Bits DRW7–DRW0 correspond to areas 7–0,
respectively. If a bit is cleared to 0, the WAIT signal is not sampled during the single-mode
DMA memory read cycle for the corresponding area. If it is set to 1, sampling takes place.
For the external memory spaces of areas 1, 3–5, and 7, single-mode DMA memory read cycles
are completed in one state when the corresponding bits are cleared to 0. When they are set to 1,
the number of wait states is 2 plus the wait states from the WAIT signal. For the external
memory space of areas 0, 2, and 6, single-mode DMA memory read cycles are completed in
one state plus the long wait state number (set in wait state controller 3 (WCR3)) when the
corresponding bits are cleared to 0. When they are set to 1, the number of wait states is 1 plus
the long wait state; when the WAIT signal is at low level as well, a wait state is inserted.
The DRAM space (area 1) finishes the column address output cycle in one state (short pitch)
when the DRW1 bit is 0, and in 2 states plus the wait states from the WAIT signal (long pitch)
when DRW1 is 1. The single-mode DMA memory read cycle of the address/data multiplexed
I/O space (area 6) is 4 states plus the wait states from the WAIT signal, regardless of the
setting of the DRW6 bit.
Table 8.4 summarizes single-mode DMA memory read cycle state information.
RENESAS 101
Table 8.4
Single-Mode DMA Memory Read Cycle States (External Memory Space)
Description
Single-Mode DMA Memory Read Cycle States
(External Memory Space)
Bits 15–8:
DRW7–DRW0
WAIT Pin Input
Signal
External Memory
Space
0
Not sampled during
single-mode DMA
memory read cycle*
Areas 1, 3–5,7: 1 state,
fixed
Sampled during
single-mode DMA
memory read cycle
(initial value)
Areas 1, 3–5, 7: 2 states
+ wait states from WAIT
1
Areas 0, 2, 6: 1 state +
long wait state
Areas 0, 2, 6: 1 state +
long wait state + Wait
state from WAIT
DRAM Space
Column address
cycle: 1 state,
fixed (short
pitch)
Multiplexed
I/O
4 states +
wait states
from WAIT
Column address
cycle: 2 states +
wait state from
WAIT (long
pitch)
Note: Sampled in the address/data multiplexed I/O space.
•
Bits 7–0 (single-mode DMA memory write wait state control (DWW7–DWW0)): DWW7–
DWW0 determine the number of states in single-mode DMA memory write cycles for each
area and whether or not to sample the WAIT signal. Bits DWW7–DWW0 correspond to areas
7–0, respectively. If a bit is cleared to 0, the WAIT signal is not sampled during the singlemode DMA memory write cycle for the corresponding area. If it is set to 1, sampling takes
place.
The number of states for areas accesses based on bit settings are the same as indicated for
single-mode DMA memory read cycles. See bits 15–8, wait state control during single-mode
DMA memory transfer (DRW7–DRW0), for details.
Table 8.5 summarizes single-mode DMA memory write cycle state information.
102 RENESAS
Table 8.5
Single-Mode DMA Memory Write Cycle States (External Memory Space)
Description
Single-mode DMA Memory Write Cycle States
(External Memory Space)
Bits 15–8:
DWW7–DWW0
WAIT Pin Input
Signal
External Memory
Space
0
Not sampled during Areas 1, 3–5,7: 1 state,
fixed
single-mode DMA
memory write cycle* Areas 0, 2, 6: 1 state +
long wait state
1
Sampled during
single-mode DMA
memory write cycle
(initial value)
Multiplexed
I/O
DRAM Space
4 states +
wait state
from WAIT
Column address
cycle: 1 state,
fixed (short
pitch)
Areas 1, 3–5, 7: 2 states
+ wait state from WAIT
Areas 0, 2, 6: 1 state +
long wait state + wait
state from WAIT
Column address
cycle: 2 states +
wait state from
WAIT (long
pitch)
Note: Sampled in the address/data multiplexed I/O space.
8.2.4
Wait State Control Register 3 (WCR3)
Wait state control register 3 is a 16-bit read/write register that controls WAIT pin pull-up and the
insertion of long wait states. WCR3 is initialized to H'F800 by a power-on reset. It is not
initialized by a manual reset or by the standby mode.
Bit:
Bit name:
Initial value:
WPU
14
13
12
A02LW1 A02LW0 A6LW1
11
10
9
8
A6LW0
—
—
—
1
1
1
1
1
0
0
0
R/W
R/W
R/W
R/W
R/W
—
—
—
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
—
—
—
Initial value:
0
0
0
0
0
0
0
0
R/W:
—
—
—
—
—
—
—
—
R/W:
•
15
Bit 15 (wait pin pull-up control (WPU)): WPU controls whether the WAIT pin is pulled up or
not. When cleared to 0, the pin is not pulled up; when set to 1, it is pulled up.
RENESAS 103
Bit 15: WPU
Description
0
WAIT pin is not pulled up
1
WAIT pin is pulled up (initial value)
•
Bits 14 and 13 (long wait insertion in areas 0 and 2, bits 1, 0 (A02LW1 and A02LW0)):
A02LW1 and A02LW0 select the long wait states to be inserted (1–4 states) when accessing
external memory space of areas 0 and 2.
Bit 14: A02LW1 Bit 13: A02LW0 Description
0
1
•
0
Inserts 1 state
1
Inserts 2 states
0
Inserts 3 states
1
Inserts 4 states (initial value)
Bits 12 and 11 (long wait insertion in area 6, bits 1, 0 (A6LW1 and A6LW0)): A6LW1 and
A6LW0 select the long wait states to be inserted (1–4 states) when accessing external memory
space of area 6.
Bit 12: A6LW1
Bit 11: A6LW0
Description
0
0
Inserts 1 state
1
Inserts 2 states
0
Inserts 3 states
1
Inserts 4 states (initial value)
1
•
Bits 10–0 (reserved): These bits always read as 0. The write value should always be 0.
8.2.5
DRAM Area Control Register (DCR)
The DRAM area control register (DCR) is a 16-bit read/write register that selects the type of
DRAM control signal, the number of precharge cycles, the burst operation mode and the use of
address multiplexing. DCR settings are valid only when the DRAME bit of BCR is set to 1. It is
initialized to H'0000 by a power-on reset, but is not initialized by a manual reset or by the standby
mode.
104 RENESAS
Bit:
15
14
13
12
11
10
9
8
CW2
RASD
TPC
BE
CDTY
MXE
MXC1
MXC0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
—
—
—
Initial value:
0
0
0
0
0
0
0
0
R/W:
—
—
—
—
—
—
—
—
Bit name:
Initial value:
R/W:
•
Bit 15 (dual-CAS or dual-WE select bit (CW2)): When accessing a 16-bit bus width space,
CW2 selects the dual-CAS or the dual-WE method. When cleared to 0, the CASH, CASL, and
WRL signals are valid ; when set to 1, the CASL, WRH, and WRL signals are valid. When
accessing an 8-bit space, only CASL and WRL signals are valid, regardless of CW2 settings.
Bit 15L: CW2
Description
0
Dual-CAS: CASH, CASL, and WRL signals are valid (initial value)
1
Dual-WE: CASL, WRH, and WRL signals are valid
•
Bit 14 (RAS down (RASD)): When DRAM access pauses, RASD determines whether to keep
RAS low while waiting for the next DRAM access (RAS down mode) or return it to high
(RAS up mode). When cleared to 0, the RAS signal returns to high; when set to 1, it stays at
low.
Bit 14: RASD
Description
0
RAS up mode: Return RAS signal to high and wait for the next DRAM
access (initial value)
1
RAS down mode: Keep RAS signal low and wait for the next DRAM
access
•
Bit 13 (RAS precharge cycle count (TPC)): TPC selects whether the RAS signal precharge
cycle (TP) will be 1 state or 2. When TPC is cleared to 0, a 1-state precharge cycle is inserted;
when 1 is set, a 2-state precharge cycle is inserted.
Bit 13: TPC
Description
0
Inserts 1-state precharge cycle (initial value)
1
Inserts 2-state precharge cycle
RENESAS 105
•
Bit 12 (burst operation enable (BE)): BE selects whether or not to perform burst operation, a
high speed page mode. When burst operation is not selected (0), the row address is not
compared but instead is transferred to the DRAM every time and full access is performed.
When burst operation is selected (1), row addresses are compared and burst operation with the
same row address as the previous is performed (in this access, no row address is output and the
column address and CAS signal alone are output).
Bit 12: BE
Description
0
Normal mode: full access (initial value)
1
Burst operation: high-speed page mode
•
Bit 11 (CAS duty (CDTY)): CDTY selects 35% or 50% of the TC state as the high-level duty
ratio of the signal CAS in the short-pitch access. When cleared to 0, the CAS signal high level
duty is 50%; when set to 1, it is 35%. Only set to 1 when the operating frequency is a
minimum of 10 MHz.
Bit 11: CDTY
Description
0
CAS signal high level duty cycle is 50% of the T C state (initial value)
1
CAS signal high level duty cycle is 35% of the T C state
•
Bit 10 (multiplex enable bit (MXE)): MXE determines whether or not DRAM row and column
addresses are multiplexed. When cleared to 0, addresses are not multiplexed; when set to 1,
they are multiplexed.
Bit 10: MXE
Description
0
Multiplex of row and column addresses disabled (initial value)
1
Multiplex of row and column addresses enabled
•
Bits 9 and 8 (multiplex shift count 1 and 0 (MXC1 and MXC0)): Shift row addresses
downward by a certain number of bits (8–10) when row and column addresses are multiplexed
(MXE = 1). Regardless of the MXE bit setting, these bits also select the range of row addresses
compared in burst operation.
106 RENESAS
Bit 9:
MXC1
Bit 8:
MXC0
Row Address Shift
(MXE = 1)
Row Address Bits Compared (in burst operation)
(MXE = 0 or 1)
0
0
8 bits (initial value)
A8–A27 (initial value)
1
9 bits
A9–A27
0
10 bits
A10–A27
1
Reserved
Reserved
1
Bits 7–0 (reserved): These bits always read as 0. The write value should always be 0.
8.2.6
Refresh Control Register (RCR)
The refresh control register (RCR) is a 16-bit read/write register that controls the start of
refreshing and selects the refresh mode and the number of wait states during refresh. It is
initialized to H'0000 by a power-on reset, but is not initialized by a manual reset or by the standby
mode.
To prevent RCR from being written incorrectly, it must be written by a different method from
most other registers. A word transfer operation is used, H'5A is written in the top byte, and the
actual data is written in the lower byte. For details, see section 8.2.11, Notes on Register Access.
Bit:
15
14
13
12
11
10
9
8
Bit name:
—
—
—
—
—
—
—
—
Initial value:
0
0
0
0
0
0
0
0
R/W:
—
—
—
—
—
—
—
—
Bit:
7
6
5
4
3
2
1
0
RLW1
RLW0
—
—
—
—
Bit name:
Initial value:
R/W:
RFSHE RMODE
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
—
—
—
—
•
Bit 15–8 (reserved): These bits always read as 0.
•
Bit 7 (refresh control (RFSHE)): RFSHE determines whether or not to perform DRAM refresh
operations. When this bit is cleared to 0, no DRAM refresh control is performed and the
refresh timer counter (RTCNT) can be used as an 8-bit interval timer. When set to 1, DRAM
refresh control is performed.
RENESAS 107
Bit 7: RFSHE
Description
0
Refresh control disabled. RTCNT can be used as an 8-bit interval
timer. (initial value)
1
Refresh control enabled
•
Bit 6 (refresh mode (RMODE)): When DRAM refresh control is selected (RFSHE = 1),
RMODE selects whether to perform CAS-before-RAS (CBR) refresh or self-refresh. When
this bit is cleared to 0, a CBR refresh is performed at the cycle set in the refresh timer
control/status register (RTCSR) and refresh time constant register (RTCOR). When set to 1, it
the DRAM does a self-refresh. When refresh control is not selected (RFSHE = 0), the RMODE
bit setting is not valid. When canceling self-refresh, set RMODE to 0 with RFSHE set to 1.
Bit 6: RMODE
Description
0
CAS-before-RAS refresh (initial value)
1
Self-refresh
•
Bits 5 and 4—Insert wait states during CBR refresh bits 1 and 0 (RLW1, RLW0): These bits
select the number of wait states to be inserted (1–4) during CAS-before-RAS refresh. When
CBR refresh is performed and the RW1 bit of WCR1 is set to 1, the number of wait states
selected in the RLW1 and RLW0 is inserted regardless of the WAIT signal. When the RW1 bit
is cleared to 0, the RLW1 and RLW0 bit settings are ignored and no wait states are inserted.
Bit 5: RLW1
Bit 4: RLW0
Description
0
0
Inserts 1 state (initial value)
1
Inserts 2 states
0
Inserts 3 states
1
Inserts 4 states
1
•
Bits 3–0 (reserved): These bits always read as 0. The write value should always be 0.
8.2.7
Refresh Timer Control/Status Register (RTCSR)
The refresh timer control/status register (RTCSR) is a 16-bit read/write register that selects the
clock input to refresh timer counter (RTCNT) and controls compare match interrupts (CMI). It is
initialized to H'0000 by a power-on reset, but is not initialized by a manual reset or by the standby
mode.
To prevent RTCSR from being written incorrectly, it must be written by a different method from
most other registers. A word transfer operation is used, H'A5 is written in the top byte and the
actual data is written in the lower byte. For details, see section 8.2.11, Notes on Register Access.
108 RENESAS
Bit:
15
14
13
12
11
10
9
8
Bit name:
—
—
—
—
—
—
—
—
Initial value:
0
0
0
0
0
0
0
0
R/W:
—
—
—
—
—
—
—
—
Bit:
7
6
5
4
3
2
1
0
CMF
CMIE
CKS2
CKS1
CKS0
—
—
—
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
—
—
—
Bit name:
Initial value:
R/W:
•
Bits 15–8 (reserved): These bits always read as 0.
•
Bit 7 (compare match flag (CMF)): CMF is a flag that indicates whether the values of RTCNT
and the refresh time constant register (RTCOR) match. When 0, the value of RTCNT and
RTCOR do not match; when 1, the value of RTCNT and RTCOR match.
Bit 7: CMF
Description
0
RTCNT does not equal the value of RTCOR (initial value)
To clear CMF, the CPU must read CMF after it has been set to 1, then write a
0 in this bit
1
•
Value RTCNT is equal to the value of RTCOR
Bit 6 (compare match interrupt enable (CMIE)): CMIE enables or disables the compare match
interrupt (CMI) generated when CMF is set to 1 in RTCSR (RTCNT value = RTCOR value).
When cleared to 0, CMI interrupt is disabled; when set to 1, it is enabled.
Bit 6: CMIE
Description
0
Compare match interrupt request (CMI) is disabled (initial value)
1
Compare match interrupt request (CMI) is enabled
•
Bits 5–3 (clock select bits 2–0 (CKS2–CKS0)): CKS2–CKS0 select the clock input to RTCNT
from among the seven types of clocks created by dividing the system clock (φ). When the input
clock is selected with the CKS2–CKS0 bits, RTCNT starts to increment.
RENESAS 109
Bit 5: CKS2
Bit 4: CKS1
Bit 3: CKS0
Description
0
0
0
Clock input disabled (initial value)
1
φ/2
0
φ/8
1
φ/32
0
φ/128
1
φ/512
0
φ/2048
1
φ/4096
1
1
0
1
•
Bits 2–0 (reserved): These bits always read as 0. The write value should always be 0.
8.2.8
Refresh Timer Counter (RTCNT)
The refresh timer counter (RTCNT) is a 16-bit read/write register that is used as an 8-bit upcounter
that generates the refresh or interrupt request. When the input clock is selected by clock select bits
2–0 (CKS2–CKS0) in RTCSR, that clock makes the RTCNT start incrementing. When the values
of RTCNT and the refresh time constant register (RTCOR) match, RTCNT is cleared to H'0000
and the CMF flag of the RTCSR is set to 1. When the RFSHE bit of the RCR is also set to 1, a
CAS-before-RAS refresh is performed. When the CMIE bit of the RTCSR is also set to 1, a
compare match interrupt (CMI) is generated.
Bits 15–8 are reserved bits and do not count. These bits always read as 0.
RTCNT is initialized to H'0000 by a power-on reset, but is not initialized by a manual reset or by
the standby mode.
To prevent RTCSR from being written incorrectly, it must be written by a different method from
most other registers. A word transfer operation is used, H'69 is written in the top byte and the
actual data is written in the lower byte. For details, see section 8.2.11, Register Access.
110 RENESAS
Bit:
15
14
13
12
11
10
9
8
Bit name:
—
—
—
—
—
—
—
—
Initial value:
0
0
0
0
0
0
0
0
R/W:
—
—
—
—
—
—
—
—
Bit:
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
RENESAS 111
8.2.9
Refresh Time Constant Register (RTCOR)
The refresh time constant register (RTCOR) is a 16-bit read/write register that sets the compare
match cycle used with RTCNT. The values in RTCOR and RTCNT are constantly compared.
When they match, the compare-match flag (CMF) is set in RTCNT and RTCSR is cleared to
H'0000. If the bit RFSHE in RCR is set to 1 when this happens, a CAS before RAS (CBR) refresh
is performed. When the CMIE bit of the RTCSR is also set to 1, a compare match interrupt (CMI)
is generated.
Bits 15–8 are reserved bits and cannot be used to set the cycle. These bits always read as 0.
RTCOR is initialized to H'00FF by a power-on reset, but is not initialized by a manual reset or by
the standby mode.
To prevent RTCOR from being written incorrectly, it must be written by a different method from
most other registers. A word transfer operation is used, H'96 is written in the top byte and the
actual data is written in the lower byte. For details, see section 8.2.11, Note on Register Access.
Bit:
15
14
13
12
11
10
9
8
Bit name:
—
—
—
—
—
—
—
—
Initial value:
0
0
0
0
0
0
0
0
R/W:
—
—
—
—
—
—
—
—
Bit:
7
6
5
4
3
2
1
0
Bit name:
Initial value:
R/W:
8.2.10
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Parity Control Register (PCR)
The parity control register (PCR) is a 16-bit read/write register that selects the parity polarity and
space to be parity checked. PCR is initialized to H'0000 by a power-on reset, but is not initialized
by a manual reset or by the standby mode.
112 RENESAS
Bit:
15
14
13
PEF
PFRC
PEO
0
0
0
0
R/W
R/W
R/W
Bit:
7
6
Bit name:
—
Initial value:
R/W:
Bit name:
Initial value:
R/W:
•
12
11
10
9
8
—
—
—
0
0
0
0
R/W
R/W
—
—
—
5
4
3
2
1
0
—
—
—
—
—
—
—
0
0
0
0
0
0
0
0
—
—
—
—
—
—
—
—
PCHK1 PCHK0
Bit 15 (parity error flag (PEF)): When a parity check is done, PEF indicates whether a parity
error has occurred. 0 indicates that no parity error has occurred; 1 indicates that a parity error
has occurred.
Bit 15: PEF
Description
0
No parity error (initial value).
Cleared by reading PEF after it has been set to 1, then writing 0 in
PEF.
1
•
Parity error has occurred.
Bit 14 (parity output force (PFRC)): PFRC selects whether to produce a forced parity output
for testing the parity error check function. When cleared to 0, there is no forced output; when
set to 1, it produces a forced output of high level from the DPH and DPL pins when data is
output, regardless of the parity.
Bit 14: PFRC
Description
0
Parity output not forced (initial value)
1
High output forced
•
Bit 13 (parity polarity (PEO)): PEO selects even or odd parity. When cleared to 0, parity is
even; when set to 1, parity is odd.
Bit 13: PEO
Description
0
Even parity (initial value)
1
Odd parity
•
Bits 12 and 11 (parity check enable bits 1, 0 (PCHK1 and PCHK0)): These bits determine
whether or not a parity is checked and generated, and select the check and generation spaces.
RENESAS 113
Bit 12: PCHK1
Bit 11: PCHK0
Description
0
0
Parity not checked and not generated (initial value)
1
Parity checked and generated only in DRAM area
0
Parity checked and generated in DRAM area and area 2
1
Reserved
1
•
Bits 10–0 (reserved): These bits always read as 0. The write value should always be 0.
8.2.11
Notes on Register Access
RCR, RTCSR, RTCNT, and RTCOR differ from other registers in being more difficult to write.
Data requires a password when it is written. This prevents data from being mistakenly overwritten
by program overruns and the like.
Writing to RCR, RTCSR, RTCNT, and RTCOR: Use only word transfer instructions. You
cannot write with byte transfer instructions. As figure 8.2 shows, when writing to RCR, place
H'5A in the upper byte and the write data in the lower byte. When writing to RTCSR, place H'A5
in the upper byte and the write data in the lower byte. When writing to RTCNT, place H'69 in the
upper byte and the write data in the lower byte. When writing to RTCOR, place H'96 in the upper
byte and the write data in the lower byte. These transfers write data in the lower byte to the
respective registers. If the upper byte differs from the above passwords, no writing occurs.
15
RCR
8 7
H'5A
15
RTCSR
8 7
15
0
Write data
8 7
H'69
15
RTCOR
Write data
H'A5
RTCNT
0
0
Write data
8 7
H'96
0
Write data
Figure 8.2 Writing to RCR, RTCSR, RTCNT, and RTCOR
Reading from RCR, RTCSR, RTCNT, and RTCORP: These registers are read like other
registers. They can be read by byte and word transfer instructions. If read by word transfer, the
value of the upper eight bits is H'00.
114 RENESAS
8.3
Address Space Subdivision
8.3.1
Address Spaces and Areas
Figure 8.3 shows the address format used in this LSI.
4 Gbyte space
128 Mbyte space
16 Mbyte space
4 Mbyte space
A31–A28 A27
A26–A24
A23,A22
A21
A0
Output address:
Output from address pins
A21–A0
Ignore: Only valid when the address multiplex
function is being used in the DRAM space (area 1);
not output in other cases. When not output,
becomes shadow.
Area selection:
Decoded to become chip select signals CS0–CS7 for areas 0–7
Basic bus width selection:
Not output externally, but used for basic bus width selection
When 0, (H'0000000–H'7FFFFFF), the basic bus width is 8 bits.
When 1, (H'8000000–H'FFFFFFF) the basic bus width is 16 bits.
Ignore: Always ignore, not output externally
Figure 8.3 Address Format
Since this LSI uses a 32-bit address, 4 Gbytes of space can be accessed on the architecture;
however, the upper 4 bits (A31–A28) are always ignored and not output. Bit A27 is basically only
used for switching the bus width. When the A27 bit is 0 (H'0000000–H'7FFFFFF), the bus width
is 8 bits; when the A27 bit is 1 (H'8000000–H'FFFFFFF), the bus width is 16 bits. Of the
remaining 27 bits (A26–A0), a total 128 Mbyte can thus be accessed.
The 128 Mbyte space is subdivided into 8 areas (areas 0–7) of 16 Mbytes each according to the
values of bits A26–A24. The space with bits A26–A24 as 000 is area 0 and the space 111 is area 7.
The A26–A24 bits are decoded and become the chip select signals (CS0–CS7) of the
corresponding areas 0–7 and output. Table 8.6 shows how the space is divided.
RENESAS 115
Table 8.6
How Space is Divided
Area Address
0
Assign-able Memory
H'0000000 – H'0FFFFFF On-chip ROM*1
External memory*2
1
H'1000000 – H'1FFFFFF External memory
DRAM*6
Capacity
(linear space)
Bus
Width
CS
Output
16 kB* 3
32 kB* 4
32
—
4 MB
8/16*5
CS0
4 MB
8
CS1
16 MB
8
RAS CAS
2
H'2000000 – H'2FFFFFF External memory
4 MB
8
CS2
3
H'3000000 – H'3FFFFFF External memory
4 MB
8
CS3
4
H'4000000 – H'4FFFFFF External memory
4 MB
8
CS4
8/16*7
—
8/16*8
CS6
8
CS7
32
—
4 MB
8/16*5
CS0
4 MB
16
CS1
16 MB
16
RAS CAS
5
H'5000000 – H'5FFFFFF On-chip peripheral module 512 B
memory*9
6
H'6000000 – H'6FFFFFF External
7
H'7000000 – H'7FFFFFF External memory
Multiplexed I/O
0
H'8000000 – H'8FFFFFF On-chip
ROM*1
External memory*2
1
H'9000000 – H'9FFFFFF External memory
DRAM*6
4 MB
4 MB
4 MB
16
32
kB* 3
kB* 4
2
H'A000000 – H'AFFFFFF External memory
4 MB
16
CS2
3
H'B000000 – H'BFFFFFF External memory
4 MB
16
CS3
4
H'C000000 – H'CFFFFFF External memory
4 MB
16
CS4
5
H'D000000 – H'DFFFFFF External memory
4 MB
16
CS5
6
H'E000000 – H'EFFFFFF External memory
4 MB
16
CS6
7
H'F000000 – H'FFFFFFF On-chip RAM
1 kB
32
—
Notes: 1.
2.
3.
4.
5.
6.
7.
When MD2–MD0 pins are 010
When MD2–MD0 pins are 000 or 001
For SH7020
For SH7021
Select with MD0 pin
Select with DRAME bit in BCR
Divided into 8-bit and 16-bit space according to value of address bit A8 (Long word
accesses are inhibited, however, in on-chip peripheral modules with bus widths of 8
bits. Some on-chip peripheral modules with bus widths of 16 bits also have registers
that are only byte-accessible and registers for which byte access is inhibited. For
details, see the sections on the individual modules.)
8. Divided into 8-bit space and 16-bit space by value of address bit A14
9. Select with IOE bit of BCR
116 RENESAS
As figure 8.4 shows, specific spaces such as DRAM space and address/data multiplexed I/O space
are allocated to the 8 areas. Each of the spaces is equipped with the necessary interfaces. The
control signals needed by DRAM and peripheral LSIs will be output by the chip to devices
connected to an area allocated to the appropriate type of space.
8.3.2
Bus Width
The primary bus width selection on for this chip is made by switching between 8-bit and 16 bit
using the A27 bit. When A27 is 0, the bus width is 8 bits and data is input/output through the
AD7–AD0 pins; when A27 is 1, the size is 16 bits and data is input/output through the AD15–
AD0 pins for word accesses. For byte access, the top byte is input/output through AD15–AD8 and
the lower byte through AD7–AD0. When the bus width is 8 bits or byte access is being performed
with a 16-bit bus width, the status of the eight AD pins that are not inputting/outputting data is as
shown in appendix B, Pin States.
Bus widths are also determined by conditions other than the A27 bit for specific areas:
•
•
•
•
Area 0 is an 8-bit external memory space when the MD2–MD0 pins are 000, a 16-bit external
memory space when the same bits are 001, and a 32-bit on-chip ROM space when they are
010.
Area 5 is an 8-bit on-chip peripheral module space when the A27 bit and A8 bit are both 0 and
a 16-bit on-chip peripheral module space when the A27 bit is 0 and the A8 bit is 1. When the
A27 bit is 1, it is a 16-bit external memory space.
Area 6 is an 8-bit bus width when the A27 bit and A14 bit are both 0 and a 16-bit bus width
when the A27 bit is 0 and the A14 bit is 1. When the A27 bit is 1, it is a 16-bit space.
Area 7 is a 32-bit on-chip RAM space when the A27 bit is 1 and an 8-bit external memory
space when the A27 bit is 0.
Word (16-bit) data accessed from 8-bit bus areas and longword (32-bit) data accessed from 16-bit
bus areas require two consecutive accesses. Longword (32-bit) data accessed from 8-bit bus areas
requires four consecutive accesses.
8.3.3
Chip Select Signals (CS0–CS7)
When the A26–A24 bits of the address are decoded, they become chip select signals (CS0–CS7)
for areas 0–7. When an area is accessed, the corresponding chip select pins are driven low. Table
8.7 shows the relationship between the A26–A24 bits and the chip select signals.
RENESAS 117
Table 8.7
A26–A24 Bits and Chip Select Signals
Address
A26
A25
A24
Area Selected
Chip Select Pin Driven Low
0
0
0
Area 0
CS0
1
Area 1
CS1
0
Area 2
CS2
1
Area 3
CS3
0
Area 4
CS4
1
Area 5
CS5
0
Area 6
CS6
1
Area 7
CS7
1
1
0
1
The chip select signal is output only for external accesses. When accessing the on-chip ROM (area
0), on-chip peripheral modules (area 5) and on-chip RAM (area 7), the CS0, CS5, and CS7 pins
are not driven low. When accessing DRAM space (area 1), select the RAS and CAS signals with
the pin function controller.
8.3.4
Shadows
The size of each area is 16 Mbytes, which can be specified with 24 address bits A23–A0 for 8-bit
spaces and 16-bit spaces alike. Bits A23 and A22, however, output externally only when the
address multiplex function is used in DRAM space (area 1); in all other cases, there is no output,
so the actually accessible area for all areas is the 4 Mbyte that can be specified with 22 bits A21–
A0. No matter what the values of A23 and A22, the same 4 Mbytes of actual space is accessed. As
illustrated in figure 8.4 (a), the A23 and A22 bit regions 00, 01, 10 and 11 are called shadows of
actual areas. Shadows are allocated in 4-Mbyte units for both 8-bit and 16-bit bus widths. When
the same addresses H'3200000, H'3600000, H'3A00000 and H'3E00000 are specified for values
A21–A0, as shown in figure 8.4 (b), the same actual space is accessed regardless of the A23 and
A22 bits.
In areas whose bus widths are switchable using the A27 address bit, the shadow of the same actual
space is allocated to both A27 = 0 spaces and A27 = 1 spaces (figure 8.4(a)). When the value of
A27 is changed, the valid AD pins switch from AD15–AD0 to AD7–AD0, but the actual space
accessed remains the same.
The spaces of on-chip ROM (area 0), DRAM (area 1), on-chip peripheral modules (area 5) and onchip RAM (area 7) have shadows of different sizes from those discussed above. See section 8.3.5,
Description of Areas, for details.
118 RENESAS
Logical address space
H'B000000
H'3000000
H'B3FFFFF
H'B400000 H'33FFFFF
Shadow
(A23, A22 = 00)
H'3400000
H'B7FFFFF
H'B800000 H'37FFFFF
Shadow
(A23, A22 = 01)
H'BBFFFFF
H'BC00000
H'BFFFFFF
H'3800000
H'3BFFFFF
H'3C00000
H'3FFFFFF
16-bit space
Shadow
(A23, A22 = 10)
Actual space
Area
accessible
with A21–A0
4 Mbytes
Shadow
(A23, A22 = 11)
8-bit space
a. Shadow allocation
Logical address space
H'3000000
H'3200000
H'33FFFFF
Location indicated
by address
H'3400000
H'3600000
H'37FFFFF
H'3800000
H'3A00000
H'3BFFFFF
H'3C00000
H'3E00000
H'3FFFFFF
Actual space
Location indicated
by address
Location actually
accessed
Location indicated
by address
Location indicated
by address
8-bit space
b. Actual space accessed when addresses are specified
Figure 8.4 Shadows
RENESAS 119
8.3.5
Area Description
Area 0: Area 0 is the area where addresses A26–A24 are 000 and its address range is H'0000000–
H'0FFFFFF and H'8000000–H'8FFFFFF. Figure 8.5 is a memory map of area 0.
Area 0 can be set for use as on-chip ROM space or external memory space with the mode pins
(MD2–MD0). The MD2–MD0 pins also determine the bus width, regardless of the A27 address
bit. When MD2–MD0 are 000, area 0 is an 8-bit external memory space; when they are 001, area
0 is a 16-bit external memory space; and when they are 010, it is a 32-bit on-chip ROM space.
In the SH7020, the capacity of the on-chip ROM is 16 kbyte, so bits A23–A14 are ignored in onchip ROM space and the shadow is in 16 kbyte units. In the SH7021, the capacity of the on-chip
ROM is 32 kbyte, so bits A23–A15 are ignored in on-chip ROM space and the shadow is in 32
kbyte units. The CSO signal is disabled in on-chip ROM space.
In external memory space, the A23 and A22 bits are not output and the shadow is in 4-Mbyte
units. When external memory space is accessed, the CS0 signal is valid. The external memory
space has a long wait function, so between 1 and 4 states can be selected for the number of long
waits inserted into the bus cycle using the areas 0 and 2 long wait insertion bits (A02LW1,
A02LW0) of wait state controller 3 (WCR3).
120 RENESAS
Logical address space
Logical address space
H'8000000
H'8003FFF(SH7020)
H'8007FFF(SH7021) H'0000000
H'8004000(SH7020) H'0003FFF
(SH7020)
H'8008000(SH7021)
H'0007FFF
(SH7021)
H'8000000
Shadow
Shadow
Shadow
H'83FFFFF
H'8400000
H'03FFFFF
H'0400000
H'004000
(SH7020)
H'0008000
(SH7021)
H'8FFC000
(SH7020)
H'8FF8000
(SH7021) H'0FFC000
(SH7020)
H'0FF8000
(SH7021)
H'0FFFFFF
H'0000000
Shadow
Shadow
Shadow
Shadow
Actual space
Shadow
H'87FFFFF
H'8800000
Actual space
H'07FFFFF
On-chip ROM
H'0800000
SH7020:
16 kbyte
SH7021:
H'8BFFFFF
32 kbyte H'8C00000
• Valid
H'0BFFFFF
addresses
H'0C00000
A15–A0
(A23–A16
H'8FFFFFF
ignored)
• CS0 not
valid
H'0FFFFFF
8 or 16
bit space
External
memory
space
(4 Mbytes)
Shadow
• MD2–MD0 =
000: 8-bit
access,
001: 16-bit
access
• Valid
addresses
A21–A0
(A23 and
A22 not
output)
• CS0 valid
• Long wait
function
Shadow
8 or 16
bit space
32-bit space 32-bit space
MD2–MD0 = 010
MD2–MD0 = 000 or 001
Note: The bus width of area 0 is determined by the MD2–MD0 pins regardless of the A27 bit
setting.
Figure 85 Memory Map of Area 0
Area 1: Area 1 is the area where addresses A26–A24 are 001 and its address range is H'1000000–
H'1FFFFFF and H'9000000–H'9FFFFFF. Figure 8.6 is a memory map of area 1.
Area 1 can be set for use as DRAM space or external memory space with the DRAM enable bit
(DRAME) of the bus control register (BCR). When the DRAME bit is 0, it is external memory
space; when DRAME is 1, it is DRAM space.
In external memory space, the bus width is 8 bits when the A27 bit is 0 and 16 bits when it is 1.
Bits A23 and A22 are not output and the shadow is in 4-Mbyte units. When external memory is
accessed, the CS1 signal is valid.
DRAM space is a type of external memory space, but it is configured especially to be connected to
DRAM so it outputs strobe signals required for this purpose. Its bus width is 8 bits when it is 0 and
16 bits when it is 1. When the multiplex enable bit (MXE) of the DRAM control register (DCR) is
RENESAS 121
set to 1 to use the address multiplex function, bits A23–A0 are multiplexed and output from pins
A15–A0, so a maximum 16-Mbyte space can be used. When DRAM space is accessed, the CS1
signal is not valid and the pin function controller should be set for access with CAS (CASH and
CASL) and RAS signals.
Logical address space
Logical address space
H'9000000
H'9000000
Actual space
H'1000000
H'1000000
Shadow
H'93FFFFF
H'9400000
H'13FFFFF
H'1400000
Actual space
Shadow
H'97FFFFF
H'9800000
H'17FFFFF
H'1800000
Shadow
H'9BFFFFF
H'9C00000
H'1BFFFFF
H'1C00000
H'9FFFFFF
H'1FFFFFF
A27 = 1:
16-bit space
Shadow
External
memory
space
(4 Mbytes)
Shadow
DRAM
space
(maximum
16 Mbytes)
• Valid
address
A21–A0
(A23 and
A22 not
output)
• CS1
H'9FFFFFF
valid
H'1FFFFFF
A27 = 0:
8-bit space
A27 = 1:
A27 = 0:
16-bit space 8-bit space
DRAME = 0 or DRAME = 1, MXE = 0
• Multiplexed
(MXE = 1):
16-bit space
• Not multiplexed
(MXE = 0):
4 Mbyte
space
• CS1 not
valid (CAS,
RAS output)
DRAME = 1
Figure 8.6 Memory Map of Area 1
Areas 2–4: Areas 2–4 are the areas where addresses A26–A24 are 010, 011 and 100, respectively,
and their address ranges are H'2000000–H'2FFFFFF and H'A000000–H'AFFFFFF (area 2),
H'3000000–H'3FFFFFF and H'B000000–H'BFFFFFF (area 3), and H'4000000–H'4FFFFFF and
H'C000000–H'CFFFFFF (area 4). Figure 8.7 is a memory map of area 2, which is representative
of areas 2–4.
122 RENESAS
Areas 2–4 are always used as external memory space. The bus width is 8 bits when the A27 bit is
0 and 16 bits when it is 1. A23 and A22 bits are not output and the shadow is in 4-Mbyte units.
When areas 2–4 are accessed, the CS2, CS3, and CS4 signals are valid. Area 2 has a long wait
function, so between 1 and 4 states can be selected for the number of long waits inserted into the
bus cycle using the bits A02LW1 and A02LW0 of WCR3.
Logical address space
H'A000000
H'2000000
Shadow
H'A3FFFFF
H'A400000
H'23FFFFF
H'2400000
Actual space
Shadow
H'A7FFFFF
H'A800000
External
memory space
(4 Mbytes)
H'27FFFFF
H'2800000
Shadow
H'ABFFFFF
H'AC00000
H'2BFFFFF
H'2C00000
• Valid addresses A21–A0
(A23 and A22 not output)
• CS2 valid
• Long wait function
Shadow
H'AFFFFFF
H'2FFFFFF
16-bit space
8-bit space
Figure 8.7 Memory Map of Area 2
RENESAS 123
Area 5: Area 5 is the area where addresses A26–A24 are 101 and its address range is H'5000000–
H'5FFFFFF and H'D000000–H'DFFFFFF. Figure 8.8 is a memory map of area 5.
Area 5 is allocated to on-chip peripheral module space when the A27 address bit is 0 and external
memory space when A27 is 1. In on-chip peripheral module space, bits A23–A9 are ignored and
the shadows are in 512-byte units. The bus width is 8 bits when the A8 bit is 0 and 16 bits when
A8 is 1. When on-chip peripheral module space is accessed, the CS5 signal is not valid. In
external memory space, the A23 and A22 bits are not output and the shadow is in 4-Mbyte units.
The bus width is always 16 bits. When external memory space is accessed, the CS5 signal is valid.
Logical address
space
H'5000000
H'50001FF Shadow
Shadow
Logical address
space
H'D000000
Shadow
Shadow
H'D3FFFFF
H'D400000
Actual
space
On chip
peripheral
module space
(512 bytes)
Shadow
Shadow
H'5FFFE00
H'5FFFFFF
Shadow
8 or 16-bit
space
Actual
space
Shadow
H'D7FFFFF
H'D800000
Shadow
A8 = 0:
H'DBFFFFF
H'DC00000
8-bit space
A8 = 1: 16-bit space*
• Ignored
Shadow
addresses:
A23–A9
(Valid addresses H'DFFFFFF
16-bit space
A8–A0)
• CS5 not valid
External
memory
space
(4 Mbytes)
• Valid
addresses
A21–A0
A23 and A22
not output)
• CS5 valid
Note: Some registers in onchip peripheral modules can only be accessed as 8-bit registers
even though they occupy 16 bits (see Appendix A).
Figure 8.8 Memory Map of Area 5
Area 6: Area 6 is the area where addresses A26–A24 are 110 and its address range is H'6000000–
H'6FFFFFF and H'E000000–H'EFFFFFF. Figure 8.9 is a memory map of area 6.
124 RENESAS
In area 6, the space when A27 is 0 is allocated to address/data multiplexed I/O space when the
multiplexed I/O enable bit (IOE) of the bus control register (BCR) is 1 and external memory space
when the IOE bit is 0. When A27 is 1, it is always external memory space.
The multiplexed I/O space is a type of external memory space but the address and data are
multiplexed and output from AD15–AD0 or AD7–AD0. The bus width is 8 bits when the A14 bit
is 0 and 16 bits when the A14 bit is 1. The A23 and A22 bits are not output and the shadow is in 4Mbyte units. When multiplexed I/O space is accessed, the CS6 signal is valid.
In external memory space, the bus width is 8 bits when both the A27 and A14 bits are 0 and 16
bits when the A27 bit is 0 and the A14 bit is 1. When the A27 bit is 1, it is always a 16-bit space.
The A23 and A22 bits are not output and the shadow is in 4-Mbyte units. When external memory
is accessed, the CS6 signal is valid. The external memory space has a long wait function so
between 1 and 4 states can be selected for the number of long waits inserted into the bus cycle
using the area 6 long wait insertion bits (A6LW1 and A6LW0) of WCR3.
RENESAS 125
Logical address
space
H'6000000
Logical address
space
H'E000000
Shadow
H'63FFFFF
H'6400000
Shadow
Actual
space
Shadow
H'67FFFFF
H'6800000
Shadow
H'6BFFFFF
H'6C00000
Shadow
H'6FFFFFF
8 or 16-bit
space
Multiplexed
I/O space
or external
memory
space
(4 Mbytes)
H'E3FFFFF
H'E400000
Actual
space
Shadow
H'E7FFFFF
H'E800000
External
memory
space
(4 Mbytes)
Shadow
• Valid
• IOE = 1:
addresses
address/data
A21–A0 (A23
multiplexed I/O
H'EBFFFFF
H'EC00000
and A22 not
space;
output)
IOE = 0: external
• CS6 valid
memory space
Shadow
• Long wait
• A14 = 0: 8-bit space
function
A14 = 1: 16-bit space
• Valid addresses
H'EFFFFFF
A21–A0 (A23 and
16-bit space
A22 not output)
• CS6 valid
• Long wait function
Figure 8.9 Memory Map of Area 6
Area 7: Area 7 is the area where addresses A26–A24 are 111 and its address range is H'7000000–
H'7FFFFFF and H'F000000–H'FFFFFFF. Figure 8.10 is a memory map of area 7.
Area 7 is allocated to external memory space when A27 is 0 and on-chip RAM space when A27 is
1. In external memory space, the bus width is 8 bits. The A23 and A22 bits are not output and the
shadow is in 4-Mbyte units. When external memory is accessed, the CS7 signal is valid.
The on-chip RAM space has an bus width of 32 bits. The on-chip RAM capacity is 1 kbytes, so
A23–A10 are ignored and the shadows are in 8-kbyte units. During on-chip RAM access, the CS7
signal is not valid.
126 RENESAS
Logical address
space
H'7000000
Logical address
space
H'F000000
H'F0003FF Shadow
Shadow
Shadow
Shadow
H'73FFFFF
H'7400000
Actual
space
Shadow
H'77FFFFF
H'7800000
Shadow
H'7BFFFFF
H'7C00000
Actual
space
External
memory
space
(4 Mbytes)
• Valid
addresses
A21–A0
(A23 and A22
not output)
• CS7 valid
Shadow
Shadow
Shadow
H'FFFFC00
H'FFFFFFF
H'7FFFFFF
8-bit space
Shadow
• On-chip
RAM space
1 Kbytes,
• Valid
addresses
A9–A0
(A23–A10
not output)
• CS7 not
valid
32-bit space
Figure 8 10 Memory Map of Area 7
RENESAS 127
8.4
Accessing External Memory Space
In external memory space, strobe signal is output based on the assumption of a directly connected
SRAM. The external memory space is allocated to the following areas:
•
•
•
•
•
•
Area 0 (when MD2–MD0 are 000 or 001)
Area 1 (when the DRAM enable bit (DRAME) of the BCR is 0)
Areas 2–4
Area 5 (space where address A27 is 1)
Area 6 (when the multiplexed I/O enable bit (IOE) bit of the BCR is 0, or space where address
A27 is 1)
Area 7 (space where address A27 is 0)
8.4.1
Basic Timing
The bus cycle for external memory space access is 1 or 2 states. The number of states is controlled
with the wait states by the settings of wait state control registers 1–3 (WCR1–WCR3). For details,
see section 8.4.2., Wait State Control. Figures 8.11 and 8.12 illustrate the basic timing of external
memory space access.
T1
CK
A21–A0
CSn
RD
(Read)
AD15–AD0
(Read)
Figure 8.11 Basic Timing of External Memory Space Access (1-state read timing)
128 RENESAS
T1
T2
CK
A21–A0
CSn
When
RDDTY = 0
RD
When
RDDTY = 1
Read
AD15–AD0
WRH, WRL
Write
AD15–AD0
Figure 8.12 Basic Timing of External Memory Space Access (2-state read timing)
High-level duties of 35% and 50% can be selected for the RD signal using the RD duty bit
(RDDTY) of the BCR. When RDDTY is set to 1, the high-level duty is 35% of the T1 state,
enabling longer access times for external devices. Only set to 1 when the operating frequency is a
minimum of 10 MHz.
8.4.2
Wait State Control
The number of external memory space access states and the insertion of wait states can be
controlled using the WCR1–WCR3 bits. The bus cycles that can be controlled are the CPU read
cycle and the DMAC dual mode read cycle. The bus cycle that can be controlled using the WCR2
is the DMAC single-mode read/write cycle.
Table 8.8 shows the number of states and number of wait states in the access cycles to external
memory spaces.
RENESAS 129
Table 8.8
Number of States and Number of Wait States in the Access Cycles to External
Memory Spaces
CPU read cycle, DMAC dual mode read cycle,
DMAC single mode read/write cycle
Area
Corresponding Bits in
WCR1 and WCR2 = 0
Corresponding Bits in
WCR1 and WCR2 = 1
CPU Write Cycle and
DMAC Dual Mode Write
Cycle (Cannot be
controlled by WCR1)*2
1, 3–5, 7
1 cycle fixed; WAIT signal 2 cycles fixed + wait state from WAIT signal
ignored
0, 2, 6 (long
wait available)
1 cycle + long wait state,
WAIT signal ignored
1 cycle + long wait state*1 + wait state from WAIT
signal
Notes: 1. The number of long wait states is set by WCR3.
2. When DRAME = 1, short pitch/long pitch is selected with the WW1 bit of the WCR1.
3. Pin wait cannot be used for the CS7 and WAIT pins of area 3 because they are
multiplexed.
For the CPU read cycle, DMAC dual mode read cycle and DMAC single mode read/write cycle,
the access cycle is completed in 1 state when the corresponding bits of WCR1 and WCR2 for
areas 1, 3–5, and 7 are cleared to 0 and the WAIT pin input signal is not sampled. When the bits
are set to 1, the WAIT signal is sampled and the number of states is 2 plus the number of wait
states in the WAIT signal. The WAIT signal is sampled at the rise of the system clock (CK)
directly preceding the second state of the bus cycle and the wait states are inserted as long as the
level is low. When a high level is detected, it shifts to the second state (final state). Figure 8.13
shows the wait state timing when accessing the external memory spaces of areas 1, 3, 4, 5, and 7.
130 RENESAS
T1
Tw (wait state)
T2
CK
A21–A0
CSn
RD
Read
AD15–AD0
WRH, WRL
Write
AD15–AD0
WAIT
Figure 8.13 Wait State Timing for External Memory Space Access (2 states plus wait states
from WAIT signal)
Areas 0, 2 and 6 have long wait functions. When the corresponding bits in WCR1 and WCR2 are
cleared to 0, the access cycle is 1 state plus the number of long wait states (set in WCR3,
selectable between 1 and 4) and the WAIT pin input signal is not sampled. When the bits are set to
1, the WAIT signal is sampled and the number of states is 1 plus the number of long wait states
plus the number of wait states in the WAIT signal. The WAIT signal is sampled at the rise of the
system clock (CK) directly preceding the last long wait state and the wait states are inserted as
long as the level is low. When a high level is detected, it shifts to the final long wait state. Figure
8.14 shows the wait state timing when accessing the external memory spaces of areas 0, 2, and 6.
RENESAS 131
Wait states
set in WCR3
T1
TLW1
TLW2
Wait state
Wait
from WAIT states set
signal input in WCR3
TW
TLW3
CK
A21–A0
CSn
RD
Read
AD15–AD0
WRH, WRL
Write
AD15–AD0
WAIT
Figure 8.14 Wait State Timing for External Memory Space Access (1 state plus long wait
state (when set to insert 3 states) plus wait states from WAIT signal)
For CPU write cycles and DMAC dual mode write cycles to external memory space, the number
of states and wait state insertion cannot be controlled by WCR1. In areas 1, 3, 4, 5, and 7, the
WAIT signal is sampled and the number of states is 2 plus the number of wait states in the WAIT
signal (figure 8.13). In areas 0, 2 and 6, the number of states is 1 state plus the number of long
wait states plus the number of wait states in the WAIT signal (figure 8.14). Never write 0 in bits
7–2 and 0 of WCR1; only write 1. When area 1 is being used as external memory space, never
write 0 to bit 1 (WW1); always write 1.
132 RENESAS
8.4.3
Byte Access Control
The upper byte and lower byte control signals when 16-bit bus width space is being accessed can
be selected from (WRH, WRL, A0) or (WR, HBS, LBS). When the byte access select bit (BAS)
of the BCR is set to 1, the WRH, WRL, and A0 pins output WR, LBS and HBS signals. Figure
8.15 illustrates the control signal output timing in the byte write cycle.
Upper byte access
T1
T2
Lower byte access
T1
T2
CK
A0
BAS = 0
WRH
WRL
HBS
BAS = 1
LBS
WR
Figure 8.15 Byte Access Control Timing For External Memory Space Access (Write Cycle)
The WRH, WRL system and the HBS, LBS system are available as byte access signals for the 16bit space in the address/data multiplexing space and the external memory space.
These strobe signals are assigned to pins in the manner: A0/HBS, WRH/LBS, WRL/WR, and the
BAS bit of the bus control register (BCR) is used to switch specify signal sending.
Note that the byte access signals are strobe signals dedicated to byte access to a 16-bit space and
not to be used for byte access to an 8-bit space. When making an access to an 8-bit space, use the
A0/HBS pin as A0 irrespective of the BAS bit value (0 or 1) to use the WRL/WR pin as the WR
pin, and avoid using the WRH/LBS pin.
RENESAS 133
8.5
DRAM Interface Operation
When the DRAM enable bit (DRAME) of the BCR is set to 1, area 1 becomes DRAM space and
the DRAM interface function is available, which permits direct connection of this LSI to DRAMs.
8.5.1
DRAM Address Multiplexing
When the multiplex enable bit (MXE) of the DRAM area control register (DCR) is set to 1, row
addresses and column addresses are multiplexed. This allows DRAMs that require multiplexing of
row and column addresses to be connected directly to the SH microprocessors without additional
multiplexing circuits. When addresses are multiplexed (MXE = 1), setting of the DCR’s multiplex
shift bits (MXC1, MXC0) allows selection of eight, nine and ten-bit row address shifting. Table
8.9 illustrates the relationship between MXC1/MXC0 bits and address multiplexing.
134 RENESAS
Table 8.9
Relationship between Multiplex Shift Count Bits (MXC1, MXC0) and Address
Multiplexing
Shift Amount 8 bits
Output Pin
Output
Row
Address
Output
Column
Address
Shift Amount 9 bits
Output
Row
Address
Output
Column
Address
Shift Amount 10 bits
Output
Row
Address
Output
Column
Address
A21
A21
A21
A21
A20
A20
A20
A20
A19
Undefined
A19
A19
A19
A18
Value
A18
Undefined
A18
A18
A17
A17
Value
A17
Undefined
A17
A16
A16
A16
Value
A16
A15
A23
A15
A15
A15
A14
A22
A14
A23
A14
A14
A13
A21
A13
A22
A13
A23
A13
A12
A20
A12
A21
A12
A22
A12
A11
A19
A11
A20
A11
A21
A11
A10
A18
A10
A19
A10
A20
A10
A9
A17
A9
A18
A9
A19
A9
A8
A16
A8
A17
A8
A18
A8
A7
A15
A7
A16
A7
A17
A7
A6
A14
A6
A15
A6
A16
A6
A5
A13
A5
A14
A5
A15
A5
A4
A12
A4
A13
A4
A14
A4
A3
A11
A3
A12
A3
A13
A3
A2
A10
A2
A11
A2
A12
A2
A1
A9
A1
A10
A1
A11
A1
A0
A8
A0
A9
A0
A10
A0
Notes: The MXC1=1, MX0=1 setting is reserved. Do not use it.
For example, when MXC1 and MXC0 are set to 00 and an 8-bit shift is selected, the A23–A8
address bit values are output to pins A15–A0 as row addresses. The values for A21–A16 are
undefined. The values of bits address A21–A0 are output to pins A21–A0 as column addresses.
Figure 8.16 depicts address multiplexing with an 8-bit shift.
RENESAS 135
RAS = Low level
Internal address A23
Address pin
A8 A7
A21
A16 A15
A0
A0
Undefined output
CAS = Low level
Internal address A23 A22 A21
Address pin
A21
A0
A0
Figure 8.16 Address Multiplexing States (8-bit shift)
8.5.2
Basic Timing
There are two types of DRAM accesses: short pitch and long pitch. Short pitch or long pitch can
be selected for the respective bus cycles using the RW1 and WW1 bits of WCR1 and the DRW1
and DWW1 bits of WCR2. When the corresponding bits are cleared to 0, DRAM access is short
pitch and column address output occurs in 1 state. When these bits are 1, DRAM access is long
pitch and column address output occurs in 2 states. Figure 8.17 shows short pitch timing; figure
8.18 shows long pitch timing.
The high-level duty of the CAS signal can also be selected between 50% and 35% of the T C state
when access is short pitch. By setting the CDTY bit to 1, high level duty becomes 35% and
DRAM access time can be lengthened. Only set to 1 when the operating frequency is a minimum
of 10 MHz.
136 RENESAS
Tp
Tr
Tc
CK
A21–A0
Row address
Column address
RAS
CDTY
=0
CAS
CDTY
=1
WRH, WRL
Read
AD15–AD0
WRH, WRL
Write
AD15–AD0
Figure 8.17 Short Pitch Access Timing
RENESAS 137
Tp
Tr
Tc1
Tc2
CK
Row address
A21–A0
Column address
RAS
CAS
WRH, WRL
Read
AD15–AD0
WRH, WRL
Write
AD15–AD0
Figure 8.18 Long Pitch Access Timing
8.5.3
Wait State Control
Precharge State Control: When the microprocessor clock frequency is raised and the cycle
period shortened, 1 cycle may not always be sufficient for the precharge time for the RAS signal
when the DRAM is accessed. The BSC allows the precharge cycle to be set to 1 state or 2 states
using the RAS signal precharge cycles bit (TPC) of the DCR. When the TPC bit is 0, the
precharge cycle is 1 state; when TPC is 1, the precharge cycle is 2 states. Figure 8.19 shows the
timing when the precharge cycle is 2 states.
138 RENESAS
Tp1
Tp2
Tr
Tc 1
Tc2
CK
Row address
A21–A0
Column address
RAS
CAS
Figure 8.19 Precharge Timing (Long Pitch)
Control of Insertion of Wait States Using the WAIT Pin Input Signal: The number of wait
states inserted into the DRAM access cycle can be controlled by setting WCR1 and WCR2. When
the corresponding bits in WCR1 and WCR2 are cleared to 0, the column address output cycle ends
in 1 state and no wait states are inserted. When the bit is 1, the WAIT pin input signal is sampled
on the rise of the system clock (CK) directly preceding the second state of the column address
output cycle and the wait state is inserted as long as the level is low. When a high level is detected,
it shifts to the second state. Figure 8.20 shows the wait state timing in a long pitch bus cycle.
Tp
Tr
Tc 1
Tcw (wait state)
Tc2
CK
A21–A0
Row address
Column address
RAS
CAS
WAIT
Figure 8.20 Wait State Timing during DRAM Access (Long Pitch)
RENESAS 139
When the RW1 bit is set to 1, the number of wait states selected by CBR refresh wait state
insertion bits 1 and 0 (RLW1, RLW0) of the refresh control register (RCR) are inserted into the
CAS-before-RAS refresh cycle.
8.5.4
Byte Access Control
16-bit width and 18-bit width DRAMs require different types of byte control signals for access. By
setting the dual CAS signals/dual WE signals select bit (CW2) in the DCR, the BSC allows
selection of either the dual CAS signals or the dual WE signals system of control signals. When
16-bit space is being accessed and the CW2 bit is cleared to 0 for dual CAS signals, CASH,
CASL, and WRL signals are output; when CW2 is set to 1 for dual WE signals, the CASL, WRH,
and WRL signals are output. When accessing 8-bit space, WRL and CASL are output regardless
of the CW2 setting.
Figure 8.21 shows the control timing of the upper byte write cycle (short pitch) in 16-bit space.
140 RENESAS
Tp
Tr
Tc
Row address
Column address
CK
A21–A0
RAS
Byte
control
CASH
CASL
High level
WRH
High level fixed
WRL
(a) Dual CAS signals (CW2 = 0)
Tp
Tr
Tc
CK
A21–A0
Row address
Column address
RAS
CASH
High level fixed
CASL
Byte
control
WRH
WRL
High level
(a) Dual WE signals (CW2 = 1)
Figure 8.21 Byte Access Control Timing for DRAM Access (Upper Byte Write Cycle, Short
Pitch)
RENESAS 141
8.5.5
DRAM Burst Mode
In addition to the normal mode of DRAM access, in which row addresses are output at every
access and data then accessed (full access), the DRAM also has a high-speed page mode for use
when continuously accessing the same row. The high speed page mode enables fast access of data
simply by changing the column address after the row address is output (burst mode). Select
between full access and burst operation by setting the burst enable bit (BE)) in the DCR. When the
BE bit is set to 1, burst operation is performed when the row address matches the previous DRAM
access row address. Figure 8.22 shows the comparison of full access and burst operation.
RAS
CAS
Column
address 2
Column address 1
A21–A0
AD15–
AD0
Row address 1
Row address 2
Data 1
Data 2
(a) Full access (read cycle)
RAS
CAS
Column
Column Column Column
address 1 address 2 address 3 address 4
A21–A0
AD15–
AD0
Row address 1
Data 1
Data 2
Data 3
Data 4
(b) Burst operation (read cycle)
Figure 8.22 Full Access and Burst Operation
Short pitch high-speed page mode or long pitch high-speed page mode burst transfers can be
selected independently for DRAM read/write cycles even when the burst operation is selected by
using the bits corresponding to area 1 in WCR1 and WCR2 (RW1, WW1, DRW1, DWW1). The
RAS down mode or RAS up mode can be selected by setting the RAS down bit (RASD) of the
DCR when there is an access outside the DRAM space during burst operation.
142 RENESAS
Short Pitch High-Speed Page Mode and Long Pitch High-Speed Page Mode: When burst
operation is selected by setting the DCR’s BE bit to 1, the short pitch high-speed page mode or
long pitch high-speed page mode can be selected by setting the RW1, WW1, DRW1, and DWW1
bits of the WCR1 and WCR2.
•
Short-pitch, high-speed page mode: When the RW1, WW1, DRW1 and DWW1 bits in the
WCR1 and WCR2 are cleared to 0, and the corresponding DRAM access cycle is continuing,
the CAS signal and column address output cycles continue as long as the row addresses
continue to match. The column address output cycle is performed in 1 state and the WAIT
signal is not sampled. Figure 8.23 shows the read cycle timing for the short pitch high-speed
page mode.
Tp
Tr
Tc
Tc
Tc
Column
address 1
Column
address 2
Tc
CK
Column
Column
address 3 address 4
A21–
A0
Row address 1
RAS
CAS
WR
AD15–
A0
Data 1
Data 2
Data 3
Data 4
Figure 8.23 Short Pitch High-Speed Page Mode (Read Cycle)
When the write cycle continues for the same row address in the short pitch high-speed page
mode, an open cycle (silent cycle) is produced for 1 cycle only. This timing is shown in figure
8.24. Likewise, when a write cycle continues after the read cycle for the same row address, a
silent cycle is produced for 1 cycle. This timing is shown in figure 8.25. Note also that when
DRAM is written to in short-pitch, high-speed page mode when using DMAC single address
mode, a silent cycle is inserted in each transfer. The details of timing are discussed in section
20.3.3, Bus Timing.
RENESAS 143
Access A
Tp
Tr
Tc
Access B
Tc
Silent
cycle
Tc
Tc
CK
Column
Column
address A-1 address A-2
Column
Column
address B-1 address B-2
A21–
A0
Row address
RAS
CAS
WR
AD15–
A0
Data A-1 Data A-2
Data B-1 Data B-2
Note: Access A and B are examples of 32-bit data accesses in their respective 16-bit bus width
spaces.
Figure 8.24 Short Pitch High-Speed Page Mode (Write Cycle)
144 RENESAS
Access A (read)
Tp
Tr
Tc
Access B (write)
Tc
Silent
cycle
Tc
Tc
CK
Column
Column
address A-1 address A-2
Column
Column
address B-1 address B-2
A21–
A0
Row address
RAS
CAS
WR
AD15–
AD0
Read data A-1 Read data A-2 Write data B-1 Write data B-2
Note: Access A and B are examples of 32-bit data accesses in their respective 16-bit bus width
spaces.
Figure 8.25 Short Pitch High-Speed Page Mode (When read and write cycle continues with
the same row address)
The high-level duty of the CAS signal can be selected in the short pitch high-speed page mode
using the CAS duty bit (CDTY) in the DCR. When the CDTY bit is cleared to 0, high-level
duty is 50% of the TC state; when CDTY is set to 1, it is 35% of the TC state.
•
Long-pitch, high-speed page mode: When the RW1, WW1, DRW1, and DWW1 bits in WCR1
and WCR2 are set to 1, and the corresponding DRAM access cycle is continuing, the CAS
signal and column address output cycles (2 states) continue as long as the row addresses
continue to match. When the WAIT signal is detected at the low level, the second cycle of the
column address output cycle is repeated as the wait state. Figure 8.26 shows the timing for the
long pitch high-speed page mode. See section 20.3.3, Bus Timing, for more information about
the timing.
RENESAS 145
Tp
Tr
Tc1
Tc2
Tc1
Tc2
CK
Column address 1 Column address 2
A21–AD0
Row address 1
RAS
CAS
WR
Read
AD15–AD0
Data 1
Data 2
WR
Write
AD15–AD0
Data 1
Data 2
Figure 8.26 Long Pitch High-Speed Page Mode (Read/Write Cycle)
RAS Down Mode and RAS Up Mode: Sometimes access to another area can occur between
accesses to the DRAM even though burst operation has been selected. Keeping the RAS signal at
low while this other access is occurring allows burst operation to continue the next time the same
row of the DRAM is accessed. The RASD bit in the DCR selects the RAS down mode when set to
1 and the RAS up mode when cleared to 0. In both RAS down mode and RAS up mode, burst
operation is continued while the same row address continues to be accessed, even if the bus master
is changed.
•
RAS Down Mode: When the RASD bit of DCR is set to 1, the DRAM access pauses and the
RAS signal is held low throughout the access of the other space while waiting for the next
access to the DRAM area. When the row address for the next DRAM access is the same as the
previous DRAM access, burst operation continues. Figure 8.27 shows the timing of the RAS
down mode when external memory space is accessed during burst operation.
The RAS signal can be held down in the DRAM for a limited time; the RAS signal must be
returned to high within the specified limits even when the RAS down mode is selected since
the critical low level period is set. In this LSI, even when the RAS down mode is selected, the
RAS signal automatically reverts to high when the DRAM is refreshed, so the BSC’s refresh
control function can be employed to set a CAS-before-RAS refresh that will keep operation
within specifications. See section 8.5.6, Refresh Control, for details.
146 RENESAS
External memory
space access
DRAM access
Tp
Tr
Tc
Tc
T1
DRAM access
Tc
Tc
CK
Column
Column
address 1 address 2
External Column Column
memory address 3 address 4
A21–
A0
Row address
RAS
CAS
WR
AD15–
AD0
Data 1
Data 2
Data 3
Data 4
External
memory data
Figure 8.27 RAS Down Mode
•
RAS Up Mode: When the RASD bit is cleared to 0, the RAS signal reverts to high whenever a
DRAM access pauses for access to another space. Burst operation continues only while
DRAM access is continuous. Figure 8.28 shows the timing when an external memory space
access occurs during burst operation in the RAS up mode.
RENESAS 147
External memory
space access
DRAM access
Tp
Tr
Tc
Tc
T1
DRAM access
Tp
Tr
Tc
CK
Column Column External memory
address 1 address 2
address
Column
address 3
A21–
A0
Row address
Row address
RAS
CAS
AD15–
AD0
Data 1
Data 2
Data 3
External
memory data
Figure 8.28 RAS Up Mode
8.5.6
Refresh Control
The BSC has a function for controlling DRAM refreshing. By setting the refresh mode bit
(RMODE) in the refresh control register (RCR), either CAS-before-RAS refresh (CBR) or selfrefresh can be selected. When no refresh is performed, the refresh timer counter (RTCNT) can be
used as an 8-bit interval timer.
CAS-Before-RAS Refresh (CBR): A refresh is performed at an interval determined by the input
clock selected in the clock select bits 2–0 (CKS2–CKS0) of the refresh timer control/status
register (RTCSR) and the value set in the refresh time constant register (RTCOR). Set the values
of RTCOR and CKS2–CKS0 so they satisfy the refresh interval specifications of the DRAM being
used.
To perform a CBR refresh, clear the RMODE bit of the RCR to 0 and then set the refresh control
bit (RFSHE) bit to 1. Also write in the required values to RTCNT and RTCOR. When the clock is
thereafter selected in the CKS2–CKS0 bits of the RTCSR, the RTCNT will begin to increment
from its current value. The RTCNT value is constantly compared to the RTCOR value and the
CBR refresh is performed when they match. The RTCNT is simultaneously cleared to H'00 and
incrementing begins again.
When the clock is selected in the CKS2–CKS0 bits, the RTCNT immediately begins to increment
from its current value. This means that when the RTCOR cycle is set after the CKS2–CKS0 bits
are set, the RTCNT count may already be higher than the RTCOR cycle. When this occurs, the
RTCNT will overflow once (H'FF goes to H'00) and incrementing will start again. Since the CBR
148 RENESAS
refresh will not be performed until the RTCNT again matches the RTCOR value, the initial refresh
interval will be rather long. It is thus advisable to set the RTCOR cycle prior to setting the CKS2–
CKS0 bits and start it incrementing. When CBR refresh control is being performed after its use as
an 8-bit interval timer, the RTCNT count value may be in excess of the refresh cycle. For this
reason, clear the RTCNT by writing H'00 before starting refresh control to assure a correct refresh
interval.
When the RW1 bit of WCR1 is set to 1 and the read cycle is set to long pitch, the number of wait
states selected by the RLW1 and RLW0 bits of the RCR will be inserted into the CBR refresh
cycle, regardless of the status of the WAIT signal. Figure 8.29 shows the RTCNT operation and
figure 8.30 shows the timing of the CBR refresh. For details on timing, see section 20.3.3, Bus
Timing.
RTCNT
value
RTCOR
value
Compare
match
with RTCOR
Compare
match
with RTCOR
Compare
match
with RTCOR
Compare
match
with RTCOR
H'00
Time
Clock
CBR
selected with
CKS2–CKS0
CBR
CBR
CBR
CBR: CAS-before-RAS refresh
Figure 8.29 Refresh Timer Counter (RTCNT) Operation
TRp
TRr
TRc
CK
RAS
CAS
Figure 8.30 Output Timing for CAS-Before-RAS Refresh Signal
Self-Refresh Mode: Some DRAMs have a self-refresh mode (parity back-up mode). This is a
type of a standby mode in which the refresh timing and refresh addresses are generated inside the
DRAM chip. When the RFSHE and RMODE bits of the RCR are both set to 1, the DRAM will
RENESAS 149
enter the self-refresh mode when the CAS and RAS signals are output as shown in figure 8.31. See
section 20.3.3, Bus Timing, for details. The DRAM self-refresh mode is cleared when the
RMODE bit in the RCR is cleared to 0 (figure 8.31). The RFSHE bit should be left at 1 when this
is done. Some DRAM vendors recommend that after exiting the self-refresh mode, all row
addresses should be refreshed again. This can be done using the BCR’s CBR refresh function to
set all row addresses for refresh in software.
To access a DRAM area in the self-refresh mode, clear the RMODE bit to 0 and exit the selfrefresh mode.
The LSI can be kept in the self-refresh state and shifted to standby mode by setting it to selfrefresh mode, setting the standby bit (SBY) of the standby control register (SBYCR) to 1, and then
executing a SLEEP instruction.
TRp
TRr
TRc
TRcc
CK
RAS
CAS
Figure 8.31 Output Timing of Self-Refresh Signal
Refresh Requests and Bus Cycle Requests: When a CAS-before-RAS refresh or self-refresh is
requested during bus cycle execution, parallel execution is sometimes possible. Table 8.10
describes operation when the refresh and bus cycle are in contention.
150 RENESAS
Table 8.10 Refresh and Bus Cycle Contention
Type of Bus Cycle
External Space Access
External Memory Space,
Multiplexed I/O Space
On-Chip ROM, On-Chip
RAM , On-Chip
Read Cycle Write Cycle Read Cycle Write Cycle Peripheral Access
Type of
Refresh
DRAM Space
CAS-beforeRAS refresh
Yes
No
No
No
Yes
Self-refresh
Yes
Yes
No
No
Yes
Yes: Can be executed in parallel
No: Cannot be executed in parallel
When parallel execution is available, the RAS and CAS signals are output simultaneously during
bus cycle execution and the refresh is executed. When parallel execution is not available, refresh
occurs after the bus cycle has ended.
Using RTCNT as an 8-Bit Interval Timer: When not performing refresh control, RTCNT can be
used as an 8-bit interval timer. Simply set the RFSHE bit of the RCR to 0. To produce a compare
match interrupt (CMI), set the compare match interrupt enable bit (CMIE) to 1 and set the
interrupt generation timing in RTCOR. When the input clock is selected with the CKS2–CKS0
bits of the RTCSR, RTCNT starts incrementing as an 8-bit interval timer. Its value is constantly
being compared to RTCOR and when a match occurs, the CMF bit of RTCSR is set to 1 and a
CMI interrupt is produced. RTCNT is cleared to H'00.
When the clock is selected with CKS2–CKS0 bits, RTCNT starts incrementing immediately. This
means that when the RTCOR cycle is set after the CKS2–CKS0 bits are set, the RTCNT count
may already be higher than the RTCOR cycle. When this occurs, the RTCNT will overflow once
(H'FF goes to H'00) and the count up will start again. No interrupt will be generated until the
RTCNT again matches the RTCOR value. It is thus advisable to set the RTCOR cycle prior to
setting the CKS2–CKS0 bits. After its use as an 8-bit interval timer, the RTCNT count value may
be in excess of the set cycle. For this reason, write H'00 to the RTCNT to clear it before starting to
use it again with new settings. RTCNT can then be restarted and an interrupt obtained after the
correct interval.
8.6
Address/Data Multiplexed I/O Space Access
The BSC is equipped with a function that multiplexes input/output of address and data to pins
AD15–AD0 in area 6. This allows the SH microprocessor to be directly connected to peripheral
LSIs that required address/data multiplexing.
RENESAS 151
8.6.1
Basic Timing
When the multiplexed I/O enable bit (IOE) of the BCR is set to 1, the area 6 space with address bit
A27 as 0 (H'6000000–H'6FFFFFF) becomes an address/data multiplexed I/O space that, when
accessed, multiplexes addresses and data. When the A14 address bit is 0, the bus width is 8 bits
and address output and data input/output are performed from the AD7–AD0 pins. When the A14
address bit is 1, the bus width is 16 bits and address output and data input/output are performed
from the AD15–AD0 pins. In the address/data multiplexed I/O space, access is controlled with the
AH, RD and WR signals. Accesses in the address/data multiplexed I/O space is performed in 4
states, regardless of the WCR settings. Figure 8.32 shows the timing when the address/data
multiplexed I/O space is accessed.
T1
T2
T3
T4
CK
A21–A0
CS
AH
RD
Read
AD15–AD0
Address
Data (input)
Address
Data (output)
WRH, WRL
Write
AD15–AD0
Figure 8.32 Access Timing For Address/Data Multiplexed I/O Space
The high-level duty of the RD signal can be selected between 35% and 50% using the RD duty bit
(RDDTY) of the BCR. When RDDTY is 1, the high-level duty is 35% of the T3 or Tw state,
lengthening the access time for external devices.
152 RENESAS
8.6.2
Wait State Control
When the address/data multiplexed I/O space is accessed, the WAIT pin input signal is sampled
and a wait state inserted whenever a low level is detected, regardless of the setting of the WCR.
Figure 8.33 shows an example in which a WAIT signal causes a wait state of 1 state to be inserted.
T1
T2
T3
Tw
(wait state)
T4
CK
A21–A0
CS
AH
RD
Read
AD15–AD0
Address
Data (input)
WRH, WRL
Write
AD15–AD0
Address
Data (output)
WAIT
Figure 8.33 Wait State Timing For Address/Data Multiplexed I/O Space Access
8.6.3
Byte Access Control
The byte access control signals when the address/data multiplexed I/O space is being accessed are
of two types (WRH, WRL, A0, or WR, HBS, LBS), just as for byte access control of external
memory space access. These types can be selected using the BAS bit of the BCR. See section
8.4.3, Byte Access Control, for details.
RENESAS 153
8.7
Parity Check and Generation
The BSC can check and generate parity for data input and output to or from in the DRAM space of
area 1 and the external memory space of area 2.
To check and generate parity, select the space (DRAM space only, or DRAM space and area 2) for
which parity is to be checked and generated using the parity check enable bits (PCHK1 and
PCHK0) of the parity control register and select odd or even parity in the parity polarity bit (PEO).
When data is input from the space selected in the PCHK1 and PCHK0 bits, the BSC checks the
PEO bit to see if the polarity of the DPH pin input (upper byte parity data) is accurate for the
AD15–AD8 pin input (upper byte data) or if the DPL pin input (lower byte parity data) is accurate
for the AD7–AD0 pin input (lower byte data). If the check indicates that either the upper or lower
byte parity is incorrect, a parity error interrupt is produced (PEI).
When outputting data to the space selected in the PCHK1 and PCHK0 bits, the BSC outputs parity
data output of the polarity set in the PEO bit from the DPH pin for the AD15–AD8 pin output
(upper byte data) or from the DPL pin for the AD7–AD0 pin input (lower byte data) using the
same timing as the data output.
The BSC is also able to force a parity output for use in testing the system’s parity error check
function. When the parity force output bit (PFRC) of the PCR is set to 1, a high level is forcibly
output from the DPH and DPL pins when data is output to the space selected in the PCHK1 and
PCHK0 bits.
8.8
Warp Mode
In warp mode, an external write cycle or DMA single address mode transfer cycle and an internal
access cycle (read/write to on-chip memory or on-chip peripheral modules) operate independently
in parallel. The warp mode is entered by setting the warp mode bit (WARP) in the BCR to 1. This
allows the LSI to be operated at high speed.
When in the warp mode an external write cycle or DMA single address mode transfer cycle
continues for at least 2 states and their is an internal access, only the external write cycle will be
performed in the initial state. The external write cycle and internal access cycle will be performed
in parallel from the next state on, without waiting for the end of the external write cycle. Figure
8.34 shows the timing when an access to an on-chip peripheral module and an external write cycle
are performed in parallel.
154 RENESAS
External space writing
On-chip peripheral module read/write
T1
T2
T3
T4
T5
CK
A21–
A0
External
space
write
External space address
CSn
WR
AD15–
AD0
Write data
External space
address
Internal address
On-chip
peripheral
module
write
On-chip
peripheral
module
read
On-chip peripheral module address
Internal
write
strobe
Internal
data bus
Write data
Internal
read
strobe
Internal
data bus
Read data
Figure 8.34 Warp Mode Timing (Access to On-Chip Peripheral Module and External Write
Cycle)
8.9
Wait State Control
The WCR1–WCR3 registers of the BSC can be set to control sampling of the WAIT signal when
accessing various areas and the number of bus cycle states. Table 8.11 shows the number of bus
cycle states when accessing various areas.
RENESAS 155
Table 8.11 Bus Cycle States when Accessing Address Spaces
CPU Read Cycle, DMAC Dual Mode Read Cycle, DMAC Single
Mode Memory Read/Write Cycle
Address Space
Corresponding Bits in WCR1
and WCR2 = 0
Corresponding Bits in WCR1
and WCR2 = 1
External memory (areas
1, 3–5, 7)
1 state fixed; WAIT signal ignored 2 states + wait states from WAIT
signal
External memory (Areas
0, 2, 6; long wait available)
1 state + long wait state*, WAIT
signal ignored
1 state + long wait state* + wait
states from WAIT signal
DRAM space (area 1)
Column address cycle: 1 state,
WAIT signal ignored (short pitch)
Column address cycle: 2 states +
wait states from WAIT signal
(long pitch)
Multiplexed I/O space
(area 6)
4 states + wait states from WAIT signal
On-chip peripheral module space (area 5)
3 states fixed, WAIT signal ignored
On-chip ROM (area 0)
1 state fixed, WAIT signal ignored
On-chip RAM (area 7)
1 state fixed, WAIT signal ignored
CPU Write Cycle, DMAC Dual Mode Memory Write Cycle
Address Space
WW1 of WCR1 = 0
External memory (areas
1, 3–5, 7)
2 states + wait states from WAIT signal
External memory (Areas
0, 2, 6; long wait
available)
1 state + long wait state* + wait states from WAIT signal
DRAM space (area 1)
Column address cycle: 1 state,
WAIT signal ignored (short pitch)
Multiplexed I/O space
(area 6)
4 states + wait states from WAIT signal
On-chip peripheral
module space (area 5)
3 states fixed, WAIT signal ignored
On-chip ROM (area 0)
1 state fixed, WAIT signal ignored
On-chip RAM (area 7)
1 state fixed, WAIT signal ignored
Note: The number of long wait states (1 to 4) is set in WCR3.
156 RENESAS
WW1 of WCR1 =1
Column address cycle: 2 states +
wait states from WAIT signal
(long pitch)
For details on bus cycles when external spaces are accessed, see section 8.4, External Memory
Space Access, section 8.5, DRAM Space Access, and section 8.6, Address/Data Multiplexed I/O
Space Access.
Accesses of on-chip spaces are as follows: On-chip peripheral module spaces (area 5 when address
bit A27 is 1) are always 3 states, regardless of the WCR, with no WAIT signal sampling. Accesses
of on-chip ROM (area 0 when MD2–MD0 are 010) and on-chip RAM (area 7 when address bit
A27 is 0) are always performed in 1 state, regardless of the WCR, with no WAIT signal sampling.
If the bus timing specifications (tWTS and tWTH) are not observed when the WAIT signal is input
in external space access, this will simply mean that WAIT signal assertion and negation will not
be detected, but will not result in misoperation. Note, however, that the inability to detect WAIT
signal assertion may result in a problem with memory access due to insertion of an insufficient
number of waits.
8.10
Bus Arbitration
The SuperH microcomputer can release the bus to external devices when they request the bus. It
has two internal bus masters, the CPU and the DMAC. Priorities for releasing the bus for these
two are as follows.
Bus request from external device > refresh > DMAC > CPU
Thus, an external device has priority when it generates a bus request, even when the DMAC is
doing a burst transfer.
Note that when a refresh request is generated while the bus is released to an external device,
BACK becomes high level and the bus right can be acquired to perform the refresh upon receipt of
a BREQ = high level response from the external device. Input all bus requests from external
devices to the BREQ pin. The signal indicating that the bus has been released is output from the
BACK pin. Figure 8.35 illustrates the bus release procedure.
RENESAS 157
External device
SuperH
BREQ = low
Bus request
BREQ received
Strobe pin:
High-level output
Address, data, strobe pin:
High impedance
BACK = low
Bus release response
BACK acknowledge
Bus acquisition
Bus released
Figure 8.35 Bus Release Procedure
8.10.1
The Operation of Bus Arbitration
This LSI has the bus arbitration function which can give bus ownership to an external device when
the device requests the bus ownership. When BREQ is input and the bus cycle being executed by
the CPU or DMAC is completed, BACK becomes low and a bus is released for an external device.
At this time, the following operates when bus arbitration conflicts with refresh.
1. If DRAM refresh is requested in this LSI when a bus is released and BACK is low, BACK
becomes high and the occurrence of the refresh request can be informed externally. At this
time, the external device may generate a bus cycle when BREQ is low even if BACK is high.
Therefore, a bus remains released to the external device. Then, when BREQ becomes high, this
LSI gets bus ownership, and executes refresh and the bus cycle of the CPU or DMAC. After
the external device gets bus ownership and BACK is low, refresh is requested when BACK
becomes high even if the low level of BREQ is input. Therefore, turn BREQ high immediately
to release a bus for this LSI to hold DRAM data (See figure 8.36).
2. When BREQ changes from high to low and internal refresh is requested at the timing of the
bus release of this LSI, BACK may remain high (do not become low). A bus is released to the
external device since the low level of BREQ is input. This operation is based on the above
specification (1). To hold DRAM data, turn BREQ high and release a bus to this LSI
immediately when the external device detects that BACK does not change to low during a
fixed time this LSI (See figure 8.37). When a refresh request is generated and BACK returns
to high, as shown in figure 8.37, a momentary narrow pulse-shaped spike may be output where
BACK was originally supposed to become low.
158 RENESAS
BREQ
BACK
Refresh execution
Refresh damand
Figure 8.36 BACK Operation by Refresh Demand (1)
If BACK has not gone low after waiting for the maximum
number of states* before the SuperH releases the bus, return
BREQ to the high level.
BREQ
BACK
BACK does not go low.
Refresh request
Note: * For details see section 8.11.3, Maximum Number of States from BREQ Input to Bus Release.
Figure 8.37 BACK Operation in Response to Refresh Request (2)
8.10.2
BACK Operation
1. BACK Operation
When an internal refresh is requested during an attempt to assert the BACK signal and BACK
is not asserted but remains high, a momentary narrow pulse-shaped spike may be output, as
shown below.
BREQ
BACK
pulse width of the spike is approx. 2 to 5 ns.
Refresh demand
RENESAS 159
2. Countermeasure against a spike on the BACK signal
The following describes the countermeasure against a spike on the BACK signal:
a. When BREQ is input to release the bus of the LSI, make sure that conflicts with a refresh
operation do not occur. Stop the refresh operation or operate the refresh timer counter
(RTCNT) or the refresh time constant register (RTCOR) of the bus controller (BSC) to
shift the refresh timing.
b. The spike on the BACK signal has a narrow pulse width of approximately 2 to 5 ns, which
can be eliminated by using a capacitor as shown in the figure below.
For example, adding a capacitance of 220 pF can raise the minimum voltage of the spike
above 2.0 V.
Note that delay of the BACK signal increases approximately in units of 0.1 ns/pF. (When a
capacitance of 220 pF is added, the delay increases approximately by 22 ns.
BACK
C
SuperH
Microcomputer
Capacitor-incorporating circuit for eliminating a spike
c. Latching the BACK signal by using a flip-flop or triggering the flip-flop may be successful
or unsuccessful due to the narrow pulse width of the spike. Implement a circuit
configuration which will cause no problems when latching BACK or using BACK as a
trigger signal.
When splitting the BACK signal into two signals and latching each of them using the flipflop or triggering the flip-flop, the flip-flop may operate for one signal but may not for
another. To capture the BACK signal using the flip-flop, receive the BACK signal using a
single flip-flop then distribute the signal (see figure below).
160 RENESAS
Trigger OK
D Q
BACK
Q
Trigger NG
D Q
BACK
Q
D Q
Q
8.11
Usage Notes
8.11.1
Usage Notes on Manual Reset
Condition: When DRAM (long-pitch mode) is used and manual reset is performed.
The low width of RAS output may be shorter than usual in rese + (2.5tcyc→1.5tcyc), causing the
specified value (tRAS) of DRAM not to be satisfied.
Corresponding DRAM conditions: long pitch/normal mode
long pitch/high-speed page mode
There are no problems regarding operations except for the above conditions.
There are the following four cases (Figure 8.38 to Figure 8.41) for the output states of DRAM
control signals (RAS, CAS, and WR) corresponding to RES latch timing. Actual output levels are
shown by solid lines (not by dashed lines).
RENESAS 161
Tp
RES latch
timing
Tr
Tc1
Tc2
CK
Manual reset
RES
A0 to A21
Row address Colum address
FFFF
RAS
CAS
WR
AD0 to AD15
Data output
Figure 8.38 Long - pitch Mode Write (1)
Tp
RES latch
timing
Tr
Tc1
CK
RES
A0 to A21
Manual reset
Row address
FFFF
RAS
CAS
WR
AD0 to AD15
Data output
Figure 8.39 Long - pitch Mode Write (2)
162 RENESAS
Tc2
RES latch
timing
Tp
Tr
Tc1
Tc2
CK
Manual reset
RES
A0 to A21
Row address Colum address
FFFF
RAS
CAS
RD
Figure 8.40 Long - pitch Mode Read (1)
Tp
RES latch
timing
Tr
Tc1
Tc2
CK
RES
A0 to A21
Manual reset
Row address
FFFF
RAS
CAS
RD
Figure 8.41 Long - pitch Mode Read (2)
For the signal output shown by solid lines, DRAM data may not be held. Therefore, when DRAM
data must be held after reset, take one of the coutermeasures described as follows.
1. When resetting manually, do this in watchdog timer (WDT) condition.
2. Even if the Low width of RAS becomes as short as 1.5 tcyc as shown above, use with a
frequency that satisfies the DRAM standard (tRAS).
3. Even in case the Low width of RAS has become 1.5 tcyc, proceed by using the external circuit
so that a RAS signal with a Low width of 2.5 tcyc is input in the DRAM (in case the Low
width of RAS is higher than 2.5 tcyc, operate so that the current waveform is input in the
DRAM).
RENESAS 163
The countermeasures are not required when DRAM data is initialized or loaded again after manual
reset.
8.11.2
Usage Notes on Parity Data Pins DPH and DPL
The following specifies the setup time tDS of the parity dada DPH and DPL to CAS signal rising
when the parity dada DPH and DPL are written to DRAM in long-pitch mode (early write).
Table 8.12 Setup Time of Parity Data DPH and DPL
Item
Symbol
min
tDS
–5 ns
Data setup time to CAS
(for only DPH and DPL in long-pitch mode)
Therefore, when writing parity data DPH and DPL to the DRAM in long-pitch mode, delay the
WRH and WRL signals of this LSI and write with delayed writing.
Nomal dada is also delayed-written, causing no problems.
SuperH
RAS
RAS
Microcomputer CAS
CAS
OE
RD
*1
WRH or WRL
CK
*1
D
*2
Q
Q
DWRH or DWRL
WE
*1: For preventing signal racing
*2: Negative edge latch
Figure 8.42 Delayed-Write Control Circuit
8.11.3
Maximum Number of States from BREQ Input to Bus Release
The maximum number of states from BREQ input to bus release is:
Maximum number of states for which bus is not released + approx. 4.5 states
Note: Breakdown of approx. 4.5 states:
1.5 states:
Until BACK output after end of bus cycle
1 state (min.):
tBACD1
1 state (max.):
tBRQS
1 state:
Sampling in 1 state before end of bus cycle
164 RENESAS
DRAM
BREQ is sampled one state before the bus cycle. If BREQ is input without satisfying tBRQS, the
bus is released after executing cycle B following the end of bus cycle A, as shown in figure 8.43.
The maximum number of states from BREQ input to bus release are used when B is a cycle
comprising the maximum number of states for which the bus is not released; the number of states
is the maximum number of states for which bus is not released + approx. 4.5 states.
The maximum number of states for which the bus is not released requires careful investigation.
CK
Bus cycle
A
B
BREQ
tBACD1
tBRQS
BACK
Bus release
Figure 8.43 When BREQ is Input without Satisfying tBRQS
1. Cycles in which bus is not released
(a) One bus cycle
The bus is never released during one bus cycle. For example, in the case of a longword
read (or write) in 8-bit ordinary space, one bus cycle consists of 4 memory accesses to 8-bit
ordinary space, as shown in figure 8.44. The bus is not released between these accesses.
Assuming one memory access to require 2 states, the bus is not released for a period of 8
states.
8 bits 8 bits 8 bits 8 bits
Cycle during which
bus is not released
Figure 8.44 One Bus Cycle
(b) TAS instruction read cycle and write cycle
The bus is never released during a TAS instruction read cycle and write cycle (figure 8.45).
The TAS instruction read cycle and write cycle should be regarded as one bus cycle during
which the bus is not released.
RENESAS 165
Read cycle
Write cycle
Cycle during which bus is
not released (1 bus cycle)
Figure 8.45 TAS Instruction Read Cycle and Write Cycle
(c) Refresh cycle + bus cycle
The bus is never released during a refresh cycle and the following bus cycle ((a) or (b)
above)) (figure 8.46).
Refresh cycle
1 bus cycle
Cycle during which bus
is not released
Figure 8.46 Refresh Cycle and Following Bus Cycle
166 RENESAS
2. Bus release procedure
The bus release procedure is shown in figure 8.47. Figure 8.47 shows the case where BREQ is
input one state before the break between bus cycles so that tBRQS is satisfied. In the SH7020
and SH7021, the bus is released after the bus cycle in which BREQ is input (if BREQ is input
between bus cycles, after the bus cycle starting next).
CK
tBRQS
tBRQS
BREQ
tBACD1
tBACD1
BACK
tBZD
RD, WR
RAS, CAS
CSn
tBZD
A21 to A0
Bus cycle
Bus release
Strobe pin:
high-level output
Bus cycle
Address & data
strobe pins:
high impedance
The bus is released after the bus
cycle in which BREQ is input
(if BREQ is input between bus cycles,
after the bus cycle starting next).
Bus cycle restart
Figure 8.47 Bus Release Procedure
RENESAS 167
Section 9 Direct Memory Access Controller (DMAC)
9.1
Overview
The SuperH microprocomputer chip includes a four-channel direct memory access controller
(DMAC). The DMAC can be used in place of the CPU to perform high speed transfers between
external devices that have DACK (transfer request acknowledge signal), external memory,
memory-mapped external devices, on-chip memory and on-chip peripheral modules (excluding
the DMAC itself). Using the DMAC reduces the burden on the CPU and increases overall
operating efficiency.
9.1.1
Features
The DMAC has the following features.
•
•
•
•
•
•
Four channels
Four Gbytes of address space on the architecture
Byte or word selectable data transfer unit
65536 transfers (maximum)
Single address mode transfers (channels 0 and 1): Either the transfer source or transfer
destination (peripheral device) is accessed by a DACK signal (selectable) while the other is
accessed by address. 1 transfer unit of data is transferred in each bus cycle.
Device combinations able to transfer:
 External devices with DACK and memory-mapped external devices (including external
memories)
 External devices with DACK and memory-mapped external memories
Dual address mode transfer: (channels 0–3): Both the transfer source and transfer destination
are accessed by address. 1 transfer unit of data is transferred in 2 bus cycles.
Device combinations able to transfer:
 Two external memories
 External memory and memory-mapped external devices
 Two memory-mapped devices
 External memory and on-chip memory
 Memory-mapped external devices and on-chip peripheral module (excluding the DMAC
itself)
 External memory and on-chip memory
 Memory-mapped external device and on-chip peripheral module (excluding the DMAC)
 Two on-chip memories
 On-chip memory and on-chip peripheral modules (excluding DMAC)
RENESAS 169
•
•
•
•
•
 Two on-chip peripheral modules (excluding DMAC)
Transfer requests
 External request (From DREQ pins (channels 0 and 1 only). DREQ can be detected either
by edge or by level)
 Requests from on-chip peripheral modules (serial communications interface (SCI), and 16bit integrated-timer pulse unit (ITU))
 Auto-request (the transfer request is generated automatically within the DMAC)
Selectable bus modes: Cycle-steal mode or burst mode
Selectable channel priority levels: Fixed, round-robin, or external-pin round-robin modes
CPU can be asked for interrupt when data transfer ends
Maximum transfer rate
 20 M words/s (320 MB/s)
For 5V and 20 MHz
Bus mode: Burst mode
Transmit size: Word
9.1.2
Block Diagram
Figure 9.1 is a block diagram of the DMAC.
170 RENESAS
On-chip
ROM
On-chip
RAM
On-chip
peripheral
module
SARn
Register
control
DREQ0, DREQ1
ITU
SCI
Start-up
control
TCRn
CHCRn
DMAC module bus
Iteration
control
Internal bus
Peripheral bus
DARn
DMAOR
Request
priority
control
DACK0, DACK1
DEIn
External
RAM
External device
(memory
mapped)
External device
(with
acknowledge)
External bus
External
ROM
Bus interface
Bus controller
DMAC
DMAOR: DMA operation register
SARn: DMA source address register
DARn: DMA destination address register
TCRn: DMA transfer count register
CHCRn: DMA channel control register
DEIn: DMA transfer-end interrupt request to CPU.
n: 0–3
Figure 9.1 DMAC Block Diagram
RENESAS 171
9.1.3
Pin Configuration
Table 9.1 shows the DMAC pins.
Table 9.1
Pin Configuration
Channel
Name
Symbol
I/O
Function
0
DMA transfer request
DREQ0
I
DMA transfer request input from
external device to channel 0
DMA transfer request
acknowledge
DACK0
O
DMA transfer request acknowledge
output from channel 0 to external
device
DMA transfer request
DREQ1
I
DMA transfer request input from
external device to channel 1
DMA transfer request
acknowledge
DACK1
O
DMA transfer request acknowledge
output from channel 1 to external
device
1
172 RENESAS
9.1.4
Register Configuration
Table 9.2 summarizes the DMAC registers. DMAC has a total of 17 registers. Each channel has
four control registers. One other control register is shared by all channels
Table 9.2
DMAC Registers
Channel
Name
Abbreviation
R/W
Initial
Value
0
DMA source address register 0
SAR0*3
R/W
Undefined H'5FFFF40
16, 32
DMA destination address
register 0
DAR0* 3
R/W
Undefined H'5FFFF44
16, 32
DMA transfer count register 0
TCR0*3
R/W
Undefined H'5FFFF4A 16, 32
DMA channel control register 0
CHCR0
R/(W)*1 H'0000
DMA source address register 1
SAR1*3
R/W
Undefined H'5FFFF50
16, 32
DMA destination address
register 1
DAR1* 3
R/W
Undefined H'5FFFF54
16, 32
DMA transfer count register 1
TCR1*3
R/W
Undefined H'5FFFF5A 16, 32
DMA channel control register 1
CHCR1
R/(W)*1
H'0000
DMA source address register 2
SAR2*3
R/W
Undefined H'5FFFF60
16, 32
DMA destination address
register 2
DAR2* 3
R/W
Undefined H'5FFFF64
16, 32
DMA transfer count register 2
TCR2*3
R/W
Undefined H'5FFFF6A 16, 32
DMA channel control register 2
CHCR2
R/(W)*1
H'0000
DMA source address register 3
SAR3*3
R/W
Undefined H'5FFFF70
16, 32
DMA destination address
register 3
DAR3* 3
R/W
Undefined H'5FFFF74
16, 32
DMA transfer count register 3
TCR3*3
R/W
Undefined H'5FFFF7A 16, 32
CHCR3
R/(W)*1
H'0000
H'5FFFF7E 8, 16, 32
DMAOR
R/(W)*2
H'0000
H'5FFFF48
1
2
3
DMA channel control register 3
Shared DMA operation register
Address
Access
Size
H'5FFFF4E 8, 16, 32
H'5FFFF5E 8, 16, 32
H'5FFFF6E 8, 16, 32
8, 16, 32
Notes: 1. Write 0 alone in bit 1 of CHCR0–CHCR3 to clear flags.
2. Write 0 alone in bits 1 and 2 of the DMAOR to clear flags.
3. Access SAR0–SAR3, DAR0–DAR3, and TCR0–TCR3 by long word or word. If byte
access is used when writing, the value of the register contents becomes undefined; if
used when reading, the value read is undefined.
RENESAS 173
9.2
Register Descriptions
9.2.1
DMA Source Address Registers 0–3 (SAR0–SAR3)
DMA source address registers 0–3 (SAR0–SAR3) are 32-bit read/write registers that specify the
source address of a DMA transfer. During a DMA transfer, these registers indicate the next source
address (in single-address mode, SAR is ignored in transfers from external devices with DACK to
memory-mapped external devices or external memory).
The initial value after resets or in standby mode is undefined.
Bit:
31
30
29
28
27
26
25
24
—
—
—
—
—
—
—
—
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
23
22
21
20
Bit name:
Initial value:
R/W:
Bit:
…
Bit name:
Initial value:
R/W:
9.2.2
0
…
—
—
—
—
…
—
R/W
R/W
R/W
R/W
…
R/W
DMA Destination Address Registers 0–3 (DAR0–DAR3)
DMA destination address registers 0–3 (DAR0–DAR3) are 32-bit read/write registers that specify
the destination address of a DMA transfer. During a DMA transfer, these registers indicate the
next destination address (in single-address mode, DAR is ignored in transfers from memorymapped external devices or external memory to external devices with DACK). The initial value
after resets or in standby mode is undefined.
Bit:
31
30
29
28
27
26
25
24
—
—
—
—
—
—
—
—
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
23
22
21
20
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
174 RENESAS
…
0
…
—
—
—
—
…
—
R/W
R/W
R/W
R/W
…
R/W
9.2.3
DMA Transfer Count Registers 0–3 (TCR0–TCR3)
DMA transfer count registers 0-3 (TCR0–TCR3) are 16-bit read/write registers that specify the
DMA transfer count (bytes or words). The number of transfers is 1 when the setting is H'0001,
65535 when the setting is H'FFFF and 65536 (the maximum) when H'0000 is set. During a DMA
transfer, these registers indicate the remaining transfer count. The initial value after resets or in
standby mode is undefined.
Bit:
15
14
13
12
11
10
9
8
—
—
—
—
—
—
—
—
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
9.2.4
—
—
—
—
—
—
—
—
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
DMA Channel Control Registers 0–3 (CHCR0–CHCR3)
DMA channel control registers 0–3 (CHCR0–CHCR3) are 16-bit read/write registers that control
the DMA transfer mode. They also indicate DMA transfer status. They are initialized to H'0000 by
a reset or standby mode.
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
15
14
13
12
11
10
9
8
DM1
DM0
SM1
SM0
RS3
RS2
RS1
RS0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
AM
AL
DS
TM
TS
IE
TE
DE
0
0
0
0
0
0
0
0
R/(W)*2
R/(W)*2
R/(W)*2
R/W
R/W
R/W
R/(W)*
R/W
Notes: 1. Write only 0 to clear the flag.
2. Writing is effective only for CHCR0 and CHCR1.
RENESAS 175
•
Bits 15 and 14 (destination address mode bits 1, 0 (DM1 and DM0)): DM1 and DM0 select
whether the DMA destination address is incremented, decremented, or left fixed (in the single
address mode, DM1 and DM0 are ignored when transfers are made from memory-mapped
external devices or external memory to external devices with DACK). DM1 and DM0 are
initialized to 00 by resets or in standby mode.
Bit 15: DM1
Bit 14: DM0
Description
0
0
Fixed destination address (initial value)
0
1
Destination address is incremented (+1 or +2 depending on if
the transfer size is word or byte)
1
0
Destination address is decremented (–1 or –2 depending on if
the transfer size is word or byte)
1
1
Reserved (illegal setting)
•
Bits 13 and 12 (source address mode bits 1, 0 (SM1 and SM0)): SM1 and SM0 select whether
the DMA source address is incremented, decremented, or left fixed (in the single address
mode, SM1 and SM0 are ignored when transfers are made from external devices with DACK
to memory-mapped external devices or external memory). SM1 and SM0 are initialized to 00
by resets or in standby mode.
Bit 13: SM1
Bit 12: SM0
Description
0
0
Fixed source address (initial value)
0
1
Source address is incremented (+1 or +2 depending on if the
transfer size is word or byte)
1
0
Source address is decremented (–1 or –2 depending on if the
transfer size is word or byte)
1
1
Reserved (illegal setting)
•
Bits 11–8 (resource select bits 3–0 (RS3-RS0)): RS3–RS0 specify which transfer requests will
be sent to the DMAC. Do not change the transfer request source unless the DMA enable bit
(DE) is 0. The RS3–RS0 bits are initialized to 0000 by resets or in standby mode.
176 RENESAS
Bit 11:
RS3
Bit 10:
RS2
Bit 9:
RS1
Bit 8:
RS0
Description
0
0
0
0
DREQ (External request*1, dual address mode) (initial value)
0
0
0
1
Reserved (illegal setting)
0
0
1
0
DREQ (External request*1, single address mode* 2)
0
0
1
1
DREQ (External request*1, single address mode* 3)
0
1
0
0
RXI0 (On-chip serial communication interface 0 receive data
full interrupt transfer request)*4
0
1
0
1
TXI0 (On-chip serial communication interface 0 transmit data
empty interrupt transfer request)*4
0
1
1
0
RXI1 (On-chip serial communication interface 1 receive data
full interrupt transfer request)*4
0
1
1
1
TXI1 (On-chip serial communication interface 1 transmit data
empty interrupt transfer request)*4
1
0
0
0
IMIA0 (On-chip ITU0 input capture/compare-match A
interrupt transfer request)*4
1
0
0
1
IMIA1 (On-chip ITU1 input capture/compare-match A
interrupt transfer request)*4
1
0
1
0
IMIA2 (On-chip ITU2 input capture/compare-match A
interrupt transfer request)*4
1
0
1
1
IMIA3 (On-chip ITU3 input capture/compare-match A
interrupt transfer request)*4
1
1
0
0
Auto-request (Transfer requests automatically generated
within DMAC)*4
1
1
0
1
Reserved (illegal setting)
1
1
1
0
Reserved (illegal setting)
1
1
1
1
Reserved (illegal setting)
SCI0, SCI1: Serial communications interface channels 0 and 1.
ITU0–ITU3: Channels 0–3 of the 16-bit integrated-timer pulse unit.
Notes: 1. These bits are valid only in channels 0 and 1. None of these request sources can be
selected in channels 2 and 3.
2. Transfer from memory-mapped external device or external memory to external device
with DACK.
3. Transfer from external device with DACK to memory-mapped external device or
external memory.
4. Dual address mode.
RENESAS 177
•
Bit 7 (acknowledge mode bit (AM)): In the dual address mode, AM selects whether the DACK
signal is output during the data read cycle or write cycle. This bit is valid only in channels 0
and 1. The AM bit is initialized to 0 by resets or in standby mode. The AM bit is not valid in
single address mode.
Bit 7: AM
Description
0
DACK is output in read cycle (initial value)
1
DACK is output in write cycle
•
Bit 6 (acknowledge Level Bit (AL)): AL selects active high signal or active low signal for the
DACK signal. This bit is valid only in channels 0 and 1. The AL bit is initialized to 0 by resets
or in standby mode.
Bit 6: AL
Description
0
DACK is active high (initial value)
1
DACK is active low
•
Bit 5 (DREQ select bit (DS)): DS selects the DREQ input detection method used. This bit is
valid only in channels 0 and 1. The DS bit is initialized to 0 by resets or in standby mode.
Bit 5: DS
Description
0
DREQ detected by low level (initial value)
1
DREQ detected by falling edge
•
Bit 4 (transfer bus mode bit (TM)): TM selects the bus mode for DMA transfers. The TM bit is
initialized to 0 by resets or in standby mode. When the source of the transfer request is an onchip peripheral module, see table 9.4, Selecting On-Chip Peripheral Module Request Modes
with the RS Bit.
Bit 4: TM
Description
0
Cycle-steal mode (initial value)
1
Burst mode
178 RENESAS
•
Bit 3 (transfer size bit (TS)): TS selects the transfer unit size. If the on-chip peripheral module
that is the source or destination of the transfer can only be accessed in bytes, byte must be
selected in this bit. The TS bit is initialized to 0 by resets or in standby mode.
Bit 3: TS
Description
0
Byte (8 bits) (initial value)
1
Word (16 bits)
•
Bit 2 (interrupt enable bit (IE)): IE determines whether or not to request a CPU interrupt at the
end of a DMA transfer. When the IE bit is set to 1, an interrupt (DEI) is requested from the
CPU when the TE bit is set. The IE bit is initialized to 0 by resets or in standby mode.
Bit 2: IE
Description
0
Interrupt request disabled (initial value)
1
Interrupt requeste enabled
•
Bit 1 (transfer end flag bit (TE)): TE indicates that the transfer has ended. When a DMA
transfer ends normally and the value in the DMA transfer count register (TCR) becomes 0, the
TE bit is set to 1. This flag is not set if the transfer ends because of an NMI interrupt or address
error, or because the DE bit or the DME bit of the DMA operation register (DMAOR) was
cleared. To clear the TE bit, read 1 from it and then write 0.
When this flag is set, setting the DE bit to 1 does not enable a DMA transfer. The TE bit is
initialized to 0 by resets or in standby mode.
Bit 1: TE
Description
0
DMA has not ended or was aborted (initial value)
To clear TE, the CPU must read TE after it has been set to 1, then
write a 0 in this bit
1
DMA has ended normally
RENESAS 179
•
Bit 0 (DMA enable bit (DE)): DE enables or disables DMA transfers. In the auto-request
mode, the transfer starts when this bit or the DME bit of the DMAOR is set to 1. The TE bit
and the NMIF and AE bits of the DMAOR must be all cleared to 0. In external request mode
or on-chip peripheral module request mode, the transfer begins when the DMA transfer request
is received from said device or on-chip peripheral module, provided this bit and the DME bit
are set to 1. As with the auto request mode, the TE bit and the NMIF and AE bits of the
DMAOR must be all cleared to 0. The transfer can be stopped by clearing this bit to 0.
The DE bit is initialized to 0 by resets or in standby mode.
Bit 0: DE
Description
0
DMA transfer disabled (initial value)
1
DMA transfer enabled
9.2.5
DMA Operation Register (DMAOR)
The DMA operation register (DMAOR) is a 16-bit read/write register that controls the DMA
transfer mode. It also indicates the DMA transfer status. It is initialized to H'0000 by a reset or the
standby mode.
Bit:
15
14
13
12
11
10
9
8
——
—
—
—
—
—
PR1
PR0
Initial value:
0
0
0
0
0
0
0
0
R/W:
R
R
R
R
R
R
R/W
R/W
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
AE
NMIF
DME
Initial value:
0
0
0
0
0
0
0
0
R/W:
R
R
R
R
R
R/(W)*
R/(W)*
R
Bit name:
Note: Write only 0 to clear the flag.
•
Bits 15–10 (reserved): These bits always read 0. The write value should always be 0.
180 RENESAS
•
Bits 9 and 8 (priority mode bits 1 and 0 (PR1 and PR0)): PR1 and PR0 select the priority level
between channels when there are transfer requests for multiple channels simultaneously.
Bit 9: PR1
Bit 8: PR0
Description
0
0
Fixed priority order (Ch. 0 > Ch. 3 > Ch. 2 > Ch. 1) (initial value)
0
1
Fixed priority order (Ch. 1 > Ch. 3 > Ch. 2 > Ch. 0)
1
0
Round-robin mode priority order (the priority order immediately
after a reset is Ch. 0 > Ch. 3 > Ch. 2 > Ch. 1)
1
1
External-pin round-robin mode priority order (the priority order
immediately after a reset is Ch. 3 > Ch. 2 > Ch. 1 > Ch. 0)
•
•
Bits 7–3 (reserved): These bits always read 0. The write value should always be 0.
Bit 2 (address error flag bit (AE)): AE indicates that an address error occurred in the DMAC.
When this flag is set to 1, the channel cannot be enabled even if the DE bit in the DMA
channel control register (CHCR) and the DME bit are set to 1. To clear the AE bit, read 1 from
it and then write 0. It is initialized to 0 by a reset or the standby mode.
Bit 2: AE
Description
0
No DMAC address error (initial value)
To clear the AE bit, read 1 from it and then write 0.
1
•
Address error by DMAC
Bit 1 (NMI Flag Bit (NMIF)): NMIF indicates that an NMI interrupt occurred. When this flag
is set to 1, the channel cannot be enabled even if the DE bit in the CHCR and the DME bit are
set to 1. To clear the NMIF bit, read 1 from it and then write 0. It is initialized to 0 by a reset or
the standby mode.
Bit 1: NMIF
Description
0
No NMI interrupt (initial value)
To clear the NMIF bit, read 1 from it and then write 0.
1
NMI has occurred
RENESAS 181
•
Bit 0 (DMA master enable bit (DME)): DME enables or disables DMA transfers on all
channels. A channel becomes enabled for a DMA transfer when the DE bit in each DMA's
CHCR and the DME bit are set to 1. For this to be effective, however, the TE bit of each
CHCR and the NMIF and AE bits must all be 0. When the DME bit is cleared, all channel
DMA transfers are aborted.
Bit 0: DME
Description
0
Disable DMA transfers on all channels (initial value)
1
Enable DMA transfers on all channels
9.3
Operation
When there is a DMA transfer request, the DMAC starts the transfer according to the
predetermined channel priority order; when the transfer end conditions are satisfied, it ends the
transfer. Transfers can be requested in three modes: auto-request, external request, and on-chip
module request. Transfer can be in either the single address mode or the dual address mode. The
bus mode can be either burst or cycle steal
9.3.1
DMA Transfer Flow
After the DMA source address registers (SAR), DMA destination address registers (DAR), DMA
transfer count registers (TCR), DMA channel control registers (CHCR), and DMA operation
register (DMAOR) are set, the DMAC transfers data according to the following procedure:
1. Checks to see if transfer is enabled (DE = 1, DME = 1, TE = 0, NMIF = 0, AE = 0)
2. When a transfer request comes and transfer is enabled, the DMAC transfers 1 transfer unit of
data (for an auto-request, the transfer begins automatically when the DE bit and DME bit are
set to 1. The TCR value will be decremented by 1). The actual transfer flows vary by address
mode and bus mode.
3. When the specified number of transfer have been completed (when TCR reaches 0), the
transfer ends normally. If the IE bit of the CHCR is set to 1 at this time, a DEI interrupt is sent
to the CPU.
4. When an address error occurs in the DMAC or an NMI interrupt is generated, the transfer is
aborted. Transfers are also aborted when the DE bit of the CHCR or the DME bit of the
DMAOR are changed to 0.
Figure 9.2 is a flowchart of this procedure.
182 RENESAS
Start
Initial settings
(SAR, DAR, TCR, CHCR, DMAOR)
DE, DME = 1 and
NMIF, AE, TE = 0?
No
Yes
Transfer request
occurs?*1
*2
No
Yes
*3
Transfer (1 transfer unit); TCR–1
→ TCR, SAR and DAR updated
TCR = 0?
Yes
DEI interrupt request
(when IE = 1)
No
Bus mode,
transfer request mode,
DREQ detection selection
system
Does
NMIF = 1, AE = 1,
DE = 0, or DME
= 0?
Yes
No
Transfer aborted
Does
NMIF = 1, AE = 1,
DE = 0, and DME
= 0?
Yes
No
Normal end
Notes: 1.
2.
3.
Transfer ends
In auto-request mode, transfer begins when NMIF, AE and TE are all and the DE
and DME bits are set to 1.
DREQ = level detection in the burst mode (external request), or cycle steal mode.
DREQ = edge detection in the burst mode (external request), or auto request mode
in burst mode.
Figure 9.2 DMA Transfer Flowchart
RENESAS 183
9.3.2
DMA Transfer Requests
DMA transfer requests are basically generated in either the data transfer source or destination, but
they can also be generated by devices and on-chip peripheral modules that are neither the source
nor the destination. Transfers can be requested in three modes: auto-request, external request, and
on-chip module request. The request mode is selected in the RS3–RS0 bits of the DMA channel
control registers 0–3 (CHCR0–CHCR3).
Auto-Request Mode: When there is no transfer request signal from an external source, as in a
memory-to-memory transfer or a transfer between memory and an on-chip peripheral module
unable to request a transfer, the auto-request mode allows the DMAC to automatically generate a
transfer request signal internally. When the DE bits of CHCR0–CHCR3 and the DME bit of the
DMAOR are set to 1, the transfer begins (so long as the TE bits of CHCR0–CHCR3 and the
NMIF and AE bits of DMAOR are all 0).
External Request Mode: In this mode a transfer is performed at the request signal (DREQ) of an
external device. Choose one of the modes shown in table 9.3 according to the application system.
When this mode is selected, if the DMA transfer is enabled (DE = 1, DME = 1, TE = 0, NMIF = 0,
AE = 0), a transfer is performed upon a request at the DREQ input. Choose to detect DREQ by
either the falling edge or low level of the signal input with the DS bit of CHCR0–CHCR3 (DS = 0
is level detection, DS = 1 is edge detection). The source of the transfer request does not have to be
the data transfer source or destination.
Table 9.3
Selecting External Request Modes with the RS Bits
RS3
RS2
RS1
RS0
Address Mode
Source
Destination
0
0
0
0
Dual address
mode
Any*
Any*
0
0
1
0
Single address
mode
External memory or
memory-mapped
external device
External device with
DACK
0
0
1
1
Single address
mode
External device with
DACK
External memory or
memory-mapped
external device
Note: External memory, memory-mapped external device, on-chip memory, on-chip peripheral
module (excluding DMAC)
On-Chip Module Request: In this mode a transfer is performed at the transfer request signal
(interrupt request signal) of an on-chip module. The transfer request signals include the receive
data full interrupt (RXI) of the serial communication interface (SCI), the transmit data empty
interrupt (TXI) of the SCI, the input capture A/compare match A interrupt request (IMIA) of the
16-bit integrated-pulse timer (ITU), (table 9.4). When this mode is selected, if the DMA transfer is
enabled (DE = 1, DME = 1, TE = 0, NMIF = 0, AE = 0), a transfer is performed upon the input of
a transfer request signal. The source of the transfer request does not have to be the data transfer
184 RENESAS
source or destination. When RXI is set as the transfer request, however, the transfer source must
be the SCI’s receive data register (RDR). Likewise, when TXI is set as the transfer request, the
transfer source must be the SCI's transmit data register (TDR).
Table 9.4
Selecting On-Chip Peripheral Module Request Modes with the RS Bit
DMA
Transfer
Request DMA Transfer Request
RS3 RS2 RS1 RS0 Source Signal
Source
Desti- Bus Mode
nation
0
1
0
0
SCI0
receiver
RXI0 (SCI0 receive data full
interrupt transfer request)
RDR0
Any*
Cycle steal
0
1
0
1
SCI0
transmitter
TXI0 (SCI0 transmit data
empty interrupt transfer
request)
Any
TDR0
Cycle steal
0
1
1
0
SCI1
receiver
RXI1 (SCI1 receive data full
interrupt transfer request)
RDR1
Any*
Cycle steal
0
1
1
1
SCI1
transmitter
TXI1 (SCI1 transmit data
empty interrupt transfer
request)
Any*
TDR1
Cycle steal
1
0
0
0
ITU0
IMIA0 (ITU0 input capture A/
compare-match A)
Any*
Any*
Burst/Cycle
steal
1
0
0
1
ITU1
IMIA1 (ITU1 input capture A/
compare-match A)
Any*
Any*
Burst/Cycle
steal
1
0
1
0
ITU2
IMIA2 (ITU2 input capture A/
compare-match A)
Any*
Any*
Burst/Cycle
steal
1
0
1
1
ITU3
IMIA3 (ITU3 input capture A/
compare-match A)
Any*
Any*
Burst/Cycle
steal
SCI0, SCI1: Serial communications interface channels 0 and 1
ITU0–ITU3: Channels 0–3 of the 16-bit integrated-timer pulse unit.
RDR0, RDR1: Receive data registers 0, 1 of SCI
TDR0, TDR1: Transmit data registers 0, 1 of SCI
Note: External memory, memory-mapped external device, on-chip memory, on-chip peripheral
module (excluding DMAC)
When outputting transfer requests from on-chip peripheral modules, the appropriate interrupt
enable bits must be set to output the interrupt signals. Note that transfer request signals from onchip peripheral modules (interrupt request signals) are sent not just to the DMAC but to the CPU
as well. When an on-chip peripheral module is specified as the transfer request source, set the
priority level values in the interrupt priority level registers (IPRC–IPRE) of the interrupt controller
(INTC) at or below the levels set in the I3–I0 bits of the CPU’s status register (SR) so that the
CPU does not acknowledge the interrupt request signal.
RENESAS 185
The DMA transfer request signals of table 9.4 are automatically withdrawn when the
corresponding DMA transfer is performed. If the cycle steal mode is being employed, the DMA
transfer request (interrupt request) will be cleared at the first transfer; if the burst mode is being
used, it will be cleared at the last transfer.
9.3.3
Channel Priority
When the DMAC receives simultaneous transfer requests on two or more channels, it selects a
channel according to a predetermined priority order. The three modes (fixed mode, round-robin
mode, and external-pin round-robin mode) are selected by the priority bits PR1 and PR0 in the
DMA operation register.
Fixed Mode: In these modes, the priority levels among the channels remain fixed. When PR1 and
PR0 bits are set 00, the priority order, high to low, is Ch. 0 > Ch. 3 > Ch. 2 > Ch. 1. When PR1
and PR0 bits are set 01, the priority order, high to low, is Ch. 1 > Ch. 3 > Ch. 2 > Ch. 0.
Round-Robin Mode: Each time one word or byte is transferred on one channel, the priority order
is rotated. The channel on which the transfer was just finished rotates to the bottom of the priority
order. When necessary, the priority order of channels other than the one that just finished the
transfer can also be shifted to keep the relationship between the channels from changing (figure
9.3). The priority order immediately after a reset is channel 0 > channel 3 > channel 2 > channel 1.
186 RENESAS
(1) When channel 0 transfers
Initial priority order
ch0 > ch3 > ch2 > ch1
Channel 0 becomes
bottom priority
Priority order
after transfer
ch3 > ch2 > ch1 > ch0
(2) When channel 3 transfers
Initial priority order
ch0 > ch3 > ch2 > ch1
Priority order
after transfer
ch2 > ch1 > ch0 > ch3
(3) When channel 2 transfers
Initial priority order
Priority order
after transfer
ch0 > ch3 > ch2 > ch1
ch1 > ch0 > ch3 > ch2
Post-transfer priority order when
there is an immediate transfer
request to channel 3 only
ch2 > ch1 > ch0 > ch3
Channel 3 becomes bottom
priority. The priority of channel
0, which was higher than
channel 3, is also shifted.
Channel 2 becomes bottom
priority. The priority of channels
0and 3, which were higher than
channel 2, are also shifted.
If immediately thereafter there
is a request to transfer channel
3 only, channel 3 becomes
bottom priority and the priority
of channels 0 and 1, which
were higher than channel 3,
are also shifted.
(4) When channel 1 transfers
Initial priority order
ch0 > ch3 > ch2 > ch1
Priority order does not change
Priority order
after transfer
ch0 > ch3 > ch2 > ch1
Figure 9.3 Round-Robin Mode
Figure 9.4 shows how the priority order changes when channel 0 and channel 1 transfers are
requested simultaneously and a channel 3 transfer is requested during the channel 0 transfer. The
DMAC operates as follows:
RENESAS 187
1. Transfer requests are generated simultaneously to channels 1 and 0.
2. Channel 0 has a higher priority, so the channel 0 transfer begins first (channel 1 waits for
transfer).
3. A channel 3 transfer request occurs during the channel 0 transfer (channels 1 and 3 are both
waiting)
4. When the channel 0 transfer ends, channel 0 becomes lowest priority.
5. At this point, channel 3 has a higher priority than channel 1, so the channel 3 transfer begins
(channel 1 waits for transfer).
6. When the channel 3 transfer ends, channel 3 becomes lowest priority.
7. The channel 1 transfer begins.
8. When the channel 1 transfer ends, channels 1 and 2 shift downward in priority so that channel
1 becomes the lowest priority.
Transfer request
Waiting channel(s)
(1) Channels 0 and 1
DMAC operation
Channel priority
(2) Channel 0
transfer starts
0>3>2>1
1
(3) Channel 3
1, 3
(4) Channel 0
transfer ends
Priority order
changes
3>2>1>0
(5) Channel 3
transfer starts
1
(6) Channel 3
transfer ends
Priority order
changes
2>1>0>3
(7) Channel 1
transfer starts
None
(8) Channel 1
transfer ends
Priority order
changes
0>3>2>1
Figure 9.4 Changes in Channel Priority in Round-Robin Mode
188 RENESAS
External-Pin Round-Robin Mode: External-pin round-robin mode switches the priority levels of
channel 0 and channel 1, which are the channels that can receive transfer requests from external
pins DREQ0 and DREQ1. The priority levels are changed after each (byte or word) transfer on
channel 0 or channel 1 is completed. The channel which just finished the transfer rotates to the
bottom of the priority order. The priority levels of channels 2 and 3 do not change. The initial
priority order after a reset is channel 3 > channel 2 > channel 1 > channel 0.
Figure 9.5 shows how the priority order changes when channel 0 and channel 1 transfers are
requested simultaneously and a channel 0 transfer is requested again after both channels finish
their transfers. The DMAC operates as follows:
1. Transfer requests are generated simultaneously to channels 1 and 0.
2. Channel 1 has a higher priority, so the channel 1 transfer begins first (channel 0 waits for
transfer).
3. When the channel 1 transfer ends, channel 1 becomes lowest priority.
4. The channel 0 transfer begins.
5. When the channel 0 transfer ends, channel 0 becomes lowest priority.
6. A channel 0 transfer request occurs again.
7. The channel 0 transfer begins.
8. When the channel 0 transfer ends, the priority order does not change, because channel 0 is
already the lowest priority.
RENESAS 189
Transfer request
Waiting channel(s)
(1) Channels
0 and 1
DMAC operation
Channel priority
(2) Channel 1
transfer starts
3>2>1>0
0
(3) Channel 1
transfer ends
Priority order
changes
3>2>0>1
(4) Channel 0
transfer starts
None
(5) Channel 0
transfer ends
(7) Channel 0
transfer starts
(6) Channel 0
None
(8) Channel 0
transfer ends
Priority order
changes
3>2>1>0
Waiting for
transfer request
Priority order
does not change
3>2>1>0
Figure 9.5 Example of Changes in Priority in External-Pin Round-Robin Mode
190 RENESAS
9.3.4
DMA Transfer Types
The DMAC supports the transfers shown in table 9.5. It can operate in the single address mode or
dual address mode, which are defined by how many bus cycles the DMAC takes to access the
transfer source and transfer destination. The actual transfer operation timing varies with the bus
mode. The DMAC has two bus modes: cycle-steal mode and burst mode.
Table 9.5
Supported DMA Transfers
Destination
External
Memory
MemoryMapped
External On-Chip
Device
Memory
On-Chip
Peripheral
Module
External device with DACK Not available
Single
Single
Not available
Not available
External memory
Single
Dual
Dual
Dual
Dual
Memory-mapped external Single
device
Dual
Dual
Dual
Dual
On-chip memory
Not available
Dual
Dual
Dual
Dual
On-chip peripheral module Not available
Dual
Dual
Dual
Dual
Source
External
Device with
DACK
Single: Single address mode
Dual: Dual address mode
RENESAS 191
Address Modes:
•
Single Address Mode
In the single address mode, both the transfer source and destination are external; one
(selectable) is accessed by a DACK signal while the other is accessed by an address. In this
mode, the DMAC performs the DMA transfer in 1 bus cycle by simultaneously outputting a
transfer request acknowledge DACK signal to one external device to access it while outputting
an address to the other end of the transfer. Figure 9.6 shows an example of a transfer between
an external memory and an external device with DACK in which the external device outputs
data to the data bus while that data is written in external memory in the same bus cycle.
External address bus External data bus
SH microcomputer
External
memory
DMAC
Read Write
*
*
(1) (2)
External device
with DACK
DACK
DREQ
: Data flow
Note:
The read/write direction is decided by the RS3-RS0 bits of the CHCRn registers. If RS3RS0=0010, the direction is shown as case 1 (circled number above); if RS3-RS0=0010,
the direction is shown as case 2. Also, DACK output (when writing) indicates case 2.
Figure 9.6 Data Flow in Single Address Mode
Two types of transfers are possible in the single address mode: 1) transfers between external
devices with DACK and memory-mapped external devices, and 2) transfers between external
devices with DACK and external memory. The only transfer requests for either of these is the
external request (DREQ). Figure 9.7 shows the DMA transfer timing for the single address
mode.
192 RENESAS
The DACK output when a transfer occurs from an external device with DACK to a memorymapped external device is the write waveform. The DACK output when a transfer occurs from
a memory-mapped external device to an external device with DACK is the read waveform.
The setting of the acknowledge mode (AM) bits in the channel control registers (CHCR0,
CHCR1) have no effect.
CK
A21–A0
Address output to external memory space
CSn
D15–D0
DACK
WRH
WRL
Data that is output from the external
device with DACK
DACK signal to external devices with
DACK (active low)
WR signal to external memory space
(a) External device with DACK to external memory space
CK
A21–A0
Address output to external memory space
CSn
D15–D0
RD
DACK
Data that is output from external memory space
RD signal to external memory space
DACK signal to external device with DACK
(active low)
(b) External memory space to external device with DACK
Figure 9.7 Example of DMA Transfer Timing in the Single Address Mode
RENESAS 193
•
Dual Address Mode
In the dual address mode, both the transfer source and destination are accessed (selectable) by
an address. The source and destination can be located externally or internally. The source is
accessed in the read cycle and the destination in the write cycle, so the transfer is performed in
two separate bus cycles. The transfer data is temporarily stored in the DMAC. Figure 9.8
shows an example of a transfer between two external memories in which data is read from one
memory in the read cycle and written to the other memory in the following write cycle.
External data bus
SuperH microcomputer
2
DMAC
External
memory
External
memory
1
: Data flow
1: Read cycle
2: Write cycle
Figure 9.8 Data Flow in Dual Address Mode
In the dual address mode transfers, external memory, memory-mapped external devices, onchip memory and on-chip peripheral modules can be mixed without restriction. Specifically,
this enables the following transfer types:
1.
2.
3.
4.
5.
6.
7.
External memory and external memory transfer
External memory and memory-mapped external devices transfer
Memory-mapped external devices and memory-mapped external devices transfer
External memory and on-chip memory transfer
External memory and on-chip peripheral modules (excluding the DMAC) transfer
Memory-mapped external devices and on-chip memory transfer
Memory-mapped external devices and on-chip peripheral modules (excluding the DMAC)
transfer
8. On-chip memory and on-chip memory transfer
9. On-chip memory and on-chip peripheral modules (excluding the DMAC) transfer
10. On-chip peripheral modules (excluding the DMAC) and on-chip peripheral modules
(excluding the DMAC) transfer
194 RENESAS
Transfer requests can be auto requests, external requests, or on-chip peripheral module
requests. When the transfer request source is either the SCI or A/D converter, however, either
the data destination or source must be the SCI or A/D converter (figure 9.4), In dual address
mode, DACK is output in read or write cycles to onchip memory or onchip peripheral
modules. The CHCR controls the cycle of DACK output.
Figure 9.9 shows the DMA transfer timing in the dual address mode.
CK
A21–A0
Source address
Destination address
CSn
D15–D0
RD
WRH
WRL
DACK
Figure 9.9 DMA Transfer Timing in the Dual Address Mode (External memory space to
external memory space transfer with DACK output in the read cycle)
Bus Modes: There are two bus modes: cycle steal and burst. Select the mode in the TM bits of
CHCR0–CHCR3.
•
Cycle-Steal Mode
In the cycle steal mode, the bus right is given to another bus master after a one-transfer-unit
(word or byte) DMA transfer. When another transfer request occurs, the bus rights are obtained
from the other bus master and a transfer is performed for one transfer unit. When that transfer
ends, the bus right is passed to the other bus master. This is repeated until the transfer end
conditions are satisfied.
The cycle steal mode can be used with all categories of transfer destination, transfer source and
transfer request. Figure 9.10 shows an example of DMA transfer timing in the cycle steal
mode. Transfer conditions shown in the figure are:
 Dual address mode
 DREQ level detection
RENESAS 195
DREQ
Bus right returned to CPU
Bus cycle
CPU
CPU
CPU
DMAC DMAC
Read
Write
CPU
DMAC
DMAC
Read
Write
CPU
Figure 9.10 Transfer Example in the Cycle-Steal Mode (Dual address mode, DREQ level
detection)
•
Burst Mode
Once the bus right is obtained, the transfer is performed continuously until the transfer end
condition is satisfied. In the external request mode with low level detection of the DREQ pin,
however, when the DREQ pin is driven high, the bus passes to the other bus master after the
bus cycle of the DMAC that currently has an acknowledged request ends, even if the transfer
end conditions have not been satisfied.
The burst mode cannot be used when the serial communications interface (SCI) is the transfer
request source. Figure 9.11 shows an example of DMA transfer timing in the burst mode. The
transfer conditions shown in the figure are:
 Single address mode
 DREQ level detection
DREQ
Bus cycle
CPU
CPU
CPU
DMAC DMAC DMAC DMAC DMAC DMAC
CPU
Figure 9.11 Transfer Example in the Burst Mode (Single address mode, DREQ level
detection)
Relationship between Request Modes and Bus Modes by DMA Transfer Category: Table 9.6
shows the relationship between request modes and bus modes by DMA transfer category.
196 RENESAS
Table 9.6
Relationship of Request Modes and Bus Modes by DMA Transfer Category
Addres
s Mode Transfer Category
Request
Mode
Bus
Mode
Transfer
Size (bits)
Usable
Channels
Single
External device with DACK and
external memory
External
B/C
8/16
0,1
External device with DACK and
memory-mapped external device
External
B/C
8/16
0, 1
External memory and external
memory
Everything*1
B/C
8/16
0–3* 5
External memory and memorymapped external device
Everything*1
B/C
8/16
0–3* 5
Memory-mapped external device and Everything*1
memory-mapped external device
B/C
8/16
0–3* 5
External memory and on-chip
memory
Everything*1
B/C
8/16
0–3* 5
External memory and on-chip
peripheral module
Everything*2
B/C* 3
8/16*4
0–3* 5
Memory-mapped external device and Everything*1
on-chip memory
B/C
8/16
0–3* 5
Memory-mapped external device and Everything*2
on-chip peripheral module
B/C* 3
8/16*4
0–3* 5
Dual
On-chip memory and on-chip
memory
Everything*1
B/C
8/16
0–3* 5
On-chip memory and on-chip
peripheral module
Everything*2
B/C* 3
8/16*4
0–3* 5
On-chip peripheral module and onchip peripheral module
Everything*2
B/C* 3
8/16*4
0–3* 5
B: Burst, C: Cycle steal
Notes: 1. External requests, auto requests and on-chip peripheral module requests are all
available. For on-chip peripheral module requests, however, SCI cannot be specified as
the transfer request source.
2. External requests, auto requests and on-chip peripheral module requests are all
available. When the SCI is also the transfer request source, however, the transfer
destination or transfer source must be the SCI respectively.
3. If the transfer request source is the SCI, cycle steal only.
4. The access size permitted when the transfer destination or source is an on-chip
peripheral module register.
5. If the transfer request is an external request, channels 0 and 1 only.
Bus Mode and Channel Priority Order: When a given channel (1) is transferring in burst mode
and there is a transfer request to a channel (2) with a higher priority, the transfer of the channel
RENESAS 197
with higher priority (2) will begin immediately. When channel 2 is also operating in the burst
mode, the channel 1 transfer will continue when the channel 2 transfer has completely finished.
When channel 2 is in the cycle steal mode, channel 1 will begin operating again after channel 2
completes the transfer of one transfer unit, but the bus will then switch between the two in the
order channel 1, channel 2, channel 1, channel 2. Since channel 1 is in burst mode, it will not give
the bus to the CPU. This example is illustrated in figure 9.12.
Bus
status
CPU
CPU
DMAC
ch1
DMAC
ch1
DMAC ch1
Burst mode
DMAC
ch2
DMAC
ch1
DMAC
ch2
ch2
ch1
ch2
DMAC ch1 and ch2
Cycle steal mode
DMAC
ch1
DMAC
ch1
DMAC ch1
Burst mode
CPU
CPU
Figure 9.12 Bus Handling when Multiple Channels Are Operating
9.3.5
Number of Bus Cycle States and DREQ Pin Sample Timing
Number of States in Bus Cycle: The number of states in the bus cycle when the DMAC is the
bus master is controlled by the bus state controller just as it is when the CPU is the bus master.
The bus cycle in the dual address mode is controlled by wait state control register 1 (WCR1) while
the single address mode bus cycle is controlled by wait state control register 2 (WCR2). For
details, see section 8.9, Wait State Control.
DREQ Pin Sampling Timing: Normally, when DREQ input is detected immediately prior to the
rise edge of the clock pulse (CK) in external request mode, a DMAC bus cycle will be generated
and the DMA transfer performed two states later at the earliest. The sampling timing after DREQ
input detection differs by bus mode, address mode and method of DREQ input detection.
•
DREQ pin sampling timing in the cycle steal mode
In the cycle steal mode, the sampling timing is the same regardless of whether the DREQ is
detected by edge or level. When edge is being detected, however, once sampled it will not be
sampled again until the next edge detection. Once DREQ input is detected, the next sampling
is not performed until the first state, among those DMAC bus cycles thereby produced, in
which a DACK signal is output (including the detection state itself). The next sampling occurs
immediately prior to the rise edge of the clock pulse(CK) of the third state after the bus cycle
previous to the bus cycle in which the DACK signal is output.
198 RENESAS
Figure 9.13 to 9.22 show the sampling timing of the pin DREQ in the cycle steal mode for
each bus cycle. When no DREQ input is detected at the sampling after the aforementioned
DREQ detection, the next sampling occurs in the next stage in which a DACK signal is output.
If no DREQ input is detected at this time, sampling occurs at every state thereafter.
CK
DREQ
Bus cycle
CPU
CPU
CPU
DMAC
CPU
CPU
CPU
CPU
DACK
Figure 9.13 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Single address mode, bus cycle = 1 state)
CK
DREQ
Bus cycle
CPU
CPU
CPU
DMAC (R) DMAC (W)
CPU
CPU
CPU
DACK
DMAC (R): DMAC read cycle
DMAC (W): DMAC write cycle
Note:
Illustrates the case when DACK is output during the DMAC Read cycle.
Figure 9.14 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Dual address mode, bus cycle = 1 state)
RENESAS 199
CK
DREQ
Bus cycle
CPU
CPU
CPU
DMAC
CPU
CPU
CPU
CPU
DACK
Figure 9.15 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Single address mode, bus cycle = 2 states)
CK
DREQ
Bus cycle
CPU
CPU
CPU
DMAC (R) DMAC (W)
CPU
CPU
CPU
DACK
DMAC (R): DMAC read cycle
DMAC (W): DMAC write cycle
Note:
Illustrates the case when DACK is output during the DMAC write cycle.
Figure 9.16 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Dual address mode, bus cycle = 2 states)
200 RENESAS
T1 Tw T2
T1 Tw T2
CK
DREQ
Bus cycle
CPU
CPU
CPU
DMAC
CPU
DMAC
CPU
DACK
Note:
When DREQ is negated at the third state of the DMAC cycle, the next DMA transfer will
be executed because the sampling is done at the second state of the DMAC cycle.
Figure 9.17 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Single address mode, bus cycle = 2 states + 1 wait state)
T1 Tw T2 T1 Tw T2
CK
DREQ
Bus cycle
CPU
CPU
CPU
DMAC (R)
DMAC (W)
CPU
CPU
DACK
DMAC (R): DMAC read cycle
DMAC (W): DMAC write cycle
Figure 9.18 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Dual address mode, bus cycle = 2 states + 1 wait state)
RENESAS 201
Tp Tr Tc Tc
Tp Tr Tc Tc
CK
DREQ
Bus cycle
CPU
CPU
CPU
DMAC
CPU
DMAC
CPU
DACK
Note:
When DREQ is negated at the fourth state of the DMAC cycle, the next DMA transfer
will be executed because the sampling is done at the second state of the DMAC cycle.
Figure 9.19 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Single address mode, bus cycle = DRAM bus cycle (long
pitch normal mode))
Tp Tr Tc Tc
Tp Tr Tc Tc
CK
DREQ
Bus cycle
CPU CPU CPU
DMAC(R)
DMAC
(W)
CPU
DMAC (R)
DMAC
(W)
CPU
DACK
DMAC (R): DMAC read cycle
DMAC (W): DMAC write cycle
Note:
When DREQ is negated at the fourth state of the DMAC cycle, the next DMA transfer
will be executed because the sampling is done at the second state of the DMAC cycle.
Figure 9.20 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Dual address mode, bus cycle = DRAM bus cycle (long
pitch normal mode))
202 RENESAS
T1 T2 T3 T4
T1 T2 T3 T4
CK
DREQ
Bus cycle
CPU
CPU
CPU
DMAC
CPU
DMAC
CPU
DACK
Note:
When DREQ is negated at the fourth state of the DMAC cycle, the next DMA transfer
will be executed because the sampling is done at the second state of the DMAC cycle.
Figure 9.21 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Single address mode, bus cycle = Address/data multiplex
I/O bus cycle)
T1 T2 T3 T4
T1 T2 T3 T4
CK
DREQ
Bus cycle
CPU CPU CPU
DMAC(R)
DMAC
(W)
CPU
DMAC (R)
DMAC
(W)
CPU
DACK
DMAC (R): DMAC read cycle
DMAC (W): DMAC write cycle
Note: When DREQ is negated at the fourth state of the DMAC cycle, the next DMA
transfer will be executed because the sampling is done at the second state of the
DMAC cycle.
Figure 9.22 DREQ Sampling Timing in Cycle Steal Mode (Output with DREQ level
detection and DACK active low) (Dual address mode, bus cycle = Address/data multiplex
I/O bus cycle)
RENESAS 203
•
DREQ pin sampling timing in the burst mode
In the burst mode, the sampling timing differs depending on whether DREQ is detected by
edge or level.
When DREQ input is being detected by edge, once the falling edge of the DREQ signal is
detected, the DMA transfer continues until the transfer end conditions are satisfied, regardless
of the status of the DREQ pin. No sampling happens during this time. After the transfer ends,
sampling occurs every state until the TE bit of the CHCR is cleared.
When DREQ input is being detected by level, once the DREQ input is detected, next sampling
is performed at the end of every CPU or DMAC bus cycle in the single address mode. In the
dual address mode, the next sampling is performed at the start of every DMAC read cycle. In
both the single address mode and dual address mode, if no DREQ input is detected at this time,
sampling thereafter occurs at every state.
Figures 9.23 and 9.24 show the DREQ pin sampling timing in burst mode when DREQ input is
detected by low level.
CK
DREQ
Bus
cycle
CPU
CPU
CPU
DMAC
DMAC
DMAC
CPU
DACK
Note: Single address DREQ level detection, DACK active low, 1 bus cycle = 2 states.
Figure 9.23 DREQ Pin Sampling Timing in Burst Mode
204 RENESAS
CK
DREQ
Bus
cycle
CPU
CPU
DMAC(R)
DMAC(W)
DMAC(R)
DMAC(W)
CPU
DACK
Note:
Dual address DREQ level detection, DACK active low, DACK output in read cycle, 1 bus
cycle = 2 states.
Figure 9.24 DREQ Pin Sampling Timing in Burst Mode
9.3.6
DMA Transfer Ending Conditions
The DMA transfer ending conditions vary for individual channels ending and all channels ending
together.
Individual Channel Ending Conditions: There are two ending conditions. A transfer ends when
the value of the channel's DMA transfer count register (TCR) is 0, or when the DE bit of the
channel's CHCR is cleared to 0.
•
When TCR is 0: When the TCR value becomes 0 and the corresponding channel’s DMA
transfer ends, the transfer end flag bit (TE) is set in the CHCR. If the IE (interrupt enable) bit
has been set, a DMAC interrupt (DEI) is requested to the CPU.
•
When DE of CHCR is 0: Software can halt a DMA transfer by clearing the DE bit in the
channel’s CHCR. The TE bit is not set when this happens.
Conditions for Ending All Channels Simultaneously: Transfers on all channels end when 1) the
NMIF (NMI flag) bit or AE (address error flag) bit is set to 1 in the DMAOR, or 2) when the
DME bit in the DMAOR is cleared to 0.
RENESAS 205
•
•
Transfers ending when the NMIF or AE bit is set to 1 in DMAOR: When an NMI interrupt or
DMAC address error occurs, the NMIF or AE bit is set to 1 in the DMAOR and all channels
stop their transfers. The SAR, DAR, TCR are all updated by the transfer immediately
preceding the halt. The TE bit is not set. To resume the transfers after NMI interrupt exception
processing or address error exception processing, clear the appropriate flag bit to 0. When a
channel’s DE bit is then set to 1, the transfer on that channel will restart. To avoid restarting a
transfer on a particular channel, keep its DE bit cleared to 0. In the dual address mode, the
DMA transfer will be halted after the completion of the write cycle that follows the initial read
cycle in which the address error occurs. SAR, DAR and TCR are updated by the final transfer.
Transfers ending when DME is cleared to 0 in DMAOR: Clearing the DME bit to 0 in the
DMAOR forcibly aborts the transfers on all channels at the end of the current cycle. The TE
bit is not set.
9.4
Examples of Use
9.4.1
DMA Transfer between On-Chip RAM and a Memory-Mapped External Device
In the following example, data is transferred from an on-chip RAM to a memory-mapped external
device with an input capture A/compare match A interrupt (IMIA0) from channel 0 of the 16-bit
integrated-timer pulse unit (ITU) as the transfer request signal. The transfer is performed by
DMAC channel 3. Table 9.7 shows the transfer conditions and register values.
Table 9.7
Transfer Conditions and Register Settings for Transfer Between On-Chip RAM
and Memory-Mapped External Device
Transfer Conditions
Register
Setting
Transfer source: on-chip RAM
SAR3
H'FFFFE00
Transfer destination: memory-mapped external device
DAR3
Destination address
Number of transfers: 8
TCR3
H'0008
Transfer destination address: fixed
CHCR3
H'1805
DMAOR
H'0001
Transfer source address: incremented
Transfer request source (transfer request signal): ITU channel
0 (IMIA0)
Bus mode: cycle steal
Transfer unit: byte
DEI interrupt request generated at end of transfer (channel 3
enabled for transfer)
Channel priority order: fixed (0 > 3 > 2 > 1) (all channels
transfer enabled)
206 RENESAS
9.4.2
Example of DMA Transfer between On-Chip SCI and External Memory
In this example, receive data of on-chip serial communications interface (SCI) channel 0 is
transferred to external memory using DMAC channel 3. Table 9.8 shows the transfer conditions
and register settings.
Table 9.8
Transfer Conditions and Register Settings for Transfer between On-Chip SCI
and External Memory
Transfer Conditions
Register
Setting
Transfer source: RDR0 of on-chip SCI0
SAR3
H'FFFFEC5
Transfer destination: external memory
DAR3
Destination address
Number of transfers: 64
TCR3
H'0040
Transfer destination address: incremented
CHCR3
H'4405
DMAOR
H'0001
Transfer source address: fixed
Transfer request source (transfer request signal): SCI0 (RXI0)
Bus mode: cycle steal
Transfer unit: byte
DEI interrupt request generated at end of transfer (channel 3
enabled for transfer
Channel priority order: fixed (0 > 3 > 2 > 1) (all channels
transfer enabled)
RENESAS 207
9.5
Cautions
1. All registers other than the DMA operations register (DMAOR) and DMA channel control
registers 0–3 (CHCR0–CHCR3) should be accessed in word or long word units.
2. Before rewriting the RS0–RS3 bits of CHCR0–CHCR3, first clear the DE bit to 0 (when
rewriting CHCR with a byte access, be sure to set the DE bit to 0 in advance).
3. Even when the NMI interrupt is input when the DMAC is not operating, the NMIF bit of the
DMAOR will be set.
4. Interrupt during DMAC Transfer
a. When an NMI interrupt is input, the DMAC stops operation and returns the bus right to the
CPU. The CPU then executes the interrupt processing.
b. When an interrupt other than an NMI occurs.
• When the DMAC is in burst mode.
The DMAC does not return the bus right to the CPU in burst mode. Therefore, even
when an interrupt is requested in DMAC operation, the CPU cannot get the bus right,
causing the interrupt processing not to be executed. When the DMAC completes
transfer and the CPU gets the bus right, the CPU executes the interrupt processing if the
interrupt requested during DMAC transfer is not cleared.*
* Clear conditions for an interrupt request.
 When an interrupt is requested from an on-chip peripheral module, the interrupt factor
flag is cleared.
 When an interrupt is requested by IRQ (edge detection), the CPU begins the IRQ
interrupt processing of the request source.
 When an interrupt is requested by IRQ (level detection), the IRQ interrupt request
signal returned to high level.
• When the DMAC is in cycle-steal mode.
The DMAC returns the bus right to the CPU every when the DMAC completes a
transfer unit in cycle-steal mode. Therefore, the CPU executes the requested interrupt
processing when getting the bus right.
5. The CPU and DMAC leaves the bus right released and the operation of the LSI is stopped
when the following conditions are satisfied.
• The warp bit (WARP) of the bus control register (BCR) of the bus controller (BSC)is set.
• The DMAC is in cycle-steal transfer mode.
• The CPU accesses (reads/writes) the on-chip I/O space.
•
Countermeasure
Set the warp bit of BCR to 0 and set it to normal mode.
208 RENESAS
6. Notes on use of the SLEEP command
a. Operation contents
When the bus cycle of DMAC is entered immediately after executing the SLEEP
command, there are cases when the DMA transfer is carried out correctly.
b. Countermeasure
• Stop the operation (for exemple, clearing of the DMA enable bit (DE) of the DMA
channel control register(CHCRn)) before entering SLEEP.
• When using DMAC during SLEEP, operate DMAC after releasing SLEEP through
interruption.
In cases when the CPU does not carry out any other processing but is waiting for DMAC to end its
transfer during DMAC operation, do not use the SLEEP command, but use the transfer end flag bit
(TE) of the channel DMA control register and the polling software loop.
Phenomenon: If the bus cycle of DMAC is entered immediately after executing the SLEEP
command, the bus cycle of DMAC may conflict with that of CPU.
Address bus
CPU
CPU
Fetch cycle of
SLEEP command
CPU
DMAC
CPU
DMAC
CPU
This is in itself a DMAC cycle but
involves CPU operation.
Accordingly, the bus cycle of DMAC which has conflicted with that of CPU may malfunction.
7. Sampling of DREQ
If DREQ is set to level detection in the DMA cycle steal mode, sampling of DREQ may take
place before DACK is output. Note that some system configurations involve unnecessary
DMA transfers.
•
Operation
As shown in Figure 9.16, sampling of DREQ is carried out immediately before the leading
edge of the third-state clock (CK) after completion of the bus cycle preceding the DMA bus
cycle where DACK is output.
If DACK is output after the third state of the DMA bus cycle, sampling of DREQ must be
carried out before DACK is output.
RENESAS 209
Number of states of
DMAC bus cycle
1
2
3
4
Sampling point
: Bus cycle of DMAC
Figure 9.16 Sampling Points of DREQ
Especially as shown in Figure 9.17, if the bus cycle of DMA is a full access to DRAM or if
refresh demand is generated, sampling of DREQ takes place before DACK is output as
mentioned above. This phenomenon is found when one of the following transfers is made with
DREQ set to the level detection in the DMA cycle steal mode, in a system which employs
DRAM (refresh enabled).
CK
Tp
Tr
Tc
Refresh
T1
T2
DACK
Sampling point
Bus cycle of DRAM
(Full access)
Sampling point
When refresh operation is entered
Sampling point of DREQ for DACK output position
differs with presence/absence of the refresh operation.
Figure 9.17 Example of DREQ Sampling before Output of DACK
• Transfer from a device having DACK to memory in the single address mode (not
restricted to DRAM)
• Transfer from DRAM to a device having DACK in the single address mode
• Output at DACK write in the dual address mode
Output at DACK read in the dual address mode and DMA transfer using DRAM as a
source
•
Countermeasure
To prevent unnecessary DMA transfers, configure the system where DREQ is used for edge
detection and the edge corresponding to the next transfer request occurs after the DACK
output.
210 RENESAS
8. When the following operations are performed in the order shown when the pin to which DREQ
input is assigned is designated as a general input pin by the pin function controller (PFC) and
inputs a low-level signal, the DREQ falling edge is detected, and a DMA transfer request
accepted, immediately after the setting in (b) is performed:
(a) A channel control register (CHCRn) setting is made so that an interrupt is detected at the
falling edge of DREQ.
(b) The function of the pin to which DREQ input is assigned is switched from general input to
DREQ input by a pin function controller (PFC) setting.
Therefore, when switching the pin function from general input pin to DREQ input, the pin
function controller (PFC) setting should be changed to DREQ input while the pin to which
DREQ input is assigned is high.
RENESAS 211
Section 10 16-Bit Integrated-Timer Pulse Unit (ITU)
10.1
Overview
The SuperH microcomputer has an on-chip 16-bit integrated-timer pulse unit (ITU) with five
channels of 16-bit timers.
10.1.1
Features
ITU features are listed below:
•
•
•
•
•
•
•
•
Can process a maximum of twelve different pulse outputs and ten different pulse inputs.
Has ten general registers (GR), two per channel, that can be set to function independently as
output compare or input capture.
Selection of eight counter input clock sources for all channels
 Internal clock: φ, φ/2, φ/4, φ/8,
 External clock: TCLKA, TCLKB, TCLKC, TCLKD
All channels can be set for the following operating modes:
 Compare match waveform output: 0 output/1 output/selectable toggle output (0 output/1
output for channel 2).
 Input capture function: Selectable rising edge, falling edge, or both rising and falling edges.
 Counter clearing function: Counters can be cleared by a compare match or input capture.
 Synchronizing mode: Two or more timer counters (TCNT) can be written to
simultaneously. Two or more timer counters can be simultaneously cleared by a compare
match or input capture. Counter synchronization functions enable synchronized
input/output.
 PWM mode: PWM output can be provided with any duty cycle. When combined with the
counter synchronizing function, enables up to five-phase PWM output.
Channel 2 can be set to the phase counting mode: Two-phase encoder output can be counted
automatically.
Channels 3 and 4 can be set in the following modes:
 Reset-synchronized PWM mode: By combining channels 3 and 4, 3-phase PWM output is
possible with positive and negative waveforms .
 Complementary PWM mode: By combining channels 3 and 4, 3-phase PWM output is
possible with non-overlapping positive and negative waveforms.
Buffer operation: Input capture registers can be double-buffered. Output compare registers can
be updated automatically.
High-speed access via internal 16-bit bus: The TCNT, GR, and buffer register (BR) 16-bit
registers can be accessed at high speed via a 16-bit bus.
RENESAS 213
•
•
•
Fifteen interrupt sources: Ten compare match/input capture interrupts (2 sources per channel)
and five overflow interrupts are vectored independently for a total of 15 sources.
Can activate DMAC: The compare match/input capture interrupts of channels 0–3 can start the
DMAC (one for each of four channels).
Output trigger can be generated for the programmable timing pattern controller (TPC): The
compare match/input capture signals of channel 0–3 can be used as output triggers for the
TPC.
Table 10.1 summarizes the ITU functions.
214 RENESAS
Table 10.1 ITU Functions
Item
Channel 0
Channel 1
Channel 2
Channel 3
Channel 4
Counter clocks
Internal: φ, φ/2, φ/4, φ/8
External: Independently selectable from TCLKA, TCLKB, TCLKC, and TCLKD
General registers
(output compare/
input capture dual
registers)
GRA0, GRB0
GRA1, GRB1
GRA2, GRB2 GRA3, GRB3
GRA4, GRB4
Buffer registers
No
No
No
BRA3, BRB3
BRA4, BRB4
Input/output pins
TIOCA0,
TIOCB0
TIOCA1,
TIOCB1
TIOCA2,
TIOCB2
TIOCA3,
TIOCB3
TIOCA4,
TIOCB4
Output pins
No
No
No
No
TOCXA4,
TOCXB4
Counter clear func- GRA0/GRB0
tion (compare match or input capture)
GRA1/GRB1
GRA2/GRB2
GRA3/GRB3
GRA4/GRB4
Compare 0
Yes
match
1
Yes
output
Toggle Yes
output
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Input capture
function
Yes
Yes
Yes
Yes
Yes
Synchronization
Yes
Yes
Yes
Yes
Yes
PWM mode
Yes
Yes
Yes
Yes
Yes
Reset-synchronized No
PWM mode
No
No
Yes
Yes
Complementary
PWM mode
No
No
No
Yes
Yes
Phase counting
mode
No
No
Yes
No
No
Buffer operation
No
No
No
Yes
Yes
DMAC activation
GRA0 comGRA1 comGRA2 comGRA3 comNo
pare match or pare match or pare match or pare match or
input capture input capture input capture input capture
Interrupt sources
(three)
• Compare
• Compare
• Compare
• Compare
• Compare
match/input
match/input
match/input
match/input
match/input
capture A0
capture A1
capture A2
capture A3
capture A4
• Compare
• Compare
• Compare
• Compare
• Compare
match/input
match/input
match/input
match/input
match/input
capture B0
capture B1
capture B2
capture B3
capture B4
• Overflow
• Overflow
• Overflow
• Overflow
• Overflow
RENESAS 215
10.1.2
Block Diagram
ITU Block Diagram (Complete): Figure 10.1 is the block diagram of the ITU.
Control
logic
IMIA0–IMIA4
IMIB0–IMIB4
OVI0–OVI4
TOCR
TSTR
TSNC
TMDR
TFCR
Module data bus
TOCR: Timer output control register (8 bits)
TSTR: Timer start regsiter (8 bits)
TSNC: Timer synchronization register (8 bits)
TMDR: Timer mode register (8 bits)
TFCR: Timer function control register (8 bits)
Figure 10.1 ITU Block Diagram
216 RENESAS
Bus interface
16-bit timer channel 0
16-bit timer channel 4
16-bit timer channel 1
Counter control and
pulse I/O control unit
TOCXA4, TOCXB4
TIOCA0–TIOCA4
TIOCB0–TIOCB4
16-bit timer channel 2
φ, φ/2, φ/4, φ/8
Clock
selection
16-bit timer channel 3
TCLKA–TCLKD
Internal
data
bus
Block Diagram of Channels 0 and 1: ITU channels 0 and 1 have the same function. Figure 10.2
is a block diagram of channels 0 and 1.
Clock selection
IMIAn
IMIBn
OVIn
TSRn
Control logic
TCRn
GRBn
GRAn
TCNTn
Comparator
TIERn
φ, φ/2,
φ/4, φ/8
TIOCAn
TIOCBn
TIORn
TCLKA–
TCLKD
Module data bus
TCNTn: Timer counter n (16 bits)
GRAn, GRBn: General registers An, Bn (input capture/output compare dual use) (16 bits × 2)
TCRn: Timer control register n (8 bits)
TIORn: Timer I/O control register n (8 bits)
TIERn: Timer interrupt enable register n (8 bits)
TSRn: Timer status register n (8 bits) (n = 0 or 1)
Figure 10.2 Channels 0 and 1 Block Diagram (One Channel Shown)
RENESAS 217
Block Diagram of Channel 2: Figure 10.3 is a block diagram of channel 2. Channel 2 is 0
output/1 output only.
Clock selection
IMIA2
IMIB2
OVI2
TSR2
Control logic
TCR2
GRB2
GRA2
TCNT2
Comparator
TIER2
φ, φ/2,
φ/4, φ/8
TIOCA2
TIOCB2
TIOR2
TCLKA–
TCLKD
Module data bus
TCNT2: Timer counter 2 (16 bits)
GRA2, GRB2: General registers A2, B2 (input capture/output compare dual use) (16 bits × 2)
TCR2: Timer control register 2 (8 bits)
TIOR2: Timer I/O control register 2 (8 bits)
TIER2: Timer interrupt enable register 2 (8 bits)
TSR2: Timer status register 2 (8 bits)
Figure 10.3 Channel 2 Block Diagram
218 RENESAS
Block Diagrams of Channels 3 and 4: Figure 10.4 is a block diagram of channel 3; figure 10.5 is
a block diagram of channel 4.
TCLKA–
TCLKD
Clock selection
TIOCA3
TIOCB3
IMIA3
IMIB3
OVI3
TSR3
TIER3
TIOR3
GRB3
Control logic
BRB3
GRA3
BRA3
TCNT3
Comparator
TCR3
φ, φ/2,
φ/4, φ/8
Module data bus
TCNT3: Timer counter 3 (16 bits)
GRA3, GRB3: General registers A3, B3 (input capture/output compare dual use) (16 bits × 2)
BRA3, BRB3: Buffer registers A3, B3 (input capture/output compare dual use) (16 bits × 2)
TCR3: Timer control register 3 (8 bits)
TIOR3: Timer I/O control register 3 (8 bits)
TIER3: Timer interrupt enable register 3 (8 bits)
TSR3: Timer status register 3 (8 bits)
Figure 10.4 Channels 3 Block Diagram
RENESAS 219
TCLKA–
TCLKD
TOCXA4
TOCXB4
Clock selection
TIOCA4
TIOCB4
IMIA4
IMIB4
OVI4
TSR4
TIER4
TIOR4
GRB4
Control logic
BRB4
GRA4
BRA4
GCNT4
Comparator
TCR4
φ, φ/2,
φ/4, φ/8
Module data bus
TCNT4: Timer counter 4 (16 bits)
GRA4, GRB4: General registers A4, B4 (input capture/output compare dual use) (16 bits × 2)
BRA4, BRB4: Buffer registers A4, B4 (input capture/output compare dual use) (16 bits × 2)
TCR4: Timer control register 4 (8 bits)
TIOR4: Timer I/O control register 4 (8 bits)
TIER4: Timer interrupt enable register 4 (8 bits)
TSR4: Timer status register 4 (8 bits)
Figure 10.5 Channel 4 Block Diagram
220 RENESAS
10.1.3
Input/Output Pins
Table 10.2 summarizes the ITU pins. External pin functions should be set with the pin function
controller to match to the ITU setting. See section 15, Pin Function Controller, for details. ITU
pins need to be set using the pin function controller (PFC) after the LSI is set to the ITU mode.
Table 10.2 Pin Configuration
Channel Name
Pin Name
I/O Function
Shared
Clock input A
TCLKA
I
External clock A input pin (A-phase input pin in
phase counting mode)
Clock input B
TCLKB
I
External clock B input pin (B-phase input pin in
phase counting mode)
Clock input C
TCLKC
I
External clock C input pin
Clock input D
TCLKD
I
External clock D input pin
Input capture/output compare A0
TIOCA0
I/O GRA0 output compare/GRA0 input capture/PWM
output pin (in PWM mode)
Input capture/output compare B0
TIOCB0
I/O GRB0 output compare/GRB0 input capture
Input capture/output compare A1
TIOCA1
I/O GRA1 output compare/GRA1 input capture/PWM
output pin (in PWM mode)
Input capture/output compare B1
TIOCB1
I/O GRB1 output compare/GRB1 input capture
Input capture/output compare A2
TIOCA2
I/O GRA2 output compare/GRA2 input capture/PWM
output pin (in PWM mode)
Input capture/output compare B2
TIOCB2
I/O GRB2 output compare/GRB2 input capture
Input capture/output compare A3
TIOCA3
I/O GRA3 output compare/GRA3 input capture/PWM
output pin (in PWM mode, complementary PWM
mode, or reset-synchronized PWM mode)
Input capture/output compare B3
TIOCB3
I/O GRB3 output compare/GRB3 input capture/PWM
output pin (in complementary PWM mode or resetsynchronized PWM mode)
Input capture/output compare A4
TIOCA4
I/O GRA4 output compare/GRA4 input capture/PWM
output pin (in PWM mode, complementary PWM
mode or reset-synchronized PWM mode)
Input capture/output compare B4
TIOCB4
I/O GRB4 output compare/GRB4 input capture/PWM
output pin (in complementary PWM mode or resetsynchronized PWM mode)
Output compare
XA4
TOCXA4
I/O PWM output pin (in complementary PWM mode or
reset-synchronized PWM mode)
Output compare
XB4
TOCXB4
I/O PWM output pin (in complementary PWM mode or
reset-synchronized PWM mode)
0
1
2
3
4
RENESAS 221
10.1.4
Register Configuration
Table 10.3 summarizes the ITU register configuration.
Table 10.3 Register Configuration
Access
Size
Channel Name
Abbreviation
R/W
Initial
Value
Shared
Timer start register
TSTR
R/W
H'E0/H'60 H'5FFFF00
8
Timer synchro register
TSNC
R/W
H'E0/H'60 H'5FFFF01
8
Timer mode register
TMDR
R/W
H'80/H'00
H'5FFFF02
8
Timer function control register TFCR
R/W
H'C0/H'40 H'5FFFF03
8
Timer output control register
TOCR
R/W
H'FF/H'7F H'5FFFF31
8
Timer control register 0
TCR0
R/W
H'80/H'00
H'5FFFF04
8
Timer I/O control register 0
TIOR0
R/W
H'88/H'08
H'5FFFF05
8
Timer interrupt enable
register 0
TIER0
R/W
H'F8/H'78 H'5FFFF06
8
Timer status register 0
TSR0
R/(W)*2
H'F8/H'78 H'5FFFF07
8
Timer counter 0
TCNT0
R/W
H'00
H'5FFFF08
8, 16, 32
H'5FFFF09
8, 16, 32
0
General register A0
GRA0
R/W
H'FF
Address* 1
H'5FFFF0A 8, 16, 32
H'5FFFF0B 8, 16, 32
General register B0
GRB0
R/W
H'FF
H'5FFFF0C 8, 16
H'5FFFF0D 8, 16
1
Timer control register 1
TCR1
R/W
H'80/H'00
H'5FFFF0E 8
Timer I/O control register 1
TIOR1
R/W
H'88/H'08
H'5FFFF0F
8
Timer interrupt enable
register 1
TIER1
R/W
H'F8/H'78 H'5FFFF10
8
Timer status register 1
TSR1
R/(W)*2
H'F8/H'78 H'5FFFF11
8
Timer counter 1
TCNT1
R/W
H'00
H'5FFFF12
8, 16
H'5FFFF13
8, 16
H'5FFFF14
8, 16, 32
H'5FFFF15
8, 16, 32
H'5FFFF16
8, 16, 32
H'5FFFF17
8, 16, 32
General register A1
General register B1
222 RENESAS
GRA1
GRB1
R/W
R/W
H'FF
H'FF
Table 10.3 Register Configuration (cont)
Channel Name
Abbreviation
R/W
Initial
Value
Address* 1
Access
Size
2
Timer control register 2
TCR2
R/W
H'80/H'00
H'5FFFF18
8
Timer I/O control register 2
TIOR2
R/W
H'88/H'08
H'5FFFF19
8
Timer interrupt enable register TIER2
2
R/W
H'F8/H'78 H'5FFFF1A 8
Timer status register 2
TSR2
R/(W)*2
H'F8/H'78 H'5FFFF1B 8
Timer counter 2
TCNT2
R/W
H'00
H'5FFFF1C 8, 16, 32
H'5FFFF1D 8, 16, 32
General register A2
General register B2
3
GRA2
GRB2
R/W
R/W
H'FF
H'FF
H'5FFFF1E 8, 16, 32
H'5FFFF1F
8, 16, 32
H'5FFFF20
8, 16
H'5FFFF21
8, 16
Timer control register 3
TCR3
R/W
H'80/H'00
H'5FFFF22
8
Timer I/O control register 3
TIOR3
R/W
H'88/H'08
H'5FFFF23
8
Timer interrupt enable register TIER3
3
R/W
H'F8/H'78 H'5FFFF24
8
Timer status register 3
TSR3
R/(W)*2
H'F8/H'78 H'5FFFF25
8
Timer counter 3
TCNT3
R/W
H'00
H'5FFFF26
8, 16
H'5FFFF27
8, 16
H'5FFFF28
8, 16, 32
H'5FFFF29
8, 16, 32
General register A3
General register B3
GRA3
GRB3
R/W
R/W
H'FF
H'FF
H'5FFFF2A 8, 16, 32
H'5FFFF2B 8, 16, 32
Buffer register A3
BRA3
R/W
H'FF
H'5FFFF2C 8, 16, 32
H'5FFFF2D 8, 16, 32
Buffer register B3
4
BRB3
R/W
H'FF
H'5FFFF2E 8, 16, 32
H'5FFFF2F
8, 16, 32
Timer control register 4
TCR4
R/W
H'80/H'00
H'5FFFF32
8
Timer I/O control register 4
TIOR4
R/W
H'88/H'08
H'5FFFF33
8
Timer interrupt enable register TIER4
4
R/W
H'F8/H'78 H'5FFFF34
8
Timer status register 4
R/(W)*2
H'F8/H'78 H'5FFFF35
8
TSR4
RENESAS 223
Table 10.3 Register Configuration (cont)
Channel Name
Abbreviation
R/W
Initial
Value
Address* 1
Access
Size
4 (cont)
TCNT4H
H'00
H'5FFFF36
8, 16
H'5FFFF37
8, 16
H'5FFFF38
8, 16, 32
H'5FFFF39
8, 16, 32
Timer counter 4
General register A4
GRA4H
General register B4
GRB4H
R/W
R/W
R/W
H'FF
H'FF
H'5FFFF3A 8, 16, 32
H'5FFFF3B 8, 16, 32
Buffer register A4
BRA4H
R/W
H'FF
H'5FFFF3C 8, 16, 32
H'5FFFF3D 8, 16, 32
Buffer register B4
BRB4H
R/W
H'FF
H'5FFFF3E 8, 16, 32
H'5FFFF3F
8, 16, 32
Notes: 1. Only the values of bits A27–A24 and A8–A0 are valid; bits A23–A9 are ignored. For
details on the register addresses, see section 8.3.5, Description of Areas.
2. Write 0 to clear flags.
10.2
ITU Register Descriptions
10.2.1
Timer Start Register (TSTR)
The timer start register (TSTR) is an eight-bit read/write register that starts and stops the timer
counters (TCNT) of channels 0–4. TSTR is initialized to H'E0 or H'60 upon reset or standby
mode.
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
STR4
STR3
STR2
STR1
STR0
Initial value:
*
1
1
0
0
0
0
0
R/W:
—
—
—
R/W
R/W
R/W
R/W
R/W
Note: Undefined
•
Bits 7–5 (reserved): Cannot be modified. Bit 7 is read as undefined. Bits 6 and 5 are always
read as 1. The write value to bit 7 should be 0 or 1, and the write value to bits 6 and 5 should
always be 1.
224 RENESAS
•
Bit 4 (count start 4 (STR4)): STR4 starts and stops TCNT4.
Bit 4: STR4
Description
0
TCNT4 is halted (initial value)
1
TCNT4 is counting
•
Bit 3 (count start 3 (STR3)): STR3 starts and stops TCNT3.
Bit 3: STR3
Description
0
TCNT3 is halted (initial value)
1
TCNT3 is counting
•
Bit 2 (count start 2 (STR2)): STR2 starts and stops TCNT2.
Bit 2: STR2
Description
0
TCNT2 is halted (initial value)
1
TCNT2 is counting
•
Bit 1 (count start 1 (STR1)): STR1 starts and stops TCNT1.
Bit 1: STR1
Description
0
TCNT1 is halted (initial value)
1
TCNT1 is counting
•
Bit 0 (count start 0 (STR0)): STR0 starts and stops TCNT0.
Bit 0: STR0
Description
0
TCNT0 is halted (initial value)
1
TCNT0 is counting
RENESAS 225
10.2.2
Timer Synchro Register (TSNC)
The timer synchro register (TSNC) is an eight-bit read/write register that selects timer
synchronizing modes for channels 0–4. Channels for which 1 is set to the corresponding bit will be
synchronized. TSNC is initialized to H'E0 or H'60 upon reset or standby mode.
Bit:
7
6
5
4
3
Bit name:
—
—
—
Initial value:
*
1
1
0
0
0
0
0
R/W:
—
—
—
R/W
R/W
R/W
R/W
R/W
SYNC4 SYNC3
2
SYNC2
1
0
SYNC1 SYNC0
Note: Undefined
•
Bits 7–5 (reserved): Bit 7 is read as undefined. Bits 6 and 5 are always read as 1. The write
value to bit 7 should be 0 or 1, and the write value to bits 6 and 5 should always be 1.
•
Bit 4 (timer synchro 4 (SYNC4)): SYNC4 selects the synchronizing mode for channel 4.
Bit 4: SYNC4
Description
0
The timer counter for channel 4 (TCNT4) operates independently
(Preset/clear of TCNT4 is independent of other channels) (initial value)
1
Channel 4 operates synchronously. Synchronized preset/clear of
TNCT4 enabled.
•
Bit 3 (timer Synchro 3 (SYNC3)): SYNC3 selects the synchronizing mode for channel 3.
Bit 3: SYNC3
Description
0
The timer counter for channel 3 (TCNT3) operates independently
(Preset/clear of TCNT3 is independent of other channels) (initial value)
1
Channel 3 operates synchronously. Synchronized preset/clear of
TNCT3 enabled.
•
Bit 2 (timer synchro 2 (SYNC2)): SYNC2 selects the synchronizing mode for channel 2.
Bit 2: SYNC2
Description
0
The timer counter for channel 2 (TCNT2) operates independently
(Preset/clear of TCNT2 is independent of other channels) (initial value)
1
Channel 2 operates synchronously. Synchronized preset/clear of
TNCT2 enabled.
226 RENESAS
•
Bit 1 (timer synchro 1 (SYNC1)): SYNC1 selects the synchronizing mode for channel 1.
Bit 1: SYNC1
Description
0
The timer counter for channel 1 (TCNT1) operates independently
(Preset/clear of TCNT1 is independent of other channels) (initial value)
1
Channel 1 operates synchronously. Synchronized preset/clear of
TNCT1 enabled.
•
Bit 0 (timer synchro 0 (SYNC0)): SYNC0 selects the synchronizing mode for channel 0.
Bit 0: SYNC0
Description
0
The timer counter for channel 0 (TCNT0) operates independently
(Preset/clear of TCNT0 is independent of other channels) (initial value)
1
Channel 0 operates synchronously. Synchronized preset/clear of
TNCT0 enabled.
10.2.3
Timer Mode Register (TMDR)
The timer mode register (TMDR) is an eight-bit read/write register that selects the PWM mode for
channels 0–4, sets the phase counting mode for channel 2, and sets the conditions for the overflow
flag (OVF). TMDR is initialized to H'80 or H'00 by a reset or the standby mode.
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
MDF
FDIR
PWM4
PWM3
PWM2
PWM1
PWM0
Initial value:
*
0
0
0
0
0
0
0
R/W:
—
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Note: Undefined
•
Bit 7 (reserved): Bit 7 is read as undefined. The write value should be 0 or 1.
•
Bit 6 (phase counting mode (MDF)): MDF selects the phase counting mode for channel 2.
Bit 6: MDF
Description
0
Channel 2 operates normally (initial value)
1
Channel 2 operates in phase counting mode
RENESAS 227
When the MDF is set to 1 to select the phase counting mode, the timer counter (TCNT2)
becomes an up/down counter and the TCLKA and TCLKB pins become count clock input
pins. TCNT2 counts on both the rising and falling edges of TCLKA and TCLKB, with the
increment/decrement chosen as follows:
Count
Direction
Decrement
Increment
TCLKA pin
Rising
High
Falling
Low
Rising
High
Falling
Low
TCLKB pin
L
Rising
High
Falling
High
Falling
Low
Rising
In the phase counting mode, selections for external clock edge made in the CKEG1 and
CKEG0 bits of the timer control register 2 (TCR2) and the selection for counter clock made in
the TPSC2 –TPSC0 bits are ignored. The phase counting mode described above takes priority.
Settings for counter clear conditions in the CCLR1 and CCLR0 bits of TCR2 and settings for
timer I/O control register 2 (TIOR2), timer interrupt enable register (TIER2) and timer status
register 2 (TSR2) compare match/input capture functions and interrupts, however, are valid
even in the phase counting mode.
•
Bit 5 (flag direction (FDIR)): FDIR selects the setting condition for the overflow flag (OVF) in
timer status register 2 (TSR2). This bit is valid no matter which mode channel 2 is operating in.
Bit 5: FDIR
Description
0
OVF of TSR2 is set to 1 when TCNT2 overflows or underflows (initial
value)
1
OVF of TSR2 is set to 1 when TCNT2 overflows
•
Bit 4 (PWM Mode 4 (PWM4)): PWM4 selects the PWM mode for channel 4. When the
PWM4 bit is set to 1 and the PWM mode entered, the TIOCA4 pin becomes a PWM output
pin. 1 is output on a compare match of general register A4 (GRA4); 0 is output on a compare
match of general register B4 (GRB4). When the complementary PWM mode or resetsynchronized PWM mode are set by the CMD1 and CMD0 bits of the timer function control
register (TFCR), the setting of this bit is ignored in favor of the settings of CMD1 and CMD0.
Bit 4: PWM4
Description
0
Channel 4 operates normally (initial value)
1
Channel 4 operates in PWM mode
228 RENESAS
•
Bit 3 (PWM Mode 3 (PWM3)): PWM3 selects the PWM mode for channel 3. When the
PWM3 bit is set to 1 and the PWM mode entered, the TIOCA3 pin becomes a PWM output
pin. 1 is output on a compare match of general register A3 (GRA3); 0 is output on a compare
match of general register B3 (GRB3). When the complementary PWM mode or resetsynchronized PWM mode are set by the CMD1 and CMD0 bits of the timer function control
register (TFCR), the setting of this bit is ignored in favor of the settings of CMD1 and CMD0.
Bit 3: PWM3
Description
0
Channel 3 operates normally (initial value)
1
Channel 3 operates in PWM mode
•
Bit 2 (PWM Mode 2 (PWM2)): PWM2 selects the PWM mode for channel 2. When the
PWM2 bit is set to 1 and the PWM mode entered, the TIOCA2 pin becomes a PWM output
pin. 1 is output on a compare match of general register A2 (GRA2); 0 is output on a compare
match of general register B2 (GRB2).
Bit 2: PWM2
Description
0
Channel 2 operates normally (initial value)
1
Channel 2 operates in PWM mode
•
Bit 1 (PWM Mode 1 (PWM1)): PWM1 selects the PWM mode for channel 1. When the
PWM1 bit is set to 1 and the PWM mode entered, the TIOCA1 pin becomes a PWM output
pin. 1 is output on a compare match of general register A1 (GRA1); 0 is output on a compare
match of general register B1 (GRB1).
Bit 1: PWM1
Description
0
Channel 1 operates normally (initial value)
1
Channel 1 operates in PWM mode
•
Bit 0 (PWM Mode 0 (PWM0)): PWM0 selects the PWM mode for channel 0. When the
PWM0 bit is set to 1 and the PWM mode entered, the TIOCA0 pin becomes a PWM output
pin. 1 is output on a compare match of general register A0 (GRA0); 0 is output on a compare
match of general register B0 (GRB0).
Bit 0: PWM0
Description
0
Channel 0 operates normally (initial value)
1
Channel 0 operates in PWM mode
RENESAS 229
10.2.4
Timer Function Control Register (TFCR)
The timer function control register (TFCR) is an 8-bit read/write register that selects
complementary PWM/reset-synchronized PWM for channels 3 and 4 and sets the buffer operation.
TFCR is initialized on a reset or standby mode to H'C0 or H'40.
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
CMD1
CMD0
BFB4
BFA4
BFB3
BFA3
Initial value:
*
1
0
0
0
0
0
0
R/W:
—
—
R/W
R/W
R/W
R/W
R/W
R/W
Note: Undefined
•
Bits 7 and 6 (reserved): Bit 7 is read as undefined. Bit 6 is always read as 1. The write value to
bit 7 should be 0 or 1. The write value to bit 6 should always be 1.
•
Bits 5 and 4 (combination mode 1 and 0 (CMD1 and CMD0)): CMD1 and CMD0 select the
complementary PWM mode or reset-synchronized mode for channels 3 and 4. Set the
complementary PWM/reset-synchronized PWM mode while the timer counter (TCNT) being
used is off. When these bits are used to set the complementary PWM/reset-synchronized PWM
mode, they take priority over the PWM4 and PWM3 bits of the TMDR. While the
complementary PWM/reset-synchronized PWM mode settings and the SYNC4 and SYNC3 bit
settings of the timer synchro register (TSNC) are valid simultaneously, when the
complementary PWM mode is set, channels 3 and 4 should not be set to operate
simultaneously (SYNC 4 and SYNC 3 bits of TSNC should not both be set to 1).
Bit 5: CMD1
Bit 4: CMD0
Description
0
0
Channels 3 and 4 operate normally (initial value)
1
Channels 3 and 4 operate normally
0
Channels 3 and 4 operate together in complementary PWM
mode
1
Channels 3 and 4 operate together in reset-synchronized
PWM mode
1
•
Bit 3 (buffer mode B4 (BFB4)): BFB4 selects the buffer mode for GRB4 and BRB4 in channel
4.
Bit 3: BFB4
Description
0
GRB4 operates normally in channel 4 (initial value)
1
GRB4 and BRB4 operate in buffer mode in channel 4
230 RENESAS
•
Bit 2 (buffer mode A4 (BFA4)): BFA4 selects the buffer mode for GRA4 and BRA4 in
channel 4.
Bit 2: BFA4
Description
0
GRA4 operates normally in channel 4 (initial value)
1
GRA4 and BRA4 operate in buffer mode in channel 4
•
Bit 1 (buffer mode B3 (BFB3)): BFB3 selects the buffer mode for GRB3 and BRB3 in channel
3.
Bit 1: BFB3
Description
0
GRB3 operates normally in channel 3 (initial value)
1
GRB3 and BRB3 operate in buffer mode in channel 3
•
Bit 0 (buffer Mode A3 (BFA3)): BFA3 selects the buffer mode for GRA3 and BRA3 in
channel 3.
Bit 0: BFA3
Description
0
GRA3 operates normally in channel 3 (initial value)
1
GRA3 and BRA3 operate in buffer mode in channel 3
10.2.5
Timer Output Control Register (TOCR)
The timer output control register (TOCR) is an eight-bit read/write register that inverts the output
level of the complementary PWM mode/reset-synchronized PWM mode. Setting bits OLS3 and
OLS4 is valid in only the complementary PWM mode and reset-synchronized PWM mode. In
other output situations, these bits are ignored. The TOCR is initialized to H'FF or H'7F by a reset
or in the standby mode.
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
—
OLS4
OLS3
Initial value:
*
1
1
1
1
1
1
1
R/W:
—
—
—
—
—
—
R/W
R/W
Note: Undefined
•
Bits 7–2 (reserved): Bit 7 is read as undefined. Bits 6–2 are always read as 1. The write value
to bit 7 should be 0 or 1. The write value to bits 6–2 should always be 1.
RENESAS 231
•
Bit 1 (output level select 4 (OLS4)): OLS4 selects the output level of the complementary PWM
mode or reset-synchronized PWM mode.
Bit 1: OLS4
Description
0
TIOCA3, TIOCA4, and TIOCB4 are inverted and output
1
TIOCA3, TIOCA4, and TIOCB4 are output directly (initial value)
•
Bit 0 (output level select 3 (OLS3)): OLS3 selects the output level of the complementary PWM
mode or reset-synchronized PWM mode.
Bit 0: OLS3
Description
0
TIOCB3, TOCXA4, and TOCXB4 are inverted and output
1
TIOCB3, TOCXA4, and TOCXB4 are output directly (initial value)
10.2.6
Timer Counters (TCNT)
The ITU has five 16-bit timer counters (TCNT), one for each channel (table 10.4).
Each TCNT is a 16-bit read/write counter that counts by input from a clock source. The clock
source is selected by timer prescalar bits 2–0 (TPSC2–TPSC0) in the timer control register (TCR).
TCNT0 and TCNT 1 are strictly upcounters. Up/down counting occurs for TCNT2 when the phase
counting mode is selected, or for TCNT3 and TCNT 4 when complementary PWM mode is
selected. In other modes, they are upcounters.
The TCNT can be cleared to H'0000 by compare match with the corresponding general register A
or B (GRA, GRB) or input capture to GRA or GRB (counter clear function).
When the TCNT overflows (changes from H'FFFF–H'0000), the overflow flag (OVF) in the timer
status register (TSR) is set to 1. The OVF of the corresponding channel TSR is also set to 1 when
the TCNT underflows (changes from H'0000–H'FFFF).
The TCNT is connected to the CPU by a 16-bit bus, so it can be written or read by either word
access or byte access. The TCNT is initialized to H'0000 by a reset or in standby mode.
232 RENESAS
Table 10.4 Timer Counters (TCNT)
Channel
Abbreviation
Function
0
TCNT0
Increment counter
1
TCNT1
2
TCNT2
Phase counting mode: Increment/decrement
All others: Increment
3
TCNT3
4
TCNT4
Complementary PWM mode: Increment/decrement
All others: Increment
Bit:
15
14
13
12
11
10
9
8
Bit name:
Initial value:
R/W:
Bit:
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
10.2.7
General Registers A and B (GRA and GRB)
Each of the five ITU channels has two 16-bit general registers (GR) for a total of ten registers
(table 10.5).
Each GR is a 16-bit read/write register that can function as either an output compare register or an
input capture register. The function is selected by settings in the timer I/O control register (TIOR).
When a general register (GRA/GRB) is used as an output compare register, its value is constantly
compared with the timer counter (TCNT) value. When the two values match (compare match), the
IMFA/IMFB bit is set to 1 in the timer status register (TSR). If compare match output is selected
in the TIOR, a specified value is output at the output compare pin.
When a general register is used as an input capture register, an external input capture signal is
detected and the TCNT value is stored. The IMFA/IMFB bit of the corresponding TSR is set to 1
at the same time. The valid edge or edges of the input capture signal are selected in the TIOR. The
TIOR setting is ignored when set for the PWM mode, complementary PWM mode or resetsynchronized PWM mode.
RENESAS 233
General registers are connected to the CPU by a 16-bit bus, so general registers can be written or
read by either word access or byte access. General registers are initialized to the output compare
register (no pin output) by a reset or in standby mode. The initial value is H'FFFF.
Table 10.5 General Registers A and B (GRA and GRB)
Channel
Abbreviation
Function
0
GRA0, GRB0
Output compare/input capture dual register
1
GRA1, GRB1
2
GRA2, GRB2
3
GRA3, GRB3
4
GRA4, GRB4
Bit:
Output compare/input capture dual register. Can also be set for
buffer operation in combination with the buffer registers (BRA, BRB)
15
14
13
12
11
10
9
8
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
10.2.8
Buffer Registers A and B (BRA, BRB)
Each buffer register is a 16-bit read/write register that is used in the buffer mode. The ITU has
four buffer registers, two each for channels 3 and 4 (table 10.6). Buffer operation can be set
independently by the timer function control register (TFCR) bits BFB4, BFA4, BFB3, and BFB3
bits. The buffer registers are paired with the general registers and their function changes
automatically to match the function of its corresponding general register.
The buffer registers are connected to the CPU by a 16-bit bus, so they can be written or read by
either word or byte access. Buffer registers are initialized to H'FFFF by a reset or in standby mode.
234 RENESAS
Table 10.6 Buffer Registers A and B (BRA, BRB)
Channel
Abbreviation
Function
3
BRA3, BRB3
4
BRA4, BRB4
When used for buffer operation:
When the corresponding GRA and GRB are output compare
registers, the buffer registers function as output compare buffer
registers that can automatically transfer the BRA and BRB values to
GRA and GRB upon a compare match.
When the corresponding GRA and GRB are input capture registers,
the buffer registers function as input capture buffer registers that can
automatically transfer the values stored until an input capture in the
GRA and GRB to the BRA and BRB.
Bit:
15
14
13
12
11
10
9
8
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
10.2.9
Timer Control Register (TCR)
The TCR is an 8-bit read/write register that selects the timer counter clock, the edges of the
external clock source, and the counter clear source. Each ITU channel has one TCR. TCR is
initialized H'80 or H'00 by a reset or the standby mode (table 10.7).
Table 10.7 Timer Control Register (TCR)
Channel
Abbreviation
0
TCR0
1
TCR1
2
TCR2
3
TCR3
4
TCR4
Function
The TCR controls the TCNTs. The TCRs have the same functions on all
channels. When channel 2 is set for phase counting mode, setting the
CKEG1, CKEG2 and TPSC2–TPSC0 bits will have no effect.
RENESAS 235
Bit:
7
6
5
4
Bit name:
—
CCLR1
CCLR0
Initial value:
*
0
0
0
R/W:
—
R/W
R/W
R/W
3
2
1
0
TPSC2
TPSC1
TPSC0
0
0
0
0
R/W
R/W
R/W
R/W
CKEG1 CKEG0
Note: Undefined
•
Bit 7 (reserved): Bit 7 is read as undefined. The write value should be 0 or 1.
•
Bits 6 and 5 (counter clear 1 and 0 (CCLR1 and CCLR0)): CCLR1 and CCLR0 select the
counter clear source.
Bit 6:
Bit 5:
CCLR1 CCLR0 Description
0
1
0
TCNT is not cleared (initial value)
1
TCNT is cleared by general register A (GRA) compare match or input capture* 1
0
TCNT is cleared by general register B (GRB) compare match or input capture* 1
1
Synchronizing clear: TCNT is cleared in synchronization with clear of other
timer counters operating in sync.* 2
Notes: 1. When GR is functioning as an output compare register, TCNT is cleared upon a
compare match. When functioning as an input capture register, TCNT is cleared upon
input capture.
2. The timer synchro register (TSNC) set the synchronization.
•
Bits 4 and 3 (external clock edge 1/0 (CKEG1 and CKEG0)): CKEG1 and CKEG0 select
external clock input edges. When channel 2 is set for phase counting mode, settings of the
CKEG1 and CKEG0 of the TCR are ignored and the phase counting mode operation takes
priority.
Bit 4:
CKEG
1
Bit 3:
CKEG
0
Description
0
0
Count rising edges (initial value)
1
Count falling edges
—
Count both rising and falling edges
1
•
Bits 2–0 (timer prescalar 2–0 (TPS2–TPS0)): TPS2–TPS0 select the counter clock source.
When TPSC2 = 0 and an internal clock source is selected, the timer counts only falling edges.
When TPSC2 = 1 and an external clock is selected, the count edge is as set by CKEG1 and
CKEG0. When the phase counting mode is selected for channel 2 (MDF bit in the timer mode
register is 1), the settings of TPSC2–TPSC0 of TCR2 are ignored and the phase counting
operation takes priority.
236 RENESAS
Bit 2:
Bit 1:
Bit 0:
TPSC2 TPSC1 TPSC0 Counter Clock (and cycle when φ = 10 MHz)
0
0
1
1
0
1
10.2.10
0
Internal clock φ (initial value)
1
Internal clock φ/2
0
Internal clock φ/4
1
Internal clock φ/8
0
External clock A (TCLKA)
1
External clock B (TCLKB)
0
External clock C (TCLKC)
1
External clock D (TCLKD)
Timer I/O Control Register (TIOR)
The timer I/O control register (TIOR) is an eight-bit read/write register that selects the output
compare or input capture function for the general registers GRA and GRB. It also selects the
function of the TIOCA and TIOCB pins. If output compare is selected, the TIOR also selects the
output settings. If input capture is selected, the TIOR also select the input capture edges. TIOR is
initialized to H'88 or H'08 on a reset or standby mode. Each ITU channel has one TIOR (table
10.8).
Table 10.8 Timer I/O Control Register (TIOR)
Channel
Abbreviation
0
TIOR0
1
TIOR1
2
TIOR2
3
TIOR3
4
TIOR4
Function
The TIOR controls the GRs. Some functions vary during PWM. When
channels 3 and 4 are set for complementary PWM mode/reset-synchronized
PWM mode, TIOR3 and TIOR4 settings are not valid.
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
IOB2
IOB1
IOB0
—
IOA2
IOA1
IOA0
Initial value:
*
0
0
0
1
0
0
0
R/W:
—
R/W
R/W
R/W
—
R/W
R/W
R/W
Note: Undefined
•
Bit 7 (reserved): Bit 7 is read as undefined. The write value should be 0 or 1.
RENESAS 237
•
Bits 6–4 (I/O control B2–B0 (IOB2–IOB0)): IOB2–IOB0 selects the GRB function.
Bit 6:
IOB2
Bit 5:
IOB1
Bit 4:
IOB0
0
0
0
1
1
0
GRB Function
GRB is an
Compare match with pin output disabled (initial value)
output compare 0 output at GRB compare match* 1
register
1 output at GRB compare match* 1
1
1
0
0
1
1
0
Output toggles at GRB compare match (1 output for
channel 2 only)* 1,* 2
GRB is an
input capture
register
GRB captures rising edge of input
GRB captures falling edge of input
GRB captures both edges of input
1
Notes: 1. After reset, the value output is 0 until the first compare match occurs.
2. Channel 2 has no compare-match driven toggle output function. If it is set for toggle, 1
is automatically selected as the output.
•
Bit 3 (reserved): Bit 3 always reads as 1. The write value should always be 1.
•
Bits 2–0 (I/O control A2–A0 (IOA2–IOA0)): IOA2–IOA0 select the GRB function.
Bit 2:
IOA2
Bit 1:
IOA1
Bit 0:
IOA0
0
0
0
1
1
0
GRA Function
GRA is an
Compare match with pin output disabled (initial value)
output compare 0 output at GRA compare match* 1
register
1 output at GRA compare match* 1
1
1
0
0
1
1
0
Output toggles at GRA compare match (1 output for
channel 2 only)* 1,* 2
GRA is an
input capture
register
GRA captures rising edge of input
GRA captures falling edge of input
GRA captures both edges of input
1
Notes: 1. After reset, the value output is 0 until the first compare match occurs.
2. Channel 2 has no compare-match driven toggle output function. If it is set for toggle, 1
is automatically selected as the output.
238 RENESAS
10.2.11
Timer Status Register (TSR)
The timer status register (TSR) is an eight-bit read/write register containing flags that indicate
timer counter (TCNT) overflow/underflow and general register (GRA/GRB) compare match or
input capture. These flags are interrupt sources. If the interrupt is enabled by the corresponding bit
in the timer interrupt enable register (TIER), an interrupt is requested of the CPU. TSR is
initialized by a reset or standby mode to H'F8 or H'78. Each ITU channel has one TSR (table
10.9).
Table 10.9 Timer Status Register (TSR)
Channel
Abbreviation
Function
0
TSR0
1
TSR1
The TSR indicates input capture, compare match and
overflow status.
2
TSR2
3
TSR3
4
TSR4
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
OVF
IMFB
IMFA
Initial value:
*1
1
1
1
1
0
0
0
R/W:
—
—
—
—
—
R/(W)*2
R/(W)*2 R/(W)*2
Notes: 1. Undefined
2. Write 0 to clear the flag.
•
Bits 7–3 (reserved): Bit 7 is read as undefined. Bits 6–3 are always read as 1. The write value
to bit 7 should be 0 or 1. The write value to bits 6–3 should always be 1.
•
Bit 2 (overflow flag (OVF)): OVF indicates a TCNT overflow/underflow has occurred.
Bit 2: OVF
Description
0
Clearing condition: Read OVF when OVF = 1, then write 0 in OVF
(initial value)
1
Setting condition: TCNT overflowed from H'FFFF–H'0000 or
underflowed from H'0000–H'FFFF.
Note: A TCNT underflow occurs when the TCNT up/down counter is functioning. It may occur in
the following cases: (1) When channel 2 is set in the phase counting mode (MDF bit of
TMDR is 1), or (2) When channel 3 and 4 are set to the complementary PWM mode (CMD1
bit of TFCR is 1 and CMD0 bit is 0).
RENESAS 239
•
Bit 1 (input capture/compare match B (IMFB)): IMFB indicates a GRB compare match or
input capture.
Bit 1: IMFB
Description
0
Clearing condition: Read IMFB when IMFB = 1, then write 0 in IMFB
(initial value)
1
Setting condition:
• GRB is functioning as an output compare register and TCNT = GRB
• GRB is functioning as an input capture register and the value of
TCNT is transferred to GRB by an input capture signal
•
Bit 0 (input capture/compare match A (IMFA)): IMFA indicates a GRA compare match or
input capture.
Bit 0: IMFA
Description
0
Read IMFA when IMFA = 1, then write 0 in IMFA (initial value).
Clearing condition: DMAC is activated by an IMIA interrupt (only
channels 0–3)
1
Setting condition:
• GRA is functions as an output compare register and TCNT = GRA
• GRA is functioning as an input capture register and the value of
TCNT is transferred to GRA by an input capture signal
10.2.12
Timer Interrupt Enable Register (TIER)
The timer status interrupt enable register (TIER) is an eight-bit read/write register that controls
enabling/disabling of overflow interrupt requests and general register compare match/input capture
interrupt requests. TIER is initialized by a reset or standby mode to H'F8 or H'78. Each ITU
channel has one TIER (table 10.10).
Table 10.10 Timer Interrupt Enable Register (TIER)
Channel
Abbreviation
Function
0
TIER0
The TIER controls interrupt enable/disable.
1
TIER1
2
TIER2
3
TIER3
4
TIER4
240 RENESAS
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
OVIE
IMIEB
IMIEA
Initial value:
*
1
1
1
1
0
0
0
R/W:
—
—
—
—
—
R/W
R/W
R/W
Note: Undefined
•
Bits 7–3 (reserved): Bit 7 is read as undefined. Bits 6–3 are always read as 1. The write value
to bit 7 should be 0 or 1. The write value to bits 6–3 should always be 1.
•
Bit 2 (overflow interrupt enable (OVIE)): When the TSR overflow flag (OVF) is set to 1,
OVIE enables or disables interrupt requests from the OVF.
Bit 2: OVIE
Description
0
Disables interrupt requests by the OVF (initial value)
1
•
Enables interrupt requests from the OVF
Bit 1 (input capture/compare match interrupt enable B (IMIEB)): When the IMFB bit of the
TSR is set to 1, IMIEB enables or disables the interrupt requests from the IMFB.
Bit 1: IMIEB
Description
0
Disables interrupt requests by the IMFB (IMIB) (initial value)
1
Enables interrupt requests from the IMFB (IMIB)
•
Bit 0 (input capture/compare match interrupt enable A (IMIEA)): When the IMFA bit of the
TSR is set to 1, IMIEA enables or disables the interrupt requests from the IMFA.
Bit 0: IMIEA
Description
0
Disables interrupt requests by the IMFA (IMIA) (initial value)
1
Enables interrupt requests from the IMFA (IMIA)
10.3
CPU Interface
10.3.1
16-Bit Accessible Registers
The timer counters (TCNT), general registers A and B (GRA, GRB), and buffer registers A and B
(BRA, BRB) are 16-bit registers. The SH CPU can access these registers a word at a time using a
16-bit data bus. Byte access is also possible. Read and write operations performed on the TCNT in
word units are shown in figures 10.6 and 10.7. Byte-unit read and write operations on TCNTH and
TCNTL are shown in figures 10.8–10.11.
RENESAS 241
Internal data bus
H
H
CPU
Bus
interface
L
L
TCNTH
Module
data bus
TCNTL
Figure 10.6 Accessing TCNT (CPU–TCNT (word))
Internal data bus
H
H
CPU
Bus
interface
L
L
TCNTH
Module
data bus
TCNTL
Figure 10.7 Accessing TCNT (TCNT–CPU (word))
Internal data bus
H
H
CPU
L
Bus
interface
L
TCNTH
TCNTL
Figure 10.8 Accessing TCNT (CPU–TCNT (upper byte))
242 RENESAS
Module
data bus
Internal data bus
H
H
CPU
Bus
interface
L
L
TCNTH
Module
data bus
TCNTL
Figure 10.9 Accessing TCNT (CPU–TCNT (lower byte))
Internal data bus
H
H
CPU
Bus
interface
L
L
TCNTH
Module
data bus
TCNTL
Figure 10.10 Accessing TCNT (TCNT–CPU (upper byte))
Internal data bus
H
H
CPU
L
Bus
interface
L
TCNTH
Module
data bus
TCNTL
Figure 10.11 Accessing TCNT (TCNT–CPU (lower byte))
10.3.2
8-Bit Accessible Registers
All registers other than the TCNT, general registers, and buffer registers are 8-bit registers. These
are connected to the CPU by an 8-bit data bus. Figures 10.12 and 10.13 illustrate reading and
writing in byte units with the timer control register (TCR). These registers must be accessed by
byte access.
RENESAS 243
Internal data bus
Module
data bus
Bus
interface
CPU
TCR
Figure 10.12 TCR Access (CPU–TCR)
Internal data bus
CPU
Module
data bus
Bus
interface
TCR
Figure 10.13 TCR Access (TCR–CPU )
10.4
Description of Operation
10.4.1
Overview
The operation modes are described below.
Ordinary Operation: Each channel has a timer counter (TCNT) and general register (GR). The
TCNT is an upcounter and can also operate as a free-running counter, periodic counter or external
event counter. General registers A and B (GRA and GRB) can be used as output compare registers
or input capture registers.
Synchronized Operation: The TCNT of a channel set for synchronized operation does a
synchronized preset. When any TCNT of a channel operating in the synchronized mode is
rewritten, the TCNTs of other channels are simultaneously rewritten as well. The CCLR1 and
CCLR0 bits of the timer control register of multiple channels set for synchronous operation can be
set to clear the TCNTs simultaneously.
244 RENESAS
PWM Mode: In PWM mode, a PWM waveform is output from the TIOCA pin. Output becomes
1 upon compare match A and 0 upon compare match B. GRA and GRB can be set so that the
PWM waveform output has a duty cycle between 0% and 100%. When set for PWM mode, the
GRA and GRB automatically become output compare registers.
Reset-synchronized PWM Mode: Three pairs of positive and negative PWM waveforms can be
obtained using channels 3 and 4 (the three phases of the PWM waveform share a transition point
on one side). When set for reset-synchronized PWM mode, GRA3, GRB3, GRA4, and GRB4
automatically become output compare registers. The TIOCA3, TIOCB3, TIOCA4, TOCXA4,
TIOCB4, and TOCXB4 pins also automatically become PWM output pins and TCNT3 becomes
an upcounter. TCNT4 functions independently (although GRA and GRB are isolated from
TCNT4).
Complementary PWM Mode: Three pairs of complementary positive and negative PWM
waveforms whose positive and negative phases do not overlap can be obtained using channels 3
and 4. When set for complementary PWM mode, GRA3, GRB3, GRA4, and GRB4 automatically
become output compare registers. The TIOCA3, TIOCB3, TIOCA4, TOCXA4, TIOCB4, and
TOCXB4 pins also automatically become PWM output pins while TCNT3 and TCNT4 become
upcounters.
Phase Counting Mode: In phase counting mode, the phase differential between two clocks input
from the TCLKA and TCLKB pins is detected and the TCNT2 operates as an up/downcounter. In
phase counting mode, the TCLKA and TCLKB pins become clock inputs and TCNT2 functions as
an up/downcounter.
Buffer Mode:
•
•
•
•
When GR is an output compare register: The BR value of each channel is transferred to the GR
when a compare match occurs.
When GR is an input capture register: The TCNT value is transferred to the GR when an input
capture occurs and simultaneously the value previously stored in the GR is transferred to the
BR.
Complementary PWM mode: When the TCNT3 and TCNT4 change count directions, the BR
value is transferred to the GR.
Reset-synchronized PWM mode: The BR value is transferred to GR upon a GRA3 compare
match.
10.4.2
Basic Functions
Counter Operation: When a start bit (STR0–STR4) in the timer start register (TSTR) is set to 1,
the corresponding timer counter (TCNT) starts counting. There are two counting modes: a freerunning mode and a periodic mode.
RENESAS 245
•
Procedure for selecting counting mode (figure 10.14):
1. Set bits TPSC2–TPSC0 in the TCR to select the counter clock source. If an external clock
source is selected, set bits CKEG1 and CKEG0 in the TCR to select the desired edges of
the external clock signal.
2. To operate as a periodic counter, set CCLR1 and CCLR0 in the TCR to select whether to
clear the TCNT at GRA compare match or GRB compare match.
3. Set the GRA or GRB selected in step 2 as an output compare register using the timer I/O
control register (TIOR).
4. Write the desired cycle value in the GRA or GRB selected in step 1.
5. Set the STR bit in the TSTR to 1 to start counting.
Counting mode selection
Select counter clock
Counting?
(1)
No
Yes
Free-running counter
Periodic counter
Select counter
clear source
(2)
Select output
compare register
(3)
Set period
(4)
Start counting
(5)
Periodic counter
Start counting
Free-running counter
Figure 10.14 Procedure for Selecting the Counting Mode
246 RENESAS
(5)
•
Free-running count and periodic count
A reset of the counters for channels 0–4 leaves them all in the free-running mode. When a
corresponding bit in the TSTR is set to 1, the corresponding timer counter operates as a freerunning counter and begins to increment. When the count wraps around from H'FFFF–H'0000,
the overflow flag (OVF) in the timer status register (TSR) is set to 1. If the OVIE bit in the
timer’s corresponding interrupt enable register (TIER) is set to 1, the CPU will be asked for an
interrupt. After the TCNT overflows, counting continues from H'0000. Figure 10.15 shows an
example of free-running counting.
Periodic counter operation is obtained for a given channel’s TCNT by selecting compare match
as a TCNT clear source. (Set the GRA or GRB for period setting to output compare register
and select counter clear upon compare match using the CCLR1 and CCLR0 bits of the timer
control register (TCR).) After setting, the TCNT begins incrementing as a periodic counter
when the corresponding bit of TSTR is set to 1. When the count matches GRA or GRB, the
IMFA/IMFB bit in the TSR is set to 1 and the counter is automatically cleared to H'0000. If the
IMIEA/IMIEB bit of the corresponding TIER is set to 1 at this point, the CPU will be asked for
an interrupt. After the compare match, TCNT continues counting from H'0000. Figure 10.16
shows an example of periodic counting.
TCNT value
H'FFFF
H'0000
Time
STR0–STR4
OVF
Figure 10.15 Free-Running Counter Operation
RENESAS 247
Counter cleared by
GR compare match
TCNT value
GR
H'0000
Time
STR0–STR4
IMF
Figure 10.16 Periodic Counter Operation
•
TCNT counter timing
Internal clock source: Bits TPSC2–TPSC0 in the TCR select the system clock (CK) or one of
three internal clock sources (φ/2, φ/4, φ/8) obtained by prescaling the system clock. Figure
10.17 shows the timing.
External clock source: The external clock input pin (TCLKA–TCLKD) source is selected by
bits TPSC2–TPSC0 in the TCR and its valid edges are selected with the CKEG1 and CKEG0
bits of the TCR. The rising edge, falling edge, or both edges can be selected. The pulse width
of the external clock signal must be at least 1.5 system clocks when a single edge is selected
and at least 2.5 system clocks when both edges are selected. Shorter pulses will not be counted
correctly. Figure 10.18 shows the timing when both edges are detected.
CK
Internal clock
TCNT input
clock
TCNT value
N–1
N
Figure 10.17 Count Timing for Internal Clock Sources
248 RENESAS
N+1
CK
External clock
input pin
TCNT input
clock
TCNT
N–1
N
N+1
Figure 10.18 Count Timing for External Clock Sources
Compare-Match Waveform Output Function: For ITU channels 0, 1, 3, and 4, the output from
the corresponding TIOCA and TIOCB pins upon compare matches A and B can be in three
modes: 0-level output, 1-level output, or toggle. Toggle output cannot be selected in channel 2.
•
Procedure for selecting the waveform output mode (figure 10.19):
1. Set the TIOR to select 0 output, 1 output, or toggle output for compare match output. The
compare match output pin will output 0 until the first compare match occurs.
2. Set a value in the GRA or GRB to select the compare match timing.
3. Set the STR bit in the TSTR to 1 to start counting.
Output selection
Select waveform
output mode
(1)
Select
output timing
(2)
Start counting
(3)
Waveform output
Figure 10.19 Procedure for Selecting the Compare Match Waveform Output Mode
RENESAS 249
•
Waveform output operation
Figure 10.20 illustrates 0 output/1 output. In the example, TCNT is a free-running counter, 0 is
output upon compare match A and 1 is output upon compare match B. When the pin level
matches the set level, the pin level does not change.
Figure 10.21 shows an example of toggle output. In the figure, the TCNT operates as a
periodic counter cleared by GRB compare match with toggle output at both compare match A
and compare match B.
TCNT value
H'FFFF
GRB
GRA
Time
TIOCB
Does not change
Does not change
1 output
TIOCA
Does not change
Does not change
Figure 10.20 Example of 0 Output/1 Output
250 RENESAS
0 output
TCNT value
Counter cleared at
GRB compare match
GRB
GRA
Time
TIOCB
Toggle
output
TIOCA
Toggle
output
Figure 10.21 Example of Toggle Output
RENESAS 251
•
Compare match output timing
The compare match signal is generated in the last state in which the TCNT and the general
register match (when the TCNT changes from the matching value to the next value). When a
compare match signal is generated, the output value set in TIOR is output to the output
compare pin (TIOCA, TIOCB). Accordingly, when the TCNT matches a general register, the
compare match signal is not generated until the next counter clock pulse. Figure 10.22 shows
the output timing of the compare match signal.
CK
TCNT input
clock
TCNT
N
GR
N
N–1
Compare
match signal
TIOCA
TIOCB
Figure 10.22 Compare Match Signal Output Timing
252 RENESAS
Input Capture Mode: In the input capture mode, the counter value is captured into a general
register when the input edge is detected at an input capture/output compare pin (TIOCA, TIOCB).
Detection can take place on the rising edge, falling edge, or both edges. Pulse width and cycle can
be measured by using the input capture function.
•
Procedure for selecting the input capture mode (figure 10.23)
1. Set the TIOR to select the input capture function of the GR and select the rising edge,
falling edge, or both edges as the input edge of the input capture signal. Put the
corresponding port into input-capture using the pin function controller before setting the
TIOR.
2. Set the STR bit in the TSTR to 1 to start the TCNT counting.
Input selection
Select input-capture input
Start counting
Capture
Figure 10.23 Procedure for Selecting Input Capture Mode
RENESAS 253
•
Input capture operation
Figure 10.24 illustrates input capture. The falling edge of TIOCB and both edges of TIOCA
are selected as input capture edges. In the example, TCNT is set to clear at the input capture of
GRB.
TCNT
value
Counter cleared
by TIOCB input
(falling edge)
H'0180
H'0160
H'0005
H'0000
Time
TIOCB
TIOCA
GRA
H'0005
H'0160
GRB
H'0180
Figure 10.24 Input Capture Operation
254 RENESAS
•
Input capture timing
Input capture on the rising edge, falling edge, or both edges can be selected by settings in the
TIOR. Figure 10.25 shows the timing when the rising edge is selected. The pulse width of the
input capture signal must be at least 1.5 system clocks for single-edge capture, and 2.5 system
clocks for capture of both edges.
CK
Input capture
input
Input capture
signal
TCNT
N
GRA/GRB
N
Figure 10.25 Input Capture Signal Timing
RENESAS 255
10.4.3
Synchronizing Mode
In the synchronizing mode, two or more timer counters can be rewritten simultaneously
(synchronized preset). Multiple timer counters can also be cleared simultaneously using TCR
settings (synchronized clear). The synchronizing mode can increase the general registers for a
single time base. All five channels can be set for synchronous operation.
Procedure for Selecting the Synchronizing Mode (figure 10.26):
1. Set 1 in the SYNC bit of the timer synchro register (TSNC) to use the channels in the
synchronizing mode.
2. When a value is written in the TCNT in any of the synchronized channels, the same value is
simultaneously written in the TCNT in the other channels.
3. Set the counter to clear with compare match/input capture using bits CCLR1 and CCLR0 in the
TCR.
4. Set the counter clear source to synchronized clear using the CCLR1 and CCLR0 bits.
5. Set the STR bits in the TSTR to 1 to start counting in the TCNT.
Select synchronizing
mode
Set synchronizing mode
(1)
Synchronized preset
Set TCNT
Synchronized clear
(2)
Channel that
generated clear
source?
Yes
No
Select counter
clear source
(3)
Select counter
clear source
(4)
Start counting
(5)
Start counting
(5)
Synchronizing preset
Counter clear
Synchronized clear
Figure 10.26 Procedure for Selecting the Synchronizing Mode
256 RENESAS
Synchronized Operation: Figure 10.27 shows an example of synchronized operation. Channels
0, 1, and 2 are set to synchronized operation and PWM output. Channel 0 is set for a counter clear
upon compare match with GRB0. Channels 1 and 2 are set for counter clears by synchronizing
clears. Accordingly, their timers are sync preset, then sync cleared by a GRB0 compare match,
and then a three-phase PWM waveform is output from the TIOCA0, TIOCA1 and TIOCA2 pins.
See section 10.4.4, PWM Mode, for details on the PWM mode.
TCNT0–TCNT2 values
Synchronized clear on GRB0 compare match
GRB0
GRB1
GRA0
GRB2
GRA1
GRA2
Time
TIOCA0
TIOCA1
TIOCA2
Figure 10.27 Synchronized Operation Example
RENESAS 257
10.4.4
PWM Mode
The PWM mode is controlled using both the GRA and GRB in pairs. The PWM waveform is
output from the TIOCA output pin. The PWM waveform’s 1 output timing is set in GRA and the 0
output timing is set in GRB. A PWM waveform with duty cycle between 0% and 100% can be
output from the TIOCA pin by having either compare match GRA or GRB be the counter clear
source for the timer counter. All five channels can be set to PWM mode.
Table 10.11 lists the combinations of PWM output pins and registers. Note that when the GRA
and GRB are set to the same value, the output will not change even if a compare match occurs.
Table 10.11 Combinations of PWM Output Pins and Registers
Channel
Output Pin
1 Output
0 Output
0
TIOCA0
GRA0
GRB0
1
TIOCA1
GRA1
GRB1
2
TIOCA2
GRA2
GRB2
3
TIOCA3
GRA3
GRB3
4
TIOCA4
GRA4
GRB4
Procedure for Selecting the PWM Mode (figure 10.28):
1. Set bits TPSC2–TPSC0 in the TCR to select the counter clock source. If an external clock
source is selected, set bits CKEG1 and CKEG0 in the TCR to select the desired edges of the
external clock signal.
2. Set CCLR1 and CCLR0 in the TCR to select the counter clear source.
3. Set the time at which the PWM waveform should go to 1 in the GRA.
4. Set the time at which the PWM waveform should go to 0 in the GRB.
5. Set the PWM bit in TMDR to select the PWM mode. When the PWM mode is selected,
regardless of the contents of TIOR, the GRA and GRB become output compare registers
specifying the times at which the PWM waveform goes high and low. TIOCA automatically
becomes a PWM output pin. TIOCB becomes whatever is set in the TIOR's IOB1 and IOB0
bits.
6. Set the STR bit in the TSTR to let the TCNT start counting.
258 RENESAS
PWM mode
Select counter clock
(1)
Select counter clear source
(2)
Set GRA
(3)
Set GRB
(4)
Select PWM mode
(5)
Start counting
(6)
PWM mode
Figure 10.28 Procedure for Selecting the PWM Mode
PWM Mode Operation: Figure 10.29 illustrates PWM mode operations. When the PWM mode
is set, the TIOCA pin becomes the output pin. Output is 1 when the TCNT matches the GRA, and
0 when the TCNT matches the GRB. The TCNT can be cleared by compare match with either
GRA or GRB. This can be used in both free-running and synchronized operation.
Figure 10.30 shows examples of PWM waveforms output with 0% and 100% duty cycles. A 0%
duty waveform can be obtained by setting the counter clear source to GRB and then setting GRA
to a larger value than GRB. A 100% duty waveform can be obtained by setting the counter clear
source to GRA and then setting GRB to a larger value than GRA.
RENESAS 259
TCNT value
Counter cleared by GRA compare match
GRA
GRB
Time
TIOCA
a. Counter cleared by GRA
TCNT value
Counter cleared by GRB compare match
GRB
GRA
Time
TIOCA
b. Counter cleared by GRB
Figure 10.29 PWM Mode Operation Example 1
260 RENESAS
TCNT value
Counter cleared on compare match B
GRB
GRA
Time
H'0000
TIOCA
GRA write
GRA write
a. 0% duty
TCNT value
Counter cleared on compare match A
GRA
GRB
Time
H'0000
TIOCA
GRB write
GRB write
b. 100% duty
Figure 10.30 PWM Mode Operation Example 2
RENESAS 261
10.4.5
Reset-Synchronized PWM Mode
In the reset-synchronized PWM mode, three pairs of complementary positive and negative PWM
waveforms that share a common wave turning point can be obtained using channels 3 and 4. When
set for reset-synchronized PWM mode, the TIOCA3, TIOCB3, TIOCA4, TOCXA4, TIOCB4, and
TOCXB4 pins automatically become PWM output pins and TCNT3 becomes an upcounter. Table
10.12 shows the PWM output pins used and table 10.13 shows the settings of the registers used.
Table 10.12 Output Pins for Reset-Synchronized PWM Mode
Channel
Output Pin
Description
3
TIOCA3
PWM output 1
TIOCB3
PWM output 1' (negative-phase waveform of PWM output 1)
TIOCA4
PWM output 2
TOCXA4
PWM output 2' (negative-phase waveform of PWM output 2)
TIOCB4
PWM output 3
TOCXB4
PWM output 3' (negative-phase waveform of PWM output 3)
4
Table 10.13 Register Settings for Reset-Synchronized PWM Mode
Register
Description of Contents
TCNT3
Initial setting of H'0000
TCNT4
Not used (functions independently)
GRA3
Set count cycle for TCNT3
GRB3
Sets the turning point for PWM waveform output by the TIOCA3 and TIOCB3 pins
GRA4
Sets the turning point for PWM waveform output by the TIOCA4 and TOCXA4 pins
GRB4
Sets the turning point for PWM waveform output by the TIOCB4 and TOCXB4 pins
Procedure for Selecting the Reset-Synchronized PWM Mode (figure 10.31):
1. Clear the STR3 bits in the TSTR to halt TCNT3. The reset-synchronized PWM mode must be
set up while TCNT3 is halted.
2. Set bits TPSC2–TPSC0 in the TCR to select the counter clock source for channel 3. If an
external clock source is selected, select the external clock edges with bits CKEG1 and CKEG0
in the TCR.
3. Set bits CCLR1 and CCLR0 in the TCR3 to select GRA3 as a counter clear source.
4. Set bits CMD1 and CMD0 in TMDB to select the reset-synchronized PWM mode. TIOCA3–
TIOCB4, TOCXA4, and TOCXB4 automatically become PWM output pins.
262 RENESAS
5. Reset the TCNT3 (to H'0000). The TCNT4 need not be set.
6. The GRA3 is the waveform period register. Set the waveform period value in the GRA3. Set
the transition times of the PWM output waveforms in the GRB3, GRA4, and GRB4. Set times
within the compare match range of the TCNT3.
X ≤ GRA3 (X: setting value)
7. Set the STR3 bits in the TSTR to 1 to let the TCNT3 start counting.
Reset synchronized
PWM mode
Stop counting
(1)
Select counter clock
(2)
Select counter clear source
(3)
Select reset-synchronized
PWM mode
(4)
Set TCNT
(5)
Set general registers
(6)
Start counting
(7)
Reset-synchronized PWM mode
Figure 10.31 Procedure for Selecting the Reset-Synchronized PWM Mode
RENESAS 263
Reset-Synchronized PWM Mode Operation: Figure 10.32 shows an example of operation in the
reset-synchronized PWM mode. TCNT3 operates as an upcounter that is cleared to H'0000 at
compare match with GRA3. TCNT4 runs independently and is isolated from GRA4 and GRB4.
The PWM waveform outputs toggle at each compare match (GRB3, GRA3, and GRB4 with
TCNT3) and when the counter is cleared.
See section 10.4.8, Buffer Mode, for details on simultaneously setting reset-synchronized PWM
mode and buffer operation.
TCNT value
Counter cleared at GRA3 compare match
GRA3
GRB3
GRA4
GRB4
Time
TIOCA3
TIOCB3
TIOCA4
TOCXA4
TIOCB4
TOCXB4
Figure 10.32 Reset-Synchronized PWM Mode Operation Example 1
10.4.6
Complementary PWM Mode
In the complementary PWM mode, three pairs of complementary, non-overlapping, positive and
negative PWM waveforms can be obtained using channels 3 and 4. In complementary PWM
mode, the TIOCA3, TIOCB3, TIOCA4, TOCXA4, TIOCB4, and TOCXB4 pins automatically
become PWM output pins and TCNT3 and TCNT4 become upcounters. Table 10.14 shows the
PWM output pins used and table 10.15 shows the settings of the registers used.
264 RENESAS
Table 10.14 Output Pins for Complementary PWM Mode
Channel
Output Pin
Description
3
TIOCA3
PWM output 1
TIOCB3
PWM output 1' (non-overlapping negative-phase waveform of
PWM output 1)
TIOCA4
PWM output 2
TOCXA4
PWM output 2' (non-overlapping negative-phase waveform of
PWM output 2)
TIOCB4
PWM output 3
TOCXB4
PWM output 3' (non-overlapping negative-phase waveform of
PWM output 3)
4
Table 10.15 Register Settings for Complementary PWM Mode
Register
Description of Contents
TCNT3
Initial setting of non-overlap cycle (the difference with TCNT4)
TCNT4
Initial setting of H'0000
GRA3
Set upper limit of TCNT3–1
GRB3
Sets the turning point for PWM waveform output by the TIOCA3 and TIOCB3 pins.
GRA4
Sets the turning point for PWM waveform output by the TIOCA4 and TOCXA4 pins.
GRB4
Sets the turning point for PWM waveform output by the TIOCB4 and TOCXB4 pins.
Procedure for Selecting the Complementary PWM Mode (figure 10.33):
1. Clear STR3 and STR4 bits in the TSTR to halt the timer counters. The complementary PWM
mode must be set up while TCNT3 and TCNT4 are halted.
2. Set bits TPSC2–TPSC0 in the TCR to select the same counter clock source for channels 3 and
4. If an external clock source is selected, select the external clock edges with bits CKEG1 and
CKEG0 in the TCR. Do not select any counter clear source with bits CCLR1 and CCLR0 in
the TCR.
3. Set bits CMD1 and CMD0 in TMDB to select the complementary PWM mode. TIOCA3–
TIOCB4, TOCXA4, and TOCXB4 automatically become PWM pins.
4. Reset TCNT4 (to H'0000). Set the non-overlap offset in TCNT3. Do not set TCNT3 and
TCNT4 to the same value.
RENESAS 265
5. GRA3 is the waveform period register. Set the upper limit of TCNT3–1*. Set the transition
times of the PWM output waveforms in GRB3, GRA4, and GRB4. Set times within the
compare match range of TCNT3 and TCNT4.
T≤X
(X: initial setting of GRB3, GRA4, and GRB4; T: initial setting of TCNT3)
Note:
GRA3 = [cycle count/2] + [count of non-overlaps] – 2cyc=[upper limit of TCNT3] – 1
6. Set the STR3 and STR4 bits in the TSTR to 1 to let TCNT3 and TCNT4 start counting.
Complementary PWM mode
Stop counting
(1)
Select counter clock
(2)
Select complementary
PWM mode
(3)
Set TCNT
(4)
Set general registers
(5)
Start counting
(6)
Complementary PWM mode
Note
To re-engage the complementary PWM mode after it has been aborted, start settings
from step 1.
Figure 10.33 Procedure for Selecting the Complementary PWM Mode
Complementary PWM Mode Operation: Figure 10.34 shows an example of operation in the
complementary PWM mode. TCNT3 and TCNT4 operate as up/downcounters, counting down
from compare match of TCNT3 and GRA3 and counting up when TCNT4 underflows. PWM
waveforms are output by repeated compare matches with each of the general registers in the
266 RENESAS
sequence TCNT3, TCNT4, TCNT4, TCNT3 (in this mode, TCNT3 starts out larger than TCNT4).
Figure 10.35 shows examples of PWM waveforms with 0% and 100% duty cycles (in one phase)
in the complementary PWM mode. In this example, the pin output changes upon GRB3 compare
match, so duty cycles of 0% and 100% can be obtained by setting GRB3 to a value larger than
GRA3. Combining buffer operation with the above operation makes it easy to change the duty
while operating. See section 10.4.8, Buffer Operation, for details.
TCNT3, TCNT4 value
GRA3
Down-counting starts at compare match
between TCNT3 and GRA3
TCNT3
GRB3
GRA4
GRB4
TCNT4
Time
Up-counting starts at TCNT4 underflow
TIOCA3
TIOCB3
TIOCA4
TOCXA4
TIOCB4
TOCXB4
Figure 10.34 Complementary PWM Mode Operation Example 1
RENESAS 267
TCNT3, TCNT4 value
GRA3
GRB3
Time
TIOCA3
TIOCB3
Duty 0%
(a) With 0% duty
TCNT3, TCNT4 value
GRA3
GRB3
Time
TIOCA3
TIOCB3
Duty 100%
(b) With 100% duty
Figure 10.35 Complementary PWM Mode Operation Example 2
At the point where the up-count/down-count changes in the complementary PWM mode, TCNT3
and TCNT4 will overshoot and undershoot, respectively. When this occurs, the setting conditions
for the IMFA bit of channel 3 and the overflow flag (OVF) of channel 4 are different from usual.
Transfer conditions for the buffer also differ. The timing is as shown in figures 10.36 and 10.37.
268 RENESAS
TCNT3
N–1
N
N+1
N
N–1
N
GRA3
Flag not set
IMFA
Set to 1
Buffer transfer
signal (BR to GR)
GR
Buffer transfer performed
Buffer transfer
not performed
Figure 10.36 Overshoot Timing
Underflow
TCNT4
H' 0001
H' 0000
H' FFFF
Overflow
H' 0000
Flag not set
OVF
Set to 1
Buffer transfer
signal (BR to GR)
GR
Buffer transfer performed
Buffer transfer
not performed
Figure 10.37 Undershoot Timing
RENESAS 269
The IMFA bit of channel 3 is set to 1 for increment pulses and the OVF bit of channel 4 is set to 1
for underflows only. The buffer register (BR) set for the buffer operation is transferred to the GR
upon compare match A3 (when incrementing) or TCNT4 underflow.
GR Setting in Complementary Mode: Be aware of the following when setting the general
registers in complementary PWM mode and when making changes during operation.
•
•
•
Initial values: Setting H'0000 to T–1 (the initial setting T: TCNT3) is prohibited. After
counting starts, this setting is allowed from the point when the first A3 compare match occurs.
Methods of changing settings: Use the buffer operation. Writing directly to general registers
may result in incorrect waveform output.
When changing settings: See figure 10.38.
GRA3
GR
H' 0000
Prohibited
BR
GR
Figure 10.38 Example of Changing GR Settings with Buffer Operation (1)
270 RENESAS
Buffer Transfers when Changing from Increment to Decrement: When the contents of the GR
are within GRA3 – T + 1 and GRA3, do not transfer a value outside this range. When the contents
of GR are outside this range, do not a transfer a value within it. Figure 10.39 illustrates a point of
caution regarding changing of GR settings with a buffer operation.
GRA3 + 1
GRA3
Changes inhibited
TCNT3
GRA3 – T + 1
GRA3 – T
TCNT4
Figure 10.39 Caution for Changing GR Settings with Buffer Operation (1)
Buffer Transfers when Changing from Decrement to Increment: When the contents of the GR
are within H'0000 to T–1, do not transfer a value outside this range. When the contents of GR are
outside this range, do not transfer a value within it. Figure 10.40 illustrates this point of caution
regarding changing of GR settings with a buffer operation
TCNT3
TCNT4
T
T–1
Changes inhibited
H' 0000
H' FFFF
Figure 10.40 Caution for Changing GR Settings with Buffer Operation (2)
When GR Settings are Outside the Count Range (H'0000–GRA3): Waveforms of duty cycle
0% and 100% can be output by setting GR outside the count area. Be sure to make the direction of
the count (increment/decrement) when writing a setting from outside the count area into the buffer
register (BR) the same as the count direction when writing the setting that returns to within the
count area in the BR.
RENESAS 271
GRA3
GR
H' 0000
Duty 0%
Duty 100%
Output pin
Output pin
BR
GR
Write on decrement
Write on increment
Figure 10.41 Example of Changing GR Settings with Buffer Operation (2)
The above settings are made by detecting the occurrence of a GRA3 compare match or underflow
of TCNT4 and then writing to BR. They can also be accomplished by starting up the DMAC with
a GRA3 compare match.
10.4.7
Phase Counting Mode
The phase counting mode detects the phase differential of two external clock inputs (TCLKA and
TCLKB) and counts TCNT2 up or down. When set in the phase counting mode, the TCLKA and
TCLKB pins automatically become external clock input pins, regardless of the settings of the
TPSC2–TPSC0 bits of TCR2 or the CKEG1 and CKEG0 bits. TCNT2 also becomes an up/down
counter. Since the TCR2 CCLR1/CCLR0 bits, TIOR2, TIER2, TSR2, GRA2 and GRB2 are all
enabled, input capture and compare match functions and interrupt sources can be used. Phase
counting is available only in channel 2.
Procedure for Selecting the Phase Counting Mode: Figure 10.42 shows the procedure for
selecting the phase counting mode.
1. Set the MDF bit of the timer mode register (TMDR) to 1 to select the phase counting mode.
2. Select the flag set conditions using the FDIR bit of the TMDR.
3. Set the STR2 bit of the timer start register (TSTR) to 1 to start the count.
272 RENESAS
Phase counting mode
Select phase counting mode
(1)
Select flag setting condition
(2)
Start counting
(3)
Phase counting mode
Figure 10.42 Procedure for Selecting the Phase Counting Mode
Phase Counting Operation: Figure 10.43 shows an example of phase counting mode operation.
Table 10.16 lists the upcounting and downcounting conditions for TCNT2. The ITU counts on
both rise and fall edges of TCLKA and TCLKB. The phase differential and overlap of TCLKA
and TCLKB must be 1.5 cycles or more and the pulse width must be 2.5 cycles or more.
TCNT2 value
Increment
Decrement
TCNT2
Time
TCLKB
TCLKA
Figure 10.43 Phase Counting Mode Operation
Table 10.16 Up/Down Counting Conditions
Counting
Direction
Increment
TCLKB
Rising
High
Falling
Low
Rising
High
Falling
Low
TCLKA
Low
Rising
High
Falling
High
Falling
Low
Rising
Decrement
RENESAS 273
Phase
differential
Phase
differential
Pulse
width
Pulse
width
TCLKA
TCLKB
Overlap
Overlap
Phase differential, overlap: 1.5 cycles minimum
Pulse width: 2.5 cycles minimum
Figure 10.44 Phase Differentials, Overlap and Pulse Width in the Phase Counting Mode
10.4.8
Buffer Mode
In the buffer mode, the buffer operation functions differ depending on whether the general
registers are set to output compare or input capture, the reset-synchronized PWM mode, or
complementary PWM mode. The buffer mode is a function of channels 3 and 4 only. Buffer
operations set this way function as follows.
GR is an Output Compare Register: The value of the buffer registers of a channel is transferred
to the GR when the channel experiences a compare match. This is illustrated in figure 10.45.
Compare match signal
BR
GR
Comparator
TCNT
Figure 10.45 Compare Match Buffer Operation
GR is an Input Capture Register: TCNT values are transferred to GR when input capture occurs
and the value previously stored in GR is transferred to BR. This operation is illustrated in figure
10.46.
274 RENESAS
Input capture signal
BR
GR
TCNT
Figure 10.46 Input Capture Buffer Operation
Complementary PWM Mode: When the count direction of TCNT3 and TCNT4 change, the BR
value is transferred to the GR. The following timing is employed for this transfer:
•
•
Whenever TCNT3 and GRA3 compare-match
Whenever TCNT4 underflows
Reset-Synchronized PWM Mode: The BR value is transferred to GR upon a GRA3 compare
match.
Procedure for Selecting the Buffer Mode (figure 10.47):
1. Set the TIOR to select the output compare or input capture function of the GR.
2. Set bits BFA3, BFB3 and BFB4 in the TFCR to select the buffer mode for the GR.
3. Set the STR bit in the TSTR to 1 to start the TCNT counting.
Buffer mode
Select general register function
(1)
Select buffer mode
(2)
Start counting
(3)
Buffer mode
Figure 10.47 Procedure for Selecting the Buffer Mode
RENESAS 275
Buffer Mode Operation: Figure 10.48 shows an example of an operation in the buffer mode with
GRA set as an output compare register and GRA and buffer register A (BRA) set for buffer
operation. The TCNT operates as a periodic counter that is cleared by GRB compare match.
TIOCA and TIOCB are set to toggle at compare matches A and B. Since buffer mode is selected,
when TIOCA toggles at a compare match A, the BRA value is simultaneously transferred to the
GRA. This operation is repeated at every compare match A. The transfer timing is shown in figure
10.49.
Counter cleared by compare match B
TCNT value
GRB
H' 0250
H' 0200
H' 0100
H' 0000
Time
BRA
H' 0200
GRA
H' 0250
H' 0100
H' 0200
H' 0200
H' 0100
H' 0200
Toggle
output
TIOCB
TIOCA
Toggle
output
Compare match A
Figure 10.48 Buffer Mode Operation Example 1 (Output Compare Register)
276 RENESAS
CK
TCNT
n
n+1
Compare
match signal
Buffer
transfer signal
BR
GR
N
n
N
Figure 10.49 Compare Match Timing Example for Buffer Operation
RENESAS 277
Figure 10.50 shows an example of input capture operation in the buffer mode between GRA and
BRA with GRA as an input capture register. The TCNT is cleared by input capture B. The falling
edge is selected as the input capture edge at TIOCB. Both edges are selected as input capture
edges at TIOCA. When the TCNT value is stored in GRA by input capture A, the previous GRA
value is transferred to the BRA. The timing is shown in figure 10.51.
TCNT value
Counter cleared
at input capture B
H' 0180
H' 0160
H' 0005
Time
TIOCB
TIOCA
GRA
H' 0005
H' 0160
H' 0005
BRA
H' 0160
H' 0180
GRB
Input capture A
Figure 10.50 Buffer Mode Operation Example 2 (Input Capture Register)
278 RENESAS
CK
TIOC pin
Input
capture signal
n
TCNT
n+1
M
GR
BR
N
n
m
N+1
n
M
N
M
n
Figure 10.51 Input Capture Timing Example for Buffer Operation
An example of buffer operation in the complementary PWM mode between GRB3 and BRB3 is as
shown in figure 10.52. By making GRB3 larger than GRA3 using the buffer operation, a PWM
waveform with a duty cycle of 0% is generated. The transfer from BRB–GRB occurs upon
TCNT3 and GRA compare match and TCNT4 underflow.
TCNT3 and TCNT4 values
TCNT3
GRB3
H' 1FFF
GRA3
H' 0999
TCNT4
H' 0000
Time
BRB3
H' 0999
GRB3
H' 0999
H' 1FFF
H' 0999
H' 1FFF
H' 0999
H' 1FFF
H' 0999
TIOCA3
TIOCB3
Figure 10.52 Buffer Operation 4 (Complementary PWM Mode)
RENESAS 279
10.4.9
ITU Output Timing
ITU outputs in channels 3 and 4 can be inverted with the TOCR.
Output Inversion Timing with the TOCR: Output levels can be inverted by inverting the output
level select bits (OLS4 and OLS3) of the TOCR in the complementary PWM mode and resetsynchronized PWM mode. Figure 10.53 illustrates the timing.
T1
T2
T3
CK
Address
TOCR address
TOCR
ITU output pin
Inversion
Figure 10.53 Example of Inverting ITU Output Levels by Writing to TOCR
280 RENESAS
10.5
Interrupts
The ITU has two interrupt sources: input capture/compare match and overflow.
10.5.1
Timing of Setting Status Flags
Timing for Setting IMFA and IMFB in a Compare Match: The IMF bits of the TSR are set to
1 by a compare match signal generated when the TCNT matches a general register. The compare
match signal is generated in the last state in which the values match (when the TCNT is updated
from the matching count to the next count). Therefore, when the TCNT matches the GRA or GRB,
the compare match signal is not generated until the next timer clock input. Figure 10.54 shows the
timing of setting the IMF bits.
CK
TCNT
input clock
TCNT
GR
N
N+1
N
Compare
match signal
IMF
IMI
Figure 10.54 Timing of Setting Compare Match Flags (IMFA, IMFB)
RENESAS 281
Timing of Setting IMFA, IMFB for Input Capture: The IMFA and IMFB are set to 1 by an
input capture signal. At this time, the TCNT contents are transferred to the GR. Figure 10.55
shows the timing.
CK
Input
capture
signal
IMF
TCNT
GR
N
N
IMI
Figure 10.55 Timing of Setting IMFA and IMFB for Input Capture
Timing of Setting Overflow Flag (OVF): The OVF is set to 1 when the TCNT overflows from
H'FFFF–H'0000 or underflows from H'0000–H'FFFF. Figure 10.56 shows the timing.
282 RENESAS
CK
TCNT
H' FFFF
H' 0000
Overflow
signal
OVF
OVI
Figure 10.56 Timing of Setting OVF
10.5.2
Clear Timing of Status Flags
The status flags are cleared by a write cycle in which 1 is read on the CPU and then 0 is written to
it. This timing is shown in figure 10.57.
TSR write cycle
T1
T2
T3
CK
Address
TSR address
IMF, OVF
Figure 10.57 Timing of Status Flag Clearing
RENESAS 283
10.5.3
Interrupt Sources and Activating the DMAC
The ITU has compare match/input capture A interrupts, compare match/input capture B interrupts
and overflow interrupts for each channel. Each of the fifteen of these three types of interrupts are
allocated their own independently vectored addresses. When the interrupt’s interrupt request flag
is set to 1 and the interrupt enable bit is set to 1, the interrupt is requested.
The channel priority order can be changed with the interrupt controller. For more information, see
section 5, Interrupt Controller. The compare match/input capture A interrupts of channels 0–3 can
start the DMAC to transfer data. Table 10.17 lists the interrupt sources.
Table 10.17 ITU Interrupt Sources
Channel
Interrupt
Source
Description
DMAC
Activation
Priority
Order*
0
IMIA0
Compare match or input capture A0
Yes
High
IMIB0
Compare match or input capture B0
No
↑
OVI0
Overflow 0
No
IMIA1
Compare match or input capture A1
Yes
IMIB1
Compare match or input capture B1
No
OVI1
Overflow 1
No
IMIA2
Compare match or input capture A2
Yes
IMIB2
Compare match or input capture B2
No
OVI2
Overflow 2
No
IMIA3
Compare match or input capture A3
Yes
IMIB3
Compare match or input capture B3
No
OVI3
Overflow 3
No
IMIA4
Compare match or input capture A4
No
IMIB4
Compare match or input capture B4
No
↓
OVI4
Overflow 4
No
Low
1
2
3
4
Note: Indicates the initial status following reset. The ranking of channels can be altered using the
interrupt controller.
284 RENESAS
10.6
Notes and Precautions
This section describes contention and other matters requiring special attention during ITU
operations.
10.6.1
Contention between TCNT Write and Clear
If a counter clear signal occurs in the T3 state of a TCNT write cycle, clearing the counter takes
priority and the write is not performed. The timing is shown in figure 10.58.
TCNT write cycle by CPU
T1
T2
T3
CK
Address
TCNT address
Internal write signal
Counter clear signal
TCNT
N
H' 0000
Figure 10.58 Contention between TCNT Write and Clear
RENESAS 285
10.6.2
Contention between TCNT Word Write and Increment
If an increment pulse occurs in the T3 state of a TCNT word write cycle, writing takes priority and
the TCNT is not incremented. The timing is shown in figure 10.59.
TCNT word write cycle by CPU
T1
T2
T3
CK
Address
TCNT address
Internal write signal
TCNT input clock
TCNT
N
M
TCNT write data
Figure 10.59 Contention between TCNT Word Write and Increment
286 RENESAS
10.6.3
Contention between TCNT Byte Write and Increment
If an increment pulse occurs in the T2 state or T3 state of a TCNT byte write cycle, counter
writing takes priority and the byte data on the side that was previously written is not incremented.
The TCNT byte data that was not written is also not incremented and retains its previous value.
The timing is shown in figure 10.60 (which shows an increment during state T2 of a byte write
cycle to TCNTH).
TCNTH byte write cycle by CPU
T1
T2
T3
CK
TCNTH address
Address
Internal write signal
TCNT input clock
N
TCNTH
M
TCNT write data
TCNTL
X
X+1
X
Figure 10.60 Contention between TCNT Byte Write and Increment
RENESAS 287
10.6.4
Contention between GR Write and Compare Match
If a compare match occurs in the T3 state of a general register (GR) write cycle, writing takes
priority and the compare match signal is inhibited. The timing is shown in figure 10.61.
GR write cycle
T1
T2
T3
CK
Address
GR address
Internal
write signal
TCNT
N
N+1
GR
N
M
GR write data
Compare
match signal
Inhibited
Figure 10.61 Contention between General Register Write and Compare Match
288 RENESAS
10.6.5
Contention between TCNT Write and Overflow/Underflow
If an overflow occurs in the T3 state of a TCNT write cycle, writing takes priority over counter
incrementing. The OVF is set to 1. The same applies to underflows. This timing is shown in figure
10.62.
TCNT write cycle
T1
T2
T3
CK
Address
TCNT address
Internal
write signal
TCNT
input clock
Overflow
signal
TCNT
H'FFFF
M
TCNT write data
OVF
Figure 10.62 Contention between TCNT Write and Overflow
RENESAS 289
10.6.6
Contention between General Register Read and Input Capture
If an input capture signal is generated during the T3 state of a general register read cycle, the value
before input capture is read. The timing is shown in figure 10.63.
GR read cycle
T1
T2
T3
CK
Address
GR address
Internal read
signal
Input capture
signal
GR
Internal
data bus
X
M
X
Figure 10.63 Contention between General Register Read and Input Capture
290 RENESAS
10.6.7
Contention Between Counter Clearing by Input Capture and Counter Increment
If an input capture signal and counter increment signal occur simultaneously, the counter is
cleared according to the input capture signal. The counter is not incremented by the increment
signal. The TCNT value before the counter is cleared is transferred to the general register. The
timing is shown in figure 10.64.
CK
Input capture
signal
Counter
clear signal
TCNT
input clock
TCNT
GR
N
H'0000
N
Figure 10.64 Contention between Counter Clearing by Input Capture and Counter
Increment
RENESAS 291
10.6.8
Contention between General Register Write and Input Capture
If an input capture signal is generated during the T3 state of a general register write cycle, the
input capture transfer takes priority and the write to the GR is not performed. The timing is shown
in figure 10.65.
GR write cycle
T1
T2
T3
CK
Address
GR address
Internal
write signal
Input capture
signal
TCNT
GR
M
M
Figure 10.65 Contention between General Register Write and Input Capture
10.6.9
Note on Waveform Cycle Setting
When a counter is cleared by compare match, the counter is cleared in the last state in which the
TCNT value matches the GR value (when the TCNT is updated from the matching count to the
next count). The actual counter frequency is therefore given by the following formula:
f = φ/(N + 1)
(f: counter frequency. φ: operating frequency. N: value set in the GR.)
292 RENESAS
10.6.10
Contention Between BR Write and Input Capture
When a buffer register (BR) is being used as an input capture register and an input capture signal
is generated in the T3 state of the write cycle, the buffer operation takes priority over the BR write.
The timing is shown in figure 10.66.
BR write cycle
T1
T2
T3
CK
Address
BR address
Internal
write signal
Input capture
signal
GR
N
X
TCNT value
BR
M
N
Figure 10.66 Contention between BR Write and Input Capture
RENESAS 293
10.6.11
Note on Writing in the Synchronizing Mode
After the synchronizing mode is selected, if the TCNT is written by byte access, all 16 bits of all
synchronized counters assume the same value as the counter that was addressed.
Example: Figures 10.67 and 10.68 show byte write and word write when channels 2 and 3 are
synchronized
TCNT2
W
X
TCNT3
Y
Z
Upper
byte
Lower
byte
Figure 10.67
W
X
TCNT3
Y
Z
Upper
byte
Lower
byte
10.6.12
TCNT2
A
X
TCNT3
A
X
Upper
byte
Lower
byte
TCNT2
Y
A
TCNT3
Y
A
Upper
byte
Lower
byte
Write A to lower
byte of channel 3
Byte Write to Channel 2 or Byte Write to Channel 3
TCNT2
Figure 10.68
Write A to upper
byte of channel 2
Word write of AB
for channel 2 or 3
TCNT2
A
B
TCNT3
A
B
Upper
byte
Lower
byte
Word Write to Channel 2 or Word Write to Channel 3
Note on Setting Reset-synchronized PWM Mode/Complementary PWM Mode
When the CMD1 and CMD0 bits of TFCR are set, note the following.
1. Writes to CMD1 and CMD0 should be done while TCNT3 and TCNT4 are halted.
2. Changes of setting from the reset-synchronized PWM mode to the complementary PWM mode
and vice versa are inhibited. Set the reset-synchronized PWM mode or complementary PWM
mode after first setting normal operation (clear CMD1 bit to 0).
294 RENESAS
10.6.13
Clearing the Complementary PWM Mode
Figure 10.69 shows the procedure for clearing the complementary PWM mode. First, reset the
combination mode bits CMD1 and CMD0 in the timer function control register (TFCR) from 10 to
either 00 or 01. The mode will switch from complementary PWM mode to normal operating
mode. Next, wait for at least 1 clock of the counter input clock being used for channels 3 and 4
and then clear the counter start bits STR3 and STR4 of the timer start register (TSTR). The
channels 3 and 4 counters TCNT3 and TCNT4 will stop counting. Clearing the complementary
PWM mode by any other procedure may result in changes other than those set for the output
waveform when complementary PWM mode is set again.
Complementary PWM mode
Clear complementary
PWM mode
Halt Count
Normal operation
1. Clear the CMD1 bit of the TFCR to 0
to set channels 3 and 4 for normal operation
2. Wait at least 1 clock after setting channels 3 and 4
for normal operation and then clear the STR3 and
STR4 bits of the TSTR to 0 to halt the TCNT3 and
TCNT4 counters
Figure 10.69 Clearing the Complementary PWM Mode
10.6.14
ITU Operating Modes
Tables 10.18–10.22 show the ITU operating modes for channels 0–4.
10.6.15
Note on Counter Clearing by Input Capture
If TCNT is cleared (to H'0000) by input capture when its value is H'FFFF, overflow will not
occur.
RENESAS 295
Table 10.18 ITU Operating Modes (Channel 0)
Register Setting
TSNC
Operating
Mode
Sync
TMDR
MDF FDIR PWM
TFCR
TOCR
TIOR0
Reset
Output
Comp Sync Buf- Level
PWM PWM fer Select IOA
TCR0
IOB
Clear Clock
Select Select
Synchronized
preset
SYNC0 —
=1
—
√
—
—
—
—
√
√
√
√
PWM
√
—
—
PWM0 —
=1
—
—
—
—
√
√
√
Output
√
compare A
function
—
—
PWM0 —
=0
—
—
—
IOA2 = √
0, others
free
√
√
Output
√
compare B
function
—
—
√
—
—
—
—
√
IOB2 = √
0, others
free
√
Input
√
capture A
function
—
—
PWM0 —
=0
—
—
—
IOA2 = √
1, others
free
Input
√
capture B
function
—
—
PWM0 —
=0
—
—
—
√
√
√
IOB2 = √
1, others
free
√
Counter Clear Function
√
Clear at
compare
match/
input
capture A
—
—
√
—
—
—
—
√
√
CCLR1 √
=0
CCLR0
=1
√
Clear at
compare
match/
input
capture B
—
—
√
—
—
—
—
√
√
CCLR1 √
=1
CCLR0
=0
SYNC0 —
=1
—
√
—
—
—
—
√
√
CCLR1 √
=1
CCLR0
=1
Synchronized
clear
√: Settable, —: Setting does not affect current mode
Note: In PWM mode, the input capture function cannot be used. When compare match A and
compare match B occur simultaneously, the compare match signal is inhibited.
296 RENESAS
Table 10.19 ITU Operating Modes (Channel 1)
Register Setting
TSNC
Operating
Mode
Sync
TMDR
MDF FDIR PWM
TFCR
TOCR
TIOR1
Reset
Output
Comp Sync Buf- Level
PWM PWM fer Select IOA
TCR1
IOB
Clear Clock
Select Select
Synchronized
preset
SYNC1 —
=1
—
√
—
—
—
—
√
√
√
√
PWM
√
—
—
PWM1 —
=1
—
—
—
—
√*1
√
√
Output
√
compare A
function
—
—
PWM1 —
=0
—
—
—
IOA2 = √
0, others
free
√
√
Output
√
compare B
function
—
—
√
—
—
—
—
√
IOB2 = √
0, others
free
√
Input
√
capture A
function
—
—
PWM1 —
=0
—
—
—
IOA2 = √
1, others
free
Input
√
capture B
function
—
—
PWM1 —
=0
—
—
—
√
√
√
IOB2 = √
1, others
free
√
Counter Clear Function
√
Clear at
compare
match/
input
capture A
—
—
√
—
—
—
—
√
√
CCLR1 √
=0
CCLR0
=1
√
Clear at
compare
match/
input
capture B
—
—
√
—
—
—
—
√
√
CCLR1 √
=1
CCLR0
=0
SYNC1 —
=1
—
√
—
—
—
—
√
√
CCLR1 √
=1
CCLR0
=1
Synchronized
clear
√: Settable, —: Setting does not affect current mode
Note: In PWM mode, the input capture function cannot be used. When compare match A and
compare match B occur simultaneously, the compare match signal is inhibited.
RENESAS 297
Table 10.20 ITU Operating Modes (Channel 2)
Register Setting
TSNC
Operating
Mode
Sync
TMDR
MDF FDIR PWM
TFCR
TOCR
TIOR2
Reset
Output
Comp Sync Buf- Level
PWM PWM fer Select IOA
TCR2
IOB
Clear Clock
Select Select
Synchronized
preset
SYNC2 —
=1
—
√
—
—
—
—
√
√
√
√
PWM
√
—
—
PWM2 —
=1
—
—
—
—
√
√
√
Output
√
compare A
function
—
—
PWM2 —
=0
—
—
—
IOA2 = √
0, others
free
√
√
Output
√
compare B
function
—
—
√
—
—
—
—
√
IOB2 = √
0, others
free
√
Input
√
capture A
function
—
—
PWM2 —
=0
—
—
—
IOA2 = √
1, others
free
Input
√
capture B
function
—
—
PWM2 —
=0
—
—
—
√
√
√
IOB2 = √
1, others
free
√
Counter Clear Function
√
Clear at
compare
match/
input
capture A
—
—
√
—
—
—
—
√
√
CCLR1 √
=0
CCLR0
=1
√
Clear at
compare
match/
input
capture B
—
—
√
—
—
—
—
√
√
CCLR1 √
=1
CCLR0
=0
Synchronized
clear
SYNC2 —
=1
—
√
—
—
—
—
√
√
CCLR1 √
=1
CCLR0
=1
Phase
counting
√
√
—
—
—
—
√
√
√
MDF √
=1
√: Settable, —: Setting does not affect current mode
298 RENESAS
—
Note: In PWM mode, the input capture function cannot be used. When compare match A and
compare match B occur simultaneously, the compare match signal is inhibited.
Table 10.21 ITU Operating Modes (Channel 3)
Register Setting
TSNC
Operating
Mode
Sync
TMDR
MDF FDIR PWM
TFCR
TOCR
TIOR3
TCR3
Reset
Output
Comp Sync Buf- Level
PWM PWM fer Select IOA
IOB
Clear Clock
Select Select
√*2
Synchronized
preset
SYNC3 —
=1
—
√
√
√
—
√
√
√
√
PWM
mode
√
—
—
PWM3 CMD1 CMD1 √
=1
=0
=0
—
—
√*1
√
√
Output
√
compare A
function
—
—
PWM3 CMD1 CMD1 √
=0
=0
=0
—
IOA2 = √
0, others
free
√
√
Output
√
compare B
function
—
—
√
CMD1 CMD1 √
=0
=0
—
√
IOB2 = √
0, others free
√
Input
√
capture A
function
—
—
PWM3 CMD1 CMD1 √
=0
=0
=0
—
IOA2 = √
1, others
free
Input
√
capture B
function
—
—
PWM3 CMD1 CMD1 √
=0
=0
=0
—
√
√
√
IOB2 = √
1, others free
√
Counter Clear Function
√
Clear at
compare
match/
input
capture A
—
—
√
CMD1 √*3
= 1,
CMD0
=0
inhibited
√
—
√
√
CCLR1 √
=0
CCLR0
=1
√
Clear at
compare
match/
input
capture B
—
—
√
CMD1 CMD1 √
=0
=0
—
√
√
CCLR1 √
=1
CCLR0
=0
SYNC3 —
=1
—
√
CMD1 √
= 1,
CMD0
=0
inhibited
—
√
√
CCLR1 √
=1
CCLR0
=1
Synchronized
clear
√
RENESAS 299
Table 10.21 ITU Operating Modes (Channel 3) (cont)
Counter Clear Function
Register Setting
TSNC
Operating
Mode
Sync
TMDR
TFCR
TOCR
Reset
Comp Sync BufMDF FDIR PWM PWM PWM fer
TIOR3
Output
Level
Select IOA
IOB
TCR3
Clear Clock
Select Select
—
—
—
CMD1 CMD1 √
=1
=1
CMD0 CMD0
=0
=0
√
—
—
CCLR1 √*4
=0
CCLR0
=0
√
Reset
synchronized PWM
mode
—
—
—
CMD1 CMD1 √
=1
=1
CMD0 CMD0
=1
=1
√
—
—
CCLR1 √
=0
CCLR0
=1
Buffer
(BRA)
√
—
—
√
√
√
BFA3 —
= 1,
others
free
√
√
√
√
Buffer
(BRB)
√
—
—
√
√
√
BFB3 —
= 1,
others
free
√
√
√
√
Complementary
PWM
mode
√*2
√: Settable, —: Setting does not affect current mode
Notes: 1. In PWM mode, the input capture function cannot be used. When compare match A and
compare match B occur simultaneously, the compare match signal is inhibited.
2. When set for complementary PWM mode, do not simultaneously set channel 3 and
channel 4 to function synchronously.
3. Counter clearing by input capture A cannot be used when the reset-synchronized PWM
mode is set.
4. Clock selection when the complementary PWM mode is set should be the same for
channels 3 and 4.
300 RENESAS
Table 10.22 ITU Operating Modes (Channel 4)
Register Setting
TSNC
Operating
Mode
Sync
TMDR
MDF FDIR PWM
TFCR
TOCR
TIOR4
TCR4
Reset
Output
Comp Sync Buf- Level
PWM PWM fer Select IOA
IOB
Clear Clock
Select Select
√*2
Synchronized
preset
SYNC4 —
=1
—
√
√
√
—
√
√
√
√
PWM
√
—
—
PWM4 CMD1 CMD1 √
=1
=0
=0
—
—
√*1
√
√
Output
√
compare A
function
—
—
PWM4 CMD1 CMD1 √
=0
=0
=0
—
IOA2 = √
0, others
free
√
√
Output
√
compare B
function
—
—
√
CMD1 CMD1 √
=0
=0
—
√
IOB2 = √
0, others
free
√
Input
√
capture A
function
—
—
PWM4 CMD1 CMD1 √
=0
=0
=0
—
IOA2 = √
1, others
free
Input
√
capture B
function
—
—
PWM4 CMD1 CMD1 √
=0
=0
=0
—
√
√
√
IOB2 = √
1, others
free
√
Counter Clear Function
√
Clear at
compare
match/
input
capture A
—
—
√
CMD1 √*3
= 1,
CMD0
=0
inhibit
ed
√
—
√
√
CCLR1 √
=0
CCLR0
=1
√
Clear at
compare
match/
input
capture B
—
—
√
CMD1 √*3
= 1,
CMD0
=0
inhibit
ed
√
—
√
√
CCLR1 √
=1
CCLR0
=0
RENESAS 301
Table 10.22 ITU Operating Modes (Channel 4) (cont)
Counter Clear Function
Register Setting
TSNC
Operating
Mode
Sync
TMDR
TFCR
TOCR
Reset
Comp Sync BufMDF FDIR PWM PWM PWM fer
Synchronized
clear
SYNC4 —
=1
—
√
CMD1 √*3
= 1,
CMD1
=0
inhibit
ed
Complementary
PWM
√*2
—
—
—
Reset
√
synchronized PWM
—
—
Buffer
(BRA)
√
—
Buffer
(BRB)
√
—
TIOR4
Output
Level
Select IOA
IOB
TCR4
Clear Clock
Select Select
—
√
√
CCLR1 √
=1
CCLR0
=1
CMD1 CMD1 √
=1
=1
CMD0 CMD0
=0
=0
√
—
—
CCLR1 √*4
=0
CCLR0
=0
—
CMD1 CMD1 √
=1
=1
CMD0 CMD0
=1
=1
√
—
—
√*5
√*5
—
√
√
√
BFA4 —
= 1,
others
free
√
√
√
√
—
√
√
√
BFB4 —
= 1,
others
free
√
√
√
√
√
√: Settable, —: Setting does not affect current mode
Notes: 1. In PWM mode, the input capture function cannot be used. When compare match A and
compare match B occur simultaneously, the compare match signal is inhibited.
2. When set for complementary PWM mode, do not simultaneously set channel 3 and
channel 4 to function synchronously.
3. Counter clearing works with the reset-synchronized PWM mode, but TCNT4 runs
independently. The output waveform is not affected.
4. Clock selection when the complementary PWM mode is set should be the same for
channels 3 and 4.
5. In the reset-synchronized PWM mode, TCNT4 runs independently. The output
waveform is not affected.
302 RENESAS
Section 11 Programmable Timing Pattern Controller (TPC)
11.1
Overview
The SuperH microcomputer has a built-in programmable timing pattern controller (TPC). The
TPC can provide pulse outputs by using the 16-bit integrated-timer pulse unit (ITU) as a time
base. The TPC pulse outputs are divided into 4-bit groups 3–0. These can operate simultaneously,
or independently.
11.1.1
Features
Features of the programmable timing pattern controller are listed below.
•
•
•
•
•
•
16-bit output data: Maximum 16-bit data can be output. TPC output can be enabled on a bitby-bit basis.
Four output groups: Output trigger signals can be selected in 4-bit groups to provide up to four
different 4-bit outputs.
Selectable output trigger signals: Output trigger signals can be selected by group from the
4-channel compare-match signals of the 16-bit integrated-timer pulse unit (ITU).
Non-overlap mode: A non-overlap interval can be set to come between multiple pulse outputs.
Can connect to DMA controller: The compare-match signals selected as output trigger signals
can activate the DMA controller for sequential output of data without CPU intervention.
RENESAS 303
11.1.2
Block Diagram
Figure 11.1 is the block diagram of the TPC.
ITU compare match signal
Control logic
TP15
TP14
TP13
TP12
TP11
TP10
TP9
TP8
TP7
TP6
TP5
TP4
TP3
TP2
TP1
TP0
PBCR1
PBCR2
NDERA
NDERB
TPMR
TPCR
Pulse output
pin group 3
NDRB
Pulse output
pin group 2
PBDR
Pulse output
pin group 1
NDRA
Pulse output
pin group 0
TPC
TPMR: TPC output mode register
PBCR1: Port B control register 1
TPCR: TPC output control register
PBCR2: Port B control register 2
NDERB: Next data enable register B
NDRB: Next data register B
NDERA: Next data enable register A
NDRA: Next data register A
PBDR: Port B data register
Figure 11.1 TPC Block Diagram
304 RENESAS
Internal
data
bus
11.1.3
Input/Output Pins
Table 11.1 summarizes the TPC input/output pins.
Table 11.1 TPC Pins
Name
Symbol
Input/Output
Function
TPC output 0
TP0
Output
Group 0 pulse output
TPC output 1
TP1
Output
TPC output 2
TP2
Output
TPC output 3
TP3
Output
TPC output 4
TP4
Output
TPC output 5
TP5
Output
TPC output 6
TP6
Output
TPC output 7
TP7
Output
TPC output 8
TP8
Output
TPC output 9
TP9
Output
TPC output 10
TP10
Output
TPC output 11
TP11
Output
TPC output 12
TP12
Output
TPC output 13
TP13
Output
TPC output 14
TP14
Output
TPC output 15
TP15
Output
Group 1 pulse output
Group 2 pulse output
Group 3 pulse output
RENESAS 305
11.1.4
Registers
Table 11.2 summarizes the TPC registers.
Table 11.2 TPC Registers
Name
Abbreviation
R/W
Initial
Value
Address* 1
Access
Size`
Port B control register 1
PBCR1
R/W
H'0000
H'5FFFFCC
8, 16
Port B control register 2
PBCR2
R/W
H'0000
H'5FFFFCE
8, 16
Port B data register
PBDR
R/(W)*2
H'0000
H'5FFFFC2
8, 16
TPC output mode register
TPMR
R/W
H'F0
H'5FFFFF0
8, 16
TPC output control register
TPCR
R/W
H'FF
H'5FFFFF1
8, 16
Next data enable register B
NDERB
R/W
H'00
H'5FFFFF2
8, 16
Next data enable register A
NDERA
R/W
H'00
H'5FFFFF3
8, 16
Next data register A
NDRA
R/W
H'00
H'5FFFFF5/
H'5FFFFF7*3
8, 16
Next data register B
NDRB
R/W
H'00
H'5FFFFF4/
H'5FFFFF6*3
8, 16
Notes: 1. Only the values of bits A27–A24 and A8–A0 are valid; bits A23–A9 are ignored. For
details on the register addresses, see section 8.3.5, Description of Areas.
2. Bits used for TPC output cannot be written to.
3. These addresses change depending on TPCR settings. When TPC output groups 0
and 1 have the same output trigger, the NDRA address is H'5FFFFF5; when their
output triggers are different, the NDRA address for group 0 is H'5FFFFF7 and the
address for group 1 is H'5FFFFF5. Likewise, when TPC output groups 2 and 3 have the
same output trigger, the NDRB address is H'5FFFFF4; when their output triggers are
different, the NDRB address for group 0 is H'5FFFFF6 and the address for group 1 is
H'5FFFFF4.
11.2
Register Descriptions
11.2.1
Port B Control Registers 1 and 2 (PBCR1, PCBR2)
The port B control register 1 and 2 (PBCR1 and PBCR2) are 16-bit read/write registers that set the
functions of port B pins. Port B consists of the dual use pins TP15–TP0. Bits corresponding to the
pins to be used for TPC output must be set to 1. For details, see the port B description in the
section 15, Pin Function Controller.
306 RENESAS
PCBR1:
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
15
14
13
12
11
10
9
8
PB15
MD1
PB15
MD0
PB14
MD1
PB14
MD0
PB13
MD1
PB13
MD0
PB12
MD1
PB12
MD0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
PB11
MD1
PB11
MD0
PB10
MD1
PB10
MD0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
15
14
13
12
11
10
9
8
PB9MD1 PB9MD0 PB8MD1 PB8MD0
PCBR2:
Bit:
Bit name: PB7MD1 PB7MD0 PB6MD1 PB6MD0 PB5MD1 PB5MD0 PB4MD1 PB4MD0
Initial value:
R/W:
Bit:
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
Bit name: PB3MD1 PB3MD0 PB2MD1 PB2MD0 PB1MD1 PB1MD0 PB0MD1 PB0MD0
Initial value:
R/W:
11.2.2
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Port B Data Register (PBDR)
The port B data register is a 16-bit read/write register that, when used for TPC output stores,
output data for groups 0–3. For details, see the port B description in section 16, I/O Ports.
Bit:
15
14
13
12
11
10
9
8
Bit name: PB15DR PB14DR PB13DR PB12DR PB11DR PB10DR PB9DR PB8DR
Initial value:
R/W:
0
0
0
0
0
0
0
0
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
Note: Bits set to TPC output by NDERA or NDERB are read-only.
RENESAS 307
Bit:
Bit name:
Initial value:
R/W:
7
PB7DR
6
5
PB6DR PB5DR
4
3
PB4DR PB3DR
2
PB2DR
1
0
PB1DR PB0DR
0
0
0
0
0
0
0
0
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
Note: Bits set to TPC output by NDERA or NDERB are read-only.
11.2.3
Next Data Register A (NDRA)
NDRA is an eight-bit read/write register that stores the next output data for TPC output groups 1
and 0 (TP7–TP0). When used for TPC output, the contents of the NDRA are transferred to the
corresponding PBDR bits when the ITU compare match specified in the TPC output control
register TPCR occurs.
The address of the NDRA differs depending on whether TPCR settings select the same trigger or
different triggers for TPC output groups 1 and 0. When reset, NDRA is initialized to H'00. It is not
initialized by standby mode.
Same Trigger for TPC Output Groups 1 and 0: If TPC output groups 1 and 0 are triggered by
the same compare match, the address of the NDRA is H'FFFFF5. The 4 upper bits becomes group
1 and the 4 lower bits become group 0. Address H'5FFFFF7 in such cases consists entirely of
reserved bits. These bits cannot be modified and always read as 1.
Address H'5FFFFF5:
•
Bits 7–4 (next data 7–4 (NDR7–NDR4)): NDR7-NDR4 store the next output data for TPC
output group 1.
•
Bits 3–0 (next data 3–0 (NDR3–NDR0)): NDR3-NDR0 store the next output data for TPC
output group 0.
Bit:
Bit name:
Initial value:
R/W:
308 RENESAS
7
6
5
4
3
2
1
0
NDR7
NDR6
NDR5
NDR4
NDR3
NDR2
NDR1
NDR0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Address H'5FFFFF7:
•
Bits 7–0 (reserved): These bits always read as 1. The write value should always be 1.
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
—
—
—
Initial value:
1
1
1
1
1
1
1
1
R/W:
—
—
—
—
—
—
—
—
Different Triggers for TPC Output Groups 1 and 0: If TPC output groups 1 and 0 are triggered by
different compare matches, the address of the upper 4 bits of NDRA (group 1) is H'5FFFFF5 and
the address of the lower 4 bits of NDRA (group 0) is H'5FFFFF7. Bits 3–0 of address H'5FFFFF5
and bits 7–4 of address H'5FFFFF7 are reserved bits. The write value should always be 1. These
bits always read as 1.
Address H'5FFFFF5:
•
Bits 7–4 (next data 7–4 (NDR7–NDR4)): NDR7–NDR4 store the next output data for TPC
output group 1.
•
Bits 3–0 (reserved): These bits always read as 1. The write value should always be 1.
Bit:
Bit name:
Initial value:
R/W:
7
6
5
4
3
2
1
0
NDR7
NDR6
NDR5
NDR4
—
—
—
—
0
0
0
0
1
1
1
1
R/W
R/W
R/W
R/W
—
—
—
—
Address H'5FFFFF7:
•
Bits 7–4 (reserved): These bits always read as 1. The write value should always be 1.
•
Bits 3–0 (next data 3–0 (NDR3–NDR0)): NDR3-NDR0 store the next output data for TPC
output group 0.
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
NDR3
NDR2
NDR1
NDR0
Initial value:
1
1
1
1
0
0
0
0
R/W:
—
—
—
—
R/W
R/W
R/W
R/W
RENESAS 309
11.2.4
Next Data Register B (NDRB)
NDRB is an eight-bit read/write register that stores the next output data for TPC output groups 3
and 2 (TP15–TP8). When used for TPC output, the contents of the NDRB are transferred to the
corresponding PBDR bits when the ITU compare match specified in the TPC output control
register TPCR occurs.
The address of the NDRB differs depending on whether TPCR settings select the same trigger or
different triggers for TPC output groups 3 and 2. When reset, NDRB is initialized to H'00. It is not
initialized by standby mode.
Same Trigger for TPC Output Groups 3 and 2: If TPC output groups 3 and 2 are triggered by the
same compare match, the address of the NDRB is H'FFFFF4. The 4 upper bits becomes group 3
and the 4 lower bits become group 2. Address H'5FFFFF6 becomes completely reserved bits.
These bits always read as 1, and the write value should always be 1.
Address H'5FFFFF4:
•
Bits 7–4 (next data 15–12 (NDR15–NDR12)): NDR15–NDR12 store next output data for TPC
output group 3.
•
Bits 3–0 (next data 11–8 (NDR11–NDR8)): NDR11–NDR8 store next output data for TPC
output group 2.
Bit:
Bit name:
Initial value:
R/W:
7
6
5
4
3
2
1
0
NDR15
NDR14
NDR13
NDR12
NDR11
NDR10
NDR9
NDR8
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Address H'5FFFFF6:
•
Bits 7–0 (reserved): These bits always read as 1. The write value should always be 1.
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
—
—
—
Initial value:
1
1
1
1
1
1
1
1
R/W:
—
—
—
—
—
—
—
—
310 RENESAS
Different Triggers for TPC Output Groups 3 and 2: If TPC output groups 3 and 2 are triggered
by different compare matches, the address of the upper 4 bits of NDRB (group 3) is H'5FFFFF4
and the address of the lower 4 bits of NDRB (group 2) is H'5FFFFF6. Bits 3-0 of address
H'5FFFFF4 and bits 7–4 of address H'5FFFFF6 are reserved bits. These bits always read as 1. The
write value should always be 1.
Address H'5FFFFF4:
•
Bits 7–4 (next data 15–12 (NDR15–NDR12)): NDR15–NDR12 store next output data for TPC
output group 3.
•
Bits 3–0 (reserved): These bits always read as 1. The write value should always be 1.
Bit:
Bit name:
Initial value:
R/W:
7
6
5
4
3
2
1
0
NDR15
NDR14
NDR13
NDR12
—
—
—
—
0
0
0
0
1
1
1
1
R/W
R/W
R/W
R/W
—
—
—
—
Address H'5FFFFF6:
•
Bits 7–4 (reserved): These bits always read as 1. The write value should always be 1.
•
Bits 3–0 (next data 11–8 (NDR11–NDR8)): NDR11–NDR8 store next output data for TPC
output group 2.
11.2.5
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
NDR11
NDR10
NDR9
NDR8
Initial value:
1
1
1
1
0
0
0
0
R/W:
—
—
—
—
R/W
R/W
R/W
R/W
Next Data Enable Register A (NDERA)
NDERA is an eight-bit read/write register that enables TPC output groups 1 and 0 (TP7–TP0) on a
bit-by-bit basis.
When the bits enabled for TPC output by NDERA generate the ITU compare match selected in the
TPC output control register, the value of the next data register A (NDRA) is automatically
transferred to the corresponding PBDR bits and the output value is updated. For disabled bits,
there is no transfer and the output value does not change. When reset, NDERA is initialized to
H'00. It is not initialized by standby mode.
RENESAS 311
Bit:
Bit name:
Initial value:
R/W:
•
7
6
NDER7
5
NDER6 NDER5
4
3
NDER4 NDER3
2
NDER2
1
0
NDER1 NDER0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bits 7–0 (next data enable 7–0 (NDER7–NDER0)): NDER7–NDER0 select enable/disable for
TPC output groups 1 and 0 (TP7–TP0) in bit units.
Bit 7–0: NDER7–NDER0 Description
0
Disables TPC outputs TP7–TP0 (transfer from NDR7–NDR0 to PB7–
PB0 is disabled) (initial value)
1
Enables TPC outputs TP7–TP0 (transfer from NDR7–NDR0 to PB7–
PB0 is enabled)
11.2.6
Next Data Enable Register B (NDERB)
NDERB is an eight-bit read/write register that enables TPC output groups 3 and 2 (TP15–TP8) on
a bit-by-bit basis.
When the bits enabled for TPC output by NDERB generate the ITU compare match selected in the
TPC output control register, the value of the next data register B (NDRB) is automatically
transferred to the corresponding PBDR bits and the output value is updated. For disabled bits,
there is no transfer and the output value does not change. When reset, NDERB is initialized to
H'00. It is not initialized by standby mode.
Bit:
7
6
5
4
3
2
1
0
Bit name: NDER15 NDER14 NDER13 NDER12 NDER11 NDER10 NDER9 NDER8
Initial value:
R/W:
•
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bits 7–0 (next data enable 15–8 (NDER15–NDER8)): NDER15–NDER8 select enable/disable
for TPC output groups 3 and 2 (TP15–TP8) in bit units.
Bit 7–0:
NDER15–NDER8
Description
0
Disables TPC outputs TP15–TP8 (transfer from NDR15–NDR8 to
PB15–PB8 is disabled) (initial value)
1
Enables TPC outputs TP15–TP8 (transfer from NDR15–NDR8 to
PB15–PB8 is enabled)
312 RENESAS
11.2.7
TPC Output Control Register (TPCR)
TPCR is an eight-bit read/write register that selects output trigger signals for TPC outputs. When
reset, TPCR is initialized to H'FF. It is not initialized by standby mode.
Bit:
7
6
5
4
3
2
1
0
Bit name: G3CMS1 G3CMS0 G2CMS1 G2CMS0 G1CMS1 G1CMS0 G0CMS1 G0CMS0
Initial value:
R/W:
•
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bits 7 and 6 (group 3 compare-match select 1 and 0 (G3CMS1 and G3CMS0)): G3CMS1 and
G3CMS0 select the compare match that triggers TPC output group 3 (TP15–TP12).
Bit 7: G3CMS1
Bit 6: G3CMS0
Description
0
0
TPC output group 3 (TP15–TP12) output is triggered by
compare-match in ITU channel 0
1
TPC output group 3 (TP15–TP12) output is triggered by
compare-match in ITU channel 1
0
TPC output group 3 (TP15–TP12) output is triggered by
compare-match in ITU channel 2
1
TPC output group 3 (TP15–TP12) output is triggered by
compare-match in ITU channel 3 (initial value)
1
•
Bits 5 and 4 (group 2 compare-match select 1 and 0 (G2CMS1 and G2CMS0)): G2CMS1 and
G2CMS0 select the ITU channel that triggers TPC output group 2 (TP11–TP8).
Bit 5: G2CMS1
Bit 4: G2CMS0
Description
0
0
TPC output group 2 (TP11–TP18) output is triggered by
compare-match in ITU channel 0
1
TPC output group 2 (TP11–TP18) output is triggered by
compare-match in ITU channel 1
0
TPC output group 2 (TP11–TP18) output is triggered by
compare-match in ITU channel 2
1
TPC output group 2 (TP11–TP18) output is triggered by
compare-match in ITU channel 3 (initial value)
1
RENESAS 313
Bits 3 and 2 (group 1 compare-match select 1 and 0 (G1CMS1 and G1CMS0)): G1CMS1 and
G1CMS0 select the ITU channel that triggers TPC output group 1 (TP7–TP4).
Bit 3: G1CMS1
Bit 2: G1CMS0
Description
0
0
TPC output group 1 (TP7–TP4) output is triggered by
compare-match in ITU channel 0
1
TPC output group 1 (TP7–TP4) output is triggered by
compare-match in ITU channel 1
0
TPC output group 1 (TP7–TP4) output is triggered by
compare-match in ITU channel 2
1
TPC output group 1 (TP7–TP4) output is triggered by
compare-match in ITU channel 3 (initial value)
1
•
Bits 1 and 0 (group 0 compare-match select 1 and 0 (G0CMS1 and G0CMS0)): G0CMS1 and
G0CMS0 select the ITU channel that triggers TPC output group 0 (TP3–TP0).
Bit 1: G0CMS1
Bit 0: G0CMS0
Description
0
0
TPC output group 0 (TP3–TP0) output is triggered by
compare-match in ITU channel 0
1
TPC output group 0 (TP3–TP0) output is triggered by
compare-match in ITU channel 1
0
TPC output group 0 (TP3–TP0) output is triggered by
compare-match in ITU channel 2
1
TPC output group 0 (TP3–TP0) output is triggered by
compare-match in ITU channel 3 (initial value)
1
11.2.8
TPC Output Mode Register (TPMR)
TPMR is an eight-bit read/write register that selects between the TPC's ordinary output and nonoverlap output modes in group units. During non-overlap operation, the output waveform cycle is
set in ITU general register B (GRB) for use as the output trigger and a non-overlap period is set in
general register A (GRA). The output value then changes on compare matches A and B. For
details, see section 11.3.4, TPC Output Non-Overlap Operation. TPMR is initialized to H'F0 on a
reset. It is not initialized in standby mode.
Bit:
7
6
5
4
Bit name:
—
—
—
—
Initial value:
1
1
1
1
0
0
0
0
R/W:
—
—
—
—
R/W
R/W
R/W
R/W
314 RENESAS
3
2
1
0
G3NOV G2NOV G1NOV G0NOV
Bits 7–4 (reserved): These bits always read as 1. The write value should always be 1.
•
Bit 3 (group 3 non-overlap mode (G3NOV)): G3NOV selects the ordinary or non-overlap
mode for TPC output group 3 (TP15–TP12).
Bit 3: G3NOV
Description
0
TPC output group 3 operates normally (output value updated according
to compare-match A of the ITU channel selected by TPCR) (initial
value)
1
TPC output group 3 operates in non-overlap mode (1 output and 0
output can be performed independently according to compare-match A
and B of the ITU channel selected by TPCR)
•
Bit 2 (group 2 non-overlap mode (G2NOV)): G2NOV selects the ordinary or non-overlap
mode for TPC output group 2 (TP11–TP8).
Bit 2: G2NOV
Description
0
TPC output group 2 operates normally (output value updated according
to compare-match A of the ITU channel selected by TPCR) (initial
value)
1
TPC output group 2 operates in non-overlap mode (1 output and 0
output can be performed independently according to compare-match A
and B of the ITU channel selected by TPCR)
•
Bit 1 (group 1 non-overlap mode (G1NOV)): G1NOV selects the ordinary or non-overlap
mode for TPC output group 1 (TP7–TP4).
Bit 1: G1NOV
Description
0
TPC output group 1 operates normally (output value updated according
to compare-match A of the ITU channel selected by TPCR) (initial
value)
1
TPC output group 1 operates in non-overlap mode (1 output and 0
output can be performed independently according to compare-match A
and B of the ITU channel selected by TPCR)
•
Bit 0 (group 0 non-overlap mode (G0NOV)): G0NOV selects the ordinary or non-overlap
mode for TPC output group 0 (TP3–TP0).
RENESAS 315
Bit 0: G0NOV
Description
0
TPC output group 0 operates normally (output value updated according
to compare-match A of the ITU channel selected by TPCR) (initial
value)
1
TPC output group 0 operates in non-overlap mode (1 output and 0
output can be performed independently according to compare-match A
and B of the ITU channel selected by TPCR)
11.3
Operation
11.3.1
Overview
When corresponding bits in the PBCR1, PBCR2, NDERA and NDERB registers are set to 1, TPC
output is enabled and the PBDR data register values are output. After that, when the comparematch event selected by TPCR occurs, the next data register contents (NDRA and NDRB) are
transferred to the PBDR and output values are updated. Figure 11.2 illustrates the TPC output
operation.
CR
NDER
Q
Q
Output trigger
signal
C
Port function
select
Q
DR
D
Q
NDR
D
Internal
data bus
TPC
output pin
Figure 11.2 TPC Output Operation
If new data is written in next data registers A and B before the next compare-match occurs, a
maximum 16 bits of data can be output at each successive compare-match. See section 11.3.4,
TPC Output Non-Overlap Operation, for details on non-overlap operation.
316 RENESAS
11.3.2
Output Timing
If TPC output is enabled, next data register (NDRA/NDRB) contents are transferred to the data
register (PBDR) and output when the selected compare-match occurs. Figure 11.3 shows the
timing of these operations. The example is of ordinary output upon compare match A with groups
2 and 3.
CK
N
TCNT
GRA
N+1
N
Compare
match A
signal
n
NDRB
PBDR
m
n
TP15–TP8
m
n
Figure 11.3 Transfer and Output Timing for NDR Data
11.3.3
Examples of Use of Ordinary TPC Output
Settings for Ordinary TPC Output (figure 11.4):
1. Select GRA as the output compare register (output disable) with the timer I/O control register
(TIOR).
2. Set the TPC output trigger cycle.
3. Select the counter clock with the TPSC2–TPSC0 bits of the timer control register (TCR).
Select the counter clear sources with the CCLR1 and CCLR0 bits.
4. Set the timer interrupt enable register (TIER) to enable IMIA interrupts. Transfers to the NDR
can also be set using the DMAC.
5. Set the initial output value in the I/O port data register to be used by TPC.
6. Set the I/O port control register to be used by TPC as the TP pin function (11).
RENESAS 317
7. Set to 1 the bit that performs TPC output to the next data enable register (NDER).
8. Select the ITU compare match that will be the TPC output trigger using the TPC output control
register (TPCR).
9. Set the next TPC output value in the NDR.
10. Set 1 in the STR bit of the timer start register (TSTR) and start the timer counter counting.
11. Set the next output value in the NDR whenever an IMIA interrupt is generated.
Original TPC
output operation
Select GR function
(1)
Set GRA
(2)
Set count operation
(3)
Select interrupt request
(4)
Set initial output value
(5)
Set port output
(6)
Set TPC output enable
(7)
ITU setting
Port and
TPC
setting
Select TPC output trigger
ITU setting
(8)
Set next TPC output value
(9)
Start count
(10)
Compare
match?
No
Yes
Set next TPC output value (11)
Figure 11.4 Example of Setting Procedure for TPC Ordinary Output
318 RENESAS
Five-Phase Pulse Output (figure 11.5):
1. Set the GRA of the ITU that serves as output trigger as the output compare register. Set the
cycle time in the GRA of the ITU and select to clear the counter upon compare match A. Set
the IMIEA bit of TIER to 1 to enable the compare match A interrupt.
2. Write H'FFC0 in the PBCR1, write H'F8 in the NDERB, and set G3CMS0, G3CMS1,
G2CMS1 and G2CMS0 in the TPCR to set the ITU compare match selected in step 1 as the
output trigger. Write output data H'80 in the NDRB.
3. When the selected ITU channel starts operating and a compare-match occurs, the values in the
NDRB are transferred to the PBDR and output. The compare-match/input capture A (IMIA)
interrupt service routine writes the next output data (H'C0) in the NDRB.
4. Five-phase pulse output can be obtained by writing H'40, H'60, H'20, H'30, H'10, H'18, H'08,
H'88… at successive compare-match interrupts. If the DMA controller is set for activation by
compare-match, pulse output can be obtained without loading the CPU.
TCNT
value TCNT
GRA
Compare matches
H'0000
NDRB
PBDR
Time
80
C0
40
60
20
30
10
18
08
88
80
C0
8000 C000 4000 6000 2000 3000 1000 1800 0800 8800 8000 C000
TP15
TP14
TP13
TP12
TP11
Figure 11.5 TPC Output Example (5-Phase Pulse Output)
RENESAS 319
11.3.4
TPC Output Non-Overlap Operation
Setting Procedures for TPC Output Non-Overlap Operation (figure 11.6):
1. Select GRA and GRB as output compare registers (output disable) with the timer I/O control
register (TIOR).
2. Set the TPC output trigger cycle to GRB and the non-overlap cycle to GRA.
3. Select the counter clock with the TPSC2–TPSC0 bits of the timer control register (TCR).
Select the counter clear sources with the CCLR1 and CCLR0 bits.
4. Set the timer interrupt enable register (TIER) to enable IMIA interrupts. Transfers to the NDR
can also be set using the DMAC.
5. Set the initial output value in the I/O port data register to be used by TPC.
6. Set the I/O port control register to be used by TPC as the TP pin function (11).
7. Set to 1 the bit that performs TPC output to the next data enable register (NDER).
8. Select the ITU compare match that will be the TPC output trigger using the TPC output control
register (TPCR).
9. Select the group that performs the non-overlap operation in the TPC output mode register
(TPMR).
10. Set the next TPC output value in the NDR.
11. Set 1 in the STR bit of the timer start register (TSTR) and start the timer counter counting.
12. Set the next output value in the NDR whenever an IMIA interrupt is generated.
320 RENESAS
TPC output nonoverlap operation
Select GR function
(1)
Set GRA
(2)
Set count operation
(3)
Select interrupt request
(4)
Set initial output value
(5)
Set TPC output
(6)
Set TPC transfer enable
(7)
Select TPC output trigger
(8)
Select non-overlap group
(9)
Set next TPC output value
(10)
Start count
(11)
ITU setting
Port and
TPC
setting
ITU setting
Compare
match A?
No
Yes
Set next TPC output value
(12)
Figure 11.6 Example of Setting Procedures for TPC Output Non-Overlap Operation
RENESAS 321
TPC Output Non-Overlap Operation (Four-Phase Complementary Non-Overlap Output)
(figure 11.7):
1. Set GRA and GRB of the ITU that serves as output trigger in the output compare registers. Set
the cycle in the GRB and the non-overlap cycle time in the GRA and select to clear the counter
upon compare match B. Set the IMIEA bit of TIER to 1 to enable the IMIA interrupt.
2. Write H'FFFF in the PBCR1, write H'FF in the NDERB, and set G3CMS1, G3CMS0,
G2CMS1 and G2CMS0 in the TPCR to set the ITU compare match selected in step 1 as the
output trigger. Set the G3NOV and G2NOV bits in the TPMR to 1 to set the non-overlap
operation. Write output data H'95 in the NDRB.
3. When the selected ITU channel starts operating and a GRB compare-match occurs, 1 output
changes to 0 output; when a GRA compare match occurs, 0 output changes to 1 output. (The
change from 0 output to 1 output is delayed by the value set in GRA.) The IMIA interrupt
service routine writes the next output data (H'65) in the NDRB.
4. Four-phase complementary non-overlap output can be obtained by writing H'59, H'56, H'95…
at successive IMIA interrupts. If the DMA controller is set for activation by compare-match,
pulse output can be obtained without loading the CPU.
322 RENESAS
TCNT
value
GRB
TCNT
GRA
Time
H'0000
NDRB
95
PBDR
00
65
95
05
59
65
41
56
59
50
95
56
14
65
95
05
65
Non-overlap cycle
TP15
TP14
TP13
TP12
TP11
TP10
TP9
TP8
Figure 11.7 Non-Overlap Output Example (Four-Phase Complementary Output)
RENESAS 323
11.3.5
TPC Output by Input Capture
TPC can also be output by using input capture rather than ITU compare matches. The general
register A (GRA) of the ITU selected by the TPCR functions as an input capture register and TPC
output occurs upon an input capture signal. Figure 11.8 shows the timing.
CK
TIOC pin
Input
capture
signal
NDR
DR
N
M
N
Figure 11.8 TPC Output by Input Capture
324 RENESAS
11.4
Usage Notes
11.4.1
Non-Overlap Operation
During non-overlap operation, transfers from the NDR to data registers (DR) occurs as follows.
1. NDR contents are always transferred to the DR on compare match A.
2. The contents of the bit transferred by the NDR are only transferred on compare match B when
they are 0. No transfer occurs for a 1.
Figure 11.9 illustrates the TPC output operation during non-overlap.
CR
NDER
Q
Q
Compare match A
Compare match B
C
Port function
select
Q
DR
D
Q
NDR
D
TPC output pin
Figure 11.9 TPC Output Non-Overlap Operation
RENESAS 325
When a compare match B occurs before the compare match A, the 0 data transfer can be
performed before the 1 data transfer, so a non-overlapping waveform can be output. In such cases,
be sure not to change the NDR contents until the compare match A after the compare match B
occurs (non-overlap period). This can be ensured by writing the next data to the NDR using the
IMIA interrupt service routine. The DMAC can also be started up using an IMIA interrupt.
However, these write operations should be performed prior to the next compare match B. The
timing is shown in figure 11.10.
Compare
match A
Compare
match B
NDR write
NDR write
NDR
DR
0 output 0/1 output
NDR write period
NDR write
disable period
0 output 0/1 output
NDR write period
NDR write
disable period
Figure 11.10 Non-Overlap Operation and NDR Write Timing
326 RENESAS
Section 12 Watchdog Timer (WDT)
12.1
Overview
The SuperH microcomputer has a one-channel watchdog timer (WDT) for monitoring system
operations. If a system becomes uncontrolled and the timer counter overflows without being
rewritten correctly by the CPU, an overflow signal (WDTOVF) is output externally . The WDT
can simultaneously generate an internal reset signal for the entire chip.
When this watchdog function is not needed, the WDT can be used as an interval timer. In the
interval timer operation, an interval timer interrupt is generated at each counter overflow. The
WDT is also used in recovering from the standby mode.
12.1.1
•
•
•
•
•
Features
Watchdog timer mode and interval timer mode can be selected.
Outputs WDTOVF in the watchdog timer mode. When the counter overflows in the watchdog
timer mode, overflow signal WDTOVF is output externally. You can select whether or not to
reset the chip internally when this happens. Either the power-on reset or manual reset signal
can be selected as the internal reset signal.
Generates interrupts in the interval timer mode. When the counter overflows, it generates an
interval timer interrupt.
Used to clear the standby mode.
Selection of eight counter clock sources
RENESAS 327
12.1.2
Block Diagram
Figure 12.1 is the block diagram of the WDT.
Overflow
Interrupt
control
Clock
WDTOVF
Internal
reset signal*
Clock
select
Reset
control
RSTCSR
TCNT
φ/2
φ/64
φ/128
φ/256
φ/512
φ/1024
φ/4096
φ/8192
Internal
clock sources
TCSR
Bus
interface
Module bus
Internal data bus
ITI
(interrupt
signal)
WDT
TCSR: Timer control/status register
TCNT: Timer counter
RSTCSR: Reset control/status register
Note: The internal reset signal can be generated by setting the register. The type of reset can
be selected (power-on or manual resets).
Figure 12.1 WDT Block Diagram
12.1.3
Pin Configuration
Table 12.1 shows the pin configuration.
RENESAS 328
Table 12.1 Pin Configuration
Pin
Abbreviation
I/O
Function
Watchdog timer overflow
WDTOVF
O
Outputs the counter overflow signal in
the watchdog mode
12.1.4
Register Configuration
Table 12.2 summarizes the three WDT registers. They are used to select the clock, switch the
WDT mode, and control the reset signal.
Table 12.2
WDT Registers
Address
Name
Abbreviation
R/W
Initial Value
Write*1
Timer
control/status
register
TCSR
R/(W)*3
H'18
H'5FFFFB8
Timer counter
TCNT
R/W
H'00
Reset
control/status
register
RSTCSR
R/(W)*3
H'3F
Read* 2
H'5FFFFB8
H'5FFFFB9
H'5FFFFBA
H'5FFFFBB
Notes: 1. Write by word transfer. It cannot be written in byte or long word.
2. Read by byte transfer. It cannot be read in word or long word.
3. Only 0 can be written in bit 7 to clear the flag.
12.2
Register Descriptions
12.2.1
Timer Counter (TCNT)
The TCNT is an eight-bit readable and writable upcounter. The TCNT differs from other registers
in that it is more difficult to write. See section 12.2.4, Register Access, for details. When the timer
enable bit (TME) in the timer control/status register (TCSR) is set to 1, the timer counter starts
counting pulses of an internal clock source selected by clock select bits 2–0 (CKS2–CKS0) in the
TCSR. When the value of the TCNT overflows (changes from H'FF–H'00), a watchdog timer
overflow signal (WDTOVF) or interval timer interrupt (ITI) is generated, depending on the mode
selected in the WT/IT bit of the TCSR. The TCNT is initialized to H'00 by a reset and when the
TME bit is cleared to 0. It is not initialized in the standby mode.
RENESAS 329
Bit:
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
12.2.2
Timer Control/Status Register (TCSR)
The timer control/status register (TCSR) is an eight-bit readable and writable register. The TCSR
differs from other registers in being more difficult to write. See section 12.2.4, Register Access,
for details. Its functions include selecting the timer mode and clock source. Bits 7–5 are initialized
to 000 by a reset or in standby mode. Bits 2–0 are initialized to 000 by a reset, but retain their
values in the standby mode.
Bit:
Bit name:
Initial value:
R/W:
•
7
6
5
4
3
2
1
0
OVF
WT/IT
TME
—
—
CKS2
CKS1
CKS0
0
0
0
1
1
0
0
0
R/(W)*
R/W
R/W
—
—
R/W
R/W
R/W
Bit 7 (overflow flag (OVF)): OVF indicates that the TCNT has overflowed from H'FF–H'00. It
is not set in the watchdog timer mode.
Bit 7: OVF
Description
0
No overflow of TCNT in interval timer mode (initial value)
Cleared by reading OVF, then writing 0 in OVF
1
•
TCNT overflow in the interval timer mode
Bit 6 (timer mode select (WT/IT)): WT/IT selects whether to use the WDT as a watchdog
timer or interval timer. When the TCNT overflows, the WDT either generates an interval timer
interrupt (ITI) or generates a WDTOVF signal, depending on the mode selected.
Bit 6: WT/IT
Description
0
Interval timer mode: interval timer interrupt to the CPU when TCNT
overflows (initial value)
1
Watchdog timer mode: WDTOVF signal output externally when TCNT
overflows. Section 12.2.3, Reset Control/Status Register (RSTCSR),
describes in detail what happens when TCNT overflows in the
watchdog timer mode.
•
Bit 5 (timer enable (TME)): TME enables or disables the timer.
RENESAS 330
Bit 5: TME
Description
0
Timer disabled: TCNT is initialized to H'00 and count-up stops (initial
value)
1
Timer enabled: TCNT starts counting. A WDTOVF signal or interrupt
is generated when TCNT overflows.
•
Bits 4 and 3 (reserved): These bits always read as 1. The write value should always be 1.
•
Bits 2–0 (clock Select 2–0 (CKS2–CKS0)): CKS2–CKS0 select one of eight internal clock
sources for input to the TCNT. The clock signals are obtained by dividing the frequency of the
system clock (φ).
Description
Bit 2:
CKS2
Bit 1:
CKS1
Bit 0:
CKS0
Clock Source
Overflow Interval* (φ = 20 MHz)
0
0
0
φ/2 (initial value)
25.6 µs
0
0
1
φ/64
819.2 µs
0
1
0
φ/128
1.6 ms
0
1
1
φ/256
3.3 ms
1
0
0
φ/512
6.6 ms
1
0
1
φ/1024
13.1 ms
1
1
0
φ/4096
52.4 ms
1
1
1
φ/8192
104.9 ms
Note: The overflow interval listed is the time from when the TCNT begins counting at H'00 until an
overflow occurs.
12.2.3
Reset Control/Status Register (RSTCSR)
The RSTCSR is an eight-bit readable and writable register that controls output of the reset signal
generated by timer counter (TCNT) overflow and selects the internal reset signal type. The
RSTCSR differs from other registers in that it is more difficult to write. See section 12.2.4
Register Access, for details. RSTCR is initialized to H'1F by input of a reset signal from the RES
pin, but is not initialized by the internal reset signal generated by the overflow of the WDT. It is
initialized to H'1F in standby mode.
Bit:
Bit name:
Initial value:
R/W:
7
6
5
4
3
2
1
0
WOVF
RSTE
RSTS
—
—
—
—
—
0
0
0
1
1
1
1
1
R/(W)*
R/W
R/W
—
—
—
—
—
Note: Only 0 can be written in bit 7 to clear the flag.
RENESAS 331
•
Bit 7 (watchdog timer overflow (WOVF)): WOVF indicates that the TCNT has overflowed
(H'FF → H'00) in the watchdog timer mode. It is not set in the interval timer mode.
Bit 7: WOVF
Description
0
No TCNT overflow in watchdog timer mode (initial value)
Cleared when software reads WOVF, then writes 0 in WOVF
1
Set by TCNT overflow in watchdog timer mode
•
Bit 6 (reset enable (RSTE)): RSTE selects whether to reset the chip internally if the TCNT
overflows in the watchdog timer mode.
Bit 6: RSTE
Description
0
Not reset when TCNT overflows (initial value). LSI not reset internally,
but TCNT and TCSR reset within WDT.
1
Reset when TCNT overflows
•
Bit 5 (reset select (RSTS)): RSTS selects the type of internal reset generated if the TCNT
overflows in the watchdog timer mode.
Bit 5: RSTS
Description
0
Power-on reset initial value)
1
Manual reset
•
Bits 4–0 (reserved): These bits always read as 1. The write value should always be 1.
RENESAS 332
12.2.4
Register Access
The watchdog timer’s TCNT, TCSR, and RSTCSR registers differ from other registers in that they
are more difficult to write. The procedures for writing and reading these registers are given below.
Writing to the TCNT and TCSR: These registers must be written by a word transfer instruction.
They cannot be written by byte transfer instructions. The TCNT and TCSR both have the same
write address. The write data must be contained in the lower byte of the written word. The upper
byte must be H'5A (for the TCNT) or H'A5 (for the TCSR) (figure 12.2). This transfers the write
data from the lower byte to the TCNT or TCSR.
Writing to the TCNT
15
Address:
H'5FFFFB8
8
H'5A
7
0
Write data
Writing to the TCSR
15
Address:
H'5FFFFB8
H'A5
8
7
0
Write data
Figure 12.2 Writing to the TCNT and TCSR
RENESAS 333
Writing to the RSTCSR: The RSTCSR must be written by a word access to address
H'5FFFFFBA. It cannot be written by byte transfer instructions. Procedures for writing 0 in
WOVF (bit 7) and for writing to RSTE (bit 6) and RSTS (bit 5) are different, as shown in figure
12.3. To write 0 in the WOVF bit, the write data must be H'A5 in the upper byte and H'00 in the
lower byte. This clears the WOVF bit to 0. The RSTE and RSTS bits are not affected. To write to
the RSTE and RSTS bits, the upper byte must be H'5A and the lower byte must be the write data.
The values of bits 6 and 5 of the lower byte are transferred to the RSTE and RSTS bits,
respectively. The WOVF bit is not affected.
Writing 0 to the WOVF bit
15
Address:
H'5FFFFBA
8
H'A5
7
0
H'00
Writing to the RSTE and RSTS bits
15
Address:
H'5FFFFBA
8
H'5A
7
0
Write data
Figure 12.3 Writing to the RSTCSR
Reading from the TCNT, TCSR, and RSTCSR: TCNT, TCSR, and RSTCSR are read like
other registers. Use byte transfer instructions. The read addresses are H'5FFFFB8 for the TCSR,
H'5FFFFB9 for the TCNT, and H'5FFFFBB for the RSTCSR.
12.3
Operation
12.3.1
Operation in the Watchdog Timer Mode
To use the WDT as a watchdog timer, set the WT/IT and TME bits of the TCSR to 1. Software
must prevent TCNT overflow by rewriting the TCNT value (normally by writing H'00) before
overflow occurs. If the TCNT fails to be rewritten and overflows due to a system crash or the like,
a WDTOVF signal is output (figure 12.4). The WDTOVF signal can be used to reset external
system devices. The WDTOVF signal is output for 128φ clock cycles.
If the RSTE bit in the RSTCSR is set to 1, a signal to reset the chip will be generated internally
simultaneous to the WDTOVF signal when TCNT overflows. Either a power-on reset or a manual
reset can be selected by the RSTS bit. The internal reset signal is output for 512φ clock cycles.
When a watchdog reset is generated simultaneously with input at the RES pin, the software
distinguishes the RES reset from the watchdog reset by checking the WOVF bit in the RSTCSR.
The RES reset takes priority. The WOVF bit is cleared to 0.
RENESAS 334
TCNT
value
Overflow
H'FF
H'00
Time
WT/IT = 1
TME = 1
H'00 written
in TCNT
WOVF = 1
WT/IT = 1 H'00 written
TME = 1
in TCNT
WDTOVF and
internal reset generated
WDTOVF
signal
128φ clock
Internal
reset signal*
512φ clock
WT/IT: Timer mode select bit
TME: Timer enable bit
Note: The internal reset signal is only generated when the RSTE bit is 1.
Figure 12.4 Operation in the Watchdog Timer Mode
RENESAS 335
12.3.2
Operation in the Interval Timer Mode
To use the WDT as an interval timer, clear WT/IT to 0 and set TME to 1. An interval timer
interrupt (ITI) is generated each time the timer counter overflows. This function can be used to
generate interval timer interrupts at regular intervals (figure 12.5).
TCNT value
Overflow
H'FF
Overflow
Overflow
Overflow
H'00
Time
WT/IT = 0
TME = 1
ITI
ITI
ITI
ITI
Figure 12.5 Operation in the Interval Timer Mode
12.3.3
Operation in the Standby Mode
The watchdog timer has a special function to clear the standby mode with an NMI interrupt. When
using the standby mode, set the WDT as described below.
Transition to the Standby Mode: The TME bit in the TCSR must be cleared to 0 to stop the
watchdog timer counter before it enters the standby mode. The chip cannot enter the standby mode
while the TME bit is set to 1. Set bits CKS2–CKS0 so that the counter overflow interval is equal
to or longer than the oscillation settling time. See section 20.3, AC Characteristics, for the
oscillation settling time.
Recovery from the Standby Mode: When an NMI request signal is received in standby mode,
the clock oscillator starts running and the watchdog timer starts counting at the rate selected by
bits CKS2–CKS0 before the standby mode was entered. When the TCNT overflows (changes
from H'FF–H'00), the system clock (φ) is presumed to be stable and usable; clock signals are
supplied to the entire chip and the standby mode ends.
For details on the standby mode, see section 19, Power Down States.
RENESAS 336
12.3.4
Timing of Setting the Overflow Flag (OVF)
In the interval timer mode, when the TCNT overflows the OVF flag is set to 1 and an interval
timer interrupt is requested (figure 12.6).
CK
H'FF
TCNT
H'00
Overflow signal
(internal signal)
OVF
Figure 12.6 Timing of Setting the OVF
12.3.5
Timing of Setting the Watchdog Timer Overflow Flag (WOVF)
When the TCNT overflows the WOVF bit of the RSTCSR is set to 1 and a WDTOVF signal is
output. When the RSTE bit is set to 1, TCNT overflow enables an internal reset signal to be
generated for the entire chip (figure 12.7).
CK
TCNT
H'FF
H'00
Overflow signal
(internal signal)
WOVF
Figure 12.7 Timing of Setting the WOVF Bit and Internal Reset
RENESAS 337
12.4
Usage Notes
12.4.1
TCNT Write and Count Up Contention
If a timer counter clock pulse is generated during the T3 state of a write cycle to the TCNT, the
write takes priority and the timer counter is not incremented (figure 12.8).
TCNT write cycle
T1
T2
T3
CK
Address
TCNT address
Internal
write signal
TCNT
input clock
TCNT
N
M
Counter write data
Figure 12.8 Contention between TCNT Write and Increment
12.4.2
Changing CKS2-CKS0 Bit Values
If the values of bits CKS2–CKS0 are altered while the WDT is running, the count may increment
incorrectly. Always stop the watchdog timer (by clearing the TME bit to 0) before changing the
values of bits CKS2–CKS0.
12.4.3
Changing Watchdog Timer/Interval Timer Modes
To prevent incorrect operation, always stop the watchdog timer (by clearing the TME bit to 0)
before switching between interval timer mode and watchdog timer mode.
RENESAS 338
12.4.4
System Reset With WDTOVF
If a WDTOVF signal is input to the RES pin, the LSI cannot initialize correctly. Avoid logical
input of the WDTOVF output signal to the RES input pin. To reset the entire system with the
WDTOVF signal, use the circuit shown in figure 12.9.
SuperH
microprocessor
Reset input
Reset signal to
entire system
RES
WDTOVF
Figure 12.9 Example of a System Reset Circuit with a WDTOVF Signal
12.4.5
Internal Reset With the Watchdog Timer
If the RSTE bit is cleared to 0 in the watchdog timer mode, the LSI will not reset internally when a
TCNT overflow occurs, but the TCNT and TCSR in WDT will reset.
RENESAS 339
Section 13 Serial Communication Interface (SCI)
13.1
Overview
The SuperH microcomputer has a serial communication interface (SCI) with two independent
channels. Both channels are functionally identical. The SCI supports both asynchronous and
clocked synchronous serial communication. It also has a multiprocessor communication function
for serial communication among two or more processors.
13.1.1
•
•
•
•
•
•
Features
Asynchronous mode
 Serial data communications are synched by start-stop in character units. The SCI can
communicate with a universal asynchronous receiver/transmitter (UART), an asynchronous
communication interface adapter (ACIA), or any other chip that employs a standard
asynchronous serial communication. It can also communicate with two or more other
processors using the multiprocessor communication function. There are twelve selectable
serial data communication formats.
 Data length: seven or eight bits
 Stop bit length: one or two bits
 Parity: even, odd, or none
 Multiprocessor bit: one or none
 Receive error detection: parity, overrun, and framing errors
 Break detection: by reading the RxD level directly when a framing error occurs
Clocked synchronous mode
 Serial data communication is synchronized with a clock signal. The SCI can communicate
with other chips having a clocked synchronous communication function. There is one serial
data communication format.
 Data length: eight bits
 Receive error detection: overrun errors
Full duplex communication: The transmitting and receiving sections are independent, so the
SCI can transmit and receive simultaneously. Both sections use double buffering, so
continuous data transfer is possible in both the transmit and receive directions.
On-chip baud rate generator with selectable bit rates
Internal or external transmit/receive clock source: baud rate generator (internal) or SCK pin
(external)
Four types of interrupts: Transmit-data-empty, transmit-end, receive-data-full, and receiveerror interrupts are requested independently. The transmit-data-empty and receive-data-full
interrupts can start the direct memory access controller (DMAC) to transfer data.
RENESAS 341
13.1.2
Block Diagram
Bus interface
Figure 13.1 shows a block diagram of the SCI.
Module data bus
RDR
TDR
BRR
SSR
SCR
RXD
RSR
TSR
SMR
Transmit/
receive control
TXD
Parity
generation
Parity check
SCK
Baud rate
generator
External clock
SMR: Serial mode register
SCR: Serial control register
SSR: Serial status register
BRR: Bit rate register
Figure 13.1 SCI Block Diagram
RENESAS 342
φ
φ/4
φ/16
φ/64
Clock
SCI
RSR: Receive shift register
RDR: Receive data register
TSR: Transmit shift register
TDR: Transmit data register
Internal
data bus
TEI
TXI
RXI
ERI
13.1.3
Input/Output Pins
Table 13.1 summarizes the SCI pins by channel.
Table 13.1 SCI Pins
Channel
Pin Name
Abbreviation
Input/Output
Function
0
Serial clock pin
SCK0
Input/output
SCI0 clock input/output
Receive data pin
RxD0
Input
SCI0 receive data input
1
13.1.4
Transmit data pin
TxD0
Output
SCI0 transmit data output
Serial clock pin
SCK1
Input/output
SCI1 clock input/output
Receive data pin
RxD1
Input
SCI1 receive data input
Transmit data pin
TxD1
Output
SCI1 transmit data output
Register Configuration
Table 13.2 summarizes the SCI internal registers. These registers select the communication mode
(asynchronous or clocked synchronous), specify the data format and bit rate, and control the
transmitter and receiver sections.
Table 13.2 Registers
Channel
Address* 1
Name
Abbreviation
R/W
Initial
Value
Access
size
0
H'05FFFEC0
Serial mode register
SMR0
R/W
H'00
8, 16
H'05FFFEC1
Bit rate register
BRR0
R/W
H'FF
8, 16
H'05FFFEC2
Serial control register
SCR0
R/W
H'00
8, 16
H'05FFFEC3
Transmit data register
TDR0
R/W
H'FF
8, 16
H'05FFFEC4
Serial status register
SSR0
R/(W)*2
H'84
8, 16
H'05FFFEC5
Receive data register
RDR0
R
H'00
8, 16
H'05FFFEC8
Serial mode register
SMR1
R/W
H'00
8, 16
H'05FFFEC9
Bit rate register
1
BRR1
R/W
H'FF
8, 16
H'05FFFECA Serial control register
SCR1
R/W
H'00
8, 16
H'05FFFECB Transmit data register
TDR1
R/W
H'FF
8, 16
H'05FFFECC Serial status register
SSR1
R/(W)*2
H'84
8, 16
H'05FFFECD Receive data register
RDR1
R
H'00
8, 16
Notes: 1. Only the values of bits A27–A24 and A8-A0 are valid; bits A23–A9 are ignored. For
details on the register addresses, see section 8.3.5, Description of Areas.
2. Write 0 to clear flags.
RENESAS 343
13.2
Register Descriptions
13.2.1
Receive Shift Register
The receive shift register (RSR) receives serial data. Data input at the RxD pin are loaded into the
RSR in the order received, LSB (bit 0) first. In this way the SCI converts received data to parallel
form. When one byte has been received, it is automatically transferred to the receive data register
(RDR). The CPU cannot read or write the RSR directly.
Bit:
7
6
5
4
3
2
1
0
—
—
—
—
—
—
—
—
Bit name:
R/W:
13.2.2
Receive Data Register
The receive data register (RDR) stores serial receive data. The SCI completes the reception of one
byte of serial data by moving the received data from the receive shift register (RSR) into the RDR
for storage. The RSR is then ready to receive the next data. This double buffering allows the SCI
to receive data continuously.
The CPU can read but not write the RDR. The RDR is initialized to H'00 by a reset or in standby
mode.
Bit:
7
6
5
4
3
2
1
0
Initial value:
0
0
0
0
0
0
0
0
R/W:
R
R
R
R
R
R
R
R
Bit name:
13.2.3
Transmit Shift Register
The transmit shift register (TSR) transmits serial data. The SCI loads transmit data from the
transmit data register (TDR) into the TSR, then transmits the data serially from the TxD pin, LSB
(bit 0) first. After transmitting one data byte, the SCI automatically loads the next transmit data
from the TDR into the TSR and starts transmitting again. If the TDRE bit of the SSR is 1,
however, the SCI does not load the TDR contents into the TSR. The CPU cannot read or write the
TSR directly.
Bit:
7
6
5
4
3
2
1
0
—
—
—
—
—
—
—
—
Bit name:
R/W:
RENESAS 344
13.2.4
Transmit Data Register
The transmit data register (TDR) is an eight-bit register that stores data for serial transmission.
When the SCI detects that the transmit shift register (TSR) is empty, it moves transmit data written
in the TDR into the TSR and starts serial transmission. Continuous serial transmission is possible
by writing the next transmit data in the TDR during serial transmission from the TSR.
The CPU can always read and write the TDR. The TDR is initialized to H'FF by a reset or in
standby mode.
Bit:
7
6
5
4
3
2
1
0
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
13.2.5
Serial Mode Register
The serial mode register (SMR) is an eight-bit register that specifies the SCI serial communication
format and selects the clock source for the baud rate generator.
The CPU can always read and write the SMR. The SMR is initialized to H'00 by a reset or in
standby mode.
Bit:
Bit name:
Initial value:
R/W:
•
7
6
5
4
3
2
1
0
C/A
CHR
PE
O/E
STOP
MP
CKS1
CKS0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit 7 (communication mode (C/A)): C/A selects whether the SCI operates in the asynchronous
or clocked synchronous mode.
Bit 7: C/A
Description
0
Synchronous mode (initial value)
1
Clocked synchronous mode
RENESAS 345
•
Bit 6 (character length (CHR)): CHR selects seven-bit or eight-bit data in the asynchronous
mode. In the clocked synchronous mode, the data length is always eight bits, regardless of the
CHR setting.
Bit 6: CHR
Description
0
Eight-bit data (initial value)
1
Seven-bit data. When seven-bit data is selected, the MSB (bit 7) of the
transmit data register is not transmitted.
•
Bit 5 (parity enable (PE)): PE selects whether to add a parity bit to transmit data and check the
parity of receive data, in the asynchronous mode. In the clocked synchronous mode, a parity
bit is neither added nor checked, regardless of the PE setting.
Bit 5: PE
Description
0
Parity bit not added or checked (initial value)
1
Parity bit added and checked. When PE is set to 1, an even or odd
parity bit is added to transmit data, depending on the parity mode (O/E)
setting. Receive data parity is checked according to the even/odd (O/E)
mode setting.
•
Bit 4 (parity mode (O/E): O/E selects even or odd parity when parity bits are added and
checked. The O/E setting is used only in asynchronous mode and only when the parity enable
bit (PE) is set to 1 to enable parity addition and check. The O/E setting is ignored in the
clocked synchronous mode, or in the asynchronous mode when parity addition and check is
disabled.
Bit 4: O/E
Description
0
Even parity. If even parity is selected, the parity bit added to transmit
data makes an even number of 1s in the transmitted character and
parity bit combined. Receive data must have an even number of 1s in
the received character and parity bit combined (initial value).
1
Odd parity. If odd parity is selected, the parity bit added to transmit
data makes an odd number of 1s in the transmitted character and
parity bit combined. Receive data must have an odd number of 1s in
the received character and parity bit combined.
RENESAS 346
•
Bit 3 (stop bit length (STOP)): STOP selects one or two bits as the stop bit length in the
asynchronous mode. This setting is used only in the asynchronous mode. It is ignored in the
clocked synchronous mode because no stop bits are added.
In receiving, only the first stop bit is checked, regardless of the STOP bit setting. If the second
stop bit is 1, it is treated as a stop bit. If the second stop bit is 0, it is treated as the start bit of
the next incoming character.
Bit 3: STOP
Description
0
One stop bit. In transmitting, a single bit of 1 is added at the end of
each transmitted character (initial value).
1
Two stop bits. In transmitting, two bits of 1 are added at the end of
each transmitted character.
•
Bit 2 (multiprocessor mode (MP)): MP selects multiprocessor format. When multiprocessor
format is selected, settings of the parity enable (PE) and parity mode (O/E) bits are ignored.
The MP bit setting is used only in the asynchronous mode; it is ignored in the clocked
synchronous mode. For the multiprocessor communication function, see section 13.3.3,
Multiprocessor Communication.
Bit 2: MP
Description
0
Multiprocessor function disabled (initial value)
1
Multiprocessor format selected
•
Bits 1 and 0 (clock select 1 and 0 (CKS1 and CKS0)): CKS1 and CKS0 select the internal
clock source of the on-chip baud rate generator. Four clock sources are available: φ, φ/4, φ/16,
and φ/64. For further information on the clock source, bit rate register settings, and baud rate,
see section 13.2.8, Bit Rate Register.
Bit 1: CKS1
Bit 0: CKS0
Description
0
0
System clock (φ) (initial value)
1
φ/4
0
φ/16
1
φ/64
1
RENESAS 347
13.2.6
Serial Control Register
The serial control register (SCR) enables the SCI transmitter/receiver, selects serial clock output in
the asynchronous mode, enables and disables interrupts, and selects the transmit/receive clock
source. The CPU can always read and write the SCR. The SCR is initialized to H'00 by a reset or
in standby mode.
Bit:
Bit name:
Initial value:
R/W:
•
7
6
5
4
3
2
1
0
TIE
RIE
TE
RE
MPIE
TEIE
CKE1
CKE0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit 7 (transmit interrupt enable (TIE)): TIE enables or disables the transmit-data-empty
interrupt (TXI) requested when the transmit data register empty bit (TDRE) in the serial status
register (SSR) is set to 1 due to transfer of serial transmit data from the TDR to the TSR.
Bit 7: TIE
Description
0
Transmit-data-empty interrupt request (TXI) is disable. The TXI
interrupt request can be cleared by reading TDRE after it has been set
to 1, then clearing TDRE to 0, or by clearing TIE to 0 (initial value).
1
Transmit-data-empty interrupt request (TXI) is enabled
•
Bit 6 (receive interrupt enable (RIE)): RIE enables or disables the receive-data-full interrupt
(RXI) requested when the receive data register full bit (RDRF) in the serial status register
(SSR) is set to 1 due to transfer of serial receive data from the RSR to the RDR. Also enables
or disables receive-error interrupt (ERI) requests.
Bit 6: RIE
Description
0
Receive-data-full interrupt (RXI) and receive-error interrupt (ERI)
requests are disabled. RXI and ERI interrupt requests can be cleared
by reading the RDRF flag or error flag (FER, PER, or ORER) after it
has been set to 1, then clearing the flag to 0, or by clearing RIE to 0
(initial value).
1
Receive-data-full interrupt (RXI) and receive-error interrupt (ERI)
requests are enabled
RENESAS 348
Bit 5 (transmit enable (TE)): TE enables or disables the SCI transmitter.
Bit 5: TE
Description
0
Transmitter disabled. The transmit data register empty bit (TDRE) in
the serial status register (SSR) is locked at 1 (initial value).
1
Transmitter enabled. Serial transmission starts when the transmit data
register empty (TDRE) bit in the serial status register (SSR) is cleared
to 0 after writing of transmit data into the TDR. Select the transmit
format in the SMR before setting TE to 1.
•
Bit 4 (receive enable (RE)): RE enables or disables the SCI receiver.
Bit 4: RE
Description
0
Receiver disabled (initial value). Clearing RE to 0 does not affect the
receive flags (RDRF, FER, PER, ORER). These flags retain their
previous values.
1
Receiver enabled. Serial reception starts when a start bit is detected in
the asynchronous mode, or serial clock input is detected in the clocked
synchronous mode. Select the receive format in the SMR before
setting RE to 1.
•
Bit 3 (multiprocessor interrupt enable (MPIE)): MPIE enables or disables multiprocessor
interrupts. The MPIE setting is used only in the asynchronous mode, and only if the
multiprocessor mode bit (MP) in the serial mode register (SMR) is set to 1 during reception.
The MPIE setting is ignored in the clocked synchronous mode or when the MP bit is cleared to
0.
Bit 3: MPIE
Description
0
Multiprocessor interrupts are disabled (normal receive operation) (initial
value)
MPE is cleared to 0 when:
1. MPIE is cleared to 0, or
2. Multiprocessor bit (MPB) is set to 1 in receive data.
1
Multiprocessor interrupts are enabled: Receive-data-full interrupt
requests (RXI), receive-error interrupt requests (ERI), and setting of
the RDRF, FER, and ORER status flags in the serial status register
(SSR) are disabled until the multiprocessor bit is set to 1.
The SCI does not transfer receive data from the RSR to the RDR, does
not detect receive errors, and does not set the RDRF, FER, and ORER
flags in the serial status register (SSR). When it receives data that
includes MPB = 1, MPB is set to 1, and the SCI automatically clears
MPIE to 0, generates RXI and ERI interrupts (if the TIE and RIE bits in
the SCR are set to 1), and allows the FER and ORER to be set.
RENESAS 349
•
Bit 2 (transmit-end interrupt enable (TEIE)): TEIE enables or disables the transmit-end
interrupt (TEI) requested if TDR does not contain new transmit data when the MSB is
transmitted.
Bit 2: TEIE
Description
0
Transmit-end interrupt (TEI) requests are disabled* (initial value)
The TEI request can be cleared by reading the TDRE bit in the serial
status register (SSR) after it has been set to 1, then clearing TDRE to
0; by clearing the transmit end (TEND) bit to 0; or by clearing the TEIE
bit to 0.
1
•
Transmit-end interrupt (TEI) requests are enabled.
Bits 1 and 0 (clock enable 1 and 0 (CKE1 and CKE0)): CKE1 and CKE0 select the SCI clock
source and enable or disable clock output from the SCK pin. Depending on the combination of
CKE1 and CKE0, the SCK pin can be used for general-purpose input/output, serial clock
output, or serial clock input.
The CKE0 setting is valid only in the asynchronous mode, and only when the SCI is internally
clocked (CKE1 = 0). The CKE0 setting is ignored in the clocked synchronous mode, or when
an external clock source is selected (CKE1 = 1). Select the SCI operating mode in the serial
mode register (SMR) before setting CKE1 and CKE0. For further details on selection of the
SCI clock source, see table 13.9 in section 13.3, Operation.
Bit 1:
CKE1
Bit 0:
CKE0
Description*1
0
0
Synchronous mode
Internal clock, SCK pin used for input pin (input
signal is ignored or output pin output level is
undefined)
Clocked synchronous mode Internal clock, SCK pin used for serial clock output*2
0
1
Synchronous mode
Internal clock, SCK pin used for clock output*3
Clocked synchronous mode Internal clock, SCK pin used for serial clock output
1
0
Synchronous mode
External clock, SCK pin used for clock input* 4
Clocked synchronous mode External clock, SCK pin used for serial clock input
1
1
Synchronous mode
External clock, SCK pin used for clock input* 4
Clocked synchronous mode External clock, SCK pin used for serial clock input
Notes: 1. The SCK pin is multiplexed with other functions. Set the pin function controller (PFC) to
select the SCK function and the SCK input/output of the SCK pin.
2. Initial value
3. The output clock frequency is the same as the bit rate.
4. The input clock frequency is 16 times the bit rate.
RENESAS 350
13.2.7
Serial Status Register
The serial status register (SSR) is an 8-bit register containing multiprocessor bit values, and status
flags that indicate SCI operating status.
The CPU can always read and write the SSR, but cannot write 1 in the status flags (TDRE, RDRF,
ORER, PER, and FER). These flags can be cleared to 0 only if they have first been read (after
being set to 1). Bits 2 (TEND) and 1 (MPB) are read-only bits that cannot be written. The SSR is
initialized to H'84 by a reset or in standby mode.
Bit:
Bit name:
Initial value:
R/W:
7
6
5
4
3
2
1
0
TDRE
RDRF
ORER
FER
PER
TEND
MPB
MPBT
1
0
0
0
0
1
0
0
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R/(W)*
R
R
R/W
Note: Write 0 to clear flag.
•
Bit 7 (transmit data register empty (TDRE)): TDRE indicates that the SCI has loaded transmit
data from the TDR into the TSR and serial transmit new data can be written in the TDR.
Bit 7: TDRE
Description
0
TDR contains valid transmit data
TDRE is cleared to 0 when:
• Software reads TDRE after it has been set to 1, then writes 0 in TDRE
• The DMAC writes data in TDR
1
TDR does not contain valid transmit data (initial value)
TDRE is set to 1 when:
• The chip is reset or enters standby mode
• The TE bit in the serial control register (SCR) is cleared to 0
• TDR contents are loaded into TSR, so new data can be written in TDR
RENESAS 351
•
Bit 6 (receive data register full (RDRF)): RDRF indicates that RDR contains received data.
Bit 6: RDRF
Description
0
RDR does not contain valid received data (initial value)
RDRF is cleared to 0 when:
• The chip is reset or enters standby mode
• Software reads RDRF after it has been set to 1, then writes 0 in RDRF
• The DMAC reads data from RDR
1
RDR contains valid received data.
RDRF is set to 1 when serial data is received normally and transferred from
RSR to RDR.
Note: The RDR and RDRF are not affected by detection of receive errors or by clearing of the RE
bit to 0 in the serial control register. They retain their previous contents. If RDRF is still set
to 1 when reception of the next data ends, an overrun error (ORER) occurs and the
received data is lost.
•
Bit 5 (overrun error (ORER)): Indicates that data reception ended abnormally due to an
overrun error.
Bit 5: ORER
Description
0
Receiving is in progress or has ended normally (initial value)* 1
ORER is cleared to 0 when:
• The chip is reset or enters standby mode
• Software reads ORER after it has been set to 1, then writes 0 in ORER
1
A receive overrun error occurred*2
ORER is set to 1 if reception of the next serial data ends when RDRF is set to 1
Notes: 1. Clearing the RE bit to 0 in the serial control register does not affect the ORER bit, which
retains its previous value.
2. RDR continues to hold the data received before the overrun error, so subsequent
receive data is lost. Serial receiving cannot continue while ORER is set to 1. In the
clocked synchronous mode, serial transmitting is disabled.
RENESAS 352
•
Bit 4 (framing error (FER)): FER indicates that data reception ended abnormally due to a
framing error in the asynchronous mode.
Bit 4: FER
Description
0
Receiving is in progress or has ended normally. Clearing the RE bit to 0 in the
serial control register does not affect the FER bit, which retains its previous
value (initial value).
FER is cleared to 0 when:
• The chip is reset or enters standby mode
• Software reads FER after it has been set to 1, then writes 0 in FER
1
A receive framing error occurred. When the stop bit length is two bits, only the
first bit is checked. The second stop bit is not checked. When a framing error
occurs, the SCI transfers the receive data into the RDR but does not set RDRF.
Serial receiving cannot continue while FER is set to 1. In the clocked
synchronous mode, serial transmitting is also disabled.
FER is set to 1 if the stop bit at the end of receive data is checked and found to
be 0.
•
Bit 3 (parity error (PER)): PER indicates that data reception (with parity) ended abnormally
due to a parity error in the asynchronous mode.
Bit 3: PER
Description
0
Receiving is in progress or has ended normally. Clearing the RE bit to 0 in the
serial control register does not affect the PER bit, which retains its previous
value (initial value).
PER is cleared to 0 when:
• The chip is reset or enters standby mode
• Software reads PER after it has been set to 1, then writes 0 in PER
1
A receive parity error occurred. When a parity error occurs, the SCI transfers the
receive data into the RDR but does not set RDRF. Serial receiving cannot
continue while PER is set to 1. In the clocked synchronous mode, serial
transmitting is also disabled.
PER is set to 1 if the number of 1s in receive data, including the parity bit, does
not match the even or odd parity setting of the parity mode bit (O/E) in the serial
mode register (SMR).
RENESAS 353
•
Bit 2 (transmit end (TEND)): TEND indicates that when the last bit of a serial character was
transmitted, the TDR did not contain new transmit data, so transmission has ended. TEND is a
read-only bit and cannot be written.
Bit 2: TEND
Description
0
Transmission is in progress
TEND is cleared to 0 when:
• Software reads TDRE after it has been set to 1, then writing 0 in TDRE
• The DMAC writes data in TDR
1
End of transmission (initial value)
TEND is set to 1 when:
• The chip is reset or enters standby mode
• TE is cleared to 0 in the serial control register (SCR)
• TDRE is 1 when the last bit of a one-byte serial character is transmitted
•
Bit 1 (multiprocessor bit (MPB)): MPB stores the value of the multiprocessor bit in receive
data when a multiprocessor format is selected for receiving in the asynchronous mode. The
MPB is a read-only bit and cannot be written.
Bit 1: MPB
Description
0
Multiprocessor bit value in receive data is 0. If RE is cleared to 0 when a
multiprocessor format is selected, the MPB retains its previous value (initial
value).
1
Multiprocessor bit value in receive data is 1
•
Bit 0 (multiprocessor bit transfer (MPBT)): MPBT stores the value of the multiprocessor bit
added to transmit data when a multiprocessor format is selected for transmitting in the
asynchronous mode. The MPBT setting is ignored in the clocked synchronous mode, when a
multiprocessor format is not selected, or when the SCI is not transmitting.
Bit 0: MPBT
Description
0
Multiprocessor bit value in transmit data is 0 (initial value)
1
Multiprocessor bit value in transmit data is 1
RENESAS 354
13.2.8
Bit Rate Register (BRR)
The bit rate register (BRR) is an eight-bit register that, together with the baud rate generator clock
source selected by the CKS1 and CKS0 bits in the serial mode register (SMR), determines the
serial transmit/receive bit rate.
The CPU can always read and write the BRR. The BRR is initialized to H'FF by a reset or in
standby mode. SCI1 and SCI2 have independent baud rate generator control, so different values
can be set in the two channels.
Bit:
7
6
5
4
3
2
1
0
1
1
1
1
1
1
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit name:
Initial value:
R/W:
Table 13.3 shows examples of BRR settings in the asynchronous mode; table 13.4 shows
examples of BBR settings in the clocked synchronous mode.
Table 13.3 Bit Rates and BRR Settings in Asynchronous Mode
φ (MHz)
2
Bit Rate
(bits/s)
n
N
110
1
141
150
1
300
2.097152
Error (%)
Error (%)
n
N
0.03
1
148
–0.04
103
0.16
1
108
0.21
0
207
0.16
0
217
0.21
600
0
103
0.16
0
108
0.21
1200
0
51
0.16
0
54
–0.70
2400
0
25
0.16
0
26
1.14
4800
0
12
0.16
0
13
–2.48
9600
—
—
—
0
6
–2.48
19200
—
—
—
—
—
—
31250
0
1
0.00
—
—
—
38400
—
—
—
—
—
—
RENESAS 355
Table 13.3 Bit Rates and BRR Settings in Asynchronous Mode (cont)
φ (MHz)
2.4576
3
3.6864
Bit Rate(bits/s)
n
N
Error (%) n
N
Error (%) n
N
Error (%)
110
1
174
–0.26
1
212
0.03
2
64
0.70
150
1
127
0.00
1
155
0.16
1
191
0.00
300
0
255
0.00
1
77
0.16
1
95
0.00
600
0
127
0.00
0
155
0.16
0
191
0.00
1200
0
63
0.00
0
77
0.16
0
95
0.00
2400
0
31
0.00
0
38
0.16
0
47
0.00
4800
0
15
0.00
0
19
–2.34
0
23
0.00
9600
0
7
0.00
0
9
–2.34
0
11
0.00
19200
0
3
0.00
0
4
–2.34
0
5
0.00
31250
—
—
—
0
2
0.00
—
—
—
38400
0
1
0.00
—
—
—
0
2
0.00
Table 13.3 Bit Rates and BRR Settings in Asynchronous Mode (cont)
φ (MHz)
4
4.9152
5
Bit Rate(bits/s)
n
N
Error (%) n
N
Error (%) n
N
Error (%)
110
2
70
0.03
2
86
0.31
2
88
–0.25
150
1
207
0.16
1
255
0.00
2
64
0.16
300
1
103
0.16
1
127
0.00
1
129
0.16
600
0
207
0.16
0
255
0.00
1
64
0.16
1200
0
103
0.16
0
127
0.00
0
129
0.16
2400
0
51
0.16
0
63
0.00
0
64
0.16
4800
0
25
0.16
0
31
0.00
0
32
–1.36
9600
0
12
0.16
0
15
0.00
0
15
1.73
19200
—
—
—
0
7
0.00
0
7
1.73
31250
0
3
0.00
0
4
–1.70
0
4
0.00
38400
—
—
—
0
3
0.00
0
3
1.73
RENESAS 356
Table 13.3 Bit Rates and BRR Settings in Asynchronous Mode (cont)
φ (MHz)
6
6.144
7.3728
Bit Rate(bits/s)
n
N
Error (%) n
N
Error (%) n
N
Error (%)
110
2
106
–0.44
2
108
0.08
2
130
–0.07
150
2
77
0.16
2
79
0.00
2
95
0.00
300
1
155
0.16
1
159
0.00
1
191
0.00
600
1
77
0.16
1
79
0.00
1
95
0.00
1200
0
155
0.16
0
159
0.00
0
191
0.00
2400
0
77
0.16
0
79
0.00
0
95
0.00
4800
0
38
0.16
0
39
0.00
0
47
0.00
9600
0
19
–2.34
0
19
0.00
0
23
0.00
19200
0
9
–2.34
0
9
0.00
0
11
0.00
31250
0
5
0.00
0
5
2.40
—
—
—
38400
0
4
–2.34
0
4
0.00
0
5
0.00
Table 13.3 Bit Rates and BRR Settings in Asynchronous Mode (cont)
φ (MHz)
8
9.8304
10
12
Bit Rate
(bits/s)
n
N
Error
(%)
n
N
Error
(%)
n
N
Error
(%)
n
N
Error
(%)
110
2
141
0.03
2
174
–0.26 2
177
–0.25
2
212
0.03
150
2
103
0.16
2
127
0.00
2
129
0.16
2
155
0.16
300
1
207
0.16
1
255
0.00
2
64
0.16
2
77
0.16
600
1
103
0.16
1
127
0.00
1
129
0.16
1
155
0.16
1200
0
207
0.16
0
255
0.00
1
64
0.16
1
77
0.16
2400
0
103
0.16
0
127
0.00
0
129
0.16
0
155
0.16
4800
0
51
0.16
0
63
0.00
0
64
0.16
0
77
0.16
9600
0
25
0.16
0
31
0.00
0
32
–1.36
0
38
0.16
19200
0
12
0.16
0
15
0.00
0
15
1.73
0
19
–2.34
31250
0
7
0.00
0
9
–1.70 0
9
0.00
0
11
0.00
38400
—
—
—
0
7
0.00
7
1.73
0
9
–2.34
0
RENESAS 357
Table 13.3 Bit Rates and BRR Settings in Asynchronous Mode (cont)
φ (MHz)
12.288
14
14.7456
16
Bit Rate
(bits/s)
n
N
Error
(%)
n
N
Error
(%)
n
N
Error
(%)
n
N
Error
(%)
110
2
217
0.08
2
248
–0.17 3
64
0.70
3
70
0.03
150
2
159
0.00
2
181
0.16
2
191
0.00
2
207
0.16
300
2
79
0.00
2
90
0.16
2
95
0.00
2
103
0.16
600
1
159
0.00
1
181
0.16
1
191
0.00
1
207
0.16
1200
1
79
0.00
1
90
0.16
1
95
0.00
1
103
0.16
2400
0
159
0.00
0
181
0.16
0
191
0.00
0
207
0.16
4800
0
79
0.00
0
90
0.16
0
95
0.00
0
103
0.16
9600
0
39
0.00
0
45
–0.93 0
47
0.00
0
51
0.16
19200
0
19
0.00
0
22
–0.93 0
23
0.00
0
25
0.16
31250
0
11
2.40
0
13
0.00
0
14
–1.70
0
15
0.00
38400
0
9
0.00
—
—
—
0
11
0.00
0
12
0.16
Table 13.3 Bit Rates and BRR Settings in Asynchronous Mode (cont)
φ (MHz)
17.2032
18
19.6608
20
Bit Rate
(bits/s)
n
N
Error
(%)
n
N
Error
(%)
n
N
Error
(%)
n
N
Error
(%)
110
3
75
0.48
3
79
–0.12 3
86
0.31
3
88
–0.25
150
2
223
0.00
2
233
0.16
2
255
0.00
3
64
0.16
300
2
111
0.00
2
116
0.16
1
127
0.00
2
129
0.16
600
1
223
0.00
1
233
0.16
1
255
0.00
2
64
0.16
1200
1
111
0.00
1
116
0.16
0
127
0.00
1
129
0.16
2400
0
223
0.00
0
233
0.16
0
255
0.00
1
64
0.16
4800
0
111
0.00
0
116
0.16
0
127
0.00
0
129
0.16
9600
0
55
0.00
0
58
–0.69 0
63
0.00
0
64
0.16
19200
0
27
0.00
0
28
1.02
0
31
0.00
0
32
–1.36
31250
0
16
1.20
0
17
0.00
0
19
–1.70
0
19
0.00
38400
0
13
0.00
0
14
–2.34 0
15
0.00
0
15
1.73
RENESAS 358
Table 13.4 Bit Rates and BRR Settings in Clocked Synchronous Mode
φ (MHz)
2
4
8
10
16
20
Bit Rate
(bits/s)
n
N
n
N
n
N
n
N
n
N
n
N
110
3
70
—
—
—
—
—
—
—
—
—
—
250
2
124
2
249
3
124
—
—
3
249
—
—
500
1
249
2
124
2
249
—
—
3
124
—
—
1k
1
124
1
249
2
124
—
—
2
249
—
—
2.5k
0
199
1
99
1
199
1
249
2
99
2
124
5k
0
99
0
199
1
99
1
124
1
199
1
249
10k
0
49
0
99
0
199
0
249
1
99
1
124
25k
0
19
0
39
0
79
0
99
0
159
0
199
50k
0
9
0
19
0
39
0
49
0
79
0
99
100k
0
4
0
9
0
19
0
24
0
39
0
49
250k
0
1
0
3
0
7
0
9
0
15
0
19
500k
0
0√
0
1
0
3
0
4
0
7
0
9
0
0*
0
1
—
—
0
3
0
4
—
—
0
0*
—
—
0
1
—
—
0
0*
1M
2.5M
5M
Note Settings with an error of 1% or less are recommended.
Blank: No setting available
—:
Setting possible, but error occurs
√:
Continuous transmit/receive not possible
The BRR setting is calculated as follows:
Asynchronous mode
N = [φ/(64 × 2 2n – 1 × B)] × 10 6 – 1
Clocked synchronous mode
N = [φ/(8 × 2 2n – 1 × B)] × 10 6 – 1
B: bit rate (bit/s)
N: BRR setting for baud rate generator (0 ≤ N ≤ 255)
φ: φ frequency (MHz)
n: baud rate generator clock source (n = 0, 1, 2, 3)
For the clock sources and values of n, see table 13.5.
RENESAS 359
SMR Settings
n
Clock Source
CKS1
CKS0
0
φ
0
0
1
φ/4
0
1
2
φ/16
1
0
3
φ/24
1
1
Find the bit rate error for the asynchronous mode by the following formula.
Error (%) = {(φ × 10 6)/[(N + 1) × B × 64 × 22n – 1 ] – 1 } × 100
RENESAS 360
Table 13.5 indicates the maximum bit rates in the asynchronous mode when the baud rate
generator is being used. Tables 13.6 and 13.7 show the maximum rates for external clock input.
Table 13.5 Maximum Bit Rates for Various Frequencies with Baud Rate Generator
(Asynchronous Mode)
Settings
φ (MHz)
Maximum Bit Rate (bits/s)
n
N
2
62500
0
0
2.097152
65536
0
0
2.4576
76800
0
0
3
93750
0
0
3.6864
115200
0
0
4
125000
0
0
4.9152
153600
0
0
5
156250
0
0
6
187500
0
0
6.144
192000
0
0
7.3728
230400
0
0
8
250000
0
0
9.8304
307200
0
0
10
312500
0
0
12
375000
0
0
12.288
384000
0
0
14
437500
0
0
14.7456
460800
0
0
16
500000
0
0
17.2032
537600
0
0
18
562500
0
0
19.6608
614400
0
0
20
625000
0
0
RENESAS 361
Table 13.6 Maximum Bit Rates during External Clock Input (Asynchronous Mode)
φ (MHz)
External Input Clock (MHz)
Maximum Bit Rate (bits/s)
2
0.5000
31250
2.097152
0.5243
32768
2.4576
0.6144
38400
3
0.7500
46875
3.6864
0.9216
57600
4
1.0000
62500
4.9152
1.2288
76800
5
1.2500
78125
6
1.5000
93750
6.144
1.5360
96000
7.3728
1.8432
115200
8
2.0000
125000
9.8304
2.4576
153600
10
2.5000
156250
12
3.0000
187500
12.288
3.0720
192000
14
3.5000
218750
14.7456
3.6834
230400
16
4.0000
250000
17.2032
4.3008
268800
18
4.5000
281250
19.6608
4.9152
307200
20
5.0000
312500
RENESAS 362
Table 13.7 Maximum Bit Rates during External Clock Input (Clocked Synchronous Mode)
φ (MHz)
External Input Clock (MHz)
Maximum Bit Rate (bits/s)
2
0.3333
333333.3
4
0.6667
666666.7
6
1.0000
1000000.0
8
1.3333
1333333.3
10
1.6667
1666666.7
12
2.0000
2000000.0
14
2.3333
2333333.3
16
2.6667
2666666.7
18
3.0000
3000000.0
20
3.3333
3333333.3
13.3
Operation
13.3.1
Overview
The SCI has an asynchronous mode in which characters are synchronized individually, and a
clocked synchronous mode in which communication is synchronized with clock pulses. Serial
communication is possible in either mode. Asynchronous/clocked synchronous mode and the
communication format are selected in the serial mode register (SMR), as shown in table 13.8. The
SCI clock source is selected by the C/A bit in the serial mode register (SMR) and the CKE1 and
CKE0 bits in the serial control register (SCR), as shown in table 13.9.
Asynchronous Mode:
•
•
•
•
Data length is selectable: seven or eight bits.
Parity and multiprocessor bits are selectable. So is the stop bit length (one or two bits). The
preceding selections constitute the communication format and character length.
In receiving, it is possible to detect framing errors (FER), parity errors (PER), overrun errors
(ORER), and the break state.
An internal or external clock can be selected as the SCI clock source.
 When an internal clock is selected, the SCI operates using the on-chip baud rate generator,
and can output a serial clock signal with a frequency matching the bit rate.
 When an external clock is selected, the external clock input must have a frequency 16 times
the bit rate. (The on-chip baud rate generator is not used.)
RENESAS 363
Clocked Synchronous Mode:
•
•
•
The communication format has a fixed eight-bit data length.
In receiving, it is possible to detect overrun errors (ORER).
An internal or external clock can be selected as the SCI clock source.
 When an internal clock is selected, the SCI operates using the on-chip baud rate generator,
and outputs a serial clock signal to external devices.
 When an external clock is selected, the SCI operates on the input serial clock. The on-chip
baud rate generator is not used.
Table 13.8 Serial Mode Register Settings and SCI Communication Formats
SMR Settings
SCI Communication Format
Mode
Bit 7: Bit 6: Bit 5: Bit 2: Bit 3:
C/A
CHR PE
MP
STOP
Data
Parity
Length Bit
MultiproStop Bit
cessor Bit Length
Asynchronous
0
8-bit
Absent
0
0
0
0
Absent
1
1
2 bits
0
Present
1 bit
1
1
0
0
2 bits
7-bit
Absent
1 bit
1
1
2 bits
0
Present
1 bit
1
Asynchronous
(multiprocessor
format)
Clocked
synchronous
0
1
1
*
*
1
0
*
1
*
0
*
1
*
*
*
Note: Asterisks (*) in the table indicate don’t-care bits.
RENESAS 364
1 bit
2 bits
8-bit
Absent
Present
1 bit
2 bits
7-bit
1 bit
2 bits
8-bit
Absent
None
Table 13.9
SMR and SCR Settings and SCI Clock Source Selection
SMR
Mode
Bit 7:
C/A
Asynchronous 0
mode
SCR Settings
SCI Transmit/Receive Clock
Bit 1:
CKE1
Bit 0:
CKE0
Clock Source SCK Pin Function*
0
0
Internal
1
1
0
SCI does not use the SCK pin
Outputs a clock with frequency
matching the bit rate
External
Inputs a clock with frequency 16
times the bit rate
Internal
Outputs the serial clock
External
Inputs the serial clock
1
Clocked
synchronous mode
1
0
0
1
1
0
1
Note: Select the function in combination with the pin function controller (PFC).
RENESAS 365
13.3.2
Operation in Asynchronous Mode
In the asynchronous mode, each transmitted or received character begins with a start bit and ends
with a stop bit. Serial communication is synchronized one character at a time.
The transmitting and receiving sections of the SCI are independent, so full duplex communication
is possible. The transmitter and receiver are both double buffered, so data can be written and read
while transmitting and receiving are in progress, enabling continuous transmitting and receiving.
Figure 13.2 shows the general format of asynchronous serial communication. In asynchronous
serial communication, the communication line is normally held in the mark (high) state. The SCI
monitors the line and starts serial communication when the line goes to the space (low) state,
indicating a start bit. One serial character consists of a start bit (low), data (LSB first), parity bit
(high or low), and stop bit (high), in that order.
When receiving in the asynchronous mode, the SCI synchronizes on the falling edge of the start
bit. The SCI samples each data bit on the eighth pulse of a clock with a frequency 16 times the bit
rate. Receive data is latched at the center of each bit.
1
Serial
data
(LSB)
0
D0
(MSB)
D1
D2
D3
D4
D5
Start
bit
D6
D7
Idling (mark state)
1
0/1
1
1
Parity
bit
Stop
bit
1 or
no bit
1 or
2 bits
Transmit/receive data
1 bit
7 or 8 bits
One unit of communication (characters or frames)
Figure 13.2 Data Format in Asynchronous Communication (Example: 8-bit data with
parity and two stop bits)
RENESAS 366
Transmit/Receive Formats: Table 13.10 shows the 12 communication formats that can be
selected in the asynchronous mode. The format is selected by settings in the serial mode register
(SMR).
Table 13.10
Serial Communication Formats (asynchronous mode)
SMR Bits
CHR PE
MP
STOP
1
2
3
4
5
6
7
8
9
10
11
12
0
0
0
0
START
8-Bit data
STOP
0
0
0
1
START
8-Bit data
STOP STOP
0
1
0
0
START
8-Bit data
P
STOP
0
1
0
1
START
8-Bit data
P
STOP STOP
1
0
0
0
START
7-Bit data
STOP
1
0
0
1
START
7-Bit data
STOP STOP
1
1
0
0
START
7-Bit data
P
STOP
1
1
0
1
START
7-Bit data
P
STOP STOP
0
—
1
0
START
8-Bit data
MPB
STOP
0
—
1
1
START
8-Bit data
MPB
STOP STOP
1
—
1
0
START
7-Bit data
MPB
STOP
1
—
1
1
START
7-Bit data
MPB
STOP STOP
—: Don't care bits.
Note: START: Start bit
STOP: Stop bit
P: Parity bit
MPB: Multiprocessor bit
Clock: An internal clock generated by the on-chip baud rate generator or an external clock input
from the SCK pin can be selected as the SCI transmit/receive clock. The clock source is selected
by the C/A bit in the serial mode register (SMR) and bits CKE1 and CKE0 in the serial control
register (SCR) (table 13.9).
When an external clock is input at the SCK pin, it must have a frequency equal to 16 times the
RENESAS 367
desired bit rate.
When the SCI operates on an internal clock, it can output a clock signal at the SCK pin. The
frequency of this output clock is equal to the bit rate. The phase is aligned as in figure 13.3 so that
the rising edge of the clock occurs at the center of each transmit data bit.
0
D0
D1
D2
D3
D4
D5
D6
D7
0/1
1
1
1 frame
Figure 13.3 Phase Relationship Between Output Clock and Serial Data (asynchronous
mode)
Transmitting and Receiving Data (SCI initialization (asynchronous mode)): Before
transmitting or receiving, software must clear the TE and RE bits to 0 in the serial control register
(SCR), then initialize the SCI as follows.
When changing the communication mode or format, always clear the TE and RE bits to 0 before
following the procedure given below. Clearing TE to 0 sets TDRE to 1 and initializes the transmit
shift register (TSR). Clearing RE to 0, however, does not initialize the RDRF, PER, FER, and
ORER flags and receive data register (RDR), which retain their previous contents.
When an external clock is used, the clock should not be stopped during initialization or subsequent
operation. SCI operation becomes unreliable if the clock is stopped.
Figure 13.4 is a sample flowchart for initializing the SCI. The procedure for initializing the SCI is
as follows:
1. Select the communication format in the serial mode register (SMR).
2. Write the value corresponding to the bit rate in the bit rate register (BRR) unless an external
clock is used.
3. Select the clock source in the serial control register (SCR). Leave RIE, TIE, TEIE, MPIE, TE
and RE cleared to 0. If clock output is selected in asynchronous mode, clock output starts
immediately after the setting is made to SCR.
4. Wait for at least the interval required to transmit or receive one bit, then set TE or RE in the
serial control register (SCR) to 1. Also set RIE, TIE, TEIE and MPIE as necessary. Setting TE
or RE enables the SCI to use the TxD or RxD pin. The initial states are the mark transmit state,
and the idle receive state (waiting for a start bit).
RENESAS 368
Start of initialization
Clear TE and RE bits to 0 in SCR
Select communication format in SMR (1)
Set value in BRR
(2)
Set CKE1 and CKE0 bits in SCR
(leaving TE and RE cleared to 0)
(3)
Wait
1-bit interval elapsed?
No
Yes
Set TE or RE to 1 in SCR; Set RIE,
TIE, TEIE, and MPIE as necessary
(4)
End
Note:
Circled numbers refer to the preceding procedure.
Figure 13.4 Sample Flowchart for SCI Initialization
Transmitting Serial Data (asynchronous mode): Figure 13.5 shows a sample flowchart for
transmitting serial data. The procedure for transmitting serial data is as follows:
1. SCI initialization: select the TxD pin function with the PFC.
2. SCI status check and transmit data write: read the serial status register (SSR), check that the
TDRE bit is 1, then write transmit data in the transmit data register (TDR) and clear TDRE to
0.
3. To continue transmitting serial data: read the TDRE bit to check whether it is safe to write (1);
if so, write data in TDR, then clear TDRE to 0. When the DMAC is started by a transmit-dataempty interrupt request (TXI) to write data in TDR, the TDRE bit is checked and cleared
automatically.
4. To output a break signal at the end of serial transmission: set the DR bit to 0 (I/O data port
register), then clear TE to 0 in SCR and set the TxD pin function as output port with the PFC.
RENESAS 369
Initialize
(1)
Start transmitting
Read TDRE bit in SSR
(2)
No
TDRE = 1?
Yes
Write transmit data in TDR and
clear TDRE bit to 0 in SSR
(3)
All data transmitted?
No
Yes
Read TEND bit in SSR
No
TEND = 1?
Yes
No
Output break signal?
(4)
Yes
Set DR = 0
Clear TE bit of SCR to 0;
Select theTxD pin function
as an output port with the PFC
Transmission ends
Note:
Circled numbers refer to the preceding procedure.
Figure 13.5 Sample Flowchart for Transmitting Serial Data
RENESAS 370
In transmitting serial data, the SCI operates as follows:
1. The SCI monitors the TDRE bit in the SSR. When TDRE is cleared to 0, the SCI recognizes
that the transmit data register (TDR) contains new data, and loads this data from the TDR into
the transmit shift register (TSR).
2. After loading the data from the TDR into the TSR, the SCI sets the TDRE bit to 1 and starts
transmitting. If the transmit-data-empty interrupt enable bit (TIE) is set to 1 in the SCR, the
SCI requests a transmit-data-empty interrupt (TXI) at this time.
Serial transmit data is transmitted in the following order from the TxD pin:
1. Start bit: one 0 bit is output.
2. Transmit data: seven or eight bits of data are output, LSB first.
3. Parity bit or multiprocessor bit: one parity bit (even or odd parity) or one multiprocessor bit is
output. Formats in which neither a parity bit nor a multiprocessor bit is output can also be
selected.
4. Stop bit: one or two 1 bits (stop bits) are output.
5. Mark state: output of 1 bits continues until the start bit of the next transmit data.
6. The SCI checks the TDRE bit when it outputs the stop bit. If TDRE is 0, the SCI loads new
data from the TDR into the TSR, outputs the stop bit, then begins serial transmission of the
next frame. If TDRE is 1, the SCI sets the TEND bit to 1 in the SSR, outputs the stop bit, then
continues output of 1 bits in the mark state. If the transmit-end interrupt enable bit (TEIE) in
the SCR is set to 1, a transmit-end interrupt (TEI) is requested.
Figure 13.6 shows an example of SCI transmit operation in the asynchronous mode.
RENESAS 371
1
Serial
data
Start
bit
0
Parity Stop Start
bit bit
bit
Data
D0
D1
D7
0/1
1
0
Data
D0
D1
Parity Stop
1
bit
bit
D7 0/1 1 Idle (mark
state)
TDRE
TEND
TXI
TXI interrupt
request handler writes
data in TDR
and clears
TDRE to 0
TXI
request
TEI request
1 frame
Figure 13.6 Example of SCI Transmit Operation in Asynchronous Mode (8-bit data with
parity and one stop bit)
Receiving Serial Data (asynchronous mode): Figure 13.7 shows a sample flowchart for receiving
serial data. The procedure for receiving serial data is listed below.
1. SCI initialization: select the RxD pin function with the PFC.
2. Receive error handling and break detection: if a receive error occurs, read the ORER, PER and
FER bits of the SSR to identify the error. After executing the necessary error handling, clear
ORER, PER and FER all to 0. Receiving cannot resume if ORER, PER or FER remains set to
1. When a framing error occurs, the RxD pin can be read to detect the break state.
3. SCI status check and receive data read: read the serial status register (SR), check that RDRF is
set to 1, then read receive data from the receive data register (RDR) and clear RDRF to 0. The
RXI interrupt can also be used to determine if the RDRF bit has changed from 0 to 1.
4. To continue receiving serial data: read RDRF and RDR, and clear RDRF to 0 before the stop
bit of the current frame is received. If the DMAC is started by a receive-data-full interrupt
(RXI) to read RDR, the RDRF bit is cleared automatically so this step is unnecessary.
RENESAS 372
Initialization
(1)
Start receiving
Read the ORER, PER, and
FER bits of the SSR
PER, FER, ORER = 1?
Yes
(2)
No
Read the RDRF bit of the SSR
No
Error handling
(3)
RDRF = 1?
Yes
Read the RDR's receive data
(4)
and clear the SSR's RDRF bit to 0
No
Total count received?
Yes
Clear the RE bit of the SCR to 0
Reception ends
Note:
Circled numbers refer to the preceding procedure.
Figure 13.7 Sample Flowchart for Receiving Serial Data
RENESAS 373
Start of error handling
No
ORER = 1?
Yes
Overrun error handling
No
FER = 1?
Yes
Break?
Yes
No
Framing error handling
No
Clear RE bit to 0 in SCR
PER = 1?
Yes
Parity error handling
Clear ORER, PER, and FER to 0 in SSR
End
Note:
Circled numbers refer to the preceding procedure.
Figure 13.7 Sample Flowchart for Receiving Serial Data (cont)
RENESAS 374
In receiving, the SCI operates as follows:
1. The SCI monitors the receive data line. When it detects a start bit (0), the SCI synchronizes
internally and starts receiving.
2. Receive data is shifted into the RSR in order from the LSB to the MSB.
3. The parity bit and stop bit are received. After receiving these bits, the SCI makes the following
checks:
a. Parity check: The number of 1s in the receive data must match the even or odd parity
setting of the O/E bit in the SMR.
b. Stop bit check: The stop bit value must be 1. If there are two stop bits, only the first stop bit
is checked.
c. Status check: RDRF must be 0 so that receive data can be loaded from the RSR into the
RDR.
If these checks all pass, the SCI sets RDRF to 1 and stores the received data in the RDR. If one
of the checks fails (receive error), the SCI operates as indicated in table 13.11.
Note:
When a receive error flag is set, further receiving is disabled. The RDRF bit is not set
to 1. Be sure to clear the error flags.
4. After setting RDRF to 1, if the receive-data-full interrupt enable bit (RIE) is set to 1 in the
SCR, the SCI requests a receive-data-full interrupt (RXI). If one of the error flags (ORER,
PER, or FER) is set to 1 and the receive-data-full interrupt enable bit (RIE) in the SCR is also
set to 1, the SCI requests a receive-error interrupt (ERI).
Figure 13.8 shows an example of SCI receive operation in the asynchronous mode.
Table 13.11 Receive Error Conditions and SCI Operation
Receive Error
Abbreviation
Condition
Data Transfer
Overrun error
ORER
Receiving of next data ends while
RDRF is still set to 1 in SSR
Receive data not loaded
from RSR into RDR
Framing error
FER
Stop bit is 0
Receive data loaded from
RSR into RDR
Parity error
PER
Parity of receive data differs from
even/odd parity setting in SMR
Receive data loaded from
RSR into RDR
RENESAS 375
1
Start
bit
Serial
data
0
Parity Stop Start
bit bit
bit
Data
D0
D1
D7
1
0/1
0
Parity Stop
bit
bit
Data
D0
D1
D7
0/1
0
1
Idle (mark
state)
TDRE
RXI request
FER
1 frame
RXI interrupt handler
reads data in RDR
and clears RDRF to 0
Framing error,
ERI request
Figure 13.8 Example of SCI Receive Operation (8-bit data with parity and one stop bit)
13.3.3
Multiprocessor Communication
The multiprocessor communication function enables several processors to share a single serial
communication line. The processors communicate in the asynchronous mode using a format with
an additional multiprocessor bit (multiprocessor format).
In multiprocessor communication, each receiving processor is addressed by a unique ID. A serial
communication cycle consists of an ID-sending cycle that identifies the receiving processor, and a
data-sending cycle. The multiprocessor bit distinguishes ID-sending cycles from data-sending
cycles. The transmitting processor starts by sending the ID of the receiving processor with which it
wants to communicate as data with the multiprocessor bit set to 1. Next the transmitting processor
sends transmit data with the multiprocessor bit cleared to 0.
Receiving processors skip incoming data until they receive data with the multiprocessor bit set to
1. When they receive data with the multiprocessor bit set to 1, receiving processors compare the
data with their IDs. The receiving processor with a matching ID continues to receive further
incoming data. Processors with IDs not matching the received data skip further incoming data
until they again receive data with the multiprocessor bit set to 1. Multiple processors can send and
receive data in this way.
Figure 13.9 shows the example of communication among processors using the multiprocessor
format.
RENESAS 376
Transmitting
processor
Serial communication line
Serial
data
Receiving
processor A
Receiving
processor B
Receiving
processor C
Receiving
processor D
(ID = 01)
(ID = 02)
(ID = 03)
(ID = 04)
H'01
H'AA
(MPB = 1)
ID-sending cycle:
receiving processor address
(MPB = 0)
Data-sending cycle:
data sent to receiving
processor specified by ID
MPB: multiprocessor bit
Figure 13.9 Example of Communication among Processors Using Multiprocessor Format
(sending data H'AA to receiving processor A)
Communication Formats: Four formats are available. Parity-bit settings are ignored when the
multiprocessor format is selected. For details see table 13.8.
Clock: See the description in the asynchronous mode section.
Transmitting Multiprocessor Serial Data: Figure 13.10 shows a sample flowchart for
transmitting multiprocessor serial data. The procedure for transmitting multiprocessor serial data is
listed below.
1. SCI initialization: select the TxD pin function with the PFC.
2. SCI status check and transmit data write: read the serial status register (SSR), check that the
TDRE bit is 1, then write transmit data in the transmit data register (TDR). Also set MPBT
(multiprocessor bit transfer) to 0 or 1 in SSR. Finally, clear TDRE to 0.
3. To continue transmitting serial data: read the TDRE bit to check whether it is safe to write (1);
if so, write data in TDR, then clear TDRE to 0. When the DMAC is started by a transmit-dataempty interrupt request (TXI) to write data in TDR, the TDRE bit is checked and cleared
automatically.
4. To output a break signal at the end of serial transmission: set the DR bit to 0 (I/O data port
register), then clear TE to 0 in SCR and set the TxD pin function as output port with the PFC.
RENESAS 377
Initialize
(1)
Start transmitting
Read TDRE bit in SSR
TDRE = 1?
(2)
No
Yes
Write transmit data in TDR
and set MPBT in SSR
Clear TDRE bit to 0
All data transmitted?
No
(3)
Yes
Read TEND bit in SSR
TEND = 1?
No
Yes
Output break signal?
No
(4)
Yes
Set DR = 0
Clear TE bit to 0 in SCR;
select theTxD pin function as
an output port with the PFC
End
Note:
Circled numbers refer to the preceding procedure.
Figure 13.10 Sample Flowchart for Transmitting Multiprocessor Serial Data
In transmitting serial data, the SCI operates as follows:
RENESAS 378
1. The SCI monitors the TDRE bit in the SSR. When TDRE is cleared to 0 the SCI recognizes
that the transmit data register (TDR) contains new data, and loads this data from the TDR into
the transmit shift register (TSR).
2. After loading the data from the TDR into the TSR, the SCI sets the TDRE bit to 1 and starts
transmitting. If the transmit-data-empty interrupt enable bit (TIE) in the SCR is set to 1, the
SCI requests a transmit-data-empty interrupt (TXI) at this time.
Serial transmit data is transmitted in the following order from the TxD pin (figure 13.11):
1.
2.
3.
4.
5.
6.
Start bit: one 0 bit is output.
Transmit data: seven or eight bits are output, LSB first.
Multiprocessor bit: one multiprocessor bit (MPBT value) is output.
Stop bit: one or two 1 bits (stop bits) are output.
Mark state: output of 1 bits continues until the start bit of the next transmit data.
The SCI checks the TDRE bit when it outputs the stop bit. If TDRE is 0, the SCI loads data
from the TDR into the TSR, outputs the stop bit, then begins serial transmission of the next
frame. If TDRE is 1, the SCI sets the TEND bit in the SSR to 1, outputs the stop bit, then
continues output of 1 bits in the mark state. If the transmit-end interrupt enable bit (TEIE) in
the SCR is set to 1, a transmit-end interrupt (TEI) is requested at this time.
1
Serial
data
Start
bit
0
Multiprocessor Stop Start
bit bit
bit
Data
D0
D1
D7
0/1
1
0
D0
Multiprocessor Stop
1
Data
bit
bit
D1
D7 0/1 1 Idle (mark
state)
TDRE
TEND
TXI
TXI interrupt
request handler writes
data in TDR and
clears TDRE to 0
TXI
request
TEI
request
1 frame
Figure 13.11 Example of SCI Multiprocessor Transmit Operation (8-bit data with
multiprocessor bit and one stop bit)
RENESAS 379
Receiving Multiprocessor Serial Data: Figure 13.12 shows a sample flowchart for receiving
multiprocessor serial data. The procedure for receiving multiprocessor serial data is listed below.
1. SCI initialization: select the RxD pin function with the PFC.
2. ID receive cycle: set the MPIE bit in the serial control register (SCR) to 1.
3. SCI status check and compare to ID reception: read the serial status register (SSR), check that
RDRF is set to 1, then read data from the receive data register (RDR) and compare with the
processor's own ID. If the ID does not match the receive data, set MPIE to 1 again and clear
RDRF to 0. If the ID matches the receive data, clear RDRF to 0.
4. Receive error handling and break detection: if a receive error occurs, read the ORER and FER
bits in SSR to identify the error. After executing the necessary error handling, clear both
ORER and FER to 0. Receiving cannot resume if ORER or FER remain set to 1. When a
framing error occurs, the RxD pin can be read to detect the break state.
5. SCI status check and data receiving: read SSR, check that RDRF is set to 1, then read data
from the receive data register (RDR).
RENESAS 380
Initialization
(1)
Start receiving
Set the MPIE bit of the SCR to 1
(2)
Read the ORER and FER
bits of the SSR
FER = 1? or ORER = 1?
No
Read the RDRF bit of the SSR
No
Yes
(3)
RDRF = 1?
Yes
Read the receive data of the RDR
No
Own ID?
Yes
Read the ORER and FER
bits of the SSR
FER = 1? or ORER = 1?
Yes
No
Read the SSR's RDRF bit
RDRF = 1?
(5)
No
Yes
Read the RDR's receive data
No
Total count received?
Yes
Clear the RE bit of the SCR to 0
(4)
Error handling
Reception ends
Figure 13.12 Sample Flowchart for Receiving Multiprocessor Serial Data
RENESAS 381
Start of error handling
No
ORER = 1?
Yes
Overrun error handling
No
FER = 1?
Yes
Break?
Yes
No
Framing error handling
Clear RE bit to 0 in SCR
Clear ORER and FER to 0 in SSR
End
Figure 13.12 Sample Flowchart for Receiving Multiprocessor Serial Data (cont)
RENESAS 382
Figure 13.13 shows an example of SCI receive operation using a multiprocessor format.
1
Serial
data
Stop Start
Data ID1 MPB bit bit
Start
bit
0
D0
D1
D7
1
1
0
Data 1
D0
D1
D7
Stop
1
MPB bit
0
1 Idle (mark
state)
MPB
MPIE
RDRF
RDR
value
RXI request,
(multiprocessor
interrupt) MPIE = 0
ID1
RXI interrupt
handler reads data
in RDR and clears
RDRF to 0
Not own ID,
so MPIE is
set to 1 again
No RXI
interrupt,
RDR maintains
state
Figure 13.13 Example of SCI Receive Operation (own ID does not match data) (8-bit data
with multiprocessor bit and one stop bit)
RENESAS 383
1
Serial
data
Start
bit
0
Stop Start
Data ID1 MPB bit bit
D0
D1
D7
1
1
0
Data 1
D0
D1
D7
Stop
1
MPB bit
0
1 Idle (mark
state)
MPB
MPIE
RDRF
RDR
value
RXI request,
(multiprocessor
interrupt) MPIE = 0
ID1
RXI interrupt
handler reads data
in RDR and clears
RDRF to 0
Not own ID,
so MPIE is
set to 1 again
No RXI
interrupt,
RDR maintains
state
Figure 13.13 Example of SCI Receive Operation (own ID matches data) (8-bit data with
multiprocessor bit and one stop bit) (cont)
13.3.4
Clocked Synchronous Operation
In the clocked synchronous mode, the SCI transmits and receives data in synchronization with
clock pulses. This mode is suitable for high-speed serial communication.
The SCI transmitter and receiver share the same clock but are otherwise independent, so full
duplex communication is possible. The transmitter and receiver are also double buffered, so
continuous transmitting or receiving is possible by reading or writing data while transmitting or
receiving is in progress.
Figure 13.14 shows the general format in clocked synchronous serial communication.
RENESAS 384
Transfer direction
One unit (character or frame) of serial data
*
*
Serial clock
LSB
Serial data
Note:
Bit 0
MSB
Bit 1
Bit 2
Bit 3
Bit 4
Bit 5
Bit 6
Bit 7
High except in continuous transmitting or receiving.
Figure 13.14 Data Format in Clocked Synchronous Communication
In clocked synchronous serial communication, each data bit is output on the communication line
from one falling edge of the serial clock to the next. Data are guaranteed valid at the rising edge of
the serial clock. In each character, the serial data bits are transmitted in order from the LSB (first)
to the MSB (last). After output of the MSB, the communication line remains in the state of the
MSB. In the clocked synchronous mode, the SCI transmits or receives data by synchronizing with
the falling edge of the serial clock.
Communication Format: The data length is fixed at eight bits. No parity bit or multiprocessor bit
can be added.
Clock: An internal clock generated by the on-chip baud rate generator or an external clock input
from the SCK pin can be selected as the SCI transmit/receive clock. The clock source is selected
by the C/A bit in the serial mode register (SMR) and bits CKE1 and CKE0 in the serial control
register (SCR). See table 13.6.
When the SCI operates on an internal clock, it outputs the clock signal at the SCK pin. Eight clock
pulses are output per transmitted or received character. When the SCI is not transmitting or
receiving, the clock signal remains in the high state.
Figure 13.15 shows an example of SCI transmit operation. In transmitting serial data, the SCI
operates as follows.
1. The SCI monitors the TDRE bit in the SSR. When TDRE is cleared to 0 the SCI recognizes
that the transmit data register (TDR) contains new data, and loads this data from the TDR into
the transmit shift register (TSR).
2. After loading the data from the TDR into the TSR, the SCI sets the TDRE bit to 1 and starts
transmitting. If the transmit-data-empty interrupt enable bit (TIE) in the SCR is set to 1, the
SCI requests a transmit-data-empty interrupt (TXI) at this time.
If clock output is selected, the SCI outputs eight serial clock pulses. If an external clock source
RENESAS 385
is selected, the SCI outputs data in synchronization with the input clock. Data are output from
the TxD pin in order from the LSB (bit 0) to the MSB (bit 7).
3. The SCI checks the TDRE bit when it outputs the MSB (bit 7). If TDRE is 0, the SCI loads
data from the TDR into the TSR, transmits the MSB, then begins serial transmission of the
next frame. If TDRE is 1, the SCI sets the TEND bit in the SSR to 1, transmits the MSB, then
holds the transmit data pin (TxD) in the MSB state. If the transmit-end interrupt enable bit
(TEIE) in the SCR is set to 1, a transmit-end interrupt (TEI) is requested at this time.
4. After the end of serial transmission, the SCK pin is held in the high state.
Transmit direction
Serial clock
LSB
Serial data
Bit 0
MSB
Bit 1
Bit 7
Bit 0
Bit 1
Bit 6
Bit 7
TDRE
TEND
TXI
request
TXI interrupt
handler writes
data in TDR and
clears TDRE to 0
TXI
request
TEI
request
1 frame
Figure 13.15 Example of SCI Transmit Operation
Transmitting and Receiving Data: SCI Initialization (clocked synchronous mode): Before
transmitting or receiving, software must clear the TE and RE bits to 0 in the serial control register
(SCR), then initialize the SCI as follows.
When changing the communication mode or format, always clear the TE and RE bits to 0 before
following the procedure given below. Clearing TE to 0 sets TDRE to 1 and initializes the transmit
shift register (TSR). Clearing RE to 0, however, does not initialize the RDRF, PER, FER, and
ORER flags and receive data register (RDR), which retain their previous contents.
Figure 13.16 is a sample flowchart for initializing the SCI. The procedure for initializing the SCI
is listed below.
RENESAS 386
1. Select the communication format in the serial mode register (SMR).
2. Write the value corresponding to the bit rate in the bit rate register (BRR) unless an external
clock is used.
3. Select the clock source in the serial control register (SCR). Leave RIE, TIE, TEIE, MPIE, TE
and RE cleared to 0.
4. Wait for at least the interval required to transmit or receive one bit, then set TE or RE in the
serial control register (SCR) to 1. Also set RIE, TIE, TEIE and MPIE. Setting the
corresponding bit of the pin function controller, TE and RE enables the SCI to use the TxD or
RxD pin.
Start of initialization
Clear TE and RE bits to 0 in SCR
Select communication format in SMR (1)
Set value in BRR
(2)
Set RIE, TIE, TEIE, MPIE, CKE1,
and CKE0 bits in SCR
(leaving TE and RE cleared to 0)
(3)
Wait
1-bit interval elapsed?
No
Yes
Set TE or RE to 1 in SCR;
Set RIE, TIE, TEIE, and MPIE
(4)
End
Note:
High except in continuous transmitting or receiving.
Figure 13.16 Sample Flowchart for SCI Initialization
RENESAS 387
Transmitting Serial Data (clocked synchronous mode): Figure 13.17 shows a sample flowchart
for transmitting serial data. The procedure for transmitting serial data is listed below.
1. SCI initialization: select the TxD pin function with the PFC.
2. SCI status check and transmit data write: read the serial status register (SSR), check that the
TDRE bit is 1, then write transmit data in the transmit data register (TDR) and clear TDRE to
0. The RXI interrupt can also be used to determine if the RDRF bit has changed from 0 to 1.
3. To continue transmitting serial data: read the TDRE bit to check whether it is safe to write (1);
if so, write data in TDR, then clear TDRE to 0. When the DMAC is started by a transmit-dataempty interrupt request (TXI) to write data in TDR, the TDRE bit is checked and cleared
automatically.
RENESAS 388
Initialize
(1)
Start transmitting
Read TDRE bit in SSR
TDRE = 1?
(2)
No
Yes
Write transmit data in TDR and
clear TDRE bit to 0 in SSR
All data transmitted?
No
(3)
Yes
Read TEND bit in SSR
TEND = 1?
No
Yes
Clear TE bit SCR to 0
Transmission ends
Figure 13.17 Sample Flowchart for Serial Transmitting
Receiving Serial Data (clocked synchronous mode): Figure 13.18 shows a sample flowchart for
receiving serial data. When switching from the asynchronous mode to the clocked synchronous
mode, make sure that ORER, PER, and FER are cleared to 0. If PER or FER is set to 1, the RDRF
bit will not be set and both transmitting and receiving will be disabled. Figure 13.19 shows an
example of SCI recieve operation.
The procedure for recieving serial data is listed below.
1. SCI initialization: select the RxD pin function with the PFC.
2. Receive error handling: if a receive error occurs, read the ORER bit in SSR to identify the
error. After executing the necessary error handling, clear ORER to 0. Transmitting/receiving
RENESAS 389
cannot resume if ORER remains set to 1.
3. SCI status check and receive data read: read the serial status register (SSR), check that RDRF
is set to 1, then read receive data from the receive data register (RDR) and clear RDRF to 0.
The RXI interrupt can also be used to determine if the RDRF bit has changed from 0 to 1.
4. To continue receiving serial data: read RDR, and clear RDRF to 0 before the frame MSB (bit
7) of the current frame is received. If the DMAC is started by a receive-data-full interrupt
(RXI) to read RDR, the RDRF bit is cleared automatically so this step is unnecessary.
Initialization
(1)
Start receiving
Read the ORER bit of the SSR
Yes
ORER = 1?
No
(2)
Error handling
Read RDRF bit of the SSR
No
(3)
RDRF = 1?
Yes
Read the RDR's receive data
(4)
and clear the SSR's RDRF bit to 0
No
Total count received?
Yes
Clear the RE bit of the SCR to 0
Reception ends
Figure 13.18 Sample Flowchart for Serial Receiving
RENESAS 390
Error handling
No
ORER = 1?
Yes
Overrun error handling
Clear ORER bit of SSR to 0
End
Figure 13.18 Sample Flowchart for Serial Receiving (cont)
Receive direction
Serial
clock
Serial
data
Bit 7
Bit 0
Bit 7
Bit 0
Bit 1
Bit 6
Bit 7
RDRF
ORER
RXI request
RXI interrupt handler
reads data in RDR
and clears RDRF to 0
RXI request
Overrun
error, ERI
request
1 frame
Figure 13.19 Example of SCI Receive Operation
In receiving, the SCI operates as follows:
1. The SCI synchronizes with serial clock input or output and initializes internally.
2. Receive data is shifted into the RSR in order from the LSB to the MSB. After receiving the
data, the SCI checks that RDRF is 0 so that receive data can be loaded from the RSR into the
RENESAS 391
RDR. If this check passes, the SCI sets RDRF to 1 and stores the received data in the RDR. If
the check does not pass (receive error), the SCI operates as indicated in table 13.8. When the
error flag is set to 1 and the RDRF bit is cleared to 0, the RDRF bit will not be set to 1 during
reception. When restarting the reception, make sure to clear the error flag to 0.
3. After setting RDRF to 1, if the receive-data-full interrupt enable bit (RIE) is set to 1 in the
SCR, the SCI requests a receive-data-full interrupt (RXI). If the ORER bit is set to 1 and the
receive-data-full interrupt enable bit (RIE) in the SCR is also set to 1, the SCI requests a
receive-error interrupt (ERI).
Transmitting and Receiving Serial Data Simultaneously (clocked synchronous mode): Figure
13.20 shows a sample flowchart for transmitting and receiving serial data simultaneously. The
procedure for transmitting and receiving serial data simultaneously is listed below.
1. SCI initialization: select the TxD and RxD pin function with the PFC.
2. SCI status check and transmit data write: read the serial status register (SSR), check that the
TDRE bit is 1, then write transmit data in the transmit data register (TDR) and clear TDRE to
0. The TXI interrupt can also be used to determine if the TDRE bit has changed from 0 to 1.
3. Receive error handling: if a receive error occurs, read the ORER bit in SSR to identify the
error. After executing the necessary error handling, clear ORER to 0. Transmitting/receiving
cannot resume if ORER remains set to 1.
4. SCI status check and receive data read: read the serial status register (SSR), check that RDRF
is set to 1, then read receive data from the receive data register (RDR) and clear RDRF to 0.
The RXI interrupt can also be used to determine if the RDRF bit has changed from 0 to 1.
5. To continue transmitting and receiving serial data: read the RDRF bit and RDR, and clear
RDRF to 0 before the frame MSB (bit 7) of the current frame is received. Also read the TDRE
bit to check whether it is safe to write (1); if so, write data in TDR, then clear TDRE to 0
before the MSB (bit 7) of the current frame is transmitted. When the DMAC is started by a
transmit-data-empty interrupt request (TXI) to write data in TDR, the TDRE bit is checked and
cleared automatically. When the DMAC is started by a receive-data-full interrupt (RXI) to read
RDR, the RDRF bit is cleared automatically.
RENESAS 392
Initialization
(1)
Start transmitting and receiving
Read TDRE bit of the SSR
No
(2)
TDRE = 1?
Yes
Write transmit data to the TDR
and clear the TDRE bit of the
SSR to 0
Read the ORER bit of the SSR
ORER = 1?
No
Read the SSR's RDRF bit
No
Yes
(3)
Error handling
(4)
RDRF = 1?
Yes
Read the receive data of the
RDR and clear the RDRF bit of
the SSR to 0
No
(5)
Total
count transmitted and
received?
Yes
Clear the TE and RE bits of
the SCR to 0
Transmitting/receiving ends
Note:
In switching from transmitting or receiving to simultaneous transmitting and receiving,
clear both TE and RE to 0, then set both TE and RE to 1.
Figure 13.20 Sample Flowchart for Serial Transmitting
RENESAS 393
13.4
SCI Interrupt Sources and the DMAC
The SCI has four interrupt sources in each channel: transmit-end (TEI), receive-error (ERI),
receive-data-full (RXI), and transmit-data-empty (TXI). Table 13.12 lists the interrupt sources and
indicates their priority. These interrupts can be enabled and disabled by the TIE, RIE, and TEIE
bits in the serial control register (SCR). Each interrupt request is sent separately to the interrupt
controller.
TXI is requested when the TDRE bit in the SSR is set to 1. TXI can start the direct memory access
controller (DMAC) to transfer data. TDRE is automatically cleared to 0 when the DMAC executes
the data transfer to the transmit data register (TDR).
RXI is requested when the RDRF bit in the SSR is set to 1. RXI can start the DMAC to transfer
data. RDRF is automatically cleared to 0 when the DMAC executes the data transfer to the receive
data register (RDR). ERI is requested when the ORER, PER, or FER bit in the SSR is set to 1. ERI
cannot start the DMAC.
TEI is requested when the TEND bit in the SSR is set to 1. TEI cannot start the DMAC. A TXI
interrupt indicates that transmit data writing is enabled. A TEI interrupt indicates that the transmit
operation is complete.
Table 13.12
SCI Interrupt Sources
Interrupt Source
Description
DMAC Availability
Priority
ERI
Receive error (ORER, PER, or FER)
No
High
RXI
Receive data full (RDRF)
Yes
↑
TXI
Transmit data empty (TDRE)
Yes
↓
TEND
Transmit end (TEND)
No
Low
13.5
Usage Notes
Note the following points when using the SCI.
TDR Write and TDRE Flags: The TDRE bit in the serial status register (SSR) is a status flag
indicating loading of transmit data from the TDR into the TSR. The SCI sets TDRE to 1 when it
transfers data from the TDR to the TSR. If new data is written in the TDR when TDRE is 0, the
old data stored in the TDR will be lost because these data have not yet been transferred to the
TSR. Before writing transmit data to the TDR, be sure to check that TDRE is set to 1.
Simultaneous Multiple Receive Errors: Table 13.13 indicates the state of the SSR status flags
when multiple receive errors occur simultaneously. When an overrun error occurs, the RSR
contents cannot be transferred to the RDR, so receive data is lost.
RENESAS 394
Table 13.13 SSR Status Flags and Transfer of Receive Data
Receive Error Status
RDRF
ORER
FER
PER
Receive Data
Transfer
RSR → RDR
Overrun error
1
1
0
0
X
Framing error
0
0
1
0
O
Parity error
0
0
0
1
O
Overrun error + framing error
1
1
1
0
X
Overrun error + parity error
1
1
0
1
X
Framing error + parity error
0
0
1
1
O
Overrun error + framing error + parity
error
1
1
1
1
X
SSR Status Flags
O:
X:
Receive data is transferred from RSR–RDR.
Receive data is not transferred from RSR–RDR.
Break Detection and Processing: Break signals can be detected by reading the RxD pin directly
when a framing error (FER) is detected. In the break state, the input from the RxD pin consists of
all 0s, so FER is set and the parity error flag (PER) may also be set. In the break state, the SCI
receiver continues to operate, so if the FER bit is cleared to 0, it will be set to 1 again.
Sending a Break Signal: When TE is cleared to 0 the TxD pin becomes an I/O port, the level and
direction (input or output) of which are determined by the data register (DR) of the I/O port and
the control register (CR) of the PFC. This feature can be used to send a break signal. The DR
value substitutes for the mark state until the PFC setting is performed. The DR bits should
therefore be set as an output port that outputs 1 beforehand. To send a break signal during serial
transmission, clear the DR bit to 0, and select output port as the TxD pin function by the PFC.
When TE is cleared to 0, the transmitter is initialized, regardless of its current state.
Receive Error Flags and Transmitter Operation (clocked synchronous mode only): When a
receive error flag (ORER, PER, or FER) is set to 1, the SCI will not start transmitting even if
TDRE is set to 1. Be sure to clear the receive error flags to 0 before starting to transmit. Note that
clearing RE to 0 does not clear the receive error flags.
Receive Data Sampling Timing and Receive Margin in the Asynchronous Mode: In the
asynchronous mode, the SCI operates on a base clock of 16 times the bit rate frequency. In
receiving, the SCI synchronizes internally with the falling edge of the start bit, which it samples on
the base clock. Receive data is latched on the rising edge of the eighth base clock pulse. See figure
13.21.
RENESAS 395
16 clocks
8 clocks
Internal
base clock
0 1 2 3 4 5 6 7 8 9 10111213 1415 0 1 2 3 4 5 6 7 8 9 10111213 1415 0 1 2 3 4 5
–7.5 clocks
Receive
data (RxD)
+7.5 clocks
Start bit
D0
D1
Synchronization
sampling
timing
Data
sampling
timing
Figure 13.21 Receive Data Sampling Timing in Asynchronous Mode
The receive margin in the asynchronous mode can therefore be expressed as in equation 1.
Equation 1:
M =
M:
N:
D:
L:
F:
0.5 –
1
2N
– (L – 0.5)F –
D
– 0.5
N
(1 + F ) × 100%
Receive margin (%)
Ratio of clock frequency to bit rate (N = 16)
Clock duty cycle (D = 0–1.0)
Frame length (L = 9–12)
Absolute deviation of clock frequency
From equation (1), if F = 0 and D = 0.5 the receive margin is 46.875%, as given by equation 2.
Equation 2:
D
M
= 0.5, F = 0
= (0.5 – 1/(2 × 16)) × 100%
= 46.875%
(2)
This is a theoretical value. A reasonable margin to allow in system designs is 20–30%.
Constraints on DMAC Use:
RENESAS 396
•
•
When using an external clock source for the serial clock, update the TDR with the DMAC, and
then after five system clocks or more elapse, input a transmit clock. If a transmit clock is input
in the first four system clocks after the TDR is written, an error may occur (figure 13.22).
Before reading the receive data register (RDR) with the DMAC, select the receive-data-full
interrupt of the SCI as a start-up source using the resource select bit (RS) in the channel control
register (CHCR).
SCK
t
TDRE
D0
Note:
D1
D2
D3
D4
D5
D6
D7
During external clock operation, an error may occur if t is 4φ or less.
Figure 13.22 Clocked Synchronous Transmitting Example with DMAC
Cautions for Clocked Synchronous External Clock Mode:
•
•
•
Set TE = RE = 1 only when the external clock SCI is 1.
Do not set TE = RE = 1 until at least 4 clocks after the external clock SCK has changed from 0
to 1.
When receiving, RDRF = 1 when RE is set to zero 2.5–3.5 clocks after the rise edge of the
RxD D7 bit SCK input, but it cannot be copied to RDR.
Caution for Clocked Synchronous Internal Clock Mode: When receiving, RDRF = 1 when RE
is set to zero 1.5 clocks after the rise edge of the RxD D7 bit SCK input, but it cannot be copied to
RDR.
RENESAS 397
Section 14 Pin Function Controller (PFC)
14.1
Overview
The pin function controller (PFC) is composed of registers for selecting the function of
multiplexed pins and the direction of input/output. The pin function and input/output direction can
be selected for each pin individually without regard to the operating mode of the LSI. Table 14.1
lists the multiplexed pins.
Table 14.1 List of Multiplexed Pins
Port
Function 1
Function 2
(related module) (related module)
Function 3
(related module)
Function 4
(related module)
Pin #
A
PA15 I/O (port)
IRQ3 input (INTC)
DREQ1 input
(DMAC)
—
68
A
PA14 I/O (port)
IRQ2 input (INTC)
DACK1 output
(DMAC)
—
67
A
PA13 I/O (port)
IRQ1 input (INTC)
TCLKB input (ITU) DREQ0 input (DMAC)
66
A
PA12 I/O (port)
IRQ0 input (INTC)
TCLKA input (ITU) DACK0 output (DMAC) 65
A
PA11 I/O (port)
DPH I/O (D bus)
TIOCB1 I/O (ITU)
—
64
A
PA10 I/O (port)
DPL I/O (D bus)
TIOCA1 I/O (ITU)
—
62
A
PA9 I/O (port)
AH output (BSC)
—
IRQOUT output (INTC) 61
A
PA8 I/O (port)
BREQ input
(system)
—
—
60
A
PA7 I/O (port)
BACK output
(system)
—
—
58
A
PA6 I/O (port)
RD output (BSC)
—
—
57
A
PA5 I/O (port)
WRH output
(BSC) (LBS
output (BSC))* 1
—
—
56
A
PA4 I/O (port)
WRL output
—
(BSC) (WR output
(BSC))*1
—
55
A
PA3 I/O (port)
CS7 output (BSC)
WAIT input (BSC)
—
54
A
PA2 I/O (port)
CS6 output (BSC)
TIOCB0 I/O (ITU)
—
53
A
PA1 I/O (port)
CS5 output (BSC)
RAS output (BSC)
—
52
A
PA0 I/O (port)
CS4 output (BSC)
TIOCA0 I/O (ITU)
—
51
RENESAS 399
Table 14.1 List of Multiplexed Pins (cont)
Port
Function 1
Function 2
(related module) (related module)
Function 3
(related module)
Function 4
(related module)
Pin #
B
PB15 I/O (port)
IRQ7 input (INTC)
—
TP15 output (TPC)
100
B
PB14 I/O (port)
IRQ6 input (INTC)
—
TP14 output (TPC)
99
B
PB13 I/O (port)
IRQ5 input (INTC)
SCK1 I/O (SCI)
TP13 output (TPC)
98
B
PB12 I/O (port)
IRQ4 input (INTC)
SCK0 I/O (SCI)
TP12 output (TPC)
97
B
PB11 I/O (port)
TxD1 output (SCI)
TP11 output (TPC)
—
96
B
PB10 I/O (port)
RxD1 input (SCI)
TP10 output (TPC)
—
95
B
PB9 I/O (port)
TxD0 output (SCI)
TP9 output (TPC)
—
94
B
PB8 I/O (port)
RxD0 input (SCI)
TP8 output (TPC)
—
93
B
PB7 I/O (port)
TCLKD input (ITU) TOCXB4 output (ITU) TP7 output (TPC)
91
B
PB6 I/O (port)
TCLKC input (ITU) TOCXA4 output (ITU) TP6 output (TPC)
90
B
PB5 I/O (port)
TIOCB4 I/O (ITU)
TP5 output (TPC)
—
89
B
PB4 I/O (port)
TIOCA4 I/O (ITU)
TP4 output (TPC)
—
87
B
PB3 I/O (port)
TIOCB3 I/O (ITU)
TP3 output (TPC)
—
86
B
PB2 I/O (port)
TIOCA3 I/O (ITU)
TP2 output (TPC)
—
85
B
PB1 I/O (port)
TIOCB2 I/O (ITU)
TP1 output (TPC)
—
84
B
PB0 I/O (port)
TIOCA2 I/O (ITU)
TP0 output (TPC)
—
83
—
CS1 output (BSC) CASH output
(BSC)
—
—
47
—
CS3 output (BSC) CASL output
(BSC)
—
—
49
INTC: Interrupt controller
DMAC: Direct memory access controller
ITU: 16-bit integrated timer pulse unit
D bus: Data bus control
BSC: Bus state controller
System: System control
SCI: Serial communications interface
TPC: Programmable timing pattern controller
Port: I/O port
Notes: 1. The bus control register of the bus state controller handles switching between the two
functions.
RENESAS 400
14.2
Register Configuration
Table 14.2 summarizes the registers of the pin function controller.
Table 14.2 Pin Function Controller Registers
Name
Abbreviation R/W
Initial Value
Address
Access Size
Port A I/O register
PAIOR
R/W
H'0000
H'5FFFFC4
8, 16, 32
Port A control register 1
PACR1
R/W
H'3302
H'5FFFFC8
8, 16, 32
Port A control register 2
PACR2
R/W
H'FF95
H'5FFFFCA
8, 16, 32
Port B I/O register
PBIOR
R/W
H'0000
H'5FFFFC6
8, 16, 32
Port B control register 1
PBCR1
R/W
H'0000
H'5FFFFCC
8, 16, 32
Port B control register 2
PBCR2
R/W
H'0000
H'5FFFFCE
8, 16, 32
Column address strobe
pin control register
CASCR
R/W
H'5FFF
H'5FFFFEE
8, 16, 32
14.3
Register Descriptions
14.3.1
Port A I/O Register (PAIOR)
The port A I/O register (PAIOR) is a 16-bit read/write register that selects input or output for
individual pins on a bit-by-bit basis. Bits PA15IOR–PA0IOR correspond to pins
PA15/IRQ3/DREQ1–PA0/CS4/TIOCA0. PAIOR is enabled when the port A pins function as
input/outputs (PA15–PA0) and for ITU input capture and output compare (TIOCA1, TIOCA0,
TIOCB1, and TIOCB0). For other functions, they are disabled. For port A pin functions PA15–
PA0 and TIOCA1, TIOCA0, TIOCB1, and TIOCB0, a given pin in port A is an output pin if its
corresponding PAIOR bit is set to 1, and an input pin if the bit is cleared to 0.
PAIOR is initialized to H'0000 by power-on resets; however, it is not initialized for manual resets,
standby mode, or sleep mode.
RENESAS 401
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
14.3.2
15
14
13
12
11
10
9
8
PA15
IOR
PA14
IOR
PA13
IOR
PA12
IOR
PA11
IOR
PA10
IOR
PA9
IOR
PA8
IOR
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
PA7
IOR
PA6
IOR
PA5
IOR
PA4
IOR
PA3
IOR
PA2
IOR
PA1
IOR
PA0
IOR
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Port A Control Registers (PACR1 and PACR2)
PACR1 and PACR2 are 16-bit read/write registers that select the functions of the sixteen
multiplexed pins of port A. PACR1 selects the function of the top eight bits of port A; PACR2
selects the function of the bottom eight bits of port A. PACR1 and PACR2 are initialized to
H'3302 and H'FF95 respectively by power-on resets but are not initialized for manual resets,
standby mode, or sleep mode.
PACR1:
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
•
15
14
13
12
11
10
9
8
PA15
MD1
PA15
MD0
PA14
MD1
PA14
MD0
PA13
MD1
PA13
MD0
PA12
MD1
PA12
MD0
0
0
1
1
0
0
1
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
PA11
MD1
PA11
MD0
PA10
MD1
PA10
MD0
—
PA8MD
0
0
0
0
0
0
1
0
R/W
R/W
R/W
R/W
R/W
R/W
—
R/W
PA9MD1 PA9MD0
Bits 15 and 14 (PA15 mode (PA15MD1 and PA15MD0)): PA15MD1 and PA15MD0 select
the function of the PA15/IRQ3/DREQ1 pin.
RENESAS 402
Bit 15:
PA15MD1
Bit 14:
PA15MD0
Function
0
0
Input/output (PA15) (initial value)
1
Interrupt request input (IRQ3)
0
Reserved
1
DMA transfer request input (DREQ1)
1
•
Bits 13 and 12 (PA14 mode (PA14MD1 and PA14MD0)): PA14MD1 and PA14MD0 select
the function of the PA14/IRQ2/DACK1 pin.
Bit 13:
PA14MD1
Bit 12:
PA14MD0
Function
0
0
Input/output (PA14)
1
Interrupt request input (IRQ2)
0
Reserved
1
DMA transfer acknowledge output (DACK1) (initial value)
1
•
Bits 11 and 10 (PA13 Mode (PA13MD1 and PA13MD0)): PA13MD1 and PA13MD0 select
the function of the PA13/IRQ1/DREQ0/TCLKB pin.
Bit 11:
PA13MD1
Bit 10:
PA13MD0
Function
0
0
Input/output (PA13) (initial value)
1
Interrupt request input (IRQ1)
0
ITU timer clock input (TCLKB)
1
DMA transfer request input (DREQ0)
1
•
Bits 9 and 8 (PA12 mode (PA12MD1 and PA12MD0)): These bits select the function of the
PA12/IRQ0/DACK0/TCLKA pin.
Bit 9:
PA12MD1
Bit 8:
PA12MD0
Function
0
0
Input/output (PA12)
1
Interrupt request input (IRQ0)
0
ITU timer clock input (TCLKA)
1
DMA transfer acknowledge output (DACK0) (initial value)
1
RENESAS 403
•
Bits 7 and 6 (PA11 mode (PA11MD1 and PA11MD0)): These bits select the function of the
PA11/DPH/TIOCB1 pin.
Bit 7:
PA11MD1
Bit 6:
PA11MD0
Function
0
0
Input/output (PA11) (initial value)
1
Upper data bus parity input/output (DPH)
0
ITU input capture/output compare (TIOCB1)
1
Reserved
1
•
Bits 5 and 4 (PA10 mode (PA10MD1 and PA10MD0)): These bits select the function of the
PA10/DPL/TIOCA1 pin.
Bit 5:
PA10MD1
Bit 4:
PA10MD0
Function
0
0
Input/output (PA10) (initial value)
1
Lower data bus parity input/output (DPL)
0
ITU input capture/output compare (TIOCA1)
1
Reserved
1
•
Bits 3 and 2 (PA9 mode (PA9MD1 and PA9MD0)): These bits select the function of the
PA9/AH/IRQOUT pin.
Bit 3:
PA9MD1
Bit 2:
PA9MD0
Function
0
0
Input/output (PA9) (initial value)
1
Address hold output (AH)
0
Reserved
1
Interrupt request output (IRQOUT)
1
•
Bit 1 (reserved): This bit always reads as 1. The write value should always be 1.
•
Bit 0 (PA8 mode (PA8MD)): PA8MD selects the function of the PA8/BREQ.
Bit 0: PA8MD
Function
0
Input/output (PA8) (initial value)
1
Bus request input (BREQ)
RENESAS 404
PACR2:
Bit:
15
14
13
12
11
10
9
8
Bit name:
—
PA7MD
—
PA6MD
—
PA5MD
—
PA4MD
Initial value:
1
1
1
1
1
1
1
1
R/W:
—
R/W
—
R/W
—
R/W
—
R/W
Bit:
7
6
5
4
3
2
1
0
Bit name: PA3MD1 PA3MD0 PA2MD1 PA2MD0 PA1MD1 PA1MD0 PA0MD1 PA0MD0
Initial value:
R/W:
1
0
0
1
0
1
0
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
•
Bit 15 (reserved): This bit always reads as 1. The write value should always be 1.
•
Bit 14 (PA7 mode (PA7MD)): PA7MD selects the function of the PA7/BACK pin.
Bit 14: PA7MD
Function
0
Input/output (PA7)
1
Bus request acknowledge output (BACK) (initial value)
•
Bit 13 (reserved): This bit always reads as 1. The write value should always be 1.
•
Bit 12 (PA6 mode (PA6MD)): PA6MD selects the function of the PA6/RD pin.
Bit 12: PA6MD
Function
0
Input/output (PA6)
1
Read output (RD) (initial value)
•
Bit 11 (reserved): This bit always reads as 1. The write value should always be 1.
•
Bit 10 (PA5 mode (PA5MD)): PA5MD selects the function of the PA5/WRH (LBS) pin.
Bit 10: PA5MD
Function
0
Input/output (PA5)
1
Upper write output (WRH) or lower byte strobe output (LBS) (initial value)
•
Bit 9 (reserved): This bit always reads as 1. The write value should always be 1.
RENESAS 405
•
Bit 8 (PA4 mode (PA4MD)): PA4MD selects the function of the PA4/WRL (WR) pin.
Bit 8: PA4MD
Function
0
Input/output (PA4)
1
Lower write output (WRL) or write output (WR) (initial value)
•
Bits 7 and 6 (PA3 mode (PA3MD1 and PA3MD0)): PA3MD1 and PA3MD0 select the
function of the PA3/CS7/WAIT pin. This pin has a pull-up MOS that is used when it functions
as a WAIT pin to allow selection of pull up or no pull up (for the WAIT pin) using the wait
state control register of the bus state controller (BSC). There is no pull up when if functions as
PA3 or CS7.
Bit 7: PA3MD1
Bit 6: PA3MD0
Function
0
0
Input/output (PA3)
1
Chip select output (CS7)
0
Wait state input (WAIT) (initial value)
1
Reserved
1
•
Bits 5 and 4 (PA2 mode (PA2MD1 and PA2MD0)): PA2MD1 and PA2MD0 select the
function of the PA2/CS6/TIOCB0 pin.
Bit 5: PA2MD1
Bit 4: PA2MD0
Function
0
0
Input/output (PA2)
1
Chip select output (CS6) (initial value)
0
ITU input capture/output compare (TIOCB0)
1
Reserved
1
•
Bits 3 and 2 (PA1 mode (PA1MD1 and PA1MD0)): PA1MD1 and PA1MD0 select the
function of the PA1/CS5/RAS pin.
Bit 3: PA1MD1
Bit 2: PA1MD0
Function
0
0
Input/output (PA1)
1
Chip select output (CS5) (initial value)
0
Row address strobe output (RAS)
1
Reserved
1
RENESAS 406
•
Bits 1 and 0 (PA0 mode (PA0MD1 and PA0MD0)): PA0MD1 and PA0MD0 select the
function of the PA0/CS4/TIOCA0 pin.
Bit 1: PA0MD1
Bit 0: PA0MD0
Function
0
0
Input/output (PA0)
1
Chip select output (CS4) (initial value)
0
ITU input capture/output compare (TIOCA0)
1
Reserved
1
14.3.3
Port B I/O Register (PBIOR)
The port B I/O register (PBIOR) is a 16-bit read/write register that selects input or output for
individual pins on a bit-by-bit basis. Bits PB15IOR–PB0IOR correspond to pins of port B. PBIOR
is enabled when the port B pins function as input/outputs (PB15–PB0), for ITU input capture and
output compare (TIOCA4, TIOCA3, TIOCA2, TIOCB4, TIOCB3, and TIOCB2), and as serial
clocks (SCK1, SCK0). For other functions, they are disabled. For port B pin functions PB15–PB0,
and TIOCA4, TIOCA3, TIOCA2, TIOCB4, TIOCB3, and TIOCB2, and SCK1/SCK0, a given pin
in port B is an output pin if its corresponding PBIOR bit is set to 1, and an input pin if the bit is
cleared to 0.
PBIOR is initialized to H'0000 by power-on resets; however, it is not initialized for manual resets,
standby mode, or sleep mode.
Bit:
Bit name:
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
15
14
13
12
11
10
9
8
PB15
IOR
PB14
IOR
PB13
IOR
PB12
IOR
PB11
IOR
PB10
IOR
PB9
IOR
PB8
IOR
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
PB7
IOR
PB6
IOR
PB5
IOR
PB4
IOR
PB3
IOR
PB2
IOR
PB1
IOR
PB0
IOR
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
RENESAS 407
14.3.4
Port B Control Registers (PBCR1 and PBCR2)
PBCR1 and PBCR2 are 16-bit read/write registers that select the functions of the sixteen
multiplexed pins of port B. PBCR1 selects the function of the top eight bits of port B; PBCR2
selects the function of the bottom eight bits of port B. PBCR1 and PBCR2 are initialized to
H'0000 by power-on resets but are not initialized for manual resets, standby mode, or sleep mode.
PBCR1:
Bit:
Bit name:
15
14
13
12
11
10
9
8
PB15
MD1
PB15
MD0
PB14
MD1
PB14
MD0
PB13
MD1
PB13
MD0
PB12
MD1
PB12
MD0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
PB11
MD1
PB11
MD0
PB10
MD1
PB10
MD0
PB9
MD1
PB9
MD0
PB8
MD1
PB8
MD0
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Initial value:
R/W:
Bit:
Bit name:
Initial value:
R/W:
•
Bits 15 and 14 (PB15 mode (PB15MD1 and PB15MD0)): PB15MD1 and PB15MD0 select the
function of the PB15/TP15/IRQ7 pin.
Bit 15: PB15MD1
Bit 14: PB15MD0
Function
0
0
Input/output (PB15) (initial value)
1
Interrupt request input (IRQ7)
0
Reserved
1
Timing pattern output (TP15)
1
•
Bits 13 and 12 (PB14 mode (PB14MD1 and PB14MD0)): PB14MD1 and PB14MD0 select the
function of the PB14/TP14/IRQ6 pin.
Bit 13: PB14MD1
Bit 12: PB14MD0
Function
0
0
Input/output (PB14) (initial value)
1
Interrupt request input (IRQ6)
0
Reserved
1
Timing pattern output (TP14)
1
RENESAS 408
•
Bits 11 and 10 (PB13 mode (PB13MD1 and PB13MD0)): PB13MD1 and PB13MD0 select the
function of the PB13/TP13/IRQ5/SCK1 pin.
Bit 11: PB13MD1
Bit 10: PB13MD0
Function
0
0
Input/output (PB13) (initial value)
1
Interrupt request input (IRQ5)
0
Serial clock input/output (SCK1)
1
Timing pattern output (TP13)
1
•
Bits 9 and 8 (PB12 mode (PB12MD1 and PB12MD0)): PB12MD1 and PB12MD0 select the
function of the PB12/TP12/IRQ4/SCK0 pin.
Bit 9: PB12MD1
Bit 8: PB12MD0
Function
0
0
Input/output (PB12) (initial value)
1
Interrupt request input (IRQ4)
0
Serial clock input/output (SCK0)
1
Timing pattern output (TP12)
1
•
Bits 7 and 6: PB11 mode (PB11MD1 and PB11MD0): PB11MD1 and PB11MD0 select the
function of the PB11/TP11/TxD1 pin.
Bit 7: PB11MD1
Bit 6: PB11MD0
Function
0
0
Input/output (PB11) (initial value)
1
Reserved
0
Transmit data output (TxD1)
1
Timing pattern output (TP11)
1
•
Bits 5 and 4 (PB10 mode (PB10MD1 and PB10MD0): PB10MD1 and PB10MD0 select the
function of the PB10/TP10/RxD1 pin.
Bit 5: PB10MD1
Bit 4: PB10MD0
Function
0
0
Input/output (PB10) (initial value)
1
Reserved
0
Receive data input (RxD1)
1
Timing pattern output (TP10)
1
RENESAS 409
•
Bits 3 and 2 (PB9 mode (PB9MD1 and PB9MD0)): PB9MD1 and PB9MD0 select the function
of the PB9/TP9/TxD0 pin.
Bit 3: PB9MD1
Bit 2: PB9MD0
Function
0
0
Input/output (PB9) (initial value)
1
Reserved
0
Transmit data output (TxD0)
1
Timing pattern output (TP9)
1
•
Bits 1 and 0 (PB8 mode (PB8MD1 and PB8MD0)): PB8MD1 and PB8MD0 select the function
of the PB8/TP8/RxD0 pin.
Bit 1: PB8MD1
Bit 0: PB8MD0
Function
0
0
Input/output (PB8) (initial value)
1
Reserved
0
Receive data input (RxD0)
1
Timing pattern output (TP8)
1
PBCR2:
Bit:
15
14
13
12
11
10
9
8
Bit name: PB7MD1 PB7MD0 PB6MD1 PB6MD0 PB5MD1 PB5MD0 PB4MD1 PB4MD0
Initial value:
R/W:
Bit:
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
Bit name: PB3MD1 PB3MD0 PB2MD1 PB2MD0 PB1MD1 PB1MD0 PB0MD1 PB0MD0
Initial value:
R/W:
RENESAS 410
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
•
Bits 15 and 14 (PB7 mode (PB7MD1 and PB7MD0)): PB7MD1 and PB7MD0 select the
function of the PB7/TP7/TOCXB4/TCLKD pin.
Bit 15: PB7MD1
Bit 14: PB7MD0
Function
0
0
Input/output (PB7) (initial value)
1
ITU timer clock input (TCLKD)
0
ITU output compare (TOCXB4)
1
Timing pattern output (TP7)
1
•
Bits 13 and 12 (PB6 mode (PB6MD1 and PB6MD0)): PB6MD1 and PB6MD0 select the
function of the PB6/TP6/TOCXA4/TCLKC pin.
Bit 13: PB6MD1
Bit 12: PB6MD0
Function
0
0
Input/output (PB6) (initial value)
1
ITU timer clock input (TCLKC)
0
ITU output compare (TOCXA4)
1
Timing pattern output (TP6)
1
•
Bits 11 and 10 (PB5 mode (PB5MD1 and PB5MD0)): PB5MD1 and PB5MD0 select the
function of the PB5/TP5/TIOCB4 pin.
Bit 11: PB5MD1
Bit 10: PB5MD0
Function
0
0
Input/output (PB5) (initial value)
1
Reserved
0
ITU input capture/output compare (TIOCB4)
1
Timing pattern output (TP5)
1
•
Bits 9 and 8 (PB4 mode (PB4MD1 and PB4MD0)): PB4MD1 and PB4MD0 select the function
of the PB4/TP4/TIOCA4 pin.
Bit 9: PB4MD1
Bit 8: PB4MD0
Function
0
0
Input/output (PB4) (initial value)
1
Reserved
0
ITU input capture/output compare (TIOCA4)
1
Timing pattern output (TP4)
1
RENESAS 411
•
Bits 7 and 6 (PB3 mode (PB3MD1 and PB3MD0)): PB3MD1 and PB3MD0 select the function
of the PB3/TP3/TIOCB3 pin.
Bit 7: PB3MD1
Bit 6: PB3MD0
Function
0
0
Input/output (PB3) (initial value)
1
Reserved
0
ITU input capture/output compare (TIOCB3)
1
Timing pattern output (TP3)
1
•
Bits 5 and 4 (PB2 mode (PB2MD1 and PB2MD0)): PB2MD1 and PB2MD0 select the function
of the PB2/TP2/TIOCA3 pin.
Bit 5: PB2MD1
Bit 4: PB2MD0
Function
0
0
Input/output (PB2) (initial value)
1
Reserved
0
ITU input capture/output compare (TIOCA3)
1
Timing pattern output (TP2)
1
•
Bits 3 and 2 (PB1 mode (PB1MD1 and PB1MD0)): PB1MD1 and PB1MD0 select the function
of the PB1/TP1/TIOCB2 pin.
Bit 3: PB1MD1
Bit 2: PB1MD0
Function
0
0
Input/output (PB1) (initial value)
1
Reserved
0
ITU input capture/output compare (TIOCB2)
1
Timing pattern output (TP1)
1
•
Bits 1 and 0 (PB0 mode (PB0MD1 and PB0MD0)): PB0MD1 and PB0MD0 select the function
of the PB0/TP0/TIOCA2 pin.
Bit 1: PB0MD1
Bit 0: PB0MD0
Function
0
0
Input/output (PB0) (initial value)
1
Reserved
0
ITU input capture/output compare (TIOCA2)
1
Timing pattern output (TP0)
1
RENESAS 412
14.3.5
Column Address Strobe Pin Control Register (CASCR)
CASCR is a 16-bit read/write register that allows selection between column address strobe and
chip select pin functions. The CASCR is initialized to H'5FFF by power-on resets but is not
initialized for manual resets, standby mode, or sleep mode.
Bit:
Bit name:
15
14
13
12
11
10
9
8
CASH
MD1
CASH
MD0
CASL
MD1
CASL
MD0
—
—
—
—
Initial value:
0
1
0
1
1
1
1
1
R/W
R/W
R/W
R/W
—
—
—
—
Bit:
7
6
5
4
3
2
1
0
Bit name:
—
—
—
—
—
—
—
—
Initial value:
1
1
1
1
1
1
1
1
R/W:
—
—
—
—
—
—
—
—
R/W:
•
Bits 15 and 14 (CASH mode (CASHMD1 and CASHMD0)): CASHMD1 and CASHMD0
select the function of the CS1/CASH pin.
Bit 15: CASHMD1
Bit 14: CASHMD0
Function
0
0
Reserved
1
Chip select output (CS1) (initial value)
0
Column address strobe output (CASH)
1
Reserved
1
•
Bits 13 and 12 (CASL mode (CASLMD1 and CASLMD0)): CASLMD1 and CASLMD0
select the function of the CS3/CASL pin.
Bit 13: CASLMD1
Bit 12: CASLMD0
Function
0
0
Reserved
1
Chip select output (CS3) (initial value)
0
Column address strobe output (CASL)
1
Reserved
1
•
Bits 11–0 (reserved): This bit always reads as 1. The write value should always be 1.
RENESAS 413
Section 15 Parallel I/O Ports
15.1
Overview
There are two ports, A and B. Ports A and B are 16-bit I/O ports. The pins of the ports are all
multiplexed for use as general-purpose I/Os or for other functions. (Use the pin function controller
(PFC) to select the function of multiplexed pins.) Ports A and B each have one data register for
storing pin data.
15.2
Port A
Port A is a 16-pin input/output port, as shown in figure 15.1. The PA3/CS7/WAIT pin of port A
has a pull-up MOS so that when it is functioning as a WAIT pin, the wait state control register of
the bus state controller can be used to select whether to pull up the WAIT pin or not. It is not
pulled up when the pin is functioning as either PA3 or CS7.
Port A
PA15 (Input/output)/IRQ3 (Input)/DREQ1 (Input)
PA14 (Input/output)/IRQ2 (Input)/DACK1 (Output)
PA13 (Input/output)/IRQ1 (Input)/DREQ0 (Input)/TCLKB (Input)
PA12 (Input/output)/IRQ0 (Input)/DACK0 (Output)/TCLKA (Input)
PA11 (Input/output)/DPH (Input/output)/TIOCB1 (Input/output)
PA10 (Input/output)/DPL (Input/output)/TIOCA1 (Input/output)
PA9 (Input/output)/AH (Output)/IRQOUT (Output)
PA8 (Input/output)/BREQ (Input)
PA7 (Input/output)/BACK (Output)
PA6 (Input/output)/RD (Output)
PA5 (Input/output)/WRH (Output) (LBS (Output))
PA4 (Input/output)/WRL (Output) (WR (Output))
PA3 (Input/output)/CS7 (Output)/WAIT (Input)
PA2 (Input/output)/CS6 (Output)/TIOCB0 (Input/output)
PA1 (Input/output)/CS5 (Output)/RAS (Output)
PA0 (Input/output)/CS4 (Output)/TIOCA0 (Input/output)
Figure 15.1 Port A Configuration
15.2.1
Register Configuration
Table 15.1 summarizes the port A register.
RENESAS 415
Table 15.1 Port A Register
Name
Abbreviation
R/W
Initial Value
Address
Access Size
Port A data register
PADR
R/W
H'0000
H'5FFFFC0
8, 16, 32
15.2.2
Port A Data Register (PADR)
PADR is a 16-bit read/write register that stores data for port A. The bits PA15DR–PA0DR
correspond to the PA15/IRQ3/DREQ1–PA0/CS4/TIOCA0 pins. When the pins are used as
ordinary outputs, they will output whatever value is written in the PADR; when PADR is read, the
register value will be output regardless of the pin status. When the pins are used as ordinary
inputs, the pin status rather than the register value is read directly when PADR is read. When a
value is written to PADR, that value can be written into PADR, but it will not affect the pin status.
Table 15.2 shows the read/write operations of the port A data register.
PADR is initialized by a power-on reset. However, PADR is not initialized for manual reset,
standby mode, or sleep mode.
Bit:
Bit name:
15
13
12
11
10
9
8
PA15DR PA14DR PA13DR PA12DR PA11DR PA10DR PA9DR PA8DR
Initial value:
R/W:
Bit:
Bit name:
14
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
PA7DR PA6DR
Initial value:
R/W:
PA5DR PA4DR
PA3DR
PA2DR PA1DR
PA0DR
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Table 15.2 Read/Write Operation of the Port A Data Register (PADR)
PAIOR
Pin Status
Read
Write
0
Input
Pin status
Can write to PADR, but it has no effect on pin
status.
Other function
Pin status
Can write to PADR, but it has no effect on pin
status.
Output
PADR value
Value written is output by pin
Other function
PADR value
Can write to PADR, but it has no effect on pin
status.
1
RENESAS 416
15.3
Port B
Port B is a 16-bit input/output port as shown in figure 16.2.
Port B
PB15 (Input/output)/TP15 (Output)/IRQ7 (Input)
PB14 (Input/output)/TP14 (Output)/IRQ6 (Input)
PB13 (Input/output)/TP13 (Output)/IRQ5 (Input)/SCK1 (Input/output)
PB12 (Input/output)/TP12 (Output)/IRQ4 (Input)/SCK0 (Input/output)
PB11 (Input/output)/TP11 (Output)/TxD1 (Output)
PB10 (Input/output)/TP10 (Output)/RxD1 (Input)
PB9 (Input/output)/TP9 (Output)/TxD0 (Output)
PB8 (Input/output)/TP8 (Output)/RxD0 (Input)
PB7 (Input/output)/TP7 (Output)/TOCXB4 (Output)/TCLKD (Input)
PB6 (Input/output)/TP6 (Output)/TOCXA4 (Output)/TCLKC (Input)
PB5 (Input/output)/TP5 (Output)/TIOCB4 (Output)
PB4 (Input/output)/TP4 (Output)/TIOCA4 (Output)
PB3 (Input/output)/TP3 (Output)/TIOCB3 (Input/output)
PB2 (Input/output)/TP2 (Output)/TIOCA3 (Input/output)
PB1 (Input/output)/TP1 (Output)/TIOCB2 (Input/output)
PB0 (Input/output)/TP0 (Output)/TIOCA2 (Input/output)
Figure 15.2 Port B Configuration
15.3.1
Register Configuration
Table 15.3 summarizes the port B register.
Table 15.3 Port B Register
Name
Abbreviation
R/W
Initial Value
Address
Access Size
Port B data register
PBDR
R/W
H'0000
H'5FFFFC2
8, 16, 32
RENESAS 417
15.3.2
Port B Data Register (PBDR)
PBDR is a 16-bit read/write register that stores data for port B. The bits PB15DR–PB0DR
correspond to the PB15/TP15/IRQ7–PB0/TP0/TIOCA2 pins. When the pins are used as ordinary
outputs, they will output whatever value is written in the PBDR; when PBDR is read, the register
value will be output regardless of the pin status. When the pins are used as ordinary inputs, the pin
status rather than the register value is read directly when PBDR is read. When a value is written to
PBDR, that value can be written into PBDR, but it will not affect the pin status. When the pin
function is set to timing pattern output and the TPC output is enabled by the TPC next data enable
register (NDER), no value can be written to PBDR. Table 15.4 shows the read/write operations of
the port B data register.
PBDR is initialized by a power-on reset. However, PBDR is not initialized for a manual reset,
standby mode, or sleep mode.
Bit:
Bit name:
15
13
12
11
10
9
8
PB15DR PB14DR PB13DR PB12DR PB11DR PB10DR PB9DR PB8DR
Initial value:
R/W:
Bit:
Bit name:
14
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
PB7DR PB6DR
Initial value:
R/W:
PB5DR PB4DR
PB3DR
PB2DR PB1DR
PB0DR
0
0
0
0
0
0
0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Table 15.4 Read/Write Operation of the Port B Data Register (PBDR)
PBIOR
Pin Status
Read
Write
0
Input
Pin status
Can write to PBDR, but it has no effect on pin
status
TPn
Pin status
Disabled
Other function
Pin status
Can write to PBDR, but it has no effect on pin
status
Output
PBDR value
Value written is output by pin
TPn
PBDR value
Disabled
Other function
PBDR value
Can write to PBDR, but it has no effect on pin
status
1
TPn: Timing pattern output
RENESAS 418
Section 16 ROM
16.1
Overview
The SH7020 microcomputer has 16 kbytes of on-chip ROM (mask ROM). The SH7021
microcomputer has 32 kbytes of on-chip ROM (mask ROM or PROM). The on-chip ROM is
connected to the CPU and the direct memory access controller (DMAC) through a 32-bit data bus
(figure 16.1). The CPU can access the on-chip ROM in 8-, 16- and 32-bit widths and the DMAC
can access the ROM in 8- and 16-bit widths. Data in the on-chip ROM can always be accessed in
one cycle.
RENESAS 419
SH7020
Internal data bus (32 bits)
H'0000000
H'0000001
H'0000002
H'0000003
H'0000004
H'0000005
H'0000006
H'0000007
On-chip ROM
H'0003FFC
H'0003FFD
H'0003FFE
H'0003FFF
SH7021
Internal data bus (32 bits)
H'0000000
H'0000001
H'0000002
H'0000003
H'0000004
H'0000005
H'0000006
H'0000007
On-chip ROM
H'0007FFC
Note:
H'0007FFD
H'0007FFE
H'0007FFF
The addresses shown in the figure are the first shadow addresses in the on-chip ROM
space.
Figure 16.1 Block Diagram of ROM
The operating mode determines whether the on-chip ROM is valid or not. The operating mode is
selected using mode-setting pins MD0-MD2 as shown in table17.1. If you are using the on-chip
ROM, select mode 2; if you are not, select mode 0 or 1. The on-chip ROM is allocated to address
H'0000000–H'0003FFF (SH7020), H'0000000–H'0007FFF (SH7021) of memory area 0. Memory
area 0 (H'0000000-H'0FFFFFF and H'8000000–H'8FFFFFF) is divided into 16-kbyte (SH7020) or
32-kbyte (SH7021) shadows. No matter which shadow is accessed, the on-chip ROM is accessed.
See section 8, Bus State Controller, for more information on shadows.
420 RENESAS
Table 16.1 Operating Modes and ROM
Mode Setting Pin
Operating Mode
MD2
MD1
MD0
Area 0
Mode 0 (MCU mode 0)
0
0
0
On-chip ROM invalid, external 8-bit space
Mode 1 (MCU mode 1)
0
0
1
On-chip ROM invalid, external 16-bit space
Mode 2 (MCU mode 2)
0
1
0
On-chip ROM valid
Mode 7 (PROM mode)
1
1
1
—
0: Low
1: High
When the SH7021 is set to PROM mode, the PROM version can write programs exactly like
ordinary EPROM using a general purpose EPROM writer.
16.2
PROM Mode
16.2.1
Setting the PROM Mode
To program the on-chip PROM, set the pins as shown in figure 16.2 and use the chip in PROM
mode.
16.2.2
Socket Adapter Pin Correspondence and Memory Map
Mount the socket adapter to the SH7021 as shown in figure 16.2. This allows the on-chip PROM
to be programmed in exactly the same way as ordinary 32-pin EPROMs (HN27C101). Figure 16.2
shows the correspondence of SH7021 pins and HN27C101 pins. Figure 16.3 shows the memory
map of the on-chip ROM.
The address range of the HN27C101 (128 kbytes) is H'00000–H'1FFFF. The on-chip PROM (34
kbytes) is not found in H'08000–H'1FFFF.
When programming with a PROM writer, the program address range must be set to H'00000–
H'07FFF. The data for the H'08000–H'1FFFF address area should all be H'FF. Set byte mode, not
page mode.
RENESAS 421
SH7021
Pin Name
Pin Number
76
RES
74
NMI
1
AD0
2
AD1
3
AD2
5
AD3
6
AD4
7
AD5
8
AD6
9
AD7
20
A0/HBS
21
A1
22
A2
23
A3
25
A4
26
A5
27
A6
28
A7
29
A8
30
A9
31
A10
33
A11
34
A12
35
A13
36
A14
37
A15
39
A16
53
PA2/CS6/TIOCB0
54
PA3/CS7/WAIT
40
A17
42
A18
VCC
13, 38, 63, 73, 80, 88
77
MD0
78
MD1
79
MD2
4, 15, 24, 32, 41,
VSS
50, 59, 70, 81, 82,92
Pin other than the above
NC (release)
EPROM Socket
HN27C101
Adapter
Pin Name Pin Number
VPP
1
A9
26
I/O0
13
I/O1
14
I/O2
15
I/O3
17
I/O4
18
I/O5
19
I/O6
20
I/O7
21
A0
12
A1
11
A2
10
A3
9
A4
8
A5
7
A6
6
A7
5
A8
27
24
OE
A10
23
A11
25
A12
4
A13
28
A14
29
A15
3
A16
2
31
PGM
22
CE
VCC
32
VSS
16
• VPP: PROM program
power adapter
• A16–A0: Address input
• I/O7–I/O0: Data input/
output
• OE: Output enable
• PGM: Program enable
• CE: Chip enable
Figure 16.2 Correspondence Between SH7021 Pins and HN27C101 Pins
422 RENESAS
Addresses in MCU
modes 0, 1, and 2*
Addresses in
PROM mode
H'0000000
H'0000
On-chip
ROM space
(area 0)
H'0007FFF
H'7FFF
Note: Addresses in the figure are the uppermost shadow addresses of the on-chip ROM
space.
Figure 16.3 Memory Map of On-chip ROM
16.3
PROM Programming
The write/verify specifications in PROM mode are the same as for the standard EPROM
HN27C101. The page program system is not supported, so do not set the PROM writer to the page
programming mode. Naturally, PROM writers that only support the page programming mode
cannot be used. When selecting a PROM writer, check that the high-speed, high-reliability
programming system for each byte is supported.
16.3.1
Selecting the Programming Mode
There are two on-chip PROM programming modes: write and verify (which reads and confirms
the data written). Use the pins to select the modes (table 16.2).
RENESAS 423
Table 16.2 Select PROM Programming Mode
Pin
Mode
CE
OE
PGM
VPP
VCC
I/O7–I/O0
A16–A0
Write
0
1
0
VPP
VCC
Data input
Address input
Verify
0
0
1
Data output
Program inhibit
0
0
0
High impedance
0
1
1
1
0
0
1
1
1
Symbols:
0: Low
1: High
VPP: VPP level
VCC: VCC level
16.3.2
Write/Verify and Electrical Characteristics
Write/Verify: Write/verify can be accomplished by an efficient high-speed high-reliability
programming system. This system can write data quickly and accurately without placing voltage
stress on the device. The basic flowchart for this high-speed, high-reliability programming system
is shown in figure 16.4.
424 RENESAS
Start
Set EPROM writer to
write/verify mode
(VCC = 6.0 V ± 0.25 V,
VPP = 12.5 V ± 0.3 V)
Address = 0
n=0
n+1→n
No
Yes
Data write
(tPW = 0.2 ms ± 5%)
n = 25?
No
Address + 1 →
Address
Verify
result OK?
Yes
Data write
(tOPW = (0.2 × n) ms)
Final
address?
No
Yes
Set EPROM
writer to read mode
(VCC = 5.0 V ± 0.25 V,
VPP = VCC)
No good
VCC:
VPP:
tPW:
tOPW:
Power supply
PROM program power supply
Initial programming pulse width
Over programming pulse width
No
Results of
reading all address
OK?
Yes
End
Figure 16.4 Basic Flowchart of High-Speed High-Reliability Programming
RENESAS 425
Electrical Characteristics: Tables 16.3 and 16.4 show the electrical characteristics of
programming. Figure 16.5 shows the timing.
Table 16.3 DC Characteristics (VCC = 6.0 V ± 0.25 V, VPP = 12.5 ± 0.3 V, VSS = 0 V, Ta =
25 ± 5˚C)
Pins
Symbol
Min
Typ Max
Input high
voltage
I/O7–I/O0, A16–A0,
OE, CE, PGM
VIH
2.4
—
VCC + 0.3 V
Input low
voltage
I/O7–I/O0, A16–A0,
OE, CE, PGM
VIL
–0.3
—
0.8
V
Output high
voltage
I/O7–I/O0
VOH
2.4
—
—
V
I OH = -200 µA
Output low
voltage
I/O7–I/O0
VOL
—
—
0.45
V
I OL = 1.6 mA
Input leak
current
I/O7–I/O0, A16–A0,
OE, CE, PGM
|ILI|
—
—
2
µA
VIN = 5.25 V/0.5 V
VCC current
I CC
—
—
40
mA
VPP current
I PP
—
—
40
mA
426 RENESAS
Unit
Measurement
Conditions
Item
Table 16.4 AC Characteristics (VCC = 6.0 V ± 0.25 V, VPP = 12.5 ± 0.3 V, VSS = 0 V, Ta =
25 ± 5˚C)
Item
Symbol
Min
Typ
Max
Unit
Measurement
Conditions
Address setup time
t AS
2
—
—
µs
Figure 16.5*1
OE setup time
t OES
2
—
—
µs
Data setup time
t DS
2
—
—
µs
Address hold time
t AH
0
—
—
µs
Data hold time
t DH
2
—
—
µs
—
—
130
ns
2
—
—
µs
0.19
0.20
0.21
ms
0.19
—
5.25
ms
*2
Data output disable time
t DF
VPP setup time
t VPS
PGM pulse width in initial programming
t PW
*3
PGM pulse width in over programming
t OPW
VCC setup time
t VCS
2
—
—
µs
CE setup time
t CES
2
—
—
µs
Data output delay time
t OE
0
—
150
ns
Notes: 1. Input pulse level: 0.45–2.4 V
Input rise, fall time ≤ 20 ns
Input timing reference levels: 0.8 V, 2.0 V
Output timing reference levels: 0.8 V, 2.0 V
2. t DF is defined as when output reaches the release state and the output level could not
be referenced.
3. t OPW is defined as the value given in the flowchart.
RENESAS 427
Write
Verify
Address
tAS
tAH
Write data
Data
VCC
tDF
tDH
tDS
VPP
Read data
VPP
VCC
tVPS
VCC + 1
VCC
tVCS
CE
tCES
PGM
tPW
tOES
tOE
(tOPW)
OE
Note: t OPW is defined as the value given in the flowchart.
Figure 16.5 Write/Verify Timing
16.3.3
Points to Note About Writing
1. Always write using the prescribed voltage and timing. The write voltage (programming
voltage) VPP is 12.5 V (when the EPROM writer is set to the Hitachi specifications for
HN27C101, VPP becomes 12.5 V.) Applying a voltage in excess of the rated voltage may
damage the device. Pay particular attention to overshooting in the EPROM writer.
2. Before programming, always check that the indexes of the EPROM writer socket, socket
adapter, and devices are consistent with each other. If they are not mounted in the proper
location, an overcurrent may be generated, damaging the device.
3. Do not touch the socket adapter or device during writing. Contact can cause malfunctions that
prevent data from being written accurately.
428 RENESAS
4. You cannot write in the page programming mode. Always set the equipment to the byte
programming mode.
5. The capacity of the on-chip ROM is 32 kbytes, so the data of PROM writer addresses
H'08000–H'1FFFF should be H'FF. Always set the range for PROM addresses to H'0000–
H'7FFF.
6. When write errors occur on consecutive addresses, stop writing. Check to see if there are any
abnormalities in the EPROM writer and socket adapter.
16.3.4
Reliability After Writing
After programming, we recommend letting the device stand at high temperature (125–150°C) for
24–48 hours to increase the reliability of data retention. Letting it stand at high temperature is a
type of screening method that can get rid of initial data retention defects of the on-chip PROM's
memory cell within a short period of time. Figure 16.6 shows the flow from programming of the
on-chip PROM, including screening, to mounting on the device board.
Writing and verification
of program
Flow chart
from figure 16.4
Let stand in nonconductive,
high temperature environment
(125–150°C, 24–48 hours)
Data read and verification
(VCC = 5.0 V)
Mount on board
Figure 16.6 Screening Flow
If abnormalities are found when the program is written and verified or the program is read and
checked after the writing/verification or letting the chip stand at high temperature, contact
Hitachi's engineering departments.
RENESAS 429
Section 17 RAM
17.1
Overview
The SH7020 and SH7021 has 1-kbytes of on-chip RAM. The on-chip RAM is linked to the CPU
and direct memory access controller (DMAC) with a 32-bit data bus (figure 17.1). The CPU can
access data in the on-chip RAM in byte, word, or long word units. The DMAC can access byte or
word data. On-chip RAM data can always be accessed in one state, making the RAM ideal for use
as a program area, stack area, or data area, which require high-speed access. The contents of the
on-chip RAM are held in both the sleep and standby modes. Memory area 7 addresses H'FFFFC00
to H'FFFFFFF are allocated to the on-chip RAM.
Internal data bus (32 bits)
H'FFFFC00
H'FFFFC01
H'FFFFC02
H'FFFFC03
H'FFFFC04
H'FFFFC05
H'FFFFC06
H'FFFFC07
On-chip RAM
H'FFFFFFC
H'FFFFFFD
H'FFFFFFE
H'FFFFFF
Figure 17.1 Block Diagram of RAM
17.2
Operation
Accesses to addresses H'FFFFC00–H'FFFFFFF are directed to the on-chip RAM. Memory area 7
(H'F000000–H'FFFFFFF) is divided into shadows in 1 kbyte units. All shadow accesses are onchip RAM accesses. For more information on shadows, see section 8, Bus State Controller.
RENESAS 431
Section 18 Power-Down States
18.1
Overview
In the power-down mode, all CPU functions are halted. This lowers power consumption
dramatically.
18.1.1
Power-Down Modes
The SH microprocessor has two power-down modes.
1. Sleep mode
2. Standby mode
The sleep mode and standby mode are entered from the program execution state according to the
transition conditions given in table 18.1. Table 18.1 also describes procedures for canceling each
mode and the states of the CPU and peripheral functions.
Table 18.1 Power-Down States
State
Mode
Entering
Procedure
Clock
CPU
Peripheral
Functions
CPU
Registers RAM
I/O
Ports
Canceling
Procedure
Sleep
mode
Execute
SLEEP
instruction
with SBY
bit set to 0
in SBYCR
Run
Halt
Run
Held
Held
Held
• Interrupt
• DMA
address
error
• Power-on
reset
• Manual
reset
Standb
y mode
Execute
SLEEP
instruction
with SBY
bit set to 1
in SBYCR
Halt
Halt
Halt*1
Held
Held
Held or
high-Z*2
• NMI
• Power-on
reset
• Manual
reset
SBYCR: Standby control register
SBY: Standby bit
Notes: 1. Some of the registers of the on-chip peripheral modules are not initialized in the
standby mode. For details, see table 18.3, Status of Registers in the Standby Mode in
section 18.4.1, Transition to the Standby Mode, or the descriptions of registers given
where the on-chip peripheral modules are covered.
2. The status of I/O ports in the standby mode are set by the port high-impedance bit (HIZ)
of the SBYCR. See section 18.2, Standby Control Register (SBYCR) for details. The
status of pins other than the I/O ports are described in appendix B, Pin States.
RENESAS 433
18.1.2
Register
Table 18.2 summarizes the register related to the power-down state.
Table 18.2 Standby Control Register (SBYCR)
Name
Abbreviation
R/W
Initial Value
Address
Access size
Standby control register
SBYCR
R/W
H'1F
H'5FFFFBC
8, 16, 32
18.2
Standby Control Register (SBYCR)
The standby control register (SBYCR) is an 8-bit register that can be read or written to. It is set in
order to enter the standby mode and also sets the port states in standby mode. The SBYCR is
initialized to H'1F when reset.
Bit:
Bit name:
Initial value:
R/W:
•
7
6
5
4
3
2
1
0
SBY
HIZ
—
—
—
—
—
—
0
0
0
1
1
1
1
1
R/W
R/W
—
—
—
—
—
—
Bit 7 (standby (SBY)): SBY enables transition to the standby mode. The SBY bit cannot be set
to 1 while the timer enable bit (bit TME) in timer control/status register TCSR of watchdog
timer WDT is set to 1. To enter the standby mode, clear the TME bit to 0 to halt the WDT and
set the SBY bit.
SBY
Description
0
Executing SLEEP instruction puts the LSI into sleep mode (initial value)
1
Executing SLEEP instruction puts the LSI into standby mode
•
Bit 6 (port high-impedance (HIZ)): HIZ selects whether I/O ports remain in their previous
states during standby, or are placed in the high-impedance state when the standby mode is
entered. The HIZ bit cannot be set to 1 while the TME bit is set to 1. To place the pins of the
I/O ports in high impedance, clear the TME bit to 0 before setting the HIZ bit.
HIZ
Description
0
Port states are maintained during standby (initial value)
1
Ports are placed in the high-impedance state in standby
RENESAS 434
•
Bits 5–0 (reserved): Bit 5 is a read-only bit that always reads as 0. Only write 0 to bit 5.
Writing to bits 4–0 is disabled. These bits always read 1.
18.3
Sleep Mode
18.3.1
Transition to the Sleep Mode
Execution of the SLEEP instruction when the standby bit (SBY) in the standby control register
(SBYCR) is cleared to 0 causes a transition from the program execution state to the sleep mode.
Although the CPU halts immediately after executing the SLEEP instruction, the contents of its
internal registers remain unchanged. The on-chip peripheral modules do not halt in the sleep
mode.
18.3.2
Canceling the Sleep Mode
The sleep mode is canceled by an interrupt, DMA address error, power-on reset, or manual reset.
Cancellation by an Interrupt: When an interrupt occurs, the sleep mode is canceled and interrupt
exception processing is executed. The sleep mode is not canceled if the interrupt cannot be
accepted because its priority level is equal to or less than the mask level set in the CPU’s status
register (SR). Likewise, the sleep mode is not canceled if the interrupt is disabled by the on-chip
peripheral module.
Cancellation by a DMA Address Error: If the DMAC operates during the sleep mode and a
DMA address error occurs, the sleep mode is canceled and DMA address error exceptionprocessing is executed.
Cancellation by a Power-On Reset: If the RES signal goes low while the NMI signal is high, the
sleep mode is canceled and the power-on reset state is entered. If the NMI signal is brought from
low to high in order to set the LSI for power-on resets, an NMI interrupt will occur whenever the
rising edge of the NMI is selected as the valid edge (in NMI edge select bit NMIE of the interrupt
control register ICR of the interrupt controller). When this occurs, the NMI interrupt cancels the
sleep mode.
Cancellation by a Manual Reset: If the RES signal goes low while the NMI signal is low, the
sleep mode is canceled and the manual reset state is entered. If the NMI signal is brought from
high to low in order to set the LSI for manual resets, the sleep mode will be canceled by an NMI
interrupt whenever the falling edge of the NMI is selected as the valid edge (in the NMIE bit).
RENESAS 435
18.4
Standby Mode
18.4.1
Transition to the Standby Mode
To enter the standby mode, set the standby bit (SBY) to 1 in the standby control register
(SBYCR), then execute the SLEEP instruction. The LSI moves from the program execution state
to the standby mode. The standby mode greatly reduces power consumption by halting not only
the CPU, but the clock and on-chip peripheral modules as well. Some registers of the on-chip
peripheral modules are initialized, others are not (See table 18.3). As long as the specified voltage
is supplied, however, CPU register contents and on-chip RAM data are held. The I/O port state
(hold or high impedance) depends on the port high-impedance bit (HIZ) in the SBYCR. For details
on the states of these pins, see appendix B. Pin States.
RENESAS 436
Table 18.3 Register States in the Standby Mode
Module
Register Initialized
Registers That Hold Data
Interrupt controller (INTC)
—
All registers
User break controller (UBC)
—
All registers
Bus state controller (BSC)
—
All registers
Pin function controller (PFC)
—
All registers
I/O ports
—
All registers
Direct memory access controller
(DMAC)
All registers
—
Watchdog timer (WDT)
• Bits 7–5 (OVF, WT/IT, TME)
of the timer control status
register (TCSR)
• Bits 2–0 (CKS2–CKS0)
of the timer control status
register (TCSR)
• Reset control/status register
(RSTCSR)
• Timer counter (TCNT)
16-bit integrated timer pulse unit
(ITU)
All registers
—
Programmable timing pattern
controller (TPC)
—
All registers
Serial communications interface
(SCI)
• Receive data register (RDR)
—
• Transmit data register (TDR)
• Serial mode register (SMR)
• Serial control register (SCR)
• Serial status register (SSR)
• Bit rate register (BBR)
Power-down state register
—
Standby control register
(SBYCR)
RENESAS 437
18.4.2
Canceling the Standby Mode
The standby mode is canceled by an NMI interrupt, a power-on reset, or a manual reset.
Cancellation by an NMI: When a rising edge or falling edge (as selected by the NMIE bit in
interrupt control register ICR of interrupt controller INTC) is detected at the NMI pin, the clock
oscillator begins operating. At first, clock pulses are supplied only to the watchdog timer. After the
time that was selected before entering the standby mode using clock select bits 2–0 (CKS2–CKS0)
in the timer control/status register TCSR of the watchdog timer WDT, the watchdog timer
overflows. After the overflow, the clock is considered stable and supplied to the entire chip. The
standby mode is canceled and the NMI exception-processing sequence begins.
When the standby mode is cleared by an NMI interrupt, bits CKS2–CKS0 must be set so that the
WDT overflow interval is equal to or greater than the clock settling time. When the standby mode
is cleared when the fall edge has been selected in the NMI bit, be sure that the NMI pin is high
when standby is entered (when the clock is halted) and low when the chip returns from standby
(clock starts up after oscillator is stabilized). Likewise, when the standby mode is cleared when the
rise edge has been selected in the NMI bit, be sure that the NMI pin is low when standby is
entered (clock halted) and high when the chip returns from standby (clock starts up after oscillator
is stabilized).
Cancellation by a Power-On Reset: If the RES signal goes low while the NMI signal is high, the
standby mode is canceled and the power-on reset state is entered. If the NMI signal is brought
from low to high in order to set the LSI for power-on resets, the standby mode will not be canceled
by an NMI interrupt, because the NMI signal is initialized for the falling edge in the standby mode
(by the NMIE bit).
Cancellation by a Manual Reset: If the RES signal goes low while the NMI signal is low, the
standby mode is canceled and the manual reset state is entered. If the NMI signal is brought from
high to low in order to set the LSI for manual resets, the standby mode will first be canceled by an
NMI interrupt, because the NMI signal is initialized for the falling edge in the standby mode (by
the NMIE bit).
RENESAS 438
18.4.3
Standby Mode Application
In this example, the standby mode is entered on the falling edge of the NMI signal and canceled
on the rising edge of the NMI signal. Figure 18.1 shows the timing.
After an NMI interrupt is accepted (high goes to low) while the NMI edge select bit NMIE in the
interrupt control register ICR is cleared to 0 to select detection of the falling edge, the NMI
exception service routine sets the NMIE to 1 (selecting detection of the rising edge) and sets the
SBY bit to 1. Finally, it executes a SLEEP instruction to enter the standby mode.
The standby mode is canceled on the rising edge of the NMI signal.
Oscillator
CK
NMI
NMIE
SSBY
Clock setting time
NMI
Exception
exception
service
processing
routine
SBY = 1
SLEEP
instruction
Standby Oscillation
mode
start
time
Time
set
in
WDT
NMI
exception
processing
Figure 18.1 NMI Timing for the Standby Mode (Example)
RENESAS 439
Section 19 Electrical Characteristics
19.1
Absolute Maximum Ratings
Table 19.1 Absolute Maximum Ratings
Item
Symbol
Rating
Unit
Power supply voltage
VCC
–0.3 to +7.0
V
Program voltage
VPP
–0.3 to +13.5
V
Input voltage
Vin
–0.3 to VCC + 0.3
V
Operating temperature
Topr
–20 to +75*
˚C
Storage temperature
Tstg
–55 to +125
˚C
Caution: Operating the LSI in excess of the absolute maximum rating may result in permanent
damage.
Note: Normal Products: Topr = –40 to +85°C for wide-temperature range products
RENESAS 441
19.2
DC Characteristics
Table 19.2 lists DC characteristics. Table 19.3 lists the permissible output current values.
Usage Conditions:
•
The current consumption value is measured under conditions of V IH min = VCC – 0.5 V and
VIL max = 0.5 V with no load on any output pin and the on-chip pull-up MOS off.
Table 19.2 DC Characteristics (1)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol Min
Typ Max
Measurement
Unit Conditions
VCC – 0.7—
VCC + 0.3 V
EXTAL
VCC × 0.7—
VCC + 0.3 V
Other input pins
2.2
—
VCC + 0.3 V
–0.3
—
0.5
V
–0.3
—
0.8
V
VT+
4.0
—
—
V
VT–
—
Input high-
RES, NMI,
MD2–MD0
level
voltage
Input low-
RES, NMI,
MD2–MD0
level
voltage
Other input pins
Schmidt
trigger
input
voltage
PA13–PA10,
PA2, PA0, PB7–
PB0
Input leak
current
RES
3-state leak
current
(while off)
Ports A and B,
CS3–CS0, A21–
A0, AD15–AD0
Input pullup MOS
current
Output
high-level
voltage
VIH
VIL
—
1.0
V
VT+–VT– 0.4
—
—
V
|Iin|
—
—
1.0
µA
Vin = 0.5 to VCC – 0.5 V
—
—
1.0
µA
Vin = 0.5 to VCC – 0.5 V
|ITSI|
—
—
1.0
µA
Vin = 0.5 to VCC – 0.5 V
PA3
–Ip
20
—
300
µA
Vin = 0V
All output pins
VOH
VCC – 0.5—
—
V
I OH = –200 µA
3.5
—
V
I OH = –1 mA
NMI, MD2–MD0
RENESAS 442
—
Table 19.2 DC Characteristics (1) (cont)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products .
Item
Symbol Min Typ Max
—
—
0.4
V
I OL = 1.6 mA
—
—
1.2
V
I OL = 8 mA
—
—
30
pF
NMI
—
—
30
pF
All other input
pins
—
—
20
pF
Vin = 0 V
Input signal
f = 1 MHz
Ta = 25°C
—
65
80
mA
f = 12.5 MHz
—
75
90
mA
f = 16.6 MHz
—
90
100
mA
f = 20 MHz
—
30
50
mA
f = 12.5 MHz
—
35
55
mA
f = 16.6 MHz
—
40
60
mA
f = 20 MHz
—
0.01 5
µA
Ta ≤ 50°C
—
—
20.0
µA
50°C < Ta
2.0
—
—
V
Output low level
voltage
All output pins
Input capacitance
RES
Current consumption
Ordinary
operation
VOL
Cin
I CC
Sleep
Standby
RAM stand-by
Measurement
Unit Conditions
voltage
VRAM
Usage Notes:
1. Current dissipation values are for VIH min = VCC - 0.5 V and VIL max = 0.5 V with all output
pins unloaded and the on-chip pull-up transistors in the off state.
5. The ZTAT and mask versions have the same functions, and the electrical characteristics of
both are within specification, but characteristic-related performance values, operating margins,
noise margins, noise emission, etc., are different. Caution is therefore required in carrying out
system design, and when switching between ZTAT and mask versions.
RENESAS 443
Table 19.2 DC Characteristics (2)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol Min
Measurement
Unit Conditions
VCC – 0.7 —
VCC + 0.3
V
EXTAL
VCC × 0.7 —
VCC + 0.3
V
Other input pins
2.2
—
VCC + 0.3
V
–0.3
—
0.5
V
–0.3
—
0.8
V
Input high-
RES, NMI,
MD2–MD0
level voltage
Input lowlevel voltage
Typ Max
RES, NMI,
MD2–MD0
VIH
VIL
Other input pins
Schmidt
trigger input
voltage
PA13–10, PA2,
PA0, PB7–PB0
Input leak
current
RES
+
VT
4.0
—
—
V
VT–
—
—
1
V
VT+–VT–
0.4
—
—
V
|Iin|
—
—
1.0
µA
Vin = 0.5 to VCC
– 0.5 V
—
—
1.0
µA
Vin = 0.5 to VCC
– 0.5 V
NMI, MD2–MD0
3-state leak Ports A and B,
current (while CS3–CS0, A21–
off)
A0, AD15–AD0
|ITSI|
—
—
1.0
µA
Vin = 0.5 to VCC
– 0.5 V
Input pull-up PA3
MOS current
–Ip
20
—
300
µA
Vin = 0 V
Output highlevel voltage
All output pins
VOH
VCC – 0.5 —
—
V
I OH = –200 µA
3.5
—
—
V
I OH = –1 mA
Output low
level voltage
All output pins
—
—
0.4
V
I OL = 1.6 mA
—
—
1.2
V
I OL = 8 mA
RENESAS 444
VOL
Table 19.2 DC Characteristics (2) (cont)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products .
Item
Input
capacitance
Current
consumption
Min
Typ
Max
Unit
Cin
—
—
30
pF
NMI
—
—
30
pF
All other input
pins
—
—
20
pF
Vin = 0 V
Input signal
f = 1 MHz
Ta = 25°C
—
65
80
mA
f = 12.5 MHz
—
75
90
mA
f = 16.6 MHz
—
30
50
mA
f = 12.5 MHz
—
35
55
mA
f = 16.6 MHz
—
0.01
5
µA
Ta ≤ 50°C
—
—
20.0
µA
50°C < Ta
2.0
—
—
V
RES
Ordinary
operation
I CC
Sleep
Standby
RAM
stand-by
Measurement
Conditions
Symbol
voltage
VRAM
RENESAS 445
Table 19.2 DC Characteristics (3)
Conditions: VCC = 3.0 V to 5.5 V, VSS = 0 V, φ = 12.5 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Typ Max
Measurement
Unit Conditions
VCC × 0.9
—
VCC + 0.3
V
EXTAL
VCC × 0.7
—
VCC + 0.3
V
Other input pins
VCC × 0.7
—
VCC + 0.3
V
–0.3
—
VCC × 0.1
V
–0.3
—
VCC × 0.2
V
VT
VCC × 0.9
—
—
V
VT–
—
—
VCC × 0.2
V
VT+–VT–
VCC × 0.07 —
—
V
|Iin|
—
—
1.0
µA
Vin = 0.5 to VCC
– 0.5 V
—
—
1.0
µA
Vin = 0.5 to VCC
– 0.5 V
Item
Input highlevel voltage
Input lowlevel voltage
Symbol Min
RES, NMI,
MD2–MD0
RES, NMI,
MD2–MD0
VIH
VIL
Other input pins
Schmidt trigger PA13–10, PA2,
input
PA0, PB7–PB0
voltage
Input leak
current
RES
+
NMI, MD2–MD0
3-state leak
Ports A and B,
current (while CS3–CS0, A21–
off)
A0, AD15–AD0
|ITSI|
—
—
1.0
µA
Vin = 0.5 to VCC
– 0.5 V
Input pull-up
MOS current
PA3
–Ip
20
—
300
µA
Vin = 0V
Output highlevel voltage
All output pins
VOH
VCC – 0.5
—
—
V
I OH = –200 µA
VCC – 1.0
—
—
V
I OH = –1 mA
Output low
level voltage
All output pins
—
—
0.4
V
I OL = 1.6 mA
—
—
1.2
V
I OL = 8 mA
RENESAS 446
VOL
Table 19.2 DC Characteristics (3) (cont)
Conditions: VCC = 3.0 V to 5.5 V, VSS = 0 V, φ = 12.5 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products .
Item
Input
capacitance
Current
consumption
Measurement
Conditions
Symbol
Min
Typ
Max
Unit
Cin
—
—
30
pF
NMI
—
—
30
pF
All other input
pins
—
—
20
pF
Vin = 0 V
Input signal
f = 1 MHz
Ta = 25°C
—
65
80
mA
f = 12.5 MHz
Sleep
—
30
50
mA
f = 12.5 MHz
Standby
—
0.01
5.0
µA
Ta ≤ 50°C
—
—
20
µA
50°C < Ta
2.0
—
—
V
RES
Ordinary
operation
RAM stand-by voltage
I CC
VRAM
RENESAS 447
Table 19.3 Permitted Output Current Values
Case A: V CC = 3.0 to 5.5 V, VSS = 0 V, φ = 12.5 MHz, Ta = –20 to +75°C*
Case B: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
Case C: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
12.5 MHz
Item
Symbol
Min
Typ
Max
Unit
Output low-level permissible
current (per pin)
I OL
—
—
10
mA
Output low-level permissible
current (total)
∑ IOL
—
—
80
mA
Output high-level permissible
current (per pin)
–I OH
—
—
2.0
mA
Output high-level permissible
current (total)
–∑ IOH
—
—
25
mA
Caution: To ensure LSI reliability, do not exceed the value for output current given in table 19.3.
RENESAS 448
19.3
AC Characteristics
The following AC timing chart represents the AC characteristics, not signal functions. For signal
functions, see the explanation in the text.
19.3.1
Clock Timing
Table 19.4 Clock Timing
Case A: V CC = 3.0 to 5.5 V, VSS = 0 V, Ta = –20 to +75°C*
Case B: VCC = 5.0 V ±10%, V SS = 0 V, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Case A
Case B
Sym- 12.5 MHz
bol
Min Max
Min
Max
Min
Max
Unit
Figures
EXTAL input high level
pulse width
t EXH
20
—
10
—
10
—
ns
19.1
EXTAL input low level
pulse width
t EXL
20
—
10
—
10
—
ns
EXTAL input rise time
t EXr
—
10
—
5
—
5
ns
EXTAL input fall time
t EXf
—
10
—
5
—
5
ns
Clock cycle time
t cyc
80
—
60
500
50
500
ns
19.1, 19.2
Clock high pulse width
t CH
30
—
20
—
20
—
ns
19.2
Clock low pulse width
t CL
30
—
20
—
20
—
ns
Clock rise time
t Cr
—
10
—
5
—
5
ns
Clock fall time
t Cf
—
10
—
5
—
5
ns
Reset oscillation settling
time
t OSC1 10
—
10
—
10
—
ms
Software standby
oscillation settling time
t OSC2 10
—
10
—
10
—
ms
Item
16.6 MHz
20 MHz
19.3
RENESAS 449
tcyc
tEXH
1/2 VCC
EXTAL
tEXL
VIH
VIL
tEXr
tEXf
Figure 19.1 EXTAL Input Timing
tCYC
tCH
tCL
CK
tCf
tCr
Figure 19.2 System Clock Timing
CK
VCC
tOSC2
tOSC1
RES
Figure 19.3 Oscillation Settling Time
RENESAS 450
19.3.2
Control Signal Timing
Table 19.5 Control Signal Timing
Case A: V CC = 3.0 to 5.5 V, VSS = 0 V, Ta = –20 to +75°C*
Case B: VCC = 5.0 V ±10%, V SS = 0 V, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Case A
12.5 MHz
Case B
16.6 MHz
20 MHz
Item
Symbol Min
Max
Min
Max
Min
Max
Unit
Figure
RES setup time
t RESS
320
—
240
—
200
—
ns
19.4
RES pulse width
t RESW
20
—
20
—
20
—
t cyc
NMI reset setup time
t NMIRS
320
—
240
—
200
—
ns
NMI reset hold time
t NMIRH
320
—
240
—
200
—
ns
NMI setup time
t NMIS
160
—
120
—
100
—
ns
NMI hold time
t NMIH
80
—
60
—
50
—
ns
IRQ0–IRQ7 setup time (edge
detection time)
t IRQES
160
—
120
—
100
—
ns
IRQ0–IRQ7 setup time (level
detection time)
t IRQLS
160
—
120
—
100
—
ns
IRQ0–IRQ7 hold time
t IRQEH
80
—
60
—
50
—
ns
IRQOUT output delay time)
t IRQOD
—
80
—
60
—
50
ns
19.6
Bus request setup time
t BRQS
80
—
60
—
50
—
ns
19.7
Bus acknowledge delay time 1
t BACD1
—
80
—
60
—
50
ns
Bus acknowledge delay time 2
t BACD2
—
80
—
60
—
50
ns
Bus 3-state delay time
t BZD
—
80
—
60
—
50
ns
19.5
RENESAS 451
CK
tRESS
tRESS
RES
tNMIRS
tRESW
tNMIRH
NMI
Figure 19.4 Reset Input Timing
CK
tNMIS
tNMIH
tIRQES
tIRQEH
NMI
IRQ edge
tIRQLS
IRQ level
Figure 19.5 Interrupt Signal Input Timing
RENESAS 452
CK
tIRQOD
tIRQOD
IRQOUT
Figure 19.6 Interrupt Signal Output Timing
CK
tBRQS
BREQ
(Input)
tBRQS
tBACD1
tBACD2
BACK
(Output)
tBZD
RD,
WR, RAS,
CAS, CSn
tBZD
A21–A0
Figure 19.7 Bus Release Timing
RENESAS 453
19.3.3
Bus Timing
Table 19.6 Bus Timing (1)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
RENESAS 454
Item
Symbol
Min
Max
Unit Figures
ns
19.8, 19.9, 19.11–19.14,
19.19, 19.20
19.8, 19.9, 19.20
Address delay time
t AD
—
20*1
CS delay time 1
t CSD1
—
25
ns
CS delay time 2
t CSD2
—
25
ns
CS delay time 3
t CSD3
—
20
ns
CS delay time 4
t CSD4
—
20
ns
t RDAC1
t cyc × 0.65 –
20
—
ns
t cyc × 0.5 – 20
—
ns
t cyc × (n+1.65)
– 20 *3
—
ns
t cyc × (n+1.5) – —
– 20 *3
ns
t cyc × (n+0.65)
– 20 *3
—
ns
t cyc × (n+0.5) – —
– 20 *3
ns
1*6
Access time
35% duty *2
from read strobe
50% duty
2*6
Access time
35%
from read strobe
duty *2
t RDAC2
50% duty
Access time 3*6 35% duty *2
from read strobe
t RDAC3
50% duty
19.19
19.8
19.9, 19.10
19.19
Read strobe delay time
t RSD
—
20
ns
19.8, 19.9, 19.11–19.15,
19.19, 19.24–19.28
19.8, 19.9, 19.11–19.14,
19.19
Read data setup time
t RDS
15
—
ns
Read data hold time
t RDH
0
—
ns
Write strobe delay time 1
t WSD1
—
20
ns
19.9, 19.13, 19.14,
19.19, 19.20
Write strobe delay time 2
t WSD2
—
20
ns
19.9, 19.13, 19.14,
19.19
Write strobe delay time 3
t WSD3
—
20
ns
19.11, 19.12
Write strobe delay time 4
t WSD4
—
20
ns
19.11, 19.12, 19.20
Write data delay time 1
t WDD1
—
35
ns
19.9, 19.13, 19.14, 19
Write data delay time 2
t WDD2
—
20
ns
19.11, 19.12
Write data hold time
t WDH
0
—
ns
19.9, 19.11–19.14
RENESAS 455
Table 19.6 Bus Timing (1) (cont)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol
Min
Max
Unit Figures
Parity output delay time 1
t WPDD1
—
40
ns
19.9, 19.13, 19.14
Parity output delay time 2
t WPDD2
—
20
ns
19.11, 19.12
Parity output hold time
t WPDH
0
—
ns
19.9, 19.11–19.14
Wait setup time
t WTS
14
—
ns
19.10, 19.15, 19.19
Wait hold time
t WTH
10
—
ns
Read data access time 1*6
t ACC1
t cyc – 30* 4
—
ns
19.8, 19.11, 19.12
Read data access time 2*6
t ACC2
t cyc × (n+2) –
30*3
—
ns
19.9, 19.10, 19.13,
19.14
RAS delay time 1
t RASD1
—
20
ns
RAS delay time 2
t RASD2
—
30
ns
19.11–19.14,
19.16–19.18
CAS delay time 1
t CASD1
—
20
ns
19.11
CAS delay time 2
t CASD2
—
20
ns
CAS delay time 3
t CASD3
—
20
ns
19.13, 19.14,
19.16–19.18
t ASC
0
—
ns
19.11, 19.12
t CAC1
t cyc × 0.65 –
19
—
ns
t cyc × 0.5 – 19
—
ns
Column address setup time
Read data
access time
from CAS 1 *6
35%
duty *2
50% duty
Read data access time from
CAS 2 *6
t CAC2
t cyc × (n+1) –
25*3
—
ns
19.13, 19.14, 19.15
Read data access time from
RAS 1 *6
t RAC1
t cyc × 1.5 – 20
—
ns
19.11, 19.12
Read data access time from
RAS 2 *6
t RAC2
t cyc × (n+2.5) – —
20*3
ns
19.13, 19.14, 19.15
High-speed page mode CAS
precharge time
t CP
t cyc × 0.25
ns
19.12
RENESAS 456
—
Table 19.6 Bus Timing (1) (cont)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol
Min
Max
Unit Figures
AH delay time 1
t AHD1
—
20
ns
AH delay time 2
t AHD2
—
20
ns
Multiplexed address delay
time
t MAD
—
30
ns
Multiplexed address hold time
t MAH
0
—
ns
DACK0, DACK1 delay time 1
t DACD1
—
23
ns
DACK0, DACK1 delay time 2
t DACD2
—
23
ns
DACK0, DACK1 delay time 3
t DACD3
—
20
ns
19.9, 19.13, 19.14,
19.19
DACK0, DACK1 delay time 4
t DACD4
—
20
ns
19.11, 19.12
DACK0, DACK1 delay time 5
t DACD5
—
20
ns
t RDD
—
t cyc × 0.35 + 12
ns
—
t cyc × 0.5 + 15
ns
Read delay time
35%
duty *2
50% duty
19.19
19.8, 19.9, 19.11–
19.14, 19.19, 19.20
19.8, 19.9, 19.1119.15, 19.19, 19.2419.28
Data setup time for CAS
t DS
0*5
—
ns
19.11, 19.13
CAS setup time for RAS
t CSR
10
—
ns
19.16, 19.17, 19.18
Row address hold time
t RAH
10
—
ns
19.11, 19.13
Write command hold time
t WCH
15
—
ns
t WCS
0
—
ns
0
—
ns
t cyc —
− 20
ns
Write command
setup time
35% duty *2
50% duty
Access time from CAS
precharge *6
Notes: 1.
2.
3.
4.
5.
6.
t ACP
19.11
19.12
HBS and LBS signals are 25 ns.
When frequency is 10 MHz or more.
n is the number of wait cycles.
Access time from addresses A0 to A21 is tcyc-25.
–5 ns for parity output of DRAM long-pitch access.
It is not necessary to meet the tRDS specification as long as the access time
specification is met.
RENESAS 457
Table 19.7 Bus Timing (2)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol
Min
Max
Unit
Figures
ns
19.8, 19.9, 19.11–19.14,
19.19, 19.20
19.8, 19.9, 19.20
Address delay time
t AD
—
25*1
CS delay time 1
t CSD1
—
30
ns
CS delay time 2
t CSD2
—
25
ns
CS delay time 3
t CSD3
—
25
ns
CS delay time 4
t CSD4
—
25
ns
Access time 1*6 35% duty *2
from read strobe
t RDAC1
t cyc × 0.65 – —
20
ns
t cyc × 0.5 –
20
—
ns
t cyc × (n +
1.65) – 20 *3
—
ns
t cyc × (n +
1.5) – 20 *3
—
ns
t cyc × (n +
0.65) – 20 *3
—
ns
t cyc × (n +
0.5) – 20 *3
—
ns
t RSD
—
25
ns
19.8, 19.9, 19.19
Read data setup time
t RDS
15
—
ns
19.8, 19.9, 19.11–19.14,
Read data hold time
t RDH
0
—
ns
19.19
Write strobe delay time 1
t WSD1
—
25
ns
19.9, 19.13, 19.14, 19.19,
19.20
Write strobe delay time 2
t WSD2
—
25
ns
19.9, 19.13, 19.14, 19.19
Write strobe delay time 3
t WSD3
—
25
ns
19.11, 19.12
Write strobe delay time 4
t WSD4
—
25
ns
19.11, 19.12, 19.20
Write data delay time 1
t WDD1
—
45
ns
19.9, 19.13, 19.14, 19.19
Write data delay time 2
t WDD2
—
25
ns
19.11, 19.12
Write data hold time
t WDH
0
—
ns
19.9, 19.11–19.14
Parity output delay time 1
t WPDD1
—
45
ns
19.9, 19.13, 19.14
Parity output delay time 2
t WPDD2
—
25
ns
19.11, 19.12
Parity output hold time
t WPDH
0
—
ns
19.9, 19.11–19.14
50% duty
Access time 2*6 35% duty *2
from read strobe
t RDAC2
50% duty
Access time 3*6 35% duty *1
from read strobe
t RDAC3
50% duty
Read strobe delay time
RENESAS 458
19.19
19.8
19.9, 19.10
19.19
Table 19.7 Bus Timing (2) (cont)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol
Min
Max
Unit Figures
Wait setup time
t WTS
19
—
ns
Wait hold time
t WTH
10
—
ns
Read data access time 1*6
t ACC1
t cyc – 30* 4
—
ns
19.8, 19.11, 19.12
Read data access time 2*6
t ACC2
t cyc × (n+2) –
30*3
—
ns
19.9, 19.10, 19.13,
19.14
RAS delay time 1
t RASD1
—
25
ns
RAS delay time 2
t RASD2
—
35
ns
19.11–19.14,
19.16–19.18
CAS delay time 1
t CASD1
—
25
ns
19.11
CAS delay time 2
t CASD2
—
25
ns
CAS delay time 3
t CASD3
—
25
ns
19.13, 19.14,
19.16–19.18
t ASC
0
—
ns
19.11, 19.12
t CAC1
t cyc × 0.65 –
19
—
ns
t cyc × 0.5 – 19
—
ns
Column address setup time
Read data access 35%
time from CAS 1*6
duty *2
50% duty
19.10, 19.15, 19.19
Read data access time from
CAS 2 *6
t CAC2
t cyc × (n + 1) – —
25*3
ns
19.13, 19.14, 19.15
Read data access time from
RAS 1 *6
t RAC1
t cyc × 1.5 – 20
—
ns
19.11, 19.12
Read data access time from
RAS 2 *6
t RAC2
t cyc × (n + 2.5) —
– 20 *3
ns
19.13, 19.14, 19.15
High-speed page mode CAS
precharge time
t CP
t cyc × 0.25
—
ns
19.12
AH delay time 1
t AHD1
—
25
ns
19.19
AH delay time 2
t AHD2
—
25
ns
Multiplexed address delay time
t MAD
—
30
ns
Multiplexed address hold time
t MAH
0
—
ns
RENESAS 459
Table 19.7 Bus Timing (2) (cont)
Conditions: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol
Min Max
Unit Figures
DACK0, DACK1 delay time 1
t DACD1
—
25
ns
DACK0, DACK1 delay time 2
t DACD2
—
25
ns
DACK0, DACK1 delay time 3
t DACD3
—
25
ns
19.9, 19.13, 19.14,
19.19
DACK0, DACK1 delay time 4
t DACD4
—
25
ns
19.11, 19.12
DACK0, DACK1 delay time 5
t DACD5
—
25
ns
t RDD
—
t cyc × 0.35 + 12 ns
—
t cyc × 0.5 + 15
ns
19.8, 19.9, 19.11-19.15,
19.19
—
ns
19.11, 19.13
Read delay
time
35%
duty *2
50% duty
19.8, 19.9, 19.11–19.14,
19.19, 19.20
Data setup time for CAS
t DS
0*5
CAS setup time for RAS
t CSR
10
—
ns
19.16, 19.17, 19.18
Row address hold time
t RAH
10
—
ns
19.11, 19.13
t WCH
15
—
ns
Write command 35%
setup time
50% duty
t WCS
0
—
ns
0
—
ns
Access time from CAS
precharge *6
t ACP
t cyc —
−20
ns
Write command hold time
duty *2
Notes 1.
2.
3.
4.
5.
6.
19.11
19.12
HBS and LBS signals are 30 ns.
When frequency is 10 MHz or more
n is the number of wait cycles.
Access time from addresses A0 to A21 is tcyc-25.
–5 ns for parity output of DRAM long-pitch access
It is not necessary to meet the tRDS specification as long as the access time
specification is met.
RENESAS 460
T1
CK
tAD
A21–A0
HBS, LBS
tCSD2
tCSD1
CSn
tRDD
tRDAC1*1
tRSD
RD (Read)
tACC1*2 tRDS
AD15–AD0
DPH, DPL
(Read)
tDACD1
tRDH*3
tDACD2
DACK0
DACK1
Notes: 1.
2.
3.
For t RDAC1 , use t cyc × 0.65 – 20 (for 35% duty) or t cyc × 0.5 – 20 (for 50% duty)
instead of tcyc – t RDD – t RDS.
For tACC1, use t cyc – 30 instead of t cyc – t AD (or tCSD1) – tRDS.
t RDH is measured from A21–A0, CSn, or RD, whichever is negated first.
Figure 19.8 Basic Bus Cycle: One-State Access
RENESAS 461
T1
CK
T2
tAD
A21–A0
HBS, LBS
tCSD2
tCSD1
CSn
tRDD
tRDAC2*1
tRSD
RD (Read)
tACC2*2
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
tDACD1
tWSD1
tRDH*3
tRDS
tDACD2
tWSD2
WRH, WRL,
WR (Write)
tWDH
tWDD1
AD15–AD0
(Write)
tWPDH
tWPDD1
DPH, DPL
(Write)
tDACD3
tDACD3
DACK0
DACK1
(Write)
Notes: 1
2
3
For tRDAC2, use t cyc × (n + 1.65) – 20 (for 35% duty) or tcyc × (n + 1.5) – 20 (for
50% duty) instead of tcyc × (n + 2) – tRDD – tRDS.
For tACC2, use t cyc × (n + 2) – 30 instead of tcyc × (n + 2) – tAD (or tCSD1) – tRDS.
t RDH is measured from A21–A0, CSn, or RD, whichever is negated first.
Figure 19.9 Basic Bus Cycle: Two-State Access
RENESAS 462
T1
TW
T2
CK
A21–A0
HBS, LBS
CSn
tRDAC2*1
RD
(Read)
tACC2*2
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
WRH, WRL,
WR (Write)
AD15–AD0
DPH, DPL
(Write)
DACK0
DACK1
(Write)
tWTS
tWTH
tWTS
tWTH
WAIT
Notes: 1.
2.
For tRDAC2, use t cyc × (n + 1.65) – 20 (for 35% duty) or tcyc × (n + 1.5) – 20 (for
50% duty) instead of tcyc × (n + 2) – tRDD – tRDS.
For tACC2, use t cyc × (n + 2) – 30 instead of tcyc × (n + 2) – tAD (or tCSD1) – tRDS.
Figure 19.10 Basic Bus Cycle: Two States + Wait State
RENESAS 463
Tp
Tr
Tc
CK
tAD
tAD
Row
A21–A0
Column
tRASD1
tRASD2
tRAH
RAS
tDS
tASC
tRDD
CAS
tRSD
tWCH
RD(Read)
WRH, WRL,
(Read)
tDACD1
DACK0
DACK1
(Read)
tCASD1
tDACD2
tCAC1*1
tACC1*2
AD15–AD0
DPH, DPL
(Read)
tRAC1*3
tWSD3
RD(Write)
WRH, WRL,
(Write)
tWCS
tWDD2
AD15–AD0
(Write)
DPH, DPL
(Write)
tRDS
tRDH*4
tWSD4
tWDH
tWPDH
tWPDD2
tDACD4 tDACD5
DACK0
DACK1
(Write)
Notes: 1.
2.
3.
4.
For tCAC1, use t cyc × 0.65 – 19 (for 35% duty) or t cyc × 0.5 – 19 (for 50% duty)
instead of tcyc – tAD – t ASC – tRDS.
For tACC1, use t cyc – 30 instead of tcyc – tAD – t RDS.
For tRAC1, use t cyc × 1.5 – 20 instead of t cyc × 1.5 – t RASD1 – tRDS.
t RDH is measured from A21–A0, RAS, or CAS, whichever is negated first.
Figure 19.11 DRAM Bus Cycle (Short Pitch, Normal Mode)
RENESAS 464
Tp
Tr
Tc
Tc
Tc
Tc
CK
tAD
A21–A0
tAD
Row address
Column address
Column address
Column address
Column address
tRASD2
tRASD1
RAS
tASC
tCP
CAS
tRDD
tRSD
RD(Read)
WRH, WRL,
WR(Read)
tCAC1*1
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
Notes: 1.
2.
3.
4.
5.
tRAC1*3
tACC1*2
tACP
tRDS tRDH*4
tRDH*5
tDACD1 tDACD2
For tCAC1, use tcyc × 0.65 – 19 (for 35% duty) or t cyc × 0.5 – 19 (for 50% duty) instead
of t cyc – t AD – t ASC – tRDS .
It is not necessary to meet the tRDS specification as long as the t CAC1 specification is
met.
For tACC1, use tcyc – 30 instead of t cyc – t AD – t RDS .
It is not necessary to meet the tRDS specification as long as the t ACC1 specification is
met.
For tRAC1, use tcyc × 1.5 – 20 instead of t cyc × 1.5 – t RASD1 – t RDS .
It is not necessary to meet the tRDS specification as long as the t RAC1 specification is
met.
t RDH is measured from A21—A0 or CAS, whichever is negated first.
t RDH is measured from A21—A0, RAS, or CAS, whichever is negated first.
Figure 19.12 (a) DRAM Bus Cycle (Short-Pitch, High-Speed Page Mode: Read)
RENESAS 465
Tp
Tr
Silent
cycle
Tc
Tc
CK
tAD
A21–A0
tAD
Row address
Column address
Column address
tRASD2
tRASD1
RAS
tASC
CAS
RD (Write)
tWSD3
tWSD4
WRH, WRL,
WR (Write)
tWDD2
tWDH
tWPDD2
tWPDH
AD15–AD0
DPH, DPL
(Write)
DPH, DPL
(Write)
tDACD4 tDACD5
tDACD5
DACK0
DACK1
(Write)
Figure 19.12 (b) DRAM Bus Cycle (Short-Pitch, High-Speed Page Mode: Write)
Note: For details of the silent cycle, see section 8.5.5, Burst Operation.
RENESAS 466
Tp
Tr
Tc1
Tc2
CK
tAD
A21–A0
tRASD1
tAD
Row
tRAH
Column
tRASD2
RAS
tDS
tCASD2
CAS
tCASD3
tRDD
tRSD
RD(Read)
WRH, WRL,
(Read)
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
tWCH
tRAC2*3
tACC2*2
tCAC2*1
tRDH*4
tRDS
tDACD2
tDACD1
RD(Write)
WRH, WRL,
(Write)
AD15–AD0
(Write)
DPH, DPL
(Write)
tWSD1
tWSD2
tWDH
tWDD1
tWPDH
tWPDD1
tDACD3
tDACD3
DACK0
DACK1
(Write)
Notes: 1.
2.
3.
4.
For tCAC2, use tcyc × (n + 1) – 25 instead of tcyc × (n + 1) – tCASD2 – tRDS.
For tACC2 , use t cyc × (n + 2) – 30 instead of tcyc × (n + 2) – tAD – t RDS.
For tRAC2, use t cyc × (n + 2.5) – 20 instead of tcyc × (n + 2.5) – tRASD1 – tRDS.
t RDH is measured from A21–A0, CAS, or RAS, whichever is negated first.
Figure 19.13 DRAM Bus Cycle: (Long Pitch, Normal Mode)
RENESAS 467
Tp
Tr
Tc1
Tc2
Tc1
Tc2
CK
tAD
A21–A0
tAD
Row
tRASD1
Column
Column
tRASD2
RAS
tCASD2
CAS
tRSD
WRH, WRL,
(Read)
DACK0
DACK1
(Read)
RD(Write)
WRH, WRL,
(Write)
tCASD3
tRDD
RD(Read)
AD15–AD0
DPH, DPL
(Read)
tCASD3
tRAC2*3
tACC2*2
tCAC2*1
tRDS tRDH*4
tDACD1
tDACD2
tWSD1
tWDD1
tWSD2
tRDH*5
tDACD1
tWSD1
tDACD2
tWSD2
tWDH tWDD1
tWDH
tWPDH tWPDD1
tWPDH
AD15–AD0
(Write)
tWPDD1
DPH, DPL
(Write)
DACK0
DACK1
(Write)
Notes: 1.
2.
3.
4.
5.
tDACD3
tDACD3
tDACD3
tDACD3
For tCAC2, use t cyc × (n + 1) – 25 instead of tcyc × (n + 1) – tCASD2 – tRDS.
For tACC2, use t cyc × (n + 2) – 30 instead of tcyc × (n + 2) – tAD – t RDS.
For tRAC2, use t cyc × (n + 2.5) – 20 instead of tcyc × (n + 2.5) – tRASD1 – tRDS.
t RDH is measured from A21–A0 or CAS, whichever is negated first.
t RDH is measured from A21–A0, RAS, or CAS whichever is negated first.
Figure 19.14 DRAM Bus Cycle: (Long Pitch, High-Speed Page Mode)
RENESAS 468
Tp
Tr
Tc1
Tw
Tc2
CK
A21–A0
Row
Column
tRSD
RAS
CAS
tRDD
RD(Read)
WRH, WRL,
(Read)
tCAC2*1
tACC2*2
tRAC2*3
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
RD(Write)
WRH, WRL,
(Write)
AD15–AD0
(Write)
DPH, DPL
(Write)
DACK0
DACK1
(Write)
tWTS tWTH
tWTS tWTH
WAIT
Notes: 1.
2.
3.
For tCAC2, use t cyc × (n + 1) – 25 instead of tcyc × (n + 1) – tCASD2 – tRDS.
For tACC2, use t cyc × (n + 2) – 30 instead of tcyc × (n + 2) – tAD – t RDS.
For tRAC2, use t cyc × (n + 2.5) – 20 instead of tcyc × (n + 2.5) – tRASD1 – tRDS.
Figure 19.15 DRAM Bus Cycle: (Long Pitch, High-Speed Page Mode + Wait State)
RENESAS 469
TRp
TRr
TRc
CK
tRASD1
tRASD2
RAS
tCSR
tCASD3
tCASD2
CAS
WRH,
WRL
Figure 19.16 CAS-before-RAS Refresh (Short Pitch)
TRp
TRr
TRc
TRc
CK
tRASD1
RAS
tRASD2
tCSR
tCASD3
tCASD2
CAS
WRH,
WRL
Figure 19.17 CAS-before-RAS Refresh (Long Pitch)
RENESAS 470
TRp
TRr
TRc
TRcc
CK
tRASD1
tRASD2
RAS
tCSR
tCASD3
tCASD2
CAS
Figure 19.18 Self Refresh
RENESAS 471
T1
T2
T3
T4
CK
tAD
A21–A0
HBS, LBS
tCSD3
tCSD4
CS6
tAHD1
tAHD2
AH
tRDD
tRSD
RD
(Read)
tMAD
AD15–AD0
(Read)
tMAH
tRDAC3
Address
tRDH
Data
(input)
tDACD1
tDACD2
DACK0
DACK1
(Read)
tWSD1
tWSD2
WRH, WRL,
WR (Write)
tMAD
AD15–AD0
(Write)
tMAH
tWDD1
Data (output)
Address
tDACD3
DACK0
DACK1
(Write)
tWDH
tDACD3
tWTH
tWTS
WAIT
Figure 19.19 Address/Data Multiplex I/O Bus Cycle
RENESAS 472
T1
CK
tAD
A21–A0
HBS, LBS
tCSD2
tCSD1
CSn
tWSD1
tWSD4
tDACD1
tDACD2
WRH, WRL,
WR (Write)
DACK0
DACK1
(Write)
Figure 19.20 DMA Single Transfer/1 State Access Write
RENESAS 473
Table 19.8 Bus Timing (3)
Conditions: VCC = 3.0 to 5.5 V, VSS = 0 V, φ = 12.5 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol Min
Max
Unit
Figures
Address delay time
t AD
—
40
ns
19.21, 19.22, 19.24–
19.27, 19.32, 19.33
CS delay time 1
t CSD1
—
40
ns
19.21, 19.22, 19.33
CS delay time 2
t CSD2
—
40
ns
CS delay time 3
t CSD3
—
40
ns
t CSD4
—
40
ns
t RDAC1
t cyc × 0.65 – 35
—
ns
—
ns
—
ns
—
ns
—
ns
—
ns
CS delay time 4
1*4
duty *1
2*4
duty *1
Access time
35%
from read strobe 50% duty
Access time
35%
from read strobe 50% duty
3*4
duty *1
t cyc × 0.5 – 35
t RDAC2
t cyc × (n+1.65) –
t cyc × (n+1.5) –
35*2
35*2
t cyc × (n+0.65) –
35*2
19.32
19.21,
19.22, 19.23
Access time
35%
from read strobe 50% duty
t RDAC3
Read strobe delay time
t RSD
—
40
ns
19.21, 19.22, 19.32
Read data set-up time
t RDS
30
—
ns
19.21, 19.22,
Read data hold time
t RDH
0
—
ns
19.24-19.27, 19.32
Write strobe delay time 1
t WSD1
—
40
ns
19.22, 19.26, 19.27,
19.32, 19.33
Write strobe delay time 2
t WSD2
—
30
ns
19.22, 19.26, 19.27,
19.32
Write strobe delay time 3
t WSD3
—
40
ns
19.24, 19.25
Write strobe delay time 4
t WSD4
—
40
ns
19.24, 19.25, 19.33
Write data delay time 1
t WDD1
—
70
ns
19.22, 19.26, 19.27,
19.32
Write data delay time 2
t WDD2
—
40
ns
19.24, 19.25
Write data hold time
t WDH
–10
—
ns
19.22, 19.24–19.27,
19.32
Parity output delay time 1
t WPDD1
—
80
ns
19.22, 19.24, 19.27
Parity output delay time 2
t WPDD2
—
40
ns
19.24, 19.25
Parity output hold time
t WPDH
–10
—
ns
19.22, 19.24–19.27
RENESAS 474
t cyc × (n+0.5) –
35*2
19.32
Table 19.8 Bus Timing (3) (cont)
Conditions: VCC = 3.0 to 5.5 V, VSS = 0 V, φ = 12.5 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Item
Symbol Min
Max
Unit
Figures
Wait setup time
t WTS
40
—
ns
19.23, 19.28, 19.32
Wait hold time
t WTH
10
—
ns
Read data access time 1*4
t ACC1
t cyc – 44
—
ns
19.21, 19.24, 19.25
Read data access time 2*4
t ACC2
t cyc × (n+2) – 44*2
—
ns
19.22, 19.23, 19.26,
19.28
RAS delay time 1
t RASD1
—
40
ns
RAS delay time 2
t RASD2
—
40
ns
19.24–19.27, 19.29–
19.31
CAS delay time 1
t CASD1
—
40
ns
19.24
CAS delay time 2
t CASD2
—
40
ns
CAS delay time 3
t CASD3
—
40
ns
19.26, 19.27, 19.29–
19.31
t ASC
0
—
ns
19.24, 19.25
Read data
35%
access time from 50% duty
CAS 1 *4
t CAC1
t cyc × 0.65 – 35
—
ns
t cyc × 0.5 – 35
—
ns
Read data access time from
CAS 2 *4
t CAC2
t cyc × (n+1) – 35*2
—
ns
19.26, 19.27, 19.28
Read data access time from
RAS 1 *4
t RAC1
t cyc × 1.5 – 35
—
ns
19.24, 19.25
Read data access time from
RAS 2 *4
t RAC2
t cyc × (n+2.5) – 35*2
—
ns
19.26, 19.27, 19.28
High-speed page mode CAS
precharge time
t CP
t cyc × 0.25
—
ns
19.25
AH delay time 1
t AHD1
—
40
ns
19.32
AH delay time 2
t AHD2
—
40
ns
Multiplexed address delay
time
t MAD
—
40
ns
Multiplexed address hold
time
t MAH
–10
—
ns
Column address setup time
duty *1
RENESAS 475
Table 19.8 Bus Timing (3) (cont)
Conditions: VCC = 3.0 to 5.5 V, VSS = 0 V, φ = 12.5 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products
Item
Symbol Min Max
Unit
Figures
19.21, 19.22, 19.24–
19.27, 19.32, 19.33
DACK0, DACK1 delay time 1 t DACD1
—
40
ns
DACK0, DACK1 delay time 2 t DACD2
—
40
ns
DACK0, DACK1 delay time 3 t DACD3
—
40
ns
19.22, 19.26, 19.27,
19.32
DACK0, DACK1 delay time 4 t DACD4
—
40
ns
19.24, 19.25
DACK0, DACK1 delay time 5 t DACD5
—
40
ns
—
t cyc × 0.35 + 35
ns
—
t cyc × 0.5 + 35
ns
19.21, 19.22, 19.2419.28, 19.32
—
ns
19.24, 19.26
Read delay time 35%
duty *1
t RDD
50% duty
Data setup time for CAS
t DS
0* 3
CAS setup time for RAS
t CSR
10
—
ns
19.29–19.31
Row address hold time
t RAH
10
—
ns
19.24, 19.26
t WCH
15
—
ns
Write command 35%
setup time
50% duty
t WCS
0
—
ns
t WCS
0
—
ns
Access time from CAS
precharge *4
t ACP
tcyc —
-20
ns
Write command hold time
duty *1
Notes: 1.
2.
3.
4.
19.24
19.25
When frequency is 10 MHz or more.
n is the number of wait cycles.
–5 ns for parity output of DRAM long-pitch access
It is not necessary to meet the tRDS specification as long as the access time
specification is met.
RENESAS 476
T1
CK
tAD
A21–A0
HBS, LBS
tCSD2
tCSD1
CSn
tRDD
tRDAC1*1
tRSD
RD (Read)
tACC1*2 tRDS
AD15–AD0
DPH, DPL
(Read)
tDACD1
tRDH*3
tDACD2
DACK0
DACK1
Notes: 1.
2.
3.
For tRDAC1, use t cyc × 0.65 – 35 (for 35% duty) or t cyc × 0.5 – 35 (for 50% duty)
instead of tcyc – tRDD – t RDS.
For tACC1, use t cyc – 44 instead of tcyc – tAD (or tCSD1) – tRDS.
t RDH is measured from A21–A0, CSn, or RD, whichever is negated first.
Figure 19.21 Basic Bus Cycle: One-State Access
RENESAS 477
T1
T2
CK
tAD
A21–A0
HBS, LBS
tCSD2
tCSD1
CSn
tRDAC2*1
tRDD
tRSD
RD (Read)
tACC2*2
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
tDACD1
tWSD1
tRDH*3
tRDS
tDACD2
tWSD2
WRH, WRL,
WR (Write)
tWDH
tWDD1
AD15–AD0
(Write)
tWPDH
tWPDD1
DPH, DPL
(Write)
tDACD3
tDACD3
DACK0
DACK1
(Write)
Notes: 1.
2.
3.
For tRDAC2, use t cyc × (n + 1.65) – 35 (for 35% duty) or tcyc × (n + 1.5) – 35 (for
50% duty) instead of tcyc × (n + 2) – tRDD – tRDS.
For tACC2, use t cyc × (n + 2) – 44 instead of tcyc × (n + 2) – tAD (or tCSD1) – tRDS.
t RDH is measured from A21–A0, CSn, or RD, whichever is negated first.
Figure 19.22 Basic Bus Cycle: Two-State Access
RENESAS 478
T1
TW
T2
CK
A21–A0
HBS, LBS
CSn
tRDAC2*1
RD
(Read)
tACC2*2
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
WRH, WRL,
WR (Write)
AD15–AD0
DPH, DPL
(Write)
DACK0
DACK1
(Write)
tWTS
tWTH
tWTS
tWTH
WAIT
Notes: 1.
2.
For tRDAC2, use t cyc × (n + 1.65) – 35 (for 35% duty) or tcyc × (n + 1.5) – 35 (for
50% duty) instead of tcyc × (n + 2) – tRDD – tRDS.
For tACC2, use t cyc × (n + 2) – 44 instead of tcyc × (n + 2) – tAD (or tCSD1) – tRDS.
Figure 19.23 Basic Bus Cycle: Two States + Wait State
RENESAS 479
Tp
Tr
Tc
CK
tAD
tAD
Row
A21–A0
tRASD1
Column
tRASD2
tRAH
RAS
tDS
tASC
CAS
tCASD1
tRSD
tRDD
RD(Read)
tWCH
WRH, WRL,
(Read)
tDACD1
tDACD2
DACK0
DACK1
(Read)
tACC1*2
tRAC1*3
AD15–AD0
DPH, DPL
(Read)
RD(Write)
WRH, WRL,
(Write)
tWSD3
tRDH*4
tRDS
tWCS
tWDD2
AD15–AD0
(Write)
DPH, DPL
(Write)
tCAC1*1
tWSD4
tWDH
tWPDH
tWPDD2
tDACD4 tDACD5
DACK0
DACK1
(Write)
Notes: 1.
2.
3.
4.
For tCAC1, use t cyc × 0.65 – 35 (for 35% duty) or t cyc × 0.5 – 35 (for 50% duty)
instead of tcyc – tAD – t ASC – tRDS.
For tACC1, use t cyc – 44 instead of tcyc – tAD – t RDS.
For tRAC1, use t cyc × 1.5 – 35 instead of t cyc × 1.5 – t RASD1 – tRDS.
t RDH is measured from A21–A0, RAS, or CAS, whichever is negated first.
Figure 19.24 DRAM Bus Cycle (Short Pitch, Normal Mode)
RENESAS 480
Tp
Tr
Tc
Tc
Tc
Tc
CK
tAD
A21–A0
tAD
Row address
Column address
Column address
Column address
Column address
tRASD2
tRASD1
RAS
tASC
tCP
CAS
tRDD
tRSD
RD(Read)
WRH, WRL,
WR(Read)
tCAC1*1
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
Notes: 1.
2.
3.
4.
5.
tRAC1*3
tACC1*2
tACP
tRDS tRDH*4
tRDH*5
tDACD1 tDACD2
For tCAC1, use tcyc × 0.65 – 35 (for 35% duty) or t cyc × 0.5 – 35 (for 50% duty) instead
of t cyc – t AD – t ASC – tRDS .
It is not necessary to meet the tRDS specification as long as the t CAC1 specification is
met.
For tACC1, use tcyc – 44 instead of t cyc – t AD – t RDS .
It is not necessary to meet the tRDS specification as long as the t ACC1 specification is
met.
For tRAC1, use tcyc × 1.5 – 35 instead of t cyc × 1.5 – t RASD1 – t RDS .
It is not necessary to meet the tRDS specification as long as the t RAC1 specification is
met.
t RDH is measured from A21—A0 or CAS, whichever is negated first.
t RDH is measured from A21—A0, RAS, or CAS, whichever is negated first.
Figure 19.25 (a) DRAM Bus Cycle (Short-Pitch, High-Speed Page Mode: Read)
RENESAS 481
Tp
Tr
Silent
cycle
Tc
Tc
CK
tAD
tAD
A21–A0
Row address
Column address
Column address
tRASD2
tRASD1
RAS
tASC
CAS
RD (Write)
tWSD3
tWSD4
WRH, WRL,
WR (Write)
tWDD2
tWDH
tWPDD2
tWPDH
AD15–AD0
DPH, DPL
(Write)
DPH, DPL
(Write)
tDACD4 tDACD5
tDACD5
DACK0
DACK1
(Write)
Figure 19.25 (b) DRAM Bus Cycle (Short-Pitch, High-Speed Page Mode: Write)
Note: For details of the silent cycle, see section 8.5.5, Burst Operation.
RENESAS 482
Tp
Tr
Tc1
Tc2
CK
tAD
A21–A0
tRASD1
tAD
Row
tRAH
Column
tRASD2
RAS
tDS
tCASD2
CAS
tCASD3
tRDD
tRDS
RD(Read)
tWCH
WRH, WRL,
(Read)
tCAC2*1
tACC2*2
tRAC2*3
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
tRDS
tRDH*4
tDACD2
tDACD1
RD(Write)
WRH, WRL,
(Write)
AD15–AD0
(Write)
DPH, DPL
(Write)
tWSD1
tWSD2
tWDH
tWDD1
tWPDH
tWPDD1
tDACD3
tDACD3
DACK0
DACK1
(Write)
Notes: 1.
2.
3.
4.
For tCAC2, use t cyc × (n + 1) – 35 instead of tcyc × (n + 1) – tCASD2 – tRDS.
For tACC2, use t cyc × (n + 2) – 44 instead of tcyc × (n + 2) – tAD – t RDS.
For tRAC2, use t cyc × (n + 2.5) – 35 instead of tcyc × (n + 2.5) – tRASD1 – tRDS.
t RDH is measured from A21–A0, CAS, or RAS, whichever is negated first.
Figure 19.26 DRAM Bus Cycle: (Long Pitch, Normal Mode)
RENESAS 483
Tp
Tr
Tc1
Tc2
Tc1
Tc2
CK
tAD
A21–A0
tAD
Row
tRASD1
Column
Column
tRASD2
RAS
tCASD2
CAS
tRSD
WRH, WRL,
(Read)
DACK0
DACK1
(Read)
RD(Write)
WRH, WRL,
(Write)
tCASD3
tRDD
RD(Read)
AD15–AD0
DPH, DPL
(Read)
tCASD3
tRAC2*3
tACC2*2
tCAC2*1
tRDS tRDH*4
tDACD1
tDACD2
tWSD1
tWDD1
tWSD2
tRDH*5
tDACD1
tWSD1
tDACD2
tWSD2
tWDH tWDD1
tWDH
tWPDH tWPDD1
tWPDH
AD15–AD0
(Write)
tWPDD1
DPH, DPL
(Write)
DACK0
DACK1
(Write)
Notes: 1.
2.
3.
4.
5.
tDACD3
tDACD3
tDACD3
tDACD3
For tCAC2, use t cyc × (n + 1) – 35 instead of tcyc × (n + 1) – tCASD2 – tRDS.
For tACC2, use t cyc × (n + 2) – 44 instead of tcyc × (n + 2) – tAD – t RDS.
For tRAC2, use t cyc × (n + 2.5) – 35 instead of tcyc × (n + 2.5) – tRASD1 – tRDS.
t RDH is measured from A21–A0 or CAS, whichever is negated first.
t RDH is measured from A21–A0, RAS, or CAS whichever is negated first.
Figure 19.27 DRAM Bus Cycle: (Long Pitch, High-Speed Page Mode)
RENESAS 484
Tp
Tr
Tc1
Tw
Tc2
CK
Row
A21–A0
Column
tRDS
RAS
tRDD
CAS
RD(Read)
WRH, WRL,
(Read)
tCAC2*1
tACC2*2
tRAC2*3
AD15–AD0
DPH, DPL
(Read)
DACK0
DACK1
(Read)
RD(Write)
WRH, WRL,
(Write)
AD15–AD0
(Write)
DPH, DPL
(Write)
DACK0
DACK1
(Write)
tWTS tWTH
tWTS tWTH
WAIT
Notes: 1.
2.
3.
For tCAC2, use t cyc × (n + 1) – 35 instead of tcyc × (n + 1) – tCASD2 – tRDS.
For tACC2, use t cyc × (n + 2) – 44 instead of tcyc × (n + 2) – tAD – t RDS.
For tRAC2, use t cyc × (n + 2.5) – 35 instead of tcyc × (n + 2.5) – tRASD1 – tRDS.
Figure 19.28 DRAM Bus Cycle: (Long Pitch, High-Speed Page Mode + Wait State)
RENESAS 485
TRp
TRr
TRc
CK
tRASD1
RAS
tRASD2
tCSR
tCASD3
tCASD2
CAS
WRH,
WRL
Figure 19.29 CAS-before-RAS Refresh (Short Pitch)
TRp
TRr
TRc
TRc
CK
tRASD1
RAS
tRASD2
tCSR
tCASD3
tCASD2
CAS
WRH,
WRL
Figure 19.30 CAS-before-RAS Refresh (Long Pitch)
RENESAS 486
TRp
TRr
TRc
TRcc
CK
tRASD1
RAS
tCSR
tRASD2
tCASD3
tCASD2
CAS
Figure 19.31 Self Refresh
RENESAS 487
T1
T2
T3
T4
CK
tAD
A21–A0
HBS, LBS
tCSD3
tCSD4
CS6
tAHD1
tAHD2
AH
tRDD
tRSD
RD
(Read)
tMAD
AD15–AD0
(Read)
tMAH
tRDAC3
Address
tRDH
Data
(input)
tDACD1
tDACD2
DACK0
DACK1
(Read)
tWSD1
tWSD2
WRH, WRL,
WR (Write)
tMAD
AD15–AD0
(Write)
tMAH
tWDD1
Data (output)
Address
tDACD3
DACK0
DACK1
(Write)
tWDH
tDACD3
tWTH
tWTS
WAIT
Figure 19.32 Address/Data Multiplex I/O Bus Cycle
RENESAS 488
T1
CK
tAD
A21–A0
HBS, LBS
tCSD2
tCSD1
CSn
tWSD1
tWSD4
tDACD1
tDACD2
WRH, WRL,
WR (Write)
DACK0
DACK1
(Write)
Figure 19.33 DMA Single Transfer/Single State Access Write
RENESAS 489
19.3.4
DMAC Timing
Table 19.9 DMAC Timing
Case A: V CC = 3.0 to 5.5 V, VSS = 0 V, Ta = –20 to +75°C*
Case B: VCC = 5.0 V ±10%, V SS = 0 V, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Case A
12.5 MHz
Case B
16.6 MHz
20 MHz
Item
Symbol Min
Max
Min
Max
Min
Max
Unit
Figure
DREQ0, DREQ1 setup time
t DRQS
80
—
40
—
27
—
ns
19.34
DREQ0, DREQ1 hold time
t DRQH
30
—
30
—
30
—
ns
DREQ0, DREQ1 low level width
t DRQW
1.5
—
1.5
—
1.5
—
t cyc
CK
tDRQS
DREQ0, DREQ1
level
tDRQS
tDRQH
DREQ0, DREQ1
edge
tDRQS
DREQ0, DREQ1
level release
Figure 19.34 DREQ0, DREQ1 Input Timing (1)
CK
DREQ0, DREQ1
edge
tDRQW
Figure 19.35 DREQ0, DREQ1 Input Timing (2)
RENESAS 490
19.35
19.3.5
16-bit Integrated Timer Pulse Unit Timing
Table 19.10 16-bit Integrated Timer Pulse Unit Timing
Case A: V CC = 3.0 to 5.5 V, VSS = 0 V, Ta = –20 to +75°C*
Case B: VCC = 5.0 V ±10%, V SS = 0 V, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Case A
12.5 MHz
Case B
16.6 MHz
20 MHz
Item
Symbol
Min
Max
Min
Max
Min
Max
Unit
Figure
Output compare delay time
t TOCD
—
100
—
100
—
100
ns
19.36
Input capture setup time
t TICS
50
—
45
—
35
—
ns
Timer clock input setup time
t TCKS
50
—
50
—
50
—
ns
Timer clock pulse width
(single edge)
t TCKWH/L
1.5
—
1.5
—
1.5
—
t cyc
Timer clock pulse width
(both edges)
t TCKWH/L
2.5
—
2.5
—
2.5
—
t cyc
19.37
CK
tTOCD
Output
compare*1
tTICS
Output
capture*2
Notes: 1.
2.
TIOCA0–TIOCA4, TIOCB0–TIOCB4, TOCXA4, TOCXB4
TIOCA0–TIOCA4, TIOCB0–TIOCB4
Figure 19.36 ITU Input/Output Timing
RENESAS 491
CK
tTCKS
TCLKA–
TCLKD
tTCKWL
tTCKS
tTCKWH
Figure 19.37 ITU Clock Input Timing
RENESAS 492
19.3.6
Programmable Timing Pattern Controller and I/O Port Timing
Table 19.11 Programmable Timing Pattern Controller and I/O Port Timing
Case A: V CC = 3.0 to 5.5 V, VSS = 0 V, φ = 12.5 MHz, Ta = –20 to +75°C*
Case B: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
Case C: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Cases A, B and C
Item
Symbol
Min
Max
Unit
Figure
Port output delay time
t PWD
—
100
ns
19.38
Port input hold time
t PRH
50
—
ns
Port input setup time
t PRS
50
—
ns
T1
T2
T3
CK
tPRS
Ports A, B
(read)
tPRH
tPWD
Ports A, B
(write)
Figure 19.38 Programmable Timing Pattern Controller Output Timing
RENESAS 493
19.3.7
Watchdog Timer Timing
Table 19.12 Watchdog Timer Timing
Case A: V CC = 3.0 to 5.5 V, VSS = 0 V, φ = 12.5 V, Ta = –20 to +75°C*
Case B: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
Case C: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products. Ta = –40 to +85°C for wide-temperature range products.
Cases A, B and C
Item
Symbol
Min
Max
Unit
Figure
WDTOVF delay time
t WOVD
—
100
ns
20.39
CK
tWOVD
tWOVD
WDTOVF
Figure 19.39 Watchdog Timer Output Timing
RENESAS 494
19.3.8
Serial Communications Interface Timing
Table 19.13 Serial Communications Interface Timing
Case A: V CC = 3.0 to 5.5 V, VSS = 0 V, φ = 12.5 MHz, Ta = –20 to +75°C*
Case B: VCC = 5.0 V ±10%, V SS = 0 V, φ = 16.6 MHz, Ta = –20 to +75°C*
Case C: VCC = 5.0 V ±10%, V SS = 0 V, φ = 20 MHz, Ta = –20 to +75°C*
*: Normal products: Ta = –40 to +85°C for wide-temperature range products.
Cases A, B and C
Item
Symbol
Min
Max
Unit
Figure
Input clock cycle
t scyc
4
—
t cyc
19.40
Input clock cycle (clocked synchronization)
t scyc
6
—
t cyc
Input clock pulse width
t sckw
0.4
0.6
t scyc
Input clock rise time
t sckr
—
1.5
t cyc
Input clock fall time
t sckf
—
1.5
t cyc
Transmission data delay time (clocked
synchronization)
t TXD
—
100
ns
Receive data setup time (clocked
synchronization)
t RXS
100
—
ns
Receive data hold time (clocked
synchronization)
t RXH
100
—
ns
tSCKW
tSCKr
tSCKf
19.41
SCK0, SCK1
tscyc
Figure 19.40 Input Clock Timing
RENESAS 495
tscyc
SCK0, SCK1
tTXD
TxD0, TxD1
(transmission
data)
tRXS
tRXH
RxD0, RxD1
(reception data)
Figure 19.41 SCI I/O Timing (Clocked Synchronization Mode)
RENESAS 496
19.3.9
AC Characteristics Measurement Conditions
IOL
LSI
output pin
Device under
test output
V Vref
CL
IOH
CL is set as follows for each pin.
30pF: CK, CASH, CASL, CS0–CS7, BREQ, BACK, AH, IRQOUT, RAS, DACK0, DACK1
50pF: A21–A0, AD15–AD0, DPH, DPL, RD, WRH, WRL, HBS, LBS, WR
70pF: All port outputs and peripheral module output pins other than the above.
I OL and IOH values are as shown in section 19.2, DC Characteristics, and table 19.3, Permitted
Output Current Values.
Figure 19.42 Output Load Circuit
RENESAS 497
19.4
Usage Note
The ZTAT version and the mask ROM version satisfy the electrical properties given in this
document. However, effective values of the electrical properties, the operating margin, and the
noise margin may differ with the manufacturing processes, on-chip ROM, and layout patterns.
When conducting a system evaluation test using the ZTAT version, conduct a similar evaluation
test of the mask ROM version before it replaces the ZTAT version.
RENESAS 498
SH7020, SH7021 Hardware Manual
Publication Date: 1st Edition, September 1994
3rd Edition, September 1998
Published by:
Electronic Devices Sales & Marketing Group
Semiconductor & Integrated Circuits Group
Hitachi, Ltd.
Edited by:
Technical Documentation Group
UL Media Co., Ltd.
Copyright © Hitachi, Ltd., 1994. All rights reserved. Printed in Japan.