TI TMS320VC5421GGU

TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
D
D
D
D
D
D
D
D
D
D
D
D
D
D
Independent Subsystems
Each Core Has an Advanced Multibus
Architecture With Three Separate 16-Bit
Data Memory Buses and One Program Bus
40-Bit Arithmetic Logic Unit (ALU)
Including a 40-Bit Barrel-Shifter and Two
40-Bit Accumulators Per Core
Each Core Has a 17-Bit × 17-Bit Parallel
Multiplier Coupled to a 40-Bit Adder for
Non-Pipelined Single-Cycle Multiply/
Accumulate (MAC) Operations
Each Core Has a Compare, Select, and
Store Unit (CSSU) for the Add/Compare
Selection of the Viterbi Operator
Each Core Has an Exponent Encoder to
Compute an Exponent Value of a 40-Bit
Accumulator Value in a Single Cycle
Each Core Has Two Address Generators
With Eight Auxiliary Registers and Two
Auxiliary Register Arithmetic Units
(ARAUs)
16-Bit Data Bus With Data Bus Holder
Feature
512K-Word × 16-Bit Extended Program
Address Space
Total of 256K-Word × 16-Bit Dual- and
Single-Access On-Chip RAM (128K-Word x
16-Bit Shared Memory)
Single-Instruction Repeat and
Block-Repeat Operations
Instructions With 32-Bit-Long Word
Operands
Instructions With 2 or 3 Operand Reads
Fast Return From Interrupts
Arithmetic Instructions With Parallel Store
and Parallel Load
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
Conditional Store Instructions
Output Control of CLKOUT
Output Control of TOUT
Power Consumption Control With IDLE1,
IDLE2, and IDLE3 Instructions
Dual 1.8-V (Core) and 3.3-V (I/O) Power
Supplies for Low-Power, Fast Operations
10-ns Single-Cycle Fixed-Point Instruction
Interprocessor Communication via Two
Internal 8-Element FIFOs
Twelve Channels of Direct Memory Access
(DMA) for Data Transfers With No CPU
Loading (Six Channels Per Subsystem With
External Access)
Six Multichannel Buffered Serial Ports
(McBSPs) With 128-Channel Selection
Capability (3 McBSPs per Subsystem)
16-Bit Host-Port Interface (HPI) Multiplexed
With External Memory Interface Pins
Software-Programmable Phase-Locked
Loop (APLL) Provides Several Clocking
Options (Requires External TTL Oscillator)
Includes JTAG Functionality for In-Circuit
Emulation
On-Chip Scan-Based Emulation Logic,
IEEE Standard 1149-1{ (JTAG) BoundaryScan Logic
Two Software-Programmable Timers
(One Per Subsystem)
Software-Programmable Wait-State
Generator (14 Wait States Maximum)
Provided in 144-pin MicroStart Ball Grid
Array (GGU Suffix) and 144-pin Thin Quad
Flatpack (TQFP) (PGE Suffix) Packages
ADVANCE INFORMATION
D 200-MIPS Dual-Core DSP Consisting of Two
description
The TMS320VC5421 fixed-point digital signal processor (DSP) is a dual-core solution running at 200-MIPS
performance. The ’5421 consists of two DSP subsystems capable of core-to-core communications and a
128K-word zero-wait-state on-chip program memory shared by the two DSP subsystems. Each subsystem
consists of one ’54x DSP core, 32K-word program/data DARAM, 32K-word data SARAM, 2K-word ROM, three
multichannel serial interfaces, xDMA logic, one timer, one APLL, and other miscellaneous circuitry.
Please be aware that an important notice concerning availability, standard warranty, and use in critical applications of
Texas Instruments semiconductor products and disclaimers thereto appears at the end of this data sheet.
MicroStar is a trademark of Texas Instruments Incorporated.
† IEEE Standard 1149.1-1990, IEEE Standard Test-Access Port and Boundary Scan Architecture.
Copyright  1999, Texas Instruments Incorporated
ADVANCE INFORMATION concerns new products in the sampling or
preproduction phase of development. Characteristic data and other
specifications are subject to change without notice.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
1
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
description (continued)
The ’5421 also contains a host-port interface (HPI) that allows the ’5421 to be viewed as a memory-mapped
peripheral to a host processor. The ’5421 is pin-compatible with the TMS320VC5420.
Each subsystem has its separate program and data spaces, allowing simultaneous accesses to program
instructions and data. Two read operations and one write operation can be performed in one cycle. Instructions
with parallel store and application-specific instructions can fully utilize this architecture. Furthermore, data can
be transferred between program and data spaces. Such parallelism supports a powerful set of arithmetic, logic,
and bit-manipulation operations that can all be performed in a single machine cycle. The ’5421 includes the
control mechanisms to manage interrupts, repeated operations, and function calls. In addition, the ’5421 has
128K words of on-chip program memory that can be shared between the two subsystems.
The ’5421 is intended as a high-performance, low-cost, high-density DSP for remote data access or voice-over
IP subsystems. It is designed to maintain the current modem architecture with minimal hardware and software
impacts, thus maximizing reuse of existing modem technologies and development efforts.
ADVANCE INFORMATION
NOTE: This data sheet is designed to be used in conjunction with the TMS320C5000 DSP Family Functional
Overview (literature number SPRU307).
migration from the ’5420 to the ’5421
Customers who are migrating from the ’5420 to the ’5421 need to take into account the following differences
between the two devices.
The memory structure of the ’5421 has been changed to incorporate 128K x 16-bit words of shared memory.
The DMA of the ’5421 has been enhanced to provide access to external, as well as internal memory.
The HPI and DMA memory maps have been changed to incorporate the new memory ’5421.
2K x 16-bit words of ROM have been added to the ’5421 for bootloading purposes only.
The VCO pin on the ’5420 has been replaced with the HOLDA pin on the ’5421 and the HOLD pin was added
to the ’5421 at a previously unused pin location.
The McBSPs have been updated with a new mode that allows 128-channel selection capability.
McBSP CLKX/R pins can be used as inputs to internal clock rate generator for CLKS-like function without
the penalty of extra pins.
The SELA/B pin on ’5421 is changed to type I/O/Z for added functionality.
NOTE: For more detailed information, see the 5420 to 5421 migration issues document .
2
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
109
74
36
73
VSS
PPD15
PPD14
VSS
PPD13
PPD12
A_BFSR0
A_BDR0
A_BCLKR0
A_BFSX0
VSS
CVDD
A_BDX0
A_BCLKX0
MSTRB
DS
PS
B_BCLKX0
B_BDX0
DV DD
V SS
B_BFSX0
B_BCLKR0
B_BDR0
CVDD
V SS
B_BFSR0
R/W
PPA2
PPA3
SELA/B
PPD8
PPD9
PPD10
PPD11
VSS
72
75
35
71
76
34
70
77
33
69
78
32
68
79
31
67
80
30
66
81
29
65
82
28
64
83
27
63
84
26
62
85
25
61
86
24
60
87
23
59
88
22
58
89
21
57
90
20
56
91
19
55
92
18
54
93
17
53
94
16
52
95
15
51
96
14
50
97
13
49
98
12
48
99
11
47
100
10
46
101
9
45
102
8
44
103
7
43
104
6
42
105
5
41
106
4
PPA14
PPA15
VSS
PPA16
PPA17
B_INT0
B_INT1
B_NMI
IS
B_GPIO2/BIO
B_GPIO1
B_GPIO0
B_BFSR1
B_BDR1
CVDD
VSS
B_BCLKR1
B_BFSX1
VSS
B_BDX1
B_BCLKX1
CVDD
VSS
TEST
XIO
B_RS
B_XF
B_CLKOUT
HMODE
HPIRS
PPA13
PPA12
VSS
DVDD
PPA11
PPA10
ADVANCE INFORMATION
111
110
112
113
114
115
116
117
119
118
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
3
40
107
39
108
2
38
1
37
PPD7
PPA8
PPA0
DVDD
PPA9
PPD1
A_INT1
A_NMI
IOSTRB
A_GPIO2/BIO
A_GPIO1
A_RS
A_GPIO0
VSS
VSS
CVDD
A_BFSR1
A_BDR1
A_BCLKR1
A_BFSX1
CVDD
VSS
A_BDX1
A_BCLKX1
A_XF
A_CLKOUT
HOLDA
TCK
TMS
TDI
TRST
EMU1/OFF
DVDD
A_INT0
EMU0
TDO
143
144
VSS
PPD0
PPD5
PPD4
PPD6
A_BFSX2
A_BDX2
A_BFSR2
A_BDR2
A_BCLKR2
VSS
CV DD
A_BCLKX2
READY
DVDD
CLKIN
HOLD
VSSA
AV DD
VSS
B_BCLKX2
B_BDX2
B_BFSX2
B_BCLKR2
CVDD
VSS
B_BDR2
B_BFSR2
PPD2
PPD3
PPA1
PPA5
DV DD
PPA4
PPA6
PPA7
PGE PACKAGE†‡
(TOP VIEW)
† DVDD is the power supply for the I/O pins while CVDD is the power supply for the core CPU. VSS is the ground for both the I/O pins and the core
CPU.
‡ Pin configuration shown for nonmultiplexed mode only. See the pin assignments table for the TMS320VC5421PGE for multiplexed functions of
specific pins and for specific pin numbers.
The TMS320VC5421PGE 144-pin thin quad flatpack (TQFP) is footprint- and pin-compatible with the ’5420.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
3
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
GGU PACKAGE
( BOTTOM VIEW )
13 12 11 10 9
8
7
6
5
4
3
2
1
A
B
C
D
E
F
G
H
J
ADVANCE INFORMATION
K
L
M
N
The pin assignments table for the TMS320VC5421GGU lists each pin name and its associated pin number for
this 144-pin ball grid array (BGA) package, which is footprint- and pin-compatible with the ’5420.
The signal descriptions table lists each pin name, function, and operating mode(s) for the ’5421 device.
4
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Pin Assignments for the TMS320VC5421PGE
(144-Pin Thin Quad Flatpack)
SIGNAL NAME
(MULTIPLEXED)
PIN
NO.
SIGNAL NAME
(NONMULTIPLEXED)
SIGNAL NAME
(MULTIPLEXED)
PIN
NO.
PPD7
HD7
1
PPA8
HA8
2
PPA9
HA9
5
PPD1
HD1
IOSTRB
A_GPIO3/A_TOUT
9
A_GPIO2/BIO
10
A_GPIO0
A_ROMEN
13
VSS
A_BDR1
14
22
6
A_BFSR1
17
CVDD
21
A_XF
25
VSS
A_CLKOUT
TMS
29
TDI
30
DVDD
33
A_INT0
34
37
PPD15
HD15
38
41
PPD12
HD12
42
A_BCLKR0
45
A_BFSX0
46
A_BDX0
49
A_BCLKX0
50
53
B_BCLKX0
54
57
B_BFSX0
58
61
62
VSS
PPD13
PS
HD13
HDS1
VSS
CVDD
18
26
PPA2
HCNTL1/HA2
65
VSS
PPA3
HCNTL0/HA3
66
PPD9
HD9
69
PPD10
HD10
70
PPA10
HA10
73
PPA11
HA11
74
PPA12
HA12
HA13
78
77
PPA13
B_CLKOUT
81
B_XF
82
TEST
85
89
VSS
VSS
86
B_BDX1
VSS
B_GPIO0
93
CVDD
94
B_ROMEN
B_NMI
90
97
B_GPIO1
98
101
B_INT1
102
106
HA6
110
B_HINT/HA1
114
PPA16
HA16
105
PPA7
HA7
109
VSS
PPA6
PPA5
HA5
113
PPA1
B_BFSR2
117
B_BDR2
118
B_BCLKR2
121
B_BFSX2
122
VSS
CLKIN
125
AVDD
DVDD
126
CVDD
133
134
137
VSS
A_BDX2
141
PPD5
129
A_BFSR2
PPD4
HD4
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
ADVANCE INFORMATION
SIGNAL NAME
(NONMULTIPLEXED)
130
138
HD5
142
5
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Pin Assignments for the TMS320VC5421PGE
(144-Pin Thin Quad Flatpack) (Continued)
SIGNAL NAME
(NONMULTIPLEXED)
PPA0
SIGNAL NAME
(MULTIPLEXED)
ADVANCE INFORMATION
SIGNAL NAME
(MULTIPLEXED)
PIN
NO.
DVDD
A_INT1
7
A_NMI
8
A_GPIO1
11
A_RS
12
VSS
A_BCLKR1
15
CVDD
16
19
A_BFSX1
20
A_BDX1
23
A_BCLKX1
24
HOLDA
27
TCK
28
TRST
31
EMU1/OFF
32
EMU0
35
TDO
36
39
VSS
A_BDR0
40
43
47
CVDD
51
DS
B_BDX0
55
DVDD
B_BCLKR0
59
B_BDR0
B_BFSR0
63
R/W
HR/W
64
HD8
68
HD14
A_BFSR0
VSS
MSTRB
HCS
4
44
48
HDS2
52
56
60
SELA/B
PPA18
67
PPD8
PPD11
HD11
71
72
DVDD
75
VSS
VSS
HPIRS
79
HMODE
80
B_RS
83
XIO
84
CVDD
87
B_BCLKX1
88
B_BFSX1
91
B_BCLKR1
92
B_BDR1
95
B_BFSR1
96
B_GPIO2/BIO
99
B_INT0
IS
76
B_GPIO3/B_TOUT
100
103
PPA17
HA17
104
PPA15
HA15
107
PPA14
HA14
108
PPA4
HAS/HA4
111
DVDD
PPD3
HD3
115
PPD2
119
CVDD
120
123
B_BCLKX2
124
127
HOLD
128
131
A_BCLKX2
132
A_BCLKR2
135
A_BDR2
A_BFSX2
139
PPD6
143
VSS
VSS
B_BDX2
VSSA
READY
PPD0
6
SIGNAL NAME
(NONMULTIPLEXED)
3
PPD14
A_HINT/HA0
PIN
NO.
HRDY
HD0
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
112
HD2
116
136
HD6
140
144
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Pin Assignments for the TMS320VC5421GGU
(144-Pin MicroStar Ball Grid Array)
PPD7
SIGNAL NAME
(MULTIPLEXED)
SIGNAL NAME
(NONMULTIPLEXED)
SIGNAL NAME
(MULTIPLEXED)
A1
PPA8
E1
CVDD
F1
A_XF
J1
TMS
K1
VSS
N1
PPD0
A_INT1
D2
A_GPIO1
E2
VSS
H2
A_CLKOUT
J2
TDO
M2
PPD15
HD15
N2
C3
PPD1
HD1
D3
A_BCLKR1
G3
A_BDX1
A_INT0
L3
PPD14
HD14
M3
A_BDX2
B4
A_BFSX2
C4
G4
HD5
A_GPIO0
A_ROMEN
F4
A_BFSX1
PPD13
HD13
HA8
BALL
NO.
A_RS
PPD5
HD7
BALL
NO.
HD0
B1
A2
H3
K4
PPD12
CVDD
A5
VSS
B5
A_BCLKR0
K5
A_BFSX0
L5
CLKIN
A6
DVDD
B6
A_BDX0
K6
A_BCLKX0
L6
AVDD
A7
VSS
B7
DVDD
K7
B_BDX0
L7
B_BCLKX2
A8
B_BDX2
B8
B_BDR0
K8
B_BCLKR0
L8
CVDD
A9
VSS
B9
L9
R/W
HR/W
K9
B_BFSR0
PPD2
HD2
A10
PPD3
IS
B_GPIO3/B_TOUT
E10
B_BFSR1
J10
PPA12
B_CLKOUT
PPA2
HCNTL1/HA2
HD12
HD3
L4
B10
F10
HA12
K10
N10
DVDD
A11
B_INT0
D11
B_GPIO2/BIO
E11
VSS
H11
B_XF
J11
M11
PPD8
PPD9
HD9
PPA16
HA16
HD8
N11
C12
B_INT1
D12
B_BDX1
G12
CVDD
H12
DVDD
L12
VSS
M12
B13
PPA17
VSS
F13
VSS
G13
HMODE
K13
VSS
L13
PPA15
HA15
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
ADVANCE INFORMATION
SIGNAL NAME
(NONMULTIPLEXED)
HA17
C13
7
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Pin Assignments for the TMS320VC5421GGU
(144-Pin MicroStar Ball Grid Array) (Continued)
SIGNAL NAME
(NONMULTIPLEXED)
SIGNAL NAME
(MULTIPLEXED)
SIGNAL NAME
(MULTIPLEXED)
BALL
NO.
C1
A_NMI
D1
A_BDR1
G1
CVDD
H1
EMU1/OFF
L1
EMU0
M1
VSS
B2
PPA0
VSS
F2
A_BFSR1
G2
TDI
K2
DVDD
L2
A3
PPD4
E3
VSS
F3
HOLDA
J3
TRST
K3
VSS
N3
A_BFSR2
A4
D4
IOSTRB
A_BCLKX1
H4
TCK
J4
A_BFSR0
M4
A_BDR0
N4
A_BCLKR2
C5
A_BDR2
D5
VSS
M5
CVDD
N5
HD6
A_GPIO2/BIO
ADVANCE INFORMATION
SIGNAL NAME
(NONMULTIPLEXED)
DVDD
PPD6
PPA9
HA9
A_HINT/HA0
HD4
A_GPIO3/A_TOUT
C2
B3
E4
READY
HRDY
C6
A_BCLKX2
MSTRB
HCS
M6
DS
C7
HOLD
D7
M7
B_BCLKX0
N7
B_BFSX2
C8
B_BCLKR2
D8
B_BFSX0
M8
VSS
N8
B_BDR2
C9
B_BFSR2
D9
VSS
M9
CVDD
N9
VSSA
PS
PPA1
HDS1
B_HINT/HA1
B_BCLKR1
C10
PPA5
G10
TEST
D6
HDS2
HA5
N6
D10
H10
SELA/B
PPA18
L10
PPA3
PPA4
HAS/HA4
B11
VSS
C11
F11
B_BFSX1
G11
B_BDR1
HCNTL0/HA3
M10
PPA13
HA13
K11
PPD10
HD10
L11
PPA6
HA6
HA14
B12
A12
PPA14
B_GPIO1
E12
CVDD
F12
B_RS
J12
HPIRS
K12
PPD11
N12
PPA7
HA7
A13
B_NMI
D13
B_GPIO0
B_ROMEN
E13
B_BCLKX1
H13
XIO
M13
PPA10
PPA11
8
BALL
NO.
HD11
HA11
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
J13
HA10
N13
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Signal Descriptions
NAME
TYPE†
DESCRIPTION
DATA SIGNALS
PPD15 (MSB)
PPD14
PPD13
PPD12
PPD11
PPD10
PPD9
PPD8
PPD7
PPD6
PPD5
PPD4
PPD3
PPD2
PPD1
PPD0 (LSB)
Parallel port address bus. The DSP can access the external memory locations by way of the external
memory interface using PPA[18:0] in external memory interface (EMIF) mode when the XIO pin is logic
high. PPA18 is a secondary output function of the SELA/B pin.
The PPA[17:0] pins are also multiplexed with the HPI interface. In HPI mode (XIO pin is low), the external
address pins PPA[17:0] are used by a host processor for access to the memory map by way of the
on-chip HPI. Refer to the Host-Port Interface (HPI) Signals section of this table for details on the
secondary functions of these pins.
These pins are placed into the high-impedance state when OFF is low.
I/O/Z
ADVANCE INFORMATION
PPA18 (MSB)
PPA17
PPA16
PPA15
PPA14
PPA13
PPA12
PPA11
PPA10
PPA9
PPA8
PPA7
PPA6
PPA5
PPA4‡§
PPA3
PPA2
PPA1
PPA0 (LSB)
Parallel port data bus. The DSP uses this bidirectional data bus to access external memory when the
device is in external memory interface (EMIF) mode (the XIO pin is logic high).
This data bus is also multiplexed with the 16-bit HPI data bus. When in HPI mode, the bus is used to
transfer data between the host processor and internal DSP memory via the HPI. Refer to the HPI section
of this table for details on the secondary functions of these pins.
I/O/Z¶
The data bus includes bus holders to reduce power dissipation caused by floating, unused pins. The bus
holders also eliminate the need for external pullup resistors on unused pins. When the data bus is not
being driven by the ’5421, the bus holders keep data pins at the last driven logic level. The data bus
keepers are disabled at reset and can be enabled/disabled via the BH bit of the BSCR register.
These pins are placed into high-impedance state when OFF is low.
INITIALIZATION, INTERRUPT, AND RESET OPERATIONS
A_INT0§
B_INT0§
A_INT1§
B_INT1§
I
External user interrupts. A_INT0–B_INT0 are prioritized and are maskable by the interrupt mask register
(IMR) and the interrupt mode bit. A_INT1 –B_INT1 can be polled and reset by way of the interrupt flag
register (IFR).
A_NMI§
B_NMI§
I
Nonmaskable interrupt. NMI is an external interrupt that cannot be masked by way of the INTM or the
IMR. When NMI is activated, the processor traps to the appropriate vector location.
A_RS§
B_RS§
I
Reset. RS causes the digital signal processor (DSP) to terminate execution and causes a reinitialization
of the CPU and peripherals. When RS is brought to a high level, execution begins at location 0FF80h
of program memory. RS affects various registers and status bits.
† I = Input, O = Output, S = Supply, Z = High Impedance
‡ This pin has an internal pullup resistor.
§ These pins are Schmitt triggered inputs.
¶ This pin has an internal bus holder controlled by way of the BSCR register in 54x cLEAD core of DSP subsystem A .
# This pin is used by Texas Instruments for device testing and should be left unconnected.
|| This pin has an internal pulldown resistor.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
9
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Signal Descriptions (Continued)
NAME
TYPE†
DESCRIPTION
INITIALIZATION, INTERRUPT, AND RESET OPERATIONS (CONTINUED)
The XIO pin is used to configure the parallel port as a host-port interface (HPI mode when XIO pin is low),
or as an asynchronous memory interface (EMIF mode when XIO pin is high).
XIO
I
At device reset, the XIO pin level determines the initialization value of the MP/MC bit (a bit in the
processor mode status (PMST) register). Refer to the memory section for details.
ADVANCE INFORMATION
GENERAL-PURPOSE I/O PINS
A_XF
B_XF
External flag output (latched software-programmable output-only signal). Bit-addressable. A_XF and
B_XF are placed into the high-impedance state when OFF is low.
O/Z
A_GPIO0
B_GPIO0
I/O/Z
A_GPIO1
B_GPIO1
I/O/Z
General-purpose I/O pins (software-programmable I/O signal). Values can be latched (output) by
writing into the GPIO register. The states of GPIO pins (inputs) can be read by reading the GPIO
register. The GPIO direction is also programmable by way of the DIRn field in the GPIO register.
A_GPIO2/BIO
B_GPIO2/BIO
I/O/Z
General-purpose I/O. These pins can be configured like GPIO0–GPIO1; however, as an input, the pins
operate as the traditional branch control bit (BIO). If application code does not perform BIO-conditional
instructions, these pins operate as general inputs.
A_ROMEN
B_ROMEN
I
General-purpose I/O pins. The secondary function of these pins. In XIO mode,
the ROM enable (ROMEN) pins are used to enable the applicable on
on-chip
chip ROM
after reset.
PRIMARY
A_GPIO3 (A_TOUT)
IOSTRB
I/O/Z
B_GPIO3 (B_TOUT)
O
IS
When the device is in HPI mode and HMODE = 0 (multiplexed), these pins act
according to the general-purpose I/O control register. TOUT bit must be set to “1”
to drive the timer output on the pin
pin. IF TOUT = 0
0, then these pins are
general-purpose I/Os. In EMIF mode (XIO = 1), these signals are active during
I/O space accesses.
MEMORY CONTROL SIGNALS
Program space select signal. The PS signal is asserted during external program space accesses. This
pin is placed into the high-impedance state when OFF is low.
PS‡§
This pin is also multiplexed with the HPI, and functions as the HDS1 data strobe input signal in HPI mode.
Refer to the HPI section of this table for details on the secondary function of this pin.
Data space select signal. The DS signal is asserted during external data space accesses. This pin is
placed into the high-impedance state when OFF is low.
DS‡§
O/Z
This pin is also multiplexed with the HPI, and functions as the HDS2 data strobe input signal in HPI mode.
Refer to the HPI section of this table for details on the secondary function of this pin.
I/O space select signal. The IS signal is asserted during external I/O space accesses. This pin is placed
into the high-impedance state when OFF is low.
IS
MSTRB‡§
This pin is also multiplexed with the general-purpose I/O feature, and functions as the B_GPIO3
(B_TOUT) input/output signal in HPI mode. Refer to the General-Purpose I/O section of this table for
details on the secondary function of this pin.
O/Z
Program and data memory strobe (active in EMIF mode). This pin is placed into the high-impedance
state when OFF is low.
† I = Input, O = Output, S = Supply, Z = High Impedance
‡ This pin has an internal pullup resistor.
§ These pins are Schmitt triggered inputs.
¶ This pin has an internal bus holder controlled by way of the BSCR register in 54x cLEAD core of DSP subsystem A .
# This pin is used by Texas Instruments for device testing and should be left unconnected.
|| This pin has an internal pulldown resistor.
10
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Signal Descriptions (Continued)
NAME
TYPE†
DESCRIPTION
MEMORY CONTROL SIGNALS (CONTINUED)
READY
I
Data-ready input signal. READY indicates that the external device is prepared for a bus transaction to
be completed. If the device is not ready (READY = 0), the processor waits one cycle and checks READY
again. The processor performs the READY detection if at least two software wait states are
programmed.
This pin is also multiplexed with the HPI, and functions as the host-port data ready (output) in HPI mode.
Refer to the HPI section of this table for details on the secondary function of this pin.
Read/write output signal. R/W indicates transfer direction during communication to an external device.
R/W is normally in the read mode (high), unless it is asserted low when the DSP performs a write
operation.
O/Z
This pin is also multiplexed with the HPI, and functions as the host-port read/write input in HPI mode.
Refer to the HPI section of this table for details on the secondary function of this pin.
This pin is placed into the high-impedance state when OFF is low.
I/O space memory strobe. External I/O space is accessible by the CPU and not the direct memory
access (DMA) controller. The DMA has its own dedicated I/O space that is not accessible by the CPU.
IOSTRB
O/Z
This pin is also multiplexed with the general-purpose I/O feature, and functions as the A_GPIO3
(A_TOUT) signal in HPI mode. Refer to the General Purpose I/O section of this table for details on the
secondary function of this pin.
This pin is placed into the high-impedance state when OFF is low.
PRIMARY
For HPI access (XIO=0), SELA/B is an input.
See Table 3 for a truth table of SELA/B, HMODE, and XIO pins and functionality.
PPA18
O/Z
HOLD‡
I
HOLDA
O/Z
SELA/B
I
For external memory accesses (XIO=1), SELA/B is multiplexed as output PPA18.
See the PPA signal descriptions. These pins are placed into the high-impedance
state when OFF is low.
Hold. HOLD is asserted to request control of the address, data, and control lines. When
acknowledged, these lines go into the high-impedance state.
Hold acknowledge. HOLDA indicates to the external circuitry that the processor is in a hold state and
that the address, data, and control lines are in the high-impedance state, allowing them to be available
to the external circuitry. HOLDA also goes into the high-impedance state when OFF is low.
CLOCKING SIGNALS
A_CLKOUT
B_CLKOUT
O/Z
Master clock output signal. CLKOUT cycles at the machine-cycle rate of the CPU. The internal machine
cycle is bounded by the falling edges of this signal. The CLKOUT pin can be turned off by writing a “1”
to the CLKOFF bit of the PMST register. CLKOUT goes into the high-impedance state when EMU1/OFF
is low.
CLKIN§
I
Input clock to the device. CLKIN connects to an oscillator circuit/device (PLL).
† I = Input, O = Output, S = Supply, Z = High Impedance
‡ This pin has an internal pullup resistor.
§ These pins are Schmitt triggered inputs.
¶ This pin has an internal bus holder controlled by way of the BSCR register in 54x cLEAD core of DSP subsystem A .
# This pin is used by Texas Instruments for device testing and should be left unconnected.
|| This pin has an internal pulldown resistor.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
11
ADVANCE INFORMATION
R/W
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Signal Descriptions (Continued)
NAME
TYPE†
DESCRIPTION
MULTICHANNEL BUFFERED SERIAL PORT 0, 1, AND 2 SIGNALS
ADVANCE INFORMATION
A_BCLKR0‡§
B_BCLKR0‡§
A_BCLKR1‡§
B_BCLKR1‡§
A_BCLKR2‡§
B_BCLKR2‡§
A_BCLKX0‡§
B_BCLKX0‡§
A_BCLKX1‡§
B_BCLKX1‡§
A_BCLKX2‡§
B_BCLKX2‡§
I/O/Z
Receive clocks. BCLKR serves as the serial shift clock for the buffered serial-port receiver. Input from
an external clock source for clocking data into the McBSP. When not being used as a clock, these pins
can be used as general-purpose I/O by setting RIOEN = 1.
BCLKR can be configured as an output by the way of the CLKRM bit in the PCR register.
These pins are placed into the high-impedance state when OFF is low.
I/O/Z
Transmit clocks. Clock signal used to clock data from the transmit register. This pin can also be
configured as an input by setting the CLKXM = 0 in the PCR register. BCLKX can be sampled as an input
by way of the IN1 bit in the SPC register. When not being used as a clock, these pins can be used as
general-purpose I/O by setting XIOEN = 1.
These pins are placed into the high-impedance state when OFF is low.
A_BDR0
B_BDR0
A_BDR1
B_BDR1
A_BDR2
B_BDR2
I
Buffered serial data receive (input) pin. When not being used as data-receive pins, these pins can be
used as general-purpose I/O by setting RIOEN = 1.
A_BDX0
B_BDX0
A_BDX1
B_BDX1
A_BDX2
B_BDX2
O/Z
Buffered serial-port transmit (output) pin. When not being used as data-transmit pins, these pins can be
used as general-purpose I/O by setting XIOEN = 1. These pins are placed into the high-impedance state
when OFF is low.
I/O/Z
Frame synchronization pin for buffered serial-port input data. The BFSR pulse initiates the receive-data
process over the BDR pin. When not being used as data-receive synchronization pins, these pins can
be used as general-purpose I/O by setting RIOEN = 1. These pins are placed into the high-impedance
state when OFF is low.
I/O/Z
Buffered serial-port frame synchronization pin for transmitting data. The BFSX pulse initiates the
transmit-data process over the BDX pin. If RS is asserted when BFSX is configured as output, then
BFSX is turned into input mode by the reset operation. When not being used as data-transmit
synchronization pins, these pins can be used as general-purpose I/O by setting XIOEN = 1. These pins
are placed into the high-impedance state when OFF is low.
A_BFSR0
B_BFSR0
A_BFSR1
B_BFSR1
A_BFSR2
B_BFSR2
A_BFSX0
B_BFSX0
A_BFSX1
B_BFSX1
A_BFSX2
B_BFSX2
HOST-PORT INTERFACE (HPI) SIGNALS
PRIMARY
HA[17:0]
I
PPA[17:0]
O
These pins are multiplexed with the external interface pins and are used by the
HPI when the subsystem is in HPI mode (XIO = 0, MP/MC = 0).
See the PPA signal descriptions. These pins are placed into the high-impedance
state when OFF is low.
† I = Input, O = Output, S = Supply, Z = High Impedance
‡ This pin has an internal pullup resistor.
§ These pins are Schmitt triggered inputs.
¶ This pin has an internal bus holder controlled by way of the BSCR register in 54x cLEAD core of DSP subsystem A .
# This pin is used by Texas Instruments for device testing and should be left unconnected.
|| This pin has an internal pulldown resistor.
12
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Signal Descriptions (Continued)
NAME
TYPE†
DESCRIPTION
HOST-PORT INTERFACE (HPI) SIGNALS (CONTINUED)
PRIMARY
Parallel bidirectional data bus. These pins are multiplexed with the external
interface pins and are used as an HPI interface when XIO = 0.
HD[15:0]
I/O/Z
PPD[15:0]
I/O/Z
The data bus includes bus holders to reduce power dissipation caused by
floating, unused pins. The bus holders also eliminate the need for external pullup
resistors on unused pins. When the data bus is not being driven by the ’5421, the
bus holders keep address pins at the last driven logic level. The data bus keepers
are disabled at reset and can be enabled/disabled via the BH bit of the BSCR
register.
HCNTL0
HCNTL1
HAS‡§
HCS‡§
HDS1‡§
HDS2‡§
HR/W
HRDY
I
I
PPA3
PPA2
PPA4‡§
I
MSTRB‡§
I
PS‡§
DS‡§
I
O/Z
R/W
READY
O/Z
HPI control inputs. Use PPA3 and PPA2 for the HCNTL0 and HCNTL1 values
during the HPI HPIC, HPIA, and HPID reads/writes. Only used in multiplexed
address/data mode (HMODE = 0).
These pins are shared with the external memory interface and are only used by
the HPI when the interface is in HPI mode (XIO pin is low). These pins are placed
into the high-impedance state when OFF is low.
O/Z
Address strobe input. Hosts with multiplexed address and data pins require HAS
to latch the address in the HPIA register. This signal is only used in HPI
multiplexed address/data mode (HMODE pin is low).
This pin is shared with the external memory interface and is only used by the HPI
when the interface is in HPI mode (XIO pin is low). This pin is placed into the
high-impedance state when OFF is low.
O/Z
HPI chip-select signal. This signal must be active during HPI transfers, and can
remain active between concurrent transfers.
This pin is shared with the external memory interface and is only used by the HPI
when the interface is in HPI mode (XIO pin is low). This pin is placed into the
high-impedance state when OFF is low.
O/Z
HPI data strobes. HDS1 and HDS2 are driven by the host read and write strobes
to control HPI transfers.
These pins are shared with the external memory interface and are only used by
the HPI when the interface is in HPI mode (XIO pin is low).
These pins are placed into the high-impedance state when OFF is low.
O/Z
HPI read/write signal. This signal is used by the host to control the direction of an
HPI transfer.
This pin is shared with the external memory interface and is only used by the HPI
when the interface is in HPI mode (XIO pin is low).
This pin is placed into the high-impedance state when OFF is low.
I
HPI data-ready output. The ready output informs the host when the HPI is ready
for the next transfer.
This pin is shared with the external memory interface and is only used by the HPI
when the interface is in HPI mode (XIO pin is low). HRDY is placed into the
high-impedance state when OFF is low.
† I = Input, O = Output, S = Supply, Z = High Impedance
‡ This pin has an internal pullup resistor.
§ These pins are Schmitt triggered inputs.
¶ This pin has an internal bus holder controlled by way of the BSCR register in 54x cLEAD core of DSP subsystem A .
# This pin is used by Texas Instruments for device testing and should be left unconnected.
|| This pin has an internal pulldown resistor.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
13
ADVANCE INFORMATION
See the PPD signal descriptions. These pins are placed into the high-impedance
state when OFF is low.
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Signal Descriptions (Continued)
NAME
TYPE†
DESCRIPTION
HOST-PORT INTERFACE (HPI) SIGNALS (CONTINUED)
PRIMARY
A_HINT
B_HINT
O/Z
HPIRS§
I
PPA0
PPA1
O/Z
Host interrupt pin. HPI can interrupt the host by asserting this low. The host can
clear this interrupt by writing a “1” to the HINT bit of the HPIC register. Only
supported in HPI multiplexed address/data mode (HMODE pin is low). These pins
are placed into the high-impedance state when OFF is low.
Host-port interface (HPI) reset pin. This signal resets the host port interface and both subsystems.
ADVANCE INFORMATION
Host mode select. When this pin is low, it selects the HPI multiplexed address/data mode. The
multiplexed address/data mode allows hosts with multiplexed address/data lines access to the HPI
registers HPIC, HPIA, and HPID. Host-to-DSP and DSP-to-host interrupts are supported in this mode.
HMODE
I
AVDD
CVDD
S
Dedicated power supply that powers the PLL. AVDD = 1.8 V. AVDD can be connected to CVDD.
S
Dedicated “clean” power supply that powers the core CPUs. CVDD = 1.8 V
DVDD
S
Dedicated “dirty” power supply that powers the I/O pins. DVDD = 3.3 V
VSS
S
Digital ground. Dedicated ground plane for the device.
VSSA
S
Analog ground. Dedicated ground for the PLL. VSSA can be connected to VSS if digital and analog
grounds are not separated.
When HMODE is high, it selects the HPI nonmultiplexed mode. HPI nonmultiplexed mode allows hosts
with separate address/data buses to access the HPI address range by way of the 18-bit address bus
and the HPI data (HPID) register via the 16-bit data bus. Host-to-DSP and DSP-to-host interrupts are
not supported in this mode.
SUPPLY PINS
TEST PIN
TEST#
No connection
EMULATION/TEST PINS
TCK‡§
I
Standard test clock. This is normally a free-running clock signal with a 50% duty cycle. Changes on the
test access port (TAP) of input signals TMS and TDI are clocked into the TAP controller, instruction
register, or selected test-data register on the rising edge of TCK. Changes at the TAP output signal (TDO)
occur on the falling edge of TCK.
TDI‡
I
Test data input. Pin with an internal pullup device. TDI is clocked into the selected register (instruction
or data) on a rising edge of TCK.
TDO
O/Z
Test data pin. The contents of the selected register is shifted out of TDO on the falling edge of TCK.
TDO is in high-impedance state except when the scanning of data is in progress. These pins are
placed into high-impedance state when OFF is low.
TMS‡
I
Test mode select. Pin with internal pullup device. This serial control input is clocked into the TAP
controller on the rising edge of TCK.
TRST||
I
Test reset. When high, TRST gives the scan system control of the operations of the device. If TRST is
driven low, the device operates in its functional mode and the IEEE 1149.1 signals are ignored. Pin with
internal pulldown device.
† I = Input, O = Output, S = Supply, Z = High Impedance
‡ This pin has an internal pullup resistor.
§ These pins are Schmitt triggered inputs.
¶ This pin has an internal bus holder controlled by way of the BSCR register in 54x cLEAD core of DSP subsystem A .
# This pin is used by Texas Instruments for device testing and should be left unconnected.
|| This pin has an internal pulldown resistor.
14
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
Signal Descriptions (Continued)
NAME
TYPE†
DESCRIPTION
EMULATION/TEST PINS (CONTINUED)
EMU0
I/O/Z
Emulator interrupt 0 pin. When TRST is driven low, EMU0 must be high for the activation of the
EMU1/OFF condition. When TRST is driven high, EMU0 is used as an interrupt to or from the emulator
system and is defined as I/O.
Emulator interrupt 1 pin. When TRST is driven high, EMU1/OFF is used as an interrupt to or from the
emulator system and is defined as I/O. When TRST transitions from high to low, then EMU1 operates
as OFF. EMU/OFF = 0 puts all output drivers into the high-impedance state.
EMU1/OFF
I/O/Z
Note that OFF is used exclusively for testing and emulation purposes (and not for multiprocessing
applications). Therefore, for the OFF condition, the following conditions apply:
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
ADVANCE INFORMATION
TRST = 0, EMU0 = 1, EMU1 = 0
† I = Input, O = Output, S = Supply, Z = High Impedance
‡ This pin has an internal pullup resistor.
§ These pins are Schmitt triggered inputs.
¶ This pin has an internal bus holder controlled by way of the BSCR register in 54x cLEAD core of DSP subsystem A .
# This pin is used by Texas Instruments for device testing and should be left unconnected.
|| This pin has an internal pulldown resistor.
15
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
functional overview
Pbus
Dbus
Ebus
32K RAM
Dual Access
Program/Data
32K RAM
Single Access
Data
54X cLEAD
(Core A)
Cbus
Pbus
Ebus
Dbus
Cbus
Ebus
Cbus
Dbus
Pbus
P, C, D, E Buses and Control Signals
2K Program
ROM
MBus
GPIO
RHEA
Bridge
TI BUS
RHEA Bus
McBSP1
ADVANCE INFORMATION
DSP Subsystem A
16 HPI
xDMA
logic
McBSP2
MBus
P bus
RHEA bus
Arbitrator
McBSP3
RHEAbus
TIMER
APLL
Arbitrator
16HPI
128K
Dual
Access
PRAM
Core-to-Core
FIFO Interface
MBus
P
Interprocessor
IRQs
Cycle
Arrangmnt
XIO
JTAG
Clocks
MBus
32K RAM
Dual Access
Program/Data
Pbus
Ebus
Dbus
MBus
32K RAM
Single Access
Data
54X cLEAD
(Core B)
Cbus
Pbus
Ebus
Dbus
Cbus
Ebus
Dbus
Cbus
Pbus
P, C, D, E Buses and Control Signals
2K
Program ROM
MBus
GPIO
TI Bus
RHEA
Bridge
RHEA Bus
McBSP1
16 HPI
Host Access Bus
McBSP2
MBus
RHEA bus
Arbitrator
xDMA
Logic
McBSP3
MBus
TIMER
JTAG
DSP Subsystem B
Figure 1. Functional Block Diagram
16
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
memory
Each ’5421 DSP subsystem maintains the peripheral register memory map and interrupt location/priorities of
the standard ’5420. Figure 2 shows the size of the required memory blocks and their link map within the program
and data space of the cLEAD core. The total on-chip memory for the ’5421 devices is 256K-word data/program.
00 005F
00 0060
Data
MemoryMapped
Registers
Hex
Program Page 2
Hex
Program Page 3
02 005F
02 0060
Reserved
03 005F
03 0060
On-Chip
DARAM A/B§
(32K Words)
Prog/Data
(OVLY=1)
On-Chip
DARAM A/B§
(32K Words)
Prog/Data
(OVLY=1)
On-Chip
DARAM A/B§
(32K Words)
Prog/Data
(OVLY=1)
External
(OVLY=0)
External
(OVLY=0)
External
(OVLY=0)
External
(OVLY=0)
On-Chip
Shared
DARAM 0
(24K Words)
Prog Only
02 7FFF
02 8000
Reserved
External‡
0n 7FFF
0n 8000
03 7FFF
03 8000
On-Chip
Shared
DARAM 3
(32K Words)
Prog Only
On-Chip
Shared
DARAM 2
(32K Words)
Prog Only
On-Chip
Shared
DARAM 1
(32K Words)
Prog Only
Program Page n
0n 005F
0n 0060
On-Chip
DARAM A/B§
(32K Words)
Prog/Data
(OVLY=1)
01 7FFF
01 8000
Hex
0n 0000
03 0000
Reserved
01 005F
01 0060
00 7FFF
00 8000
External
(DROM=0)
Hex
02 0000
Reserved
Reserved
On-Chip
SARAM A/B
(32K Words)
Data Only
(DROM=1)
Program Page 1
01 0000
00 005F
00 0060
On-Chip
DARAM A/B§
(32K Words)
Prog/Data
00 7FFF
00 8000
Hex
Program Page 0
00 0000
External‡
Shared 0
00 DFFF
00 E000
Reserved
00 F7FF
00 F800
00 FFFF
00 FFFF
ROM
(ROMEN=1)†
Shared 2
Shared 1
01 FFFF
(extended)
Shared 3
0n FFFF
03 FFFF
02 FFFF
(extended)
(extended)
(n = 4 – 127)
† ROM enabled after reset.
‡ When cLEAD PMST register bit MP/MC=0 and an address is generated outside the on-chip memory bound or the address reach, i.e.,
XPC > 3h, access is always external, if XIO = 1. Pages 8–127 are mapped over pages 4–7. When XIO = 1 and MP/MC = 1, program pages 0,
1, 2, and 3 are external. Pages 4–127 are mapped over pages 0–3.
§ On-chip DARAM A and SARAM A are for core A. Likewise, on-chip DARAM B and SARAM B are for core B.
NOTES: A. Clearing the ROMEN bit (GPIO[7]) enables an 8K-word block (0E000h – 0FFFFh) of DARAM .
B. All external accesses require the XIO pin to be high.
C. CPU I/O space is a single page of 64K words. Access is always external.
D. All internal memory is divided into 8K blocks.
Figure 2. Memory Map Relative to CPU Subsystems A and B
on-chip dual-access RAM (DARAM)
The ’5421 subsystems A and B each have 32K 16-bit words of on-chip DARAM (4 blocks of 8K words). Each
of these DARAM blocks can be accessed twice per machine cycle. This memory is intended primarily to store
data values; however, it can be used to store program as well. At reset, the DARAM is mapped into data memory
space. The DARAM can be mapped into program/data memory space by setting the OVLY bit in the
processor-mode status (PMST) register of the 54X cLEAD CPU in each DSP subsystem.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
17
ADVANCE INFORMATION
Hex
00 0000
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
on-chip single-access RAM (SARAM)
The ’5421 subsystems A and B each have 32K 16-bit words of on-chip SARAM (4 blocks of 8K words). Each
of these SARAM blocks can be accessed once per machine cycle. This memory is intended to store data values
only. At reset, the SARAM is disabled. The SARAM can be enabled in data memory space by setting the DROM
bit in the PMST register.
on-chip shared RAM (DARAM)
The ’5421 has 128K 16-bit words of on-chip DARAM (16 blocks of 8K words) that is shared between the two
DSP subsystems. This memory is intended to store program only. Each subsystem is able to make one
instruction fetch from any location in shared memory each cycle. Neither subsystem CPU can write to the shared
memory as only the DMA can write to shared memory.
ADVANCE INFORMATION
on-chip boot ROM
The ’5421 subsystems A and B each have 2K 16-bit words of on-chip ROM. This ROM is used for bootloading
functions only. Enabling the ROM maps out one 8K-word block of the shared program memory. The ROM can
be disabled by clearing bit 7 (ROMEN) of the general-purpose I/O (GPIO) register. Table 1 shows the
XIO/ROMEN modes. The ROM is enabled or disabled at reset for each subsystem depending on the state of
the GPIO0 pin for that subsystem.
Table 1. XIO/ROMEN Modes
XIO
ROMEN/GPIO0
0
x
Fetch internal from RAM
MODE
1
0
Fetch external
1
1
ROM enabled
extended program memory
The program memory space on the ’5421 device addresses up to 512K 16-bit words. The ’5421 device uses
a paged extended memory scheme in program space to allow access of up to 512K of program memory . This
extended program memory (each subsystem) is organized into eight pages (0–7), pages 0–3 are internal,
pages 4–7 are external, each 64K in length. (Pages 8–127 as defined by the program counter extension register
(XPC) are aliases for pages 4–7.) Access to the extended program memory is similar to the ’5420. To implement
the extended program memory scheme, the ’5421 device includes the following feature:
Two ’54x instructions are extended to use the additional two bits in the ’5421 device.
–
–
READA – Read program memory addressed by accumulator A and store in data memory
WRITA – Write data to program memory addressed by accumulator A
(Writes not allowed for CPUs to shared program memory)
program memory
The program memory is accessible on multiple pages, depending on the XPC value. Within these pages,
memory is accessible, depending on the address range.
Access in the lower 32K of each page is dependent on the state of OVLY.
–
–
OVLY = 0 – Program memory is accessed externally for all values of XPC.
OVLY = 1 – Program memory is accessed from local data/program DARAM for all values of XPC.
Access in the upper 32K of each page is dependent on the state of MP/MC and the value of XPC.
–
–
18
MP/MC = 0 – Program memory is accessed internally from shared DARAM for XPC = 0–3. Program
memory is accessed externally for XPC = 4–127.
MP/MC = 1 – Program memory is accessed externally for all values of XPC.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
data memory
The data memory space is a single page of 64K. Access is dependent on the address range. Access in the lower
32K of data memory is always from local DARAM.
Access in the upper 32K of data memory is dependent on the state of DROM.
DROM = 0 – Data memory is accessed externally
DROM = 1 – Data memory is accessed internally from local SARAM
I/O memory
The I/O space is a single page of 64K. Access is always external.
When XIO = 0 and an access to external memory is attempted, any write is ignored and any read is an unknown
value.
The ’5421 device includes three reset signals: A_RS, B_RS, and HPIRS. The A_RS and B_RS pins function
as the CPU reset signal for subsystem A and subsystem B, respectively. These signals reset the state of the
CPU registers and upon release, initiate the reset function. Additionally, the A_RS signal resets the on-chip PLL
and initializes the CLKMD register to bypass mode.
The HPI reset signal (HPIRS) places the HPI peripheral into a reset state. It is necessary to wait three clock
cycles after the rising edge of HPIRS before performing an HPI access. The HPIRS signal also resets the PLL
by turning off the PLL and initializing the CLKMD register to bypass mode.
bootloader
The on-chip bootloader is used to automatically transfer user code from an external source to anywhere in
program memory after reset. The XIO pin is sampled during a hardware reset and the results indicate the
operating mode as shown in Table 2.
Table 2. Bootloader Operating Modes
XIO
AFTER RESET
HPI mode, bootload is controlled by host. The external host holds the ’5421 in reset while it loads the on-chip
memory of one or both subsystems as determined by the SELA/B pin.
The host can release the ’5421 from reset by either of the following methods:
1.
If the RS_A/B pins are held low while HPIRS transitions from low to high, the subsystem cores reset will be
controlled by the RS_A/B pins. When the host has finished downloading code, it drives RS_A/B high to release
the cores from reset.
2.
If the RS_A/B pins are held high while HPIRS transitions from low to high, the subsystems stay in reset until
a HPI data write to address 0x2F occurs. This means the host can download code to subsystem A and then
release core A from reset by writing any data to core A address 0x2F via the HPI. The host can then repeat
the sequence for core B. This mode allows the host to control the ’5421 reset without additional hardware.
0
1
XIO mode. ROM is mapped in, if ROMEN pin = 1 during reset.
The ’5421 bootloader provides the following options for the source of code to download:
Parallel from 8-bit or 16-bit-wide EPROM
Serial boot from McBSPs, 8-bit mode
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
19
ADVANCE INFORMATION
multicore reset signals
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
bootloader (continued)
GPIO register bit 7 (ROMEN) is used to enable/disable the ROM after reset. The ROMEN bit reflects the status
of the ROMEN/GPIO0 pin for each core. ROMEN = 1 indicates that the ROM and the 8K-word program memory
block (00 E000h–00 FFFFh) are not available for the CPU. When ROMEN = 0, this 8K-word program memory
is available and the ROM are disabled.
A combination of interrupt flags and the bit values of an external memory location determine the selection of
the various boot options.
external interface (XIO)
ADVANCE INFORMATION
The external interface (XIO) supports the ’5421 master boot modes and other external accesses. Its features
include:
Multiplexed with the HPI pins
Selection of XIO or HPI mode is determined by a dedicated pin (XIO)
Provides 512K words of external program space, 64K words of external data space, and 64K words of
external I/O space.
Different boot modes are selectable by the XIO, HMODE, and RS_A/B pins.
After reset, the control register bit ROMEN is always preset to 1.
While XIO = 0 during reset, host HPI mode is on, the host sees all RAM, and ROM is disabled. A host write to
002Fh releases the CPUs from reset; the 002Fh write by the host clears the ROMEN bit in the GPIO register.
While XIO = 1 and ROMEN = 1 during reset, the CPU starts from ROM (0FF80h) to do boot selection. After
branching to non-ROM area, the code changes the ROMEN bit to enable the RAM area occupied by ROM. While
XIO = 1 and ROMEN = 0 during reset, the CPU starts from external (0FF80h) to do boot selection.
Table 3 provides a complete description of HMODE, SELA/B, and XIO pin functionality.
Table 3. XIO/HPI Modes
HMODE
SELA/B
0
0
HPI muxed address/data subsystem A slave to host
HPI MODES (XIO = 0)
SELA/B pin is multiplexed as PPA18 output.
XIO MODES (XIO = 1)
0
1
HPI muxed address/data subsystem B slave to host
SELA/B pin is multiplexed as PPA18 output.
1
0
HPI non-muxed address/data subsystem A slave to host
SELA/B pin is multiplexed as PPA18 output.
1
1
HPI non-muxed address/data subsystem B slave to host
SELA/B pin is multiplexed as PPA18 output.
on-chip peripherals
All the ’54x devices have the same CPU structure; however, they have different on-chip peripherals connected
to their CPUs. The on-chip peripheral options provided are:
Software-programmable wait-state generator
Programmable bank-switching
Parallel I/O ports
Multichannel buffered serial ports (McBSPs)
A hardware timer
A software-programmable clock generator using a phase-locked loop (PLL)
20
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
software-programmable wait-state generators
The software-programmable wait-state generator can be used to extend external bus cycles up to fourteen
machine cycles to interface with slower off-chip memory and I/O devices. The software wait-state register
(SWWSR) controls the operation of the wait-state generator. The SWWSR of a particular DSP subsystem
(A or B) is used for the external memory interface, depending on the state of the xDMA/XIO arbitration logic (see
Direct Memory Access (DMA) Controller section and Table 13. The 14 least significant bits (LSBs) of the
SWWSR specify the number of wait states (0–7) to be inserted for external memory accesses to five separate
address ranges. This allows a different number of wait states for each of the five address ranges.
Additionally, the software wait-state multiplier (SWSM) bit of the software wait-state control register (SWCR)
defines a multiplication factor of 1 or 2 for the number of wait states. At reset, the wait-state generator is initialized
to provide seven wait states on all external memory accesses. The SWWSR bit fields are shown in Figure 3
and described in Table 4.
14
13
12
11
10
9
8
7
6
5
4
3
2
1
XPA
I/O
Data
Data
Program
Program
R/W=0
R/W=111
R/W=111
R/W=111
R/W=111
R/W=111
0
LEGEND: R = Read, W = Write
Figure 3. Software Wait-State Register (SWWSR) [Memory-Mapped Register (MMR) Address 0028h]
Table 4. Software Wait-State Register (SWWSR) Bit Fields
BIT
NO.
NAME
RESET
VALUE
15
XPA
0
Extended program address control bit. XPA is used in conjunction with the program space fields
(bits 0 through 5) to select the address range for program space wait states.
14–12
I/O
1
I/O space. The field value (0–7) corresponds to the base number of wait states for I/O space accesses
within addresses 0000–FFFFh. The SWSM bit of the SWCR defines a multiplication factor of 1 or 2 for
the base number of wait states.
11–9
Data
1
Upper data space. The field value (0–7) corresponds to the base number of wait states for external data
space accesses within addresses 8000–FFFFh. The SWSM bit of the SWCR defines a multiplication
factor of 1 or 2 for the base number of wait states.
8–6
Data
1
Lower data space. The field value (0–7) corresponds to the base number of wait states for external data
space accesses within addresses 0000–7FFFh. The SWSM bit of the SWCR defines a multiplication
factor of 1 or 2 for the base number of wait states.
FUNCTION
Upper program space. The field value (0–7) corresponds to the base number of wait states for external
program space accesses within the following addresses:
5–3
Program
1
XPA = 0: x8000–xFFFFh
XPA = 1: The upper program space bit field has no effect on wait states.
The SWSM bit of the SWCR defines a multiplication factor of 1 or 2 for the base number of wait states.
Program space. The field value (0–7) corresponds to the base number of wait states for external program
space accesses within the following addresses:
2–0
Program
1
XPA = 0: x0000–x7FFFh
XPA = 1: 00000–3FFFFh
The SWSM bit of the SWCR defines a multiplication factor of 1 or 2 for the base number of wait states.
The software wait-state multiplier bit of the software wait-state control register (SWCR) is used to extend the
base number of wait states selected by the SWWSR. The SWCR bit fields are shown in Figure 4 and described
in Table 5.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
21
ADVANCE INFORMATION
15
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
software-programmable wait-state generators (continued)
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Reserved
SWSM
R/W=0
R/W=0
LEGEND: R = Read, W = Write
Figure 4. Software Wait-State Control Register (SWCR) [MMR Address 002Bh]
Table 5. Software Wait-State Control Register (SWCR) Bit Fields
ADVANCE INFORMATION
PIN
NO.
NAME
RESET
VALUE
15–1
Reserved
0
FUNCTION
These bits are reserved and are unaffected by writes.
Software wait-state multiplier. Used to multiply the number of wait states defined in the SWWSR by a factor
of 1 or 2.
0
SWSM
0
SWSM = 0: wait-state base values are unchanged (multiplied by 1).
SWSM = 1: wait-state base values are multiplied by 2 for a maximum of 14 wait states.
programmable bank-switching
Programmable bank-switching can be used to insert one cycle automatically when crossing memory-bank
boundaries inside program memory or data memory space. One cycle can also be inserted when crossing from
program-memory space to data-memory space (’54x) or one program memory page to another program
memory page. This extra cycle allows memory devices to release the bus before other devices start driving the
bus, thereby avoiding bus contention. The size of the memory bank for the bank-switching is defined by the
bank-switching control register (BSCR), as shown in Figure 5. The BSCR of a particular DSP subsystem
(A or B) is used for the external memory interface based on the xDMA/XIO arbitration logic.
15
14
13
12
11
10
9
8
7
BNKCMP
PS-DS
Reserved
IPIRQ
R/W
R/W
R/W
R/W
6
5
Reserved
4
3
2
1
0
BH
Reserved
EXIO
R/W
LEGEND: R = Read, W = Write
Figure 5. BSCR Register Bit Layout for Each DSP Subsystem
22
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
R/W
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
programmable bank-switching (continued)
BIT
NO.
BIT
NAME
RESET
VALUE
FUNCTION
15–12
BNKCMP
1111
Bank compare. BNKCMP determines the external memory-bank size. BNKCMP is used to mask the four
most significant bits (MSBs) of an address. For example, if BNKCMP = 1111b, the four MSBs (bits 12–15)
are compared, resulting in a bank size of 4K words. Bank sizes of 4K words to 64K words are allowed.
11
PS-DS
1
Program read – data read access. PS-DS inserts an extra cycle between consecutive accesses of
program read and data read or data read and program read.
PS-DS = 0
No extra cycles are inserted by this feature.
PS-DS = 1
One extra cycle is inserted between consecutive data and program reads.
10–9
Reserved
0
These bits are reserved and are unaffected by writes.
8
IPIRQ
0
The IPIRQ bit is used to send an interprocessor interrupt to the other subsystem. IPIRQ=1 sends the
interrupt. IPIRQ must be cleared before subsequent interrupts can be made. Refer to the interrupts section
for more details.
7–3
Reserved
0
These bits are reserved and are unaffected by writes.
2
BH
0
Bus holder. BH controls the data bus holder feature: BH is cleared to 0 at reset.
BH = 0
The bus holder is disabled.
BH = 1
The bus holder is enabled. When not driven, the data bus (PPD[15:0]) is held in the
previous logic level.
1
Reserved
0
This bit is reserved and is unaffected by writes.
0
External bus interface off. The EXIO bit controls the external bus-off function.
EXIO = 0
The external bus interface functions as usual.
EXIO = 1
The address bus, data bus, and control signals become inactive after completing the
current bus cycle. Note that the DROM, MP/MC, and OVLY bits in the PMST and the
HM bit of ST1 cannot be modified when the interface is disabled.
0
EXIO
parallel I/O ports
The ’5421 has a total of 64K words of I/O port address space. These ports can be addressed by PORTR and
PORTW. The IS signal indicates the read/write access through an I/O port. The devices can interface easily with
external devices through the I/O ports while requiring minimal off-chip address-decoding logic. The SELA/B pin
selects which subsystem is accessing the external I/O space.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
23
ADVANCE INFORMATION
Table 6. BSCR Register Bit Functions for Each DSP Subsystem
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
16-bit bidirectional host-port interface (HPI16)
HPI16 memory map
The HPI16 is an enhanced 16-bit version of the ’C54x 8-bit host-port interface (HPI). The HPI16 is designed
to allow a 16-bit host to access the DSP on-chip memory, with the host acting as the master of the interface.
Figure 6 illustrates the available memory accessible by the HPI. Neither the CPU nor DMA I/O spaces can be
accessed using the host-port interface.
Hex
Page 0
00 0000
Hex
Page 1
01 0000
Hex
02 0000
Reserved
00 001F
00 0020
ADVANCE INFORMATION
00 005F
00 0060
02 001F
02 0020
McBSP
DXR/DRR
MMRegs Only
Subsystem A
02 005F
02 0060
On-Chip
Shared
DARAM 0
(32K Words)
Program Only
01 7FFF
01 8000
On-Chip
SARAM A
(32K Words)
Data Only
00 FFFF
Subsystem A
Hex
Page 3
03 0000
Reserved
On-Chip
DARAM A
(32K Words)
Prog/Data
00 7FFF
00 8000
Page 2
Shared 0
Shared 1
On-Chip
Shared
DARAM 2
(32K Words)
Program Only
On-Chip
DARAM B
(32K Words)
Prog/Data
02 7FFF
02 8000
Subsystem B
03 7FFF
03 8000
On-Chip
SARAM B
(32K Words)
Data Only
On-Chip
Shared
DARAM 1
(32K Words)
Program Only
01 FFFF
McBSP
DXR/DRR
MMRegs Only
02 FFFF
Subsystem B
Shared 2
On-Chip
Shared
DARAM 3
(32K Words)
Program Only
03 FFFF
Shared 3
NOTES: A. All local memory is available to the HPI
B. The encoder maps CPU A Data Page 0 into the HPI Page 0. CPU B Data Page 0 is mapped into the HPI Page 2. Pages 1 and 3
are the on-chip shared program memory.
C. In pages 00 and 02, in the range of 0020–005F, only the following memory mapped registers are accessible: 20,21,30,31,40,41 (read
only), 22,23,32,33,42,43 (write only).
Figure 6. Memory Map Relative to Host-Port Interface HPI16
24
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 memory map (continued)
Some of the features of the HPI16 include:
The HPI16 acts as a slave to a 16-bit host processor and allows access to the on-chip memory of the DSP. There
are two modes of operation as determined by the HMODE signal: multiplexed mode and nonmultiplexed mode.
HPI multiplexed mode
In multiplexed mode, HPI16 operation is very similar to that of the standard 8-bit HPI, which is available with
other ’C54x products. A host with a multiplexed address/data bus can access the HPI16 data register (HPID),
address register (HPIA), or control register (HPIC) via the HD bidirectional data bus. The host initiates the
access with the strobe signals (HDS1, HDS2, HCS) and controls the type of access with the HCNTL, HR/W,
and HAS signals. The DSP can interrupt the host via the HINT signal, and can stall host accesses via the HRDY
signal.
host/DSP interrupts
In multiplexed mode, the HPI16 offers the capability for the host and DSP to interrupt each other through the
HPIC register.
For host-to-DSP interrupts, the host must write a “1” to the DSPINT bit of the HPIC register. This generates an
interrupt to the DSP. This interrupt can also be used to wake the DSP from any of the IDLE 1,2, or 3 states. Note
that the DSPINT bit is always read as “0” by both the host and DSP. The DSP cannot write to this bit (see
Figure 7).
For DSP-to-host interrupts, the DSP must write a “1” to the HINT bit of the HPIC register to interrupt the host
via the HINT pin. The host acknowledges and clears this interrupt by also writing a “1” to the HINT bit of the HPIC
register. Note that writing a “0” to the HINT bit by either host or DSP has no effect.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
25
ADVANCE INFORMATION
16-bit bidirectional data bus
Multiple data strobes and control signals to allow glueless interfacing to a variety of hosts
Multiplexed and nonmultiplexed address/data modes
18-bit address bus used in nonmultiplexed mode to allow access to all internal memory (including internal
extended address pages)
18-bit address register used in multiplexed mode. Includes address autoincrement feature for faster
accesses to sequential addresses
Interface to on-chip DMA module to allow access to entire internal memory space
HRDY signal to hold off host accesses due to DMA latency
Control register available in multiplexed mode only. Accessible by either host or DSP to provide host/DSP
interrupts, extended addressing, and data prefetch capability
Maximum data rate of 33 megabytes per second (MBps) at 100-MHz DSP clock rate (no other DMA
channels active)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI nonmultiplexed mode
In nonmultiplexed mode, a host with separate address/data buses can access the HPI16 data register (HPID)
via the HD 16-bit bidirectional data bus, and the address register (HPIA) via the 18-bit HA address bus. The host
initiates the access with the strobe signals (HDS1, HDS2, HCS) and controls the direction of the access with
the HR/W signal. The HPI16 can stall host accesses via the HRDY signal. Note that the HPIC register is not
available in nonmultiplexed mode since there are no HCNTL signals available. All host accesses initiate a DMA
read or write access. Figure 7 shows a block diagram of the HPI16 in nonmultiplexed mode.
ÎÎ
ÎÎ
ÎÎ
ÎÎ
ÎÎ
ÎÎ
ÎÎ
ÎÎ
ÎÎ
HPI-16
HOST
Internal
memory
HD[15:0]
Data[15:0]
HPID[15:0]
ADVANCE INFORMATION
Address[n:0]
HA[ n :0]
R/W
Data strobes
Ready
HRDY
HR/W
HDS1, HDS2, HCS
DMA
’54x
CPU
Figure 7. Interfacing to the HPI-16 in Non-Multiplexed Mode
other HPI16 system considerations
operation during IDLE
The HPI16 can continue to operate during IDLE1 or IDLE2 by using special clock management logic that turns
on relevant clocks to perform a synchronous memory access, and then turns the clocks back off to save power.
The DSP CPU does not wake up from the IDLE mode during this process.
downloading code during reset
The HPI16 can download code while the DSP is in reset. However, the system provides a pin (HPIRS) that
provides a way to take the HPI16 module out of reset while leaving the DSP in reset. The maximum HPI16 data
rate is 33 MBps assuming no other DMA activity (100-MIPS DSP subsystem).
emulation considerations
The HPI16 can continue operation even when the DSP CPU is halted due to debugger breakpoints or other
emulation events.
26
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port (McBSP)
The ’5421 device provides high-speed, full-duplex serial ports that allow direct interface to other ’C54x/’LC54x
devices, codecs, and other devices in a system. There are six multichannel buffered serial ports (McBSPs) on
board (three per subsystem).
The McBSP provides:
Full-duplex communication
Double-buffer data registers, which allow a continuous data stream
Independent framing and clocking for receive and transmit
In addition, the McBSP has the following capabilities:
Direct interface to:
T1/E1 framers
MVIP switching-compatible and ST-BUS compliant devices
IOM-2 compliant device
AC97-compliant device
Serial port interface (SPI)
Multichannel transmit and receive of up to 128 channels
A wide selection of data sizes, including: 8, 12, 16, 20, 24, or 32 bits
µ-law and A-law companding
Programmable polarity for both frame synchronization and data clocks
Programmable internal clock and frame generation
The ’5421 McBSPs have been enhanced to provide more flexibility in the choice of the sample rate generator
input clock source. On previous C5000 devices, the McBSP sample rate input clock can be driven from one of
two possible choices: the internal CPU clock , or the external CLKS pin. However, most C5000 devices have
only the internal CPU clock as a possible source because the CLKS pin is not implemented on most device
packages.
To accomodate applications that require an external reference clock for the sample rate generator, the ’5421
McBSPs allow either the receive clock pin (BCLKR) or the transmit clock pin (BCLKX) to be configured as the
input clock to the sample rate generator. This enhancement is enabled through two register bits: pin control
register (PCR) bit 7 – enhanced sample clock mode (SCLKME), and sample rate generator register 2 (SRGR2)
bit 13 – McBSP sample rate generator clock mode (CLKSM). SCLKME is an addition to the PCR contained in
the McBSPs on previous C5000 devices. The new bit layout of the PCR is shown in Figure 8. For a description
of the remaining bits, see TMS320C54x DSP Reference Set, Volume 5: Enhanced Peripherals (literature
number SPRU302).
15
14
13
12
11
10
9
8
Reserved
XIOEN
RIOEN
FSXM
FSRM
CLKXM
CLKRM
R,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
7
6
5
4
3
2
1
0
SCLKME
CLKS_STAT
DX_STAT
DR_STAT
FSXP
FSRP
CLKXP
CLKRP
RW,+0
R,+0
R,+0
R,+0
RW,+0
RW,+0
RW,+0
RW,+0
Note:
R = Read, W = Write, +0 = Value at reset
Figure 8. Pin Control Register (PCR)
SPI is a trademark of Motorola Incorporated.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
27
ADVANCE INFORMATION
–
–
–
–
–
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port (McBSP) (continued)
The selection of the sample rate generator (SRG) clock input source is made by the combination of the CLKSM
and SCLKME bit values as shown in Table 7.
ADVANCE INFORMATION
Table 7. Sample Rate Generator Clock Source Selection
SRG Clock Source
SCLKME
CLKSM
0
0
CLKS (not available as a pin on ’5421)
0
1
CPU clock
1
0
BCLKR pin
1
1
BCLKX pin
When either of the bidirectional pins, BCLKR or BCLKX, is configured as the clock input, its output buffer is
automatically disabled. For example, with SCLKME = 1 and CLKSM = 0, the BCLKR pin is configured as the
SRG input. In this case, both the transmitter and receiver circuits can be synchronized to the SRG output by
setting the PCR bits (9:8) for CLKXM = 1 and CLKRM = 1. However, the SRG output is only driven onto the
BCLKX pin because the BCLKR output is automatically disabled.
The McBSP supports independent selection of multiple channels for the transmitter and receiver. When multiple
channels are selected, each frame represents a time-division multiplexed (TDM) data stream. In using
time-division multiplexed data streams, the CPU may only need to process a few of them. Thus, to save memory
and bus bandwidth, multichannel selection allows independent enabling of particular channels for transmission
and reception. Up to a maximum of 128 channels in a bit stream can be enabled or disabled.
The ’5421 McBSPs have two working modes that are selected by setting the RMCME and XMCME bits in the
multichannel control registers (MCR1x and MCR2x, respectively). See Figure 9 and Figure 10. For a
description of the remaining bits, see TMS320C54x DSP Reference Set, Volume 5: Enhanced Peripherals
(literature number SPRU302).
In the first mode, when RMCME = 0 and XMCME = 0, there are two partitions (A and B), with each containing
16 channels as shown in Figure 9 and Figure 10. This is compatible with the McBSPs used in the ’5420,
where only 32-channel selection is enabled (default).
15
14
Note:
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Reserved
XMC
ME
XPBBLK
XPABLK
XCBLK
XMCM
R,+0
RW,+0
RW,+0
RW,+0
R,+0
RW,+0
R = Read, W = Write, +0 = Value at reset; x = McBSP 0,1, or 2
Figure 9. Multichannel Control Register 2x (MCR2x)
15
14
Note:
13
12
11
10
9
8
7
6
5
4
3
1
0
Reserved
RMC
ME
RPBBLK
RPABLK
RCBLK
RMCM
R,+0
RW,+0
RW,+0
RW,+0
R,+0
RW,+0
R = Read, W = Write, +0 = Value at reset; x = McBSP 0,1, or 2
Figure 10. Multichannel Control Register 1x (MCR1x)
28
2
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port (McBSP) (continued)
15
14
13
12
11
10
9
8
RCERyz15
RCERyz14
RCERyz13
RCERyz12
RCERyz11
RCERyz10
RCERyz9
RCERyz8
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
7
6
5
4
3
2
1
0
RCERyz7
RCERyz6
RCERyz5
RCERy4
RCERyz3
RCERyz2
RCERyz1
RCERyz0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
Note:
ADVANCE INFORMATION
In the second mode, with RMCME = 1 and XMCME = 1, the McBSPs have 128 channel selection capability.
Twelve new registers (RCERCx–RCERHx and XCERCx–XCERHx) are used to enable the 128 channel
selection. The subadresses of the new registers are shown in Table 20. These new registers, functionally
equivalent to the RCERA0–RCERB1 and XCERA0–XCERB1 registers in the ’5420, are used to
enable/disable the transmit and receive of additional channel partitions (C,D,E,F,G, and H) in the
128 channel stream. For example, XCERH1 is the transmit enable for channel partition H (channels 112
to 127) of MCBSP1 for each DSP subsystem. See Figure 11, Table 8, Figure 12, and Table 9 for bit layout
and function of the receive and transmit registers .
R = Read, W = Write, +0 = Value at reset; y = Partition A,B,C,D,E,F,G, or H; z = McBSP 0,1, or 2
Figure 11. Receive Channel Enable Registers Bit Layout for Partitions A to H
Table 8. Receive Channel Enable Registers for Partitions A to H
Bit
15–0
Note:
Name
Function
RCERyz(15:0)
Receive Channel Enable Register
RCERyz n = 0
Disables reception of nth channel in partition y.
RCERyz n = 1
Enables reception of nth channel in partition y.
y = Partition A,B,C,D,E,F,G, or H; z = McBSP 0,1, or 2; n = bit 15–0
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
29
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port (McBSP) (continued)
15
14
13
12
11
10
9
8
XCERyz15
XCERyz14
XCERyz13
XCERyz12
XCERyz11
XCERyz10
XCERyz9
XCERyz8
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
7
6
5
4
3
2
1
0
XCERyz7
XCERyz6
XCERyz5
XCERy4
XCERyz3
XCERyz2
XCERyz1
XCERyz0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
RW,+0
Note:
R = Read, W = Write, +0 = Value at reset; y = Partition A,B,C,D,E,F,G, or H; z = McBSP 0,1, or 2
Figure 12. Transmit Channel Enable Registers Bit Layout for Partitions A to H
ADVANCE INFORMATION
Table 9. Transmit Channel Enable Registers for Partitions A to H
Bit
15–0
Note:
Name
Function
XCERyz(15:0)
Transmit Channel Enable Register
XCERyz n = 0
Disables transmit of nth channel in partition y.
XCERyz n = 1
Enables transmit of nth channel in partition y.
y = Partition A,B,C,D,E,F,G, or H; z = McBSP 0,1, or 2; n = bit 15–0
The clock stop mode (CLKSTP) in the McBSP provides compatibility with the serial port interface (SPI) protocol.
Clock stop mode works with only single-phase frames and one word per frame. The word sizes supported by
the McBSP are programmable for 8-, 12-, 16-, 20-, 24-, or 32-bit operation. When the McBSP is configured to
operate in SPI mode, both the transmitter and the receiver operate together as a master or as a slave.
The McBSP is fully static and operates at arbitrarily low clock frequencies. The maximum McBSP multichannel
operating frequency on the ’5421 is 9 MBps. Nonmultichannel operation is limited to 38 MBps.
direct memory access (DMA) controller
The ’5421 includes two 6-channel direct memory access (DMA) controllers for performing data transfers
independent of the CPU, one for each subsystem. The DMA controller controls accesses to off-chip
program/data/IO and internal data/program memory. The primary function of the ’5421 DMA controller is to
provide code overlays and manage data transfers between on-chip memory, the peripherals, and off-chip
memory.
In the background of CPU operation,the ’5421 DMA allows movement of data between internal and external
program/data memory, and internal peripherals, such as the McBSPs and the HPI. Each subsystem has its own
independent DMA with six programmable channels, which allows for six different contexts for DMA operation.
The HPI has a dedicated auxiliary DMA channel. Figure 13 illustrates the memory map accessible by the DMA.
30
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
direct memory access (DMA) controller (continued)
Data†
Reserved
00 001F
00 0020
00 005F
00 0060
McBSP
DXR/DRR
MMRegs Only
Program Page 0‡
Hex
00 0000
Reserved
00 001F
00 0020
00 005F
00 0060
00 7FFF
00 8000
McBSP
DXR/DRR
MMRegs Only
On-Chip
Shared
DARAM 0
(32K Words)
Program
Only
Subsystem A
01 7FFF
01 8000
On-Chip
SARAM A
(32K Words)
Data Only
On-Chip
SARAM A/B
(32K Words)
Data Only
00 FFFF
Program Page 1‡ Hex
On-Chip
DARAM A
(32K Words)
Prog/Data
On-Chip
DARAM A/B
(32K Words)
Prog/Data
00 7FFF
00 8000
Hex
01 0000
00 FFFF
Subsystem A
Shared 0
Program Page 2‡ Hex Program Page 3‡
02 0000
03 0000
Reserved
02 001F
02 0020
McBSP
DXR/DRR
02 005F MMRegs Only
On-Chip
02 0060
Shared
DARAM 2
On-Chip
(32K Words)
DARAM B
Program
(32K Words)
Only
Prog/Data
02 7FFF Subsystem B
02 8000
On-Chip
SARAM B
(32K Words)
Data Only
On-Chip
Shared
DARAM 1
(32K Words)
Program
Only
01 FFFF
Shared 1
03 7FFF
03 8000
02 FFFF
Subsystem B
Shared 2
On-Chip
Shared
DARAM 3
(32K Words)
Program
Only
03 FFFF
Shared 3
† DMD/DMS = 01
‡ DMD/DMS = 00
NOTES: A. All local memory is available to the DMA.
B. All I/O memory accesses by the DMA (DMD/DMS = 10) are mapped to the core-to-core FIFO.
C. In pages 00 and 02, in the range of 0020–005F, only the following memory mapped registers are accessible: 20,21,30,31,40,41
(read only), 22,23,32,33,42,43 (write only).
Figure 13. On-Chip Memory Map Relative to DMA (DLAXS/SLAXS = 0)
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
31
ADVANCE INFORMATION
Hex
00 0000
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
direct memory access (DMA) controller (continued)
xDMA External Program Memory Map†
00 0000
02 0000
01 0000
Page 0
Lower
32K†
External
...
03 0000
07 0000
Page 1
Lower
32K†
Page 2
Lower
32K†
Page 3
Lower
32K†
Page7
Lower
32K†
External
External
External
External
07 7FFF
00 7FFF
01 7FFF
02 7FFF
03 7FFF
00 8000
01 8000
02 8000
03 8000
Page 1
Upper
32K
External
Page 0
Upper
32K
External
Page 2
Upper
32K
External
...
...
07 8000
Page 3
Upper
32K
External
Page 7
Upper
32K
External
ADVANCE INFORMATION
...
00 FFFF
02 FFFF
01 FFFF
03 FFFF
07 FFFF
xDMA External Data Memory Map†
00 0000
02 0000
01 0000
03 0000
07 0000
..
Page 1
Lower
32K†
External
Page 0
Lower
32K†
External
Page 2
Lower
32K†
External
Page 3
Lower
32K†
External
Page 7
Lower
32K†
External
00 7FFF
01 7FFF
02 7FFF
03 7FFF
...
07 7FFF
00 8000
01 8000
02 8000
03 8000
...
07 8000
Page 1
Upper
32K
External
Page 0
Upper
32K
External
Page 2
Upper
32K
External
Page 3
Upper
32K
External
Page 7
Upper
32K
External
...
00 FFFF
02 FFFF
01 FFFF
03 FFFF
† Pages 8 – 127 are overlaid over pages 0 – 7.
Figure 14. DMA External Program Memory Map
32
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
07 FFFF
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
DMA controller features
The ’5421 DMA has the following features:
A 16-bit DMA transfer requires four CPU clock cycles to complete — two cycles for reads and two cycles for
writes. This gives a maximum DMA throughput of 50 MBps. Since the DMA controller shares the DMA bus with
the HPI module, the DMA access rate is reduced when the HPI is active.
DMA accesses to external memory
The ’5421 DMAs supports external accesses to extended program, extended data, and extended I/O memory.
These overlay pages are only visible to the DMA controller. A maximum of two channels (one for reads, one
for writes) per DMA can be used for external memory accesses. The DMA external accesses require 9 cycles
(minimum) for external writes and 13 cycles (minimum) for external reads.
The control of the bus is arbitrated between the two CPUs and the two DMAs. While one DMA or CPU is in control
of the external bus, the other three components will be held off (via wait-states) until the current transfer is
complete. The DMA takes precedence over XIO requests. The HOLD/HOLDA feature of the ’5421 affects
external CPU transfers, as well as external DMA transfers. When an external processor asserts the HOLD pin
to gain control of the memory interface, the HOLDA signal is not asserted until all pending DMA transfers are
completed. To prevent a DMA from blocking out the CPUs or HOLD/HOLDA feature from accessing the external
bus, uninterrupted burst transfers are not supported by the DMAs. Subsequently, CPU and DMA arbitration
testing is performed for each external bus cycle, regardless of the bus activity. With the completion of each block,
the highest priority will be swapped.
For arbitration at the DSP subsystem level, the DMA requests (DMA_REQ_A or DMA_REQ_B) from either DMA
will be sent to both CPUs as shown in Figure 15. Regardless of which CPU controls the external pin interface
(XIO), both CPUs must send a grant (GRANT_A, GRANT_B) for control of the bus to be released to the DMAs.
Arbitration between CPUs is done using a request/grant scheme. Prior to accessing XIO of one of the CPUs,
software is responsible for asserting a request for access to the device pins and polling grant status until the
pins are granted to the requestor. If both CPUs request the bus simultaneously, subsystem A is granted priority.
For details on memory-mapped register bits pertaining to CPU XIO arbitration, see the general-purpose I/O
control register bits 6:4 (CORE SEL, XIO GRANT, XIO REQ) in Table 13.
At reset, the default is that subsystem A has access to the device pins. Accesses without a grant will be allowed,
but do not show up on the device pins.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
33
ADVANCE INFORMATION
The DMA operates independently of the CPU.
The DMA has six channels. The DMA can keep track of the contexts of six independent block transfers.
Two DMA channels are available for external accesses: one for reads and one for writes.
The DMA has higher priority than the CPU for internal accesses.
Each channel has independently programmable priorities.
Each channel’s source and destination address registers include configurable indexing modes. The
address can be held constant, postincremented, postdecremented, or adjusted by a programmable value.
For internal accesses, each read or write transfer can be initialized by selected events.
Supports 32-bit transfers for internal accesses only.
Single-word (16-bit) transfers are supported for external accesses.
The DMA does not support transfers from peripherals to external memory.
The DMA does not support transfers from external memory to the peripherals.
The DMA does not support external to external transfers.
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
DMA accesses to external memory (continued)
DMA_REQUEST
DMA_REQUEST
54x cLEAD CPU
(DSP Subsystem B)
54x cLEAD CPU
(DSP Subsystem B)
GRANT_B
GRANT_A
XIO
Grant
XIO
Req
Core
SEL=0
GPIO Control
Register
(DSP Subsystem B)
GPIO Control
Register
(DSP Subsystem A)
XIO
Grant
XIO
Req
Core
SEL=1
VCC
CPU_ARB
EMIF
Controller
DMA_REQ_A
DMA_REQ_B
XDMA
(DSP
Subsystem A)
XDMA
(DSP
Subsystem B)
DMA_ARB
DMA_GRANT_A
DMA_GRANT_B
ADVANCE INFORMATION
GRANT
XDMA
XDMA
XCPU
XCPU
SEL
XHOLDA
XHOLD
XIO
Figure 15. Arbitration Between XIO and xDMA for External Access
The HM bit in the ST1 indicates whether the processor continues internal execution when acknowledging an
active HOLD signal.
HM = 0, the processor continues execution from internal program memory but places its external interface
in the high-impedance state.
When HM = 1, the processor halts internal execution.
To ensure that proper arbitration occurs, the HM bit should be set to 0 in the memory-mapped ST1 registers
for both CPUs.
To allow the DMA access to extended data pages, the SLAXS and DLAXs bits are added to the DMMCRn
register. For a description of the remaining bits, see TMS320C54x DSP Reference Set, Volume 5: Enhanced
Peripherals (literature number SPRU302).
15
14
13
12
11
AUTO
INIT
DINM
IMOD
CT
MOD
SLAXS
10
9
SIND
8
7
6
DMS
5
DLAXS
4
3
DIND
Figure 16. DMA Transfer Mode Control Register (DMMCRn)
34
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
2
1
0
DMD
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
DMA accesses to external memory (continued)
These new bit fields were created to allow the user to define the space-select for the DMA (internal/external).
Also, a new extended destination data page (XDSTDP[6:0], subaddress 029h) and extended source data page
(XSRCDP[6:0], subaddress 028h) have been created.
DLAXS(DMMCRn[5])
Destination
0 = No external access (default internal)
SLAXS(DMMCRn[11])
Source
0 = No external access (default internal)
1 = External access
1 = External access
DMA controller synchronization events
The transfers associated with each DMA channel can be synchronized to one of several events. The DSYN bit
field of the DMA channel x sync select and frame count (DMSFCx) register selects the synchronization event
for a channel. The list of possible events and the DSYN values are shown in Table 10.
Table 10. DMA Synchronization Events
DSYN VALUE
DMA SYNCHRONIZATION EVENT
0000b
No synchronization used
0001b
McBSP0 Receive Event
0010b
McBSP0 Transmit Event
0011b
McBSP2 Receive Event
0100b
McBSP2 Transmit Event
0101b
McBSP1 Receive Event
0110b
McBSP1 Transmit Event
0111b
FIFO Receive Buffer Not Empty Event
1000b
FIFO Transmit Buffer Not Full Event
1001b – 1111b
Reserved
DMA channel interrupt selection
The DMA controller can generate a CPU interrupt for each of the six channels. However, channels 0, 1, 2,
and 3 are multiplexed with other interrupt sources. DMA channels 0 and 1 share an interrupt line with the receive
and transmit portions of McBSP2 (IMR/IFR bits 6 and 7), and DMA channels 2 and 3 share an interrupt line with
the receive and transmit portions of McBSP1 (IMR/IFR bits 10 and 11). When the ’5421 is reset, the interrupts
from these four DMA channels are deselected. The INTSEL bit field in the DMA channel priority and enable
control (DMPREC) register can be used to select these interrupts, as shown in Table 11.
Table 11. DMA Channel Interrupt Selection
INTSEL Value
IMR/IFR[6]
IMR/IFR[7]
IMR/IFR[10]
IMR/IFR[11]
00b (reset)
BRINT2
BXINT2
BRINT1
BXINT1
01b
BRINT2
BXINT2
DMAC2
DMAC3
10b
DMAC0
DMAC1
DMAC2
DMAC3
11b
Reserved
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
35
ADVANCE INFORMATION
For the CPU external access, software can configure the memory cells to reside inside or outside the program
address map. When the cells are mapped into program space, the device automatically accesses them when
their addresses are within bounds. When the program address generation (PAGEN) logic generates an address
outside its bounds, the device automatically generates an external access. All DMA I/O space accesses are
mapped to the core-to-core FIFO.
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
subsystem communications
The ’5421 device provides two options for efficient core-to-core communications:
Core-to-core FIFO communications
DMA global memory transfer
FIFO data communications
ADVANCE INFORMATION
The subsystems’ FIFO communications interface is shown in the ’5421 functional block diagram (Figure 1). Two
unidirectional 8-word-deep FIFOs are available in the device for efficient interprocessor communication: one
configured for core A-to-core B data transfers, and the other configured for core B-to-core A data transfers. Each
subsystem, by way of DMA control, can write to its respective output data FIFO and read from its respective
input data FIFO. The FIFOs are accessed using the DMAs I/O space, which is completely independent of the
CPU I/O space. The DMA transfers to or from the FIFOs can be synchronized to “receive FIFO not empty” and
“transmit FIFO not full” events, providing protection from overflow and underflow. Subsystems can interrupt
each other to flag when the FIFOs are either full or empty. The interprocessor interrupt request bit (IPIRQ) (bit
8 in the BSCR register (BSCR.8)) is set to 1 to generate a PINT in the other subsystem’s IFR.14. See the
interrupts section for more information.
DMA global memory transfers
The ’5421 enables each subsystem to transfer data directly between the memories that are CPU local via DMA
global memory transfers. The DMA global memory map is shown in Figure 13.
chip subsystem ID Register
The chip subsystem ID Register (CSIDR) is a read-only memory-mapped register located at 3Eh within each
DSP subsystem. This register contains three elements for electrically readable device identification. The ChipID
bits identify the type of 54x device (21h for 5421). The ChipRev bits contain the revision number of the device.
Lastly, the SubSysID contains a unique subsystem identifier.
15
14
13
12
11
10
9
8
7
Chip ID
6
5
4
Chip Rev
Figure 17. Chip subsystem ID Register
Table 12. Chip subsystem ID Register Bit Functions
BIT
NO.
36
BIT FIELD
NAME
FUNCTION
15 8
15–8
Chip ID
7–4
Chip Rev
type Contains 21h for ’5421.
’5421
54x device type.
Revision number of device (i.e., 0h for revision 0).
3 0
3–0
SubSysID
Identifier for DSP subsystem: A = 0h
0h, B = 1h
1h.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
3
2
1
SubSysID
0
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
general-purpose I/O
In addition to the A_XF and B_XF pins, the ’5421 has eight general-purpose I/O pins. These pins are:
A_GPIO0, A_GPIO1, A_GPIO2, A_GPIO3
B_GPIO0, B_GPIO1, B_GPIO2, B_GPIO3
Four general-purpose I/O pins are available to each core. Each GPIO pin can be individually selected as either
an input or an output. Additionally, the timer output is selectable on GPIO pin 3. At core reset, all GPIO pins are
configured as inputs. GPIO data and control bits are accessible through a memory-mapped register at 3Ch with
the format shown in Figure 18.
14
11
10
9
8
7
6
5
4
3
2
1
0
GPIO
DIR3
GPIO
DIR2
GPIO
DIR1
GPIO
DIR0
ROM
EN
CORE
SEL
XIO
GRAN
T
XIO
REQ
GPIO
DAT3
GPIO
DAT2
GPIO
DAT1
GPIO
DAT0
R/W
R/W
R/W
R/W
R/W
R/W
R
R
R/W,
R/W,
R/W,
R/W,
R/W,
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
+0
TOUT
Note:
13
12
Reserved
ADVANCE INFORMATION
15
R = Read, W = Write, +0 = Value at reset
Figure 18. General-Purpose I/O Control Register
Table 13. General-Purpose I/O Control Register Bit Functions
BIT
NO.
BIT
NAME
15
TOUT
14-12
11 8
11–8
7
BIT
VALUE
FUNCTION
0
Timer output disable. Uses GPIO3 as general-purpose I/O.
1
Timer output enable. Overrides DIR3. Timer output is driven on GPIO3 and readable in DAT3.
Reserved
X
Register bit is reserved. Read 0, write has no effect.
GPIO
DIRn†
0
GPIOn pin is used as an input.
1
GPIOn pin is used as an output.
0
ROM is mapped out
1
ROM is mapped in
cLEAD core A is selected for XIO REQ bit. DSP subsystem A is tied low internally for this bit.
ROMEN
6
CORE
SEL
0
1
cLEAD core B is selected for XIO REQ bit. DSP subsystem B is tied high internally for this bit.
0
EMIF is not available to the cLEAD core determined by the CORE SEL bit.
5
XIO
GRANT
1
EMIF is granted to the cLEAD core determined by the CORE SEL bit.
XIO
REQ
0
EMIF is not requested for the cLEAD core indicated by the CORE SEL bit.
4
1
Request EMIF for the cLEAD core indicated by the CORE SEL bit.
GPIO
DATn†
0
GPIOn is driven with a 0 (DIRn = 1). GPIOn is read as 0 (DIRn = 0).
1
GPIOn is driven with a 1 (DIRn = 1). GPIOn is read as 1 (DIRn = 0).
3 0
3–0
† n = 3, 2, 1, or 0
Register bit 7 is used as ROMEN to enable and disable ROM space. In XIO mode, ROM enable (ROMEN)
reflects the state of the A_GPIO0 and B_GPIO0 pins (GPIODAT0 input) to enable the applicable on-chip ROM
after reset. Register bits (6:4) are used for XIO arbitration of external memory interface (EMIF) control between
DSP subsystems. The timer out (TOUT) bit is used to multiplex the output of the timer and GPIO3. All GPIO
pins are programmable as an input or output by the direction bit (DIRn). Data is either driven or read from the
data bit field (DATn). DIR3 has no affect when TOUT = 1.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
37
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
general-purpose I/O (continued)
GPIO2 is a special case where the logic level determines the operation of BIO-conditional instructions on the
CPU. GPIO2 is always mapped as a general-purpose I/O, but the BIO function exists when this pin is configured
as an input.
hardware timer
The ’54x devices feature a 16-bit timing circuit with a 4-bit prescaler. The timer counter decrements by one at
every CLKOUT cycle. Each time the counter decrements to zero, a timer interrupt is generated. The timer can
be stopped, restarted, reset, or disabled by specific status bits. The timer output pulse is driven on GPIO3 when
the TOUT bit is set to one in the general-purpose I/O control register. The device must be in HPI mode
(XIO = 0) to drive TOUT on the GPIO3 pin.
software-programmable phase-locked loop (PLL)
ADVANCE INFORMATION
The clock generator provides clocks to the ’5421 device, and consists of a phase-locked loop (PLL) circuit. The
clock generator requires a reference clock input, which must be provided by using an external clock source. The
reference clock input is then divided by two (DIV mode) to generate clocks for the ’5421 device. Alternately, the
PLL circuit can be used (PLL mode) to generate the device clock by multiplying the reference clock frequency
by a scale factor, allowing use of a clock source with a lower frequency than that of the CPU. The PLL is an
adaptive circuit that, once synchronized, locks onto and tracks an input clock signal. When the PLL is initially
started, it enters a transitional mode during which the PLL acquires lock with the input signal. Once the PLL is
locked, it continues to track and maintain synchronization with the input signal. Then, other internal clock
circuitry allows the synthesis of new clock frequencies for use as master clock for the ’5421 device. Only
subsystem A controls the PLL. Subsystem B cannot access the PLL registers.
The software-programmable PLL features a high level of flexibility, and includes a clock scaler that provides
various clock multiplier ratios, capability to directly enable and disable the PLL, and a PLL lock timer that can
be used to delay switching to PLL clocking mode of the device until lock is achieved. Devices that have a built-in
software-programmable PLL can be configured in one of two clock modes:
PLL mode. The input clock (CLKIN) is multiplied by 1 of 31 possible ratios. These ratios are achieved using
the PLL circuitry.
DIV (divider) mode. The input clock is divided by 2 or 4. Note that when DIV mode is used, the PLL can be
completely disabled in order to minimize power dissipation.
The software-programmable PLL is controlled using the 16-bit memory-mapped (address 0058h) clock mode
register (CLKMD). The CLKMD register is used to define the clock configuration of the PLL clock module.
Figure 19 shows the bit layout of the clock mode register and Table 14 describes the bit functions.
15
2
1
0
PLLMUL†
12
11
PLLDIV†
10
PLLCOUNT†
3
PLLON/OFF†
PLLNDIV
STATUS
R/W
R/W
R/W
R/W
R/W
R/W
† When in DIV mode (PLLSTATUS is low), PLLMUL, PLLDIV, PLLCOUNT, and PLLON/OFF are don’t cares, and their contents are indeterminate.
LEGEND: R = Read, W = Write
Figure 19. Clock Mode Register (CLKMD)
38
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
software-programmable phase-locked loop (PLL) (continued)
Table 14. Clock Mode Register (CLKMD) Bit Functions
BIT
NO.
BIT
NAME
15–12
PLLMUL†
11
PLLDIV†
FUNCTION
PLL multiplier. PLLMUL defines the frequency multiplier in conjunction with PLLDIV and PLLNDIV. See Table 15.
PLLDIV = 0
Means that an integer multiply factor is used
PLLDIV = 1
Means that a noninteger multiply factor is used
10–3
PLLCOUNT†
PLL counter value. PLLCOUNT specifies the number of input clock cycles (in increments of16 cycles) for the
PLL lock timer to count before the PLL begins clocking the processor after the PLL is started. The PLL counter
is a down-counter, which is driven by the input clock divided by 16; therefore, for every 16 input clocks, the PLL
counter decrements by one.
The PLL counter can be used to ensure that the processor is not clocked until the PLL is locked, so that only valid
clock signals are sent to the device.
2
PLLON/OFF†
PLL on/off. PLLON/OFF enables or disables the PLL part of the clock generator in conjunction with the PLLNDIV
bit (see Table 16). Note that PLLON/OFF and PLLNDIV can both force the PLL to run; when PLLON/OFF is high,
the PLL runs independently of the state of PLLNDIV.
1
PLLNDIV
PLLNDIV configures PLL mode when high or DIV mode when low. PLLNDIV defines the frequency multiplier in
conjunction with PLLDIV and PLLMUL. See Table 15.
0
STATUS
Indicates the PLL mode.
STATUS = 0
Indicates DIV mode
STATUS = 1
Indicates PLL mode
† When in DIV mode (PLLSTATUS is low), PLLMUL, PLLDIV, PLLCOUNT, and PLLON/OFF are don’t cares, and their contents are indeterminate.
Table 15. Multiplier Related to PLLNDIV, PLLDIV, and PLLMUL
MULTIPLIER‡
PLLNDIV
PLLDIV
PLLMUL
0
x
0–14
0.5
0
x
15
0.25
1
0
0–14
1
0
15
PLLMUL + 1
bypass (multiply by 1)§
1
1
0 or even
(PLLMUL + 1)/2
odd
PLLMUL/4
1
1
‡ CLKOUT = CLKIN * Multiplier
§ Indicates the default clock mode after reset
Table 16. VCO Truth Table
PLLON/OFF
PLLNDIV
VCO STATE
0
0
off
1
0
on
0
1
on
1
1
on
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
39
ADVANCE INFORMATION
PLL divider. PLLDIV defines the frequency multiplier in conjunction with PLLMUL and PLLNDIV. See Table 15.
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
PLL clock programmable timer
During the lockup period, the PLL should not be used to clock the ’5421. The PLLCOUNT programmable lock
timer provides a convenient method of automatically delaying clocking of the device by the PLL until lock is
achieved.
The PLL lock timer is a counter, loaded from the PLLCOUNT field in the CLKMD register, that decrements from
its preset value to 0. The timer can be preset to any value from 0 to 255, and its input clock is CLKIN divided
by 16. The resulting lockup delay can therefore be set from 0 to 255 × 16 CLKIN cycles.
The lock timer is activated when the operating mode of the clock generator is switched from DIV to PLL. During
the lockup period, the clock generator continues to operate in DIV mode; after the PLL lock timer decrements
to zero, the PLL begins clocking the ’5421.
Accordingly, the value loaded into PLLCOUNT is chosen based on the following formula:
PLLCOUNT +
Lockup Time
16
T CLKIN
ADVANCE INFORMATION
where TCLKIN is the input reference clock period and lockup time is the required VCO lockup time, as shown in
Table 17.
Table 17. VCO Lockup Time
40
CLKOUT FREQUENCY (MHz)
LOCKUP TIME (µs)
5
23
10
17
20
16
40
19
60
24
80
29
100
35
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
memory-mapped registers
Table 18. Processor Memory-Mapped Registers for Each DSP Subsystem
IMR
IFR
ADDRESS
DEC
HEX
0
DESCRIPTION
0
Interrupt Mask Register
Interrupt Flag Register
1
1
2–5
2–5
ST0
6
6
Status Register 0
ST1
7
7
Status Register 1
AL
8
8
Accumulator A Low Word (15–0)
—
Reserved for testing
AH
9
9
Accumulator A High Word (31–16)
AG
10
A
Accumulator A Guard Bits (39–32)
BL
11
B
Accumulator B Low Word (15–0)
BH
12
C
Accumulator B High Word (31–16)
BG
13
D
Accumulator B Guard Bits (39–32)
TREG
14
E
Temporary Register
TRN
15
F
Transition Register
AR0
16
10
Auxiliary Register 0
AR1
17
11
Auxiliary Register 1
AR2
18
12
Auxiliary Register 2
AR3
19
13
Auxiliary Register 3
AR4
20
14
Auxiliary Register 4
AR5
21
15
Auxiliary Register 5
AR6
22
16
Auxiliary Register 6
AR7
23
17
Auxiliary Register 7
SP
24
18
Stack Pointer
BK
25
19
Circular Buffer Size Register
BRC
26
1A
Block-Repeat Counter
RSA
27
1B
Block-Repeat Start Address
REA
28
1C
Block-Repeat End Address
PMST
29
1D
Processor Mode Status Register
XPC
30
1E
Extended Program Counter
—
31
1F
Reserved
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
ADVANCE INFORMATION
NAME
41
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
memory-mapped registers (continued)
Table 19. Peripheral Memory-Mapped Registers for Each DSP Subsystem
ADVANCE INFORMATION
NAME
ADDRESS
DEC
HEX
DESCRIPTION
DRR20
32
20
McBSP 0 Data Receive Register 2
DRR10
33
21
McBSP 0 Data Receive Register 1
DXR20
34
22
McBSP 0 Data Transmit Register 2
DXR10
35
23
McBSP 0 Data Transmit Register 1
TIM
36
24
Timer Register
PRD
37
25
Timer Period Register
TCR
38
26
Timer Control Register
—
39
27
Reserved
SWWSR
40
28
Software Wait-State Register
BSCR
41
29
Bank-Switching Control Register
—
42
2A
Reserved
SWCR
43
2B
Software Wait-State Control Register
HPIC
44
2C
HPI Control Register (HMODE=0 only)
—
45–47
2D–2F
DRR22
48
30
McBSP 2 Data Receive Register 2
DRR12
49
31
McBSP 2 Data Receive Register 1
DXR22
50
32
McBSP 2 Data Transmit Register 2
DXR12
51
33
McBSP 2 Data Transmit Register 1
SPSA2
52
34
SPSD2
53
35
McBSP 2 Subbank Address Register†
McBSP 2 Subbank Data Register†
54–55
36–37
56
38
—
SPSA0
SPSD0
Reserved
Reserved
McBSP 0 Subbank Address Register†
McBSP 0 Subbank Data Register†
57
39
58–59
3A–3B
GPIO
60
3C
General-Purpose I/O Register
—
61
3D
Reserved
CSIDR
62
3E
Chip Subsystem ID register
—
63
3F
Reserved
DRR21
64
40
McBSP 1 Data Receive Register 2
DRR11
65
41
McBSP 1 Data Receive Register 1
DXR21
66
42
McBSP 1 Data Transmit Register 2
DXR11
67
43
McBSP 1 Data Transmit Register 1
68–71
44–47
SPSA1
72
48
SPSD1
73
49
—
—
—
Reserved
Reserved
McBSP 1 Subbank Address Register†
McBSP 1 Subbank Data Register†
74–83
4A–53
DMPREC
84
54
DMSA
85
55
DMSDI
86
56
DMSDN
87
57
DMA Subbank Data Register with Autoincrement‡
DMA Subbank Data Register‡
88
58
Clock Mode Register (CLKMD)
89–95
59–5F
CLKMD
—
Reserved
DMA Priority and Enable Control Register
DMA Subbank Address Register‡
Reserved
† See Table 20 for a detailed description of the McBSP control registers and their subaddresses.
‡ See Table 21 for a detailed description of the DMA subbank addressed registers.
42
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
McBSP control registers and subaddresses
The control registers for the multichannel buffered serial port (McBSP) are accessed using the subbank
addressing scheme. This allows a set or subbank of registers to be accessed through a single memory location.
The McBSP subbank address register (SPSA) is used as a pointer to select a particular register within the
subbank. The McBSP data register (SPSDx) is used to access (read or write) the selected register. Table 20
shows the McBSP control registers and their corresponding subaddresses.
Table 20. McBSP Control Registers and Subaddresses
McBSP1
McBSP2
NAME
ADDRESS
NAME
ADDRESS
SUBADDRESS
39h
SPCR11
49h
SPCR12
35h
00h
Serial port control register 1
39h
SPCR21
49h
SPCR22
35h
01h
Serial port control register 2
39h
RCR11
49h
RCR12
35h
02h
Receive control register 1
39h
RCR21
49h
RCR22
35h
03h
Receive control register 2
XCR10
39h
XCR11
49h
XCR12
35h
04h
Transmit control register 1
NAME
ADDRESS
SPCR10
SPCR20
RCR10
RCR20
DESCRIPTION
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁ
XCR20
39h
XCR21
49h
XCR22
35h
05h
Transmit control register 2
SRGR10
39h
SRGR11
49h
SRGR12
35h
06h
Sample rate generator register 1
SRGR20
39h
SRGR21
49h
SRGR22
35h
07h
Sample rate generator register 2
MCR10
39h
MCR11
49h
MCR12
35h
08h
Multichannel register 1
MCR20
39h
MCR21
49h
MCR22
35h
09h
Multichannel register 2
RCERA0
39h
RCERA1
49h
RCERA2
35h
0Ah
Receive channel enable register partition A
RCERB0
39h
RCERB1
49h
RCERB2
35h
0Bh
Receive channel enable register partition B
XCERA0
39h
XCERA1
49h
XCERA2
35h
0Ch
Transmit channel enable register partition A
XCERB0
39h
XCERB1
49h
XCERB2
35h
0Dh
Transmit channel enable register partition B
PCR0
39h
PCR1
49h
PCR2
35h
0Eh
Pin control register
RCERC0
39h
RCERC1
49h
RCERC2
35h
010h
Receive channel enable register partition C
RCERD0
39h
RCERD1
49h
RCERD2
35h
011h
Receive channel enable register partition D
XCERC0
39h
XCERC1
49h
XCERC2
35h
012h
Transmit channel enable register partition C
XCERD0
39h
XCERD1
49h
XCERD2
35h
013h
Transmit channel enable register partition D
RCERE0
39h
RCERE1
49h
RCERE2
35h
014h
Receive channel enable register partition E
RCERF0
39h
RCERF1
49h
RCERF2
35h
015h
Receive channel enable register partition F
XCERE0
39h
XCERE1
49h
XCERE2
35h
016h
Transmit channel enable register partition E
XCERF0
39h
XCERF1
49h
XCERF2
35h
017h
Transmit channel enable register partition F
RCERG0
39h
RCERG1
49h
RCERG2
35h
018h
Receive channel enable register partition G
RCERH0
39h
RCERH1
49h
RCERH2
35h
019h
Receive channel enable register partition H
XCERG0
39h
XCERG1
49h
XCERG2
35h
01Ah
Transmit channel enable register partition G
XCERH0
39h
XCERH1
49h
XCERH2
35h
01Bh
Transmit channel enable register partition H
DMA subbank addressed registers
The direct memory access (DMA) controller has several control registers associated with it. The main control
register (DMPREC) is a standard memory-mapped register. However, the other registers are accessed using
the subbank addressing scheme. This allows a set or subbank of registers to be accessed through a single
memory location. The DMA subbank address (DMSA) register is used as a pointer to select a particular register
within the subbank, while the DMA subbank data (DMSD) register or the DMA subbank data register with
autoincrement (DMSDI) is used to access (read or write) the selected register.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
43
ADVANCE INFORMATION
McBSP0
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
DMA subbank addressed registers (continued)
When the DMSDI register is used to access the subbank, the subbank address is automatically
postincremented so that a subsequent access affects the next register within the subbank. This autoincrement
feature is intended for efficient, successive accesses to several control registers. If the autoincrement feature
is not required, the DMSDN register should be used to access the subbank. Table 21 shows the DMA controller
subbank addressed registers and their corresponding subaddresses.
Table 21. DMA Subbank Addressed Registers
ÁÁÁÁÁ
ÁÁÁÁÁ
Á
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
Á
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
Á
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
Á
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
Á
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
Á
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
Á
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
Á
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ADDRESS
SUBADDRESS
DMSRC0
56h/57h
00h
DMA channel 0 source address register
DMDST0
56h/57h
01h
DMA channel 0 destination address register
DMCTR0
56h/57h
02h
DMA channel 0 element count register
DMSFC0
56h/57h
03h
DMA channel 0 sync select and frame count register
DMMCR0
56h/57h
04h
DMA channel 0 transfer mode control register
DMSRC1
56h/57h
05h
DMA channel 1 source address register
DMDST1
56h/57h
06h
DMA channel 1 destination address register
DMCTR1
56h/57h
07h
DMA channel 1 element count register
DMSFC1
56h/57h
08h
DMA channel 1 sync select and frame count register
DMMCR1
56h/57h
09h
DMA channel 1 transfer mode control register
DMSRC2
56h/57h
0Ah
DMA channel 2 source address register
DMDST2
56h/57h
0Bh
DMA channel 2 destination address register
DMCTR2
56h/57h
0Ch
DMA channel 2 element count register
DMSFC2
56h/57h
0Dh
DMA channel 2 sync select and frame count register
DMMCR2
56h/57h
0Eh
DMA channel 2 transfer mode control register
DMSRC3
56h/57h
0Fh
DMA channel 3 source address register
DMDST3
56h/57h
10h
DMA channel 3 destination address register
DMCTR3
56h/57h
11h
DMA channel 3 element count register
DMSFC3
56h/57h
12h
DMA channel 3 sync select and frame count register
DMMCR3
56h/57h
13h
DMA channel 3 transfer mode control register
DMSRC4
56h/57h
14h
DMA channel 4 source address register
DMDST4
56h/57h
15h
DMA channel 4 destination address register
DMCTR4
56h/57h
16h
DMA channel 4 element count register
DMSFC4
56h/57h
17h
DMA channel 4 sync select and frame count register
DMMCR4
56h/57h
18h
DMA channel 4 transfer mode control register
DMSRC5
56h/57h
19h
DMA channel 5 source address register
DMDST5
56h/57h
1Ah
DMA channel 5 destination address register
DMCTR5
56h/57h
1Bh
DMA channel 5 element count register
DMSFC5
56h/57h
1Ch
DMA channel 5 sync select and frame count register
DMMCR5
56h/57h
1Dh
DMA channel 5 transfer mode control register
DMSRCP
56h/57h
1Eh
DMA source program page address (common channel)
DMDSTP
56h/57h
1Fh
DMA destination program page address (common channel)
DMIDX0
56h/57h
20h
DMA element index address register 0
DMIDX1
56h/57h
21h
DMA element index address register 1
DMFRI0
56h/57h
22h
DMA frame index register 0
DMFRI1
56h/57h
23h
DMA frame index register 1
ADVANCE INFORMATION
NAME
44
DESCRIPTION
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
DMA subbank addressed registers (continued)
Table 21. DMA Subbank Addressed Registers (Continued)
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ADDRESS
SUBADDRESS
DMGSA0
56h/57h
24h
DMA channel 0 global source address reload register
DMGDA0
56h/57h
25h
DMA channel 0 global destination address reload register
DMGCR0
56h/57h
26h
DMA channel 0 global count reload register
DMGFR0
56h/57h
27h
DMA channel 0 global frame count reload register
XSRCDP
56h/57h
28h
DMA extended source data page
XDSTDP
56h/57h
29h
DMA extended destination data page
DMGSA1
56h/57h
2Ah
DMA channel 1 global source address reload register
DMGDA1
56h/57h
2Bh
DMA channel 1 global destination address reload register
DMGCR1
56h/57h
2Ch
DMA channel 1 global count reload register
DMGFR1
56h/57h
2Dh
DMA channel 1 global frame count reload register
DMGSA2
56h/57h
2Eh
DMA channel 2 global source address reload register
DMGDA2
56h/57h
2Fh
DMA channel 2 global destination address reload register
DMGCR2
56h/57h
30h
DMA channel 2 global count reload register
DMGFR2
56h/57h
31h
DMA channel 2 global frame count reload register
DMGSA3
56h/57h
32h
DMA channel 3 global source address reload register
DMGDA3
56h/57h
33h
DMA channel 3 global destination address reload register
DMGCR3
56h/57h
34h
DMA channel 3 global count reload register
DMGFR3
56h/57h
35h
DMA channel 3 global frame count reload register
DMGSA4
56h/57h
36h
DMA channel 4 global source address reload register
DMGDA4
56h/57h
37h
DMA channel 4 global destination address reload register
DMGCR4
56h/57h
38h
DMA channel 4 global count reload register
DMGFR4
56h/57h
39h
DMA channel 4 global frame count reload register
DMGSA5
56h/57h
3Ah
DMA channel 5 global source address reload register
DMGDA5
56h/57h
3Bh
DMA channel 5 global destination address reload register
DMGCR5
56h/57h
3Ch
DMA channel 5 global count reload register
DMGFR5
56h/57h
3Dh
DMA channel 5 global frame count reload register
DESCRIPTION
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
45
ADVANCE INFORMATION
NAME
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
interrupts
Vector-relative locations and priorities for all internal and external interrupts are shown in Table 22.
Table 22. ’5421 Interrupt Locations and Priorities for Each DSP Subsystem
ADVANCE INFORMATION
NAME
LOCATION
PRIORITY
FUNCTION
DECIMAL
HEX
RS, SINTR
0
00
1
Reset (Hardware and Software Reset)
NMI, SINT16
4
04
2
Nonmaskable Interrupt
SINT17
8
08
—
Software Interrupt #17
SINT18
12
0C
—
Software Interrupt #18
SINT19
16
10
—
Software Interrupt #19
SINT20
20
14
—
Software Interrupt #20
SINT21
24
18
—
Software Interrupt #21
SINT22
28
1C
—
Software Interrupt #22
SINT23
32
20
—
Software Interrupt #23
SINT24
36
24
—
Software Interrupt #24
SINT25
40
28
—
Software Interrupt #25
SINT26
44
2C
—
Software Interrupt #26
SINT27
48
30
—
Software Interrupt #27
SINT28
52
34
—
Software Interrupt #28
SINT29
56
38
—
Software Interrupt #29
SINT30
60
3C
—
Software Interrupt #30
INT0, SINT0
64
40
3
External User Interrupt #0
INT1, SINT1
68
44
4
External User Interrupt #1
INT2, SINT2
72
48
5
Reserved
TINT, SINT3
76
4C
6
External Timer Interrupt
BRINT0, SINT4
80
50
7
BSP #0 Receive Interrupt
BXINT0, SINT5
84
54
8
BSP #0 Transmit Interrupt
BRINT2, DMAC0
88
58
9
BSP #2 Receive Interrupt or DMA Channel 0
BXINT2, DMAC1
92
5C
10
BSP #2 Receive Interrupt or DMA Channel 1
INT3, SINT8
96
60
11
Reserved
HPINT, SINT9
100
64
12
HPI Interrupt (from DSPINT in HPIC)
BRINT1, DMAC2
104
68
13
BSP #1 Receive Interrupt or DMA Channel 2
BXINT1, DMAC3
108
6C
14
BSP #1 transmit Interrupt or DMA channel 3
DMAC4, SINT12
112
70
15
DMA Channel 4
DMAC5, SINT13
116
74
16
DMA Channel 5
IPINT, SINT14
120
78
17
Interprocessor Interrupt
124–127
7C–7F
—
Reserved
—
46
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
interrupts (continued)
The interprocessor interrupt (IPINT) bit of the interrupt mask register (IMR) and the interrupt flag register (IFR)
allows the subsystem to perform interrupt service routines based on the other subsystem activity. Incoming
IPINT interrupts are latched in IFR.14. Generating an interprocessor interrupt is performed by writing a “1” to
the IPIRQ field of the bank-switching control register (BSCR). Subsequent interrupts must first clear the interrupt
by writing “0” to the IPIRQ field. Figure 20 shows the bit layout of the IMR and the IFR. Table 23 describes the
bit functions.
15
14
13
12
11
10
9
8
Reserved
IPINT
DMAC5
DMAC4
XINT1 or
DMAC3
RINT1 or
DMAC2
HPINT
Reserved
R/W
R/W
R/W
R/W
R/W
7
6
5
4
3
2
1
0
XINT2 or
DMAC1
RINT2 or
DMAC0
XINT0
RINT0
TINT
Reserved
INT1
INT0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
ADVANCE INFORMATION
For example, if subsystem A is required to notify subsystem B of a completed task, subsystem A must write a
“1” to the IPIRQ field to generate a IPINT interrupt on subsystem B. On subsystem B, the IPINT interrupt is
latched in IFR.14. Figure 5 shows the bit layout of the BSCR and Table 6 describes the bit functions.
LEGEND: R = Read, W = Write
Figure 20. Bit Layout of the IMR and IFR Registers for Subsystems A and B
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
47
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
interrupts (continued)
Table 23. Bit Functions for IMR and IFR Registers for Each DSP Subsystem
BIT
NO.
BIT
NAME
BIT
VALUE
15
Reserved
X
Register bit is reserved.
0
IFR/IMR: Interprocessor IRQ has no interrupt pending/is disabled (masked).
14
13
12
IPINT
DMAC5
DMAC4
XINT1
11
ADVANCE INFORMATION
DMAC3
RINT1
10
DMAC2
9
HPINT
8
Reserved
XINT2
7
DMAC1
RINT2
6
DMAC0
5
4
48
XINT0
RINT0
3
TINT
2
Reserved
1
INT1
0
INT0
FUNCTION
1
IFR/IMR: Interprocessor IRQ has an interrupt pending/is enabled.
0
IFR/IMR: DMA Channel 5 has no interrupt pending/is disabled (masked).
1
IFR/IMR: DMA Channel 5 has an interrupt pending/is enabled.
0
IFR/IMR: DMA Channel 4 has no interrupt pending/is disabled (masked).
1
IFR/IMR: DMA Channel 4 has an interrupt pending/is enabled.
0
IFR/IMR: McBSP_1 has no transmit interrupt pending/is disabled (masked).
1
IFR/IMR: McBSP_1 has a transmit interrupt pending/is enabled.
0
IFR/IMR: DMA Channel 3 has no interrupt pending/is disabled (masked).
1
IFR/IMR: DMA Channel 3 has an interrupt pending/is enabled.
0
IFR/IMR: McBSP_1 has no receive interrupt pending/is disabled (masked).
1
IFR/IMR: McBSP_1 has a receive interrupt pending/is enabled.
0
IFR/IMR: DMA Channel 2 has no interrupt pending/is disabled (masked).
1
IFR/IMR: DMA Channel 2 has an interrupt pending/is enabled.
0
IFR/IMR: Host-port interface has no DSPINT interrupt pending/is disabled (masked).
1
IFR/IMR: Host-port interface has an DSPINT interrupt pending/is enabled.
X
Register bit is reserved.
0
IFR/IMR: McBSP_2 has no transmit interrupt pending/is disabled (masked).
1
IFR/IMR: McBSP_2 has a transmit interrupt pending/is enabled.
0
IFR/IMR: DMA Channel 1 has no interrupt pending/is disabled (masked).
1
IFR/IMR: DMA Channel 1 has an interrupt pending/is enabled.
0
IFR/IMR: McBSP_2 has no receive interrupt pending/is disabled (masked).
1
IFR/IMR: McBSP_2 has a receive interrupt pending/is enabled.
0
IFR/IMR: DMA Channel 0 has no interrupt pending/is disabled (masked).
1
IFR/IMR: DMA Channel 0 has an interrupt pending/is enabled.
0
IFR/IMR: McBSP_0 has no receive interrupt pending/is disabled (masked).
1
IFR/IMR: McBSP_0 has a receive interrupt pending/is enabled.
0
IFR/IMR: McBSP_0 has no receive interrupt pending/is disabled (masked).
1
IFR/IMR: McBSP_0 has a receive interrupt pending/is enabled.
0
IFR/IMR: Timer has no interrupt pending/is disabled (masked).
1
IFR/IMR: Timer has an interrupt pending/is enabled.
X
Register bit is reserved.
0
IFR/IMR: Ext user interrupt pin 1 has no interrupt pending/is disabled (masked).
1
IFR/IMR: Ext user interrupt pin 1 has an interrupt pending/is enabled.
0
IFR/IMR: Ext user interrupt pin 0 has no interrupt pending/is disabled (masked).
1
IFR/IMR: Ext user interrupt pin 0 has an interrupt pending/is enabled.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
IDLE3 power-down mode
The IDLE1 and IDLE2 power-down modes operate as described in the TMS320C54x DSP Reference Set,
Volume 1: CPU and Peripherals (literature number SPRU131). The IDLE3 mode is special in that the clocking
circuitry is shut off to conserve power. The ’5421 cannot enter an IDLE3 mode unless both subsystems execute
an IDLE3 instruction. The power-reduced benefits of IDLE3 cannot be realized until both subsystems enter the
IDLE3 state and the internal clocks are automatically shut off. The order in which subsystems enter IDLE3 does
not matter.
emulating the ’5421 device
The ’5421 is a single device, but actually consists of two independent subboundary systems that contain
register/status information used by the emulator tools. The emulator tools must be informed of the multicore
device by modifying the board.cfg file. The board.cfg file is an ASCII file that can be modified with most editors.
This provides the emulator with a description of the JTAG chain. The board.cfg file must identify two processors
when using the ’5421. The file contents would look something like this:
“CPU_A” TI320C5xx
Use the compose program to make this file into a binary file (board.dat), readable by the emulation tools. Place
the board.dat file in the directory that contains the emulator software.
The subsystems are serially connected together internally. Emulation information is serially transmitted into the
device using TDI. The device responds with serial scan information transmitted out the TDO pin.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
49
ADVANCE INFORMATION
“CPU_B” TI320C5xx
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
documentation support
Extensive documentation supports all TMS320 family generations of devices from product announcement
through applications development. The following types of documentation are available to support the design
and use of the ’C5000 family of DSPs:
TMS320C5000 DSP Family Functional Overview (literature number SPRU307)
Device-specific data sheets (such as this document)
Complete User Guides
Development-support tools
Hardware and software application reports
The five-volume TMS320C54x DSP Reference Set (literature number SPRU210) consists of:
ADVANCE INFORMATION
Volume 1: CPU and Peripherals (literature number SPRU131)
Volume 2: Mnemonic Instruction Set (literature number SPRU172)
Volume 3: Algebraic Instruction Set (literature number SPRU179)
Volume 4: Applications Guide (literature number SPRU173)
Volume 5: Enhanced Peripherals (literature number SPRU302)
The reference set describes in detail the ’54x TMS320 products currently available and the hardware and
software applications, including algorithms, for fixed-point TMS320 devices.
For general background information on DSPs and Texas Instruments (TI) devices, see the three-volume
publication Digital Signal Processing Applications with the TMS320 Family (literature numbers SPRA012,
SPRA016, and SPRA017).
A series of DSP textbooks is published by Prentice-Hall and John Wiley & Sons to support digital signal
processing research and education. The TMS320 newsletter, Details on Signal Processing, is published
quarterly and distributed to update TMS320 customers on product information.
Information regarding TI DSP products is also available on the Worldwide Web at http://www.ti.com uniform
resource locator (URL).
TI is a trademark of Texas Instruments Incorporated.
50
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
absolute maximum ratings over specified temperature range (unless otherwise noted)†
Supply voltage I/O range, DVDD‡ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 0.5 V to 4.0 V
Supply voltage core range, CVDD‡ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 0.5 V to 2.4 V
Supply voltage analog PLL, AVDD‡ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 0.5 V to 2.4 V
Input voltage range, VI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 0.5 V to DVDD + 0.5 V
Output voltage range, Vo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 0.5 V to DVDD + 0.5 V
Operating case temperature range, TC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0°C to 85°C
Storage temperature range Tstg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 65 C to 150 C
† Stresses beyond those listed under “absolute maximum ratings” may cause permanent damage to the device. These are stress ratings only, and
functional operation of the device at these or any other conditions beyond those indicated under “recommended operating conditions” is not
implied. Exposure to absolute-maximum-rated conditions for extended periods may affect device reliability.
‡ All voltage values are with respect to VSS.
recommended operating conditions
NOM
MAX
3.0
3.3
3.6
V
Device supply voltage, core
1.71
1.80
1.98
V
AVDD
VSS
Device supply voltage, PLL
1.71
1.80
1.98
V
VIH
High-level
High
level in
input
ut voltage, I/O
Device supply voltage, I/O
CVDD
VIL
Supply voltage, GND
Low-level
Low
level in
input
ut voltage, I/O
IOH
High-level output current
IOL
TC
Low-level output current
0
Schmitt triggered inputs
DVDD = 3.3 ± 0.3 V
V
0.7DVDD
DVDD
All other inputs
2
DVDD
Schmitt triggered inputs
DVDD = 3.3 ± 0.3 V
0
0.3DVDD
All other inputs
0
0.8
Operating case temperature
0
UNIT
ADVANCE INFORMATION
MIN
DVDD
V
V
–300
µA
1.5
mA
85
°C
See Figure 21 for 3.3-V device test load circuit values.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
51
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
electrical characteristics over recommended operating case temperature range (unless otherwise
noted)
VOH
VOL
IIZ
II
PARAMETER
High-level output voltage‡
TEST CONDITIONS
MAX
UNIT
2.4
0.4
V
10
–10
10
A
µA
TRST
VDD = MAX
MAX, VO = VSS to VDD
With internal pulldown
–10
35
See pin descriptions
With internal pullups
–35
10
PPD[15:0]
Bus holders enabled, VDD = MAX,
VI = VSS to VDD
–175
175
–10
10
Input current in high impedance
All other input-only pins
ADVANCE INFORMATION
TYP†
VDD = 3.3 ± 0.3 V, IOH = MAX
IOL = MAX
Low-level output voltage‡
Input current
(VI = VSS to VDD)
MIN
IDDC
Supply current, both core CPUs
CVDD = 1.8 V, fx = 100 MHz§,
TC = 25°C
IDDP
Supply current, pins
DVDD = 3.3 V, fclock = 100 MHz¶,
TC = 25°C#
IDDA
Supply current, PLL
IDDC
Supply current,
current standby
Ci
Input capacitance
V
µA
90¶
mA
54
mA
5
mA
IDLE2
PLL × n mode, 20 MHz input
2
mA
IDLE3
PLL x n mode, 20 MHz input
600
µA
10
pF
Co
Output capacitance
10
pF
† All values are typical unless otherwise specified.
‡ All input and output voltage levels except RS, INT0, INT1, NMI, CLKIN, BCLKX, BCLKR, HAS, HCS, HDS1, HDS2, and HPIRS are
LVTTL-compatible.
§ Clock mode: PLL × 1 with external source
¶ This value is based on 50% usage of MAC and 50% usage of NOP instructions. Actual operating current varies with the program being executed.
# This value was obtained using the following conditions: external memory writes at a rate of 20 million writes per second, CLKOFF = 0, full-duplex
operation of all six McBSPs at a rate of 10 million bits per second each, and 15-pF loads on all outputs. For more details on how this calculation
is performed, refer to the Calculation of TMS320LC54x Power Dissipation Application Report (literature number SPRA164).
PARAMETER MEASUREMENT INFORMATION
IOL
50 Ω
Tester Pin
Electronics
VLoad
CT
IOH
Where:
IOL
IOH
VLoad
CT
=
=
=
=
1.5 mA (all outputs)
300 µA (all outputs)
1.5 V
40 pF typical load circuit capacitance
Figure 21. 3.3-V Test Load Circuit
52
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
Output
Under
Test
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
divide-by-two, divide-by-four, and bypass clock option (PLL Disabled)
The frequency of the reference clock provided at the CLKIN pin can be divided by a factor of two or four to
generate the internal machine cycle. The selection of the clock mode is described in the software-programmable
phase-locked loop (PLL) section.
switching characteristics over recommended operating conditions [H = 0.5tc(CO)]† (see Figure 22,
and the recommended operating conditions table)
’5421-200
MIN
TYP
MAX
UNIT
†
ns
40
2tc(CI)
2tc(CI)
†
ns
3
6
10
ns
tc(CO)
tc(CO)
Cycle time, CLKOUT
40
Cycle time, CLKOUT – bypass mode
td(CIH-CO)
tf(CO)
Delay time, CLKIN high to CLKOUT high/low
Fall time, CLKOUT
2
ns
tr(CO)
tw(COL)
Rise time, CLKOUT
2
ns
Pulse duration, CLKOUT low
H–2
H–1
H
ns
tw(COH)
Pulse duration, CLKOUT high
H–2
H–1
H
ns
† This device utilizes a fully static design and therefore can operate with tc(CI) approaching ∞. The device is characterized at frequencies
approaching 0 Hz.
timing requirements (see Figure 22)
’5421-200
UNIT
MIN
MAX
20
†
ns
tc(CI)
tf(CI)
Cycle time, CLKIN
Fall time, CLKIN
8
ns
tr(CI)
Rise time, CLKIN
8
ns
tw(CIL)
Pulse duration, CLKIN low
5
ns
tw(CIH) Pulse duration, CLKIN high
5
ns
† This device utilizes a fully static design and therefore can operate with tc(CI) approaching ∞. The device is characterized at frequencies
approaching 0 Hz.
tr(CI)
tw(CIH)
tw(CIL)
tc(CI)
tf(CI)
CLKIN
tw(COH)
tf(CO)
tc(CO)
tr(CO)
td(CIH-CO)
tw(COL)
CLKOUT
Figure 22. External Divide-by-Two Clock Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
53
ADVANCE INFORMATION
PARAMETER
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multiply-by-N clock option (PLL Enabled)
The frequency of the reference clock provided at the CLKIN pin can be multiplied by a factor of N to generate
the internal machine cycle. The selection of the clock mode and the value of N is described in the
software-programmable phase-locked loop (PLL) section.
switching characteristics
(see Figure 23)
over
recommended
operating
conditions
[H
=
0.5tc(CO)]
’5421-200
ADVANCE INFORMATION
PARAMETER
MIN
MAX
10
TYP
tc(CI)/N†
4
10
16
UNIT
tc(CO)
td(CI-CO)
Cycle time, CLKOUT
tf(CO)
tr(CO)
Fall time, CLKOUT
tw(COL)
tw(COH)
Pulse duration, CLKOUT low
H–2
H–1
H
ns
Pulse duration, CLKOUT high
H–2
H–1
H
ns
30
ms
Delay time, CLKIN high/low to CLKOUT high/low
ns
2
Rise time, CLKOUT
ns
2
tp
Transitory phase, PLL lock up time
† N = Multiplication factor
ns
ns
timing requirements (see Figure 23)†
’5421-200
Integer PLL multiplier N (N = 1–15)
PLL multiplier N = x.5
MIN
20‡
MAX
20‡
20‡
100
UNIT
200
tc(CI)
Cycle time, CLKIN
tf(CI)
tr(CI)
Fall time, CLKIN
8
ns
Rise time, CLKIN
8
ns
PLL multiplier N = x.25, x.75
ns
50
tw(CIL) Pulse duration, CLKIN low
5
ns
tw(CIH) Pulse duration, CLKIN high
5
ns
† N = Multiplication factor
‡ The multiplication factor and minimum CLKIN cycle time should be chosen such that the resulting CLKOUT cycle time is within the specified range
(tc(CO))
tf(CI)
tw(CIH)
tw(CIL) tr(CI)
tc(CI)
CLKIN
td(CI-CO)
tc(CO)
tw(COL)
tp
CLKOUT
tf(CO)
tw(COH)
Unstable
Figure 23. External Multiply-by-One Clock Timing
54
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
tr(CO)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
external memory interface timing
switching characteristics over recommended operating conditions for a memory read
(MSTRB = 0)† (see Figure 24)
td(CLKL-A)
td(CLKH-A)
PARAMETER
Delay time, CLKOUT low to address valid‡
Delay time, CLKOUT high (transition) to address valid§
td(CLKL-MSL) Delay time, CLKOUT low to MSTRB low
td(CLKL-MSH) Delay time, CLKOUT low to MSTRB high
th(CLKL-A)R Hold time, address valid after CLKOUT low‡
th(CLKH-A)R Hold time, address valid after CLKOUT high§
† Address, PS, and DS timings are all included in timings referenced as address.
‡ In the case of a memory read preceded by a memory read
§ In the case of a memory read preceded by a memory write
MIN
MAX
–1
4
UNIT
ns
–1
4
ns
–1
4
ns
–1
ns
–1
4
4§
–1
4§
ns
ns
MIN
MAX
UNIT
ta(A)M
ta(MSTRBL)
Access time, read data access from address valid
2H
ns
Access time, read data access from MSTRB low
2H
ns
tsu(D)R
th(D)R
Setup time, read data before CLKOUT low
H
ns
Hold time, read data after CLKOUT low
0
ns
0
ns
0
ns
th(A-D)R
Hold time, read data after address invalid
th(D)MSTRBH Hold time, read data after MSTRB high
† Address, PS, and DS timings are all included in timings referenced as address.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
ADVANCE INFORMATION
timing requirements for a memory read (MSTRB = 0) [H = 0.5 tc(CO)]† (see Figure 24)
55
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
external memory interface timing (continued)
CLKOUT
td(CLKL-A)
th(CLKL-A)R
PPA[18:0]
th(A-D)R
tsu(D)R
ta(A)M
th(D)R
PPD[15:0]
ADVANCE INFORMATION
th(D)MSTRBH
td(CLKL-MSL)
td(CLKL-MSH)
ta(MSTRBL)
MSTRB
R/W
PS, DS
Figure 24. Memory Read (MSTRB = 0)
56
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
external memory interface timing for a memory write
switching characteristics over recommended operating conditions for a memory write
(MSTRB = 0) [H = 0.5 tc(CO)]† (see Figure 25)
PARAMETER
Delay time, CLKOUT high to address valid‡
Delay time, CLKOUT low to address valid§
td(CLKL-MSL)
td(CLKL-D)W
Delay time, CLKOUT low to MSTRB low
td(CLKL-MSH)
td(CLKH-RWL)
Delay time, CLKOUT low to MSTRB high
Delay time, CLKOUT low to data valid
Delay time, CLKOUT high to R/W low
td(CLKH-RWH) Delay time, CLKOUT high to R/W high
td(RWL-MSTRBL) Delay time, R/W low to MSTRB low
th(A)W
Hold time, address valid after CLKOUT high‡
† Address, PS, and DS timings are all included in timings referenced as address.
‡ In the case of a memory write preceded by a memory write
§ In the case of a memory write preceded by an I/O cycle
MIN
MAX
–1
4
UNIT
ns
–1
4
ns
–1
4
ns
0
7
ns
–1
5
ns
0
4
ns
–1
4
ns
H–2
H+2
ns
–1
4
ns
timing requirements for a memory write (MSTRB = 0) [H = 0.5 tc(CO)]† (see Figure 25)
th(D)MSH
tw(SL)MS
Hold time, write data valid after MSTRB high
Pulse duration, MSTRB low§
MAX
UNIT
H +3§
ns
4H–4
tsu(A)W
Setup time, address valid before MSTRB low
tsu(D)MSH
Setup time, write data valid before MSTRB high
† Address, PS, and DS timings are all included in timings referenced as address.
§ In the case of a memory write preceded by an I/O cycle
POST OFFICE BOX 1443
MIN
H–3
• HOUSTON, TEXAS 77251–1443
ns
4H–4
4H–5
ns
4H+5§
ns
57
ADVANCE INFORMATION
td(CLKH-A)
td(CLKL-A)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
external memory interface timing for a memory write (continued)
CLKOUT
td(CLKH-A)
td(CLKL-A)
th(A)W
PPA[18:0]
td(CLKL-D)W
th(D)MSH
tsu(D)MSH
PPD[15:0]
td(CLKL-MSL)
td(CLKL-MSH)
tsu(A)W
ADVANCE INFORMATION
MSTRB
td(CLKH-RWH)
td(CLKH-RWL)
tw(SL)MS
td(RWL-MSTRBL)
R/W
PS, DS
Figure 25. Memory Write (MSTRB = 0)
58
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
ready timing for externally generated wait states
timing requirements for externally generated wait states [H = 0.5 tc(CO)]† (see Figure 26 and
Figure 27)
MIN
tsu(RDY)
th(RDY)
Setup time, READY before CLKOUT low
Hold time, READY after CLKOUT low
tv(RDY)MSTRB Valid time, READY after MSTRB low‡
th(RDY)MSTRB Hold time, READY after MSTRB low‡
MAX
UNIT
5
ns
0
ns
4H–8
ns
4H
ns
† The hardware wait states can be used only in conjunction with the software wait states to extend the bus cycles. To generate wait states by
READY, at least two software wait states must be programmed. READY is not sampled until the completion of the internal software wait states.
‡ These timings are included for reference only. The critical timings for READY are those referenced to CLKOUT
ADVANCE INFORMATION
CLKOUT
PPA[18:0]
tsu(RDY)
th(RDY)
READY
tv(RDY)MSTRB
th(RDY)MSTRB
MSTRB
Wait States
Generated Internally
Wait State
Generated
by READY
Figure 26. Memory Read With Externally Generated Wait States
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
59
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
ready timing for externally generated wait states (continued)
CLKOUT
PPA[18:0]
PPD[15:0]
th(RDY)
tsu(RDY)
READY
ADVANCE INFORMATION
tv(RDY)MSTRB
th(RDY)MSTRB
MSTRB
Wait States
Generated Internally
Wait State Generated
by READY
Figure 27. Memory Write With Externally Generated Wait States
60
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
parallel I/O interface timing
switching characteristics over recommended operating conditions for a parallel I/O port read
(IOSTRB = 0)† (see Figure 28)
PARAMETER
td(CLKL-A)
Delay time, CLKOUT low to address valid
td(CLKH-ISTRBL) Delay time, CLKOUT high to IOSTRB low
td(CLKH-ISTRBH) Delay time, CLKOUT high to IOSTRB high
th(A)IOR
Hold time, address after CLKOUT low
† Address and IS timings are included in timings referenced as address.
MIN
MAX
0
4
UNIT
ns
0
4
ns
0
4
ns
0
4
ns
MIN
MAX
UNIT
5H–3
ns
2H–3
ns
ta(A)IO
ta(ISTRBL)IO
Access time, read data access from address valid
tsu(D)IOR
th(D)IOR
Setup time, read data before CLKOUT high
4
ns
Hold time, read data after CLKOUT high
0
ns
0
ns
Access time, read data access from IOSTRB low
th(ISTRBH-D)R Hold time, read data after IOSTRB high
† Address and IS timings are included in timings referenced as address.
ADVANCE INFORMATION
timing requirements for a parallel I/O port read (IOSTRB = 0) [H = 0.5 tc(CO)]† (see Figure 28)
CLKOUT
th(A)IOR
td(CLKL–A)
PPA[18:0]
tsu(D)IOR
ta(A)IO
th(D)IOR
PPD[15:0]
th(ISTRBH–D)R
td(CLKH–ISTRBH)
ta(ISTRBL)IO
td(CLKH–ISTRBL)
IOSTRB
R/W
IS
Figure 28. Parallel I/O Port Read (IOSTRB=0)
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
61
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
parallel I/O interface timing (continued)
switching characteristics over recommended operating conditions for a parallel I/O port write
(IOSTRB = 0) [H = 0.5 tc(CO)]† (see Figure 29)
PARAMETER
ADVANCE INFORMATION
MIN
MAX
td(CLKL-A)
td(CLKH-ISTRBL)
Delay time, CLKOUT low to address valid
–1
4
ns
Delay time, CLKOUT high to IOSTRB low
0
4
ns
td(CLKH-D)IOW
td(CLKH-ISTRBH)
Delay time, CLKOUT high to write data valid
H–5
H+3
ns
–1
4
ns
td(CLKL-RWL)
td(CLKL-RWH)
Delay time, CLKOUT low to R/W low
0
4
ns
th(A)IOW
th(D)IOW
Hold time, address valid after CLKOUT low
Delay time, CLKOUT high to IOSTRB high
Delay time, CLKOUT low to R/W high
Hold time, write data after IOSTRB high
tsu(D)IOSTRBH
Setup time, write data before IOSTRB high
tsu(A)IOSTRBL
Setup time, address valid before IOSTRB low
† Address and IS timings are included in timings referenced as address.
0
4
ns
–1
4
ns
H–3
H+7
ns
H–5
H+1
ns
H–3
H+3
ns
CLKOUT
tsu(A)IOSTRBL
th(A)IOW
td(CLKL-A)
PPA[18:0]
td(CLKH-D)IOW
th(D)IOW
PPD[15:0]
td(CLKH-ISTRBL)
td(CLKH-ISTRBH)
tsu(D)IOSTRBH
IOSTRB
td(CLKL-RWH)
td(CLKL-RWL)
R/W
IS
Figure 29. Parallel I/O Port Write (IOSTRB=0)
62
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
UNIT
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
externally generated wait states timing
timing requirements for externally generated wait states [H = 0.5 tc(CO)]† (see Figure 30 and
Figure 31)
MIN
tsu(RDY)
th(RDY)
tv(RDY)IOSTRB
th(RDY)IOSTRB
Setup time, READY before CLKOUT low
5
Hold time, READY after CLKOUT low
Valid time, READY after IOSTRB low‡
0
Hold time, READY after IOSTRB low‡
5H
MAX
UNIT
ns
ns
5H–8
ns
ns
† The hardware wait states can be used only in conjunction with the software wait states to extend the bus cycles. To generate wait states using
READY, at least two software wait states must be programmed.
‡ These timings are included for reference only. The critical timings for READY are those referenced to CLKOUT.
ADVANCE INFORMATION
CLKOUT
PPA[18:0]
th(RDY)
tsu(RDY)
READY
tv(RDY)IOSTRB
th(RDY)IOSTRB
IOSTRB
Wait
States
Generated
Internally
Wait State Generated
by READY
Figure 30. I/O Port Read With Externally Generated Wait States
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
63
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
externally generated wait states timing (continued)
CLKOUT
PPA[18:0]
PPD[15:0]
th(RDY)
tsu(RDY)
READY
ADVANCE INFORMATION
tv(RDY)IOSTRB
th(RDY)IOSTRB
IOSTRB
Wait States
Generated
Internally
Wait State Generated
by READY
Figure 31. I/O Port Write With Externally Generated Wait States
64
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
reset, BIO, interrupt, and XIO timings
timing requirements for reset, BIO, interrupt, and XIO [H = 0.5 tc(CO)] (see Figure 32, Figure 33,
and Figure 34)
MAX
UNIT
Hold time, RS after CLKOUT low
0
ns
Hold time, BIO after CLKOUT low
0
ns
th(INT)
th(XIO)
Hold time, INTn, NMI, after CLKOUT low†
0
ns
Hold time, XIO after CLKOUT low
Pulse duration, RS low‡§
0
ns
4H+5
ns
Pulse duration, BIO low, asynchronous†
5H
ns
Pulse duration, INTn, NMI high (asynchronous)†
Pulse duration, INTn, NMI low (asynchronous)†
4H
ns
tw(RSL)
tw(BIO)A
tw(INTH)A
tw(INTL)A
4H
ns
tw(INTL)WKP Pulse duration, INTn, NMI low for IDLE2/IDLE3 wakeup†
tsu(RS)
Setup time, RS before CLKIN low§
8
ns
5
ns
tsu(BIO)
tsu(INT)
Setup time, BIO before CLKOUT low
9
12
ns
Setup time, INTn, NMI, RS before CLKOUT low
Setup time, XIO before CLKOUT low†
9
13
ns
tsu(XIO)
10
8
ns
† The external interrupts (INT0–INT1, NMI) are synchronized to the core CPU by way of a two flip-flop synchronizer which samples these inputs
with consecutive falling edges of CLKOUT. The input to the interrupt pins is required to represent a 1-0-0 sequence at the timing that is
corresponding to a three-CLKOUT sampling sequence.
‡ If the PLL mode is selected, then at power-on sequence, or at wakeup from IDLE3, RS must be held low for at least 50 µs to ensure
synchronization and lock-in of the PLL.
§ RS can cause a change in clock frequency, changing the value of H (see the software-programmable phase-locked loop (PLL) section).
CLKIN
tsu(RS)
tw(RSL)
A_RS, B_RS,
INTn, NMI
tsu(INT)
th(RS)
CLKOUT
tsu(BIO)
th(BIO)
BIO
tw(BIO)A
Figure 32. Reset and BIO Timings
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
65
ADVANCE INFORMATION
MIN
th(RS)
th(BIO)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
reset, BIO, interrupt, and XIO timings (continued)
CLKOUT
tsu(INT)
tsu(INT)
th(INT)
INTn, NMI
tw(INTH)A
tw(INTL)A
Figure 33. Interrupt Timing
ADVANCE INFORMATION
CLKOUT
th(XIO)
tsu(XIO)
XIO
Figure 34. XIO Timing
66
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HOLD and HOLDA timing
switching characteristics over recommended operating conditions for memory control signals
and HOLDA [H = 0.5 tc(CO)] (see Figure 35)
PARAMETER
tv(HOLDA)
tw(HOLDA)
MIN
MAX
UNIT
Valid time, HOLDA low after CLKOUT low
0
5
ns
Valid time, HOLDA high after CLKOUT low
–2
3
ns
Pulse duration, HOLDA low duration
2H–3
ns
timing requirements for HOLD [H = 0.5 tc(CO)] (see Figure 35)
MAX
tw(HOLD)
tsu(HOLD)
Pulse duration, HOLD low duration
Setup time, HOLD low/high before CLKOUT low
MIN
UNIT
4H+10
ns
10
ns
tsu(HOLD)
ADVANCE INFORMATION
CLKOUT
tsu(HOLD)
tw(HOLD)
HOLD
tv(HOLDA)
tv(HOLDA)
tw(HOLDA)
HOLDA
Figure 35. HOLD and HOLDA Timing (HM = 1)
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
67
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
external flag and timer output timing
switching characteristics over recommended operating conditions for external flag (XF) and TOUT
[H = 0.5 tc(CO)] (see Figure 36 and Figure 37)
PARAMETER
Delay time, CLKOUT low to XF high
td(XF)
Delay time, CLKOUT low to XF low
MIN
MAX
–2
3
0
5
UNIT
ns
td(TOUTH)
td(TOUTL)
Delay time, CLKOUT low to TOUT high
–2
3
ns
Delay time, CLKOUT low to TOUT low
–2
3
ns
tw(TOUT)
Pulse duration, TOUT
2H–10
2H
ns
CLKOUT
ADVANCE INFORMATION
td(XF)
XF
Figure 36. External Flag (XF) Timing
CLKOUT
td(TOUTL)
td(TOUTH)
TOUT
tw(TOUT)
Figure 37. Timer (TOUT) Timing
68
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
general-purpose input output (GPIO) timing
timing requirements for GPIO (see Figure 38)
MIN
MAX
UNIT
tsu(GPIO-COH)
Setup time, GPIOx input valid before CLKOUT high, GPIOx configured as
general-purpose input.
7
ns
th(GPIO-COH)
Hold time, GPIOx input valid after CLKOUT high, GPIOx configured as general-purpose
input.
0
ns
PARAMETER
td(COH-GPIO)
Delay time, CLKOUT high to GPIOx output change. GPIOx configured as
general-purpose output.
MIN
MAX
UNIT
0
5
ns
ADVANCE INFORMATION
switching characteristics for GPIO (see Figure 38)
CLKOUT
tsu(GPIO-COH)
th(GPIO-COH)
GPIOx Input Mode
td(COH-GPIO)
GPIOx Output Mode
Figure 38. GPIO Timings
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
69
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port timing
timing requirements for the McBSP† [H=0.5tc(CO)] (see Figure 39 and Figure 40)
MIN
ADVANCE INFORMATION
tc(BCKRX)
tw(BCKRX)
MAX
UNIT
Cycle time, BCLKR/X
BCLKR/X ext
4H
ns
Pulse duration, BCLKR/X low or BCLKR/X high
BCLKR/X ext
6
ns
BCLKR int
0
BCLKR ext
4
th(BCKRL-BFRH)
Hold time,
time external BFSR high after BCLKR low
th(BCKRL-BDRV)
Hold time,
time BDR valid after BCLKR low
th(BCKXL-BFXH)
Hold time,
time external BFSX high after BCLKX low
tsu(BFRH-BCKRL)
Setup time,
time external BFSR high before BCLKR low
tsu(BDRV-BCKRL)
time BDR valid before BCLKR low
Setup time,
tsu(BFXH-BCKXL)
Setup time,
time external BFSX high before BCLKX low
BCLKR int
0
BCLKR ext
5
BCLKX int
0
BCLKX ext
4
BCLKR int
10
BCLKR ext
4
BCLKR int
10
BCLKR ext
3
BCLKX int
10
BCLKX ext
6
ns
ns
ns
ns
ns
ns
tr(BCKRX)
Rise time, BCLKR/X
BCLKR/X ext
8
ns
tf(BCKRX)
Fall time, BCLKR/X
BCLKR/X ext
8
ns
† Polarity bits CLKRP = CLKXP = FSRP = FSXP = 0. If the polarity of any of the signals is inverted, then the timing references of that signal are
also inverted.
70
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port timing (continued)
switching characteristics for the McBSP† [H=0.5tc(CO)] (see Figure 39 and Figure 40)
MIN
Cycle time, BCLKR/X
BCLKR/X int
Pulse duration, BCLKR/X high
BCLKR/X int
4H
D–4‡
tw(BCKRXL)
td(BCKRH-BFRV)
Pulse duration, BCLKR/X low
BCLKR/X int
Delay time, BCLKR high to internal BFSR valid
td(BCKXH-BFXV)
Delay time,
time BCLKX high to internal BFSX valid
tdis(BCKXH-BDXHZ)
time BCLKX high to BDX high impedance following last data bit
Disable time,
Delay time, BCLKX high to BDX valid. This applies
a lies to all bits except
exce t the first
bit transmitted.
td(BCKXH-BDXV)
ten(BCKXH-BDX)
td(BFXH-BDXV)
ten(BFXH-BDX)
Delay time, BCLKX high to BDX valid.§
Only applies to first bit transmitted when in Data Delay 1
or 2 (XDATDLY=01b or 10b) modes
Enable time, BCLKX high to BDX driven.§
Only applies to first bit transmitted when in Data Delay 1
or 2 (XDATDLY=01b or 10b) modes
Delay time, BFSX high to BDX valid.§
Only applies to first bit transmitted when in Data Delay 0
(XDATDLY=00b) mode.
Enable time, BFSX high to BDX driven.§
Only applies to first bit transmitted when in Data Delay 0
(XDATDLY=00b) mode
DXENA = 0
DXENA = 1
DXENA = 0
DXENA = 1
DXENA = 0
DXENA = 1
DXENA = 0
MAX
UNIT
ns
ns
C–4‡
D+1‡
C+1‡
BCLKR int
–3
3
ns
BCLKX int
–3
8
BCLKX ext
2
15
BCLKX int
–8
5
BCLKX ext
1
19
BCLKX int
0
11
BCLKX ext
5
20
BCLKX int
11
BCLKX ext
20
BCLKX int
25
BCLKX ext
ns
ns
ns
ns
27
BCLKX int
–4
BCLKX ext
2
BCLKX int
6
BCLKX ext
12
ns
BFSX int
9
BFSX ext
12
BFSX int
25
BFSX ext
ns
26
BFSX int
–1
BFSX ext
2
BFSX int
9
ns
DXENA = 1
BFSX ext
13
† Polarity bits CLKRP = CLKXP = FSRP = FSXP = 0. If the polarity of any of the signals is inverted, then the timing references of that signal are
also inverted.
‡ T=BCLKRX period = (1 + CLKGDV) * 2H
C=BCLKRX low pulse width = T/2 when CLKGDV is odd or zero and = (CLKGDV/2) * 2H when CLKGDV is even
D=BCLKRX high pulse width = T/2 when CLKGDV is odd or zero and = (CLKGDV/2 + 1) * 2H when CLKGDV is even
§ See the TMS320C54x Enhanced Peripherals Reference Set, Volume 5 (literature number SPRU302) for a description of the DX enable (DXENA)
and data delay features of the McBSP.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
71
ADVANCE INFORMATION
PARAMETER
tc(BCKRX)
tw(BCKRXH)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port timing (continued)
tc(BCKRX)
tw(BCKRXH)
tr(BCKRX)
tw(BCKRXL)
BCLKR
td(BCKRH–BFRV)
td(BCKRH–BFRV)
tr(BCKRX)
BFSR (int)
tsu(BFRH–BCKRL)
th(BCKRL–BFRH)
BFSR (ext)
th(BCKRL–BDRV)
tsu(BDRV–BCKRL)
ADVANCE INFORMATION
BDR
(RDATDLY=00b)
Bit (n–1)
(n–2)
tsu(BDRV–BCKRL)
(n–3)
(n–4)
th(BCKRL–BDRV)
BDR
(RDATDLY=01b)
Bit (n–1)
(n–2)
tsu(BDRV–BCKRL)
BDR
(RDATDLY=10b)
(n–3)
th(BCKRL–BDRV)
Bit (n–1)
(n–2)
Figure 39. McBSP Receive Timings
tc(BCKRX)
tw(BCKRXH)
tw(BCKRXL)
tr(BCKRX)
tf(BCKRX)
BCLKX
td(BCKXH–BFXV)
td(BCKXH–BFXV)
BFSX (int)
tsu(BFXH–BCKXL)
th(BCKXL–BFXH)
BFSX (ext)
ten(BDFXH–BDX)
BDX
(XDATDLY=00b)
Bit 0
td(BFXH–BDXV)
Bit (n–1)
td(BCKXH–BDXV)
(n–2)
ten(BCKXH–BDX)
BDX
(XDATDLY=01b)
Bit (n–1)
Bit 0
td(BCKXH–BDXV)
(n–2)
(n–3)
ten(BCKXH–BDX)
Bit 0
Bit (n–1)
Figure 40. McBSP Transmit Timings
72
(n–4)
td(BCKXH–BDXV)
tdis(BCKXH–BDXHZ)
BDX
(XDATDLY=10b)
(n–3)
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
(n–2)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port timing (continued)
timing requirements for McBSP general-purpose I/O (see Figure 41)
MIN
tsu(BGPIO-COH)
th(COH-BGPIO)
Setup time, BGPIOx input mode before CLKOUT high†
Hold time, BGPIOx input mode after CLKOUT high†
MAX
UNIT
9
ns
0
ns
† BGPIOx refers to BCLKRx, BFSRx, BDRx, BCLKXx, or BFSXx when configured as a general-purpose input.
switching characteristics for McBSP general-purpose I/O (see Figure 41)
Delay time, CLKOUT high to BGPIOx output mode‡
td(COH-BGPIO)
‡ BGPIOx refers to BCLKRx, BFSRx, BCLKXx, BFSXx, or BDXx when configured as a general-purpose output.
tsu(BGPIO-COH)
MIN
MAX
–10
10
UNIT
ns
ADVANCE INFORMATION
PARAMETER
td(COH-BGPIO)
CLKOUT
th(COH-BGPIO)
BGPIOx Input
Mode†
BGPIOx Output
Mode‡
† BGPIOx refers to BCLKRx, BFSRx, BDRx, BCLKXx, or BFSXx when configured as a general-purpose input.
‡ BGPIOx refers to BCLKRx, BFSRx, BCLKXx, BFSXx, or BDXx when configured as a general-purpose output.
Figure 41. McBSP General-Purpose I/O Timings
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
73
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port timing (continued)
timing requirements for McBSP as SPI master or slave: [H=0.5tc(CO)] CLKSTP = 10b, CLKXP = 0†
(see Figure 42)
MASTER
MIN
tsu(BDRV-BCKXL)
th(BCKXL-BDRV)
Setup time, BDR valid before BCLKX low
tsu(BFXL-BCKXH)
Setup time, BFSX low before BCLKX high
SLAVE
MAX
MIN
MAX
UNIT
12
2 – 12H
ns
4
6 + 12H
ns
10
ns
32H
ns
Hold time, BDR valid after BCLKX low
tc(BCKX)
Cycle time, BCLKX
12H
† For all SPI slave modes, CLKG is programmed as 1/2 of the CPU clock by setting CLKSM = CLKGDV = 1.
switching characteristics for McBSP as SPI master or slave: [H=0.5tc(CO)] CLKSTP = 10b,
CLKXP = 0† (see Figure 42)
ADVANCE INFORMATION
MASTER‡
PARAMETER
MIN
th(BCKXL-BFXL)
td(BFXL-BCKXH)
Hold time, BFSX low after BCLKX low§
Delay time, BFSX low to BCLKX high¶
td(BCKXH-BDXV)
Delay time, BCLKX high to BDX valid
tdis(BCKXL-BDXHZ)
Disable time, BDX high impedance following last data bit from BCLKX
low
tdis(BFXH-BDXHZ)
Disable time, BDX high impedance following last data bit from BFSX
high
SLAVE
MAX
T–5
T+5
C–5
C+5
–2
12
C–2
C +10
MIN
MAX
UNIT
ns
ns
6H + 4
10H + 19
ns
ns
4H+ 4
8H + 17
ns
td(BFXL-BDXV)
Delay time, BFSX low to BDX valid
4H + 4
8H + 17
ns
† For all SPI slave modes, CLKG is programmed as 1/2 of the CPU clock by setting CLKSM = CLKGDV = 1.
‡ T = BCLKX period = (1 + CLKGDV) * 2H
C = BCLKX low pulse width = T/2 when CLKGDV is odd or zero and = (CLKGDV/2) * 2H when CLKGDV is even
§ FSRP = FSXP = 1. As a SPI master, BFSX is inverted to provide active-low slave-enable output. As a slave, the active-low signal input on BFSX
and BFSR is inverted before being used internally.
CLKXM = FSXM = 1, CLKRM = FSRM = 0 for master McBSP
CLKXM = CLKRM = FSXM = FSRM = 0 for slave McBSP
¶ BFSX should be low before the rising edge of clock to enable slave devices and then begin a SPI transfer at the rising edge of the master clock
(BCLKX).
tc(BCKX)
MSB
tsu(BFXL-BCKXH)
LSB
BCLKX
th(BCKXL-BFXL)
td(BFXL-BCKXH)
BFSX
tdis(BFXH-BDXHZ)
td(BFXL-BDXV)
td(BCKXH-BDXV)
tdis(BCKXL-BDXHZ)
BDX
Bit 0
Bit(n-1)
tsu(BDRV-BCKXL)
BDR
Bit 0
(n-2)
(n-3)
(n-4)
th(BCKXL-BDRV)
Bit(n-1)
(n-2)
(n-3)
Figure 42. McBSP Timing as SPI Master or Slave: CLKSTP = 10b, CLKXP = 0
74
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
(n-4)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port timing (continued)
timing requirements for McBSP as SPI master or slave: [H=0.5tc(CO)] CLKSTP = 11b, CLKXP = 0†
(see Figure 43)
MASTER
MIN
tsu(BDRV-BCKXH)
th(BCKXH-BDRV)
Setup time, BDR valid before BCLKX high
tsu(BFXL-BCKXH)
Setup time, BFSX low before BCLKX high
Hold time, BDR valid after BCLKX high
SLAVE
MAX
MIN
MAX
UNIT
12
2 – 12H
ns
4
5 +12H
ns
10
ns
32H
ns
tc(BCKX)
Cycle time, BCLKX
12H
† For all SPI slave modes, CLKG is programmed as 1/2 of the CPU clock by setting CLKSM = CLKGDV = 1.
MASTER‡
PARAMETER
MIN
MAX
C–5
C+5
T–5
T+5
SLAVE
MIN
MAX
UNIT
th(BCKXL-BFXL)
td(BFXL-BCKXH)
Hold time, BFSX low after BCLKX low§
Delay time, BFSX low to BCLKX high¶
td(BCKXL-BDXV)
Delay time, BCLKX low to BDX valid
–2
12
6H + 4
10H + 19
ns
tdis(BCKXL-BDXHZ)
Disable time, BDX high impedance following last data bit from BCLKX
low
–2
10
6H + 4
10H + 17
ns
ns
ns
td(BFXL-BDXV)
Delay time, BFSX low to BDX valid
D – 2 D +10
4H – 4
8H + 17
ns
† For all SPI slave modes, CLKG is programmed as 1/2 of the CPU clock by setting CLKSM = CLKGDV = 1.
‡ T = BCLKX period = (1 + CLKGDV) * 2H
C = BCLKX low pulse width = T/2 when CLKGDV is odd or zero and = (CLKGDV/2) * 2H when CLKGDV is even
D = BCLKX high pulse width = T/2 when CLKGDV is odd or zero and = (CLKGDV/2 + 1) * 2H when CLKGDV is even
§ FSRP = FSXP = 1. As a SPI master, BFSX is inverted to provide active-low slave-enable output. As a slave, the active-low signal input on BFSX
and BFSR is inverted before being used internally.
CLKXM = FSXM = 1, CLKRM = FSRM = 0 for master McBSP
CLKXM = CLKRM = FSXM = FSRM = 0 for slave McBSP
¶ BFSX should be low before the rising edge of clock to enable slave devices and then begin a SPI transfer at the rising edge of the master clock
(BCLKX).
tsu(BFXL-BCKXH)
LSB
tc(BCKX)
MSB
BCLKX
td(BFXL-BCKXH)
th(BCKXL-BFXL)
BFSX
tdis(BCKXL-BDXHZ)
BDX
td(BCKXL-BDXV)
td(BFXL-BDXV)
Bit 0
Bit(n-1)
tsu(BDRV-BCKXH)
BDR
Bit 0
(n-2)
(n-3)
(n-4)
th(BCKXH-BDRV)
Bit(n-1)
(n-2)
(n-3)
(n-4)
Figure 43. McBSP Timing as SPI Master or Slave: CLKSTP = 11b, CLKXP = 0
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
75
ADVANCE INFORMATION
switching characteristics for McBSP as SPI master or slave: [H=0.5tc(CO)] CLKSTP = 11b,
CLKXP = 0† (see Figure 43)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port timing (continued)
timing requirements for McBSP as SPI master or slave: [H=0.5tc(CO)] CLKSTP = 10b, CLKXP = 1†
(see Figure 44)
MASTER
MIN
tsu(BDRV-BCKXH)
th(BCKXH-BDRV)
Setup time, BDR valid before BCLKX high
tsu(BFXL-BCKXL)
Setup time, BFSX low before BCLKX low
Hold time, BDR valid after BCLKX high
SLAVE
MAX
MIN
MAX
UNIT
12
2 – 12H
ns
4
6 + 12H
ns
10
ns
32H
ns
tc(BCKX)
Cycle time, BCLKX
12H
† For all SPI slave modes, CLKG is programmed as 1/2 of the CPU clock by setting CLKSM = CLKGDV = 1.
ADVANCE INFORMATION
switching characteristics for McBSP as SPI master or slave: [H=0.5tc(CO)] CLKSTP = 10b,
CLKXP = 1†‡ (see Figure 44)
MASTER
PARAMETER
MIN
th(BCKXH-BFXL)
td(BFXL-BCKXL)
Hold time, BFSX low after BCLKX high§
Delay time, BFSX low to BCLKX low¶
td(BCKXL-BDXV)
Delay time, BCLKX low to BDX valid
tdis(BCKXH-BDXHZ)
Disable time, BDX high impedance following last data bit from BCLKX
high
tdis(BFXH-BDXHZ)
Disable time, BDX high impedance following last data bit from BFSX
high
SLAVE
MAX
T–5
T+5
D–5
D+5
–2
12
D–2
D +10
MIN
MAX
UNIT
ns
ns
6H + 4
10H + 19
ns
ns
4H + 4
8H + 17
ns
td(BFXL-BDXV)
Delay time, BFSX low to BDX valid
4H – 4
8H + 17
ns
† For all SPI slave modes, CLKG is programmed as 1/2 of the CPU clock by setting CLKSM = CLKGDV = 1.
‡ T = BCLKX period = (1 + CLKGDV) * 2H
D = BCLKX high pulse width = T/2 when CLKGDV is odd or zero and = (CLKGDV/2 + 1) * 2H when CLKGDV is even
§ FSRP = FSXP = 1. As a SPI master, BFSX is inverted to provide active-low slave-enable output. As a slave, the active-low signal input on BFSX
and BFSR is inverted before being used internally.
CLKXM = FSXM = 1, CLKRM = FSRM = 0 for master McBSP
CLKXM = CLKRM = FSXM = FSRM = 0 for slave McBSP
¶ BFSX should be low before the rising edge of clock to enable slave devices and then begin a SPI transfer at the rising edge of the master clock
(BCLKX).
LSB
tsu(BFXL-BCKXL)
tc(BCKX)
MSB
BCLKX
th(BCKXH-BFXL)
td(BFXL-BCKXL)
BFSX
td(BFXL-BDXV)
tdis(BFXH-BDXHZ)
tdis(BCKXH-BDXHZ)
BDX
td(BCKXL-BDXV)
Bit 0
Bit(n-1)
tsu(BDRV-BCKXH)
BDR
Bit 0
(n-2)
(n-3)
th(BCKXH-BDRV)
Bit(n-1)
(n-2)
(n-3)
Figure 44. McBSP Timing as SPI Master or Slave: CLKSTP = 10b, CLKXP = 1
76
POST OFFICE BOX 1443
(n-4)
• HOUSTON, TEXAS 77251–1443
(n-4)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
multichannel buffered serial port timing (continued)
timing requirements for McBSP as SPI master or slave: [H=0.5tc(CO)] CLKSTP = 11b, CLKXP = 1†
(see Figure 45)
MASTER
MIN
tsu(BDRV-BCKXL)
th(BCKXL-BDRV)
Setup time, BDR valid before BCLKX low
tsu(BFXL-BCKXL)
Setup time, BFSX low before BCLKX low
Hold time, BDR valid after BCLKX low
SLAVE
MAX
MIN
UNIT
MAX
12
2 – 12H
ns
4
6 + 12H
ns
10
ns
32H
ns
tc(BCKX)
Cycle time, BCLKX
12H
† For all SPI slave modes, CLKG is programmed as 1/2 of the CPU clock by setting CLKSM = CLKGDV = 1.
MASTER‡
PARAMETER
MIN
MAX
D–5
D+5
T–5
T+5
SLAVE
MIN
UNIT
MAX
th(BCKXH-BFXL)
td(BFXL-BCKXL)
Hold time, BFSX low after BCLKX high§
Delay time, BFSX low to BCLKX low¶
td(BCKXH-BDXV)
Delay time, BCLKX high to BDX valid
–2
12
6H + 4
10H + 19
ns
tdis(BCKXH-BDXHZ)
Disable time, BDX high impedance following last data bit from BCLKX
high
–2
10
6H + 4
10H + 17
ns
ns
ns
td(BFXL-BDXV)
Delay time, BFSX low to BDX valid
C – 2 C +10
4H – 4
8H + 17
ns
† For all SPI slave modes, CLKG is programmed as 1/2 of the CPU clock by setting CLKSM = CLKGDV = 1.
‡ T = BCLKX period = (1 + CLKGDV) * 2H
C = BCLKX low pulse width = T/2 when CLKGDV is odd or zero and = (CLKGDV/2) * 2H when CLKGDV is even
D = BCLKX high pulse width = T/2 when CLKGDV is odd or zero and = (CLKGDV/2 + 1) * 2H when CLKGDV is even
§ FSRP = FSXP = 1. As a SPI master, BFSX is inverted to provide active-low slave-enable output. As a slave, the active-low signal input on BFSX
and BFSR is inverted before being used internally.
CLKXM = FSXM = 1, CLKRM = FSRM = 0 for master McBSP
CLKXM = CLKRM = FSXM = FSRM = 0 for slave McBSP
¶ BFSX should be low before the rising edge of clock to enable slave devices and then begin a SPI transfer at the rising edge of the master clock
(BCLKX).
tsu(BFXL-BCKXL)
LSB
MSB
tc(BCKX)
BCLKX
th(BCKXH-BFXL)
td(BFXL-BCKXL)
BFSX
tdis(BCKXH-BDXHZ)
BDX
td(BCKXH-BDXV)
td(BFXL-BDXV)
Bit 0
Bit(n-1)
tsu(BDRV-BCKXL)
BDR
Bit 0
(n-2)
(n-3)
(n-4)
th(BCKXL-BDRV)
Bit(n-1)
(n-2)
(n-3)
(n-4)
Figure 45. McBSP Timing as SPI Master or Slave: CLKSTP = 11b, CLKXP = 1
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
77
ADVANCE INFORMATION
switching characteristics for McBSP as SPI master or slave: [H=0.5tc(CO)] CLKSTP = 11b,
CLKXP = 1†‡ (see Figure 45)
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 timing
switching characteristics over recommended operating conditions†‡§ [H = 0.5tc(CO)]
(see Figure 46 and Figure 53)
PARAMETER
td(DSL-HDD)
Delay time, DS low to HD driven§
MIN
MAX
UNIT
3
20
ns
Case 1a: Memory accesses when
DMAC is active in 16-bit mode and
tw(DSH) < 18H
Case 1b: Memory accesses when
DMAC is active in 16-bit mode and
tw(DSH) ≥ 18H
ADVANCE INFORMATION
td(DSL-HDV1)
d(DSL HDV1)
Delay time, DS low to HD valid for first
byte of an HPI read
18H+20 – tw(DSH)
20
Case 1c: Memory access when
DMAC is active in 32-bit mode and
tw(DSH) < 26H
Case 1d: Memory access when
DMAC is active in 32-bit mode and
tw(DSH) ≥ 26H
26H+20 – tw(DSH)
ns
20
Case 2a: Memory accesses when
DMAC is inactive and tw(DSH) < 10H
10H+20 – tw(DSH)
Case 2b: Memory accesses when
DMAC is inactive and tw(DSH) ≥ 10H
20
Case 3: Register accesses
td(DSL-HDV2)
td(DSH-HYH)
Multiplexed reads with autoincrement. Prefetch completed.
Delay time, DS high to HRDY high§
(writes and autoincrement reads)
20
No DMA channel active
12H+5
One or more 16-bit DMA channels
active
18H+5
One or more 32-bit DMA channels
active
26H+5
Writes to DSPINT and HINT
4H + 5
tv(HYH-HDV)
th(DSH-HDV)R
Valid time, HD valid after HRDY high
td(COH-HYH)
td(DSL-HYL)
Delay time, CLKOUT rising edge to HRDY high
Delay time, HDS low to HRDY low§
Hold time, HD valid after DS rising edge, read§
td(COH–HTX)
Delay time, CLKOUT rising edge to HINT change
† HAD stands for HCNTL0, HCNTL1, and HR/W.
‡ HDS refers to either HDS1 or HDS2.
§ DS refers to the logical OR of HCS and HDS.
78
20
3
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
0
ns
ns
7
ns
10
ns
5
ns
12
ns
5
ns
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 timing (continued)
timing requirements [H = 0.5tc(CO)] (see Figure 46 and Figure 53)
MAX
UNIT
Setup time, HAD valid before DS falling edge†‡
Hold time, HAD valid after DS falling edge†‡
5
ns
5
ns
tsu(HBV-HSL)
th(HSL-HBV)
Setup time, HAD valid before HAS falling edge†
Hold time, HAD valid after HAS falling edge†
5
ns
5
ns
tsu(HAV-DSH)
tsu(HAV-DSL)
Setup time, address valid before DS rising edge (nonmultiplexed write)
th(DSH-HAV)
tsu(HSL-DSL)
th(HSL-DSL)
tw(DSL)
tw(DSH)
Setup time, address valid before DS falling edge (nonmultiplexed mode)‡
Hold time, address valid after DS rising edge (nonmultiplexed mode)‡
ns
1
ns
5
ns
2
ns
Pulse duration, DS low‡
Pulse duration, DS high‡
30
ns
10
ns
Nonmulti lexed or multi
Nonmultiplexed
multiplexed
lexed mode
(no increment) with no DMA activity.
Reads
12H
ns
Writes
14H
ns
Nonmulti lexed or multi
Nonmultiplexed
multiplexed
lexed mode
(no increment) with 16-bit DMA activity.
Reads
18H
ns
Writes
20H
Nonmulti lexed or multi
Nonmultiplexed
multiplexed
lexed mode
(no increment) with 32-bit DMA activity.
Reads
26H
Writes
28H
Multiplexed (autoincrement) with no DMA
activity.
12H
ns
Multiplexed (autoincrement) with 16-bit DMA
activity.
18H
ns
Multiplexed (autoincrement) with 32-bit DMA
activity.
26H
ns
8H
ns
10
ns
Hold time, HD valid after DS rising edge, write‡
0
ns
Setup time, SELA/B valid before DS falling edge‡
Hold time, SELA/B valid after DS rising edge‡
5
ns
0
ns
Cycle time, DS rising edge to next DS
rising edge‡
(
(In
au
autoincrement
o ce e
mode,
ode, WRITE
timings are the same as READ
timings.)
tsu(HDV-DSH)
th(DSH-HDV)W
ns
Setup time, HAS low before DS falling edge‡
Hold time, HAS low after DS falling edge‡
Cycle time, DS rising edge to next DS
rising
g edge
g ‡
tc(DSH-DSH)
5
–4H – 5
Cycle time, DS rising edge to next DS rising edge writes to DSPINT and HINT
Setup time, HD valid before DS rising edge‡
tsu(SELV-DSL)
th(DSH-SELV)
† HAD stands for HCNTL0, HCNTL1, and HR/W.
‡ DS refers to the logical OR of HCS and HDS.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
ADVANCE INFORMATION
MIN
tsu(HBV-DSL)
th(DSL-HBV)
ns
79
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 timing (continued)
HCS
tsu(HSL–DSL)
th(HSL–DSL)
HAS
tc(DSH–DSH)
tsu(HBV–HSL)
HDS
tw(DSH)
th(HSL–HBV)
tw(DSL)
HR/W
HCNTL[1:0]
01
01
td(DSL–HDV1)
th(DSH–HDV)R
ADVANCE INFORMATION
td(DSL–HDV2)
Data 1
HD[15:0]
PF Data
td(DSL–HDD)
HRDY
tv(HYH–HDV)
td(DSL–HYL)
td(DSH–HYH)
Figure 46. Multiplexed Read Timings Using HAS
HCS
tsu(HBV–DSL)
tc(DSH–DSH)
HDS
th(DSL–HBV)
tw(DSH)
tw(DSL)
HR/W
01
HCNTL[1:0]
01
td(DSL–HDV1)
th(DSH–HDV)R
td(DSL–HDV2)
Data 1
HD[15:0]
PF Data
td(DSL–HDD)
HRDY
tv(HYH–HDV)
td(DSH–HYH)
td(DSL–HYL)
Figure 47. Multiplexed Read Timings With HAS Held High
80
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
Á
Á
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 timing (continued)
HCS
tsu(HBV–HSL)
th(HSL–DSL)
HAS
tsu(HSL–DSL)
HR/W
th(HSL–HBV)
HCNTL[1:0]
01
01
ADVANCE INFORMATION
tc(DSH–DSH)
HDS
tw(DSH)
tw(DSL)
tsu(HDV–DSH)
HD[15:0]
Data 1
Data n
th(DSH–HDV)W
HRDY
td(DSL–HYL)
td(DSH–HYH)
Figure 48. Multiplexed Write Timings Using HAS
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
81
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 timing (continued)
HCS
tc(DSH–DSH)
tw(DSH)
HDS
tw(DSL)
tsu(HBV–DSL)
HR/W
th(DSL–HBV)
HCNTL[1:0]
01
01
tsu(HDV–DSH)
th(DSH–HDV)W
ADVANCE INFORMATION
HD[15:0]
Data 1
Data n
td(DSL–HYL)
HRDY
td(DSH–HYH)
Figure 49. Multiplexed Write Timings With HAS Held High
82
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 timing (continued)
HCS
tw(DSH)
tc(DSH–DSH)
HDS
tw(DSL)
tsu(HBV–DSL)
th(DSL–HBV)
HR/W
tsu(HAV–DSL)
th(DSH–HAV)
Valid Address
Valid Address
td(DSL–HDV1)
HD[15:0]
Data
ADVANCE INFORMATION
HA[17:0]
th(DSH–HDV)R
Valid Data
td(DSH–HYH)
tv(HYH–HDV)
HRDY
td(DSL–HYL)
Figure 50. Nonmultiplexed Read Timings
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
83
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 timing (continued)
HCS
tw(DSH)
tc(DSH–DSH)
HDS
tsu(HBV–DSL)
tw(DSL)
th(DSL–HBV)
HR/W
tsu(HAV–DSH)
th(DSH–HAV)
ADVANCE INFORMATION
HA[17:0]
Valid Address
Valid Address
tsu(HDV–DSH)
HD[15:0]
Data Valid
Data Valid
td(DSH–HYH)
HRDY
td(DSL–HYL)
Figure 51. Nonmultiplexed Write Timings
84
th(DSH–HDV)W
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
HPI16 timing (continued)
HRDY
td(COH–HYH)
CLKOUT
td(COH–HTX)
HINT
Figure 52. HRDY and HINT Relative to CLKOUT
HCS
tsu(DSL–SELV)
ADVANCE INFORMATION
th(DSH–SELV)
SELA/B
HDS
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁÁ
Figure 53. SELA/B Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
85
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
MECHANICAL DATA
PGE (S-PQFP-G144)
PLASTIC QUAD FLATPACK
108
73
109
72
0,27
0,17
0,08 M
ADVANCE INFORMATION
0,50
144
0,13 NOM
37
1
36
Gage Plane
17,50 TYP
20,20 SQ
19,80
22,20
SQ
21,80
0,25
0,05 MIN
0°–ā7°
0,75
0,45
1,45
1,35
Seating Plane
0,08
1,60 MAX
4040147 / C 10/96
NOTES: A. All linear dimensions are in millimeters.
B. This drawing is subject to change without notice.
C. Falls within JEDEC MO-136
Thermal Resistance Characteristics
86
PARAMETER
°C / W
RΘJA
56
RΘJC
5
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320VC5421
DIGITAL SIGNAL PROCESSOR
SPRS098 – DECEMBER 1999
MECHANICAL DATA
GGU (S-PBGA-N144)
PLASTIC BALL GRID ARRAY PACKAGE
12,10
SQ
11,90
A
B
C
D
E
F
G
H
J
K
L
M
N
0,80
0,80
0,95
0,85
0,12
0,08
1,40 MAX
0,55
0,45
0,08 M
0,45
0,35
0,10
4073221/A 11/96
NOTES: A. All linear dimensions are in millimeters.
B. This drawing is subject to change without notice.
Thermal Resistance Characteristics
PARAMETER
°C / W
RΘJA
38
RΘJC
5
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
87
ADVANCE INFORMATION
13 12 11 10 9 8 7 6 5 4 3 2 1
IMPORTANT NOTICE
Texas Instruments and its subsidiaries (TI) reserve the right to make changes to their products or to discontinue
any product or service without notice, and advise customers to obtain the latest version of relevant information
to verify, before placing orders, that information being relied on is current and complete. All products are sold
subject to the terms and conditions of sale supplied at the time of order acknowledgement, including those
pertaining to warranty, patent infringement, and limitation of liability.
TI warrants performance of its semiconductor products to the specifications applicable at the time of sale in
accordance with TI’s standard warranty. Testing and other quality control techniques are utilized to the extent
TI deems necessary to support this warranty. Specific testing of all parameters of each device is not necessarily
performed, except those mandated by government requirements.
CERTAIN APPLICATIONS USING SEMICONDUCTOR PRODUCTS MAY INVOLVE POTENTIAL RISKS OF
DEATH, PERSONAL INJURY, OR SEVERE PROPERTY OR ENVIRONMENTAL DAMAGE (“CRITICAL
APPLICATIONS”). TI SEMICONDUCTOR PRODUCTS ARE NOT DESIGNED, AUTHORIZED, OR
WARRANTED TO BE SUITABLE FOR USE IN LIFE-SUPPORT DEVICES OR SYSTEMS OR OTHER
CRITICAL APPLICATIONS. INCLUSION OF TI PRODUCTS IN SUCH APPLICATIONS IS UNDERSTOOD TO
BE FULLY AT THE CUSTOMER’S RISK.
In order to minimize risks associated with the customer’s applications, adequate design and operating
safeguards must be provided by the customer to minimize inherent or procedural hazards.
TI assumes no liability for applications assistance or customer product design. TI does not warrant or represent
that any license, either express or implied, is granted under any patent right, copyright, mask work right, or other
intellectual property right of TI covering or relating to any combination, machine, or process in which such
semiconductor products or services might be or are used. TI’s publication of information regarding any third
party’s products or services does not constitute TI’s approval, warranty or endorsement thereof.
Copyright  1999, Texas Instruments Incorporated