TI TMS320C80GGP60

TMS320C80
Digital Signal Processor
Data Sheet
1997
Digital Signal Processing Solutions
Printed in U.S.A., October 1997
SPRS023B
Book
Type
Data Sheet
TMS320C80 DSP
1997
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
D
D
D
D
D
D
D
D
D
D
D
D
Single-Chip Parallel Multiple
Instruction / Multiple Data (MIMD) DSP
More Than Two Billion RISC-Equivalent
Operations per Second
Master Processor (MP)
– 32-Bit Reduced Instruction Set
Computing (RISC) Processor
– IEEE-754 Floating-Point Capability
– 4K-Byte Instruction Cache
– 4K-Byte Data Cache
Four Parallel Processors (PP)
– 32-Bit Advanced DSPs
– 64-Bit Opcode Provides Many Parallel
Operations per Cycle
– 2K-Byte Instruction Cache and 8K-Byte
Data RAM per PP
Transfer Controller ( TC)
– 64-Bit Data Transfers
– Up to 480M-Byte / s Transfer Rate
– 32-Bit Addressing
– Direct DRAM / VRAM Interface With
Dynamic Bus Sizing
– Intelligent Queuing and Cycle
Prioritization
Video Controller ( VC)
– Provides Video Timing and VRAM
Control
– Dual-Frame Timers for Two Simultaneous
Image-Capture and / or Display Systems
Big- or Little-Endian Operation
50K-Byte On-Chip RAM
4G-Byte Address Space
16.6-ns Cycle Time
3.3-V Operation
IEEE Standard 1149.1† Test Access Port
(JTAG)
GF PACKAGE
( BOTTOM VIEW )
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
AR
AP
AN
AM
AL
AK
AH
AF
AJ
AG
AD
AB
Y
V
T
P
M
K
H
F
D
B
AE
AC
AA
W
U
R
N
L
J
G
E
C
A
GGP PACKAGE
( BOTTOM VIEW )
26 24 22 20 18 16 14 12 10 8
6
4
2
5
25 23 21 19 17 15 13 11 9
7
3
1
A
B
C
D
E
F
G
H
J
K
L
M
N
P
R
T
U
V
W
Y
AA
AB
AC
AD
AE
AF
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.
† IEEE Standard 1149.1–1990, IEEE Standard Test Access Port and Boundary-Scan Architecture
Copyright  1997, Texas Instruments Incorporated
PRODUCTION DATA information is current as of publication date.
Products conform to specifications per the terms of Texas Instruments
standard warranty. Production processing does not necessarily include
testing of all parameters.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
description
The TMS320C80 is a single chip, MIMD parallel processor capable of performing over two billion operations
per second. It consists of a 32-bit RISC master processor with a 120-MFLOP IEEE floating-point unit, four 32-bit
parallel processing digital signal processors (DSPs), a transfer controller with up to 480M-byte/s off-chip
transfer rate, and a video controller. All the processors are coupled tightly through an on-chip crossbar that
provides shared access to on-chip RAM. This performance and programmability make the ’C80 ideally suited
for video, imaging, and high-speed telecommunications applications.
2
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GF Terminal Assignments – Numerical Listing
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
A5
CT1
C21
HSYNC0
L5
VDD
HACK
C23
VDD
W
E33
A7
E35
TCK
L31
VSS
VSS
C25
DBEN
F2
L33
TRST
C27
L35
XPT1
C29
VSS
CAREA0
F4
A13
VSS
CAS / DQM7
VDD
VSS
M2
A15
CAS / DQM5
C31
CBLNK0 / VBLNK0
F10
VDD
VSS
VDD
VSS
A17
D2
RETRY
F12
D4
F14
A21
RAS
D6
VDD
VSS
VDD
PS0
M32
A19
VDD
VSS
DSF
D8
AS0
F18
VSS
CT2
N1
A23
A25
VSS
SCLK1
D10
UTIME
F20
N5
D12
F22
N31
VSS
VSS
D14
TMS
D16
REQ0
F26
VDD
VSS
N33
A31
VDD
EINT1
VSS
RESET
VDD
VSS
N35
B2
No Connect
D18
F28
VDD
A4
BS1
D20
VDD
VSS
P2
B4
VSS
CAS / DQM0
P4
A9
B6
VDD
PS1
D22
FCLK1
F34
TDO
D24
G1
P34
XPT0
B10
REQ1
D26
VSS
CAREA1
VDD
VDD
P32
B8
G3
A2
R1
B12
D28
SCLK0
G5
A1
R3
VSS
VDD
B14
VDD
CAS / DQM6
D30
G31
EINT2
R5
B16
CAS / DQM3
D32
VSS
VDD
G33
CBLNK1 / VBLNK1
R31
B18
D34
VSYNC0
G35
E1
AS1
H2
VDD
STATUS0
R33
B20
VDD
CAS / DQM1
R35
VDD
VSS
B22
TRG / CAS
E3
FAULT
H4
A3
T2
A5
B24
E5
CSYNC1 / HBLNK1
T4
A13
E7
VSS
STATUS2
H32
B26
VDD
DDIN
H34
TDI
T32
D62
B28
FCLK0
E9
READY
J1
STATUS1
T34
EMU0
B30
VDD
CSYNC0 / HBLNK0
E11
BS0
J3
U1
E13
J5
U3
VDD
A10
E15
PS3
E17
CAS / DQM4
J33
VDD
VSS
U5
C5
VSS
STATUS3
VSS
HREQ
VSS
VDD
U31
FF1
C7
AS2
E19
RL
J35
EMU1
U33
D61
C9
E21
STATUS5
K2
STATUS4
U35
C11
VSS
CT0
E23
K4
A6
V2
C13
PS2
E25
VSS
CLKOUT
VDD
VDD
K32
VSYNC1
V4
C15
VDD
CLKIN
E27
LINT4
K34
HSYNC1
V32
C17
E29
EINT3
L1
A0
V34
C19
CAS / DQM2
E31
VSS
L3
A7
W1
A9
A11
A27
A29
B32
C3
POST OFFICE BOX 1443
F8
F16
F24
F32
J31
• HOUSTON, TEXAS 77251–1443
M4
M34
N3
VSS
VDD
VDD
A8
VDD
VDD
VSS
VSS
VDD
A11
3
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GF Terminal Assignments – Numerical Listing (Continued)
NO.
NO.
TERMINAL
NAME
TERMINAL
NO.
TERMINAL
NAME
NO.
NAME
W3
A18
AG1
A16
AL17
D20
AN29
D35
W5
AG3
VSS
VDD
AL19
D21
AN31
D45
W31
VSS
VSS
AL21
D24
AN33
W33
D59
AG31
AL23
D63
AG33
AL25
VSS
D29
AP4
W35
VDD
VSS
VDD
A27
Y2
A12
AG35
D57
AL27
D32
AP8
VDD
D5
Y4
A19
AH2
A20
AL29
D38
AP10
D8
Y32
XPT2
AH4
A30
AL31
AP12
Y34
D56
AH32
D44
AL33
VSS
D48
AP14
VDD
D13
AA1
VSS
VDD
AH34
D54
AL35
D53
AP16
D17
AJ1
AM2
A24
AP18
VDD
VDD
AJ3
VDD
A31
AM4
AP20
AM6
AP22
D34
AJ31
AM8
D2
AP24
AA35
VDD
VSS
VSS
VSS
VDD
VSS
VDD
D26
AJ33
D42
AM10
D6
AP26
VDD
D39
AB2
A14
AJ35
AM12
D41
A21
AK2
AM14
VSS
D14
AP28
AB4
VDD
VDD
AP30
AB32
D55
AK4
AM16
D19
AP32
VDD
D47
AB34
D60
AK8
VSS
VDD
AM18
D0
VDD
A22
AK10
VSS
VDD
AM20
VSS
D23
AR5
AC1
AR7
AM22
D25
AR9
VDD
D7
VSS
VSS
AK14
AM26
VSS
D31
AR11
AK16
VSS
VDD
AM24
AC31
AR13
VSS
D11
AC33
D52
AK18
FF2
AM28
D33
AR15
D15
AC35
VDD
VDD
AK20
VSS
D27
AM30
AR17
AM32
VSS
VDD
AR19
VSS
VDD
VSS
VSS
AK24
VDD
VSS
AM34
D50
AR21
D30
AN5
A29
AR23
D36
AK28
VDD
VSS
AN7
D1
AR25
AE1
VDD
A15
AN9
AR27
AE3
A26
AK34
AN11
AE5
AL1
AN13
D12
AR31
AE31
VSS
VSS
VDD
A23
VSS
D9
VSS
D40
AL3
A25
AN15
AE33
D51
AL5
AN17
AE35
D58
AL7
VSS
D3
VDD
D18
AN19
D22
AF2
A17
AL9
D4
AN21
AF4
A28
AL11
D10
AN23
VDD
D28
AF32
D46
AL13
AN25
D37
AF34
D49
AL15
VSS
D16
AN27
VSS
AA3
AA5
AA31
AA33
AC3
AC5
AD2
AD4
AD32
AD34
4
TERMINAL
NAME
AG5
AJ5
AK12
AK22
AK26
AK32
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
AP6
AR29
VDD
D43
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GF Terminal Assignments – Alphabetical Listing
TERMINAL
NAME
NO.
TERMINAL
NAME
TERMINAL
NAME
NO.
NO.
TERMINAL
NAME
NO.
A0
L1
CAS/DQM0
D20
D22
AN19
D61
U33
A1
G5
CAS/DQM1
B20
D23
AM20
D62
T32
A2
G3
CAS/DQM2
C19
D24
AL21
D63
W35
A3
H4
CAS/DQM3
B16
D25
AM22
DBEN
C25
A4
P2
CAS/DQM4
E17
D26
AP20
DDIN
B26
A5
T2
CAS/DQM5
A15
D27
AK22
DSF
A23
A6
K4
CAS/DQM6
B14
D28
AN23
EINT1
A31
A7
L3
CAS/DQM7
A13
D29
AL25
EINT2
G31
A8
N3
CBLNK0/VBLNK0
C31
D30
AR21
EINT3
E29
A9
P4
CBLNK1/VBLNK1
G33
D31
AM26
EMU0
T34
A10
U3
CLKIN
C17
D32
AL27
EMU1
J35
A11
W1
CLKOUT
E25
D33
AM28
FAULT
E3
A12
Y2
CSYNC0/HBLNK0
B32
D34
AP22
FCLK0
B28
A13
T4
CSYNC1/HBLNK1
H32
D35
AN29
FCLK1
D22
A14
AB2
CT0
C11
D36
AR23
FF1
U31
A15
AE1
CT1
A5
D37
AN25
FF2
AK18
A16
AG1
CT2
F18
D38
AL29
HACK
A9
A17
AF2
D0
AR5
D39
AP26
HREQ
E15
A18
W3
D1
AN7
D40
AR27
HSYNC0
E33
A19
Y4
D2
AM8
D41
AP28
HSYNC1
K34
A20
AH2
D3
AL7
D42
AJ33
LINT4
E27
A21
AB4
D4
AL9
D43
AR31
PS0
F14
A22
AC3
D5
AP8
D44
AH32
PS1
B8
A23
AL1
D6
AM10
D45
AN31
PS2
C13
A24
AM2
D7
AR9
D46
AF32
PS3
U5
A25
AL3
D8
AP10
D47
AP32
RAS
A21
A26
AE3
D9
AN11
D48
AL33
READY
E9
A27
AP4
D10
AL11
D49
AF34
REQ0
D16
A28
AF4
D11
AR13
D50
AM34
REQ1
B10
A29
AN5
D12
AN13
D51
AE33
RESET
D14
A30
AH4
D13
AP14
D52
AC33
RETRY
D2
A31
AJ3
D14
AM14
D53
AL35
RL
E19
AS0
D8
D15
AR15
D54
AH34
SCLK0
D28
AS1
E1
D16
AL15
D55
AB32
SCLK1
A27
AS2
C7
D17
AP16
D56
Y34
STATUS0
H2
BS0
E11
D18
AN17
D57
AG35
STATUS1
J1
BS1
B4
D19
AM16
D58
AE35
STATUS2
E7
CAREA0
C29
D20
AL17
D59
W33
STATUS3
C5
CAREA1
D26
D21
AL19
D60
AB34
STATUS4
K2
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
5
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GF Terminal Assignments – Alphabetical Listing (Continued)
TERMINAL
NAME
NO.
STATUS5
E21
TCK
E35
TDI
H34
TDO
P32
TMS
N33
TRG/CAS
B22
TRST
L33
UTIME
D10
VDD
VDD
A17
A7
VDD
VDD
A29
VDD
VDD
B12
VDD
VDD
B24
VDD
VDD
C15
VDD
VDD
B6
B18
B30
C21
D4
D32
VDD
VDD
F2
VDD
VDD
F12
VDD
VDD
F24
VDD
VDD
F34
VDD
VDD
G35
VDD
VDD
J31
VDD
VDD
M34
VDD
VDD
N35
VDD
6
F8
F20
F28
G1
J5
M2
TERMINAL
NAME
NO.
VDD
VDD
R31
VDD
VDD
U1
VDD
VDD
R33
U35
V2
V34
VDD
VDD
AA3
VDD
VDD
AA31
VDD
VDD
AC1
VDD
VDD
AA5
AA33
AC35
AD2
AD34
VDD
VDD
AG5
AG31
VDD
VDD
AJ35
AJ1
VDD
VDD
AK2
VDD
VDD
AK12
VDD
VDD
AK24
VDD
VDD
AK34
VDD
VDD
AM32
VDD
VDD
AN21
VDD
VDD
AP6
AK8
AK16
AK28
AM4
AN15
AN33
AP12
VDD
VDD
AP18
VDD
VDD
AP30
R3
R5
VDD
N1
TERMINAL
NAME
NO.
VDD
VSS
AR29
VSS
VSS
A19
VSS
VSS
C3
VSS
VSS
C27
VSS
VSS
D12
VSS
VSS
D24
VSS
VSS
E5
VSS
VSS
VSS
VSS
A11
A25
C9
D6
D18
D30
E13
E23
E31
F4
F10
VSS
VSS
F16
VSS
VSS
F26
VSS
VSS
J3
F22
F32
J33
VSS
VSS
L31
VSS
VSS
M32
VSS
VSS
VSS
VSS
L5
M4
N5
N31
R1
VSS
VSS
AC31
VSS
VSS
AD32
VSS
VSS
AE31
VSS
VSS
AG33
VSS
VSS
AJ31
VSS
VSS
AK10
VSS
VSS
AK20
VSS
VSS
AK32
VSS
VSS
AL13
VSS
VSS
AL31
VSS
VSS
AM12
VSS
VSS
AM24
VSS
VSS
VSS
VSS
AC5
AD4
AE5
AG3
AJ5
AK4
AK14
AK26
AL5
AL23
AM6
AM18
AM30
AN9
AN27
AR11
AR17
V4
VSYNC1
K32
V32
W
C23
W5
XPT0
P34
W31
XPT1
L35
XPT2
Y32
VSS
VSS
AR19
VSS
AA1
• HOUSTON, TEXAS 77251–1443
AA35
R35
AR7
POST OFFICE BOX 1443
VSS
VSS
VSS
VSYNC0
VSS
VSS
AP24
TERMINAL
NAME
NO.
AR25
D34
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GGP Terminal Assignments – Numerical Listing
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
A1
No Connect
B1
No Connect
C1
No Connect
D1
A0
A2
No Connect
B2
No Connect
C2
No Connect
D2
A3
No Connect
B3
No Connect
C3
No Connect
D3
VDD
STATUS4
A4
STATUS1
B4
STATUS2
C4
STATUS3
AS1
B5
AS2
C5
VSS
STATUS0
D4
A5
D5
A6
RETRY
B6
READY
C6
FAULT
D6
VDD
AS0
A7
CT1
B7
BS0
C7
BS1
D7
UTIME
A8
PS0
B8
PS1
C8
PS2
D8
CT0
A9
VDD
VSS
B9
HACK
C9
HREQ
D9
RESET
B10
C10
REQ1
B11
C11
VSS
VDD
D10
D11
REQ0
B12
CAS / DQM7
C12
CLKIN
D12
A13
VSS
VDD
No Connect
VSS
VDD
VDD
CAS / DQM6
No Connect
B14
C14
VSS
CAS / DQM5
D13
A14
No Connect
VSS
C13
A15
B15
CAS / DQM4
C15
CAS / DQM3
D15
A16
VDD
CAS / DQM2
B16
CAS / DQM1
D16
CAS / DQM0
B17
VSS
RL
C16
A17
C17
D17
VDD
VSS
A18
VSS
FCLK1
B18
RAS
C18
VSS
VDD
D18
TRG / CAS
B19
W
D19
STATUS5
B20
VDD
DSF
C19
VDD
VDD
C20
D20
DBEN
B21
DDIN
C21
VSS
CLKOUT
D21
CAREA1
VSS
VSS
B22
SCLK1
C22
FCLK0
B23
SCLK0
C23
VDD
VDD
D22
A23
D23
CAREA0
A24
No Connect
B24
No Connect
C24
No Connect
D24
LINT4
A25
No Connect
B25
No Connect
C25
No Connect
D25
EINT3
A26
No Connect
B26
No Connect
C26
No Connect
D26
EINT2
A10
A11
A12
A19
A20
A21
A22
B13
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
D14
VDD
CT2
7
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GGP Terminal Assignments – Numerical Listing (Continued)
NO.
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
E1
A3
J23
HSYNC0
P1
No Connect
V23
D52
E2
A2
J24
TRST
P2
No Connect
V24
E3
J25
TCK
P3
J26
TMS
P4
VSS
VSS
V25
E4
VSS
A1
VSS
VSS
V26
D53
E23
EINT1
K1
A12
P23
W1
A24
E24
CBLNK1 / VBLNK1
K2
P24
W2
E25
CBLNK0 / VBLNK0
K3
VDD
A11
VSS
D59
W3
VSS
VSS
E26
VSS
A4
K4
P26
K23
VSS
TDI
VDD
No Connect
W23
VDD
VDD
K24
TDO
R2
VSS
A17
VSS
D49
W24
D50
K25
EMU1
R3
A18
W25
D51
VDD
VSS
K26
XPT0
R4
L1
R23
Y1
VDD
A25
F24
CSYNC1 / HBLNK1
L2
VDD
A13
VSS
XPT2
W26
F23
R24
D57
Y2
A26
F25
VDD
CSYNC0 / HBLNK0
L3
R25
VDD
D58
Y3
A27
L4
VSS
VSS
Y4
VSS
VSS
L23
XPT1
T1
A19
Y23
VDD
VDD
L24
T2
D48
A5
L25
T3
VDD
VDD
Y24
G3
VSS
VSS
Y25
G4
L26
EMU0
T4
A20
Y26
VSS
VSS
G23
VSS
VSYNC1
M1
T23
VSYNC0
M2
T24
VDD
VDD
AA1
G24
VDD
A15
AA2
VDD
VDD
G25
VSS
VSS
M3
PS3
T25
D56
AA3
A28
M4
A14
T26
A8
M23
U1
AA23
VSS
D45
H2
M24
D46
M25
D62
U3
VSS
A21
AA24
H3
VDD
A7
VDD
D63
VSS
VSS
AA4
H1
AA25
D47
H4
A6
M26
D61
U4
AA26
H23
HSYNC1
N1
No Connect
U23
VDD
VDD
AB1
VDD
VSS
H24
VDD
VDD
N2
A16
U24
D54
AB2
A29
N3
U25
A30
N4
U26
VSS
D55
AB3
VDD
A10
VDD
VDD
AB4
V1
A22
AB23
N24
V2
A23
AB24
VSS
VDD
D44
J3
VDD
A9
VSS
D60
N25
No Connect
V3
VSS
N26
No Connect
V4
VDD
VDD
AB25
J4
F1
F2
F3
F4
F26
G1
G2
G26
H25
H26
J1
J2
8
TERMINAL
NAME
N23
POST OFFICE BOX 1443
P25
R1
R26
U2
• HOUSTON, TEXAS 77251–1443
W4
AB26
VSS
VSS
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GGP Terminal Assignments – Numerical Listing (Continued)
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
NO.
TERMINAL
NAME
AC1
A31
AD1
No Connect
AE1
No Connect
AF1
No Connect
AC2
AD2
No Connect
AE2
No Connect
AF2
No Connect
AC3
VDD
VDD
AD3
No Connect
AE3
No Connect
AF3
No Connect
AC4
D0
AD4
D1
AF4
D2
D3
AD5
AE5
D4
AF5
AC6
VSS
VDD
D9
AD6
VSS
VDD
D5
AE4
AC5
AE6
D6
AF6
VSS
D7
VDD
D10
AE7
D8
AF7
AE8
D11
AF8
AC9
D12
AD9
D13
AF9
AD10
AE10
D15
AF10
AC11
VDD
D16
VSS
VDD
VSS
AE9
AC10
AE11
AF11
AC12
D18
AD12
D19
AE12
VSS
VDD
AC13
D20
AD13
AE13
VDD
No Connect
AD14
AE14
AF14
No Connect
AC15
VDD
D24
VSS
No Connect
AF13
AC14
VSS
D21
AE15
D23
AF15
D22
AC16
D27
AD16
VDD
D26
AE16
D25
AF16
AC17
VSS
D31
AD17
VSS
D30
AE17
D28
AF17
VSS
VDD
AE18
AF18
D29
AD19
VDD
D34
AE19
AF19
AC20
VDD
D35
VDD
D32
AE20
D33
AF20
VSS
VSS
AC21
D37
AD21
AE21
D36
AF21
AC22
D40
AD22
VSS
VDD
AE22
D39
AF22
AC23
D42
AD23
D41
AE23
AF23
AC24
D43
AD24
No Connect
AE24
VSS
No Connect
AF24
VSS
No Connect
AC25
VDD
VDD
AD25
No Connect
AE25
No Connect
AF25
No Connect
AD26
No Connect
AE26
No Connect
AF26
No Connect
AC7
AC8
AC18
AC19
AC26
AD7
AD8
AD11
AD15
AD18
AD20
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
AF12
VSS
VDD
D14
VSS
D17
VDD
D38
9
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GGP Terminal Assignments – Alphabetical Listing
TERMINAL
NAME
10
NO.
TERMINAL
NAME
TERMINAL
NAME
NO.
NO.
TERMINAL
NAME
NO.
A0
D1
CAS / DQM0
A17
D22
AF15
D61
M26
A1
E4
CAS / DQM1
C16
D23
AE15
D62
M25
A2
E2
CAS / DQM2
A16
D24
AC15
D63
M24
A3
E1
CAS / DQM3
C15
D25
AE16
DBEN
D20
A4
F1
CAS / DQM4
B15
D26
AD16
DDIN
B21
A5
G3
CAS / DQM5
C14
D27
AC16
DSF
B20
A6
H4
CAS / DQM6
D13
D28
AE17
EINT1
E23
A7
H3
CAS / DQM7
B12
D29
AF18
EINT2
D26
A8
H1
CBLNK0 / VBLNK0
E25
D30
AD18
EINT3
D25
A9
J3
CBLNK1 / VBLNK1
E24
D31
AC18
EMU0
L26
A10
J1
CLKIN
C12
D32
AE19
EMU1
K25
A11
K3
CLKOUT
C21
D33
AE20
FAULT
C6
A12
K1
CSYNC0 / HBLNK0
F26
D34
AD20
FCLK0
D22
A13
L2
CSYNC1 / HBLNK1
F24
D35
AC20
FCLK1
A19
A14
M4
CT0
D8
D36
AE21
HACK
B9
A15
M2
CT1
A7
D37
AC21
HREQ
C9
A16
N2
CT2
D15
D38
AF22
HSYNC0
J23
A17
R2
D0
AC4
D39
AE22
HSYNC1
H23
A18
R3
D1
AE4
D40
AC22
LINT4
D24
A19
T1
D2
AF4
D41
AD23
PS0
A8
A20
T4
D3
AC5
D42
AC23
PS1
B8
A21
U3
D4
AE5
D43
AC24
PS2
C8
A22
V1
D5
AD6
D44
AB24
PS3
M3
A23
V2
D6
AE6
D45
AA23
RAS
B18
A24
W1
D7
AF6
D46
AA24
READY
B6
A25
Y1
D8
AE7
D47
AA25
REQ0
D11
A26
Y2
D9
AC8
D48
Y24
REQ1
D10
A27
Y3
D10
AD8
D49
W23
RESET
D9
A28
AA3
D11
AE8
D50
W24
RETRY
A6
A29
AB2
D12
AC9
D51
W25
RL
B17
A30
AB3
D13
AE9
D52
V23
SCLK0
B23
A31
AC1
D14
AF9
D53
V26
SCLK1
B22
AS0
D6
D15
AE10
D54
U24
STATUS0
C5
AS1
A5
D16
AC11
D55
U26
STATUS1
A4
AS2
B5
D17
AF11
D56
T25
STATUS2
B4
BS0
B7
D18
AC12
D57
R24
STATUS3
D4
BS1
C7
D19
AD12
D58
R26
STATUS4
D3
CAREA0
D23
D20
AC13
D59
P24
STATUS5
D19
CAREA1
D21
D21
AD14
D60
N24
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GGP Terminal Assignments – Alphabetical Listing (Continued)
TERMINAL
NAME
NO.
TERMINAL
NAME
TERMINAL
NAME
NO.
NO.
TERMINAL
NAME
NO.
TCK
J25
A11
VSS
VSS
U25
R25
VSS
VSS
A10
K23
VDD
VDD
P25
TDI
TDO
K24
VDD
T2
VSS
A18
VSS
V25
VDD
VDD
T3
VSS
VSS
A22
VSS
VSS
W2
VDD
VDD
T24
VSS
VSS
B10
VSS
VSS
W4
VDD
VDD
U23
VSS
VSS
B16
VSS
VSS
Y26
VDD
VDD
V4
VSS
VSS
C10
VSS
VSS
AB1
VDD
VDD
Y4
VSS
VSS
C17
VSS
VSS
AB25
VDD
VDD
AA1
VSS
VSS
D17
VSS
VSS
AC6
VDD
VDD
AA26
VSS
VSS
E26
VSS
VSS
AD4
VDD
VDD
AC2
VSS
VSS
G1
VSS
VSS
AD11
VDD
VDD
AC7
VSS
VSS
G4
VSS
VSS
AD17
VDD
VDD
AC14
VSS
VSS
G26
VSS
VSS
AE11
VDD
VDD
AC25
VSS
VSS
K4
VSS
VSS
AE23
VDD
VDD
AD5
VSS
VSS
L4
VSS
VSS
AF7
VDD
VDD
AD10
VSS
VSS
L25
VSS
VSS
AF16
VDD
VDD
AD19
VSS
VSS
P3
AF20
P4
VSS
VSS
VDD
VDD
AE12
VSS
VSS
P23
VSYNC0
G24
R1
VSYNC1
G23
VDD
VDD
AF8
VSS
VSS
R4
W
C19
T26
XPT0
K26
VDD
VDD
AF17
VSS
VSS
U1
XPT1
L23
U2
XPT2
R23
TMS
J26
TRG / CAS
D18
TRST
J24
UTIME
D7
VDD
VDD
A9
VDD
VDD
A15
VDD
VDD
A21
VDD
VDD
B19
VDD
VDD
C18
VDD
VDD
C23
VDD
VDD
D5
VDD
VDD
D14
VDD
VDD
F2
VDD
VDD
F4
VDD
VDD
H2
VDD
VDD
H25
VDD
VDD
J2
VDD
VDD
L1
VDD
VDD
M23
VDD
N4
A12
A20
B11
C11
C22
D2
D12
D16
F3
F25
H24
H26
K2
M1
N3
T23
U4
V3
W26
Y23
AA2
AB23
AC3
AC10
AC19
AC26
AD7
AD15
AD22
AE18
AF12
AF21
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
A23
B14
C4
C13
C20
E3
F23
G2
G25
J4
L3
L24
N23
V24
W3
Y25
AA4
AB4
AB26
AC17
AD9
AD13
AD21
AE13
AF5
AF10
AF19
AF23
11
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
Terminal Functions
TERMINAL
NAME
DESCRIPTION
TYPE†
LOCAL MEMORY INTERFACE
A31 – A0
O
Address bus. A31 – A0 output the 32-bit byte address of the external memory cycle. The address can be
multiplexed for DRAM accesses.
AS2 – AS0
I
Address-shift selection. AS2 – AS0 determine how the column address appears on the address bus. Eight
shift values are supported, including zero.
BS1 – BS0
I
Bus-size selection. BS1 – BS0 indicate the bus size of the memory or other device being accessed, allowing
dynamic bus sizing for data buses less than 64-bits wide.
I
Cycle-timing selection. CT2 – CT0 signals determine the timing of the current memory access.
CT2 – CT0
D63 – D0
I/O
Data bus. D63 – D0 transfer up to 64 bits of data per memory cycle into or out of the ’C80.
DBEN
O
Data-buffer enable. DBEN drives the active-low output-enables of bi-directional transceivers that can be
used to buffer input and output data on D63 – D0.
DDIN
O
Data-direction indicator. DDIN indicates the direction of the data that passes through the transceivers. When
DDIN is low, the transfer is from external memory into the ’C80.
FAULT
I
Fault. FAULT is driven low by external circuitry to inform the ’C80 that a fault has occurred on the current
memory row-access.
PS3 – PS0
I
Page-size indication. PS3 – PS0 indicate the page size of the memory device(s) being accessed by the
current cycle. The ’C80 uses this information to determine when to begin a new row-access.
READY
I
Ready. READY indicates that the external device is ready to complete the memory cycle. READY is driven
low by external circuitry to insert wait states into a memory cycle.
RL
O
Row latch. The high-to-low transition of RL can be used to latch the valid 32-bit byte address that is present
on A31 – A0.
RETRY
I
Retry. RETRY is driven low by external circuitry to indicate that the addressed memory is busy. The ’C80
memory cycle is rescheduled.
STATUS5 – STATUS0
O
Status code. At row time, STATUS5 – STATUS0 indicate the type of cycle being performed. At column time,
they identify the processor and type of request that initiated the cycle.
UTIME
I
User-timing selection. UTIME causes the timing of RAS and CAS / DQM7 – CAS / DQM0 to be modified so
that custom memory timings can be generated. During reset, UTIME selects the endian mode in which the
’C80 operates.
DRAM, VRAM, AND SDRAM CONTROL
CAS / DQM7 –
CAS / DQM0
O
Column-address strobes. CAS / DQM7 – CAS / DQM0 drive the CAS inputs of DRAMs and VRAMs, or the
DQM input of SDRAMs. The eight strobes provide byte-write access to memory.
DSF
O
Special function. DSF selects special VRAM functions such as block-write, load color register, split-register
transfer, and SGRAM block write.
RAS
O
Row-address strobe. RAS drives the RAS inputs of DRAMs, VRAMs, and SDRAMs.
TRG / CAS
O
Transfer / output enable or column-address strobe. TRG / CAS is used as an output-enable for DRAMs and
VRAMs, and also as a transfer-enable for VRAMs. TRG / CAS also drives the CAS inputs of SDRAMs.
W
O
Write enable. W is driven low before CAS during write cycles. W controls the direction of the transfer during
VRAM transfer cycles.
† I = input, O = output, Z = high impedance
12
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
Terminal Functions (Continued)
TERMINAL
NAME
TYPE†
DESCRIPTION
O
Host acknowledge. The ’C80 drives HACK output low following an active HREQ to indicate that it has driven
the local-memory-bus signals to the high-impedance state and is relinquishing the bus. HACK is driven high
asynchronously following HREQ being detected inactive, and then the ’C80 resumes driving the bus.
HREQ
I
Host request. An external device drives HREQ low to request ownership of the local-memory bus. When
HREQ is high, the ’C80 owns and drives the bus. HREQ is synchronized internally to the ’C80’s internal clock.
Also, HREQ is used at reset to determine the power-up state of the MP. If HREQ is low at the rising edge
of RESET, the MP comes up running. If HREQ is high, the MP remains halted until the first interrupt
occurrence on EINT3.
REQ1, REQ0
O
Internal cycle request. REQ1 and REQ0 provide a two-bit code indicating the highest-priority memory-cycle
request that is being received by the TC. External logic can monitor REQ1 and REQ0 to determine if it is
necessary to relinquish the local-memory bus to the ’C80.
HOST INTERFACE
HACK
SYSTEM CONTROL
CLKIN
I
Input clock. CLKIN generates the internal ’C80 clocks to which all processor functions (except the frame
timers) are synchronous.
CLKOUT
O
Local output clock. CLKOUT provides a way to synchronize external circuitry to internal timings. All ’C80
output signals (except the VC signals) are synchronous to this clock.
EINT1, EINT2, EINT3
I
Edge-triggered interrupts. EINT1, EINT2 and EINT3 allow external devices to interrupt the master processor
(MP) on one of three interrupt levels (EINT1 is the highest priority). The interrupts are rising-edge triggered.
EINT3 also serves as an unhalt signal. If the MP is powered-up halted, the first rising edge on EINT3 causes
the MP to unhalt and fetch its reset vector (the EINT3 interrupt-pending bit is not set in this case).
LINT4
I
Level-triggered interrupt. LINT4 provides an active-low level-triggered interrupt to the MP. Its priority falls
below that of the edge-triggered interrupts. Any interrupt request should remain low until it is recognized by
the ’C80.
RESET
I
Reset. RESET is driven low to reset the ’C80 (all processors). During reset, all internal registers are set to
their initial state and all outputs are driven to their inactive or high-impedance levels. During the rising edge
of RESET, the MP reset mode and the ’C80’s operating endian mode are determined by the levels of HREQ
and UTIME pins, respectively.
XPT2 – XPT0
I
External packet transfer. XPT2 – XPT0 are used by external devices to request a high-priority XPT by the TC.
EMULATION CONTROL
I/O
Emulation pins. EMU0 and EMU1 are used to support emulation host interrupts, special functions targeted
at a single processor, and multiprocessor halt-event communications.
I
Test clock. TCK provides the clock for the ’C80 IEEE-1149.1 logic, allowing it to be compatible with other
IEEE-1149.1 devices, controllers, and test equipment designed for different clock rates.
TDI‡
I
Test data input. TDI provides input data for all IEEE-1149.1 instructions and data scans of the ’C80.
TDO
TMS‡
O
Test data output. TDO provides output data for all IEEE-1149.1 instructions and data scans of the ’C80.
I
Test-mode select. TMS controls the IEEE-1149.1 state machine.
TRST§
I
Test reset. TRST resets the ’C80 IEEE-1149.1 module. When low, all boundary-scan logic is disabled,
allowing normal ’C80 operation.
EMU0, EMU1‡
TCK‡
† I = input, O = output, Z = high impedance
‡ This pin has an internal pullup and can be left unconnnected during normal operation.
§ This pin has an internal pulldown and can be left unconnnected during normal operation.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
13
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
Terminal Functions (Continued)
TERMINAL
NAME
TYPE†
DESCRIPTION
O
Composite area. CAREA0 and CAREA1 define a special area such as an overscan boundary. This area
represents the logical OR of the internal horizontal and vertical area signals.
VIDEO INTERFACE
CAREA0, CAREA1
Composite blanking / vertical blanking. Each of CBLNK0 / VBLNK0 and VBLNK1 provides one of two
blanking functions, depending on the configuration of the CSYNC / HBLNK pin:
CBLNK0 / VBLNK0,
CBLNK1 / VBLNK1
O
Composite blanking disables pixel display /capture during both horizontal and vertical retrace periods
and is enabled when CSYNC is selected for composite sync video systems.
Vertical blanking disables pixel display /capture during vertical retrace periods and is enabled when
HBLNK is selected for separate-sync video systems.
Following reset, CBLNK0 / VBLNK0 and CBLNK1 / VBLNK1 are configured as CBLNK0 and CBLNK1,
respectively.
Composite sync / horizontal blanking. CSYNC0 / HBLNK0 and CSYNC1 / HBLNK1 can be programmed
for one of two functions:
CSYNC0 / HBLNK0,
CSYNC1 / HBLNK1
I/O/Z
Composite sync is for use on composite-sync video systems and can be programmed as an input,
output, or high-impedance signal. As an input, the ’C80 extracts horizontal and vertical sync
information from externally generated active-low sync pulses. As an output, the active-low composite
sync pulses are generated from either external HSYNC and VSYNC signals or the ’C80’s internal
video timers. In the high-impedance state, the pin is neither driven nor allowed to drive circuitry.
Horizontal blank disables pixel display / capture during horizontal retrace periods in separate-sync
video systems and can be used as an output only.
Immediately following reset, CSYNC0 / HBLNK0 and CSYNC1 / HBLNK1 are configured as
high-impedance CSYNC0 and CSYNC1, respectively.
FCLK0, FCLK1
HSYNC0,
HSYNC1
SCLK0, SCLK1
VSYNC0,
VSYNC1
I
Frame clock. FCLK0 and FCLK1 are derived from the external video system’s dotclock and are used to
drive the ’C80 video logic for frame timer 0 and frame timer 1.
I/O/Z
Horizontal sync. HSYNC0 and HSYNC1 control the video system. They can be programmed as input,
output, or high impedance signals. As an input, HSYNC synchronizes the video timer to externally
generated horizontal sync pulses. As an output, HSYNC is an active-low horizontal sync pulse generated
by the ’C80 on-chip frame timer. In the high-impedance state, the pin is not driven, and no internal
synchronization is allowed to occur. Immediately following reset, HSYNC0 and HSYNC1 are in the
high-impedance state.
I
Serial-data clock. SCLK0 and SCLK1 are used by the ’C80 SRT controller to track the VRAM tap point
when using midline reload. SCLK0 and SCLK1 should be the same signals that clock the serial register
on the VRAMs controlled by frame timer 0 and frame timer 1, respectively.
I/O/Z
Vertical sync. VSYNC0 and VSYNC1 control the video system. They can be programmed as inputs,
outputs, or high-impedance signals. As inputs, VSYNCx synchronizes the frame timer to externally
generated vertical-sync pulses. As outputs, VSYNCx are active-low vertical-sync pulses generated by the
’C80 on-chip frame timer. In the high-impedance state, the pin is not driven and no internal synchronization
is allowed to occur. Immediately following reset, VSYNCx is in the high-impedance state.
POWER
VSS‡
VDD‡
I
Ground. Electrical ground inputs
I
Power. Nominal 3.3-V power supply inputs
MISCELLANEOUS
No Connect
No connect serves as an alignment key and must be left unconnected.
FF2 – FF1
FF2 – FF1 (GF package only) are reserved for factory use and should be left unconnected.
† I = input, O = output, Z = high-impedance
‡ For proper operation, all VDD and VSS pins must be connected externally.
14
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
architecture
Figure 1 shows the major components of the ’C80: the master processor (MP), the parallel digital signal
processors (PPs), the transfer controller ( TC), the video controller ( VC), and the IEEE-1149.1 emulation
interface. Shared access to on-chip RAM is achieved through the crossbar. Crossbar connections are
represented by . Each PP can perform three accesses per cycle through its local ( L ), global ( G ), and
instruction ( I ) ports. The MP can access two RAMs per cycle through its crossbar/data ( C/ D) and instruction
( I ) ports, and the TC can access one RAM through its crossbar interface. Up to 15 simultaneous accesses are
supported in each cycle. Addresses can be changed every cycle, allowing the crossbar matrix to be changed
on a cycle-by-cycle basis. Contention between processors for the same RAM in the same cycle is resolved by
a round-robin priority scheme. In addition to the crossbar, a 32-bit datapath exists between the MP and the TC
and VC. This allows the MP to access TC and VC on-chip registers that are memory mapped into the MP
memory space.
The ’C80 has a 4G-byte address space as shown in Figure 2. The lower 32M bytes are used to address internal
RAM and memory-mapped registers.
PP3
PP2
PP1
PP0
MP
VC
OCR
L G
I
32
64
L G
I
32
64
32
L G
I
32
64
32
L G
I
32
C/D
I
64
32
32
32
IEEE1149.1
(JTAG)
32
64
64
Instruction Cache
Data RAM0
Data RAM1
Data RAM2
Parameter RAM
Data RAM0
Instruction Cache
Data RAM1
Data RAM2
Parameter RAM
Instruction Cache
Data RAM0
Data RAM1
Data RAM2
Parameter RAM
Data RAM0
Instruction Cache
Data RAM1
Data RAM2
Parameter RAM
Data RAM0
Instruction Cache
Data RAM1
Data RAM2
Parameter RAM
64
TC
Figure 1. Block Diagram Showing Datapaths
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
15
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
architecture (continued)
PP0 Data RAM0
(2K Bytes)
PP0 Data RAM1
(2K Bytes)
PP1 Data RAM0
(2K Bytes)
PP1 Data RAM1
(2K Bytes)
PP2 Data RAM0
(2K Bytes)
PP2 Data RAM1
(2K Bytes)
PP3 Data RAM0
(2K Bytes)
PP3 Data RAM1
(2K Bytes)
Reserved
(16K Bytes)
PP0 Data RAM2
(2K Bytes)
Reserved
(2K Bytes)
PP1 Data RAM2
(2K Bytes)
Reserved
(2K Bytes)
PP2 Data RAM2
(2K Bytes)
Reserved
(2K Bytes)
PP3 Data RAM2
(2K Bytes)
Reserved
(16 730 112 Bytes)
PP0 Parameter RAM
(2K Bytes)
Reserved
(2K Bytes)
PP1 Parameter RAM
(2K Bytes)
Reserved
(2K Bytes)
PP2 Parameter RAM
(2K Bytes)
Reserved
(2K Bytes)
PP3 Parameter RAM
(2K Bytes)
0x01003800
0x00000000
Reserved
(51 200 Bytes)
0x000007FF
0x00000800
0x00000FFF
0x00001000
MP Parameter RAM
(2K Bytes)
0x000017FF
0x00001800
PP0 Instruction Cache
(2K Bytes)
0x00002FFF
0x00003000
0x000037FF
0x00003800
Reserved
(6K Bytes)
PP1 Instruction Cache
(2K Bytes)
0x00003FFF
0x00004000
0x00007FFF
0x00008000
Reserved
(6K Bytes)
PP2 Instruction Cache
(2K Bytes)
0x000087FF
0x00008800
Reserved
(6K Bytes)
0x00008FFF
0x00009000
PP3 Instruction Cache
(2K Bytes)
0x000097FF
0x00009800
Reserved
(32K Bytes)
0x00009FFF
0x0000A000
MP Data Cache
(4K Bytes)
0x0000A7FF
0x0000A800
Reserved
(28K Bytes)
0x0000AFFF
0x0000B000
MP Instruction Cache
(4K Bytes)
0x0000B7FF
0x0000B800
0x00FFFFFF
0x01000000
0x010007FF
0x01000800
Reserved
(28K Bytes)
Memory-Mapped TC Registers
(512 Bytes)
Memory-Mapped VC Registers
(512 Bytes)
0x01000FFF
0x01001000
0x010017FF
0x01001800
Reserved
(8 327 168 Bytes)
0x018017FF
0x01801800
0x01801FFF
0x01802000
0x018037FF
0x01803800
0x01803FFF
0x01804000
0x018057FF
0x01805800
0x01805FFF
0x01806000
0x018077FF
0x01807800
0x01807FFF
0x01808000
0x0180FFFF
0x01810000
0x01810FFF
0x01811000
0x01817FFF
0x01818000
0x01818FFF
0x01819000
0x0181FFFF
0x01820000
0x018201FF
0x01820200
0x018203FF
0x01820400
0x01FFFFFF
0x02000000
0x01001FFF
0x01002000
External Memory
(4 064M Bytes)
0x010027FF
0x01002800
0x01002FFF
0x01003000
0x010037FF
0xFFFFFFFF
Figure 2. Memory Map
16
0x010107FF
0x01010800
Reserved
(8 327 168 Bytes)
0x00001FFF
0x00002000
0x000027FF
0x00002800
0x0100FFFF
0x01010000
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
master processor (MP) architecture
The master processor (MP) is a 32-bit RISC processor with an integral IEEE-754 floating-point unit. The MP
is designed for effective execution of C code and is capable of performing at well over 130K dhrystones/s. Major
tasks which the MP typically performs are:
D
D
D
Task control and user interface
Information processing and analysis
IEEE-754 floating point (including graphics transforms)
MP functional block diagram
Figure 3 shows a block diagram of the master processor. Key features of the MP include:
D
D
D
D
D
D
D
D
D
D
D
32-bit RISC processor
–
Load/store architecture
–
Three operand arithmetic and logical instructions
4K-byte instruction cache and 4K-byte data cache
–
Four-way set associative
–
LRU replacement
–
Data writeback
2K-byte non-cached parameter RAM
Thirty-one 32-bit general-purpose registers
Register and accumulator scoreboard
15-bit or 32-bit immediate constants
32-bit byte addressing
Scalable timer
Leftmost-one and rightmost-one logic
IEEE-754 floating-point hardware
–
Four double-precision floating-point vector accumulators
–
Vector floating-point instructions
Floating-point operation and parallel load or store
Multiply and accumulate
High performance
–
60 million instructions per second (MIPS)
–
120 million floating-point operations per second (MFLOPS)
–
Over 130K dhrystones/s
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
17
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP functional block diagram (continued)
Register File
(Thirty-One 32-Bit Registers)
Scoreboard
Barrel Rotator
Mask Generator
Double-Precision
Floating-Point Multiplier
(Single-Precision Core)
Zero Comparator
Integer ALU
Leftmost / Rightmost One
Double-Precision Floating-Point
Accumulators
Timer
Control Registers
Double-Precision
Floating-Point Adder
Instruction Register
Program Counters
PC Incrementer
Emulation Logic
Endian Multiplexers
Instruction Cache
Controller
Data/Cache
Controller
Crossbar Interface
Figure 3. MP Block Diagram
MP general-purpose registers
The MP contains 31 32-bit general-purpose registers, R1 – R31. Register R0 always reads as zero and writes
to it are discarded. Double precision values are always stored in an even-odd register pair with the higher
numbered register always holding the sign bit and exponent. The R0/R1 pair is not available for this use. A
scoreboard keeps track of which registers are awaiting loads or the result of a previous instruction and stalls
the instruction pipeline until the register contains valid data. As a recommended software convention, typically
R1 is used as a stack pointer and R31 as a return-address link register.
Figure 4 shows the MP general-purpose registers.
18
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP general-purpose registers (continued)
Zero/Discard
Not Available
R1
R2
R2 R3
R2,
R3
R4
R4 R5
R4,
R5
•••
•••
R30
R30 R31
R30,
R31
32-Bit Registers
64-Bit Register Pairs
Figure 4. MP General-Purpose Registers
The 32-bit registers can contain signed-integer, unsigned-integer, or single precision floating-point values.
Signed and unsigned bytes and halfwords are sign extended or zero-filled. Doublewords may be stored in a
64-bit even/odd register pair. Double-precision floating-point values are referenced using the even register
number or the register pair. Figure 5 through Figure 7 show the register data formats.
31
Si
Single
l P
Precision
i i
Floating Point
22
0
S E E E E E E E E M M M M M M M M M M M M M M M M M M M M M M M
MS
LS
31
Si
d 32
bit
Signed
32-bit
Integer
S
0
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
MS
I
I
I
I
I
I
I
I
I
I
LS
31
Unsigned 32-Bit
32 Bit
Integer
I
0
U U U U U U U U U U U U U U U U U U U U U U U U U U U U U U U U
MS
LS
Figure 5. MP Register 32-Bit Data Formats
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
19
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP general-purpose registers (continued)
31
Signed Byte
7
S S S S S S S S S S S S S S S S S S S S S S S S S
0
I
I
I
I
I
I
MS
31
Unsigned Byte
0
LS
7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
U U U U U U U U
MS
31
Signed Halfword
0
I
I
I
I
I
I
I
MS
31
Unsigned
Halfword
0
LS
15
S S S S S S S S S S S S S S S S S
15
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
I
I
I
I
I
I
I
I
I
LS
0
U U U U U U U U U U U U U U U U
MS
LS
Figure 6. MP Register 8-Bit and 16-Bit Data
31
Odd Register
0
Most Significant 32-Bit Word
MS
31
Even Register
0
Least Significant 32-Bit Word
LS
19
0
Double Precision 31
Double-Precision
S E E E E E E E E E E E M M M M M M M M M M M M M M M M M M M M
Floating-Point
Odd Register
MS
0
Double-Precision
Double
Precision 31
M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M
Floating-Point
Even Register
LS
Figure 7. MP Register 64-Bit Data
MP double-precision floating-point accumulators
There are four double-precision floating-point registers (see Figure 8) to accumulate intermediate floating-point
results.
64
0
a0
Accumulator 0
a1
Accumulator 1
a2
Accumulator 2
a3
Accumulator 3
MSB
LSB
Figure 8. Double-Precision Floating-Point Accumulators
20
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP control registers
In addition to the general-purpose registers, there are a number of control registers that are used to represent
the state of the processor. Table 1 shows the control register numbers of the accessible registers.
Table 1. Control Register Numbers
NO.
NAME
DESCRIPTION
NO.
NAME
0x0000
EPC
Exception Program Counter
0x0015–0x001F
—
0x0001
EIP
Exception Instruction Pointer
0x0020
SYSSTK
System Stack Pointer
0x0002
CONFIG
Configuration
0x0021
SYSTMP
System Temporary Register
0x0003
—
0x0004
INTPEN
0x0005
Reserved
DESCRIPTION
Reserved
0x0022–0x002F
—
Interrupt Pending
0x0030
MPC
Emulator Exception Program Cntr
—
Reserved
0x0031
MIP
Emulator Exception Instruction Ptr
0x0006
IE
Interrupt Enable
0x0032
—
0x0007
—
Reserved
0x0033
ECOMCNTL
0x0008
FPST
0x0009
—
0x000A
PPERROR
Floating-Point Status
Reserved
Reserved
Reserved
Emulator Communication Control
0x0034
ANASTAT
0x0035–0x0038
—
0x0039
BRK1
Emulation Breakpoint 1 Reg.
Emulation Breakpoint 2 Reg.
PP Error Indicators
Emulation Analysis Status Reg
Reserved
0x000B
—
Reserved
0x003A
BRK2
0x000C
—
Reserved
0x003B–0x01FF
—
0x000D
PKTREQ
Packet Request Register
0x0200 – 0x020F
ITAG0 – 15
Instruction Cache Tags 0 to 15
0x000E
TCOUNT
Current Counter Value
0x0300
ILRU
Instruction Cache LRU Register
0x000F
TSCALE
Counter Reload Value
0x0400–0x040F
DTAG0 – 15
Data Cache Tags 0 to 15
Reserved
0x0010
FLTOP
Faulting Operation
0x0500
DLRU
Data Cache LRU Register
0x0011
FLTADR
Faulting Address
0x4000
IN0P
Vector Load Pointer 0
0x0012
FLTTAG
Faulting Tag
0x4001
IN1P
Vector Load Pointer 1
0x0013
FLTDTL
Faulting Data (low)
0x4002
OUTP
Vector Store Pointer
0x0014
FLTDTH
Faulting Date (high)
MP pipeline registers
The MP uses a three-stage fetch, execute, access (FEA) pipeline. The primary pipeline registers are
manipulated implicitly by branch and trap instructions and are not accessible by the user. The exception and
emulation pipeline registers are user accessible as control registers. All pipeline registers are 32 bits.
Program Execution Mode
Normal
Exception
Emulation
PC
EPC
MPC
Instruction Pointer
IP
EIP
MIP
Instruction Register
IR
Program Counter
•
•
•
Instruction register (IR) contains the instruction being
executed.
Instruction pointer (IP) points to the instruction being
executed.
Program counter (PC) points to the instruction being
fetched.
•
•
Exception/emulator instruction pointer (EIP/MIP) points to the
instruction that would have been executed had the exception /
emulation trap not occurred.
Exception/emulator program counter (EPC/MPC) points to the
instruction to be fetched on returning from the exception/emulation
trap.
Figure 9. MP FEA Pipeline Registers
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
21
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
configuration (CONFIG) register (0x0002)
The CONFIG register controls or reflects the state of certain options as shown in Figure 10.
31
30
29
28
27
E
R
T
H
X
E
R
T
H
X
Type
Release
26
25
24
23
22
21
20
19
18
17
16
15
Reserved
14
13
12
11
Type
10
9
8
7
Reserved
6
5
4
3
Release
2
1
0
Reserved
Endian mode; 0 = big-endian, 1 = little-endian, read only
PPData RAM round robin; 0 = fixed, 1 = variable, read / write
TC PT round robin; 0 = variable, 1 = fixed, read / write.
High priority MP events; 0 = disabled, 1 = enabled, read / write
Externally initiated packet transfers; 0 = disabled, 1 = enabled, read / write
Number of PPs in device, read only
TMS320C80 version number
Figure 10. CONFIG Register
interrupt-enable (IE) register (0x0006)
The IE register contains enable bits for each of the interrupts/traps as shown in Figure 11. The
global-interrupt-enable (ie) bit and the appropriate individual interrupt-enable bit must be set in order for an
interrupt to occur.
31
30
29
28
27
26
25
pe
x4
x3
bp
pb
pc
mi
pe
x4
x3
bp
pb
pc
mi
p3
24
23
22
21
20
19
18
17
16
15
14
p3
p2
p1
p0
io
mf
p2
p1
p0
io
mf
x2
x1
ti
PP error
External interrupt 4 (LINT4)
External interrupt 3 (EINT3)
Bad packet transfer
Packet transfer busy
Packet transfer complete
Message (MP self) interrupt
PP3 message interrupt
13
12
11
10
9
8
7
6
5
x2
x1
ti
f1
f0
fx
fu
fo
PP2 message interrupt
PP1message interrupt
PP0 message interrupt
Integer overflow
Memory fault
External interrupt 2 (EINT2)
External interrupt 1 (EINT1)
MP timer interrupt
f1
f0
fx
fu
fo
fz
fi
ie
4
3
2
fz
fi
1
0
ie
Frame-timer 1 interrupt
Frame-timer 0 interrupt
Floating-point inexact
Floating-point underflow
Floating-point overflow
Floating-point divide-by-zero
Floating-point invalid
Global-interrupt enable
Figure 11. IE Register
interrupt-pending (INTPEN) register (0x0004)
The bits in INTPEN register show the current state of each interrupt/trap. Pending interrupts do not occur unless
the ie bit and corresponding interrupt-enable bit are set. Software must write a 1 to the appropriate INTPEN bit
to clear an interrupt. Figure 12 shows the INTPEN register locations.
31
30
29
28
27
26
25
pe
x4
x3
bp
pb
pc
mi
24
23
22
21
20
19
18
17
16
15
14
p3
p2
p1
p0
io
mf
13
12
11
10
9
8
7
6
5
x2
x1
ti
f1
f0
fx
fu
fo
Figure 12. INTPEN Register
22
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
4
3
2
fz
fi
1
0
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
floating-point status register (FPST) (0x0008)
FPST contains status and control information for the FPU as shown in Figure 13. Bits 17–21 are read/write
floating-point unit (FPU) control bits. Bits 22 – 26 are read/write accumulated status bits. All other bits show the
status of the last FPU instruction to complete and are read only.
31
30
29
28
27
destination
dest
ai
az
ao
au
ax
sm
fs
vm
drm
opcode
e1
26
25
24
23
22
21
20
19
ai
az
ao
au
ax
sm
fs
vm
18
17
16
drm
15
14
13
opcode
11
e0
10
9
8
pd
7
6
rm
The ninth MSB of exponent
Destination precision
00 – single float
01 – double float
Rounding mode
00 – nearest
01 – zero
Int multiply overflow
Invalid
Divide-by-zero
Overflow
Underflow
Inexact
e0
pd
Destination register value
Accumulated value invalid
Accumulated divide-by-zero
Accumulated overflow
Accumulated underflow
Accumulated inexact
Sequential mode select
Floating-point stall
Vector fast mode
Rounding mode
00 – nearest
10 – positive ∞
01 – zero
11 – negative ∞
Last opcode
The tenth MSB of exponent
12
e1
rm
mo
i
z
o
u
x
5
4
3
2
1
0
mo
i
z
o
u
x
10 – signed int
11 – unsigned int
10 – positive ∞
11 – negative ∞
Figure 13. FPSTS Register
PP error register (PPERROR) (0x000A)
The bits in the PPERROR register reflect parallel processor errors (see Figure 14). The MP can use these when
a PP error interrupt occurs to determine the cause of the error.
31
30
29
28
27
26
25
24
23
22
21
20
Reserved
PP#
h – PP halted
19
18
17
16
h
h
h
h
3
2
1
0
15
14
13
12
11
10
9
8
i
i
i
i
3
2
1
0
Reserved
PP#
i – PP illegal instruction
7
6
5
4
Reserved
PP#
3
2
1
0
f
f
f
f
3
2
1
0
f – PP fault type; 0 = icache, 1 = DEA
Figure 14. PPERROR Register
packet-transfer request register (PKTREQ) (0x000D)
PKTREQ controls the submission and priority of packet-transfer requests as shown in Figure 15. It also
indicates that a packet transfer is currently active.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
Reserved
I – Immediate (urgent) priority selected
F – High (foreground) priority selected
S – Suspend packet transfer
Q – Packet transfer queued; read only
9
8
7
6
5
4
3
2
1
0
I
F
S
Q
P
P – Submit packet-transfer
request
Figure 15. PKTREQ Register
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
23
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
memory-fault registers
The five read-only memory-fault registers contain information about memory address exceptions, as shown in
Figure 16.
FLTOP
(0x0010)
FLTTAG
(0x0012)
31
30
29
28
27
26
25
24
Dest
31
30
29
23
22
21
20
19
18
17
Reserved
28
27
26
25
24
23
22
21
16
15
K
20
19
18
17
14
13
SZ
16
15
14
13
12
11
10
9
i
d
x
R
12
11
10
22-Bit Cache Tag Address
8
7
6
5
4
3
2
Reserved
9
8
7
6
5
4
3
2
P
D
P
D
P
D
P
D
3
2
1
0
Block
1
1
0
0
Sub-Block
31
0
FLTADR
(0x0011)
Faulting Address Accessed by the Instruction
FLTDTH
(0x0013)
Faulting Write Most-Significant-Data Word
FLTDTL
(0x0014)
Faulting Write Least-Significant-Data Word
Dest
K
SZ
Destination Register Number
Kind of Operation:
00 – load
01 – unsigned load
10 – store
11 – cache flush/clean
Size of Data:
00 – 8-bit
01 – 16-bit
10 – 32-bit
11 – 64-bit
i
d
x
R
Block
P
D
MP icache fault
MP dcache fault
DEA Fault
Modified return sequence
Faulting block number
Sub-block is present.
Dirty bit set
Figure 16. Memory-Fault Registers
24
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP cache registers
The ILRU and DLRU registers track least-recently-used (LRU) information for the sixteen instruction-cache and
sixteen data-cache blocks. The ITAGxx registers contain block addresses and the present flags for each
sub-block. DTAGxx registers are identical to ITAGxx registers but include dirty bits for each sub-block. Figure 17
shows the cache registers.
ILRU (0x0300)
DLRU (0x0500)
31
30
mru
29
28
nmru
27
26
25
nlru
24
lru
23
22
mru
21
20
nmru
19
18
17
nlru
16
lru
15
14
mru
13
Set 2
Set 3
12
nmru
11
10
9
nlru
8
7
lru
6
mru
5
4
nmru
Set 1
3
2
1
nlru
0
lru
Set 0
ITAG0 –ITAG15 (0x0200 – 0x020F)
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
22-Bit Cache Tag Address
9
8
P
7
6
P
3
5
4
P
2
3
2
1
0
1
0
P
1
0
Sub-Block
DTAG0 – DTAG15 (0x0400 – 0x040F)
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
22-Bit Cache Tag Address
9
8
7
6
5
4
3
2
P
D
P
D
P
D
P
D
3
2
1
0
Sub-Block
mru
nmru
nlru
Most-recently-used block
Next most-recently-used block
Next least-recently-used block
lru
P
D
Least-recently-used block
Sub-block present
Sub-block dirty
mru, nmru, nlru, and lru have the value 0, 1, 2, or 3 representing the block number and are mutually exclusive for each set.
Figure 17. Cache Registers
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
25
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP cache architecture
The MP contains two four-way set-associative, 4K caches for instructions and data. Each cache is divided into
four sets with four blocks in each set. Each block represents 256 bytes of contiguous instructions or data and
is aligned to a 256-byte address boundary. Each block is partitioned into four sub-blocks that each contain
sixteen 32-bit words and are aligned to 64-byte boundaries within the block. Cache misses cause one sub-block
to be loaded into cache. Figure 18 shows the cache architecture for one of the four sets in each cache. Figure 19
shows how addresses map into the cache using the cache tags and address bits.
Block 0
Tag Reg 0 (Block 0)
Block 1
Tag Reg 1 (Block 1)
LRU in SET 0
Sub-Blocks
NLRU in SET 0
NMRU in SET 0
Set 0
Block 2
Tag Reg 2 (Block 2)
Block 3
Tag Reg 3 (Block 3)
MRU in SET 0
LRU Stack for SET 0
Figure 18. MP Cache Architecture (x4 Sets)
32-Bit Logical Address
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
S
S
s
s
5
4
3
2
W W W W
1
0
B
B
On-Chip MP 2K Cache RAMS
Bank 0
Bank 1
Set 0
Set 2
Set 1
Set 3
11 10 9 8 7 6 5 4 3 2 1 0
S S A A s s W W W W B B
Address in On-Chip
Cache Bank
T – Tag Address Bits
s – Sub-Block (within block) Select (0 – 3)
B – Byte (within word) Select (0 – 3)
S – Set Select Bits (0 – 3)
W – Word (within sub-block) Select (0 – 15)
A – Block Select (which tag matched) (0 – 3)
Figure 19. MP Cache Addressing
26
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP parameter RAM
The parameter RAM is a noncachable, 2K-byte, on-chip RAM which contains MP-interrupt vectors,
MP-requested TC task buffers, and a general-purpose area. Figure 20 shows the parameter RAM address map.
0x01010000– 0x0101007F
Suspended PT Parameters
(128 Bytes)
0x01010080– 0x010100DF
Reserved
(96 Bytes)
0x010100E0– 0x010100FB
XPT7 / SOF0 Linked List Start Add.
0x010100E0
XPT6 / SAM0 Linked List Start Add. 0x010100E4
XPT5 / SOF1 Linked List Start Add.
XPT Linked List Start Addresses
(28 Bytes)
0x010100FC – 0x010100FF
MP Linked List Start Address
0x01010100– 0x0101017F
Off-Chip to Off-Chip PT Buffer
(128 Bytes)
0x01010180– 0x0101021F
Interrupt and Trap Vectors
(160 Bytes)
0x01010220– 0x0101029F
XPT Off-Chip to Off-Chip PT Buffer
(128 Bytes)
0x010102A0– 0x010107FF
General-Purpose RAM
(1376 Bytes)
0x010100E8
XPT4 / SAM1 Linked List Start Add. 0x010100EC
XPT3 Linked List Start Add.
0x010100F0
XPT2 Linked List Start Add.
0x010100F4
XPT1 Linked List Start Add.
0x010100F8
Figure 20. MP Parameter RAM
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
27
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP interrupt vectors
Table 2 and Table 3 show the MP interrupts and traps and their vector addresses.
Table 2. Maskable Interrupts
IE BIT
(TRAP#)
NAME
VECTOR
ADDRESS
0
ie
0x01010180
2
fi
0x01010188
Floating-point invalid
3
fz
0x0101018C
Floating-point divide-by-zero
5
fo
0x01010194
Floating-point overflow
6
fu
0x01010198
Floating-point underflow
7
fx
0x0101019C
Floating-point inexact
8
f0
0x010101A0
Frame timer 0
9
f1
0x010101A4
Frame timer 1
MASKABLE INTERRUPT
10
ti
0x010101A8
MP timer
11
x1
0x010101AC
External interrupt 1 (EINT1)
12
x2
0x010101B0
External interrupt 2 (EINT2)
14
mf
0x010101B8
Memory fault
15
io
0x010101BC
Integer overflow
16
p0
0x010101C0
PP0 message
17
p1
0x010101C4
PP1 message
18
p2
0x010101C8
PP2 message
19
p3
0x010101CC
PP3 message
25
mi
0x010101E4
MP message
26
pc
0x010101E8
Packet transfer complete
27
pb
0x010101EC
Packet transfer busy
28
bp
0x010101F0
Bad packet transfer
29
x3
0x010101F4
External interrupt 3 (EINT3)
30
x4
0x010101F8
External interrupt 4 (LINT4)
31
pe
0x010101FC
PP error
Table 3. Nonmaskable Traps
TRAP
NO.
NAME
VECTOR
ADDRESS
32
e1
0x01010200
Emulator trap1 (reserved)
33
e2
0x01010204
Emulator trap2 (reserved)
34
e3
0x01010208
Emulator trap3 (reserved)
35
e4
0x0101020C
Emulator trap4 (reserved)
36
fe
0x01010210
Floating-point error
0x01010214
Reserved
0x01010218
Illegal MP instruction
39
0x0101021C
Reserved
72
to
415
0x010102A0 to
0x010107FC
37
38
28
er
POST OFFICE BOX 1443
NONMASKABLE TRAP
System or user defined
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP opcode formats
The three basic classes of MP instruction opcodes are; short immediate, three register, and long immediate.
Figure 21 shows the opcode structure for each class of instruction.
31
Short
Immediate
27 26
Dest
31
Three
Register
Dest
Source 2
1
1
15-Bit Immediate
13 12 11
1
Opcode
22 21 20 19
Source 2
0
Opcode
22 21 20 19
27 26
Dest
15 14
Source 2
27 26
31
Long
immediate
22 21
0
5 4
Options
13 12 11
1
Opcode
1
0
Source 1
5 4
Options
0
Source 1
32-Bit Long Immediate
Figure 21. MP Opcode Formats
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
29
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP opcode summary
Table 4 through Table 6 show the opcode formats for the MP. Table 7 summarizes the master processor
instruction set.
Table 4. Short-Immediate Opcodes
31
30
29
illop0
28
27
26
25
24
23
22
Source
Dest
21
20
19
18
17
16
15
0
0
0
0
0
0
0
14
13
12
11
10
Unsigned Immediate
9
8
7
6
5
4
3
2
trap
–
–
–
–
E
–
–
–
–
–
0
0
0
0
0
0
1
Unsigned Trap Number
cmnd
–
–
–
–
–
–
–
–
–
–
0
0
0
0
0
1
0
Unsigned Immediate
–
–
–
–
–
0
0
0
0
1
0
0
Unsigned Control Register Number
0
0
0
0
1
0
1
Unsigned Control Register Number
0
0
0
0
1
1
0
Unsigned Control Register Number
rdcr
Dest
swcr
Dest
brcr
–
–
–
Source
–
–
–
–
–
–
–
shift.dz
Dest
Source
0
0
0
1
0
0
0
–
–
–
i
n
Endmask
Rotate
shift.dm
Dest
Source
0
0
0
1
0
0
1
–
–
–
i
n
Endmask
Rotate
shift.ds
Dest
Source
0
0
0
1
0
1
0
–
–
–
i
n
Endmask
Rotate
shift.ez
Dest
Source
0
0
0
1
0
1
1
–
–
–
i
n
Endmask
Rotate
shift.em
Dest
Source
0
0
0
1
1
0
0
–
–
–
i
n
Endmask
Rotate
shift.es
Dest
Source
0
0
0
1
1
0
1
–
–
–
i
n
Endmask
Rotate
shift.iz
Dest
Source
0
0
0
1
1
1
0
–
–
–
i
n
Endmask
Rotate
shift.im
Dest
Source
0
0
0
1
1
1
1
–
–
–
i
n
Endmask
Rotate
and.tt
Dest
Source2
0
0
1
0
0
0
1
Unsigned Immediate
and.tf
Dest
Source2
0
0
1
0
0
1
0
Unsigned Immediate
and.ft
Dest
Source2
0
0
1
0
1
0
0
Unsigned Immediate
xor
Dest
Source2
0
0
1
0
1
1
0
Unsigned Immediate
or.tt
Dest
Source2
0
0
1
0
1
1
1
Unsigned Immediate
and.ff
Dest
Source2
0
0
1
1
0
0
0
Unsigned Immediate
xnor
Dest
Source2
0
0
1
1
0
0
1
Unsigned Immediate
or.tf
Dest
Source2
0
0
1
1
0
1
1
Unsigned Immediate
or.ft
Dest
Source2
0
0
1
1
1
0
1
Unsigned Immediate
or.ff
Dest
Source2
0
0
1
1
1
1
0
Unsigned Immediate
ld
Dest
Base
0
1
0
0
M
SZ
Signed Offset
ld.u
Dest
Base
0
1
0
1
M
SZ
Signed Offset
Base
0
1
1
0
M
SZ
Signed Offset
Source2
0
1
1
1
M
0
0
Signed Offset
1
0
0
0
0
0
A
Signed Offset
st
dcache
bsr
–
A
E
F
i
30
Source
–
–
–
–
Link
F
–
–
–
–
–
jsr
Link
Base
1
0
0
0
1
0
A
Signed Offset
bbz
BITNUM
Source
1
0
0
1
0
0
A
Signed Offset
bbo
BITNUM
Source
1
0
0
1
0
1
A
Signed Offset
bcnd
Cond
Source
1
0
0
1
1
0
A
Signed Offset
cmp
Dest
Source2
1
0
1
0
0
0
0
Signed Immediate
add
Dest
Source2
1
0
1
1
0
0
U
Signed Immediate
sub
Dest
Source2
1
0
1
1
0
1
U
Signed Immediate
Reserved bit (code as 0)
Annul delay slot instruction if branch taken
Emulation trap bit
Clear present flags
Invert endmask
POST OFFICE BOX 1443
M
n
SZ
U
1
0
Modify, write modified address back to register
Rotate sense for shifting
Size (0 = byte, 1 = halfword, 2 = word, 3 = doubleword)
Unsigned form
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP opcode summary (continued)
Table 5. Long-Immediate and Three-Register Opcodes
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
trap
–
–
–
–
E
–
–
–
–
–
1
1
0
0
0
0
0
0
1
I
–
–
–
–
–
–
–
IND TR
cmnd
–
–
–
–
–
–
–
–
–
–
1
1
0
0
0
0
0
1
0
I
–
–
–
–
–
–
–
Source1
–
–
–
–
–
1
1
0
0
0
0
1
0
0
I
–
–
–
–
–
–
–
IND CR
1
1
0
0
0
0
1
0
1
I
–
–
–
–
–
–
–
IND CR
1
1
0
0
0
0
1
1
0
I
–
–
–
–
–
–
–
IND CR
rdcr
Dest
swcr
Dest
brcr
–
–
–
Source
–
–
–
–
–
–
–
4
3
2
shift.dz
Dest
Source
1
1
0
0
0
1
0
0
0
0
i
n
Endmask
Rotate
shift.dm
Dest
Source
1
1
0
0
0
1
0
0
1
0
i
n
Endmask
Rotate
shift.ds
Dest
Source
1
1
0
0
0
1
0
1
0
0
i
n
Endmask
Rotate
shift.ez
Dest
Source
1
1
0
0
0
1
0
1
1
0
i
n
Endmask
Rotate
shift.em
Dest
Source
1
1
0
0
0
1
1
0
0
0
i
n
Endmask
Rotate
shift.es
Dest
Source
1
1
0
0
0
1
1
0
1
0
i
n
Endmask
Rotate
shift.iz
Dest
Source
1
1
0
0
0
1
1
1
0
0
i
n
Endmask
Rotate
shift.im
Dest
Source
1
1
0
0
0
1
1
1
1
0
i
n
and.tt
Dest
Source2
1
1
0
0
1
0
0
0
1
I
–
–
–
Endmask
–
–
–
–
Source1
and.tf
Dest
Source2
1
1
0
0
1
0
0
1
0
I
–
–
–
–
–
–
–
Source1
and.ft
Dest
Source2
1
1
0
0
1
0
1
0
0
I
–
–
–
–
–
–
–
Source1
xor
Dest
Source2
1
1
0
0
1
0
1
1
0
I
–
–
–
–
–
–
–
Source1
or.tt
Dest
Source2
1
1
0
0
1
0
1
1
1
I
–
–
–
–
–
–
–
Source1
Dest
Source2
1
1
0
0
1
1
0
0
0
I
–
–
–
–
–
–
–
Source1
xnor
Dest
Source2
1
1
0
0
1
1
0
0
1
I
–
–
–
–
–
–
–
Source1
or.tf
Dest
Source2
1
1
0
0
1
1
0
1
1
I
–
–
–
–
–
–
–
Source1
or.ft
Dest
Source2
1
1
0
0
1
1
1
0
1
I
–
–
–
–
–
–
–
Source1
or.ff
Dest
Source2
1
1
0
0
1
1
1
1
0
I
–
–
–
–
–
–
–
Source1
ld
Dest
Base
1
1
0
1
0
0
M
SZ
I
S
D
–
–
–
–
–
Offset
ld.u
Dest
Base
1
1
0
1
0
1
M
SZ
I
S
D
–
–
–
–
–
Offset
Base
1
1
0
1
1
0
M
SZ
I
S
D
–
–
–
–
–
Offset
Source2
1
1
0
1
1
1
M
0
0
I
0
0
–
–
–
–
–
Source1
1
1
1
0
0
0
0
0
A
I
–
–
–
–
–
–
–
Offset
st
bsr
–
D
E
F
i
Source
–
–
–
–
F
Link
–
–
–
–
–
jsr
Link
Base
1
1
1
0
0
0
1
0
A
I
–
–
–
–
–
–
–
Offset
bbz
BITNUM
Source
1
1
1
0
0
1
0
0
A
I
–
–
–
–
–
–
–
Target
Target
bbo
BITNUM
Source
1
1
1
0
0
1
0
1
A
I
–
–
–
–
–
–
–
bcnd
Cond
Source
1
1
1
0
0
1
1
0
A
I
–
–
–
–
–
–
–
Target
cmp
Dest
Source2
1
1
1
0
1
0
0
0
0
I
–
–
–
–
–
–
–
Source1
add
Dest
Source2
1
1
1
0
1
1
0
0
U
I
–
–
–
–
–
–
–
Source1
sub
Dest
Source2
1
1
1
0
1
1
0
1
U
I
–
–
–
–
–
–
–
Source1
Reserved bit (code as 0)
Direct external access bit
Emulation trap bit
Clear present flags
Invert endmask
l
M
n
S
SZ
POST OFFICE BOX 1443
0
Rotate
and.ff
dcache
1
Long immediate
Modify, write modified address back to register
Rotate sense for shifting
Scale offset by data size
Size (0 = byte, 1 = halfword, 2 = word, 3 = doubleword
• HOUSTON, TEXAS 77251–1443
31
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP opcode summary (continued)
Table 6. Miscellaneous Instruction Opcodes
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
vadd
31
Mem Src / Dst
30
29
28
27
26
Source2 / Dest
25
24
23
22
1
1
1
1
0
–
0
0
0
I
–
m
P
–
d
m
s
4
3
Source1
2
1
vsub
Mem Src / Dst
Source2 / Dest
1
1
1
1
0
–
0
0
1
I
–
m
P
–
d
m
s
Source1
vmpy
Mem Src / Dst
Source2 / Dest
1
1
1
1
0
–
0
1
0
I
–
m
P
–
d
m
s
Source1
vmsub
Mem Src / Dst
Dest
1
1
1
1
0
a
0
1
1
I
a
m
P
Z
–
m
–
Source1
vrnd(FP)
Mem Src / Dst
Dest
1
1
1
1
0
a
1
0
0
I
a
m
P
m
s
Source1
PD
0
vrnd(Int)
Mem Src / Dst
Dest
1
1
1
1
0
–
1
0
1
I
–
m
P
–
d
m
s
Source1
vmac
Mem Src / Dst
Source2
1
1
1
1
0
a
1
1
0
I
a
m
P
Z
–
m
–
Source1
vmsc
Mem Src / Dst
Source2
1
1
1
1
0
a
1
1
1
I
a
m
P
Z
–
m
–
Source1
fadd
Dest
Source2
1
1
1
1
1
0
0
0
0
I
–
PD
P2
P1
Source1
fsub
Dest
Source2
1
1
1
1
1
0
0
0
1
I
–
PD
P2
P1
Source1
fmpy
Dest
Source2
1
1
1
1
1
0
0
1
0
I
–
PD
P2
P1
Source1
fdiv
Dest
Source2
1
1
1
1
1
0
0
1
1
I
–
PD
P2
P1
Source1
frndx
Dest
1
1
1
1
1
0
1
0
0
I
–
PD
RM
P1
Source1
fcmp
Dest
1
1
1
1
1
0
1
0
1
I
–
P2
P1
Source1
fsqrt
Dest
1
1
1
1
1
0
1
1
1
I
–
P1
Source1
lmo
Dest
Source
1
1
1
1
1
1
0
0
0
–
–
–
–
–
–
–
–
–
–
–
–
–
rmo
Dest
Source
1
1
1
1
1
1
0
0
1
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
Source2
–
–
–
–
–
–
–
PD
–
–
estop
–
–
–
–
–
–
–
–
–
–
1
1
1
1
1
1
1
1
0
–
–
–
–
–
–
–
–
–
–
–
–
–
illopF
–
–
–
–
–
–
–
–
–
–
1
1
1
1
1
1
1
1
1
C
–
–
–
–
–
–
–
–
–
–
–
–
–
a
C
d
l
m
Mem Src / Dst
Dest
32
–
Reserved bit (code as 0)
Floating-point accumulator select
Constant operands rather than register
Destination precision for vector (0 = sp, 1 = dp)
Long immediate 32-bit data
Parallel memory operation specifier
Vector store or load source / dst register
Destination register
POST OFFICE BOX 1443
P
P1
P2
PD
RM
s
Z
Dest precision for parallel load / store (0 = single, 1 = double)
Precision of source1 operand
Precision of source2 operand
Precision of destination result
Rounding Mode (0 = N, 1 = Z, 2 = P, 3 = M)
Scale offset by data size
Use 0 rather than accumulator
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MP opcode summary (continued)
Table 7. Summary of MP Opcodes
INSTRUCTION
add
DESCRIPTION
INSTRUCTION
DESCRIPTION
Signed integer add
or.ff
Bitwise OR with 1s complement
and.tt
Bitwise AND
or.ft
Bitwise OR with 1s complement
and.ff
Bitwise AND with 1s complement
or.tf
Bitwise OR with 1s complement
and.ft
Bitwise AND with 1s complement
rdcr
Read control register
and.tf
Bitwise AND with 1s complement
rmo
Rightmost one
bbo
Branch bit one
shift.dz
Shift, disable mask, zero extend
bbz
Branch bit zero
shift.dm
Shift, disable mask, merge
bcnd
Branch conditional
shift.ds
Shift, disable mask, sign extend
Branch always
shift.ez
Shift, enable mask, zero extend
brcr
br
Branch control register
shift.em
Shift, enable mask, merge
bsr
Branch and save return
shift.es
Shift, enable mask, sign extend
cmnd
Send command
shift.iz
Shift, invert mask, zero extend
cmp
Integer compare
shift.im
Shift, invert mask, merge
dcache
Flush data cache sub-block
st
Store register into memory
estop
Emulation stop
sub
Signed integer subtract
fadd
Floating-point add
swcr
Swap control register
fcmp
Floating-point compare
trap
Trap
Floating-point divide
vadd
Vector floating-point add
fmpy
Floating-point multiply
vmac
Vector floating-point multiply and add to accumulator
frndx
Floating-point convert/round
vmpy
Vector floating-point multiply
fsqrt
Floating-point square root
vmsc
Vector floating-point multiply and subtract
from accumulator
fsub
FLoating-point subtract
vmsub
Vector floating-point subtract accumulator
from source
illop
fdiv
Illegal operation
vrnd(FP)
Vector round with floating-point input
jsr
Jump and save return
vrnd(Int)
Vector round with integer input
ld
Load signed into register
vsub
Vector floating-point subtract
ld.u
Load unsigned into register
xnor
Bitwise exclusive NOR
lmo
Leftmost one
xor
Bitwise exclusive OR
or.tt
Bitwise OR
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
33
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP architecture
The parallel processor (PP) is a 32-bit integer DSP optimized for imaging and graphics applications. Each PP
can execute in parallel; a multiply, ALU operation, and two memory accesses within a single instruction. This
internal parallelism allows a single PP to achieve over 500 million operations per second for certain algorithms.
The PP has a three-input ALU that supports all 256 three input Boolean combinations and many combinations
of arithmetic and Boolean functions. Data-merging and bit-to-byte, bit-to-word, and bit-to-halfword translations
are supported by hardware in the input data path to the ALU. Typical tasks performed by a PP include:
D
D
D
D
D
34
Pixel-intensive processing
–
Motion estimation
–
Convolution
–
PixBLTs
–
Warp
–
Histogram
–
Mean square error
Domain transforms
–
DCT
–
FFT
–
Hough
Core graphics functions
–
Line
–
Circle
–
Shaded fills
–
Fonts
Image Analysis
–
Segmentation
–
Feature extraction
Bit-stream encoding/decoding
–
Data merging
–
Table look-ups
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP functional block diagram
Figure 22 shows a block diagram of a parallel processor. Key features of the PP include:
D
D
D
D
D
D
D
64-bit instruction word (supports multiple parallel operations)
Three-stage pipeline for fast instruction cycle
Numerous registers
–
8 data, 10 address, 6 index registers
–
20 other user-visible registers
Data Unit
–
16x16 integer multiplier (optional dual 8x8)
–
Splittable 3-input ALU
–
32-bit barrel rotator
–
Mask generator
–
Multiple-status flag expander for translations to/from 1 bit-per-pixel space.
–
Conditional assignment of data unit results
–
Conditional source selection
–
Special processing hardware
Leftmost one / rightmost one
Leftmost bit change / rightmost bit change
Memory addressing
–
Two address units (global & local) provide up to two 32-bit accesses in parallel with data unit operation.
–
12 addressing modes (immediate and indexed)
–
Byte, halfword, and word addressability
–
Scaled indexed addressing
–
Conditional assignment for loads
–
Conditional source selection for stores
Program flow
–
Three hardware loop controllers
Zero overhead looping / branching
Nested loops
Multiple loop endpoints
–
Instruction cache management
–
PC mapped to register file
–
Interrupts for messages and context switching
Algebraic assembly language
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
35
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP functional block diagram (continued)
Data Unit
ALU Data Path
Multiplier
Data Path
d0 – d7
Expander
Mask Generator Barrel
Rotator
Three-Input ALU
mf & sr
Registers
Local Address Unit
Global Destination
Global Source
Local Destination / Source
a0 – a4, a7
x0 – x2
Global Address Unit
a8 – a12,
a15
sp = a6 = a14
Local Data Path
x8 – x10
Global Data Path
Program Flow Control Unit
Three Zero-Overhead
Loop / Branch Controllers
Repl
Instruction and Cache Control
Repl
32
A/S
A/S
64
Local
Data Port
Instruction
Port
Global
Data Port
Repl Replicate hardware
A/S Align/sign extend hardware
IAP Instruction address port
LAP Local address port
GAP Global address port
Figure 22. PP Block Diagram
36
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
32
IAP
LAP GAP
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP registers
The PP contains many general-purpose registers, status registers, and configuration registers. All PP registers
are 32-bit registers. Figure 23 shows the accessible registers of the PP blocks.
Data-Unit Registers
Data Registers
Multiple Flags
Status
d0 / EALU Operation
mf
sr
d1
d2
d3
d4
d5
d6
d7
Address-Unit Registers
Global-address Unit
Address Registers
Index Flags
Local-Address Unit
Address Registers
Index Flags
a8
x8
a0
x0
a9
x9
a1
x1
a10
x10
a2
x2
a11
a3
a12
a14 / sp
a15 = 0
PFC-Unit Registers
PC-Related Registers
a4
Stack Pointer
Same Physical
Register
a6/ sp
a7 = 0
Loop Addresses
Loop Counts
Communications
pc (br, call)
ls0
lr0
comm
iprs
ls1
lr1
Interrupts
ipa (read only)
ls2
lr2
lntflg
ipe (read only)
le0
lc0
inten
Cache Tags
le1
lc1
tag0 (read only)
le2
lc2
tag1 (read only)
tag2 (read only)
Loop Control
tag3 (read only)
lctl
Figure 23. PP Registers
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
37
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP data-unit registers
The data unit contains eight 32-bit general-purpose data registers (d0–d7) referred to as the D registers. The
d0 register also acts as the control register for EALU operations.
d0 register
Figure 24 shows the format when d0 is used as the EALU control register.
31
30
29
28
FMOD
FMOD
A
C
I
S
N
27
26
A
25
24
23
22
21
20
19
EALU Function Code
18
17
16
15
14
13
12
11
10
9
C
I
S
N
E
F
–
–
–
DMS
E
F
DMS
M
R
DBR
Function modifiers
Arithmetic enable
EALU carry-In
Invert-carry-In
Sign extend
Nonmultiple mask
8
7
6
M R
5
4
3
–
2
1
0
DBR
Explicit multiple carry-In
Expanded mf
Default multiply shift amount
Split multiply
Rounded multiply
Default barrel rotate amount
Figure 24. d0 Format for EALU Operations
multiple flags (mf) register
The mf register records status information from each split ALU segment for multiple arithmetic operations. The
mf register may be expanded to generate a mask for the ALU. Figure 25 shows the mf register format.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
3
2
1
0
1
0
Figure 25. mf Register Format
status register (sr)
The sr contains status and control bits for the PP ALU. Figure 26 shows the sr register format.
31
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
N C V
30
Z
LV
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
MSS
N
C
V
Z
LV
R
MSS
Negative status bit
Carry status bit
Overflow status bit
Zero status bit
Latched overflow
Rotation bit
Msize
Asize
7
6
R
5
4
Msize
Asize
mf Status selection
00 – set by zero
10 – set by extended result
01 – set by sign
11 – reserved
Expander data size
Split ALU data size
Figure 26. sr Format
PP address-unit registers
address registers
The address unit contains ten 32-bit address registers which contain the base address for address
computations or which can be used for general-purpose data. The registers a0 – a4 are used for local address
computations and registers a8–a12 are used for global-address computations.
38
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
index registers
The six 32-bit index registers contain index values for use with the address registers in address computations
or they can be used for general-purpose data. Registers x0 – x2 are used by the local-address unit and registers
x8 – x10 are used by the global-address unit.
stack pointer (sp)
The sp contains the address of the bottom of the PP’s system stack. The stack pointer is addressed as a6 by
the local-address unit and as a14 by the global-address unit. Figure 27 shows the sp register format.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
Word-Aligned Address
1
0
0
0
Figure 27. sp Register Format
zero register
The zero registers are read-as-zero address registers for the local address unit (a7) and global-address unit
(a15). Writes to the registers are ignored and can be specified when operational results are to be discarded.
Figure 28 shows the zero register format.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Figure 28. zero Registers
PP program flow control (PFC) unit registers
loop registers
The loop registers control three levels of zero-overhead loops. The 32-bit loop start registers (ls0 – ls2) and
loop-end registers (le0 – le2) contain the starting and ending addresses for the loops. The loop-counter registers
(lc0 – lc2) contain the number of repetitions remaining in their associated loops. The lr0 – lr2 registers are loop
reload registers used to support nested loops. The format for the loop-control (lctl) register is shown in Figure 29.
There are also six special write-only mappings of the loop-reload registers. The lrs0 – lrs2 codes are used for
fast initialization of lsn, lrn, and lcn registers for multi-instruction loops while the lrse0 – lrse2 codes are used
for single instruction-loop fast initialization.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
E
E
LCDn
10
Loop End Enable
Loop Counter Designator
000 – None
010 – lc1
001 – lc0
011 – lc2
1xx – reserved
9
LCD2
le2
8
7
E
6
5
LCD1
le1
4
3
E
2
1
0
LCD0
le0
Figure 29. lctl Register
pipeline registers
The PFC unit contains a pointer to each stage of the PP pipeline. The pc contains the program counter which
points to the instruction being fetched. The ipa points to the instruction in the address stage of the pipeline and
the ipe points to the instruction in the execute stage of the pipeline. The instruction pointer
return-from-subroutine (iprs) register contains the return address for a subroutine call. Figure 30 shows the
variable pipeline register format.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
39
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
pipeline registers (continued)
31
30
29
28
27
26
25
24
23
22
21
pc
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
PC (29-Bit Doubleword Address)
31
30
29
28
27
26
25
24
23
22
21
20
19
1
0
– G L
G – Global Interrupt Enable
ipa
2
L – Loop Inhibit
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
9
8
7
6
5
4
3
2
1
0
9
8
7
6
5
4
3
32-Bit Copy of the Previous pc Register Value
31
30
29
28
27
26
25
24
23
22
ipe
21
20
19
18
17
16
15
14
13
12
11
10
32-Bit Copy of the Previous ipa Register Value
31
30
29
28
27
26
25
24
23
22
iprs
21
20
19
18
17
16
15
14
13
12
11
10
29-Bit Doubleword Return Address
2
1
0
–
–
–
Figure 30. Pipeline Registers
interrupt registers
The interrupt-enable (inten) register allows individual interrupts to be enabled and configures the interrupt flag
(intflg) register operation. The intflg register contains the interrupt flag bits. Interrupt priority increases moving
from left to right on intflg. Figure 31 shows the PP-interrupt register format.
inten
intflg
31
30
29
28
r
r
r
r
27
26
25
24
23
22
21
E E E E
–
–
–
P P
P P
3 2
M M
S S
G G
P
P
1
M
S
G
P
P
0
M
S
G
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
E E E E
–
–
E
–
–
–
–
–
–
–
–
–
–
–
–
– W
M
P
M
S
G
P
T
E
N
D
P
T
E
R
R
P
T
Q
T
A
S
K
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
r
r
r
r
I
I
I
I
–
–
–
I
I
I
I
–
–
I
–
–
–
–
–
–
–
–
–
–
–
–
–
–
r
E
W
PPnMSG
Reserved (write as 0)
Enable Interrupt
Write Mode
0 – writing 1 clears intflg
1 – writing 1 sets intflg
PPn message Interrupt
MPMSG
PTEND
PTERR
PTQ
TASK
MP Message Interrupt
Packet Transfer Complete
Packet-Transfer Error
Packet Transfer Queued
MP Task Interrupt
Figure 31. PP-Interrupt Registers
40
0
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
communication (comm) register
The comm register contains the packet-transfer handshake bits and PP indicator bits. Figure 32 shows the
comm register format.
31
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
H S Q P
30
29
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
MAREV
PP#
High-priority packet transfer
Packet-transfer suspend
Packet transfer queued
Submit packet transfer request
Major revision #
Minor revision #
00 – 320C80
10 – 320C82
H
S
Q
P
MAREV
MIREV
DEV Device
MIREV
2
1
0
PP#
DEV
pp Number (read only)
000 – pp0
010 – pp2
001 – pp1
011 – pp3
1xx – not implemented
Figure 32. comm Register
cache-tag registers
The tag0 – tag3 registers contain the tag address and sub-block present bits for each cache block. Figure 33
shows the cache tag registers.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
23-Bit Tag Address
P
LRU
Present bit
Least-recently-used code
00 – mru
10 – next lru
01 – next mru 11 – lru
Sub-Block #
8
7
6
5
4
3
2
1
0
P
P
P
P
–
–
–
LRU
3
2
1
0
Figure 33. Cache Tag Registers
PP cache architecture
Each PP has its own 2K-byte instruction cache. Each cache is divided into four blocks and each block is divided
into four sub-blocks containing 16 64-bit instructions each. Cache misses cause one sub-block to be loaded
into cache. Figure 34 shows the cache architecture for one of the four sets in each cache. Figure 35 shows how
addresses map into the cache using the cache tags and address bits.
Block 0
tag 0 (Block 0)
Block 1
tag 1 (Block 1)
Block 2
tag 2 (Block 2)
Block 3
tag 3 (Block 3)
LRU
Sub-Blocks
NLRU
NMRU
MRU
LRU Stack
Figure 34. PP Cache Architecture
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
23-Bit Tag Value
8
7
sub
6
5
4
instruction
3
2
1
0
ignored
sub – sub-block
Figure 35. pc Register Cache-Address Mapping
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
41
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP parameter RAM
The parameter RAM is a, 2K-byte, on-chip RAM which contains PP-interrupt vectors, PP-requested TC task
buffers, and a general-purpose area. The parameter RAM does not use the cache memory. Figure 36 shows
the parameter RAM address map.
Suspended PT Parameters
(128 Bytes)
0x0100#000– 0x0100#07F
Reserved
(96 Bytes)
0x0100#080– 0x0100#0DF
Restricted for Operating System Use
(24 Bytes)
0x0100#0E0– 0x0100#0F7
Cache Fault Address
0x0100#0F8– 0x0100#0FB
PP-Linked List Start Address
0x0100#0FC – 0x0100#0FF
Off-Chip to Off-Chip PT Buffer
(128 Bytes)
0x0100#100– 0x0100#17F
Interrupt Vectors
(128 Bytes)
0x0100#180– 0x0100#1FF
General-Purpose RAM
(1524 Bytes Less Stack Size)
0x0100#200
Application-Dependent Boundary
Stack
0x0100#7F0
Stack State Information After Reset
(12 Bytes)
Stack Pointer After Reset
0x0100#7F4– 0x0100#7FF
# – PP Number
Figure 36. PP Parameter RAM
PP interrupt vectors
The PP interrupts and their vector addresses are shown in Table 8.
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁÁÁÁÁ
Table 8. PP-Interrupt Vectors
NAME
42
VECTOR
ADDRESS
INTERRUPT
TASK
0x0100#1B8
Task Interrupt
PTQ
0x0100#1C4
Packet Transfer Queued
PTERR
0x0100#1C8
Packet-Transfer Error
PTEND
0x0100#1CC
Packet-Transfer End
MPMSG
0x0100#1D0
MP Message
PP0MSG
0x0100#1E0
PP0 Message
PP1MSG
0x0100#1E4
PP1 Message
PP2MSG
0x0100#1E8
PP2 Message
PP3MSG
0x0100#1EC
PP3 Message
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP data-unit architecture
The data unit has independent data paths for the ALU and the multiplier, each with its own set of hardware
functions. The multiplier data path includes a 16 × 16 multiplier, a halfword swapper, and rounding hardware.
The ALU data path includes a 32-bit three-input ALU, a barrel rotator, mask generator, mf expander,
left/rightmost one and left/rightmost bit-change logic, and several multiplexers. Figure 37 shows the data-unit
block diagram.
src1 / src2 / dstc / 0
dst2
src3
src4
src4 / src2
0
src1 / 0x1
d0
Rotate Amount
Multiplexer
mf
dst / dst1
Mask Generator
Multiplexer
Expander
LMO, RMO,
LMBC, RMBC
Barrel Rotator
Mask
Generator
C Port
Multiplexer
Multiplier
(Splittable)
Barrel
Rotator Input
Sign Bit
Scale
Round
A
B
Three-Input ALU (Splittable)
Swap / Merge
C
ALU
Function
Code Logic
N, C, V, Z, LV mf
Legend:
src1
scr2
scr3
scr4
dst/dst1
Any register, D reg only for l/rmo, l/rmbc hardware
D reg or sometimes 15/32-bit immediate
D reg only
D reg only
Any register
dst2
dstc
0x1
0
d0
D reg only
D reg only (destination companion reg source)
Constant
Constant
5 LSBs of d0
Figure 37. Data Unit Block Diagram
The PP’s ALU can be split into one 32-bit ALU, two 16-bit ALUs, or four 8-bit ALUs. Figure 38 shows the multiple
arithmetic data flow for the case of a four 8-bit split of the ALU (called multiple-byte arithmetic). The ALU
operates as independent parallel ALUs where each ALU receives the same function code.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
43
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP data-unit architecture (continued)
32
Rotate
Clear
mf Register
4
Expander (Replicate)
8
8
8
8
sr(C)
A
B
C
C-Out
A
C-IN
8
B
C
C-Out
C-IN
Logic
A
C-IN
8
C, Z, S,
or E
C, Z, S,
or E
B
C
C-Out
C-IN
Logic
C-IN
A
8
C, Z, S,
or E
B
C
C-Out
C-IN
Logic
C-IN
C-IN
Logic
8
C, Z, S,
or E
Figure 38. Multiple-Byte Arithmetic Data Flow
PP multiplier
The PP’s hardware multiplier can perform one 16x16 multiply with a 32-bit result or two 8x8 multiplies with two
16-bit results in a single cycle. A 16x16 multiply can use signed or unsigned operands as shown in Figure 39.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
S
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
S
S
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Signed Input
14
13
12
11
10
11
10
9
8
7
6
5
4
3
2
1
0
9
8
7
6
5
4
3
2
1
0
5
4
3
2
1
0
Signed × Signed Result
15
14
13
12
Unsigned Input
15
14
13
12
11
10
9
8
7
6
Unsigned × Unsigned Result
Figure 39. 16 x 16 Multiplier Data Formats
When performing two simultaneous 8x8 split multiplies, the first input word contains unsigned byte operands
and the second input word contains signed or unsigned byte operands. These formats are shown in Figure 40
and Figure 41.
44
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP multiplier (continued)
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
S
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
1b × 2b Signed Result
S
15
14
13
12
11
10
9
8
7
6
Unsigned Input 1b
14
13
12
11
10
9
13
12
11
10
9
4
3
2
1
0
Unsigned Input 1a
8
Signed Input 2b
14
5
7
6
S
8
7
5
4
3
2
1
0
Signed Input 2a
6
5
4
3
2
1
0
4
3
2
1
0
1a × 2a Signed Result
S
Figure 40. Signed Split Multiply Data Formats
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
Unsigned Input 1b
15
14
13
12
11
10
9
8
7
6
Unsigned Input 2b
15
14
13
12
11
1b × 2b Unsigned Result
10
9
5
Unsigned Input 1a
5
4
3
2
1
0
Unsigned Input 2a
8
7
6
5
4
3
2
1
0
1a × 2a Unsigned Result
Figure 41. Unsigned Split Multiply Data Formats
PP program-flow-control unit architecture
The PP has a three-stage fetch, address, execute (FAE) pipeline as shown in Figure 42. The pc, ipa, and ipe
registers point to the address of the instruction in each stage of the pipeline. On each cycle in which the pipeline
advances, ipa is copied into ipe, pc is copied into ipa, and the pc is incremented by one instruction (8 bytes).
The program-flow-control (pfc) unit performs instruction fetching and decoding, loop control, and handshaking
with the transfer controller. The pfc unit architecture is shown in Figure 43.
pc
Instruction
One
Two
Three
T1
T2
T3
T4
Fetch
Address
Execute
Fetch
Address
Execute
Fetch
Address
T5
ipa
Execute
ipe
Figure 42. FAE-Instruction Pipeline
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
45
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP program-flow-control unit architecture (continued)
pc
lprs
incrementer
Cache Controller
ipa
Tag Comparators
ipe
Tag Registers
Present Bits
LRU Stack
Loop Controller 0
ls0
le0
lctl
Comparator
Instruction Decode
lr0
FAE Pipeline Control
decr.
Control Signal Generation
lc0
zero
Loop Control
Loop Controller 1
Instruction
Control
Signal
Instruction
Address
Loop Controller 2
Figure 43. Program-Flow-Control Unit Block Diagram
PP address-unit architecture
The PP has both a local- and global-address unit which operate independently of each other. The address units
support twelve different addressing modes. In place of performing a memory access, either or both of the
address units can perform an address computation that is written directly to a PP register instead of being used
for a memory access. This address-unit arithmetic provides additional arithmetic operation to supplement the
data unit during compute-intensive algorithms. Figure 44 shows the address-out architecture.
46
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP address-unit architecture (continued)
From Global
Destination Bus
Offset
To Global
Source Bus
From Global
Destination Bus
Offset
To Global
Source Bus
sp = a6 (local)
sp = a14 (global)
a0 – a4
(a7 = 0)
a8 – a12
(a15 = 0)
x0 – x2
pba dba
PP-Relative
Multiplexer
pba, dba
Index Multiplexer
Index Scaler
Scale
Data Size
32-Bit Adder / Subtracter Unit
Preindex / Postindex
Multiplexer
x8 – x10
Preindex / Postindex
PP-Relative
Multiplexer
Index Multiplexer
Index Scaler
Scale
Data Size
32-Bit Adder / Subtracter Unit
Preindex / Postindex
Multiplexer
Preindex / Postindex
Global-Address Port
Local-Address Port
Figure 44. Address-Unit Architecture
PP instruction set
PP instructions are represented by algebraic expressions for the operations performed in parallel by the
multiplier, ALU, global-address unit, and local-address unit. The expressions use the || symbol to indicate
operations that are to be performed in parallel. The PP ALU operator syntax is shown in Table 9. The data unit
operations (multiplier and ALU) are summarized in Table 10 and the parallel transfers (global and local) are
summarized in Table 11.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
47
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP instruction set (continued)
Table 9. PP Operators by Precedence
OPERATOR
48
FUNCTION
src1 [n] src1–1
Select odd (n=true) or even (n=false) register of D register pair
based on negative condition code
()
Subexpression delimiters
@mf
Expander operator
%
Mask generator
%%
Nonmultiple mask generator (EALU only)
%!
Modified mask generator (0xFFFFFFFF output for 0 input)
%%!
Nonmultiple shift right mask generator (EALU only)
\\
Rotate left
<<
Shift left (pseudo-op for rotate and mask)
>>u
Unsigned shift right
>> or >>s
Signed shift right
&
Bitwise AND
^
Bitwise XOR
|
Bitwise OR
+
Addition
–
Subtraction
=[cond]
Conditional assignment
=[cond.pro]
Conditional assignment with status protection
=
Equate
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP instruction set (continued)
Table 10. Summary of Data-Unit Operations
Operation
Base set ALUs
Description
Perform an ALU operation specifying ALU function, 2 src and 1 dest operand, and operand routing. ALU function is one of
256 three-input Boolean operations or one of 17 arithmetic operations combined with one of 15 function modifiers.
Syntax
dst = [fmod] [ [[cond [.pro] ]] ] ALU_EXPRESSION
Examples
d6 = (d6 ^ d4) & d2
d3 = [nn.nv] d1 –1
Operation
EALU || ROTATE
Description
Perform an extended ALU (EALU) operation (specified in d0) with one of two data routings to the ALU and optionally write
the barrel rotator output to a second dest register. ALU Function is one of 256 Boolean or 256 arithmetic.
Syntax
dst1 = [ [[cond [.pro] ]] ] ealu (src2, [dst2 = ] [ [[cond]] src1 [[[n]] src1–1] \\ src3, [%] src4)
dst1 = [fmod] [ [[cond [.pro] ]] ] ealu (label:EALU_EXPRESSION [ || dst2 = [[cond]] src1 [ [[n]] src1–1] \\ src3])
Examples
d7 = [nn] ealu(d2, d6 = [nn] d3\\d1, %d4)
d3 = mzc ealu(mylabel: d4 + (d5\\d6 & %d7) || d1 = d5\\d6)
Operation
MPY || ADD
Description
Perform a 16x16 multiply with optional parallel add or subtract. Condition code applies to both multiply and add.
Syntax
dst2 = [sign] [ [[cond]] ] src3 * src4 [ || dst = [ [[cond[.pro] ]] ] src2 + src1 [ [[n]] src1 –1] ]
dst2 = [sign] [ [[cond]] ] src3 * src4 [ || dst = [ [[cond[.pro] ]] ] src2 – src1 [ [[n]] src1 –1] ]
Example
d7 = u d6 * d5 || d5 = d4 – d1
Operation
MPY || SADD
Description
Perform a 16x16 multiply with a parallel right-shift and add or subtract. Condition code applies to multiply, shift, and add.
Syntax
dst2 = [sign] [ [[cond]] ] src3 * src4 || dst = [ [[ cond [.pro] ]] ] src2 + src1 [ [[n]] src1 –1] >> –d0
dst2 = [sign] [ [[cond]] ] src3 * src4 || dst = [ [[ cond [.pro] ]] ] src2 – src1 [ [[n]] src1 –1] >> –d0
Examples
d7 = u d6 * d5 || d5 = d4 – d1 >> –d0
Operation
MPY || EALU
Description
Perform a multiply and an optional parallel EALU. Multiply can use rounding, scaling, or splitting features.
Syntax
Generic Form:
dst2 = [sign] [ [[cond]] ] src3 * src4 || dst = [ [[cond [.pro] ]] ] ealu[f] (src2, src1 [ [[n]] src1 –1] \\ d0, %d0)
dst2 = [sign] [ [[cond]] ] src3 * src4 || ealu()
Explicit Form:
dst2 = [sign] [opt] [ [[cond]] ] src3 * src4 [<<dms] || dst1 = [fmod] [ [[cond [.pro] ]] ] ealu (label: EALU_EXPRESSION)
dst2 = [sign] [opt] [ [[cond]] ] src3 * src4 [<<dms] || ealu (label)
Examples
d7 = [p] d5 * d3 || d2 = [p] ealu(d1, d6\\d0, %d0)
d2 = m d4 * d7 || d3 = ealu (mylabel: d3 + d2 >> 9)
Operation
divi
Description
Perform one iteration of unsigned divide algorithm. Generates one quotient bit per execution using iterative subtraction.
Syntax
dst1 = [ [[cond [.pro] ]] ] divi (src2, dst2 = [[cond]] src1 [ [[n]] src1 –1])
Examples
d3 = divi (d1, d2 = d2)
d3 = divi (d1, d2 = d3[n]d2)
Legend:
[]
[[ ]]
pro
f
Optional parameter extension
Square brackets ([ ]) must be used
Protect status bits
Use 1s compliment of d0
cond
fmod
dms
sign
POST OFFICE BOX 1443
; generic form
; explicit form
Condition code
Function modifier
Default multiply shift amount
u = unsigned, s = signed
• HOUSTON, TEXAS 77251–1443
49
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP instruction set (continued)
Table 10. Summary of Data-Unit Operations (Continued)
Misc.
Operations
dint; eint; nop; dloop; eloop; qwait
Description
Globally disable interrupts; globally enable interrupts; do nothing in the data unit; globally enable looping; globally disable
looping; wait until comm register Q bit is zero.
Syntax
dint dloop
eint eloop
nop qwait
Legend:
[]
[[ ]]
pro
f
50
Optional parameter extension
Square brackets ([ ]) must be used
Protect status bits
Use 1s compliment of d0
POST OFFICE BOX 1443
cond
fmod
dms
sign
Condition code
Function modifier
Default multiply shift amount
u = unsigned, s = signed
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP instruction set (continued)
Table 11. Summary of Parallel Transfers
Operation
Load
Description
Transfer from memory into PP register
Syntax
dst = [sign] [size] [ [[cond]] ]* addrexp
dst = [sign] [size] [ [[cond]] ]* an.element
Examples
d3 = uh[n]* (a9++=[2])
d1 = * a2.sMY_ELEMENT
Operation
Store
Description
Transfer from PP register into memory
Syntax
* addrexp = [size] src [ [[n]] src–1]
* an.element = [size] src [ [[n]] src–1]
Examples
*––a2 = d3
*a9.sMY_ELEMENT = a3
Operation
Address unit arithmetic
Description
Compute address and store in PP register
Syntax
dst = [size] [ [[cond]] ] & * addrexp
dst = [size] [ [[cond]] ] & * an.element
Examples
d2 = &*(a3 + x0)
a1 = &*a9.sMY_ELEMENT
Operation
Move
Description
Transfer from PP register to PP register
Syntax
dst = [g] [ [[cond]] ] src
Examples
x2 = mf
d1 = g d3
Operation
Field extract move
Description
Transfer from PP register to PP register extracting and right-aligning one byte or halfword
Syntax
dst = [sign] [size item]
Example
d3 = ub2 d1
Operation
Field replicate move
Description
Transfer from PP register to PP register replicating the LSbyte or LShalfword to 32 bits
Syntax
dst = r [size] [[cond]] src
Example
d7 = rh d3
Legend:
[]
[[ ]]
g
item
Optional parameter extension
Square brackets ([ ]) must be used
Use global unit
0 = byte0/halfword0, 1 = byte1/halfword1, 2 = byte2, 3 = byte3
POST OFFICE BOX 1443
cond
sign
size
Condition code
u = unsigned, s = signed
b = byte, h = halfword, w = word (default)
• HOUSTON, TEXAS 77251–1443
51
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP opcode formats
A PP instruction uses a 64-bit opcode. The opcode is divided essentially into a data unit portion and a parallel
transfer portion. There are five data unit opcode formats comprising bits 38–63 of the opcode. Bits 0–38 of the
opcode specify one of 10 parallel transfer formats. An alphabetical list of the mnemonics used in Figure 45 for
the data unit and parallel transfer portions of the opcode are shown in Table 12 and Table 13, respectively.
Data Unit Formats
6 6 6 6 5 5 5 5 5 5 5 5 5 5 4 4 4 4 4 4 4 4 4 4 3 3 3 3 3 3 3 3 3 3 2
3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9
0 1 1
oper
dst1
src1
1
class
A
ALU Operation
dst
src1
0
1
class
A
ALU Operation
dst
src1
1 0 –
1
class
A
ALU Operation
dst
src1
1 1
src3
dst2
src4
1 0 0 0 1 – 0 – 0 – 0 – 0 – – – – – – 0
0 0
3 2 1 0
src2
Parallel Transfers
A. Six-Operand (MPYIIADD, etc.)
imm. src2
Parallel Transfers
B. Base Set ALU (5-Bit Immediate)
Parallel Transfers
C. Base Set ALU (Register src2)
src2
dstbank
s1bnk
cond
Operation
32-Bit Immediate
Parallel Transfers
D. Base Set ALU (32-Bit Immediate)
E. Miscellaneous
Reserved
0 1 0
Reserved
Transfer Formats
3 3 3 3 3 3 3 3 3 2 2 2 2 2 2 2 2 2
8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1
2 1 1 1 1 1 1 1 1 1 1
9 8 7 6 5 4 3 2
0 9 8 7 6 5 4 3 2 1 0
0
Lmode
d
e size s
La
Gim/X
reg
e size s
Ga
Lim/X
1. Double Parallel
Lmode
d
e size s
La
0 Lrm
dstbank
L 0 0 0 0
src
srcbank
dst
Lim/X
2. Move II Local
Lmode
d
e size s
La
0 Lrm
dstbank
L 0 0 0 1
src
e size D
dst
Lim/X
3. Field Move II Local
Lmode
reg
e size s
La
1 Lrm
bank
L 0 0
bank
L
0 0
Lmode
Global Long Offset /X
d
e size s
La
0 Lrm
L
0bank
Gmode
Local Long Offset / X
Gmode
reg
e size s
Ga
4. Local (Long Offset)
0
L 0 0 1 – – – –
As1bank
– – –
cond
c
r g N C V Z 0 – –
dstbank
– 0 0 0 0
src
srcbank
dst
–
0 0 –
cond
c
r g N C V Z 0
dstbank
– 0 0 0 1
src
e size D
dst
0 0 –
cond
c
r g N C V Z Gim/X
bank
L
reg
e size s
Ga
0 0 –
cond
c
r – N C V Z 0 – –
Adstbank
itm
Adstbank
L
0 0 –
Gmode
– 0 0 1 – – – –
As1bank
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
Grm
Lim/X
5. Global (Long Offset)
6. Non-D DU II Local
–
– 7. Conditional DU II Conditional Mode
–
–
– 8. Conditional DU II Conditional Field Move
1
Grm
– – – –
Figure 45. PP Opcode Formats
52
1
–
9. Conditional DU II Conditional Global
– 10. Conditional Non-D DU
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP opcode formats (continued)
Table 12. Data Unit Mnemonics
MNEMONIC
FUNCTION
A
A = 1 selects arithmetic operations, A = 0 selects Boolean operations
ALU Operation
For Boolean operation (A = 0) select the 8 ALU function signals. For arithmetic operation (A = 1), odd bits specify the ALU
function and even bits define the ALU function modifiers.
class
Operation class, determines routing of ALU operands
cond
condition code
dst
D register destination or lower 3 bits of non-D register code
dst1
ALU dest. for MPY||ADD, MPY||EALU, or EALU||ROTATE operation. D register or lower 3 bits of non-D register code
dst2
Multiply dest. for MPY||ADD or MPY||EALU operation or rotate dest. for EALU||ROTATE operation. D register
dstbank
ALU register bank
imm.src2
5-bit immediate for src2 of ALU operation
32-Bit Immediate
32-bit immediate for src2 of ALU operation
oper
Six-operand data unit operation (MPY||ADD, MPY||SADD, MPY||EALU, EALU||ROTATE, divi)
Operation
Miscellaneous operation
src1
ALU source 1 register code (D register unless srcbank or s1bank is used)
src2
D register used as ALU source 2
src3
D register for multiplier source (MPY||ADD or MPY||EALU) or rotate amount (EALU||ROTATE)
src4
D reg for ALU C port operand or EALU||ROTATE mask generator input or multiplier source 2 for MPY||ADD, MPY||EALU
s1bnk
Bits 5-3 of src1 register code (bit 6 assumed to be 0)
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
53
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP opcode formats (continued)
Table 13. Parallel Transfer Mnemonics
MNEMONIC
FUNCTION
0bank
Bits 5–3 of global transfer source/destination register code (bit 6 assumed to be 0)
Adstbnk
Bits 6–3 of ALU destination register code
As1bank
Bits 6–3 of ALU source 1 register code
bank
Bits 6–3 of global (or local) store source or load destination
c
Conditional choice of D register for src1 operand of the ALU
C
Protect status register’s carry bit
cond
Condition code
d
D register or lower 3 bits of register code for local transfer source/destination
D
Duplicate least significant data during moves
dst
The three lowest bits of the register code for move or field move destination
dstbank
Bits 6–3 of move destination register code
e
Sign extend local (bit 31), sign extend global (bit 9)
g
Conditional global transfer
Ga
Global address register for load, store, or address unit arithmetic
Gim / X
Global address unit immediate offset or index register
Gmode
Global unit addressing mode
Grm
Global PP-relative addressing mode
itm
Number of item selected for field extract move
L
L = 1 selects load operation, L = 0 selects store / address unit arithmetic operation
La
Local address register for load, store, or address unit arithmetic
Lim / X
Local address unit immediate offset or index register
Lmode
Local unit addressing mode
Lrm
Local PP-relative addressing mode
N
Protect status register’s negative bit
r
Conditional write of ALU result
reg
Register number used with bank or 0bank for global load, store, or address unit arithmetic
s
Enable index scaling. Additional index bit for byte accesses or arithmetic operations (bit 28, local; bit 6, global)
size
Size of data transfer (bits 30 – 29, local; bits 8 – 7, global)
src
Three lowest bits of register code for register-register move source or non-field moves. D register source for field move
srcbank
Bits 6–3 of register code for register-register move source
V
Protect status register’s overflow bit
Z
Protect status register’s zero bit
–
Unused bit (fill with 0)
54
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP opcode formats (continued)
Table 14 summarizes the supported parallel-transfer formats, their formats, and whether the transfers are local
or global. It also lists the allowed ALU operations and states whether conditions and status protection are
supported.
Table 14. Parallel-Transfer Format Summary
FORMAT
GLOBAL TRANSFER
ALU
OPERANDS
Move
LOCAL TRANSFER
Load / Store / AUA
Load / Store / AUA
dst1
src1
Cond
Status
Protection
Double parallel
D
D
No
No
Move || Local
D
D
No
No
Field move || Local
D
D
No
No
Global (long offset)
D
D
No
No
—
Any
X/long
Yes
—
src → dst
s/d
Index
Rel
s/d
Index
Rel
Port
—
Lower
X/short
No
D
X/short
No
Local
Any Any
—
—
—
D
X/short
Yes
Local
—
—
—
D
X/short
No
Local
—
—
—
D
³
³Any
Local (long offset)
D
D
No
No
—
—
—
—
Any
X/long
Yes
Global
Non-D DU || Local
Any
Any
No
No
—
—
—
—
D
X/short
Yes
Global
Conditional move
D
D
Yes
Yes
Any Any
—
—
—
—
—
—
—
Conditional field move
D
D
Yes
Yes
—
—
—
—
—
—
—
Conditional global
D
D
Yes
Yes
—
Any
X/short
Yes
—
—
—
—
—
—
—
—
—
—
—
—
Conditional non-D DU
Any
Any
Yes
Yes
32-bit imm. base ALU
Any
Lower
Yes
No
Legend:
DU
AUA
s/d
Rel
D
³
³Any
—
Data unit
Address unit arithmetic
Source/destination register
Relative addressing support
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
55
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PP opcode formats (continued)
Table 15 shows the encoding used in the opcodes to specify particular PP registers. A 3-bit register field
contains the three LSBs. The register codes are used for the src, src1, src2, src3, src4, dst, dst1, dst2, d, reg,
Ga, La, Gim/X, and Lim/X opcode fields. The four MSBs specify the register bank which is concatenated to the
register field for the full 7-bit code. The register bank codes are used for the dstbank, s1bnk, srcbank, 0bank,
bank, Adstbnk, and As1bank opcode fields. When no associated bank is specified for a register field in the
opcode, the D register bank is assumed. When the MSB of the bank code is not specified in the opcode (as in
0bank and s1bnk) it is assumed to be 0, indicating a lower register.
Table 15. PP Register Codes
LOWER REGISTERS (MSB OF BANK = 0)
CODING
BANK
REG
0000
000
0000
0000
REGISTER
CODING
BANK
REG
a0
0100
000
001
a1
0100
010
a2
0100
0000
011
a3
0000
100
0000
101
0000
110
0000
111
0001
000
0001
0001
UPPER REGISTERS (MSB OF BANK = 1)
REGISTER
CODING
BANK
REG
d0
1000
000
001
d1
1000
010
d2
1000
0100
011
d3
a4
0100
100
reserved
0100
101
a6 (sp)
0100
a7 (zero)
0100
a8
001
010
CODING
REGISTER
BANK
REG
reserved
1100
000
lc0
001
reserved
1100
001
lc1
010
reserved
1100
010
lc2
1000
011
reserved
1100
011
reserved
d4
1000
100
reserved
1100
100
lr0
d5
1000
101
reserved
1100
101
lr1
110
d6
1000
110
reserved
1100
110
lr2
111
d7
1000
111
reserved
1100
111
reserved
0101
000
reserved
1001
000
reserved
1101
000
lrse0
a9
0101
001
sr
1001
001
reserved
1101
001
lrse1
a10
0101
010
mf
1001
010
reserved
1101
010
lrse2
0001
011
a11
0101
011
reserved
1001
011
reserved
1101
011
reserved
0001
100
a12
0101
100
reserved
1001
100
reserved
1101
100
lrs0
0001
101
reserved
0101
101
reserved
1001
101
reserved
1101
101
lrs1
0001
110
a14 (sp)
0101
110
reserved
1001
110
reserved
1101
110
lrs2
0001
111
a15 (zero)
0101
111
reserved
1001
111
reserved
1101
111
reserved
0010
000
x0
0110
000
reserved
1010
000
reserved
1110
000
ls0
0010
001
x1
0110
001
reserved
1010
001
reserved
1110
001
ls1
0010
010
x2
0110
010
reserved
1010
010
reserved
1110
010
ls2
0010
011
reserved
0110
011
reserved
1010
011
reserved
1110
011
reserved
0010
100
reserved
0110
100
reserved
1010
100
reserved
1110
100
le0
0010
101
reserved
0110
101
reserved
1010
101
reserved
1110
101
le1
0010
110
reserved
0110
110
reserved
1010
110
reserved
1110
110
le2
0010
111
reserved
0110
111
reserved
1010
111
reserved
1110
111
reserved
0011
000
x8
0111
000
pc/call
1011
000
reserved
1111
000
reserved
0011
001
x9
0111
001
ipa/br
1011
001
reserved
1111
001
reserved
0011
010
x10
0111
010
ipe #
1011
010
reserved
1111
010
reserved
0011
011
reserved
0111
011
iprs
1011
011
reserved
1111
011
reserved
0011
100
reserved
0111
100
inten
1011
100
reserved
1111
100
tag0 #
0011
101
reserved
0111
101
intflg
1011
101
reserved
1111
101
tag1 #
0011
110
reserved
0111
110
comm
1011
110
reserved
1111
110
tag2 #
0011
111
reserved
0111
111
lctl
1011
111
reserved
1111
111
tag3 #
# Read only
56
REGISTER
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
data unit operation code
For data unit opcode format A, a 4-bit operation code specifies one of 16 six-operand operations and an
associated data path, as shown in Table 16.
Table 16. Six Operand Format Operation Codes
oper FIELD BIT
60
59
58
57
OPERATION TYPE
0
u
0
s
MPY || ADD
0
u
1
f
MPYU || EALU
1
0
f
k
EALU || ROTATE
1
0
1
0
divi
1
1
u
s
MPY || SADD
Legend:
u
Unsigned
f
1s complement EALU function code
s
Subtract
k
Use mask or mf expander
operation class code
The base set ALU opcodes (formats B, C, D) use an operation class code to specify one of eight different
routings to the A, B, and C ports of the ALU, as shown in Table 17.
Table 17. Base Set ALU Class Summary
CLASS
DESTINATION
A PORT
000
dst
src2
src1
001
dst
dstc
src1
010
dst
dstc
src1
011
dst
dstc
src1
\\
src2
100
dst
src2
src1
\\
d0
%d0
101
dst
src2
src1
\\
d0
@mf
110
dst
dstc
src1
111
dst
src1
1
\\
src2
Legend:
\\
@mf
%
dstc
dst
src2
srd1
B PORT
C PORT
@mf
\\
d0
src2
%src2
%src2
src2
src2
Rotate left
Expand function
Mask generation
Companion D reg
Destination Dreg or any reg if dstbank or Adstbnk is used with destination.
Source D reg or immediate
Source D reg or any if As1bank is used or any lower reg if s1bnk is used
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
57
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
ALU-operation code
For base-set ALU Boolean opcodes (A=0), the ALU function is formed by a sum of Boolean products selected
by the ALU operation opcode bits as shown in Table 18. For base-set arithmetic opcodes (A=1), the four odd
ALU operation bits specify an arithmetic operation as described in Table 19 while the four even bits specify one
of the ALU function modifiers as shown in Table 20.
Table 18. Base-Set ALU Boolean Function Codes
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
OPCODE BIT
PRODUCT TERM
58
A&B&C
57
~A & B & C
56
A & ~B & C
55
~A & ~B & C
54
A & B & ~C
53
~A & B & ~C
52
A & ~B & ~C
51
~A & ~B & ~C
Table 19. Base-Set Arithmetics
OPCODE BITS
CARRY
IN
ALGEBRAIC DESCRIPTION
NATURAL
FUNCTION
MODIFIED FUNCTION
(IF DIFFERENT FROM
NATURAL FUNCTION)
57
55
53
51
0
0
0
0
x
0
0
0
1
1
A – (B | C)
A – B <1<
0
0
1
0
0
A + (B & ~C)
A + B <0<
0
0
1
1
1
A–C
A–C
0
1
0
0
1
A – (B | ~C)
A – B >1>
0
1
0
1
1
A–B
A–B
0
1
1
0
C(n)
A – (B & @mf | –B & ~@mf)
A+B/A–B
if class 0 or 5
1/0
A + |B|
A+B/A–B
if class 1–4 or 6–7, A–B if sign=1
0
1
1
1
1
A – (B & C)
A – B>0>
1
0
0
0
0
A + (B & C)
A + B>0>
1
0
0
1
~C(n)
A + (B & @mf | –B & ~@mf)
A–B/A+B
if class 0 or 5
0/1
A – |B|
A–B/A+B
if class 1–4 or 6–7, A+B if sign=1
1
0
1
0
0
A+B
A+B
1
0
1
1
0
A + (B | ~C)
A + B >1>
1
1
0
0
0
A+C
A+C
1
1
0
1
1
A – (B & ~C)
A – B <0<
1
1
1
0
0
A + (B | C)
A + B <1<
1
1
1
1
0
(A & C) + (B & C)
field A + B
Legend:
C(n)
>0>
<0<
>1>
<1<
58
(A – (B & C)) if sign=0
LSB of each part of C port register
Zero-extend shift right
Zero-extend shift left
One-extend shift right
One-extend shift left
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
(A + (B & C)) if sign=0
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
ALU-operation code (continued)
Table 20. Function Modifier Codes
FUNCTION
MODIFIER BITS
MODIFICATION PERFORMED
58
56
54
52
0
0
0
0
Normal operation
0
0
0
1
cin
0
0
1
0
%! if maskgen instruction, lmo if not maskgen
0
0
1
1
%! and cin if maskgen instruction, rmo if not maskgen
0
1
0
0
A port = 0
0
1
0
1
A port = 0 and cin
0
1
1
0
A port = 0 and %! if maskgen, lmbc if not maskgen
0
1
1
1
A port = 0, %! and cin if maskgen, rmbc if not maskgen
1
0
0
0
mf bit(s) set by carry out(s). (mc)
1
0
0
1
mf bit(s) set based on status register MSS field. (me)
1
0
1
0
Rotate mf by Asize, mf bit(s) set by carry out(s). (mrc)
1
0
1
1
Rotate mf by Asize, mf bit(s) set based on status register MSS field. (mre)
1
1
0
0
Clear mf, mf bit(s) set by carry out(s). (mzc)
1
1
0
1
Clear mf, mf bit(s) set based on status register MSS field. (mze)
1
1
1
0
No setting of bits in mf register. (mx)
1
1
1
1
Reserved
Legend:
cin
lmbc
lmo
Carry in from sr(C)
Leftmost-bit change
Leftmost one
%!
rmbc
rmo
Modified mask generator
Rightmost-bit change
Rightmost one
miscellaneous operation code
For data-unit opcode format E, the operation field selects one of the miscellaneous operations codes as shown
in Table 21.
Table 21. Miscellaneous Operation Codes
OPCODE BITS
MNEMONIC
OPERATION
0
nop
No data-unit operation. Status not modified
0
1
qwait
1
0
eint
Global-interrupt enable
0
1
1
dint
Global-interrupt disable
0
1
0
0
eloop
Global loop enable
0
1
0
1
dloop
Global loop disable
0
0
1
1
x
reserved
0
1
x
x
x
reserved
1
x
x
x
x
reserved
43
42
41
40
39
0
0
0
0
0
0
0
0
0
0
0
0
0
0
POST OFFICE BOX 1443
Wait until comm Q bit is clear
• HOUSTON, TEXAS 77251–1443
59
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
addressing-mode codes
The Lmode (bits 35–38) and Gmode (bits 13–16) of the opcode specify the local and global transfer for various
parallel transfer opcode formats (Lmode in formats 1, 2, 3, 4, and 6 and Gmode in formats 1, 5, and 9). Table 22
shows the coding for the addressing-mode fields.
Table 22. Addressing-Mode Codes
CODING
EXPRESSION
DESCRIPTION
00xx
Legend:
an
imm
xm
Nop (nonaddressing mode operation)
0100
*(an ++= xm)
Postaddition of index register, with modify
0101
*(an – – = xm)
Postsubtraction of index register, with modify
0110
*(an ++= imm)
Postaddition of immediate, with modify
Postsubtraction of immediate, with modify
0111
*(an – – = imm)
1000
*(an + xm)
Preaddition of index register
1001
*(an – xm)
Presubtraction of index register
1010
*(an + imm)
Preaddition of immediate
1011
*(an – imm)
Presubtraction of immediate
1100
*(an += xm)
Preaddition of index register, with modify
1101
*(an – = xm)
Presubtraction of index register, with modify
1110
*(an += imm)
Preaddition of immediate, with modify
1111
*(an – = imm)
Presubtraction of immediate, with modify
Address register in l/g address unit
Immediate offset
Index register in same unit as an register
L, e codes
The L and e bits combine to specify the type of parallel transfer performed, as shown in Table 23. For the local
transfer, L and e are bits 21 and 31, respectively. For the global transfer, L and e are bits 17 and 9, respectively.
Table 23. Parallel Transfer Type
L
e
1
0
Zero-extend load
PARALLEL TRANSFER
1
1
Sign-extend load
0
0
Store
0
1
Address unit arithmetic
size codes
The size code specifies the data transfer size. For field moves (parallel transfer format 3), only byte and halfword
data sizes are valid, as shown in Table 24.
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
ÁÁÁÁÁ
ÁÁÁÁÁÁ
Table 24. Transfer Data Size
CODING
60
DATA SIZE
00
Byte (8 bits)
01
Halfword (16 bits)
10
Word (32 bits)
11
Reserved
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
relative-addressing mode codes
The Lrm and Grm opcode fields allow the local-address or global-address units, respectively, to select
PP-relative addressing as shown in Table 25.
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁ
ÁÁÁÁ
ÁÁÁÁÁÁÁÁÁÁ
Table 25. Relative-Addressing Mode Codes
CODING
RELATIVE-ADDRESSING
MODE
00
Normal (absolute addressing)
01
Reserved
10
PP-relative dba
11
PP-relative pba
Legend:
dba – Data RAM 0 base is base address
pba – Paramater RAM base is base address
condition codes
In the four conditional parallel transfer opcodes (formats 7–10), the condition code field specifies one of 16
condition codes to be applied to the data-unit operation source, data-unit result, or global transfer based on the
setting of the c, r, and g bits, respectively. Table 26 shows the condition codes. For the 32-bit immediate data
unit opcode (format D), the condition applies to the data-unit result only.
Table 26. Condition Codes
CONDITION
BITS
MNEMONIC
DESCRIPTION
STATUS BIT COMBINATION
35
34
33
32
0
0
0
0
u
Unconditional (default)
None
0
0
0
1
p
Positive
~N & ~Z
0
0
1
0
ls
Lower than or same
~C | Z
0
0
1
1
hi
Higher than
C & ~Z
0
1
0
0
lt
Less than
(N & ~V) | (~N & V)
0
1
0
1
le
Less than or equal
(N & ~V) | (~N & V) | Z
0
1
1
0
ge
Greater than or equal
(N & V) | (~N & ~V)
0
1
1
1
gt
Greater than
(N & V & ~Z) | (~N & ~V & ~Z)
1
0
0
0
hs, c
Higher than or same, carry
C
1
0
0
1
lo, nc
Lower than, no carry
~C
1
0
1
0
eq, z
Equal, zero
Z
1
0
1
1
ne, nz
Not equal, not zero
~Z
1
1
0
0
v
Overflow
V
1
1
0
1
nv
No overflow
~V
1
1
1
0
n
Negative
N
1
1
1
1
nn
Nonnegative
~N
EALU operations
Extended ALU (EALU) operations allow the execution of more advanced ALU functions than those specified
in the base set ALU opcodes. The opcode for EALU instructions contains the operands for the operation while
the d0 register extends the opcode by specifying the EALU operation to be performed. The format of d0 for EALU
operations is shown in Figure 24.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
61
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
EALU Boolean functions
EALU operations support all 256 Boolean ALU functions plus the flexibility to add 1 or a carry-in to Boolean sum.
The Boolean function performed by the ALU are shown below and in Table 27.
(F0 & (~A & ~B & ~C)) | (F1 & (A & ~B & ~C)) | (F2 & (~A & B & ~C))
| (F3 & (A & B & ~C)) | (F4 & (~A & ~B & C)) | (F5 & (A & ~B & C))
| (F6 & (~A & B & C)) | (F7 & (A & B & C)) [+1 | +cin]
Table 27. EALU Boolean Function Codes
d0 BIT
ALU FUNCTION SIGNAL
26
F7
PRODUCT TERM
A&B&C
25
F6
~A & B & C
24
F5
A & ~B & C
23
F4
~A & ~B & C
22
F3
A & B & ~C
21
F2
~A & B & ~C
20
F1
A & ~B & ~C
19
F0
~A & ~B & ~C
EALU arithmetic functions
EALU operations support all 256 arithmetic functions provided by the three-input ALU plus the flexibility to add
1 or a carry-in to the result. The arithmetic function performed by the ALU is:
f(A,B,C) = A & f1(B,C) + f2(B,C) [+1 | cin]
f1(B,C) and f2(B,C) are independent Boolean combinations of the B and C ALU inputs. The ALU function is
specified by selecting the desired f1 and f2 subfunction and then XORing the f1 and f2 code from Table 28 to
create the ALU function code for bits 19–26 of d0. Additional operations such as absolute values and signed
shifts can be performed using d0 bits which control the ALU function based on the sign of one of the inputs.
Table 28. ALU f1(B,C) and f2(B,C) Subfunctions
62
f1
CODE
f2
CODE
00
00
0
Zero the term
AA
FF
–1
–1 (All 1s)
88
CC
B
B
SUBFUNCTION
COMMON USAGE
22
33
–B –1
Negate B
A0
F0
C
C
0A
0F
–C –1
Negate C
80
C0
B&C
Force bits in B to 0 where bits in C are 0
2A
3F
–(B & C) – 1
Force bits in B to 0 where bits in C are 0 and negate
A8
FC
B|C
Force bits in B to 1 where bits in C are 1
02
03
–(B | C) – 1
Force bits in B to 1 where bits in C are 1 and negate
08
0C
B & ~C
Force bits in B to 0 where bits in C are 1
A2
F3
–(B & ~C) –1
Force bits in B to 0 where bits in C are 1 and negate
8A
CF
B | ~C
Force bits in B to 1 where bits in C are 0
20
30
–(B | ~C) –1
Force bits in B to 1 where bits in C are 0 and negate
28
3C
(B & ~C) | ((–B – 1) & C)
Choose B if C = all 0s and –B if C = all 1s
82
C3
(B & C) | ((–B – 1) & ~C)
Choose B if C = all 1s and –B if C = all 0s
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
video controller architecture
The video controller (VC) provides a method for handling the video or graphics capture, or display portions of
a TMS320C80 system. It provides simultaneous control over two independent capture or display systems and
frame grabber or frame buffer image storage.
VC functional block diagram
Figure 46 shows a functional block diagram of the video controller. Key features of the VC include:
D
D
D
Dual-frame timers
–
Independent or locked operation
–
Programmable horizontal and vertical timing
–
Separate or composite sync and blanking control
–
Synchronization to external timing signals
–
Interlaced or noninterlaced frame control
–
Virtually limitless screen resolutions
Programmable timing and control registers
Programmable line interrupt to MP
Shift register transfer (SRT) controller
–
Generates VRAM serial register transfer requests to the TC
–
Tracks VRAM tap point and schedules midline reloads
–
Generates packet-transfer requests for DRAM-based buffer updates
–
Supports two display or capture buffers
On-Chip Register Bus (MP)
32
VC Register Interface
HSYNC0
FT0 Events
Frame
Timer
0
VC Request (to TC)
VSYNC0
CSYNC0 / HBLNK0
CBLNK0 / VBLNK0
CAREA0
SRT Controller
MUX
D
HSYNC1
Frame
Timer
1
FT1 Events
SCLK1
SCLK0
VSYNC1
CSYNC1 / HBLNK1
CBLNK1 / VBLNK1
CAREA1
FCLK1
FCLK0
Figure 46. VC Block Diagram
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
63
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
frame-timer registers
Each frame timer has twenty-one 16-bit registers to control its horizontal and vertical timing signals. The
registers are on-chip memory-mapped registers accessible by the MP only. Each horizontal / vertical register
pair can be accessed as a single 32-bit quantity. The register map for Frame-Timer 0 is shown in Figure 47. The
Frame-Timer 1 register map is shown in Figure 48.
Address
Address
FTCTL0
0x01820200
SETVCT0
0x01820206
SETHCT0
0x01820204
VFTINT0
0x0182020A
HESERR0
0x01820208
VESYNC0
0x0182020E
HESYNC0
0x0182020C
VEBLNK0
0x01820212
HEBLNK0
0x01820210
VSAREA0
0x01820216
HSAREA0
0x01820214
VEAREA0
0x0182021A
HEAREA0
0x01820218
VSBLNK0
0x0182021E
HSBLNK0
0x0182021C
VTOTAL0
0x01820222
HTOTAL0
0x01820220
HALINE0
0x01820224
HBLINE0
0x01820228
HCOUNT0
0x0182023C
VCOUNT0
0x0182023E
Figure 47. Frame-Timer 0 Register Map
Address
Address
FTCTL1
0x01820240
SETVCT1
0x01820246
SETHCT1
0x01820244
VFTINT1
0x0182024A
HESERR1
0x01820248
VESYNC1
0x0182024E
HESYNC1
0x0182024C
VEBLNK1
0x01820252
HEBLNK1
0x01820250
VSAREA1
0x01820256
HSAREA1
0x01820254
VEAREA1
0x0182025A
HEAREA1
0x01820258
VSBLNK1
0x0182025E
HSBLNK1
0x0182025C
VTOTAL1
0x01820262
HTOTAL1
0x01820260
HALINE1
0x01820264
HBLINE1
0x01820268
HCOUNT1
0x0182027C
VCOUNT1
0x0182027E
Figure 48. Frame-Timer 1 Register Map
frame-timer register programming
The register format for the frame-timer control registers is shown in Figure 49. All other registers are 16-bit
values. For programming details, see the TMS320C80 Video Controller User’s Guide (literature number
SPRU111).
64
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
frame-timer control (FTCTLx) register
The FTCTLx register contains mode bits to determine frame-timer behavior.
15
14
13
FTE
IFD
IIM
FTE
IFD
IIM
SSE
FLE
CPM
12
11
10
9
8
SSE
FLE
7
6
5
4
3
CPM
Frame-timer enable
Interlaced frame disable
Interlace interrupt mode
Set synchronization enable
Frame lock enable
CSYNC / HBLNK pin mode
00 – CSYNC Hi-Z
10 – CSYNC output
11 – HBLNK output
01 – CSYNC input
VPM
HPM
2
1
VPM
VSYNC Pin Mode
00 – Hi-Z
01 – Input
HSYNC Pin Mode
00 – Hi-Z
01 – Input
0
HPM
10 – Output
11 – Reserved
10 – Output
11 – Reserved
Figure 49. FTCTLx Register
SRT controller registers
The SRT controller has two sets of 32-bit registers, one for each of the supported frame memory regions. The
location of these registers in on-chip memory-mapped register space is shown in Figure 50.
Address
Address
FMEMCTL0
0x01820300
FMEMCTL1
0x01820340
F1STADR0
0x01820304
F1STADR1
0x01820344
F0STADR0
0x01820308
F0STADR1
0x01820348
LINEINC0
0x0182030C
LINEINC1
0x0182034C
SAMMASK0
0x01820310
SAMMASK1
0x01820350
NEXTADR0
0x01820314
NEXTADR1
0x01820354
CRNTADR0
0x0182033C
CRNTADR1
0x0182037C
Figure 50. SRT Controller Register Map
SRT controller register programming
The register format for the frame memory control registers is shown in Figure 51. All other registers are 32-bit
values. For programming details, see the TMS320C80 Video Controller User’s Guide (literature number
SPRU111).
FMEMCTLx Register
The frame memory control (FMEMCTLx) register contains mode bits to determine operation of the associated
frame memory.
31
30
29
HSS
ILR
PTS
TMS
28
27
26
25
24
23
Half-SAM select
Interlaced line repeat
Packet transfer select
Transfer mode select
00 – Display
01 – Reserved
22
21
20
19
18
17
16
15
14
13
12
H
S
S
I
L
R
P
T
S
EMS
10 – Capture
11 – Merge capture
UED
FTS
11
10
9
8
TMS
7
6
5
EMS
4
3
U
E
D
2
1
0
FTS
Event mode select
00 – sof, line, sam
10 – sof, line
01 – sof, eof, sam
11 – none
Unblanked event disable
Frame timer sequencer
00 – ft0 / disabled
10 – ft1 / disabled
01 – ft0 / enabled
11 – ft1 / enabled
Figure 51. FMEMCTLx Register
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
65
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
TC architecture
The transfer controller (TC) is a combined memory controller and DMA (direct memory access) machine. It
handles the movement of data within the ’C80 system as requested by the master processor, parallel
processors, video controller, and external devices. The transfer controller performs the following data
movement and memory control functions:
D
D
D
D
D
D
D
D
D
MP and PP instruction cache fills
MP data cache fills and dirty block write-back
MP and PP direct external accesses (DEAs)
MP and PP packet transfers
Externally initiated packet transfers (XPTs)
VC packet transfers (VCPTs)
VC shift register transfers (SRTs)
DRAM/ SDRAM refresh
Host bus request
TC functional block diagram
Figure 52 shows a functional block diagram of the transfer controller. Key features of the TC include:
D
D
Crossbar interface
–
64-bit data path
–
Single-cycle access
External memory interface
–
4G-Byte address range
–
Dynamically configurable memory cycles
8-, 16-, 32-, or 64-bit bus size
Selectable memory page size
Selectable address multiplexing
Selectable cycle timing
–
D
D
D
66
Big- or little-endian operation
Cache, VRAM, refresh controller
–
Programmable refresh rate
–
VRAM block write support
Independent Src and Dst addressing
–
Autonomous addressing based on packet-transfer parameters
–
Data read and write at different rates
–
Numerous data merging and alignment functions performed during transfer
Intelligent request prioritization
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
TC functional block diagram (continued)
Packet Transfer
FIFO
Src MUX and
Alignment
Dst MUX and
Alignment
Cache Buffer
64
Crossbar
Interface
External
Memory
Interface
64
64
Src
Controller
64
Dst
Controller
Cache, VRAM, and
Refresh Controller
Src Control
Registers
Dst Control
Registers
Request Queuing and Prioritization
MP
Requests
PP
Requests
VC
Requests
XPT
Requests
Host
Requests
Figure 52. TC Block Diagram
TC registers
The TC contains four on-chip memory-mapped registers accessible by the MP. TC registers are shown in
Figure 53.
refresh control (REFCNTL) register (0x01820000)
The REFCNTL register controls refresh cycles.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
RPARLD
RPARLD
8
7
6
5
4
3
2
1
0
REFRATE
Refresh Pseudo-Address Reload Value
REFRATE
Refresh Interval (in clock cycles)
Figure 53. REFCNTL Register
packet-transfer minimum (PTMIN) register (0x01820004)
The PTMIN register determines the minimum number of cycles that a packet transfer executes before being
suspended by a higher priority packet transfer. Figure 54 shows the PTMIN register.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
PTMIN
Figure 54. PTMIN Register
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
67
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PT maximum (PTMAX) register (0x01820008)
The PTMAX register determines the maximum number of cycles after PTMIN has elapsed that a packet transfer
executes before timing out. Figure 55 showns the format of the PTMAX register.
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
PTMAX
Figure 55. PTMAX Register
fault status (FLTSTS) register (0x0182000C)
The FLTSTS register indicates the cause of a memory access fault. Fault status bits are cleared by writing a
1 to the appropriate bit. Figure 56 shows the format of the fault status (FLTSTS) register.
31
30
29
28
PP #
PC
PP
27
26
25
24
P
C
P
C
P
C
P
C
3
2
1
0
23
22
21
20
PP#
19
18
17
16
P
P
P
P
P
P
P
P
3
2
1
0
PPx Cache / DEA Fault
PPx Packet-Transfer Fault
15
14
13
12
11
10
9
8
7
6
XPT
XPT
M
5
4
3
2
1
0
M
Faulting XPT
MP Packet-Transfer Fault
Figure 56. FLTSTS Register
packet-transfer parameters
The most efficient method for data movement in a TMS320C80 system is through the use of packet transfers
(PTs). Packet transfers allow the TC to move blocks of data autonomously between a specified src and dst
memory region. Requests for the TC to execute a packet transfer may be made by the MP, PPs, VC, or external
devices. A packet-transfer parameter table describing the data packet and how it is to be transferred must be
programmed in on-chip memory before the transfer is requested. The parameter table formats for long-form and
short-form packet transfers are shown in Figure 57.
68
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
packet-transfer parameters (continued)
Byte
Address
PT
Long-Form Parameter Table
31
0
Word
Number
Byte
Address
Short-Form Parameter Table
31
0
Next Entry Address
0
PT
PT+4
PT Options
1
PT+4
PT+8
Src Start Address
0
PT+8
Src Start Address
0
Dst Start Address
Dst Start Address
1
PT+12
Next Entry Address
Word
Number
PT Options
Count
0
1
1
PT+12
PT+16
Src B Count
Src A Count
0
PT+20
Dst B Count
Dst A Count
1
PT - 16-byte aligned on-chip starting address of
parameter table.
PT+24
Src C Count
0
PT+28
Dst C Count
1
PT+32
Src B Pitch
0
PT+36
Dst B Pitch
1
PT+40
Src C Pitch
0
PT+44
Dst C Pitch
1
†PT+48
Transparency/Color Word 0
0†
†PT+52
Transparency/Color Word 1
1†
PT+56
Don’t Care
0
PT+60
Don’t Care
1
PT - 64-byte aligned on-chip starting address of
parameter table.
† These words are swapped in big-endian mode.
Figure 57. Packet-Transfer Parameter Table
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
69
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PT-options field
The PT-options field of the parameter table controls the type of src and dst transfer that the TC performs. The
formats of the options field for long-form and short-form packet transfers are shown in Figure 58.
3
1
S
30
29
PTS
S
PTS
I
RDC
RDB
RA
RSC
RSB
SF
X
ÉÉÉ
ÉÉÉ
ÉÉÉ
ÉÉÉ
28
I
27
26
Long-Form PT Options
25
24
23
22
21
20
19
R
D
C
R
D
B
R
A
R
S
C
R
S
B
S
F
X
18
17
16
PAM
ÉÉ
ÉÉ
ÉÉ
ÉÉ
15
14
13
STM
PAM
Stop Bit
PT Status
00 – Active
10 – Fault on Src
01 – Suspended
11 – Fault on Dst
Stop Bit
Reverse Dst C Addressing
Reverse Dst B Addressing
Reverse A Addressing
Reverse Src C Addressing
Reverse Src B Addressing
Short Form
Exchange Src and Dst Parameters
12
STM / DTM
SUM / DTM
ÉÉÉ
ÉÉÉ
ÉÉÉ
ÉÉÉ
11
10
9
8
SUM
ÉÉ
ÉÉ
ÉÉ
ÉÉ
7
6
5
4
DTM
ÉÉÉ
ÉÉÉ
ÉÉÉ
ÉÉÉ
3
2
1
0
DUM
PT Access Mode
000 – Normal
100 – 8-Bit Transparency
001 – PDT
101 – 16-Bit Transparency
010 – Block Write
110 – 32-Bit Transparency
011 – SRT
111 – 64-Bit Transparency
Src / Dst Transfer Mode
000 – Dimensioned 100 – Variable, Delta-Guided
001 – Fill (Src Only) 101 – Variable,
Offset-Guided
010 – Reserved
110 – Fixed, Delta-Guided
011 – LUT (Src Only) 111 – Fixed, Offset-Guided
Src / Dst Update Mode
00 – None
01 – Add C Pitch
01 – Add B Pitch
11 – Add C Pitch / Reverse
Short-Form PT Options
3
1
S
30
29
PTS
S
PTS
I
RA
28
I
ÉÉÉÉÉÉÉÉ
ÉÉÉÉÉÉÉÉ
27
26
25
24
Stop Bit
PT Status
00 – Active
01 – Suspended
Stop Bit
Reverse Addressing
23
R
A
22
21
20
19
S
F
X
18
17
16
15
14
13
12
11
PAM
10 – Fault on Src
11 – Fault on Dst
POST OFFICE BOX 1443
9
8
7
6
5
4
3
Count
RA
X
PAM
Reverse Addressing
Exchange Src & Dst Parameters
PT Access Mode
000 – Normal
100 – Reserved
001 – PDT
101 – Reserved
010 – Reserved
110 – Reserved
011 – SRT
111 – Reserved
Figure 58. PT-Options Field
70
10
• HOUSTON, TEXAS 77251–1443
2
1
0
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
LOCAL MEMORY INTERFACE
status codes
Status codes are output on STATUS[5:0] to describe the cycle being performed. During row time, the
STATUS[5:0] pins indicate the type of cycle being performed. The cycle type can be latched using RL or RAS
and used by external logic to perform memory bank decoding or to enable special hardware features. During
column time, the STATUS[5:0] pins indicate the requesting processor or special column information. See
Table 29 for a listing of the Row-time status codes and Table 30 for a listing of column-time status codes.
Table 29. Row-Time Status Codes
STATUS[5:0]
CYCLE TYPE
STATUS[5:0]
CYCLE TYPE
0 0 0 0 0
0
Normal Read
1 0 0 0 0
0
Reserved
0 0 0 0 0
1
Normal Write
1 0 0 0 0
1
Reserved
0 0 0 0 1
0
Refresh
1 0 0 0 1
0
Reserved
0 0 0 0 1
1
SDRAM DCAB
1 0 0 0 1
1
Reserved
0 0 0 1 0
0
Peripheral Device PT Read
1 0 0 1 0
0
XPT1 Read
0 0 0 1 0
1
Peripheral Device PT Write
1 0 0 1 0
1
XPT1 Write
0 0 0 1 1
0
Reserved
1 0 0 1 1
0
XPT1 PDPT Read
0 0 0 1 1
1
Reserved
1 0 0 1 1
1
XPT1 PDPT Write
0 0 1 0 0
0
Reserved
1 0 1 0 0
0
XPT2 Read
0 0 1 0 0
1
Block Write PT
1 0 1 0 0
1
XPT2 Write
0 0 1 0 1
0
Reserved
1 0 1 0 1
0
XPT2 PDPT Read
0 0 1 0 1
1
Reserved
1 0 1 0 1
1
XPT2 PDPT Write
0 0 1 1 0
0
SDRAM MRS
1 0 1 1 0
0
XPT3 Read
0 0 1 1 0
1
Load Color Register
1 0 1 1 0
1
XPT3 Write
0 0 1 1 1
0
Reserved
1 0 1 1 1
0
XPT3 PDPT Read
0 0 1 1 1
1
Reserved
1 0 1 1 1
1
XPT3 PDPT Write
0 1 0 0 0
0
Frame 0 Read Transfer
1 1 0 0 0
0
XPT4 / SAM1 Read
0 1 0 0 0
1
Frame 0 Write Transfer
1 1 0 0 0
1
XPT4 / SAM1 Write
0 1 0 0 1
0
Frame 0 Split Read Transfer
1 1 0 0 1
0
XPT4 / SAM1 PDPT Read
0 1 0 0 1
1
Frame 0 Split Write Transfer
1 1 0 0 1
1
XPT4 / SAM1 PDPT Write
0 1 0 1 0
0
Frame 1 Read Transfer
1 1 0 1 0
0
XPT5 / SOF1 Read
0 1 0 1 0
1
Frame 1 Write Transfer
1 1 0 1 0
1
XPT5 / SOF1 Write
0 1 0 1 1
0
Frame 1 Split Read Transfer
1 1 0 1 1
0
XPT5 / SOF1 PDPT Read
0 1 0 1 1
1
Frame 1 Split Write Transfer
1 1 0 1 1
1
XPT5 / SOF1 PDPT Write
0 1 1 0 0
0
Reserved
1 1 1 0 0
0
XPT6 / SAM0 Read
0 1 1 0 0
1
Reserved
1 1 1 0 0
1
XPT6 / SAM0 Write
0 1 1 0 1
0
Reserved
1 1 1 0 1
0
XPT6 / SAM0 PDPT Read
0 1 1 0 1
1
Reserved
1 1 1 0 1
1
XPT6 / SAM0 PDPT Write
0 1 1 1 0
0
PT Read Transfer
1 1 1 1 0
0
XPT7 / SOF0 Read
0 1 1 1 0
1
PT Write Transfer
1 1 1 1 0
1
XPT7 / SOF0 Write
0 1 1 1 1
0
Reserved
1 1 1 1 1
0
XPT7 / SOF0 PDPT Read
0 1 1 1 1
1
Idle
1 1 1 1 1
1
XPT7 / SOF0 PDPT Write
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
71
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
status codes (continued)
Table 30. Column-Time Status Codes
STATUS[5:0]
CYCLE TYPE
0 0 0 0 0 0
PP0 Low Priority Packet Transfer
1 0 0 0 0
STATUS[5:0]
0
Reserved
0 0 0 0 0 1
PP0 High Priority Packet Transfer
1 0 0 0 0
1
Reserved
0 0 0 0 1 0
PP0 Instruction Cache
1 0 0 0 1
0
Reserved
0 0 0 0 1 1
PP0 DEA
1 0 0 0 1
1
Reserved
0 0 0 1 0 0
PP1 Low Priority Packet Transfer
1 0 0 1 0
0
Reserved
0 0 0 1 0 1
PP1 High Priority Packet Transfer
1 0 0 1 0
1
Reserved
0 0 0 1 1 0
PP1 Instruction Cache
1 0 0 1 1
0
Reserved
0 0 0 1 1 1
PP1 DEA
1 0 0 1 1
1
Reserved
0 0 1 0 0 0
PP2 Low Priority Packet Transfer
1 0 1 0 0
0
Reserved
0 0 1 0 0 1
PP2 High Priority Packet Transfer
1 0 1 0 0
1
Reserved
0 0 1 0 1 0
PP2 Instruction Cache
1 0 1 0 1
0
Reserved
0 0 1 0 1 1
PP2 DEA
1 0 1 0 1
1
Reserved
0 0 1 1 0 0
PP3 Low Priority Packet Transfer
1 0 1 1 0
0
Reserved
0 0 1 1 0 1
PP3 High Priority Packet Transfer
1 0 1 1 0
1
Reserved
0 0 1 1 1 0
PP3 Instruction Cache
1 0 1 1 1
0
Reserved
0 0 1 1 1 1
PP3 DEA
1 0 1 1 1
1
Reserved
0 1 0 0 0 0
MP Low Priority Packet Transfer
1 1 0 0 0
0
Reserved
0 1 0 0 0 1
MP High Priority Packet Transfer
1 1 0 0 0
1
Reserved
0 1 0 0 1 0
MP Urgent Packet Transfer (Low)
1 1 0 0 1
0
Reserved
0 1 0 0 1 1
MP Urgent Packet Transfer (High)
1 1 0 0 1
1
Reserved
0 1 0 1 0 0
XPT / VCPT in Progress
1 1 0 1 0
0
Reserved
0 1 0 1 0 1
XPT / VCPT Complete
1 1 0 1 0
1
Reserved
0 1 0 1 1 0
MP Instruction Cache (Low)
1 1 0 1 1
0
Reserved
0 1 0 1 1 1
MP Instruction Cache (High)
1 1 0 1 1
1
Reserved
0 1 1 0 0 0
MP DEA (Low)
1 1 1 0 0
0
Reserved
0 1 1 0 0 1
MP DEA (High)
1 1 1 0 0
1
Reserved
0 1 1 0 1 0
MP Data Cache (Low)
1 1 1 0 1
0
Reserved
0 1 1 0 1 1
MP Data Cache (High)
1 1 1 0 1
1
Reserved
0 1 1 1 0 0
Frame 0
1 1 1 1 0
0
Reserved
0 1 1 1 0 1
Frame 1
1 1 1 1 0
1
Reserved
0 1 1 1 1 0
Refresh
1 1 1 1 1
0
Reserved
0 1 1 1 1 1
Idle
1 1 1 1 1
1
Write Drain / SDRAM DCAB
Low – MP operating in low (normal) priority mode
High – MP operating in high priority mode
72
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
CYCLE TYPE
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
address multiplexing
To support various RAM devices, the TMS320C80 can provide multiplexed row and column addresses on its
address bus. A full 32-bit address is always output at row time. The alignment of column addresses is configured
by the value input on the AS[2:0] pins at row time (see Figure 59).
A[31:0] Pins
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Row Time
A[31:0] Pins
AS [2:0]
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
000
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
001
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
x
x
x
x
x
2
1
0
010
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
x
x
x
x
x
x
2
1
0
011
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
x
x
x
x
x
x
x
2
1
0
100
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
x
x
x
x
x
x
x
x
2
1
0
101
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
x
x
x
x
x
x
x
x
x
2
1
0
110
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
x
x
x
x
x
x
x
x
x
x
2
1
0
111
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
x
x
x
x
x
x
x
x
x
x
x
2
1
0
Column Time
Figure 59. Address Multiplexing
dynamic bus sizing
The ’C80 supports data bus sizes of 8, 16, 32, or 64 bits as shown in Table 31. The value input on the BS[1:0]
pins at row time indicates the bus size of the addressed memory. This determines the maximum number of bytes
which the ’C80 can transfer during each column access. If the number of bytes to be transferred exceeds the
bus size, multiple accesses are performed automatically to complete the transfer.
Table 31. Bus Size Selection
BS[1:0]
BUS SIZE
00
8 bits
01
16 bits
10
32 bits
11
64 bits
The selected bus size also determines which portion of the data bus is used for the transfer. For 64-bit memory,
the entire data bus is used. For 32-bit memory, D[31:0] are used in little-endian mode and D[63:32] are used
in big-endian mode. 16-bit buses use D[15:0] and D[63:48] and 8-bit buses use D[7:0] and D[63:56] for littleand big-endian modes, respectively. The ’C80 always aligns data to the proper portion of the bus and activates
the appropriate CAS / DQM strobes to ensure that only valid bytes are transferred.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
73
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
cycle time selection
The ’C80 supports eight basic sets of memory timings to support various memory types directly. The cycle timing
is selected by the value input on the CT[2:0] and UTIME pins at row time. The selected timing remains in effect
until the next row access. See Table 32 for Cycle-timing selections.
Table 32. Cycle-Timing Selection
UTIME
CT[2:0]
MEMORY TIMING
0
0
0
0
Reserved
0
0
0
1
SDRAM: burst length 1, read latency 4
0
0
1
0
Reserved
0
0
1
1
SDRAM: burst length 2, read latency 4
0
1
0
0
User timed DRAM: pipelined 1 cycle / column
0
1
0
1
User timed DRAM: 1 cycle / column
0
1
1
0
User timed DRAM: 2 cycle / column
0
1
1
1
User timed DRAM: 3 cycle / column
1
0
0
0
SDRAM: burst length 1, read latency 2
1
0
0
1
SDRAM: burst length 1, read latency 3
1
0
1
0
SDRAM: burst length 2, read latency 2
1
0
1
1
SDRAM: burst length 2, read latency 3
1
1
0
0
DRAM: pipelined 1 cycle / column
1
1
0
1
DRAM: 1 cycle / column
1
1
1
0
DRAM: 2 cycle / column
1
1
1
1
DRAM: 3 cycle / column
page sizing
Whenever an external memory access occurs, the TC records the 26 most significant bits of the address in its
internal LASTPAGE register. The address of each subsequent (column) access is compared to this value. The
page size value input on the PS[3:0] pins determines which bits of LASTPAGE are used for this comparison.
If a difference exists between the enabled LASTPAGE bits and the corresponding bits of the next access then
the page has changed and the next memory access begins with a new row-address cycle (see Table 33).
74
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
page sizing (continued)
Table 33. Page-Size Selection
PS[3:0]
ADDRESS BITS COMPARED
PAGE SIZE (BYTES)
0 0 0 0
A(31:6)
64
0 0 0 1
A(31:7)
128
0 0 1 0
A(31:8)
256
0 0 1 1
A(31:9)
512
0 1 0 0
A(31:10)
1K
0 1 0 1
A(31:18)
256K
0 1 1 0
A(31:19)
512K
0 1 1 1
A(31:20)
1 0 0 0
A(31:0)
1M
1–8†
1 0 0 1
A(31:11)
2K
1 0 1 0
A(31:12)
4K
1 0 1 1
A(31:13)
8K
1 1 0 0
A(31:14)
16K
1 1 0 1
A(31:15)
32K
1 1 1 0
A(31:16)
64K
1 1 1 1
A(31:17)
128K
† PS[3:0] = 1000 disables page-mode cycles so that the effective page size is the same
as the bus size
block write support
The TMS320C80 supports three modes of VRAM block write. The block-write mode is dynamically selectable
so that software may specify block writes without knowing what type of block write the addressed memory
supports. Block writes are supported only for 64-bit buses. During block-write and load-color-register cycles,
the BS[1:0] inputs determine which block mode will be used (see Table 34).
Table 34. Block-Write Selection
BS[1:0]
BLOCK-WRITE MODE
0 0
Simulated
0 1
Reserved
1 0
4x
1 1
8x
SDRAM support
The TMS320C80 provides direct support for synchronous DRAM (SDRAM), VRAM (SVRAM), and graphics
RAM (SGRAM). During ’C80 power-up refresh cycles, the external system must signal the presence of these
memories by inputting a CT2 value of 0. This causes the ’C80 to perform special deactivate (DCAB) and mode
register set (MRS) commands to initialize the synchronous RAMs. Figure 60 shows the MRS value generated
by the ’C80. Note that read latency 4 timing programs the mode register for a read latency of 3. See Figure 60
for a listing of MRS values.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
75
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM support (continued)
SDRAM Mode Register Bit
11
10
Meaning
Value
0
0
9
8
7
WB
0
0
0
0
0
6
5
4
3
Read Latency
!(UTIME)
UTIME
2
S/I
UTIME&CT0
0
1
0
Burst Length
0
0
CT1
UTIME, CT0, CT1 values as input at the start of the MRS cycle
Figure 60. MRS Value
Because the MRS register is programmed through the SDRAM address inputs, the alignment of the MRS data
to the ’C80 logical-address bits is adjusted for the bus size (see Figure 61). The appearance of the MRS bits
on the ’C80 physical-address bus is dependent on the address multiplexing as selected by the AS[2:0] inputs.
’C80 LOGICAL ADDRESS BITS
BS[1:0]
A15
A14
A13
0 0
X
X
X
0 1
X
X
X
1 0
X
X
11
1 1
X
11
10
A12
A11
A10
A9
A8
A7
A6
A5
A4
A3
A2
A1
A0
X
11
10
9
8
7
6
5
4
3
2
1
0
11
10
9
8
7
6
5
4
3
2
1
0
X
10
9
8
7
6
5
4
3
2
1
0
X
X
9
8
7
6
5
4
3
2
1
0
X
X
X
Figure 61. MRS Value Alignment
memory cycles
TMS320C80 external memory cycles are generated by the TC’s external memory controller. The controller’s
state machine generates a sequence of states which define the transition of the memory interface signals. The
state sequence is dependent on the cycle timing selected for the memory access being performed as shown
in Figure 62. Memory cycles consist of row states and the column pipeline (see Figure 62).
76
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
memory cycles (continued)
rhiz
bus release
idle or abort
bus request
r1
r2
any cycle
r8
wait
always
r5
new page & CT = 100 & write
r4
or
(CT = 0xx & !SRS)
CT = 110
CT = 111
CT = 10x or (0xx & SRS)
refresh & CT = 10x or 0xx
refresh & CT = 110
r7
refresh & CT = 111
dw
MRS or DCAB
dcab
new page & CT = 0xx
r3
new page
always
always
always
r9
always
any cycle
fault, retry, or abort
always
!MRS & !DCAB
spin
spin
r6
col access
rspin
wait
col access
Column Pipeline
Figure 62. Memory Cycle State Diagram
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
77
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
row states
The row states make up the row time of each memory access. They occur when each new page access begins.
The transition indicators determine the conditions that cause transitions to another state. See Table 35 and
Table 36.
Table 35. Row States
STATE
DESCRIPTION
r1
Beginning state for all memory accesses. Outputs row address (A[31:0]) and cycle type (STATUS[5:0]) and drives control
signals to their inactive state
r2
Common to all memory accesses. Asserts RL and drives DDIN according to the data transfer direction. AS[2:0], BS[1:0],
CT[2:0], PS[3:0] and UTIME inputs are sampled
r3
Common to all memory accesses. DBEN is driven to its active level. For non-SDRAM, W, TRG / CAS, and DSF are driven to
their active levels and for non-SDRAM refreshes, all CAS / DQM strobes are activated. FAULT, READY, and RETRY inputs are
sampled.
r4
Inserted for 3 cycle / column accesses (CT=111) only. No signal transitions occur. RETRY input is sampled.
r5
Common to SDRAM and 2 or 3 cycle / column accesses (CT=0xx or 11x). RAS is driven low. W is driven low for DCAB and MRS
cycles and TRG / CAS is driven low for MRS and SDRAM refresh cycles.
r6
Common to all memory accesses. For SDRAM cycles, RAS, TRG / CAS, and W are driven high. For non-SDRAM, RAS is driven
low (if not already) and W, TRG / CAS, and DSF are driven to their appropriate levels. DBEN is driven low and READY and
RETRY are sampled.
rspin
Additional state to allow TC column time pipeline to load. No signal transitions occur. RETRY is sampled. The rspin state can, on
occasion, repeat multiple times.
r7
Common to 2 and 3 cycle / column refreshes (CT=11x). Processor activity code is output on STATUS[5:0]. RETRY input is
sampled.
r8
For 3 cycle / column refreshes only (CT=111). No signal transitions occur. RETRY input is sampled.
r9
Common to all refresh cycles. Processor activity code is output on STATUS[5:0] and RETRY input is sampled.
dw
Occurs for pipelined 1 cycle / column writes only. All CAS / DQM strobes are activated.
dcab
Occurs for SDRAM cycles (CT = 0xx). RAS and W are activated to perform a DCAB command.
rhiz
High impedance state. Occurs during host requests and repeats until bus is released by the host
Table 36. State Transition Indicators
INDICATOR
any cycle
CT=xxx
DESCRIPTION
Continuation of current cycle
State change occurs for indicated CT[2:0] value (as latched in r2 state)
abort
Current cycle aborted by TC in favor of higher priority cycle
fault
FAULT input sampled low (in r3 state), memory access faulted
retry
RETRY input sampled low (in r3 state), row-time retry
wait
READY input sampled low (in r3, r6, or last column state) repeat current state
spin
Internally generated wait state to allow TC pipeline to load
new page
The next access requires a page change (new row access).
external memory timing examples
The following sections contain descriptions of the ’C80 memory cycles and illustrate the signal transitions for
those cycles. Memory cycles may be separated into two basic categories; DRAM-type cycles for use with
DRAM-like devices, SRAM, and peripherals, and SDRAM-type cycles for use with SDRAM-like devices.
78
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
DRAM-type cycles
The DRAM-type cycles are page-mode accesses consisting of a row access followed by one or more column
accesses. Column accesses may be one, two, or three clock cycles in length with two and three cycle accesses
allowing the insertion of wait states to accommodate slow devices. Idle cycles can occur after necessary column
accesses have completed or between column accesses due to “bubbles” in the TC data-flow pipeline. The
pipeline diagrams in Figure 63 show the pipeline stages for each access type and when the CAS / DQM signal
corresponding to the column access is activated.
CAS / DQM - / A A / B B / C C / Col A
c1
Col B
CAS / DQM
Col A
c2
c3
c1
c2
c3
c1
c2
c3
ci
ci
Col C
Idle
A
B
C
c1
Col B
c1
Col C
c1
Idle
ci
ci
Pipelined 1 cycle / column (CT = 100)
Pipelined 1 cycle / column (CT = 100)
reads, read transfers, split read transfers
writes, LCRs, block writes, write transfers, split write transfers
CAS / DQM
A
B
Col A
c1
c2
Col B
c1
Col C
C
-
CAS / DQM
Col A
c2
c1
B
Col B
ci
C
c1
Col C
c2
ci
Idle
A
c1
c1
ci
Idle
Nonpipelined 1 cycle / column (CT = 101)
Nonpipelined 1 cycle / column (CT = 101)
reads, read transfers, split read transfers
writes, LCRs, block writes, write transfers, split write transfers
A
CAS / DQM
Col A
c1
Col B
B
CAS / DQM
C
Col A
c2
c1
A
A
c2
c3
-
-
c2
-
Col C
c1
Col B
c1
-
-
c1
c2
c3
-
-
ci
B
ci
C
C
-
c2
-
Idle
B
Col C
-
-
c1
c2
c3
-
-
-
-
-
-
ci
ci
Idle
2 cycle / column (CT = 110)
3 cycle / column (CT = 111)
all cycles (except refresh)
all cycles (except refresh)
ci
Figure 63. DRAM Cycle Column Pipelines
read cycles
Read cycles transfer data or instructions from external memory to the ’C80. The cycles can occur as a result
of a packet transfer, cache request, or DEA request. During the cycle, W is held high, TRG / CAS is driven low
after RAS to enable memory output drivers and DDIN is low so that data transceivers drive into the ’C80. The
TC places D[63:0] in high impedance allowing it to be driven by the memory and latches input data during the
appropriate column state. The TC always reads 64 bits and extracts and aligns the appropriate bytes. Invalid
bytes for bus sizes of less than 64 bits are discarded. During peripheral device packet transfers, DBEN remains
high. Read cycles are shown in Figure 64 through Figure 67.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
79
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
read cycles (continued)
State
Col A
Col B
Col C
Col D
r1
r2
r3
r6
col
c1
col
c2
c1
col
c3
c2
c1
col
col
col
c3
c2
c1
c3
c2
c3
CLKOUT
CT[2:0]
4
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
PAC
PAC
Row
Col A
Col B
Col C
Col D
Idle
RL
A[31:0]
RAS
–/A
B/C
A/B
C/D
D/–
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
A
DBEN
B
C
0 For Normal Reads, 1 For PDPT Reads
DDIN
For user-modified timing:
UTIME
RAS
–/A
CAS / DQM[7:0]
A/B
B/C
C/D
D/–
Figure 64. Pipelined 1 Cycle / Column Read-Cycle Timing
80
D
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
read cycles (continued)
State
r1
r2
r3
Col A
r6
col
col
c1
c2
Col B
col
c1
col
r1
c2
Col C
c1
c2
Idle
CLKOUT
CT[2:0]
5
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
PAC
Row
Col A
Col B
Col C
A
B
C
RL
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
A
B
C
0 For Normal Reads, 1 For PDPT Reads
DBEN
DDIN
For user-modified timing:
UTIME
RAS
A
CAS / DQM[7:0]
B
C
Figure 65. Nonpipelined 1 Cycle / Column Read-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
81
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
read cycles (continued)
State
r1
r2
r3
r5
r6
Col A
col
col
c1
c2
Col B
col
col†
col
c1
c2
c2
ci‡
Col C
col
col
c1
c2
CLKOUT
CT[2:0]
6
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
Row
Col A
PAC
Idle
PAC
RL
A[31:0]
Col B
Col C
RAS
A
CAS / DQM[7:0]
B
C
DSF
TRG / CAS
W
A
D[63:0]
B
C
0 For Normal Reads, 1 For PDPT Reads
DBEN
DDIN
For user-modified timing:
UTIME
RAS
A
CAS / DQM[7:0]
B
† Wait state inserted by external logic (example)
‡ Internally generated pipeline bubble (example)
Figure 66. 2 Cycles / Column Read-Cycle Timing
82
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
C
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
read cycles (continued)
State
r1
r2
r3
r4
Col A
r5
r6
col
col
col
c1
c2
c3
Col B
col
col
col†
col
c1
c2
c3
c3
ci‡
Col C
col
col
col
c1
c2
c3
r1
CLKOUT
CT[2:0]
7
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
Idle
PAC
RL
A[31:0]
Row
Column A
Column B
Column C
A
B
C
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
A
D[63:0]
DBEN
B
C
0 For Normal Reads, 1 For PDPT Reads
DDIN
For user-modified timing:
UTIME
RAS
A
CAS / DQM[7:0]
B
C
† Wait state inserted by external logic (example)
‡ Internally generated pipeline bubble (example)
Figure 67. 3 Cycles / Column Read-Cycle Timing
write cycles
Write cycles transfer data from the ’C80 to external memory. These cycles can occur as a result of a packet
transfer, a DEA request, or an MP data cache write-back. During the cycle TRG/ CAS is held high, W is driven
low after the fall of RAS to enable early-write cycles, and DDIN is high so that data transceivers drive toward
memory. The TC drives data out on D[63:0] and indicates valid bytes by activating the appropriate CAS/ DQM
strobes. During peripheral device packet transfers, DBEN remains high and D[63:0] is placed in high impedance
so that the peripheral device can drive data into the memory. Write cycles are shown in Figure 68 through
Figure 71.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
83
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
write cycles (continued)
State
Col A
Col B
Col C
r1
r2
r3
r6
rspin
rspin
col
col
ci†
col
drn
c1
c1
c1
CLKOUT
CT[2:0]
4
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
Idle
PAC
Col A
Col B
Col C
A
B
C
Drain
RL
Row
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
A
D[63:0]
B
C
0 For Normal Write, 1 For PDPT Write
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
A
B
† Internally generated pipeline bubble (example)
Figure 68. Pipelined 1 Cycle / Column Write-Cycle Timing
84
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
C
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
write cycles (continued)
State
Col A
Col B
Col C
r1
r2
r3
r6
rspin
rspin
col
c1
col
ci†
col
r1
c1
c1
CLKOUT
CT[2:0]
5
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
Idle
Row
Col A
Col B
Col C
A
B
C
PAC
RL
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
A
B
C
0 For Normal Write, 1 For PDPT Write
DBEN
DDIN
For user-modified timing:
UTIME
RAS
A
CAS / DQM[7:0]
B
C
† Internally generated pipeline bubble (example)
Figure 69. Nonpipelined 1 Cycle / Column Write-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
85
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
write cycles (continued)
State
r1
r2
r3
r5
r6
rspin
Col A
col
col
c1
c2
Col B
col
col†
col
c1
c2
c2
ci‡
Col C
col
col
c1
c2
CLKOUT
CT[2:0]
6
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
Cycle Type
STATUS[5:0]
PAC
PAC
Col A
Col B
Col C
B
C
Idle
PAC
RL
Row
A[31:0]
RAS
A
CAS / DQM[7:0]
DSF
TRG / CAS
W
A
D[63:0]
B
C
0 For Normal Write, 1 For PDPT Write
DBEN
DDIN
For user-modified timing:
UTIME
RAS
A
CAS / DQM[7:0]
B
† Wait state inserted by external logic (example)
‡ Internally generated pipeline bubble (example)
Figure 70. 2 Cycles / Column Write-Cycle Timing
86
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
C
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
write cycles (continued)
State
r1
r2
r3
r4
r5
Col A
r6
col
col
col
c1
c2
c3
Col B
col
col
col†
col
c1
c2
c3
c3
ci‡
col
col
col
c1
c2
c3
Col C
r1
CLKOUT
CT[2:0]
7
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
Row
Col A
Col B
Idle
PAC
RL
A[31:0]
Col C
RAS
A
CAS / DQM[7:0]
B
C
DSF
TRG / CAS
W
A
D[63:0]
B
C
0 For Normal Write, 1 For PDPT Write
DBEN
DDIN
For user-modified timing:
UTIME
RAS
A
CAS / DQM[7:0]
B
C
† Wait state inserted by external logic (example)
‡ Internally generated pipeline bubble (example)
Figure 71. 3 Cycles / Column Write-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
87
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
load-color-register cycles
Load-color-register (LCR) cycles are used to load a VRAM’s color register prior to performing a block write. LCR
cycles are supported only on 64-bit data buses. An LCR cycle closely resembles a normal write cycle because
it writes into a VRAM. The difference is that the DSF output is high at both the fall of RAS and the fall of
CAS/ DQM. Also, because the VRAM color register is a single location, only one column access occurs.
The row address that is output by the TC is used for bank decode only. Normally all VRAM banks should be
selected during an LCR cycle because another LCR cycle will not occur when a block-write memory page
change occurs. The column address that is output during an LCR is likewise irrelevant because the VRAM color
register is the only location written. All CAS / DQM strobes are active during an LCR cycle.
The RETRY input is sampled during LCR column states and must be valid high or low. Asserting RETRY at
column time has no effect, however, because only one column access is performed.
If the BS[1:0] inputs indicate that the addressed memory supports only simulated block writes, the LCR cycle
will be changed into a normal write cycle at the start of the simulated block write. Load color register cycles timing
is shown in Figure 72 through Figure 75.
88
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
load-color-register cycles (continued)
State
r1
r2
r3
r6
rspin
rspin
c1
drn
PAC
Drain
r1
CLKOUT
CT[2:0]
4
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
Cycle Type
STATUS[5:0]
RL
Row
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
Color
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 72. Pipelined 1 Cycle / Column Load-Color-Register-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
89
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
load-color-register cycles (continued)
State
r1
r2
r3
r6
rspin
rspin
c1
r1
CLKOUT
CT[2:0]
5
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
Cycle Type
STATUS[5:0]
PAC
RL
A[31:0]
Row
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
Color
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 73. Nonpipelined 1 Cycle / Column Load-Color-Register-Cycle Timing
90
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
load-color-register cycles (continued)
State
r1
r2
r3
r5
r6
rspin
c1
c2
r1
CLKOUT
CT[2:0]
6
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
RL
Row
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
Color Value
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 74. 2 Cycles / Column Load-Color-Register-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
91
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
load-color-register cycles (continued)
State
r1
r2
r3
r4
r5
r6
c1
c2
c3
CLKOUT
CT[2:0]
7
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
Cycle Type
STATUS[5:0]
PAC
RL
Row
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
Color Value
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 75. 3 Cycles / Column Load-Color-Register-Cycle Timing
92
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
block-write cycles
Block-write cycles cause the data stored in the VRAM color registers to be written to the memory locations
enabled by the appropriate data bits output on the D[63:0] bus. This allows up to a total of 64 bytes (depending
on the type of block write being used) to be written in a single-column access. This cycle is identical to a standard
write cycle with the following exceptions:
D
D
D
D
D
DSF is active (high) at the fall of CAS, enabling the block-write function within the VRAMs.
Only 64-bit bus sizes are supported during block write; therefore, BS[1:0] inputs are used to indicate the
type of block write that is supported by the addressed VRAMs, rather than the bus size.
The two or three LSBs (depending on the type of block write) of the column address are ignored by the
VRAMs because these column locations are specified by the data inputs.
The values output by the TC on D[63:0] represent the column locations to be written to, using the
color-register value. Depending on the type of block write supported by the VRAM, all of the data bits
are not necessarily used by the VRAMs.
Block writes always begin with a row access. Upon completion of a block write, the memory interface
returns to state r1 to await the next access.
See Figure 76 through Figure 79 for block-write cycle timing.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
93
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
block-write cycles (continued)
State
Col A
Col B
Col C
r1
r2
r3
r6
rspin
rspin
col
c1
ci†
col
col
drn
c1
c1
CLKOUT
CT[2:0]
4
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
Row
Col A
Col B
Col C
A
B
C
Sel A
Sel B
Sel C
A
B
C
Idle
PAC
RL
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
† Internally generated pipline bubble (example)
Figure 76. Pipelined 1 Cycle / Column Block-Write-Cycle Timing
94
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
Drain
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
block-write cycles (continued)
State
Col A
Col B
Col C
r1
r2
r3
r6
rspin
rspin
col
c1
col
ci†
col
r1
c1
c1
CLKOUT
CT[2:0]
5
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
Row
Col A
Col B
Col C
A
B
C
Sel A
Sel B
Sel C
A
B
C
Idle
PAC
RL
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
† Internally generated pipline bubble (example)
Figure 77. Nonpipelined 1 Cycle / Column Block-Write-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
95
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
block-write cycles (continued)
State
r1
r2
r3
r5
r6
rspin
Col A
Col B
Col C
col
col
c1
c2
col
c1
col†
col
c2
c2
ci‡
col
col
c1
c2
CLKOUT
CT[2:0]
6
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
Col A
Col B
Idle
PAC
RL
Row
A[31:0]
Col C
RAS
A
CAS / DQM[7:0]
B
C
DSF
TRG / CAS
W
Col Sel A
D[63:0]
Col Sel B
Col Sel C
DBEN
DDIN
For user-modified timing:
UTIME
RAS
A
CAS / DQM[7:0]
B
† Wait state inserted by external logic (example)
‡ Internally generated pipeline bubble (example)
Figure 78. 2 Cycles / Column Block-Write-Cycle Timing
96
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
C
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
block-write cycles (continued)
State
Col A
Col B
Col C
r1
r2
r3
r4
r5
r6
col
col
col
c1
c2
c3
col
col
col
col†
c1
c2
c3
c3
ci‡
col
col
col
c1
c2
c3
r1
CLKOUT
CT[2:0]
7
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
Row
Col A
Col B
Idle
PAC
RL
A[31:0]
Col C
RAS
A
CAS / DQM[7:0]
C
B
DSF
TRG / CAS
W
D[63:0]
Col Sel A
Col Sel B
Col Sel C
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
A
B
C
† Wait state inserted by external logic (example)
‡ Internally generated pipeline bubble (example)
Figure 79. 3 Cycles / Column Block-Write-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
97
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles
Read-transfer (memory-to-register) cycles transfer a row from the VRAM memory array into the VRAM shift
register (SAM). This causes the entire SAM (both halves of the split SAM) to be loaded with the array data.
Split-register read-transfer (memory-to-split-register) cycles also transfer data from a row in the memory array
to the SAM. However, these transfers cause only half of the SAM to be written. Split-register read transfers allow
the inactive half of the SAM to be loaded with the new data while the other active half continues to shift data
in or out.
Write-transfer (register-to-memory) cycles transfer data from the SAM into a row of the VRAM array. This
transfer causes the entire SAM (both halves of the split SAM) to be written into the array.
Split-register write-transfer (split-register-to-memory) cycles also transfer data from the SAM to a row in the
memory array. However, these transfers write only half of the SAM into the array. Split-register write transfers
allow the inactive half of the SAM to be transferred into memory while the other (active) half continues to shift
serial data in or out.
Read and split-read transfers resemble a standard read cycle. Write and split-write transfers resemble a
standard write cycle. The TRG / CAS output is driven low prior to the fall of RAS to indicate a transfer cycle. Only
a single column access is performed so RETRY, while required to be at a valid level, has no effect if asserted
at column time. The value output on A[31:0] at column time represents the SAM tap point (see Figure 80 through
Figure 86 for transfer cycle timing.
98
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles (continued)
State
r1
r2
r3
r6
c1
c2
c3
r1
CLKOUT
CT[2:0]
4
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
Idle
RL
A[31:0]
Row
Tap Point
RAS
CAS / DQM[7:0]
DSF
0 for Full Transfer, 1 for Split Transfer
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 80. Pipelined 1 Cycle/Column Read-Transfer and Split-Register Read-Transfer-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
99
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles (continued)
State
r1
r2
r3
c1
c2
PAC
Idle
r6
r1
CLKOUT
CT[2:0]
5
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
RL
A[31:0]
Row
Tap Point
RAS
CAS / DQM[7:0]
DSF
0 for Full Transfer, 1 for Split Transfer
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 81. Nonpipelined 1 Cycle/Column Read-Transfer and Split-Register Read-Transfer-Cycle Timing
100
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles (continued)
State
r1
r2
r3
r5
r6
c1
c2
r1
CLKOUT
CT[2:0]
6
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
RL
A[31:0]
Row
Tap Point
RAS
CAS / DQM[7:0]
0 for Full Transfer, 1 for Split Transfer
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 82. 2 Cycles/Column Read-Transfer and Split-Register Read-Transfer-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
101
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles (continued)
State
r1
r2
r3
r4
r5
r6
c1
c2
c3
r1
CLKOUT
CT[2:0]
7
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
Cycle Type
STATUS[5:0]
PAC
RL
Row
A[31:0]
Tap Point
RAS
CAS / DQM[7:0]
DSF
0 for Full Transfer, 1 for Split Transfer
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 83. 3 Cycles/Column Read-Transfer and Split-Register Read-Transfer-Cycle Timing
102
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles (continued)
State
r1
r2
r3
r6
rspin
rspin
c1
drn
PAC
Drain
r1
CLKOUT
CT[2:0]
4
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
RL
A[31:0]
Row
Tap Point
RAS
CAS / DQM[7:0]
0 for Full Transfer, 1 for Split Transfer
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 84. Pipelined 1 Cycle/Column Write-Transfer and Split-Register Write-Transfer-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
103
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles (continued)
State
r1
r2
r3
r6
rspin
rspin
c1
r1
CLKOUT
CT[2:0]
5
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
RL
Row
A[31:0]
Tap Point
RAS
CAS / DQM[7:0]
0 for Full Transfer, 1 for Split Transfer
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 85. Nonpipelined 1 Cycle/Column Write-Transfer and Split-Register Write-Transfer-Cycle Timing
104
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles (continued)
State
r1
r2
r3
r5
r6
rspin
c1
c2
rl
CLKOUT
CT[2:0]
6
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
Row
Tap Point
RL
A[31:0]
RAS
CAS / DQM[7:0]
0 for Full Transfer, 1 for Split Transfer
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 86. 2 Cycles/Column Write-Transfer and Split-Register Write-Transfer-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
105
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
transfer cycles (continued)
State
r1
r2
r3
r4
r5
r6
c1
c2
c3
r1
CLKOUT
CT[2:0]
7
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
Row
Tap Point
RL
A[31:0]
RAS
CAS / DQM[7:0]
0 for Full Transfer, 1 for Split Transfer
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
For user-modified timing:
UTIME
RAS
CAS / DQM[7:0]
Figure 87. 3 Cycles/Column Write-Transfer and Split-Register Write-Transfer-Cycle Timing
106
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
refresh cycles
Refresh cycles are generated by the TC at the programmed refresh interval. They are characterized by the
following signal activity:
D
D
D
D
D
D
D
D
D
CAS falls prior to RAS.
All CAS pins ( CAS/ DQM[7:0]) are active.
TRG, W, and DBEN all remain inactive (high) because no data transfer occurs.
DSF remains inactive (low).
The data bus is driven to the high-impedance state.
The upper half of the address bus (A[31:16]) contains the refresh pseudo-address and the lower half
(A[15:0]) is driven to all zeros.
If RETRY is asserted at any sample point during the cycle, the cycle timing is not modified. Instead, the
pseudo-address and backlog counters are simply not decremented.
Selecting user-modified timing has no effect on the cycles.
Upon completion of the refresh cycle, the memory interface returns to state r1 to await the next access.
See Figure 88 through Figure 90 for refresh cycle timing.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
107
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
refresh cycles (continued)
State
r1
r2
r3
r6
r9
CLKOUT
CT[2:0]
4/5
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
RL
A[31:16]
Refresh Pseudo Address
A[15:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Figure 88. 1-Cycle/Column Refresh-Cycle Timing
108
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
refresh cycles (continued)
State
r1
r2
r3
r5
r6
r7
r9
r1
CLKOUT
CT[2:0]
6
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
RL
A[31:16]
Refresh Pseudo Address
A[15:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Figure 89. 2 Cycles/Column Refresh-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
109
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
refresh cycles (continued)
State
r1
r2
r3
r4
r5
r6
r7
r8
CLKOUT
CT[2:0]
7
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
RL
A[31:16]
Refresh Pseudo Address
A[15:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Figure 90. 3 Cycles/Column Refresh-Cycle Timing
110
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r9
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM type cycles
The SDRAM type cycles support the use of SDRAM, SGRAM, or SVRAM devices for single-cycle memory
accesses. While SDRAM cycles use the same state sequences as DRAM cycles, the memory-control signal
transitions are modified to perform SDRAM command cycles. The supported SDRAM commands are:
DCAB
Deactivate (precharge) all banks
ACTV
Activate the selected bank and select the row
READ
Input starting column address and start read operation
WRT
Input starting column address and start write operation
MRS
Set SDRAM mode register
REFR
Auto-refresh cycle with internal address
SRS
Set special register (color register)
BLW
Block write
SDRAM cycles begin with an activate (ACTV) command followed by the requested column accesses. When
a memory-page change occurs, the selected bank is deactivated with a DCAB command.
The TMS320C80 supports read latencies of 2, 3, or 4 cycles and burst lengths of 1 or 2. These are selected
by the CT code and UTIME value input at the start of the access.
The column pipelines for SDRAM accesses are shown in Figure 91. Idle cycles can occur after necessary
column accesses have completed or between column accesses due to “bubbles” in the TC data flow pipeline.
The pipeline diagrams show the pipeline stages for each access type and when the CAS/ DQM signal
corresponding to the column access is activated.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
111
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM type cycles (continued)
CAS / DQM
Col A
A
B
C
CAS / DQM
c1
c2
c3
Col A
c1
c2
c3
c1
c2
c3
ci
ci
Col B
Col C
Idle
Col B
Col A
c1
A
B
C
c2
c3
c4
c1
Col B
Col C
c3
c4
c5
c2
c3
c4
ci
C
c4
c1
c2
c3
c4
c1
c2
c3
c4
ci
ci
ci
CAS / DQM
A
(B)
C
Col A, B
c1
c2
c3
-
c5
ci
ci
Col C, D
ci
(D)
-
-
c1
c2
-
Col E, F
Burst-length 1, 4 cycle latency reads, read
transfers,split-read transfers.
CAS / DQM
A
Col A
c1
Col B
B
E
(F)
c3
-
-
c1
c2
c3
-
-
-
ci
ci
Idle
C
ci
Burst-length 2, 2 cycle latency reads, read transfers,
split-read transfers
c1
c1
Col C
Idle
ci
CAS / DQM
A
(B)
Col A, B
c1
c2
-
Burst-length 1 writes, block writes, SRSs, write transfers,
split-write transfers
Col C, D
C
(D)
CAS / DQM
c1
A
(B)
C
c2
c3
c4
-
-
-
-
c1
c2
c3
Col C, D
Col E, F
(D)
E
E
(F)
c1
c2
-
Col E, F
Col A, B
ci
Burst-length 1, 3 cycle latency reads, read
transfers, split-read transfers
c5
c1
ci
B
c3
Idle
ci
c2
Idle
A
c2
Col C
Burst-length 1, 2 cycle latency reads, read
transfers, split-read transfers
CAS / DQM
c1
c1
(F)
c2
-
ci
Burst-length 2, writes
c4
-
-
-
c1
c2
c3
c4
-
-
-
-
ci
ci
ci
Idle
ci
Idle
ci
Burst-length 2, 3 cycle latency reads, read transfers,
split-read transfers
CAS / DQM
Col A
A
B
C
CAS / DQM
A
(B)
C
(D)
E
Col A, B
c1
c2
c3
c4
c5
c1
-
Col B
c1
Col C, D
-
-
-
-
c1
c2
c3
c4
-
-
-
-
-
c1
c2
c3
c4
c5
-
-
-
-
-
ci
ci
ci
ci
Col C
c1
Col E, F
-
Idle
ci
Burst-length 2, block writes, write transfers, split-write
transfers
Idle
Burst-length 2, 4 cycle latency reads, read
transfers,split-read transfers.
Figure 91. SDRAM Column Pipelines
112
POST OFFICE BOX 1443
(F)
• HOUSTON, TEXAS 77251–1443
c5
ci
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
special SDRAM cycles
To initialize the SDRAM properly, the TMS320C80 performs two special SDRAM cycles after reset. The ’C80
first performs a deactivate cycle on all banks (DCAB) and then initializes the SDRAM mode register with a mode
register set (MRS) cycle. The CT code input at the start of the MRS cycle determines the burst length and latency
that is programmed into the SDRAM mode register (see Figure 92 and Figure 93).
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
113
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
special SDRAM cycles (continued)
State
r1
r2
r3
r5
CLKOUT
CT[2:0]
0xx
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
Cycle Type
STATUS[5:0]
RL
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Command
DCAB
Figure 92. SDRAM Power-Up Deactivate Cycle Timing
114
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
special SDRAM cycles (continued)
State
r1
r2
r3
r5
r1
CLKOUT
CT[2:0]
0xx
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
Cycle Type
STATUS[5:0]
RL
MRS
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Command
MRS
Figure 93. SDRAM Mode Register Set-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
115
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM read cycles
Read cycles begin with an activate (ACTV) command to activate the bank and to select the row. The TC outputs
the column address and activates the TRG/ CAS strobe for each read command. For burst length 1 accesses,
a read command can occur on each cycle. For burst-length 2 accesses, a read command can occur every two
cycles. The TC places D[63:0] into the high-impedance state, allowing it to be driven by the memory, and latches
input data during the appropriate column state. The TC always reads 64 bits and extracts and aligns the
appropriate bytes. Invalid bytes for bus sizes of less than 64 bits are discarded. The CAS/ DQM strobes are
activated two cycles before input data is latched. If the second column in a burst is not required, then CAS/ DQM
is not activated. During peripheral device packet transfers, DBEN remains high (see Figure 94 through
Figure 99).
116
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM read cycles (continued)
State
r1
r2
r3
r5
Col Pipe
r6
Col A
Col B
Col C
Col D
col
c1
col
c2
c1
col
c3
c2
c1
col
col
c3
c2
c1
c3
c2
DCAB
r1
c3
CLKOUT
CT[2:0]
0
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
PAC
PAC
Row
Col A
Col B
Col C
Col D
A
B
C
D
A
B
C
D
A
B
RL
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
C
D
0 For Normal Read, 1 For PDPT read
DBEN
DDIN
Command
ACTV
READ
READ
READ
READ
DCAB
Figure 94. SDRAM Burst-Length 1, 2 Cycle Latency Read-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
117
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM read cycles (continued)
State
r1
r2
r3
r5
r6
Col A
Col B
Col C
Col D
Col Pipe
col
c1
col
col
col
col
col
r1
c2
c1
c3
c2
c1
c4
c3
c2
c1
c4
c3
c2
c4
c3
c4
PAC
Idle
DCAB
CLKOUT
CT[2:0]
1
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
PAC
Row
Col A
Col B
Col C
A
B
C
B
C
D
RL
A[31:0]
Col D
RAS
CAS / DQM[7:0]
D
DSF
A
TRG / CAS
W
D[63:0]
A
B
C
0 For Normal Read, 1 For PDPT Read
DBEN
DDIN
Command
ACTV
READ
READ
READ
READ
Figure 95. SDRAM Burst-Length 1, 3 Cycle Latency Read-Cycle Timing
118
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
DCAB
D
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM read cycles (continued)
State
Col Pipe
r1
r2
r3
r5
r6
Col A
Col B
Col C
Col D
col
c1
col
c2
c1
col
c3
c2
c1
col
c4
c3
c2
c1
col
c5
c4
c3
c2
col
col
r1
c5
c4
c3
c5
c4
c5
CLKOUT
CT[2:0]
1
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS
[5:0]
Cycle Type
PAC
PAC
PAC
PAC
Row
Col A
Col B
Col C
A
B
C
B
C
D
DCAB
Idle
RL
A[31:0]
Col D
RAS
CAS / DQM
[7:0]
D
DSF
A
TRG / CAS
W
D[63:0]
A
B
C
D
0 For Normal Read, 1 For PDPT Read
DBEN
DDIN
Command
ACTV
READ
READ
READ
READ
DCAB
Figure 96. SDRAM Burst-Length 1, 4 Cycle Latency Read-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
119
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM read cycles (continued)
State
r1
r2
r3
r5
Col Pipe
r6
Col A
Col B
Col C
Col D
col
col
col
c1
c2
c1
c3
c2
c1
col
c3
c2
c1
col
c3
c2
r1
c3
CLKOUT
CT[2:0]
2
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
DCAB
PAC
RL
A[31:0]
Row
Col A
(Col B)
Col C
(Col D)
A
B
C
D
RAS
CAS / DQM[7:0]
DSF
A, B
TRG / CAS
C, D
W
D[63:0]
A
B
C
0 For Normal Read, 1 For PDPT Read
DBEN
DDIN
Command
ACTV
READ
READ
DCAB
Figure 97. SDRAM Burst-Length 2, 2 Cycle Latency Read-Cycle Timing
120
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
D
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM read cycles (continued)
State
r1
r2
r3
r5
Col Pipe
r6
Col A
Col B
Col C
Col D
col
col
col
col
col
col
r1
c1
c2
c1
c3
c2
c1
c4
c3
c2
c1
c4
c3
c2
c4
c3
c4
Idle
DCAB
CLKOUT
CT[2:0]
3
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
RL
A[31:0]
Row
Col A
(Col B)
Col C
A
B
(Col D)
RAS
CAS / DQM[7:0]
C
D
A
B
DSF
A, B
TRG / CAS
C, D
W
D[63:0]
C
D
0 For Normal Read, 1 For PDPT Read
DBEN
DDIN
Command
ACTV
READ
READ
DCAB
Figure 98. SDRAM Burst-Length 2, 3 Cycle Latency Read-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
121
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM read cycles (continued)
State
Col Pipe
r1
r2
r3
r5
r6
Col A
Col B
Col C
Col D
col
c1
col
c2
c1
col
c3
c2
c1
col
c4
c3
c2
c1
col
c5
c4
c3
c2
col
col
r1
c5
c4
c3
c5
c4
c5
CLKOUT
CT[2:0]
3
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS
[5:0]
Cycle Type
PAC
PAC
DCAB
Idle
RL
A[31:0]
Row
Col A
(Col B)
Col C
A
B
(Col D)
RAS
CAS / DQM
[7:0]
C
D
DSF
A, B
TRG / CAS
C, D
W
D[63:0]
A
B
C
0 For Normal Read, 1 For PDPT Read
DBEN
DDIN
Command
ACTV
READ
READ
Figure 99. SDRAM Burst-Length 2, 4 Cycle Latency Read-Cycle Timing
122
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
DCAB
D
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM write cycles
Write cycles begin with an activate (ACTV) command to activate the bank and select the row. The TC outputs
the column address and activates the TRG / CAS and W strobes for each write command. For burst-length 1
accesses, a write command can occur on each cycle. For burst-length 2 accesses, a write command can occur
every two cycles. The TC drives data out on D[63:0] during each cycle of an active-write command and indicates
valid bytes by driving the appropriate CAS / DQM strobes low. During peripheral device packet transfers, DBEN
remains high and D[63:0] are placed in the high-impedance state so that the peripheral can drive data into the
memories. For SDRAM write cycles, see Figure 100 and Figure 101.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
123
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM write cycles (continued)
State
Col Pipe
r1
r2
r3
r5
r6
rspin
Col A
Col B
Col C
Col D
col
c1
col
col
col
col
col
Idle
DCAB
c1
c1
c1
CLKOUT
CT[2:0]
0,1
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
PAC
PAC
Col A
Col B
Col C
A
B
C
D
A
B
C
D
A
B
C
D
RL
A[31:0]
Row
Col D
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
0 For Normal Read, 1 For PDPT Read
DBEN
DDIN
Command
ACTV
WRT
WRT
WRT
WRT
Figure 100. SDRAM Burst-Length 1 Write-Cycle Timing
124
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
DCAB
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM write cycles (continued)
State
Col Pipe
r1
r2
r3
r5
r6
rspin
Col A
Col B
Col C
Col D
col
c1
col
col
col
col
col
Idle
DCAB
r1
c1
c1
c1
CLKOUT
CT[2:0]
2, 3
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
PAC
Cycle Type
STATUS[5:0]
PAC
RL
Row
A[31:0]
Col A
(Col B)
Col C
A
B
C
(Col D)
RAS
CAS / DQM[7:0]
D
DSF
A, B
TRG / CAS
C, D
W
D[63:0]
A
B
C
D
DBEN
DDIN
ACTV
Command
WRT
WRT
DCAB
Figure 101. SDRAM Burst-Length 2 Write-Cycle Timing
special register set cycles
Special register set (SRS) cycles are used to program control registers within an SVRAM or SGRAM. The ’C80
only supports programming of the color register for use with block writes. The cycle is similar to a single burst
length 1 write cycle but DSF is driven high. The values output on the ’C80 address bits cause the color register
to be selected as shown in Figure 102 (see Figure 103).
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
125
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
special register set cycles (continued)
SDRAM Address Pin
BS
A8
A7
A6
A5
A4
SDRAM Function
0
0
0
LC
LM
LS
TMS320C80 Output Value
0
0
0
1
0
0
A3
A2
A1
A0
Stop Register
0
0
0
0
Figure 102. Special-Register-Set Value
State
r1
r2
r6
r3
rspin
rspin
col
c1
Col Pipe
CLKOUT
CT[2:0]
0xx
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
Row
SRS
RL
A[31:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
Color
D[63:0]
DBEN
DDIN
SRS
Command
Figure 103. SDRAM SRS-Cycle Timing
126
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM block-write cycles
Block-write cycles allow SVRAMs and SGRAMs to write a stored color value to multiple column locations in a
single access. Block-write cycles are similar to write cycles except that DSF is driven high to indicate a
block-write command. Because burst is not supported for block write, burst length 2 accesses generate a single
block-write every other clock cycle (see Figure 104 and Figure 105).
State
Col Pipe
r1
r2
r3
r5
r6
rspin
Col A
Col B
Col C
Col D
col
c1
col
col
col
col
col
Idle
DCAB
r1
c1
c1
c1
CLKOUT
CT[2:0]
0,1
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
PAC
PAC
PAC
Col A
Col B
Col C
A
B
C
D
A
B
C
D
Sel A
Sel B
Sel C
Sel D
BLKW
BLKW
BLKW
BLKW
RL
A[31:0]
Row
Col D
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Command
ACTV
DCAB
Figure 104. SDRAM Burst-Length 1 Block-Write Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
127
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM block-write cycles (continued)
State
Col Pipe
r1
r2
r3
r5
r6
rspin
ColA
ColB
col
c1
col
col
col
col
col
Idle
DCAB
r1
c1
CLKOUT
CT[2:0]
2, 3
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
PAC
Cycle Type
PAC
RL
Row
A[31:0]
Col A
Col B
A
B
A
B
Sel A
Sel B
BLKW
BLKW
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
ACTV
Command
DCAB
Figure 105. SDRAM Burst-Length 2 Block-Write Cycle Timing
SVRAM transfer cycles
The SVRAM read- and write-transfer cycles transfer data between the SVRAM memory-array and the serial
register (SAM). The TMS320C80 supports both normal and split transfers for SVRAMs. Read-and split-read
transfers resemble a standard read cycle. Write-and split-write transfers resemble a standard write cycle.
Because the ’C80’s TRG output is used as CAS, external logic must generate a TRG signal (by decoding
STATUS) to enable the SVRAM transfer cycle. The value output on A[31:0] at column time represents the SAM
tap point (see Figure 106 through Figure 113).
128
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SVRAM transfer cycles (continued)
State
Col Pipe
r1
r2
r5
r3
r6
col
c1
col
c2
r1
c3
CLKOUT
CT[2:0]
000
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
Row
Tap Pt.
DCAB
RL
A[31:0]
RAS
CAS / DQM[7:0]
0 For Full, 1 For Split
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
ACTV
Command
RTR
DCAB
Figure 106. SVRAM Burst-Length 1, 2 Cycle Latency Read-Transfer Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
129
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SVRAM transfer cycles (continued)
State
Col Pipe
r1
r2
r3
col
c1
col
c2
col
c3
Cycle Type
PAC
Idle
DCAB
Row
Tap Pt.
r5
r6
CLKOUT
CT[2:0]
001
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
RL
A[31:0]
RAS
CAS / DQM[7:0]
0 For Full, 1 For Split
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
ACTV
Command
RTR
DCAB
Figure 107. SVRAM Burst-Length 1, 3 Cycle Latency Read-Transfer Cycle Timing
130
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SVRAM transfer cycles (continued)
State
Col Pipe
r1
r2
r3
r5
r6
col
c1
col
c2
col
c3
col
c4
r1
c5
CLKOUT
CT[2:0]
001
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS
[5:0]
Cycle Type
PAC
Row
Tap Pt.
Idle
DCAB
RL
A[31:0]
RAS
CAS / DQM
[7:0]
0 For Full, 1 For Split
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Command
ACTV
RTR
DCAB
Figure 108. SVRAM Burst-Length 1, 4 Cycle Latency Read-Transfer Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
131
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SVRAM transfer cycles (continued)
State
Col Pipe
r1
r2
r3
col
c1
col
c2
col
c3
Cycle Type
PAC
Idle
DCAB
Row
Tap Pt.
r5
r6
CLKOUT
CT[2:0]
010
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
RL
A[31:0]
RAS
CAS / DQM[7:0]
0 For Full, 1 For Split
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
ACTV
Command
RTR
DCAB
Figure 109. SVRAM Burst-Length 2, 2 Cycle Latency Read-Transfer Cycle Timing
132
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SVRAM transfer cycles (continued)
State
Col Pipe
r1
r2
r3
r5
r6
col
col
c1
col
c2
col
c3
r1
CLKOUT
CT[2:0]
011
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
Row
Tap Pt.
Idle
DCAB
RL
A[31:0]
RAS
CAS / DQM[7:0]
0 For Full, 1 For Split
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Command
ACTV
RTR
DCAB
Figure 110. SVRAM Burst-Length 2, 3 Cycle Latency Read-Transfer Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
133
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SVRAM transfer cycles (continued)
State
Col Pipe
r1
r2
r3
r5
r6
col
c1
col
c2
col
c3
col
c4
col
c5
CLKOUT
CT[2:0]
011
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS
[5:0]
Cycle Type
PAC
Row
Tap Pt.
Idle
DCAB
RL
A[31:0]
RAS
CAS / DQM
[7:0]
DSF
0 For Full, 1 For Split
TRG / CAS
W
D[63:0]
DBEN
DDIN
Command
ACTV
RTR
DCAB
Figure 111. SVRAM Burst-Length 2, 4 Cycle Latency Read-Transfer Cycle Timing
134
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
r1
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SVRAM transfer cycles (continued)
State
Col Pipe
r1
r2
r3
col
c1
col
col
Cycle Type
PAC
Idle
DCAB
Row
Tap Pt.
r5
r6
r1
CLKOUT
CT[2:0]
00x
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
RL
A[31:0]
RAS
CAS / DQM[7:0]
0 For Full, 1 For Split
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
ACTV
Command
WTR
DCAB
Figure 112. SVRAM Burst-Length 1, Write-Transfer Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
135
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SVRAM transfer cycles (continued)
r1
State
Col Pipe
r2
r3
r5
r6
col
c1
col
col
col
r1
CLKOUT
01x
CT[2:0]
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
Row
Tap Pt.
Idle
DCAB
RL
A[31:0]
RAS
CAS / DQM[7:0]
0 For Full, 1 For Split
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
ACTV
Command
WTR
DCAB
Figure 113. SVRAM Burst-Length 2, Write-Transfer Cycle Timing
SDRAM refresh cycle
The SDRAM refresh cycle is performed when the TC receives an SDRAM-cycle timing input (CT=0xx) at the
start of a refresh cycle. The RAS and TRG / CAS outputs are driven low for one cycle to strobe a refresh
command (REFR) into the SDRAM. The refresh address is generated internal to the SDRAM. The ’C80 outputs
a 16-bit pseudo-address (used for refresh bank decode) on A[31:16] and drives A[15:0] low (see Figure 114).
136
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
SDRAM refresh cycle (continued)
State
r1
r2
r3
r5
r6
r9
r1
CLKOUT
CT[2:0]
0xx
AS[2:0]
BS[1:0]
PS[3:0]
UTIME
FAULT
READY
RETRY
STATUS[5:0]
Cycle Type
PAC
RL
A[31:16]
Refresh Pseudo-Address
A[15:0]
RAS
CAS / DQM[7:0]
DSF
TRG / CAS
W
D[63:0]
DBEN
DDIN
Command
REFR
Figure 114. SDRAM Refresh-Cycle Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
137
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
host interface
The ’C80 contains a simple four-pin mechanism by which a host or another device can gain control of the ’C80
local memory bus. The HREQ input can be driven low by the host to request the ’C80’s bus. Once the TC has
completed the current memory access, it places the local bus (except CLKOUT) into a high-impedance state.
It then drives the HACK output low to indicate that the host device owns the bus and can drive it. The REQ[1:0]
outputs reflect the highest priority cycle request being received internally by the TC. The host can monitor these
outputs to determine if it needs to relinquish the local bus back to the ’C80 (see Table 37).
Table 37. TC Priority Cycles
REQ[1:0]
ASSOCIATED INTERNAL TC REQUEST
11
SRT, urgent refresh, XPT, or VCPT
10
Cache / DEA request, urgent packet transfer
01
High-priority packet transfer
00
Low-priority packet transfer, trickle refresh, idle
device reset
The TMS320C80 is reset when the RESET input is driven low. The ’C80 outputs immediately go into a
high-impedance state with the exception of CLKOUT, HACK, and REQ[1:0]. While RESET is low, all internal
registers are set to their default values and internal logic is reset.
On the rising edge of RESET, the state of UTIME is sampled to determine if big-endian (UTIME = 0) or
little-endian (UTIME = 1) operation is selected. Also on the rising edge of RESET, the state of HREQ is sampled
to determine if the master processor comes up running (HREQ = 0) or halted (HREQ = 1).
Once RESET is high, the ’C80 drives the high-impedance signals to their inactive values. The TC then performs
32 refresh cycles to initialize system memory. If, during initialization refresh, the TC receives an SDRAM cycle
timing code (CT = 0xx), it performs an SDRAM DCAB cycle and a MRS cycle to initialize the SDRAM, and then
continues the refresh cycles.
After completing initialization refresh, if the MP is running, the TC performs its instruction-cache-fill request to
fetch the cache block beginning at 0xFFFFFFC0. This block contains the starting MP instruction located at
0xFFFFFFF8. If the MP comes up halted, the instruction cache fill does not take place until the first occurrence
of an EINT3 interrupt to unhalt the MP.
138
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
absolute maximum ratings over specified temperature ranges (unless otherwise noted)†
Supply voltage range, VDD (see Note 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 0.3 V to 4 V
Input voltage range, VI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 0.3 V to 4 V
Output voltage range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 0.3 V to 4 V
Operating case temperature range, TC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0°C to 85°C
Storage temperature range, Tstg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – 55°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.
NOTE 1: All voltage values are with respect to VSS.
recommended operating conditions
MIN
NOM
MAX
UNIT
3.135
3.3
3.465
V
VDD
VSS
Supply voltage
IOH
IOL
High-level output current
– 400
µA
Low-level output current
2
mA
Supply voltage (see Note 2)
0
V
TC
Operating case temperature
0
85
°C
NOTE 2: In order to minimize noise on VSS, care should be taken to provide a minimum inductance path between the VSS pins and system ground.
electrical characteristics over recommended ranges of supply voltage and operating case
temperature (unless otherwise noted)
TEST CONDITIONS‡
PARAMETER
VIH
VIL
High-level input voltage
VOH
VOL
High-level output voltage
Low-level input voltage
– 0.3
VDD = MIN,
VDD = MAX,
Low-level output voltage
Output current,, leakage
g ((high
g impedance))
(except EMU0 and EMU1)
II
Input current (except TCK, TDI, and TMS)
Ci
Supply current (see Note 3)
TYP§
2
IO
IDD
MIN
IOH = MAX
IOH = MIN
2.6
20
40 MHz
0.9#
10
V
V
– 20
50 MHz
Input capacitance
V
0.6
1.2#
1.0#
60 MHz
UNIT
¶
VDD = MAX,
VO = 2.8 V
VDD = MAX,
VO = 0.6 V
VI = VSS to VDD
VDD = MAX,
VDD = MAX,
VDD = MAX,
MAX
VDD + 0.3
0.8
V
µA
± 20
2.5#
µA
2.3#
1.9#
A
pF
Co
Output capacitance
10
pF
‡ For conditions shown as MIN / MAX, use the appropriate value specified under the recommended operating conditions.
§ All typical values are at VDD = 3.3 V, ambient air temperature = 25°C
¶ Typical steady-state VOH will not exceed VDD
# Parameter value is representative of revision 4.x and higher devices.
NOTE 3: Maximum supply current is derived from a test case that generates the theoretical maximum data flow using a worst case checkerboard
data pattern on a sustained cycle by cycle basis. Actual maximum IDD varies in real applications based on internal and external data
flow and transitions. Typical supply current is derived from a test case which attempts to emulate typical use conditions of the on-chip
processors with random data. Typical IDD varies from application to application based on data flow and transitions and on-chip processor
utilization.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
139
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PARAMETER MEASUREMENT INFORMATION
IOL
Tester Pin
Electronics
50 Ω
VLOAD
Output
Under
Test
CT
IOH
Where:
IOL
IOH
VLOAD
CT
=
=
=
=
2.0 mA (all outputs)
400 µA (all outputs)
1.5 V
60 pF typical load circuit capacitance
Figure 115. Test Load Circuit
signal transition levels
TTL-output levels are driven to a minimum logic-high level of 2.4 V and to a maximum logic-low level of 0.6 V.
Figure 116 shows the TTL-level outputs.
2.4 V
2V
0.8 V
0.6 V
Figure 116. TTL-Level Outputs
TTL-output transition times are specified as follows:
D
D
For a high-to-low transition, the level at which the output is said to be no longer high is 2 V, and the level
at which the output is said to be low is 0.8 V.
For a low-to-high transition, the level at which the output is said to be no longer low is 0.8 V, and the level
at which the output is said to be high is 2 V.
Figure 117 shows the TTL-level inputs.
2V
0.8 V
Figure 117. TTL-Level Inputs
TTL-compatible input transition times are specified as follows:
D
D
140
For a high-to-low transition on an input signal, the level at which the input is said to be no longer high is
2 V, and the level at which the input is said to be low is 0.8 V.
For a low-to-high transition on an input signal, the level at which the input is said to be no longer low is
0.8 V, and the level at which the input is said to be high is 2 V.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
PARAMETER MEASUREMENT INFORMATION
timing parameter symbology
Timing parameter symbols used herein were created in accordance with JEDEC Standard 100-A. In order to
shorten the symbols, some of the pin names and other related terminology have been abbreviated as follows:
A
A[31 : 0]
RDY
READY
CAS
CAS / DQM[7 : 0]
RST
RESET
CFG
AS[2:0], BS[1:0], CT[2:0], PS[3:0], UTIME
RTY
RETRY
CKI
CLKIN
REQ
REQ[1:0]
CKO
CLKOUT
RL
RL
CMP
RETRY, READY, FAULT
RR
READY, RETRY
D
D[63:0]
SCK
SCLK0, SCLK1
EIN
EINT1, EINT2, EINT3, or EINTx
TCK
TCK
EMU
EMU0, EMU1
TDI
TDI
FCK
FCLK0, FCLK1
TDO
TDO
HAK
HACK
TMS
TMS
HRQ
HREQ
TRS
TRST
LIN
LINT4
UTM
UTIME
MID
A[31:0], STATUS[5:0]
SI
HSYNC0, VSYNC0, CSYNC0, HSYNC1, VSYNC1,
or CSYNC1
OUT
A[31:0], CAS / DQM[7 : 0], D[63:0], DBEN, DDIN,
DSF, RAS, RL, STATUS[5:0], TRG / CAS, W
SY
HSYNC0, VSYNC0, CSYNC0 / HBLNK0,
CBLNK0 / VBLNK0, HSYNC1, VSYNC1,
CSYNC1 / HBLNK1, CBLNK1 / VBLNK1, CAREA0,
or CAREA1
RAS
RAS
XPT
XPT[2:0] OR XPTx
Lowercase subscripts and their meanings are:
The following letters and symbols and their meanings are:
a
access time
H
High
c
cycle time (period)
L
Low
d
delay time
V
Valid
h
hold time
Z
High impedance
su
setup time
X
Unknown, changing, or don’t care level
t
transition time
w
pulse duration (width)
general notes on timing parameters
The period of the output clock (CLKOUT) is twice the period of the input clock (CLKIN), or 2 × tc(CKI). The half
cycle time (tH) that appears in the following tables is one-half of the output clock period, or equal to the input
clock period, tc(CKI).
All output signals from the ’C80 (including CLKOUT) are derived from an internal clock such that all output
transitions for a given half cycle occur with a minimum of skewing relative to each other.
The signal combinations shown in the following timing diagrams may not necessarily represent actual cycles.
For actual cycle examples, refer to the appropriate cycle description section of this data sheet.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
141
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
CLKIN timing requirements (see Figure 118)
’C80-40
NO
NO.
1
2
3
4
MIN
’C80-50
MAX
MIN
’C80-60
MAX
MIN
MAX
UNIT
tc(CKI)
tw(CKIH)
Period of CLKIN (tH)
12.5
10
8.3
ns
Pulse duration of CLKIN high
4.8
4.2
3.9
ns
tw(CKIL)
tt(CKI)
Pulse duration of CLKIN low
Transition time of CLKIN†
4.8
4.2
3.9
1.5
ns
1.5
1.5
ns
† This parameter is verified by computer simulation and is not tested.
1
4
2
4
CLKIN
3
Figure 118. CLKIN Timing
local-bus switching characteristics over full operating range: CLKOUT‡(see Figure 119)
NO
NO.
5
6
7
’C80-40
PARAMETER
MIN
’C80-50
MAX
MIN
’C80-60
MAX
MIN
Pulse duration of CLKOUT high
2tc(CKI)§
tH – 5.5
2tc(CKI) §
tH – 4.5
2tc(CKI) §
tH – 3.7
Pulse duration of CLKOUT low
tH – 5.5
tH – 4.5
tH – 3.7
tc(CKO)
tw(CKOH)
Period of CLKOUT
tw(CKOL)
tt(CKO)
2¶
2¶
MAX
UNIT
ns
ns
ns
2¶
8
Transition time of CLKOUT
ns
‡ The CLKOUT output has twice the period of CLKIN. No propagation delay or phase relationship to CLKIN is assured. Each state of a memory
access begins on the falling edge of CLKOUT.
§ This is a functional minimum and is not tested. This parameter may also be specified as 2tH.
¶ This parameter is verified by computer simulation and is not tested.
tH
5
tH
tH
tH
8
6
8
CLKOUT
7
Figure 119. CLKOUT Timing
142
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
device reset timing requirements (see Figure 120)
NO.
MIN
Initial reset during power-up
MAX
UNIT
6th
6th
ns
ns
9
tw(RSTL)
(RSTL)
P lse d
ration RESET low
lo
Pulse
duration,
10
Setup time of HREQ low to RESET high to configure self-bootstrap mode
11
tsu(HRQL-RSTH)
th(RSTH-HRQL)
Hold time, HREQ low after RESET high to configure self-bootstrap mode
4th
0
12
tsu(UTML-RSTH)
Setup time of UTIME low to RESET high to configure big-endian operation
4th
ns
13
th(RSTH-UTML)
Hold time, UTIME low after RESET high to configure big-endian operation
0
ns
Reset during active operation
ns
ns
9
RESET
10
11
HREQ
13
12
UTIME
Figure 120. Device-Reset Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
143
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
local bus timing requirements: cycle configuration inputs (see Figure 121)
The cycle configuration inputs are sampled at the beginning of each row access during the r2 state. The inputs
typically are generated by a static decode of the A[31:0] and STATUS[5:0] outputs.
NO.
14
15
16
MIN
tsu(CFGV-CKOH)
th(CKOH-CFGV)
ta(MIDV-CFGV)
UNIT
8
ns
Hold time, AS, BS, CT, PS, and UTIME valid after CLKOUT high
2
ns
Access time, AS, BS, CT, PS, and UTIME valid after memory identification (A,
STATUS) valid
tH
tH
tH
tH
tH
CLKOUT
Cycle Type
STATUS[5:0]
Row Address
A[31:0]
15
RL
16
14
AS[2:0]
Valid
BS[1:0]
Valid
CT[2:0]
Valid
PS[3:0]
Valid
UTIME
Valid
Figure 121. Local Bus Timing: Cycle Configuration Inputs
144
MAX
Setup time, AS, BS, CT, PS, and UTIME valid to CLKOUT no longer low
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
3tH – 10
tH
ns
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
local bus timing: cycle completion inputs (see Figure 122 and Figure 123)
The cycle completion inputs are sampled at the beginning of each row access at the start of the r3 state. The
READY input is also sampled at the start of the r6 state and during each column access (2 and 3 cyc/col
accesses only). The RETRY input is sampled on each CLKOUT falling edge following r3. The value n as used
in the parameters represents the integral number of half cycles between the transitions of the two signals in
question.
’C80-40
NO
NO.
MIN
MAX
’C80-50
MIN
MAX
’C80-60
MIN
MAX
UNIT
17
ta(MIDV-CMPV)
Access time, RETRY, READY, FAULT valid
after memory identification (A, STATUS)
valid
18
tsu(CMPV-CKOL)
Setup time, RETRY, READY, FAULT valid to
CLKOUT no longer high
8.0
7.5
7.5
ns
19
th(CKOL-CMPV)
Hold time, RETRY, READY, FAULT valid
after CLKOUT low
1.2
1.2
1.2
ns
20
ta(RASL-RRV)
Access time RETRY, READY valid from
RAS low
ntH–8
ntH–7.5
ntH–7.5
ns
21
ta(RLL-RRV)
Access time, RETRY, READY valid from RL
low
ntH–8
ntH–7.5
ntH–7.5
ns
ta(CASL-RRV)
a(CASL RRV)
Access time, READY
valid from CAS low
(non-usertime mode)
2 cyc/col accesses
22
tH–13.5
tH–12
tH–12
3 cyc/col accesses
2tH–9
2tH–8
2tH–7
POST OFFICE BOX 1443
ntH–9
• HOUSTON, TEXAS 77251–1443
ntH–8
ntH–7
ns
ns
145
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
local bus timing: cycle completion inputs (continued)
tH
tH
tH
tH
tH
tH
tH
tH
CLKOUT
STATUS[5:0]
A[31:0]
RL
RAS
17
18
21
19
20
RETRY
READY
FAULT
Figure 122. Local Bus Timing: Row-Time Cycle Completion Inputs
146
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
tH
tH
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
local bus timing: cycle completion inputs (continued)
tH
tH
tH
tH
tH
tH
tH
tH
tH
tH
CLKOUT
STATUS[5:0]
A[31:0]
CAS/DQM[7:0]
18
22
19
17
READY
RETRY
Figure 123. Local Bus Timing: Column-Time Cycle Completion Inputs
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
147
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
general output signal characteristics over full range of operating conditions
The following general timing parameters apply to all TMS320C80 output signals unless otherwise specifically
given. The value n as used in the parameters represents the integral number of half cycles between the
transitions of the two outputs in question. For timing purposes, outputs fall into one of three groups – the data
bus (D[63:0]); the other output buses (A[31:0], STATUS[5:0], CAS/DQM[7:0]); and non-bus outputs (DBEN,
DDIN, DSF, RAS, RL, TRG/CAS, W). When measuring output to output, the named group refers to the first
output to transition (output A), and the second output (output B) refers to any output group (see Figure 124).
148
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
general output signal characteristics over full range of operating conditions†(continued)
NO
NO.
23
24
’C80-40
PARAMETER
th(OUTV-CKOL)
th(OUTV-CKOH)
MIN
Hold time, CLKOUT high after
output valid
D[63:0]
A[31:0], STATUS[5:0],
CAS/DQM[7:0‡
DBEN, DDIN, DSF, RAS, RL,
TRG/CAS, W
’C80-50
MAX
MIN
’C80-60
MAX
MIN
ntH–7
ntH–5.3
ntH–5.3
ntH–6.5
ntH–4.3
ntH–4.3
ntH–5.5
ntH–3.9
ntH–3.9
MAX
UNIT
ns
Hold time, CLKOUT low after
output valid
D[63:0]
A[31:0]
STATUS[5:0], CAS/DQM[7:0]‡
DBEN, DDIN, DSF, RAS, RL,
TRG/CAS, W
ntH–7
ntH–6.5
ntH–6.5
ntH–5.5
ntH–4.5
ntH–4.5
ntH–4
ntH–4
ntH–4.5
ntH–5.5
ntH–4.1
ntH–4.1
ns
25
th(CKOL-OUTV)
Hold time, output valid after
CLKOUT low
ntH–5.5
ntH–5
ntH–5
ns
26
th(CKOH-OUTV)
Hold time, output valid after
CLKOUT high
ntH–5.5
ntH–5
ntH–4
ns
ntH–7
ntH–6.5
ntH–5.9
th(OUTV-OUTV)
Hold time, output valid after output
valid
D[63:0]
A[31:0], STATUS[5:0],
CAS/DQM[7:0]‡
DBEN, DDIN, DSF, RAS, RL,
TRG/CAS, W
ntH–6.5
ntH–5.5
ntH–5.5
ntH–5.5
ntH–5
ntH–4.7
27
28
29
td(CKOH-OUTV)
td(CKOL-OUTV)
Delay time, CLKOUT no longer
low to output valid
D[63:0]
A[31:0], STATUS[5:0],
CAS/DQM[7:0]‡
DBEN, DDIN, DSF, RAS, RL,
TRG/CAS, W
ntH+7
ntH+6.5
ntH+5.9
ntH+6.5
ntH+5.5
ntH+5.5
ntH+5.5
ntH+5
ntH+4.7
ntH+7
ntH+6.5
ntH+5.9
ns
Delay time, CLKOUT no longer
high to output valid
D[63:0]
A[31:0], STATUS[5:0],
CAS/DQM[7:0]‡
DBEN, DDIN, DSF, RAS, RL,
TRG/CAS, W
ntH+6.5
ntH+5.5
ntH+5.5
ntH+5.5
ntH+5
ntH+4.7
ntH+5.5
ntH+5
ntH+5
ns
ntH+5.5
ntH+5
ntH+5
ns
ntH+7
ntH+6.5
ntH+6.1
ntH+6.5
ntH+5.5
ntH+5.5
30
td(OUTV-CKOH)
Delay time, output no longer valid
to CLKOUT high
31
td(OUTV-CKOL)
Delay time, output no longer valid
to CLKOUT low
td(OUTV-OUTV)
Delay time, output no longer valid
to output valid
D[63:0]
A[31:0], STATUS[5:0],
CAS/DQM[7:0]‡
DBEN, DDIN, DSF, RAS, RL,
TRG/CAS, W
32
ns
ns
ns
ntH+5.5
ntH+5
ntH+5
† Tested across full voltage. Test temperature is selected by manufacturing test flow. Compliance across full temperature range is ensured by
device characterization.
‡ Except for CAS/DQM[7:0] during non user-timed 2 cycle/column accesses
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
149
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
general output signal characteristics over full range of operating conditions† (continued)
NO
NO.
33
’C80-40
PARAMETER
MIN
Pulse duration, output valid
D[63:0]
A[31:0], STATUS[5:0],
CAS/DQM[7:0]‡
DBEN, DDIN, DSF, RAS, RL,
TRG/CAS, W
tw(OUTV)
’C80-50
MAX
MIN
’C80-60
MAX
MIN
MAX
ntH–7
ntH–6.5
ntH–6.1
ntH–6.5
ntH–5.5
ntH–5.5
UNIT
ns
ntH–5.5
ntH–5
ntH–5
† Tested across full voltage. Test temperature is selected by manufacturing test flow. Compliance across full temperature range is ensured by
device characterization.
‡ Except for CAS/DQM[7:0] during non user-timed 2 cycle/column accesses
tH
tH
tH
tH
tH
tH
tH
tH
tH
tH
tH
CLKOUT
30
26
28
24
OutputA
25
31
32
23
27
29
OutputB
33
Figure 124. General Output-Signal Timing
150
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
tH
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
data input timing
The following general timing parameters apply to the D[63:0] inputs unless otherwise specifically given. The
value n as used in the parameters represents the integral number of half cycles between the transitions of the
output and input in question (see Figure 125).
NO
NO.
’C80-40
PARAMETER
MIN
’C80-50
MAX
MIN
’C80-60
MAX
MIN
MAX
UNIT
34
ta(CKOH-DV)
Access time, CLKOUT high to D[63:0]
valid
ntH–8
ntH–5.3
ntH–4.0
ns
35
ta(CKOL-DV)
Access time, CLKOUT low to D[63:0]
valid
ntH–8
ntH–6.5
ntH–6.5
ns
36
tsu(DV-CKOH)
Setup time, D[63:0] valid to CLKOUT no
longer low
8
6.1
6.1
ns
37
tsu(DV-CKOL)
Setup time, D[63:0] valid to CLKOUT no
longer high
8
6.1
6.1
ns
38
th(CKOL-DV)
Hold time, D[63:0] valid after CLKOUT
low
2
2
2
ns
39
th(CKOH-DV)
Hold time, D[63:0] valid after CLKOUT
high
2
2
2
ns
ta(OUTV-DV)
Access time, output valid to D[63:0]
inputs valid
A[31:0], CAS/DQM[7:0]†, STATUS[5:0]
DBEN, DDIN, DSF, RAS, RL,
TRG/CAS, W
th(OUTV-DV)‡
Hold time, D[63:0] valid after output valid
RAS, CAS/DQM[7:0]
A[31:0]
40
41
ntH–9
ntH–7
ntH–7
ntH–8
ntH–6.5
ntH–6.5
2
3
2
3
ns
2
3
ns
† Except CAS/DQM[7:0] during non user-timed 2 cycle/column accesses
‡ Applies to RAS, CAS/DQM[7:0], and A[31:0] transitions that occur on CLKOUT edge coincident with input data sampling
tH
tH
tH
tH
tH
tH
tH
tH
tH
tH
tH
tH
CLKOUT
38
39
37
36
34
35
D[63:0]
41
40
Output
Figure 125. Data-Input Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
151
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
local bus timing: 2 cycle/column CAS timing
These timing parameters apply to the CAS/DQM[7:0] signals during 2 cycle per column memory accesses only.
They should be used in place of the general output and data input timing parameters when the 2 cycle/column
(non user-timed) cycle timing is selected (CT[2:0] inputs = 0b110). The value n as used in the parameters
represents the integral number of half cycles between the transitions of the signals in question (see Figure 126).
MIN
42
43
’C80-50
’C80-60
’C80-40
NO.
tw(CASH)
tw(CASL)
Pulse duration, CAS/DQM high
MAX
MIN
tH–2
3tH–11
tH–2
3tH–9.5
Hold time, CAS/DQM high after output valid
D[63:0]
A[31:0], STATUS[5:0]
DBEN, DDIN, DSF, RAS, RL, TRG/CAS, W
ntH–5
ntH–4.5
ntH–3.5
ntH–11
ntH–4.5
ntH–3.5
ntH–3
ntH–9.5
Pulse duration, CAS/DQM low
44
th(OUTV-CASL)
45
th(CASL-OUTV)
ta(CASL-DV)
Hold time, output valid after CAS/DQM low
46
47
th(CASH-DV)
Hold time, data valid after CAS/DQM high
Access time, data valid from CAS/DQM low
tH
tH
tH
tH
3tH–12
2
tH
tH
tH
CLKOUT
42
43
CAS/DQM[7:0]
44
45
47
46
D[63:0]
Figure 126. 2 Cycle / Column CAS Timing
152
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
ns
ns
ns
ns
3tH–12
2
tH
Output
UNIT
MAX
tH
ns
ns
tH
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
external-interrupt timing
The following description defines the timing of the edge-triggered interrupts EINT1 – EINT3 and the level
triggered interrupt LINT4 (see Note 4). See Figure 127.
’C80-40
NO.
MIN
48
49
50
tw(EINL)
tsu(EINH-CKOH)
tw(EINH)
tsu(LINL-CKOL)
Pulse duration, EINTx low†
Setup time, EINTx high before CLKOUT no longer low‡
Pulse duration, EINTx high†
MAX
’C80-50
’C80-60
MIN
UNIT
MAX
6
6
ns
11.5
9.5
ns
6
6
ns
8
ns
Setup time, LINT4 low before CLKOUT no longer high‡
51
10
† This parameter is assured by characterization and is not tested.
‡ This parameter must only be met to ensure that the interrupt is recognized on the indicated cycle.
NOTE 4: In order to assure recognition, LINT4 must remain low until cleared by the interrupt service routine.
Interrupt Recognized
CLKOUT
49
50
EINTx
51
48
LINT4
Figure 127. External-Interrupt Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
153
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
XPT input timing
The following description defines the sampling of the XPT[2:0] inputs. The value encoded on the XPT[2:0] inputs
is synchronized over multiple cycles to ensure that a stable value is present (see Figure 128 and Figure 129).
MIN
52
53
54
’C80-50
’C80-60
’C80-40
NO.
tw(XPTV)
tsu(XPTV-CKOH)
Pulse duration, XPTx valid†
th(CKOH-XPTV)
th(RLL-XPTV)
Hold time, XPT[2:0] valid after CLKOUT high
Hold time, XPT[2:0] valid after RL low§
Setup time, XPT[2:0] valid before CLKOUT no longer low‡
MAX
MIN
UNIT
MAX
12tH
13.5
12tH
12
ns
5
5
ns
ns
55
6tH
6tH
ns
† This parameter is a functional minimum assured by logic and is not tested.
‡ This parameter must only be met to ensure that the XPT input is recognized on the indicated cycle.
§ This parameter must be met to ensure that a second XPT request does nor occur. This parameter is a functional maximum assured by logic and
is not tested.
XPT Inputs Sampled
XPT Inputs Recognized
CLKOUT
53
54
52
XPT[2:0]
Figure 128. XPT Input Timing – XPT Recognition
CLKOUT
STATUS[5:0]
XPTn Row Status
RL
55
XPT[2:0]
XPTn
XPTz
Figure 129. XPT Input Timing – XPT Service
154
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
host-interface timing (see Figure 130)
NO
NO.
’C80-40
’C80-50
MIN
MIN
MAX
’C80-60
MAX
MIN
MAX
UNIT
56
tsu(REQV-CKOH)
Setup time, REQ1 – REQ0 valid to CLKOUT no
longer low
tH – 7
tH – 7
tH – 5.5
ns
57
th(CKOH-REQV)
Hold time, REQ1 – REQ0 valid after CLKOUT
high
tH – 7
tH – 7
tH – 5.5
ns
58
th(HRQL-HAKL)
Hold time for HACK high after HREQ goes low†
4tH – 12
4tH – 12
4tH – 12
ns
59
60
td(HAKL
OUTZ)
d(HAKL-OUTZ)
Delay time, HACK low
to output
out ut hi-Z‡
All signals except
D[63:0]
0
0
0
D[63:0]
1
1
1
61
td(HRQH- HAKH)
td(HAKH-OUTD)
Delay time, HREQ high to HACK no longer low
Delay time, HACK high to outputs driven†
62
tsu(HRQL-CKOH)
Setup time, HREQ low to CLKOUT no longer
low (see Note 5)
10
10
10
6tH
6tH
6tH
10.5
8.5
8.5
ns
ns
ns
† This parameter is a functional minimum assured by logic and is not tested.
‡ This parameter is assured by characterization and is not tested.
NOTE 5: Parameter must be met only to ensure HREQ recognition during the indicated clock cycle.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
155
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
host-interface timing (see Figure 130) (continued)
HREQ Sampled
CLKOUT
56
57
REQ[1:0]
HREQ
62
58
60
HACK
61
59
A[31:0]
RL, TRG,
WE, DSF,
DSF2, DBEN
RAS
CAS[7:0]
D[63:0]
DDIN
Figure 130. Host-Interface Timing
156
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
video interface timing: SCLK timing (see Figure 131)
NO.
63
64
65
MIN
tc(SCK)
tw(SCKH)
SCLK period
tw(SCKL)
tt(SCK)
MAX
UNIT
13
ns
Pulse duration, SCLK high
5
ns
Pulse duration, SCLK low
5
ns
Transition time, SCLK (rise and fall)†
66
† This parameter is assured by simulation and is not tested.
2
ns
63
66
64
66
SCLK0
SCLK1
65
Figure 131. Video Interface Timing: SCLK Timing
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
157
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
video interface timing: FCLK input and video outputs (see Note 6 and Figure 132)
NO.
67
68
69
70
MIN
tc(FCK)
tw(FCKH)
FCLK period
tw(FCKL)
tt(FCK)
MAX
UNIT
25
ns
Pulse duration, FCLK high
8
ns
Pulse duration, FCLK low
8
ns
Transition time, FCLK (rise and fall)†
2
ns
71
th(FCKL-SYL)
Hold time, HSYNC, VSYNC, CSYNC / HBLNK, CBLNK / VBLNK, or CAREA high after
FCLK low
72
th(FCKL-SYH)
Hold time, HSYNC, VSYNC, CSYNC / HBLNK, CBLNK / VBLNK, or CAREA low after FCLK
low
73
td(FCKL-SYL)
Delay time, FCLK no longer high to HSYNC, VSYNC, CSYNC / HBLNK, CBLNK / VBLNK,
or CAREA low
20
ns
74
td(FCKL-SYH)
Delay time, FCLK no longer high to HSYNC, VSYNC, CSYNC / HBLNK, CBLNK / VBLNK,
or CAREA high
20
ns
0
ns
0
ns
† This parameter is assured by simulation and is not tested.
NOTE 6: Under certain circumstances these outputs also can transition asynchronously. These transitions occur when controller timing register
values are modified by user programming. If the new register value forces the output to change states then this transition occurs without
regard to FCLK inputs.
70
68
67
70
69
FCLK0
FCLK1
74
73
72
HSYNCn, VSYNCn,
CSYNCn / HBLNKn
CBLNKn / VBLNKn
CAREAn
71
Figure 132. Video Interface Timing: FCLK Input and Video Outputs
158
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
video interface timing: external sync inputs
When configured as inputs, the HSYNCn, VSYNCn, and CSYNCn signals may be driven asynchronously. The
following parameters apply only when the inputs are being generated synchronous to FCLKn in order to ensure
recognition on a particular FLCKn edge (see Figure 133).
NO.
75
76
77
78
MIN
MAX
UNIT
tsu(SIL-FCKH)
th(FCKH-SIL)
Setup time, HSYNC, VSYNC, or CSYNC low to FCLK no longer low†
Hold time, HSYNC, VSYNC, or CSYNC high after FCLK high‡
5
ns
7
ns
tsu(SIH-FCKH)
th(FCKH-SIH)
Setup time, HSYNC, VSYNC, or CSYNC high to FCLK no longer low§
Hold time, HSYNC, VSYNC, or CSYNC low after FCLK high¶
5
ns
7
ns
† This parameter must be met only to ensure the input is recognized as low at FLCK edge B.
‡ This parameter must be met only to ensure the input is recognized as high at FLCK edge A.
§ This parameter must be met only to ensure the input is recognized as high at FLCK edge D.
¶ This parameter must be met only to ensure the input is recognized as low at FLCK edge C.
A
B
C
D
FCLK0
FCLK1
76
77
78
HSYNC0, HSYNC1
VSYNC0, VSYNC1
CSYNC0, CSYNC1
(Inputs)
75
Figure 133. Video Interface Timing: External Sync Inputs
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
159
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
thermal resistance
Figure 134 illustrates the maximum ambient temperature allowed for various air flow rates across the
TMS320C80 to ensure that the case temperature is kept below the maximum operating temperature (85°C) (see
Note A). Values for the GF package include integral heat sink. Values for the GGP package are with no heat
sink.
Max T A ( °C)
Maximum Ambient Temperature Versus Airflow
For GF and GGP Packages
85
40 MHz (GF)
80
50 MHz (GF)
75
60 MHz (GF)
40 MHz (GGP)
70
50 MHz (GGP)
65
60 MHz (GGP)
60
55
50
45
40
0
100
200
300
400
500
600
700
800
900
1000
Air Flow (Linear Ft / Min)
NOTE A: TMS320C80 power consumption is based on the “typical” values of IDD measured at VDD = 3.3 V. Power consumption varies by
application based on TMS320C80 processor activity and I/O pin loadings. User must ensure that the case temperature (TC)
specifications are met when defining airflow and other thermal constraints of their system.
Figure 134. Airflow Requirements
160
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
emulator interface connection
The ’C80 supports emulation through a dedicated emulation port that is a superset of the IEEE Standard 1149.1
(JTAG) Standard. To support the ’C80 emulator, a target system must include a 14-pin header (2 rows of 7 pins)
with the connections shown in Figure 135.
TMS
2
TRST
3
4
GND
PD(+3.3V)
5
6
No pin (key)
TDO
7
8
GND
TCKRET
TDI
1
9
10
GND
TCK
11
12
GND
EMU0
13
14
EMU1
Pin Spacing: 0.100 in. (X,Y)
Pin Width: 0.025 in, square post
Pin Length: 0.235 in. nominal
(see Table 38)
Figure 135. Target System Header
Table 38. Target Connectors
XDS 510
SIGNAL
XDS 510
STATE
TARGET
STATE
TMS
O
I
TDI
O
I
Test-mode select†
Test-data input†
TDO
I
O
Test-data output†
TCK
O
I
TRST
O
I
Test clock – 10 MHz clock source from emulator. Can be used to drive system-test clock.†
Test reset†
EMU0
I
I/O
Emulation pin 0
EMU1
I
I/O
Emulation pin 1
PD (3.3 V)
I
O
Presence detect. Indicates that the target is connected and powered up. Should be tied to
+ 3.3 V on target system.
TCKRET
I
O
Test clock return. Test clock input to the XDS 510 emulator. Can be buffered or unbuffered
version of TCK.†
DESCRIPTION
† IEEE Standard 1149.1.
For best results, the emulation header should be located as close as possible to the ’C80. If the distance exceeds
six inches, the emulation signals should be buffered. See Figure 136.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
161
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
emulator-interface connection (continued)
3.3 V
EMU0
EMU1
TRST
TMS
TDI
TDO
T34 (L26)
13
J35 (K25)
14
L33 (J24)
2
N33 (J26)
1
H34 (K23)
3
P32 (K24)
7
E35 (J25)
11
TCK
’C80
GF (GGP)
Signal
Pins
9
6 in. or less
3.3 V
3.3 V
EMU0
3.3 V
5
PD
EMU1
EMU0
EMU1
4
TRST
GND
TMS
GND
TDI
GND
TDO
GND
TCK
GND
TRST
6
TMS
8
TDI
10
TDO
12
T34 (L26)
13
J35 (K25)
14
L33 (J24)
2
N33 (J26)
1
H34 (K23)
3
P32 (K24)
7
E35 (J25)
11
TCK
TCKRET
Emulator
Header
9
’C80
GF (GGP)
Signal
Pins
More than 6 in.
3.3 V
5
EMU0
PD
EMU1
4
TRST
GND
TMS
GND
TDI
GND
TDO
GND
TCK
GND
6
8
10
12
TCKRET
Emulator
Header
Figure 136. Emulation Header Connections – Emulator Driven Test Clock
The target system also can generate the test clock. This allows the user to:
D
D
Set the test clock frequency to match the system requirements. (The emulator provides only a 10-MHz
test clock.)
Have other devices in the system that require a test clock when the emulator is not connected
3.3 V
EMU0
EMU1
TRST
TMS
TDI
TDO
TCK
3.3 V
T34 (L26)
13
J35
(K25)
14
L33 (J24)
2
N33 (J26)
1
H34 (K23)
3
P32 (K24)
7
E35 (J25)
11
9
’C80
GF (GGP) Signal
Pins
System
Test
Clock
PD
EMU0
EMU1
TRST
GND
TMS
GND
TDI
GND
TDO
GND
TCK
GND
3.3 V
5
4
6
8
10
12
TCKRET
Emulator
Header
More than 6 in.
Figure 137. Emulation Header Connections – System Driven Test Clock
162
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
emulator-interface connection (continued)
For multiprocessor applications, the following conditions are recommended:
D
D
D
D
To reduce timing skew, buffer TMS, TDI, TDO, and TCK through the same physical package.
If buffering is used, 4.7 kΩ resistors are recommended for TMS, TDI, and TCK which should be pulled
high (3.3 V).
Buffering EMU0 and EMU1 is highly recommended to provide isolation. The buffers need not be in the
same physical package as TMS, TCK, TDI, or TDO. Pullups to 3.3 V are required and should provide
a signal rise time of less than 10 µs. A 4.7 kΩ resistor is suggested for most applications.
To ensure high quality signals, special printed wire board (PWB) routing and use of termination resistors
may be required. The emulator provides fixed series termination (33 Ω) on TMS and TDI and optional
parallel terminators (180 Ω pullup and 270 Ω pulldown) on TCKRET and TDO.
3.3 V
EMU0
EMU1
TRST
TMS
TDI
TDO
TCK
3.3 V
T34 (L26)
13
J35 (K25)
14
L33 (J24)
2
N33 (J26)
1
H34 (K23)
3
P32 (K24)
7
E35 (J25)
11
9
EMU0
EMU1
TRST
GND
TMS
GND
TDI
GND
TDO
GND
TCK
GND
EMU1
TRST
TMS
TDI
TDO
TCK
4
6
8
10
12
TCKRET
’C80
GF (GGP) Signal
Pins
EMU0
PD
3.3 V
5
Emulator
Header
T34 (L26)
J35 (K25)
L33 (J24)
N33 (J26)
H34 (K23)
P32 (K24)
E35 (J25)
’C80
GF (GGP) Signal
Pins
Figure 138. Emulation Header Connections – Multiprocessor Applications
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
163
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
GF package drawing
AR
AP
AN
AM
AL
AK
AJ
AH
AG
AF
AE
AD
AC
AB
AA
Y
W
V
U
T
R
P
N
M
L
K
J
H
G
F
E
D
C
B
A
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31323334 35
Index Mark
(Gold Plate)
Index Pin
M5 x 0.8
STUD
1.895 (48,13)
DIA
1.905 (48,39)
ÉÉÉ
ÉÉÉ
ÉÉÉ
0.279 (7,09)
0.299 (7,59)
~
0.236 (6,00)
NOTES: A. Pins are located within 0,13 (0.005) radius of the true position relative to each other at maximum material condition and within
0,457 (0.018) radius of the center of the ceramic.
B. Dimensions do not include solder finish.
Figure 139. Assembled Package Drawing Showing Integral Heatsink
164
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MECHANICAL DATA
GF (S-CPGA-P305)
CERAMIC PIN GRID ARRAY PACKAGE
1.717 (43,61)
TYP
1.683 (42,75)
1.879 (47,73)
SQ
1.841 (46,76)
0.100 (2,54)
AR
AN
AL
AJ
AG
AE
AC
AA
W
U
R
N
L
J
G
E
C
A
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
Heatsink
0.050 (1,27)
0.060 (1,52)
0.040 (1,02)
0.019 (0,48)
0.014 (0,36)
0.026 (0,660)
0.006 (0,152)
0.045 (1,14) DIA 4 Places
0.040 (1,02)
0.190 (4,83)
0.170 (4,32)
0.150 (3,81)
0.180 (4,57)
0.110 (2,79)
0.140 (3,56)
4040035-3/E 03/97
NOTES: A. All linear dimensions are in inches (millimeters).
B. This drawing is subject to change without notice.
C. Package thickness of 0.150 (3,81) / 0.110 (2,79) includes package body and lid, but does not include integral heatsink or attached
features.
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
165
TMS320C80
DIGITAL SIGNAL PROCESSOR
SPRS023B – JULY 1994 – REVISED OCTOBER 1997
MECHANICAL DATA
GGP (S-PBGA-N352)
PLASTIC BALL GRID ARRAY (CAVITY DOWN) PACKAGE
31,75 SQ
35,20
SQ
34,80
1,27
26 24 22 20 18 16 14 12 10 8
6
4
2
25 23 21 19 17 15 13 11 9
7
5
3
1
A
B
C
D
E
F
G
H
J
K
L
M
N
P
R
T
U
V
W
Y
AA
AB
AC
AD
AE
AF
Heat Slug
1,70 MAX
0,91 NOM
Seating Plane
0,90
0,60
0,50 MIN
0,15
0,30 M
4073223/A 11/96
NOTES: A. All linear dimensions are in millimeters.
B. This drawing is subject to change without notice.
C. Thermally enhanced die down plastic package with top surface metal heat slug.
166
POST OFFICE BOX 1443
• HOUSTON, TEXAS 77251–1443
PACKAGE OPTION ADDENDUM
www.ti.com
9-May-2005
PACKAGING INFORMATION
Orderable Device
Status (1)
Package
Type
Package
Drawing
Pins Package Eco Plan (2)
Qty
Lead/Ball Finish
MSL Peak Temp (3)
TMS320C80GF
OBSOLETE
CPGA
GF
305
TBD
Call TI
Call TI
TMS320C80GF50
NRND
CPGA
GF
305
10
TBD
Call TI
Level-NC-NC-NC
TMS320C80GF60
NRND
CPGA
GF
305
10
TBD
Call TI
Level-NC-NC-NC
TMS320C80GGP50
NRND
BGA
GGP
352
24
TBD
SNPB
Level-4-220C-72HR
TMS320C80GGP60
NRND
BGA
GGP
352
1
TBD
SNPB
Level-4-220C-72HR
TMX320C80GF50
OBSOLETE
CPGA
GF
305
TBD
Call TI
Call TI
TMX320C80GF60
OBSOLETE
CPGA
GF
305
TBD
Call TI
Call TI
TMX320C80GGP
OBSOLETE
BGA
GGP
352
TBD
Call TI
Call TI
TMX320C80GGP50
OBSOLETE
BGA
GGP
352
TBD
Call TI
Call TI
TMX320C80GGP60
OBSOLETE
BGA
GGP
352
TBD
Call TI
Call TI
(1)
The marketing status values are defined as follows:
ACTIVE: Product device recommended for new designs.
LIFEBUY: TI has announced that the device will be discontinued, and a lifetime-buy period is in effect.
NRND: Not recommended for new designs. Device is in production to support existing customers, but TI does not recommend using this part in
a new design.
PREVIEW: Device has been announced but is not in production. Samples may or may not be available.
OBSOLETE: TI has discontinued the production of the device.
(2)
Eco Plan - The planned eco-friendly classification: Pb-Free (RoHS) or Green (RoHS & no Sb/Br) - please check
http://www.ti.com/productcontent for the latest availability information and additional product content details.
TBD: The Pb-Free/Green conversion plan has not been defined.
Pb-Free (RoHS): TI's terms "Lead-Free" or "Pb-Free" mean semiconductor products that are compatible with the current RoHS requirements
for all 6 substances, including the requirement that lead not exceed 0.1% by weight in homogeneous materials. Where designed to be soldered
at high temperatures, TI Pb-Free products are suitable for use in specified lead-free processes.
Green (RoHS & no Sb/Br): TI defines "Green" to mean Pb-Free (RoHS compatible), and free of Bromine (Br) and Antimony (Sb) based flame
retardants (Br or Sb do not exceed 0.1% by weight in homogeneous material)
(3)
MSL, Peak Temp. -- The Moisture Sensitivity Level rating according to the JEDEC industry standard classifications, and peak solder
temperature.
Important Information and Disclaimer:The information provided on this page represents TI's knowledge and belief as of the date that it is
provided. TI bases its knowledge and belief on information provided by third parties, and makes no representation or warranty as to the
accuracy of such information. Efforts are underway to better integrate information from third parties. TI has taken and continues to take
reasonable steps to provide representative and accurate information but may not have conducted destructive testing or chemical analysis on
incoming materials and chemicals. TI and TI suppliers consider certain information to be proprietary, and thus CAS numbers and other limited
information may not be available for release.
In no event shall TI's liability arising out of such information exceed the total purchase price of the TI part(s) at issue in this document sold by TI
to Customer on an annual basis.
Addendum-Page 1
IMPORTANT NOTICE
Texas Instruments Incorporated and its subsidiaries (TI) reserve the right to make corrections, modifications,
enhancements, improvements, and other changes to its products and services at any time and to discontinue
any product or service without notice. Customers should obtain the latest relevant information before placing
orders and should verify that such information is current and complete. All products are sold subject to TI’s terms
and conditions of sale supplied at the time of order acknowledgment.
TI warrants performance of its hardware products to the specifications applicable at the time of sale in
accordance with TI’s standard warranty. Testing and other quality control techniques are used to the extent TI
deems necessary to support this warranty. Except where mandated by government requirements, testing of all
parameters of each product is not necessarily performed.
TI assumes no liability for applications assistance or customer product design. Customers are responsible for
their products and applications using TI components. To minimize the risks associated with customer products
and applications, customers should provide adequate design and operating safeguards.
TI does not warrant or represent that any license, either express or implied, is granted under any TI patent right,
copyright, mask work right, or other TI intellectual property right relating to any combination, machine, or process
in which TI products or services are used. Information published by TI regarding third-party products or services
does not constitute a license from TI to use such products or services or a warranty or endorsement thereof.
Use of such information may require a license from a third party under the patents or other intellectual property
of the third party, or a license from TI under the patents or other intellectual property of TI.
Reproduction of information in TI data books or data sheets is permissible only if reproduction is without
alteration and is accompanied by all associated warranties, conditions, limitations, and notices. Reproduction
of this information with alteration is an unfair and deceptive business practice. TI is not responsible or liable for
such altered documentation.
Resale of TI products or services with statements different from or beyond the parameters stated by TI for that
product or service voids all express and any implied warranties for the associated TI product or service and
is an unfair and deceptive business practice. TI is not responsible or liable for any such statements.
Following are URLs where you can obtain information on other Texas Instruments products and application
solutions:
Products
Applications
Amplifiers
amplifier.ti.com
Audio
www.ti.com/audio
Data Converters
dataconverter.ti.com
Automotive
www.ti.com/automotive
DSP
dsp.ti.com
Broadband
www.ti.com/broadband
Interface
interface.ti.com
Digital Control
www.ti.com/digitalcontrol
Logic
logic.ti.com
Military
www.ti.com/military
Power Mgmt
power.ti.com
Optical Networking
www.ti.com/opticalnetwork
Microcontrollers
microcontroller.ti.com
Security
www.ti.com/security
Telephony
www.ti.com/telephony
Video & Imaging
www.ti.com/video
Wireless
www.ti.com/wireless
Mailing Address:
Texas Instruments
Post Office Box 655303 Dallas, Texas 75265
Copyright  2005, Texas Instruments Incorporated