TSPC106AVGSB/T83CE [ATMEL]
PCI Bus Controller, CMOS, CBGA303, 21 X 25 MM, 3.84 MM HEIGHT, 1.27 MM PITCH, CICGA-303;型号: | TSPC106AVGSB/T83CE |
厂家: | ATMEL |
描述: | PCI Bus Controller, CMOS, CBGA303, 21 X 25 MM, 3.84 MM HEIGHT, 1.27 MM PITCH, CICGA-303 PC |
文件: | 总39页 (文件大小:504K) |
中文: | 中文翻译 | 下载: | 下载PDF数据表文档文件 |
Features
• Processor Bus Frequency Up to 66 MHz and 83.3 MHz
• 64-bit Data Bus and 32-bit Address Bus
• L2 Cache Control for 256-Kbyte, 512-Kbyte, 1-Mbyte Sizes
• Provides Support for Either Asynchronous SRAM, Burst SRAM
or Pipelined Burst SRAM
• Compliant with PCI Specification, Revision 2.1
• PCI Interface Operates at 20 to 33 MHz, 3.3V/5.0V-compatible
• IEEE 1149.1-compliant, JTAG Boundary-scan Interface
• PD Max = 1.7 Watts (66 MHz), Full Operating Conditions
• Nap, Doze and Sleep Modes Reduce Power Consumption
• Fully Compliant with MIL-STD-883 Class Q or According to Atmel Standards
• Upscreenings Based on Atmel Standards
• Full Military Temperature Range (-55°C ≤ TC ≤ +125°C)
Industrial Temperature Range (-40°C ≤ TC ≤ +110°C)
• VCC = 3.3V 5%
PCI Bridge/
Memory
Controller
• Available in a 303-ball CBGA or a 303-ball CBGA with Solder Column Interposer (SCI)
(CI-CGA) Package
TSPC106
Description
The TSPC106 provides an integrated, high-bandwidth, high-performance, TTL-com-
patible interface between a 60x processor, a secondary (L2) cache or up to a total of
four additional 60x processors, the PCI bus and main memory.
PCI support allows system designers to rapidly design systems using peripherals
already designed for PCI.
The TSPC106 uses an advanced 3.3V CMOS-process technology and maintains full
interface compatibility with TTL devices.
The TSPC106 integrates system testability and debugging features via JTAG bound-
ary-scan capability.
Note:
In this document, the term “60x” is used to denote a 32-bit microprocessor from the
PowerPC family that conforms to the bus interface of the PowerPC 601®, PowerPC
603™ or PowerPC 604™ microprocessors. This does not include the PowerPC 602™
microprocessor that has a multiplexed address/data bus. 60x processors implement
the PowerPC architecture as it is specified for 32-bit addressing, providing 32-bit effec-
tive (logical) addresses, integer data types of 8, 16 and 32 bits, and floating-point data
types of 32 and 64 bits (single-precision and double-precision).
Rev. 2102A–06/01
Figure 1. TSPC106 Block Diagram
L2 Cache
Interface
Memory
Interface
60x Processor
Interface
Power Management
Error/Interrupt
Control
Target
Master
Configuration
Registers
PCI Interface
Functional Description
The TSPC106 provides a PowerPC microprocessor
CHRP-compliant bridge between the PowerPC micropro-
cessor family and the PCI bus. CHRP is a set of
specifications that defines a unified personal computer
architecture and brings the combined advantages of the
Power Macintosh platform and the standard PC environ-
ment to both system vendors and users. PCI support
allows system designers to rapidly design systems using
peripherals already designed for PCI and other standard
interfaces available in the personal computer hardware
environment. These open specifications make it easier for
system vendors to design computers capable of running
multiple operating systems. The TSPC106 integrates sec-
ondary cache control and a high-performance memory
controller. The TSPC106 uses an advanced 3.3V CMOS
process technology and is fully compatible with TTL
devices.
is 64 bits wide. The 60x processor interface of the
TSPC106 uses a subset of the 60x bus protocol, support-
ing single-beat and burst data transfers. The address and
data buses are decoupled to support pipelined
transactions.
The TSPC106 provides support for the following configura-
tions of 60x processors and L2 cache:
• Up to four 60x processors with no L2 cache
• A single 60x processor plus a direct-mapped, lookaside
L2 cache using the internal L2 cache controller of the
TSPC106
• Up to four 60x processors plus an externally controlled
L2 cache (e.g., the Motorola MPC2604GA integrated L2
lookaside cache)
The memory interface controls processor and PCI interac-
tions to main memory and is capable of supporting a
variety of configurations using DRAM, EDO, or SDRAM
and ROM or Flash ROM.
The TSPC106 supports a programmable interface to a vari-
ety of PowerPC microprocessors operating at select bus
speeds. The 60x address bus is 32 bits wide; the data bus
TSPC106
2
TSPC106
The PCI interface of the TSPC106 complies with the PCI
local bus specification Revision 2.1 and follows the guide-
lines in the PCI System Design Guide Revision 1.0 for host
bridge architecture. The PCI interface connects the proces-
sor and memory buses to the PCI bus to which I/O
components are connected. The PCI bus uses a 32-bit
multiplexed address/data bus plus various control and error
signals.
and write operations to the PCI memory space, the PCI I/O
space and the PCI configuration space. The TSPC106 also
supports PCI special-cycle and interrupt-acknowledge
commands. As a target, the TSPC106 supports read and
write operations to system memory.
The TSPC106 provides hardware support for four levels of
power reduction: doze, nap, sleep and suspend. The
design of the TSPC106 is fully static, allowing internal logic
states to be preserved during all power saving modes.
The PCI interface of the TSPC106 functions as both a mas-
ter and target device. As a master, the 106 supports read
3
Pin Description
Figure 2. TSPC106 in 303-ball CBGA Package
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
W
V
U
T
DL26
DL28
DL30
DH31
DH29
DH27
DH25
DH23
DH21
DH19
DH17
DH15
DH13
DH11
DH9
DH7
DL24
DL27
DL23
WE
DL29
DL25
DH0
DH2
DH4
DL31
DL14
DL15
DH1
DH30
PLL2
PLL3
DL16
VSS
DH28
PLL0
PLL1
VSS
DH26
DL12
DL13
VDD
VSS
DH24
DL10
DL11
DL9
DH20
DL4
DL3
DL5
DL6
DH18
DL2
DH16
DL0
DH14
DH12
DBG1
HIT
DH10
DH6
DH8
DL21
DL19
A0
DL22
DL20
MA1/
SDBA0/
AR9
DOE/
DBGL2
MA2/
SDMA2/
AR10
TV/
BR2
BA0/
BR3
DIRTY_IN/
BR1
DCS/
BG3
DL1
MA3/
SDMA3/
AR11
ADS/
DALE/
BRL2
TWE/
BG2
DIRTY_OUT/
BG1
R
P
RCS0
VSS
VDD
VDD
VSS
TS
MA5/
SDMA5/
AR13
MA4/
SDMA4/
AR12
BA1/
BAA
BGL2
DWE0/
DBG2
XATS/
SDMA1
DH3
VDD
DL8
VDD
VSS
A1
MA0/
SDBA1/
SDMA0/
AR0
MA6/
SDMA6/
AR14
N
DL17
DH5
VDD
VSS
VDD
VDD
VSS
DL7
DH22
VSS
VDD
LBCLAIM
CI
A2
TA
MA8/
SDMA8/
AR16
MA7/
SDMA7/
AR15
RAS0/
CS0
DL18
VSS
VDD
NC
VSS
AVDD
VSS
NC
VDD
VSS
VDD
VSS
VSS
SYSCLK
A9
VDD
DBG0
A8
WT
TBST
A7
GBL
BR0
BG0
A12
A3
A4
TT4
TT3
TT2
TEA
M
L
MA9/
SDMA9/
AR17
RAS1/
CS1
CKO/
DWE2
RAS5/
CS5
HRST
MA11/
SDMA11/ SDMA10/
AR19
QACK
VDD
VSS
VDD
MA10/
RAS3/
CS3
RAS2/
CS2
RAS4/
CS4
RAS7/
CS7
K
J
VDD
A5
AR18
MA12/
SDMA12/
AR20
CAS0/
DQM0
RAS6/
CS6
DBGLB/
CKE
PPEN
RCS1
TRST
MCP
A11
A6
A13
A10
CAS1/
DQM1
SUS-
PEND
DWE1/
DBG3
PIRQ/
SDRAS
H
G
F
QREQ
VSS
VDD
VSS
TDO
NC
NC
VDD
VSS
VDD
VSS
VSS
VDD
VSS
VDD
VDD
VSS
VDD
A31
A15
TSIZ1
A21
A14
TSIZ0
TSIZ2
A22
A16
A17
TT1
TT0
A18
A19
CAS2/
DQM2
CAS4/
DQM4
CAS5/
DQM5
LSSD_
MODE
RTC
BCTL1
NMI
VDD
VSS
VDD
PAR
LOCK
CAS6/
DQM6
DEV-
SEL
BCTL0
TCK
VDD
VSS
PERR
SERR
ARTRY
A20
CAS3/
DQM3
CAS7/
DQM7
MDLE/
SDCAS
E
IRDY
A29
PAR0/
AR1
PAR1/
AR2
TMS
FOE
AD30
TDI
AD28
AD26
AD7
AD24
AD23
AD11
AD13
AD21
AD19
AD17
C/BE2
TRDY
STOP
AD14
C/BE1
AD18
AD16
AD10
AD12
AD22
AD20
C/BE0
AD8
AD4
AD6
AD0
AD2
A30
A27
AACK
A25
A23
A24
A26
D
C
B
A
PAR2/
AR3
PAR3/
AR4
PAR5/
AR6
PAR4/
AR5
PAR7/
AR8
ISA_MASTER/
BERR
AD1
AD3
AD15
AD25
C/BE3
AD29
AD27
REQ
AD31
A28
PAR6/
AR7
GNT
AD5
AD9
FRAME
FLSHREQ
MEMACK
TSPC106
4
TSPC106
Pinout
Table 1. TSPC106 Pinout in 303-ball CBGA Package
Signal Name
Pin Number
Active
I/O
60x Processor Interface Signals
A[0:31]
R2, P2, N2, M2, L2, K2, J5, K4, K5, K6, J2, J6, J3, J4, H3, H4, H2,
G2, F1, E1, E2, F4, E3, D1, C1, C2, B1, C3, B2, E4, D3, E5
High
I/O
AACK
D2
F2
K3
R4
R5
T1
L3
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
High
I/O
ARTRY
I/O
BG0
Output
Output
Output
Output
Input
BG1 (DIRTY_OUT)
BG2 (TWE)
BG3 (DCS)
BR0
BR1 (DIRTY_IN)
BR2 (TV)
T3
T6
T5
N3
L5
Input
Input
BR3 (BA0)
CI
Input
I/O
DBG0
Output
Output
Output
Output
Output
I/O
DBG1 (TOE)
DBG2 (DWE0)
DBG3 (DWE1)
DBGLB (CKE)
DH[0:31]
U4
P3
H11
J10
T14, R13, R14, P13, P14, N13, U3, W1, V2, W2, V3, W3, V4, W4,
V5, W5, V6, W6, V7, W7, V8, W8, N8, W9, V9, W10, V10, W11,
V11, W12, V12, W13
DL[0:31]
U6, T7, U7, T8, U8, R8, P8, N9, P9, R9, U9, T9, U10, T10, U13,
T13, R12, N14, M13, T2, U1, U2, V1, U15, V16, U14, W16, V15,
W15, V14, W14, V13
High
I/O
GBL
M3
Low
Low
Low
Low
Low
Low
Low
High
High
Low
I/O
Input
Output
I/O
LBCLAIM
MCP
TA
N4
J11
N1
TBST
TEA
L4
I/O
J1
Output
I/O
TS
R1
TSIZ[0:2]
TT[0:4]
WT
G3, G4, F3
G1, H1, K1, L1, M1
M4
I/O
I/O
I/O
5
Table 1. TSPC106 Pinout in 303-ball CBGA Package
Signal Name
Pin Number
Active
I/O
XATS (SDMA1)
P1
Low
Input
L2 Cache Interface Signals
ADS/DALE/BRL2
BA0 (BR3)
R3
T5
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
High
Low
Output
Output
Output
Output
Output
Input
BA1/BAA/BGL2
DBGL2/DOE
DCS (BG3)
P4
U5
T1
DIRTY_IN (BR1)
DIRTY_OUT (BG1)
DWE0 (DBG2)
DWE1 (DBG3)
DWE2 (CKO)
HIT
T3
R4
P3
H11
L11
T4
Output
Output
Output
Output
Input
TOE (DBG1)
TV (BR2)
U4
T6
Output
I/O
TWE (BG2)
R5
Output
Memory Interface Signals
BCTL [0:1]
F16, F15
B3
Low
Low
Low
High
Low
High
High
High
High
Low
High
Low
Low
Low
Low
High
Low
Low
Output
Input
BERR (ISA_MASTER)
CAS/DQM[0:7]
CKE/DBGLB
FOE
J15, H15, G16, E16, G14, G13, F14, E14
Output
Output
Output
Output
Output
Output
Output
Output
I/O
J10
D13
MA0/SDBA1/SDMA0/AR0
SDMA1 (XATS)
MA1/SDBA0/AR9
MA[2:12]/SDMA[2:12]/AR[10:20]
MDLE/SDCAS
PAR[0:7]/AR[1:8]
PPEN
N15
P1
U16
T16, R16, P15, P16, N16, M15, M16, L15, K15, K16, J16
E13
D16, D15, C16, C15, B16, C14, A16, B15
J14
Output
Output
I/O
RAS/CS[0:7]
RCS0
M14, L13, K13, K14, K12, L10, J12, K11
R15
J13
RCS1
Output
Input
RTC
G15
H10
T15
SDRAS (PIRQ)
WE
Output
Output
TSPC106
6
TSPC106
Table 1. TSPC106 Pinout in 303-ball CBGA Package
Signal Name
Pin Number
Active
I/O
PCI Interface Signals(2)
AD[31:0](2)
A4, C13, B5, D12, A5, C12, B6, D11, C11, B7, D10, A7, C10, B8,
D9, A8, B10, D8, A11, C7, B11, D7, A12, C6, B12, C5, A13, D5,
A14, C4, B14, D4
High
I/O
C/BE[3:0](2)
DEVSEL
FLSHREQ
FRAME
GNT
A6, C9, C8, D6
Low
Low
Low
Low
Low
Low
Low
Low
Low
High
Low
Low
Low
Low
Low
Low
I/O
I/O
F8
A3
Input
I/O
A10
A15
Input
I/O
IRDY
E8
ISA_MASTER (BERR)
LOCK
B3
Input
Input
Output
I/O
G8
MEMACK
PAR
A2
G9
PERR
F9
I/O
PIRQ (SDRAS)
REQ
H10
Output
Output
I/O
B4
SERR
E9
STOP
A9
I/O
TRDY
B9
I/O
Interrupt, Clock and Power Management Signals
CK0 (DWE2)
HRST
L11
High
Low
Output
Input
Input
Output
Input
Input
L16
NMI
E15
High
Low
QACK
L14
QREQ
H16
Low
SYSCLK
SUSPEND
L6
Clock
Low
H14
Input
Test/Configuration Signals
PLL[0:3]
TCK
U11, T11, U12, T12
High
Clock
High
High
High
Low
Input
Input
Input
Output
Input
Input
F13
B13
E12
D14
H13
TDI
TDO
TMS
TRST
7
Table 1. TSPC106 Pinout in 303-ball CBGA Package
Signal Name
Pin Number
Active
I/O
Power and Ground Signals
AVDD
K9
High
Low
High
Clock
Input
LSSD_MODE(3)
G11
VDD
E10, E6, F11, F5, F7, G10, G12, G6, H5, H7, K10, K7, L12, M11,
M5, M7, N10, N12, N6, P11, P5, P7, R10, R6, J8, L8
Power
VSS
NC
E11, E7, F10, F12, F6, G5, G7, H12, H6, J7, L7, M10, M12, M6,
N11, N5, N7, P10, P12, P6, R11, R7, K8, J9, L9
Low
Ground
H8, H9, M8, M9
–
–
Notes: 1. Some signals have dual functions and are shown more than once in this table.
2. All PCI signals are in little-endian bit order.
3. This test signal is for factory use only. It must be pulled up to VDD for normal device operation.
TSPC106
8
TSPC106
Note:
Note:
A bar over a signal name indicates that the signal is
active low, for example, address retry (ARTRY) and
transfer start (TS). Active-low signals are referred to as
asserted (active) when they are low and negated when
they are high. Signals that are not active low such as tag
valid (TV) and nonmaskable interrupt (NMI) are referred
to as asserted when they are high and negated when
they are low.
Signal Description
The signals on the TSPC106 are grouped as follows:
• 60x processor interface signals
• L2 cache/multiple processor interface signals
• Memory interface signals
• PCI interface signals
For multiple-function signals, outlined signal names refer
to the alternate function(s) of the signal being described.
For example, the L2 controller signal, (tag output enable
TOE), has the alternate function data bus grant 1
(DBG1) when the TSPC106 is configured for a second
60x processor.
• Interrupt, clock and power management signals
• IEEE 1149.1 interface signals
• Configuration signals
Figure 3. Symbol
AD[31:0]
32
BR0
1
C/BE[3:0]
BG0
1
4
PAR
1
TS
1
XATS
TRDY
1
1
A[0:31]
32
IRDY
1
FRAME
TT[0:4]
5
1
TSIZ[0:2]
3
STOP
1
LOCK
TBST
1
3
60x
Processor
Interface
1
PCI
Interface
DEVSEL
WT, CI, GBL
AACK
1
SERR, PERR
1
2
REQ
1
ARTRY
DBG0(1)
1
GNT
1
1
DH[0:31], DL[0:31]
FLSHREQ
1
64
1
MEMACK
TA
1
ISA_MASTER
1
LBCLAIM
DBGLB
1
PIRQ
1
1
MCP, TEA
2
RAS[0:7]
8
ADS/DALE/BRL2
BAA/BA1/BGL2
CAS[0:7]
8
1
1
1
3
1
1
1
1
1
1
1
1
WE
1
MA0/AR0
DOE/DBGL2
DWE[0:2]/DBG2, DBG3, CKO(1)
1
MA[1:12]/AR[9:20]
12
Memory
Interface
L2 Cache/
Multiple
PAR[0:7]/AR[1:8]
HIT
DCS/BG3
8
FOE(1)
1
Processor
Interface
MDLE, PPEN
BA0/BR3
2
TWE/BG2
RCS1
1
RCS0(1)
TV/BR2
1
BCTL0, BCTL1
DIRTY_IN/BR1
DIRTY_OUT/BG1
2
RTC
1
TOE/DBG1
SYSCLK, CKO(1)
2
TDO
TDI
HRST
1
1
1
1
1
1
Interrupt, Clock
and Power
Management
Signals
NMI
1
1
1
1
IEEE 1149.1
JTAG
Interface
TCK
TMS
QREQ
QACK
SUSPEND
TRST
DBG0
FOE(1)
RCS0(1)
1
1
1
4
Configuration
PLL[0:3]
Note:
1. Some signals have dual functions and are shown more than once in this figure.
9
60x Processor Interface Signals
Table 2. 60x Processor Interface Signals
Number of
Signal
Signal Name
Pins
I/O
O
I
Signal Description
A[0:31]
Address bus
32
Specifies the physical address for 60x bus snooping.
Specifies the physical address of the bus transaction. For burst reads,
the address is aligned to the critical double-word address that missed
in the instruction or data cache. For burst writes, the address is aligned
to the double-word address of the cache line being pushed from the
data cache.
AACK
Address
acknowledge
1
O
I
Indicates that the address tenure of a transaction is terminated. On the
cycle following the assertion of AACK, the bus master releases the
address-tenure-related signals to a high impedance state and samples
ARTRY.
Indicates that an externally-controlled L2 cache is terminating the
address tenure. On the cycle following the assertion of AACK, the bus
master releases the address-tenure-related signals to a high
impedance state and samples ARTRY.
ARTRY
Address retry
1
1
O
I
Indicates that the initiating 60x bus master must retry the current
address tenure.
During a snoop operation, indicates that the 60x either requires the
current address tenure to be retried due to a pipeline collision or needs
to perform a snoop copy-back operation. During normal 60x bus cycles
in a multiprocessor system, indicates that the other 60x or external L2
controller requires the address tenure to be retried.
BG0
Bus grant 0
O
Indicates that the primary 60x may, with the proper qualification, begin
a bus transaction and assume mastership of the address bus.
BR0
CI
Bus request 0
Cache inhibit
1
1
1
I
Indicates that the primary 60x requires the bus for a transaction.
Indicates that an access is caching-inhibited.
I/O
O
DBG0
Data bus grant 0
Indicates that the 60x may, with the proper qualification, assume
mastership of the data bus.
DBGLB
Local bus slave
data bus grant
1
O
Indicates that the 60x processor is prepared to accept data and the
local bus slave should drive the data bus.
TSPC106
10
TSPC106
Table 2. 60x Processor Interface Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
DH[0:31],
DL[0:31]
Data bus
64
The data bus is comprised of two halves - data bus high (DH[0:31])
and data bus low (DL[0:31]). The data bus has the following byte lane
assignments:
Data Byte Byte Lane
DH[0:7]
DH[8:15]
DH[16:23]
DH[24:31]
DL[0:7]
0
1
2
3
4
5
6
7
DL[8:15]
DL[16:23]
DL[24:31]
O
I
Represents the value of data being driven by the TSPC106.
Represents the state of data being driven by a 60x processor, the local
bus slave, the L2 cache or the memory subsystem.
GBL
Global
1
1
1
1
I/O
I
Indicates that an access is global and hardware needs to enforce
coherency.
LBCLAIM
MCP
TA
Local bus slave
cycle claim
Indicates that the local bus slave claims the transaction and is
responsible for driving TA during the data tenure.
Machine check
O
O
Indicates that the TSPC106 detected a catastrophic error and the 60x
processor should initiate a machine check exception.
Transfer
acknowledge
Indicates that the data has been latched for a write operation or that
the data is valid for a read operation, thus terminating the current data
beat. If it is the last (or only) data beat, this also terminates the data
tenure.
I
Indicates that the external L2 cache or local bus slave has latched data
for a write operation or is indicating the data is valid for a read
operation. If it is the last (or only) data beat, then the data tenure is
terminated.
TBST
TEA
Transfer burst
1
1
O
I
Indicates that a burst transfer is in progress.
Indicates that a burst transfer is in progress.
Transfer error
acknowledge
O
Indicates that a bus error has occurred. Assertion of TEA terminates
the transaction in progress. An unsupported memory transaction, such
as a direct-store access or a graphics read or write, causes the
assertion of TEA (provided TEA is enabled).
TS
Transfer start
1
O
I
Indicates that the TSPC106 has started a bus transaction and that the
address and transfer attribute signals are valid. Note that the
TSPC106 only initiates a transaction to broadcast the address of a PCI
access to memory for snooping purposes.
Indicates that a 60x bus master has begun a transaction and that the
address and transfer attribute signals are valid.
11
Table 2. 60x Processor Interface Signals
Number of
Signal
Signal Name
Pins
I/O
O
I
Signal Description
TSIZ[0:2]
Transfer size
3
Specifies the data transfer size for the 60x bus transaction.
Specifies the data transfer size for the 60x bus transaction.
Specifies the type of 60x bus transfer in progress.
Specifies the type of 60x bus transfer in progress.
Indicates that an access is write-through.
TT[0:4]
Transfer type
Write-through
5
O
I
WT
1
1
I/O
I
XATS
Extended address
transfer start
Indicates that the 60x has started a direct-store access (using the
extended transfer protocol). Since direct-store accesses are not
supported by the TSPC106, the TSPC106 automatically asserts when
TEA and XATS are asserted (provided TEA is enabled).
L2 Cache/Multiple Processor Interface Signals
The TSPC106 provides support for either an internal L2 cache controller or an external L2 cache controller and/or addi-
tional 60x processors.
Internal L2 Controller Signals
Table 3 lists the interface signals for the internal L2 controller and provides a brief description of their functions. The internal
L2 controller supports either burst SRAMs or asynchronous SRAMs. Some of the signals perform different functions
depending on the SRAM configuration.
Table 3. Internal L2 Controller Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
ADS
DALE
BRL2
Address strobe
1
O
For a burst SRAM configuration, indicates to the burst SRAM that the
address is valid to be latched.
BA0
BR3
Burst address 0
Burst address 1
1
1
I/O
O
For an asynchronous SRAM configuration, indicates bit 0 of the burst
address counter.
BA1
BAA
BGL2
For an asynchronous SRAM configuration, indicates bit 1 of the burst
address counter.
BAA
BA1
BGL2
Bus address
advance
1
1
O
O
For a burst SRAM configuration, indicates that the burst RAMs should
increment their internal addresses.
DALE
ADS
Data address latch
enable
For an asynchronous SRAM configuration, indicates that the external
address latch should latch the current 60x bus address.
BRL2
DCS
BG3
Data RAM chip
select
1
1
1
1
O
I
Enables the L2 data RAMs for a read or write operation.
DIRTY_IN
BR1
Dirty in
Indicates that the selected L2 cache line is modified. The polarity of
DIRTY_IN is programmable.
DIRTY_OUT Dirty out
BG1
O
O
Indicates that the L2 cache line should be marked as modified. The
polarity of DIRTY_OUT is programmable.
DOE
DBGL2
Data RAM output
enable
Indicates that the L2 data RAMs should drive the data bus.
TSPC106
12
TSPC106
Table 3. Internal L2 Controller Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
DWE[0:2]
DBG2
Data RAM write
enable
3
O
Indicates that a write to the L2 data RAMs is in progress. Multiple pins
are provided to reduce loading.
DBG3
HIT
Hit
1
1
1
1
I
Indicates that the L2 cache has detected a hit. The polarity of HIT is
programmable.
TOE
DBG1
Tag output enable
Tag valid
O
Indicates that the tag RAM should drive the L2 tag address onto the
address bus.
TV
BR2
I/O
O
Indicates that the current L2 cache line should be marked valid. The
polarity of TV is programmable.
TWE
BG2
Tag write enable
Indicates that the L2 tag address, valid, and dirty bits should be
updated.
External L2 Controller Signals
When an external L2 cache controller is used instead of the internal L2 cache controller, four signals change their
functions.
Table 4. External L2 Controller Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
BGL2
BA1
External L2 bus
grant
1
O
Indicates that the external L2 controller has been granted mastership
of the 60x address bus.
BAA
BRL2
ADS
External L2 bus
request
1
I
Indicates that the external L2 controller requires mastership of the 60x
bus for a transaction.
DALE
DBGL2
DOE
External L2 data
bus grant
1
1
O
I
Indicates that the external L2 controller has been granted mastership
of the 60x data bus.
HIT
External L2 hit
Indicates that the current transaction is claimed by the external L2
controller. The external L2 controller will assert AACK and TA for the
transaction.
13
Multiple Processor Signals
When a system implementation uses more than one 60x processor, nine of the internal L2 cache controller signals change
their functions.
Note that in a multi-processor system, with the exception of the bus grant, bus request and data bus grant signals, all of the
60x processor interface signals are shared by all 60x processors.
Table 5. Multiple Processor Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
BG1
DIRTY_OUT
Bus grant 1
1
O
Indicates that processor 1 may, with the proper qualification, begin a
60x bus transaction and assume mastership of the address bus.
BG2
TWE
Bus grant 2
1
1
1
1
1
1
1
1
O
O
I
Indicates that processor 2 may, with the proper qualification, begin a
60x bus transaction and assume mastership of the address bus.
BG3
DCS
Bus grant 3
Indicates that processor 3 may, with the proper qualification, begin a
60x bus transaction and assume mastership of the address bus.
BR1
DIRTY_IN
Bus request 1
Bus request 2
Bus request 3
Data bus grant 1
Data bus grant 2
Data bus grant 3
Indicates that processor 1 requires mastership of the 60x bus for a
transaction.
BR2
TV
I
Indicates that processor 2 requires mastership of the 60x bus for a
transaction.
BR3
BA0
I
Indicates that processor 3 requires mastership of the 60x bus for a
transaction.
DBG1
TOE
O
O
O
Indicates that processor 1 may, with the proper qualification, assume
mastership of the 60x data bus.
DBG2
DWE0
Indicates that processor 2 may, with the proper qualification, assume
mastership of the 60x data bus.
DBG3
DWE1
Indicates that processor 3 may, with the proper qualification, assume
mastership of the 60x data bus.
TSPC106
14
TSPC106
Memory Interface Signals
Table 6 lists the memory interface signals and provides a brief description of their functions. The memory interface sup-
ports either standard DRAMs or EDO DRAMs, and either standard ROMs or Flash ROMs. Some of the memory interface
signals perform different functions depending on the RAM and ROM configurations.
Table 6. Memory Interface Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
AR0
MA0
ROM address 0
8
O
Represents address bit 0 of the 8-bit ROM/Flash. Note that AR0 is only
supported for ROM bank 0 when configured for an 8-bit ROM/Flash
data bus width. The extra address bit allows for up to 2 Mbytes of ROM
when using the 8-bit wide data path. Bits 1 - 8 of the ROM address are
provided by AR[1:8] and bits 9 - 20 of the ROM address are provided
by AR[9:20].
AR[1:8]
PAR[0:7]
ROM address 1 - 8
8
O
O
Represents bits 1 - 8 of the ROM/Flash address. The other ROM
address bits are provided by AR0 and AR[9:20].
AR[9:20]
MA[1:12[
ROM address
9 - 20
12
Represents bits 9 - 20 of the ROM/Flash address (the 12 lowest order
bits, with AR20 as the least significant bit (lsb)). Bits 0 - 8 of the ROM
address are provided by AR0 and AR[1:8].
BCTL[0:1]
Buffer control 0 - 1
2
8
O
O
Used to control external data bus buffers (directional control and high-
impedance state) between the 60x bus and memory. Note that external
data buffers may be optional for lightly loaded data buses, but buffers
are required whenever an L2 cache and ROM/Flash (on the
60x/memory bus) are both in the system or when ECC is used.
CAS[0:7]
FOE
Column address
strobe 0 - 7
Indicates a memory column address is valid and selects one of the
columns in the row. CAS0 connects to the most significant byte select.
CAS7 connects to the least significant byte select.
Flash output
enable
1
O
O
Enables Flash output for the current read access.
MA0
MA[1:12]
AR0
Memory address
0 - 12
13
Represents the row/column multiplexed physical address for DRAMs or
EDOs (MA0 is the most significant address bit; MA12 is the least
significant address bit). Note that MA0 also functions as the ROM
address signal AR0 and MA[1:12] function as the ROM address
signals AR[9:20].
AR[9:20]
MDLE
Memory data latch
enable
1
8
O
O
I
Enables the external, latched data buffer for read operations, if such a
buffer is used in the system.
PAR[0:7]
AR[1:8]
Data parity/ECC
Represents the byte parity or ECC being written to memory (PAR0 is
the most significant bit).
Represents the byte parity or ECC being read from memory (PAR0 is
the most significant bit).
PPEN
Parity path read
enable
1
8
1
O
O
O
Enables external parity/ECC bus buffer or latch for memory read
operations.
RAS[0:7]
RCS0
Row address
strobe 0 - 7
Indicates a memory row address is valid and selects one of the rows in
the bank.
ROM/Flash bank 0
select
Selects ROM/Flash bank 0 for a read access or Flash bank 0 for a read
or write access.
15
Table 6. Memory Interface Signals
Number of
Pins
Signal
Signal Name
I/O
Signal Description
RCS1
ROM/Flash bank 1
select
1
1
1
O
Selects ROM/Flash bank 1 for a read access or Flash bank 1 for a read
or write access.
RTC
WE
Real-time clock
I
External clock source for the memory refresh logic when the TSPC106
is in the suspend power-saving mode.
Write enable
O
Enables writing to DRAM, EDO or Flash.
PCI Interface Signals
Table 7 lists the PCI interface signals and provides a brief description of their functions. Note that the bits in Table 7 are ref-
erenced in little-endian format.
The PCI specification defines a sideband signal as any signal, not part of the PCI specification, that connects two or more
PCI-compliant agent, and has meaning only to those agents. The TSPC106 implements four PCI sideband signals -
FLSHREQ, ISA_MASTER, MEMACK and PIRQ.
Table 7. PCI Interface Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
AD[31:0]
Address/data
32
I/O
Represents the physical address during the address phase of a
transaction. During the data phase(s) of a PCI transaction, AD[31:0]
contain data. AD[7:0] define the least significant byte and AD[31:24]
the most significant byte.
C/BE[3:0]
DEVSEL
Command/byte
enable
4
1
O
I
During the address phase, C/BE[3:0] define the PCI bus command.
During the data phase, C/BE[3:0] are used as byte enables. Byte
enables determine which byte lanes carry meaningful data. C/BE0
applies to the least significant byte.
During the address phase, C/BE[3:0] indicates the PCI bus command
that another master is sending. During the data phase C/BE[3:0]
indicate which byte lanes are valid.
Device select
O
I
Indicates that the TSPC106 has decoded the address and is the target
of the current access.
Indicates that some PCI agent (other than the TSPC106) has decoded
its address as the target of the current access.
FLSHREQ
FRAME
Flush request
Frame
1
1
I
Indicates that a device needs to have the TSPC106 flush all of its
current operations.
O
Indicates that the TSPC106, acting as a PCI master, is initiating a bus
transaction.
I
I
Indicates that another PCI master is initiating a bus transaction.
GNT
PCI bus grant
1
Indicates that the TSPC106 has been granted control of the PCI bus.
Note that GNT is a point-to-point signal. Every master has its own GNT
SIGNAL.
TSPC106
16
TSPC106
Table 7. PCI Interface Signals
Number of
Pins
Signal
Signal Name
I/O
Signal Description
IRDY
Initializer ready
1
O
Indicates that the TSPC106, acting as a PCI master, can complete the
current data phase of a PCI transaction. During a write, the TSPC106
asserts IRDY to indicate that valid data is present on AD[31:0]. During
a read, the TSPC106 asserts IRDY to indicate that it is prepared to
accept data.
I
I
Indicates another PCI master is able to complete the current data
phase of the transaction.
ISA_
MASTER
ISA master
Lock
1
1
1
Indicates that an ISA master is requesting system memory.
LOCK
I
Indicates that a master is requesting exclusive access to memory,
which may require multiple transactions to complete.
MEMACK
Memory
acknowledge
O
Indicates that the TSPC106 has flushed all of its current operations
and has blocked all 60x transfers except snoop copy-back operations.
The TSPC106 asserts MEMACK in response to assertion of
FLSHREQ after the flush is complete.
PAR
Parity
1
O
I
Asserted indicates odd parity across the AD[31:0] and C/BE[3:0]
signals during address and data phases. Negated indicates even
parity.
Asserted indicates odd parity driven by another PCI master or the PCI
target during read data phases. Negated indicates even parity.
PERR
PIRQ
Parity error
1
1
O
I
Indicates that another PCI agent detected a data parity error.
Indicates that another PCI agent detected a data parity error.
Modified memory
interrupt request
O
In emulation mode (see “Address Maps” on page 19), indicates that a
PCI write has occurred to system memory that has not been recorded
by software.
REQ
PCI bus request
System error
1
1
O
O
Indicates that the TSPC106 is requesting control of the PCI bus to
perform a transaction. Note that REQ is a point-to-point signal. Every
master has its own REQ signal.
SERR
Indicates that an address parity error or some other system error
(where the result will be a catastrophic error) was detected.
I
Indicates that another target has detected a catastrophic error.
STOP
TRDY
Stop
1
1
O
Indicates that the TSPC106, acting as the PCI target, is requesting that
the PCI bus master stop the current transaction.
I
Indicates that some other PCI agent is requesting that the PCI initiator
stop the current transaction.
Target ready
O
Indicates that the TSPC106, acting as a PCI target, can complete the
current data phase of a PCI transaction. During a read, the TSPC106
asserts TRDY to indicate that valid data is present on AD[31:0]. During
a write, the TSPC106 asserts TRDY to indicate that it is prepared to
accept data.
I
Indicates that another PCI target is able to complete the current data
phase of a transaction.
17
Interrupt, Clock and Power Management Signals
The TSPC106 coordinates interrupt, clocking, and power management signals across the memory bus, the PCI bus and
the 60x processor bus.
Table 8. Interrupt, Clock and Power Management Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
CKO
DWE2
Test clock
1
1
1
1
1
O
CKO provides a means to monitor the internal PLL output or the bus
clock frequency. The CKO clock should be used for testing purposes
only. It is not intended as a reference clock signal.
HRST
NMI
Hard reset
I
Initiates a complete hard reset of the TSPC106. During assertion, all
bidirectional signals are released to a high-impedance state and all
output signals are either in a high impedance or inactive state.
Nonmaskable
interrupt
I
Indicates that an external device (typically an interrupt controller) has
detected a catastrophic error. In response, the TSPC106 asserts MCP
on the 60x processor bus.
QACK
QREQ
Quiesce
acknowledge
O
I
Indicates that the TSPC106 is in a low-power state. All bus activity that
requires snooping has terminated and the 60x processor may enter a
low-power state.
Quiesce request
Indicates that a 60x processor is requesting that all bus activity
involving snoop operations pause or terminate so that the 60x
processor may enter a low-power state.
SUSPEND
SYSCLK
Suspend
1
1
I
I
Activates the suspend power-saving mode.
System clock
SYSCLK sets the frequency of operation for the PCI bus and provides
a reference clock for the phase-locked loop (PLL) in the TSPC106.
SYSCLK is used to synchronize bus operations. Refer to section
“Clocking” on page 19 for more information.
IEEE 1149.1 Interface Signals
To facilitate system testing, the TSPC106 provides a JTAG test access port that complies with the IEEE 1149.1 boundary-
scan specification.
Table 9. IEEE 1149.1 Interface Signals
Number of
Signal
Signal Name
Pins
I/O
Signal Description
TCK
JTAG test clock
1
I
Input signals to the test access port (TAP) are clocked in on the rising
edge of TCK. Changes to the TAP output signals occur on the falling
edge of TCK. The test logic allows TCK to be stopped.
TDO
JTAG test data
output
1
O
The contents of the selected internal instructions or data register are
shifted out onto this signal on the falling edge of TCK. TDO will remain
in a high-impedance state except when scanning of data is in progress.
TDI
JTAG test data
input
1
1
1
I
I
I
The value presented on this signal on the rising edge of TCK is clocked
into the selected JTAG test instruction or data register.
TMS
TRST
JTAG test mode
select
This signal is decoded by the internal JTAG TAP controller to
distinguish the primary operation of the test support circuitry.
JTAG test reset
This input causes asynchronous initialization of the internal JTAG TAP
controller.
TSPC106
18
TSPC106
Configuration Signals
The configuration signals select the ROM/Flash options, the clock mode of the TSPC106 and how the TSPC106 responds
to addresses on the 60x and PCI buses.
Table 10. Configuration Signals
Number of
Signal
Pins
I/O
Configuration
DBG0(1)
1
I
High configures the TSPC106 for address map A.
Low configures the TSPC106 for address map B.
FOE(1)
1
I
High configures ROM bank 0 for an 8-bit data bus width.
Low configures ROM bank 0 for an 64-bit data bus width.
Note that the data bus width for ROM bank 1 is always 64 bits.
PLL[0:3](2)
RCS0(1)
4
1
I
I
High/Low – configuration for the PLL clock mode.
High indicates ROM is located on the 60x processor/memory data bus.
Low indicates ROM is located on the PCI bus.
Note:
1. The TSPC106 samples these signals during a power-on reset or hard reset operation to determine the configuration. Weak
pull-up or pull-down resistors should be used to avoid interference with the normal signal operations.
2. The TSPC106 continuously samples the phase-locked loop (PLL) configuration signals to allow the switching of clock modes
or the bypass of the PLL without a hard reset operation.
Clocking
Address Maps
The TSPC106 requires a single system clock input,
SYSCLK. The frequency of SYSCLK dictates the operating
frequency of the PCI bus. An internal PLL on the TSPC106
generates a master clock that is used for all of the internal
(core) logic. The master clock provides the core frequency
reference and is phase-locked to the SYSCLK input. The
60x processor, L2 cache and memory interfaces operate at
the core frequency.
The TSPC106 supports three address mapping configura-
tions designated address map A, address map B, and
emulation mode address map. Address map A conforms to
the “PowerPC Reference Platform Specification”. Address
map B conforms to the “PowerPC Common Hardware Ref-
erence Platform Architecture (CHRP)”. The emulation
mode address map is provided to support software emula-
tion fx86 hardware. The configuration signal DBG0,
sampled during power-on reset, selects between address
map A and address map B. After reset, the address map
can be changed by a programmable parameter. The emu-
lation mode address map can only be selected by software
after reset.
The PLL[0:3] signals configure the core-to-SYSCLK fre-
quency ratio. The TSPC106 supports core-to-SYSCLK
frequency ratios of 1:1, 2:1 and 3:1, although not all ratios
are supported for all frequencies. The TSPC106 supports
changing the clock mode and bypassing the PLL without
requiring a hard reset operation, provided the system
design allows sufficient time for the PLL to relock.
19
Detailed Specifications
Design and Construction
Scope
Terminal Connections
This drawing describes the specific requirements of the
TSPC106 in compliance with MIL-STD-883 class B or man-
ufacturer’s standard screening.
The terminal connections are as shown in “Pin Description”
on page 4.
Lead Material and Finish
Applicable Documents
Documents applicable to the information contained in this
datasheet are:
Lead material and finish are as specified in “Package
Mechanical Data” on page 36.
Package
1. MIL-STD-883: Test methods and procedures for
electronics
The macrocircuits are packaged in 303-pin ceramic ball
grid array packages.
2. MIL-PRF-38535: General specifications for
microcircuits
The precise package drawings are described at the end of
the specification. “CBGA Package Parameters” on page 36
and “CI_CGA Package Parameters” on page 37.
Requirements
General
The microcircuits are in accordance with the applicable
documents and as specified herein.
Absolute Maximum Ratings
Stresses above the absolute maximum rating may cause
permanent damage to the device. Extended operation at
the maximum levels may degrade performance and affect
reliability.
Table 11. Absolute Maximum Ratings
Symbol
VDD
Parameter
Min
-0.3
-0.3
-0.3
-55
Max
3.6
Unit
V
Supply Voltage
AVDD
VIN
PLL Supply Voltage
Input Voltage
3.6
V
5.5
V
TSTG
Storage Temperature Range
+150
°C
Notes: 1. Functional operating conditions are given in AC and DC electrical specifications. Stresses beyond the absolute maximums
listed may affect device reliability or cause permanent damage to the device.
2. Caution: Input voltage must not be greater than the VDD supply voltage by more than 2.5V at all times including during
power-on reset.
TSPC106
20
TSPC106
Thermal Characteristics
This section provides thermal management information for
the C4/CBGA package. Proper thermal control design is
primarily dependent upon the system-level design.
strate. After the C4 solder bump is reflowed, epoxy
(encapsulant) is under-filled between the die and the sub-
strate. Under-fill material is commonly used on large high-
power die; however, this is not a requirement of the C4
technology. The package substrate is a multilayer-co-fired
ceramic. The package-to-board interconnection is via an
array of orthogonal 90/10 (lead/tin) solder balls on 1.27 mm
pitch. During assembly of the C4/CBGA package to the
board, the high-melt balls do not collapse.
The use of C4 die on CBGA interconnect technology offers
significant reduction in both the signal delay and the micro-
electronic packaging volume. Figure 4 shows the salient
features of the C4/CBGA interconnect technology. The C4
interconnection provides both the electrical and the
mechanical connections for the die to the ceramic sub-
Figure 4. Exploded Cross-section
CI_CGA Package
CBGA Package
Chip with C4 Encapsulant
Ceramic Substrate
BGA Joint
Printed Circuit Board
Internal Package Conduction Resistance
For the C4/CBGA packaging technology, the intrinsic con-
duction thermal resistance paths are as follows:
• the die junction-to-case thermal resistance
These parameters are shown in Table 12. In the C4/CBGA
package, the silicon chip is exposed; therefore, the pack-
age case is the top of the silicon.
• the die junction-to-lead thermal resistance
Table 12. Thermal Resistance
Thermal Metric
Effective Thermal Resistance
0.133°C/W
Junction-to-case thermal resistance
Junction-to-lead (ball) thermal resistance
Junction-to-lead (column) thermal resistance
3.8°C/W (CBGA package)
4.0°C/W (CI_CGA package)
Figure 5 shows a simplified thermal network in which a C4/CBGA package is mounted on a printed-circuit board.
Figure 5. C4/CBGA Package Mounted on a Printed Circuit Board
External Resistance
Radiation
Convection
Heat Sink
Thermal Interface Material
Die/Package
Die Junction
Package/Leads
Internal Resistance
Printed Circuit Board
Radiation
Internal package resistance differs from external package resistance.
Convection
External Resistance
Note:
21
Power Consumption
The TSPC106 provides hardware support for four levels of
power reduction - the doze, nap and sleep modes are
invoked by register programming and the suspend mode is
invoked by assertion of an external signal. The design of
the TSPC106 is fully static, allowing internal logic states to
be preserved during all power-saving modes. The following
sections describe the programmable power modes pro-
vided by the TSPC106.
sor access, the access is completed but the TSPC106
remains in the full-on state. When in the nap mode, the PLL
is required to be running and locked to the system clock
(SYSCLK).
Sleep Mode
Sleep mode provides further power saving compared to the
nap mode. As in nap mode, both the processor and the
TSPC106 are placed in a reduced power mode concur-
rently. In sleep mode, no functional units are operating
except the system RAM refresh logic, which can continue
(optionally) to perform the refresh cycles. A hard reset or a
bus request wakes the TSPC106 from sleep mode. The
PLL and SYSCLK inputs may be disabled by an external
power management controller (PMC). For additional power
savings, the PLL can be disabled by configuring the
PLL[0:3] signals into the PLL-bypass mode. The external
PMC must enable the PLL, turn on SYSCLK and allow the
PLL time to lock before waking the system from sleep
mode.
Full-power Mode
This is the default power state of the TSPC106 following a
hard reset with all internal functional units fully powered
and operating at full clock speed.
Doze Mode
In this power-saving mode, all the TSPC106 functional
units are disabled except for PCI address decoding, sys-
tem RAM refreshing and 60x bus request monitoring
(through BRn). Once the doze mode is entered, a hard
reset, a PCI transaction referenced to system memory or a
60x bus request can bring the TSPC106 out of the doze
mode and into the full-on state. If the TSPC106 is awak-
ened for a processor or PCI bus access, the access is
completed and the MC106 returns to the doze mode. The
TSPC106’s doze mode is totally independent of the power
saving mode of the processor.
Suspend Mode
Suspend mode is activated through assertion of the SUS-
PEND signal. In suspend mode, the TSPC106 may have its
clock input and PLL shut down for additional power sav-
ings. Memory refresh can be accomplished in two ways,
either by using self-refresh mode DRAMs or by using the
RTC input on the TSPC106. To exit the suspend mode, the
system clock must be turned on in sufficient time to restart
the PLL. After this time, SUSPEND may be negated. In
suspend mode, all outputs (except memory refresh) are
released to a high-impedance state and all inputs, including
hard reset (HRST), are ignored.
Nap Mode
Nap mode provides further power savings compared to
doze mode. In nap mode, both the processor and the
TSPC106 are placed in a power reduction mode. In this
mode, only the PCI address decoding, system RAM refresh
and the processor bus request monitoring are still operat-
ing. Hard reset, a PCI bus transaction referenced to system
memory or a 60x bus request can bring the TSPC106 out
of the nap mode. If the TSPC106 is awakened by a PCI
access, the access is completed and the TSPC106 returns
to the nap mode. If the TSPC106 is awakened by a proces-
Power Dissipation
Table 13 provides figures on power consumption for the
TSPC106.
Table 13. Power Consumption
Mode
SYSCLK/Core33/66 MHz
SYSCLK/Core33/83.3 MHz
Unit
Full-on Mode
Typical
1.2
1.4
2.2
2.4
W
W
Maximum
Doze Mode
Typical
1.0
1.2
1.1
1.4
W
W
Maximum
Nap Mode
Typical
1.0
1.2
1.1
1.4
W
W
Maximum
TSPC106
22
TSPC106
Table 13. Power Consumption (Continued)
Mode
SYSCLK/Core33/66 MHz
SYSCLK/Core33/83.3 MHz
Unit
Sleep Mode
Typical
260
360
330
450
W
W
Maximum
Suspend Mode
Typical
140
190
220
270
W
W
Maximum
Notes: 1. Power consumption for common system configurations assuming 50 pF loads.
2. Suspend power saving mode assumes SYSCLK off and PLL in bypass mode.
3. Typical power is an average value measured at VDD = AVDD = 3.30 V and TA = 25 °C.
4. Maximum power is measured at VDD = AVDD = 3.45 V and TA = 25 °C.
Marking
The documents that define the marking are identified in the
related reference documents. Each microcircuit is legibly
and permanently marked with the following information as
a minimum:
• ESD identifier (if available)
• Country of manufacture
Electrical Characteristics
• Manufacturer logo
• Manufacturer part number
General Requirements
• Class B identification (if applicable)
• Date-code of inspection lot
All static and dynamic electrical characteristics specified for
inspection purposes and the relevant measurement condi-
tions are given in Table 14.
Table 14. Clock DC Timing Specifications (VDD = 3.3V 5% dc, GND = 0V dc, -55°C ≤ TC ≤ 125°C)
Symbol
VIH
Characteristic
Min
2
Max
5.5
Unit
V
Input high voltage (all inputs except SYSCLK)
Input low voltage (all inputs except SYSCLK)
SYSCLK input high voltage
VIL
GND
2.4
0.8
V
CVIH
CVIL
Iin
5.5
V
SYSCLK input low voltage
GND
0.4
V
Input leakage current, VIN = 3.3V(1)
Hi-Z (off-state) leakage current, VIN = 3.3V(1)
Output high voltage, IOH = -7 mA
Output low voltage, IOL = 7 mA
15.0
15.0
µA
µA
V
ITSI
VOH
VOL
2.4
2.7
0.5
V
VOH
PCI 3.3V signaling output high voltage,
V
I
OH = -0.5 mA
VOL
Cin
PCI 3.3V signaling output low voltage, IOL = 1.5 mA
Capacitance, Vin = 0V, f = 1 MHz(2)
0.3
7.0
V
pF
Notes: 1. Excludes test signals (LSSD_MODE and JTAG signals).
2. Capacitance is periodically sampled rather than 100% tested.
23
Dynamic Characteristics
This section provides the AC electrical characteristics for
the TPSC106. After fabrication, parts are sorted by maxi-
mum 60x processor bus frequency as shown in Table 15
and tested for conformance to the AC specifications for that
frequency. These specifications are for operation between
16.67 and 33.33 MHz PCI bus (SYSCLK) frequencies. The
60x processor bus frequency is determined by the PCI bus
(SYSCLK) frequency and the settings of the PLL[0:3] sig-
nals. All timings are specified relative to the rising edge of
SYSCLK.
Clock AC Specifications
Table 15 provides the clock AC timing specifications as
defined in Figure 6.
Table 15. Clock AC Timing Specifications (VDD = 3.3V 5% dc, GND = 0V dc, -55°C ≤ TC ≤ 125°C)
SYSCLK/Core 33/66 MHz SYSCLK/Core 33/83.3 MHz
Ref
Characteristic
Min
16.67
120
Max
66
Min
16.67
120
Max
83.3
200
33.33
60.0
2.0
Unit
MHz
MHz
MHz
ns
60x processor bus (core) frequency(1)
VCO frequency(1)
200
33.33
60.0
2.0
SYSCLK frequency(1)
16.67
30.0
16.67
30.0
1
2, 3
4
SYSCLK cycle time
SYSCLK rise and fall time(2)
SYSCLK duty cycle measured at 1.4V(3)
SYSCLK jitter(4)
ns
40
60
40
60
%
200
100
200
100
ps
106 internal PLL relock time(3, 5)
µs
Notes: 1. The SYSCLK frequency and PLL[0:3] settings must be chosen so that the resulting SYSCLK (bus) frequency, CPU (core)
frequency and PLL (VCO) frequency do not exceed their respective maximum or minimum operating frequencies. Refer to
the PLL[0:3] signal description in “System Design Information” on page 32 for valid PLL[0:3] settings.
2. Rise and fall times for the SYSCLK input are measured from 0.4V to 2.4V.
3. Timing is guaranteed by design and characterization and is not tested.
4. The total input jitter (short-term and long-term combined) must be under 200 ps.
5. PLL-relock time is the maximum time required for PLL lock after a stable VDD, AVDD, and SYSCLK are reached during the
power-on reset sequence. This specification also applies when the PLL has been disabled and subsequently reenabled dur-
ing the sleep and suspend power-saving modes. Also note that HRST must be held asserted for a minimum of 255 bus
clocks after the PLL-relock time (100 ms) during the power-on reset sequence.
Figure 6. SYSCLK Input Timing Diagram
1
2
3
4
4
CVIH
SYSCLK
VM
VM
VM
CVIL
Note:
VM = Midpoint Voltage (1.4V)
TSPC106
24
TSPC106
Input AC Specifications
Table 16 provides the input AC timing specifications as shown in Figure 7 and Figure 8.
Table 16. Input AC Timing Specifications (VDD = 3.3V 5% dc, GND = 0V dc, CL = 50 pF, -55°C ≤ TC ≤ 125°C)
66 MHz 83.3 MHz
Ref
Characteristic
Min
Max
Min
Max
Unit
10a
Group I input signals valid to SYSCLK
(input setup)(1, 2, 3)
4.0
3.5
ns
10a
10a
10a
10b
10b
11a
11b
Group II input signals valid to SYSCLK
(input setup)(1, 2, 4)
3.5
3.0
5.0
7.0
7.0
0
3.5
2.5
4.0
7.0
7.0
0
ns
ns
ns
ns
ns
ns
ns
Group III input signals valid to SYSCLK
(input setup)(1, 2, 5)
Group IV input signals valid to SYSCLK
(input setup)(1, 2, 6)
Group V input signals valid to SYSCLK
(input setup)(7, 8)
Group VI input signals valid to SYSCLK
(input setup)(7, 9)
60x Bus Clock to group I - IV inputs invalid
(input hold)(3, 4, 5, 6)
SYSCLK to group V - VI inputs invalid
(input hold)(8, 9)
-0.5
-0.5
HRST pulse width
255 x tSYSCLK
+100 µs
255 x tSYSCLK
+100 µs
10c
11c
Mode select inputs valid to HRST (input
setup)(10, 11, 12)
3 x tSYSCLK
3 x tSYSCLK
ns
ns
HRST to mode select input invalid (input
hold)(10, 12)
1.0
1.0
Notes: 1. Input specifications are measured from the TTL level (0.8 or 2.0V) of the signal in question to the 1.4V of the rising edge of
the input SYSCLK. Input and output timings are measured at the pin.
2. Processor and memory interface signals are specified from the rising edge of the 60x bus clock.
3. Group I input signals include the following processor, L2 and memory interface signals: A[0:31], PAR[0:7]/AR[1:8],BR[0:4],
BRL2, XATS, LBCLAIM, ADS, BA0, TV and HIT (when configured for external L2).
4. Group II input signals include the following processor and memory interface signals: TBST, TT[0:4], TSIZ[0:2], WT, CI, GBL,
AACK and TA.
5. Group III input signals include the following processor and memory interface signals: DL[0:31] and DH[0:31]
6. Group IV input signals include the following processor and L2 interface signals: TS, ARTRY, DIRTY_IN and HIT (when con-
figured for internal L2 controller).
7. PCI 3.3 V signaling environment signals are measured from 1.65V (VDD ÷ 2) on the rising edge of SYSCLK to VOH = 3.0V or
VOL = 0.3V.
PCI 5V signaling environment signals are measured from 1.65V (VDD ÷ 2) on the rising edge of SYSCLK to VOH = 2.4V or
VOL = 0.55V.
8. Group V input signals include the following bussed PCI interface signals: FRAME, C/BE[0:3], AD[0:31], DEVSEL, IRDY,
TRDY, STOP, PAR, PERR, SERR, LOCK, FLSHREQ, and ISA_MASTER.
9. Group VI input signal is the point-to-point PCI GNT input signal.
10. The setup and hold time is with respect to the rising edge of HRST. Mode select inputs include the RCS0, FOE and DBG0
configuration inputs.
11. tSYSCLK is the period of the external clock (SYSCLK) in nanoseconds (ns). When the unit is given as tSYSCLK the numbers
given in the table must be multiplied by the period of SYSCLK to compute the actual time duration (in nanoseconds) of the
parameter in question.
12. These values are guaranteed by design and are not tested.
25
Figure 7. Input Timing Diagram
VM
SYSCLK
All Inputs
10a
10b
11a
Note:
VM = Midpoint Voltage (1.4V)
Figure 8. Mode Select Input Timing Diagram
HRST
VM
10c
11c
MODE Pins
Note:
VM = Midpoint Voltage (1.4V)
TSPC106
26
TSPC106
Output AC Specifications
Table 17 provides the output AC timing specifications as shown in Figure 9.
Table 17. Output AC Timing Specifications (VDD = 3.3V 5% dc, GND = 0V dc, CL = 50 pF, -55°C ≤ TC ≤ 125°C)
66 MHz 83.3 MHz
Ref
Characteristic
Min
Max
Min
Max
Unit
12
SYSCLK to output driven (output enable
time)(9)
2.0
2.0
ns
13a
13b
SYSCLK to output valid (for TS and
ARTRY)(1, 2, 3, 4)
7.0
7.0
6.0
6.0
ns
ns
SYSCLK to output valid (for all non-PCI
signals except TS, ARTRY, RAS[0:7] and
CAS[0:7]) and DWE[0:2](1, 2, 3, 5)
14a
14b
15a
15b
18
SYSCLK to output valid (for RAS[0:7] and
CAS[0:7])(1, 2, 3)
7.0
6.0
ns
ns
ns
ns
ns
ns
ns
SYSCLK to output valid
(for PCI signals)(3, 6)
11.0
11.0
SYSCLK to output invalid for all non-PCI
signals (output hold)(7, 10)
1.0
1.0
1.0
1.0
SYSCLK to output valid for PCI signals
(output hold)(7)
SYSCLK to ARTRY high impedance
before precharge (output hold)(9)
8.0
8.0
19
SYSCLK to ARTRY precharge enable(8, 9)
(0.4 x tSYSCLK
+ 2.0
)
(0.4 x tSYSCLK
+ 2.0
)
21
SYSCLK to ARTRY high impedance after
precharge(8, 9)
(1.5 x tSYSCLK
)
(1.5 x tSYSCLK
+ 8.0
)
+ 8.0
Notes: 1. Processor and memory interface signals are specified from the rising edge of the 60x bus clock.
2. Output specifications are measured from 1.4V on the rising edge of SYSCLK to the TTL level (0.8V or 2.0V) of the signal in
question. Both input and output timings are measured at the pin.
3. The maximum timing specification assumes CL = 50 pF.
4. The shared outputs TS and ARTRY require pull-up resistors to hold them negated when there is no bus master driving them.
5. When the TPSC106 is configured for asynchronous L2 cache SRAMs, the DWE[0:2] signals have a maximum SYSCLK to
output valid time of (0.5 x tPROC) + 8.0 ns (where tPROC is the 60x bus clock cycle time).
6. PCI 3.3V signaling environment signals are measured from 1.65V (Vdd ÷ 2) on the rising edge of SYSCLK to VOH = 3.0V or
VOL = 0.3V.
7. The minimum timing specification assumes CL = 50 pF.
8. tSYSCLK is the period of the external bus clock (SYSCLK) in nanoseconds (ns). When the unit is given as tSYSCLK, the num-
bers given in the table must be multiplied by the period of SYSCLK to compute the actual duration in nanoseconds of the
parameter in question.
9. These values are guaranteed by design and are not tested.
10. PCI devices which require more than the PCI-specified hold time of TH = 0 ns or systems where clock skew approaches the
PCI-specified allowance of 2 ns may not work with the TSPC106. For workarounds, see Motorola application note “Design-
ing PCI 2.1-compliant MPC106 Systems” (order number AN1727/D).
27
Figure 9. Output Timing Diagram
VM
VM
VM
SYSCLK
14
15
16
12
All Outputs
(except TS and ARTRY)
15
13
13
16
TS
21
20
19
18
ARTRY
Note:
VM = Midpoint Voltage (1.4V)
JTAG AC Timing Specifications
Table 18. JTAG AC Timing Specifications (Independent of SYSCLK) (VDD = 3.3V 5% dc, GND = 0V dc,CL = 50 pF,
-55°C ≤ TC ≤ 125°C)
Ref
Characteristic
Min
0
Max
Unit
MHz
ns
TCK frequency of operation
TCK cycle time
25
1
2
40
20
0
TCK clock pulse width measured at 1.4 V
ns
3
TCK rise and fall times
3
ns
4
TRST setup time to TCK rising edge(1)
TRST assert time
10
10
5
ns
5
ns
6
Boundary-scan input data setup time(2)
Boundary-scan input data hold time(2)
TCK to output data valid(3)
TCK to output high impedance(3)
TMS, TDI data setup time
TMS, TDI data hold time
ns
7
15
0
ns
8
30
30
ns
9
0
ns
10
11
12
13
5
ns
15
0
ns
TCK to TDO data valid
15
15
ns
TCK to TDO high impedance
0
ns
Notes: 1. TRST is an asynchronous signal. The setup time is for test purposes only.
2. Non-test signal input timing with respect to TCK.
3. Non-test signal output timing with respect to TCK.
TSPC106
28
TSPC106
Figure 10. JTAG Clock Input Timing Diagram
1
2
2
VM
VM
VM
TCK
3
3
Figure 11. TRST Timing Diagram
TCK
4
TRST
5
Figure 12. Boundary-scan Timing Diagram
TCK
6
7
Input Data Valid
Data Inputs
8
Data Outputs
Output Data Valid
9
8
Data Outputs
Data Outputs
Output Data Valid
29
Figure 13. Test Access Port Timing Diagram
TCK
10
11
Input Data Valid
TDI, TMS
TDO
12
Output Data Valid
13
12
TDO
TDO
Output Data Valid
TSPC106
30
TSPC106
Architectural Overview
60x Processor Interface
Memory Interface
The TPSC106 supports a programmable interface to a vari-
ety of PowerPC microprocessors operating at select bus
speeds.The address bus is 32 bits wide and the data bus is
64 bits wide. The 60x processor interface of the TPSC106
uses a subset of the 60x bus protocol, supporting single-
beat and burst data transfers. The address and data buses
are decoupled to support pipelined transactions.
The memory interface controls processor and PCI interac-
tions to main memory. It is capable of supporting a variety
of DRAM or extended data out (EDO) DRAM and ROM or
Flash ROM configurations as main memory. The maximum
supported memory size is 1G byte of DRAM or EDO
DRAM, with 16M bytes of ROM or Flash ROM. The mem-
ory controller of the TPSC106 supports the various
memory sizes through software initialization of on-chip con-
figuration registers. Parity or ECC is provided for error
detection.
Two signals on the TPSC106, local bus slave claim
(LBCLAIM) and data bus grant local bus slave (DBGLB),
are provided for an optional local bus slave. However, the
local bus slave must be capable of generating the transfer
acknowledge (TA) signal to interact with the 60x
processor(s).
The TPSC106 controls the 64-bit data path to main mem-
ory (72-bit data path with parity or ECC). To reduce loading
on the data bus, system designers must implement buffers
between the 60x bus and memory. The TPSC106 features
configurable data/parity buffer control logic to accommo-
date several buffer types.
Depending on the system implementation, the processor
bus may operate at the PCI bus clock rate or at two or three
times the PCI bus clock rate. The 60x processor bus is syn-
chronous with all timing relative to the rising edge of the
60x bus clock.
The TPSC106 is capable of supporting a variety of
DRAM/EDO configurations. DRAM/EDO banks can be built
of SIMMS, DIMMs or directly-attached memory devices.
Thirteen multiplexed address signals provide for device
densities up to 16 Mbits. Eight row address strobe
(RAS[0:7]) signals support up to eight banks of memory.
The TPSC106 supports bank sizes from 2M bytes to 128M
bytes. Eight column address strobe (CAS[0:7]) signals are
used to provided byte selection for memory bank access
(note that all CAS signals are driven in ECC mode).
Secondary (L2) Cache/Multiple Processor
Interface
The 106 provides support for the following configurations of
60x processors and L2 cache:
• Up to four 60x processors with no L2 cache
• A single 60x processor plus a direct-mapped, lookaside,
L2 cache using the internal L2 cache controller of the
TPSC106
The TPSC106 provides parity checking and generation in
two forms, normal parity and read-modify-write (RMW) par-
ity. As an alternative to simple parity, the TPSC106
supports error checking and correction (ECC) for system
memory. Using ECC, the TPSC106 detects and corrects all
single-bit errors and detects all double-bit errors and all
errors within a nibble (i.e., four bits or one-half byte).
• Up to four 60x processors plus an externally-controlled
L2 cache
The internal L2 cache controller generates the arbitration
and support signals necessary to maintain a write-through
or write-back L2 cache. The internal L2 cache controller
supports either asynchronous SRAMs, pipelined burst
SRAMs or synchronous burst SRAMs, using byte parity for
data error detection.
For ROM/Flash support, the TPSC106 provides 20 address
bits (21 address bits for the 8-bit wide ROM interface), two
bank selects, one output enable, and one Flash ROM write
enable. The 16-Mbyte ROM space is subdivided into two 8-
Mbyte banks. Bank 0 (selected by RCS0) is addressed
from 0xFF80_0000 to 0xFFFF_FFFF. Bank 1 (selected by
RCS1) is addressed from 0xFF00_0000 to 0xFF7F_FFFF.
A configuration signal, flash output enable (FOE) sampled
at reset, determines the bus width of the ROM or Flash
device (8-bit or 64-bit) in bank 0. The data bus width for
ROM bank 1 is always 64 bits. For systems using the 8-bit
interface to bank 0, the ROM/Flash device must be con-
nected to the most-significant byte lane of the data bus
(DH[0:7]).
When more than one 60x processor is used, nine signals of
the L2 interface change their functions (to BR[1:3], BG[1:3]
and DBG[1:3]) to allow for arbitration between the 60x pro-
cessors. The 60x processors share all 60x interface signals
of the TPSC106, except the bus request (BR), bus grant
(BG) and the data bus grant (DBG) signals.
When an external L2 controller (or integrated L2 cache
module) is used, three signals of the L2 interface change
their functions (to BRL2, BGL2 and DBGL2) to allow the
TPSC106 to arbitrate between the external cache and the
60x processor(s).
31
The TPSC106 also supports a mixed ROM system configu-
ration. That is, the system can have the upper 8M bytes
(bank 0) of ROM space located on the PCI bus and the
lower 8M bytes (bank 1) of ROM located on the 60x/mem-
ory bus.
ory operation has one 32-byte read buffer and two 32-byte
write buffers.
System Design Information
This section provides electrical and thermal design recom-
mendations for successful application of the TPSC106.
PCI Interface
The TPSC106’s PCI interface is compliant with the PCI
Local Bus Specification, Revision 2.1, and follows the
guidelines in the PCI System Design Guide, Revision 1.0
for host bridge architecture. The PCI interface connects the
processor and memory buses to the PCI bus, to which I/O
components are connected. The PCI bus uses a 32-bit
multiplexed address/data plus various control and error
signals.
PLL Configuration
The TPSC106 requires a single system clock input,
SYSCLK. The SYSCLK frequency dictates the frequency of
operation for the PCI bus. An internal PLL on the TSPC106
generates a master clock that is used for all of the internal
(core) logic. The master clock provides the core frequency
reference and is phase-locked to the SYSCLK input. The
60x processor, L2 cache and memory interfaces operate at
the core frequency. In the 5:2 clock mode (Rev. 4.0 only),
the TSPC106 needs to sample the 60x bus clock (on the
LBCLAIM configuration input) to resolve clock phasing with
the PCI bus clock (SYSCLK).
The PCI interface of the TPSC106 functions as both a mas-
ter and target device. As a master, the TPSC106 supports
read and write operations to the PCI memory space, the
PCI I/O space, and the PCI configuration space. The
TPSC106 also supports PCI special-cycle and interrupt-
acknowledge commands. As a target, the TPSC106 sup-
ports read and write operations to system memory. Mode
selectable big-endian to little-endian conversion is supplied
at the PCI interface.
The PLL is configured by the PLL[0:3] signals. For a given
SYSCLK (PCI bus) frequency, the clock mode configura-
tion signals (PLL[0:3]) set the core frequency (and the
frequency of the VCO controlling the PLL lock). The sup-
ported core and VCO frequencies and the corresponding
PLL[0:3] settings are provided in Table 19.
Internal buffers are provided for I/O operation between the
PCI bus and the 60x processor or memory. Processor read
and write operations each have a 32-byte buffer and mem-
TSPC106
32
TSPC106
Table 19. Core/VCO Frequencies and PLL Settings
Core Frequency (VCO Frequency) in MHz
Core/SYSCL
PCI Bus
16.6 MHz
PCI Bus
20 MHz
PCI Bus
25 MHz
PCI Bus
33.3 MHz
PLL[0:3](1)
0001
Ratio
VCO Multiplier
1:1
x4
x2
x4
x2
x4
x2
33.3 (133)
66.6 (133)
0100
2:1
0101
2:1
33.3 (133)
41.6 (166)
40 (160)
50 (200)
0110
5:2(2)
5:2(2)
3:1
83.3 (166)
0111
1000
60 (120)
75 (150)
0011
PLL Bypass(3)
PLL off
SYSCLK clocks core circuitry directly
1 x core/SYSCLK ratio implied
1111
Clock off(4)
PLL off
No core clocking occurs
Notes: 1. PLL[0:3] settings not listed are reserved. Some PLL configurations may select bus, CPU or VCO frequencies which are not
useful, not supported or not tested. See “Input AC Specifications” on page 25 for valid SYSCLK and VCO frequencies.
2. 5:2 clock modes are only supported by TSPC106 Rev 4.0; earlier revisions do not support 5:2 clock modes. The 5:2 modes
require a 60x bus clock applied to the 60x clock phase (LBCLAIM) configuration input signal during power-on reset, hard
reset and coming out of sleep and suspend power saving modes.
3. In PLL-bypass mode, the SYSCLK input signal clocks the internal circuitry directly, the PLL is disabled and the
core/SYSCLK ratio is set for 1:1 mode operation. This mode is intended for factory use only.
Note: The AC timing specifications given in this document do not apply in PLL-bypass mode.
4. In clock-off mode, no clocking occurs inside the TSPC106 regardless of the SYSCLK input.
5. PLL[0:3] = 0010 (1:1 Core/SYSCLK Ratio; X8 VCO Multiplier) exists on the chip but will fail to lock 50% of the time. There-
fore this configuration should not be used and 1:1 modes between 16 and 25 MHz are not supported.
power supply, especially while driving large capacitive
loads. This noise must be prevented from reaching other
PLL Power Supply Filtering
The AVDD power signal is provided on the 106 to provide
components in the system, and the TSPC106 itself
power to the clock generation phase-locked loop. To
requires a clean, tightly regulated source of power.
ensure stability of the internal clock, the power supplied to
It is strongly recommended that the system design include
six to eight 0.1 µF (ceramic) and 10 µF (tantalum) decou-
pling capacitors to provide both high- and low-frequency
filtering. These capacitors should be placed closely around
the perimeter of the TSPC106 package (or on the under-
side of the PCB). It is also recommended that these
decoupling capacitors receive their power from separate
VDD and GND power planes in the PCB, utilizing short
traces to minimize inductance. Only SMT (surface mount
technology) capacitors should be used to minimize lead
inductance.
the AVDD input signal should be filtered using a circuit simi-
lar to the one shown in Figure 14. The circuit should be
placed as close as possible to the AVDD pin to ensure it fil-
ters out as much noise as possible.
Figure 14. PLL Power Supply Filter Circuit
10Ω
VDD
(3.3V)
AVDD
10µF
0.1µF
GND
In addition, it is recommended that there be several bulk
storage capacitors distributed around the PCB, feeding the
VDD plane, to enable quick recharging of the smaller chip
capacitors. These bulk capacitors should have a low ESR
(equivalent series resistance) rating to ensure the quick
response time necessary. They should also be connected
Decoupling Recommendations
Due to the TSPC106's large address and data buses, and
high operating frequencies, the TSPC106 can generate
transient power surges and high frequency noise in its
33
to the power and ground planes through two vias to mini-
mize inductance. Suggested bulk capacitors are 100 µF
(AVX TPS tantalum) or 330 µF (AVX TPS tantalum).
LSSD_MODE, must be pulled up for normal device
operation.
During inactive periods on the bus, the address and trans-
fer attributes on the bus (A[0:31], TT[0:4], TBST, WT, CI
and GBL) are not driven by any master and may float in the
high-impedance state for relatively long periods of time.
Since the TSPC106 must continually monitor these signals,
this float condition may cause excessive power draw by the
input receivers on the TSPC106 or by other receivers in the
system. It is recommended that these signals be pulled up
or restored in some manner by the system.
Connection Recommendations
To ensure reliable operation, it is recommended to connect
unused inputs to an appropriate signal level. Unused active
low inputs should be tied (using pull-up resistors) to VDD
.
Unused active high inputs should be tied (using pull-down
resistors) to GND. All NC (no-connect) signals must remain
unconnected.
The 60x data bus input receivers on the TSPC106 do not
require pull-up resistors on the data bus signals (DH[0:31],
DL[0:31] and PAR[0:7]). However, other data bus receivers
in the system may require pull-up resistors on these
signals.
Power and ground connections must be made to all exter-
nal VDD, AVDD and GND pins of the TSPC106.
Pull-up Resistor Recommendations
The TSPC106 requires pull-up (or pull-down) resistors on
several control signals of the 60x and PCI buses to main-
tain the control signals in the negated state after they have
been actively negated and released by the TSPC106 or
other bus masters. The JTAG test reset signal, TRST,
should be pulled down during normal system operation.
Also, as indicated in Table 20, the factory test signal,
In general, the 60x address and control signals are pulled
up to 3.3 Vdc and the PCI control signals are pulled up to 5
Vdc through weak (2-10 kΩ) resistors. Resistor values may
need to be adjusted stronger to reduce induced noise on
specific board designs. Table 20 summarizes the pull-
up/pull-down recommendations for the TSPC106.
Table 20. Pull-up/Pull-down Recommendations
Signal Type
Signals
Pull-up/Pull-down
60x bus control
BRn, TS, XATS, AACK, ARTRY, TA
Pull-up to 3.3V dc
Pull-up to 3.3V dc
Pull-up to 3.3V dc
60x bus address/transfer attributes
Cache control
A[0:31], TT[0:4], TBST, WT, CI, GBL
ADS
HIT, TV
Pull-up to 3.3V dc or pull-down to GND
depending on programmed polarity
PCI bus control
REQ, FRAME, IRDY, DEVSEL, TRDY,
STOP, SERR, PERR, LOCK, FLSHREQ,
ISA_MASTER
Pull-up to 5V dc
JTAG
TRST
Pull-down to GND (during normal system
operation)
Factory test
LSSD_MODE
Pull-up to 3.3V dc
TSPC106
34
TSPC106
Preparation for Delivery
Handling
MOS devices must be handled with certain precautions to
avoid damage due to accumulation of static charge. Input
protection devices have been designed in the chip to mini-
mize the effect of this static buildup. However, the following
handling practices are recommended:
Packaging
Microcircuits are prepared for delivery in accordance with
MIL-PRF-38535.
• Devices should be handled on benches with conductive
and grounded surfaces.
Certificate of Compliance
Atmel offers a certificate of compliance with each shipment
of parts, confirming that the products are in compliance
with MIL-STD-883 and guaranteeing the parameters not
tested at temperature extremes for the entire temperature
range.
• Ground test equipment, tools and operator.
• Do not handle devices by the leads.
• Store devices in conductive foam or carriers.
• Avoid use of plastic, rubber or silk in MOS areas.
• Maintain relative humidity above 50 percent if practical.
35
Package Mechanical Data
CBGA Package Parameters
Table 21. CBGA Package Parameters
Parameter
Min
Max
Parameter
Min
Max
Package outline
Interconnects
21 mm x 25 mm
A
B
C
D
G
H
K
N
P
25.0 0.2
21.0 0.2
303 (16 x 19 ball array minus one)
1.27 mm
Pitch
2.3
3.16
0.93
Solder attach
63/37 Sn/Pb
0.82
Solder balls
10/90 Sn/Pb, 0.89 mm diameter
3.16 mm
1.27 BASIC
Maximum module height
Co-planarity specification
0.79
0.99
0.15 mm
0.635 BASIC
5.8
7.2
6.0
7.4
Figure 15. CBGA Mechanical Drawing
Figure 16 provides the mechanical dimensions and bottom surface nomenclature of the CBGA package.
2X
0.200
– F –
Br
– E –
5
1
A1
*Not to scale
3
8
2
7
6
2
6
Top view
– T –
A
P
0.150
T
2X
0.200
N
1
2 3 4 5 6 7 8 9 1011 1213141516
W
V
U
T
R
P
N
M
L
Bottom View
K
J
H
G
F
C
H
E
D
C
B
A
G
S
S
T
T
E
S
F
S
∅ 0.300
∅ 0.150
K
303X∅ D
G
TSPC106
36
TSPC106
CI_CGA Package Parameters
Table 22. CI-CGA Package Parameters
Parameter
Min
Max
Parameter
Min
Max
Package outline
Interconnects
21 mm x 25 mm
A
B
C
D
G
H
K
N
P
25.0 BASIC
303 (16 x 19 ball array minus one)
1.27 mm
21.0 BASIC
3.84 BASIC
Pitch
Solder attach
63/37 Sn/Pb
0.79
0.99
Solder balls
10/90 Sn/Pb, 0.89 mm diameter
3.84 mm
1.27 BASIC
Maximum module height
Co-planarity specification
1.545
1.695
0.15 mm
0.635 BASIC
6.2 BASIC
7.6 BASIC
Figure 16. CI_CGA Mechanical Drawing
2X
0.200
– F –
Br
– E –
A1
*Not to scale
Top view
A
– T –
P
0.150
T
2X
0.200
N
1
2 3 4 5 6 7 8 9 1011 1213141516
W
V
U
T
R
P
N
M
L
Bottom View
K
J
H
G
F
C
H
E
D
C
B
A
G
S
S
T
T
E
S
F
S
∅ 0.300
∅ 0.150
K
303X∅ D
G
37
Ordering Information
TS
PC106A
M
GS
B/Q
66
CG
Manufacturer's
Prefix
Revision Level(1)
CE: Rev. 3.0
CG: Rev. 4.0
Type
Operating Frequency(1)
66: 66 MHz
83: 83.3 MHz
Temperature Range: TC
M: -55 C, +125 C
V: -40 C, +110 C
Package
G: CBGA
GS: CI_CGA
Screening Level(1)
_: Standard
B/Q: MIL-STD-883, Class Q
B/T: According to MIL-STD-883
U: Upscreening
U/T: Upscreening + burn-in
Note:
1. For availability of different versions, contact your Atmel sales office.
TSPC106
38
Atmel Headquarters
Atmel Operations
Corporate Headquarters
2325 Orchard Parkway
San Jose, CA 95131
TEL (408) 441-0311
FAX (408) 487-2600
Atmel Colorado Springs
1150 E. Cheyenne Mtn. Blvd.
Colorado Springs, CO 80906
TEL (719) 576-3300
FAX (719) 540-1759
Europe
Atmel Irving
Atmel SarL
6431 Longhorn Drive
Irving, TX 75063
TEL (972) 756-3000
FAX (972) 756-3445
Route des Arsenaux 41
Casa Postale 80
CH-1705 Fribourg
Switzerland
TEL (41) 26-426-5555
FAX (41) 26-426-5500
Atmel Rousset
Zone Industrielle
13106 Rousset Cedex
France
TEL (33) 4-4253-6000
FAX (33) 4-4253-6001
Asia
Atmel Asia, Ltd.
Room 1219
Chinachem Golden Plaza
77 Mody Road Tsimhatsui
East Kowloon
Atmel Smart Card ICs
Scottish Enterprise Technology Park
East Kilbride, Scotland G75 0QR
TEL (44) 1355-803-000
Hong Kong
TEL (852) 2721-9778
FAX (852) 2722-1369
FAX (44) 1355-242-743
Atmel Grenoble
Japan
Avenue de Rochepleine
BP 123
Atmel Japan K.K.
9F, Tonetsu Shinkawa Bldg.
1-24-8 Shinkawa
Chuo-ku, Tokyo 104-0033
Japan
38521 Saint-Egreve Cedex
France
TEL (33) 4-7658-3000
FAX (33) 4-7658-3480
TEL (81) 3-3523-3551
FAX (81) 3-3523-7581
Fax-on-Demand
North America:
1-(800) 292-8635
International:
1-(408) 441-0732
e-mail
literature@atmel.com
Web Site
http://www.atmel.com
BBS
1-(408) 436-4309
© Atmel Corporation 2001.
Atmel Corporation makes no warranty for the use of its products, other than those expressly contained in the Company’s standard war-
ranty which is detailed in Atmel’s Terms and Conditions located on the Company’s web site. The Company assumes no responsibility for
any errors which may appear in this document, reserves the right to change devices or specifications detailed herein at any time without
notice, and does not make any commitment to update the information contained herein. No licenses to patents or other intellectual prop-
erty of Atmel are granted by the Company in connection with the sale of Atmel products, expressly or by implication. Atmel’s products are
not authorized for use as critical components in life support devices or systems.
ATMEL® is the registered trademark of Atmel Corporation.
The PowerPC names and logos are trademarks of IBM Corporation used under licence. Other terms
Printed on recycled paper.
and product names in this document may be trademarks of others.
2102A–06/01/0M
相关型号:
TSPC106AVGSB/T83CG
PCI Bus Controller, CMOS, CBGA303, 21 X 25 MM, 3.84 MM HEIGHT, 1.27 MM PITCH, CICGA-303
ATMEL
©2020 ICPDF网 联系我们和版权申明