ATMEL TSPC106A

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 ≤ Tj ≤ +125°C)
– Industrial Temperature Range (-40°C ≤ Tj ≤ +110°C)
VCC = 3.3V ± 5%
Available in a 303-ball CBGA or a 303-ball CBGA with Solder Column Interposer (SCI)
(CI-CGA) Package
Description
The TSPC106 provides an integrated, high-bandwidth, high-performance, TTL-compatible 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 Bus Bridge
Memory
Controller
66-83 MHz
TSPC106
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 boundary-scan capability.
G suffix
CBGA 303
Ceramic Ball Grid Array
GS suffix
CI-CGA 303
Ceramic Ball Grid Array
with Solder Column Interposer (SCI)
Rev. 2102C–HIREL–01/05
Figure 1. TSPC106 Block Diagram
L2 Cache
Interface
Memory
Interface
L2
60x Processor
Interface
Memory
60x Bus
Power Management
Error/Interrupt
Control
Target
Master
PCI Interface
Configuration
Registers
PCI Bus
Functional Description
The TSPC106 provides a PowerPC® microprocessor CHRP-compliant bridge between
the PowerPC microprocessor 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 environment 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 secondary 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.
The TSPC106 supports a programmable interface to a variety of PowerPC microprocessors operating at select bus speeds. The 60x address bus is 32 bits wide; the data bus
is 64 bits wide. The 60x processor interface of the TSPC106 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.
2
TSPC106A
2102C–HIREL–01/05
TSPC106A
The TSPC106 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 TSPC106
•
Up to four 60x processors plus an externally controlled L2 cache (e.g., the Freescale
MPC2604GA integrated L2 lookaside cache)
The memory interface controls processor and PCI interactions to main memory and is
capable of supporting a variety of configurations using DRAM, EDO, or SDRAM and
ROM or Flash ROM.
The PCI interface of the TSPC106 complies 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 bus plus various control and error signals.
The PCI interface of the TSPC106 functions as both a master and target device. As a
master, the 106 supports read and write operations to the PCI memory space, the PCI
I/O space and the PCI configuration space. The TSPC106 also supports PCI specialcycle 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.
3
2102C–HIREL–01/05
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
DL26
DL28
DL30
DH31
DH29
DH27
DH25
DH23
DH21
DH19
DH17
DH15
DH13
DH11
DH9
DH7
V
DL24
DL27
DL29
DL31
DH30
DH28
DH26
DH24
DH20
DH18
DH16
DH14
DH12
DH10
DH8
DL22
U
MA1/
SDBA0/
AR9
DL23
DL25
DL14
PLL2
PLL0
DL12
DL10
DL4
DL2
DL0
DOE/
DBGL2
DBG1
DH6
DL21
DL20
T
MA2/
SDMA2/
AR10
WE
DH0
DL15
PLL3
PLL1
DL13
DL11
DL3
DL1
TV/
BR2
BA0/
BR3
HIT
DIRTY_IN/
BR1
DL19
DCS/
BG3
R
MA3/
SDMA3/
AR11
RCS0
DH2
DH1
DL16
VSS
VDD
DL9
DL5
VSS
VDD
TWE/
BG2
DIRTY_OUT/
BG1
ADS/
DALE/
BRL2
A0
TS
P
MA5/
SDMA5/
AR13
MA4/
SDMA4/
AR12
DH4
DH3
VSS
VDD
VSS
DL8
DL6
VDD
VSS
VDD
BA1/
BAA
BGL2
DWE0/
DBG2
A1
XATS/
SDMA1
N
MA6/
SDMA6/
AR14
MA0/
SDBA1/
SDMA0/
AR0
DL17
DH5
VDD
VSS
VDD
DL7
DH22
VSS
VDD
VSS
LBCLAIM
CI
A2
TA
M
MA8/
SDMA8/
AR16
MA7/
SDMA7/
AR15
RAS0/
CS0
DL18
VSS
VDD
VSS
NC
NC
VDD
VSS
VDD
WT
GBL
A3
TT4
L
HRST
MA9/
SDMA9/
AR17
QACK
RAS1/
CS1
VDD
CKO/
DWE2
RAS5/
CS5
VSS
VDD
VSS
SYSCLK
DBG0
TBST
BR0
A4
TT3
K
MA11/
MA10/
SDMA11/ SDMA10/
AR19
AR18
RAS3/
CS3
RAS2/
CS2
RAS4/
CS4
RAS7/
CS7
VDD
AVDD
VSS
VDD
A9
A8
A7
BG0
A5
TT2
J
MA12/
SDMA12/
AR20
CAS0/
DQM0
PPEN
RCS1
RAS6/
CS6
MCP
DBGLB/
CKE
VSS
VDD
VSS
A11
A6
A13
A12
A10
TEA
H
QREQ
CAS1/
DQM1
SUSPEND
TRST
VSS
DWE1/
DBG3
PIRQ/
SDRAS
NC
NC
VDD
VSS
VDD
A15
A14
A16
TT1
G
CAS2/
DQM2
RTC
CAS4/
DQM4
CAS5/
DQM5
VDD
LSSD_
MODE
VDD
PAR
LOCK
VSS
VDD
VSS
TSIZ1
TSIZ0
A17
TT0
F
BCTL0
BCTL1
CAS6/
DQM6
TCK
VSS
VDD
VSS
PERR
DEVSEL
VDD
VSS
VDD
A21
TSIZ2
ARTRY
A18
E
CAS3/
DQM3
NMI
CAS7/
DQM7
MDLE/
SDCAS
TDO
VSS
VDD
SERR
IRDY
VSS
VDD
A31
A29
A22
A20
A19
D
PAR0/
AR1
PAR1/
AR2
TMS
FOE
AD28
AD24
AD21
AD17
AD14
AD10
C/BE0
AD4
AD0
A30
AACK
A23
C
PAR2/
AR3
PAR3/
AR4
PAR5/
AR6
AD30
AD26
AD23
AD19
C/BE2
C/BE1
AD12
AD8
AD6
AD2
A27
A25
A24
B
PAR4/
AR5
PAR7/
AR8
AD1
TDI
AD7
AD11
AD15
TRDY
AD18
AD22
AD25
AD29
REQ
ISA_MASTER/
BERR
A28
A26
A
PAR6/
AR7
GNT
AD3
AD5
AD9
AD13
FRAME
STOP
AD16
AD20
C/BE3
AD27
AD31
FLSHREQ
MEMACK
4
TSPC106A
2102C–HIREL–01/05
TSPC106A
Figure 3. Pin Assignments Shading Ley
VIEW
NC
VSS
VDD Power Supply Positive
No connect
AVDD Clock Power Supply Positive (K9)
Power Supply Ground
Signals
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
Low
I/O
ARTRY
F2
Low
I/O
BG0
K3
Low
Output
BG1 (DIRTY_OUT)
R4
Low
Output
BG2 (TWE)
R5
Low
Output
BG3 (DCS)
T1
Low
Output
BR0
L3
Low
Input
BR1 (DIRTY_IN)
T3
Low
Input
BR2 (TV)
T6
Low
Input
BR3 (BA0)
T5
Low
Input
CI
N3
Low
I/O
DBG0
L5
Low
Output
DBG1 (TOE)
U4
Low
Output
DBG2 (DWE0)
P3
Low
Output
DBG3 (DWE1)
H11
Low
Output
DBGLB (CKE)
J10
Low
Output
DH[0:31]
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
High
I/O
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
I/O
LBCLAIM
N4
Low
Input
5
2102C–HIREL–01/05
Table 1. TSPC106 Pinout in 303-ball CBGA Package (Continued)
Signal Name
Pin Number
Active
I/O
MCP
J11
Low
Output
TA
N1
Low
I/O
TBST
L4
Low
I/O
TEA
J1
Low
Output
TS
R1
Low
I/O
TSIZ[0:2]
G3, G4, F3
High
I/O
TT[0:4]
G1, H1, K1, L1, M1
High
I/O
WT
M4
Low
I/O
XATS (SDMA1)
P1
Low
Input
ADS/DALE/BRL2
R3
Low
Output
BA0 (BR3)
T5
Low
Output
BA1/BAA/BGL2
P4
Low
Output
DBGL2/DOE
U5
Low
Output
DCS (BG3)
T1
Low
Output
DIRTY_IN (BR1)
T3
Low
Input
DIRTY_OUT (BG1)
R4
Low
Output
DWE0 (DBG2)
P3
Low
Output
DWE1 (DBG3)
H11
Low
Output
DWE2 (CKO)
L11
Low
Output
HIT
T4
Low
Input
TOE (DBG1)
U4
Low
Output
TV (BR2)
T6
High
I/O
TWE (BG2)
R5
Low
Output
BCTL [0:1]
F16, F15
Low
Output
BERR (ISA_MASTER)
B3
Low
Input
CAS/DQM[0:7]
J15, H15, G16, E16, G14, G13, F14, E14
Low
Output
CKE/DBGLB
J10
High
Output
FOE
D13
Low
Output
MA0/SDBA1/SDMA0/AR0
N15
High
Output
SDMA1 (XATS)
P1
High
Output
MA1/SDBA0/AR9
U16
High
Output
MA[2:12]/SDMA[2:12]/AR[10:20]
T16, R16, P15, P16, N16, M15, M16, L15, K15, K16, J16
High
Output
MDLE/SDCAS
E13
Low
Output
PAR[0:7]/AR[1:8]
D16, D15, C16, C15, B16, C14, A16, B15
High
I/O
L2 Cache Interface Signals
Memory Interface Signals
6
TSPC106A
2102C–HIREL–01/05
TSPC106A
Table 1. TSPC106 Pinout in 303-ball CBGA Package (Continued)
Signal Name
Pin Number
Active
I/O
PPEN
J14
Low
Output
RAS/CS[0:7]
M14, L13, K13, K14, K12, L10, J12, K11
Low
Output
RCS0
R15
Low
I/O
RCS1
J13
Low
Output
RTC
G15
High
Input
SDRAS (PIRQ)
H10
Low
Output
WE
T15
Low
Output
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)
A6, C9, C8, D6
Low
I/O
DEVSEL
F8
Low
I/O
FLSHREQ
A3
Low
Input
FRAME
A10
Low
I/O
GNT
A15
Low
Input
IRDY
E8
Low
I/O
ISA_MASTER (BERR)
B3
Low
Input
LOCK
G8
Low
Input
MEMACK
A2
Low
Output
PAR
G9
High
I/O
PERR
F9
Low
I/O
PIRQ (SDRAS)
H10
Low
Output
REQ
B4
Low
Output
SERR
E9
Low
I/O
STOP
A9
Low
I/O
TRDY
B9
Low
I/O
PCI Interface Signals(2)
Interrupt, Clock and Power Management Signals
CK0 (DWE2)
L11
High
Output
HRST
L16
Low
Input
NMI
E15
High
Input
QACK
L14
Low
Output
QREQ
H16
Low
Input
SYSCLK
L6
Clock
Input
SUSPEND
H14
Low
Input
U11, T11, U12, T12
High
Input
Test/Configuration Signals
PLL[0:3]
7
2102C–HIREL–01/05
Table 1. TSPC106 Pinout in 303-ball CBGA Package (Continued)
Signal Name
Pin Number
Active
I/O
TCK
F13
Clock
Input
TDI
B13
High
Input
TDO
E12
High
Output
TMS
D14
High
Input
TRST
H13
Low
Input
AVDD
K9
High
Clock
LSSD_MODE(3)
G11
Low
Input
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
High
Power
VSS
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
–
–
Power and Ground Signals
NC
Notes:
H8, H9, M8, M9
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.
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
•
Interrupt, clock and power management signals
•
IEEE 1149.1 interface signals
•
Configuration signals
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.
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.
8
TSPC106A
2102C–HIREL–01/05
TSPC106A
Figure 4. Symbol
AD[31:0]
C/BE[3:0]
PAR
TRDY
IRDY
FRAME
STOP
PCI
Interface
LOCK
DEVSEL
SERR, PERR
REQ
GNT
FLSHREQ
MEMACK
ISA_MASTER
PIRQ
RAS[0:7]
CAS[0:7]
WE
MA0/AR0
Memory
Interface
MA[1:12]/AR[9:20]
PAR[0:7]/AR[1:8]
FOE
(1)
MDLE, PPEN
RCS1
RCS0
(1)
BCTL0, BCTL1
RTC
SYSCLK, CKO(1)
HRST
Interrupt, Clock
and Power
Management
Signals
NMI
QREQ
QACK
SUSPEND
DBG0
FOE
Configuration
(1)
RCS0
(1)
PLL[0:3]
Note:
32
1
4
1
1
1
1
1
1
32
1
5
1
3
1
1
1
3
2
1
1
1
1
1
1
64
1
1
1
1
1
1
8
2
8
1
1
1
1
1
12
8
3
1
1
1
2
1
1
1
1
1
2
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
BR0
BG0
TS
XATS
A[0:31]
TT[0:4]
TSIZ[0:2]
TBST
60x
Processor
Interface
WT, CI, GBL
AACK
ARTRY
(1)
DBG0
DH[0:31], DL[0:31]
TA
LBCLAIM
DBGLB
MCP, TEA
ADS/DALE/BRL2
BAA/BA1/BGL2
DOE/DBGL2
(1)
DWE[0:2]/DBG2, DBG3, CKO
HIT
DCS/BG3
BA0/BR3
TWE/BG2
L2 Cache/
Multiple
Processor
Interface
TV/BR2
DIRTY_IN/BR1
DIRTY_OUT/BG1
TOE/DBG1
TDO
TDI
TCK
TMS
IEEE 1149.1
JTAG
Interface
TRST
1
1
1
4
Some signals have dual functions and are shown more than once in this figure.
9
2102C–HIREL–01/05
60x Processor Interface Signals
Table 2. 60x Processor Interface Signals
Signal
A[0:31]
AACK
ARTRY
Signal Name
Address bus
Address
acknowledge
Address retry
Number of
Pins
I/O
Signal Description
O
Specifies the physical address for 60x bus snooping.
I
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.
O
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.
I
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.
O
Indicates that the initiating 60x bus master must retry the current
address tenure.
I
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.
32
1
1
BG0
Bus grant 0
1
O
Indicates that the primary 60x may, with the proper qualification, begin
a bus transaction and assume mastership of the address bus.
BR0
Bus request 0
1
I
Indicates that the primary 60x requires the bus for a transaction.
CI
Cache inhibit
1
I/O
Indicates that an access is caching-inhibited.
DBG0
Data bus grant 0
1
O
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.
10
TSPC106A
2102C–HIREL–01/05
TSPC106A
Table 2. 60x Processor Interface Signals (Continued)
Signal
Signal Name
Number of
Pins
I/O
Signal Description
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:
DH[0:31],
DL[0:31]
Data bus
64
Data Byte
Byte Lane
DH[0:7]
0
DH[8:15]
1
DH[16:23]
2
DH[24:31]
3
DL[0:7]
4
DL[8:15]
5
DL[16:23]
6
DL[24:31]
7
O
Represents the value of data being driven by the TSPC106.
I
Represents the state of data being driven by a 60x processor, the local
bus slave, the L2 cache or the memory subsystem.
Indicates that an access is global and hardware needs to enforce
coherency.
GBL
Global
1
I/O
LBCLAIM
Local bus slave
cycle claim
1
I
Indicates that the local bus slave claims the transaction and is
responsible for driving TA during the data tenure.
MCP
Machine check
1
O
Indicates that the TSPC106 detected a catastrophic error and the 60x
processor should initiate a machine check exception.
O
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.
O
Indicates that a burst transfer is in progress.
I
Indicates that a burst transfer is in progress.
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).
O
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.
I
Indicates that a 60x bus master has begun a transaction and that the
address and transfer attribute signals are valid.
TA
TBST
TEA
TS
Transfer
acknowledge
Transfer burst
Transfer error
acknowledge
Transfer start
1
1
1
1
11
2102C–HIREL–01/05
Table 2. 60x Processor Interface Signals (Continued)
Signal
Signal Name
Number of
Pins
TSIZ[0:2]
Transfer size
3
TT[0:4]
Transfer type
WT
Write-through
XATS
Extended address
transfer start
L2 Cache/Multiple
Processor Interface
Signals
Internal L2 Controller Signals
I/O
Signal Description
O
Specifies the data transfer size for the 60x bus transaction.
I
Specifies the data transfer size for the 60x bus transaction.
O
Specifies the type of 60x bus transfer in progress.
I
Specifies the type of 60x bus transfer in progress.
5
1
1
I/O
I
Indicates that an access is write-through.
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).
The TSPC106 provides support for either an internal L2 cache controller or an external
L2 cache controller and/or additional 60x processors.
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
Pins
I/O
Signal Description
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
1
I/O
For an asynchronous SRAM configuration, indicates bit 0 of the burst
address counter
BA1
BAA
BGL2
Burst address 1
1
O
For an asynchronous SRAM configuration, indicates bit 1 of the burst
address counter
BAA
BA1
BGL2
Bus address
advance
1
O
For a burst SRAM configuration, indicates that the burst RAMs should
increment their internal addresses
DALE
ADS
BRL2
Data address latch
enable
1
O
For an asynchronous SRAM configuration, indicates that the external
address latch should latch the current 60x bus address
DCS
BG3
Data RAM chip
select
1
O
Enables the L2 data RAMs for a read or write operation
DIRTY_IN
BR1
Dirty in
1
I
Indicates that the selected L2 cache line is modified. The polarity of
DIRTY_IN is programmable
DIRTY_OUT
BG1
Dirty out
1
O
Indicates that the L2 cache line should be marked as modified. The
polarity of DIRTY_OUT is programmable
Signal
Signal Name
ADS
DALE
BRL2
12
TSPC106A
2102C–HIREL–01/05
TSPC106A
Table 3. Internal L2 Controller Signals (Continued)
Number of
Pins
I/O
Data RAM output
enable
1
O
Indicates that the L2 data RAMs should drive the data bus
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
Hit
1
I
Indicates that the L2 cache has detected a hit. The polarity of HIT is
programmable
TOE
DBG1
Tag output enable
1
O
Indicates that the tag RAM should drive the L2 tag address onto the
address bus.
TV
BR2
Tag valid
1
I/O
Indicates that the current L2 cache line should be marked valid. The
polarity of TV is programmable
TWE
BG2
Tag write enable
1
O
Indicates that the L2 tag address, valid, and dirty bits should be
updated
Signal
Signal Name
DOE
DBGL2
DWE[0:2]
DBG2
DBG3
HIT
External L2 Controller Signals
Signal Description
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
Pins
I/O
Signal Description
External L2 bus
grant
1
O
Indicates that the external L2 controller has been granted mastership
of the 60x address bus
BRL2
ADS
DALE
External L2 bus
request
1
I
Indicates that the external L2 controller requires mastership of the 60x
bus for a transaction
DBGL2
DOE
External L2 data
bus grant
1
O
Indicates that the external L2 controller has been granted mastership
of the 60x data bus
HIT
External L2 hit
1
I
Indicates that the current transaction is claimed by the external L2
controller. The external L2 controller will assert AACK and TA for the
transaction
Signal
Signal Name
BGL2
BA1
BAA
13
2102C–HIREL–01/05
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
Pins
I/O
Signal Description
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
O
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
1
O
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
1
I
Indicates that processor 1 requires mastership of the 60x bus for a
transaction.
BR2
TV
Bus request 2
1
I
Indicates that processor 2 requires mastership of the 60x bus for a
transaction.
BR3
BA0
Bus request 3
1
I
Indicates that processor 3 requires mastership of the 60x bus for a
transaction.
DBG1
TOE
Data bus grant 1
1
O
Indicates that processor 1 may, with the proper qualification, assume
mastership of the 60x data bus.
DBG2
DWE0
Data bus grant 2
1
O
Indicates that processor 2 may, with the proper qualification, assume
mastership of the 60x data bus.
DBG3
DWE1
Data bus grant 3
1
O
Indicates that processor 3 may, with the proper qualification, assume
mastership of the 60x data bus.
Signal
Signal Name
BG1
DIRTY_OUT
14
TSPC106A
2102C–HIREL–01/05
TSPC106A
Memory Interface
Signals
Table 6 lists the memory interface signals and provides a brief description of their functions. The memory interface supports 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
Signal
Signal Name
Number of
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
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
O
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
O
Used to control external data bus buffers (directional control and highimpedance 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]
Column address
strobe 0 - 7
8
O
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.
FOE
Flash output
enable
1
O
Enables Flash output for the current read access.
MA0
MA[1:12]
AR0
AR[9:20]
Memory address
0 - 12
13
O
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].
MDLE
Memory data latch
enable
1
O
Enables the external, latched data buffer for read operations, if such a
buffer is used in the system.
O
Represents the byte parity or ECC being written to memory (PAR0 is the
most significant bit).
I
Represents the byte parity or ECC being read from memory (PAR0 is the
most significant bit).
PAR[0:7]
AR[1:8]
Data parity/ECC
8
PPEN
Parity path read
enable
1
O
Enables external parity/ECC bus buffer or latch for memory read
operations.
RAS[0:7]
Row address
strobe 0 - 7
8
O
Indicates a memory row address is valid and selects one of the rows in the
bank.
RCS0
ROM/Flash bank 0
select
1
O
Selects ROM/Flash bank 0 for a read access or Flash bank 0 for a read or
write access.
RCS1
ROM/Flash bank 1
select
1
O
Selects ROM/Flash bank 1 for a read access or Flash bank 1 for a read or
write access.
RTC
Real-time clock
1
I
External clock source for the memory refresh logic when the TSPC106 is in
the suspend power-saving mode.
WE
Write enable
1
O
Enables writing to DRAM, EDO or Flash.
15
2102C–HIREL–01/05
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 referenced 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
Signal
Signal Name
AD[31:0]
Address/data
Command/byte
enable
C/BE[3:0]
DEVSEL
Device select
Number of
Pins
32
1
FRAME
Frame
1
Initializer ready
IRDY
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.
O
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.
I
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.
O
Indicates that the TSPC106 has decoded the address and is the target
of the current access.
I
Indicates that some PCI agent (other than the TSPC106) has decoded
its address as the target of the current access.
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
Indicates that another PCI master is initiating a bus transaction.
I
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.
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
Indicates another PCI master is able to complete the current data
phase of the transaction.
1
Flush request
PCI bus grant
Signal Description
4
FLSHREQ
GNT
I/O
1
1
ISA_
MASTER
ISA master
1
I
Indicates that an ISA master is requesting system memory.
LOCK
Lock
1
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.
16
1
TSPC106A
2102C–HIREL–01/05
TSPC106A
Table 7. PCI Interface Signals
Signal
PAR
PERR
Signal Name
Parity
Parity error
Number of
Pins
I/O
Signal Description
O
Asserted indicates odd parity across the AD[31:0] and C/BE[3:0]
signals during address and data phases. Negated indicates even
parity.
I
Asserted indicates odd parity driven by another PCI master or the PCI
target during read data phases. Negated indicates even parity.
O
Indicates that another PCI agent detected a data parity error.
I
Indicates that another PCI agent detected a data parity error.
1
1
PIRQ
Modified memory
interrupt request
1
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
1
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.
O
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.
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.
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.
SERR
STOP
TRDY
System error
Stop
Target ready
1
1
1
17
2102C–HIREL–01/05
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
Pins
I/O
Test clock
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
Hard reset
1
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.
NMI
Nonmaskable
interrupt
1
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
Quiesce
acknowledge
1
O
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.
QREQ
Quiesce request
1
I
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
Suspend
1
I
Activates the suspend power-saving mode.
I
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.
Signal
Signal Name
CKO
DWE2
SYSCLK
System clock
1
Signal Description
IEEE 1149.1 Interface Signals
To facilitate system testing, the TSPC106 provides a JTAG test access port that complies with the IEEE 1149.1 boundaryscan specification.
Table 9. IEEE 1149.1 Interface Signals
Number of
Pins
I/O
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
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
JTAG test mode
select
1
I
This signal is decoded by the internal JTAG TAP controller to
distinguish the primary operation of the test support circuitry.
TRST
JTAG test reset
1
I
This input causes asynchronous initialization of the internal JTAG TAP
controller.
Signal
Signal Name
TCK
18
Signal Description
TSPC106A
2102C–HIREL–01/05
TSPC106A
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
Signal
Number of Pins
I/O
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)
4
I
High/Low – configuration for the PLL clock mode.
RCS0(1)
1
I
High indicates ROM is located on the 60x processor/memory data bus.
Low indicates ROM is located on the PCI bus.
Notes:
Configuration
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
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 PLL[0:3] signals configure the core-to-SYSCLK frequency 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.
Address Maps
The TSPC106 supports three address mapping configurations 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 Reference Platform Architecture (CHRP)”. The emulation
mode address map is provided to support software emulation 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 emulation mode address map can only be selected by software after
reset.
19
2102C–HIREL–01/05
Detailed
Specifications
Scope
This drawing describes the specific requirements of the TSPC106 in compliance with
MIL-STD-883 class B or manufacturer’s standard screening.
Applicable Documents
Documents applicable to the information contained in this datasheet are:
1. MIL-STD-883: Test Methods and Procedures for Electronics
2. MIL-PRF-38535: General Specifications for Microcircuits
Requirements
General
The microcircuits are in accordance with the applicable documents and as specified
herein.
Design and Construction
Terminal Connections
The terminal connections are as shown in “Pin Description” on page 4.
Lead Material and Finish
Lead material and finish are as specified in “Package Mechanical Data” on page 37.
Package
The macrocircuits are packaged in 303-pin ceramic ball grid array packages.
The precise package drawings are described at the end of the specification. “CBGA
Package Parameters” on page 37 and “CI_CGA Package Parameters” on page 38.
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
Parameter
Min
Max
Unit
VDD
Supply Voltage
-0.3
3.6
V
AVDD
PLL Supply Voltage
-0.3
3.6
V
VIN
Input Voltage
-0.3
5.5
V
TSTG
Storage Temperature Range
-55
+150
°C
Notes:
Thermal Characteristics
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.
This section provides thermal management information for the C4/CBGA package.
Proper thermal control design is primarily dependent upon the system-level design.
The use of C4 die on CBGA interconnect technology offers significant reduction in both
the signal delay and the microelectronic packaging volume. Figure 5 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 substrate.
20
TSPC106A
2102C–HIREL–01/05
TSPC106A
After the C4 solder bump is reflowed, epoxy (encapsulant) is under-filled between the
die and the substrate. 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.
Figure 5. 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 conduction thermal resistance
paths are as follows:
•
the die junction-to-case thermal resistance
•
the die junction-to-lead thermal resistance
These parameters are shown in Table 12. In the C4/CBGA package, the silicon chip is
exposed; therefore, the package case is the top of the silicon.
Table 12. Thermal Resistance
Thermal Metric
Effective Thermal Resistance
Junction-to-case thermal resistance
0.133°C/W
Junction-to-lead (ball) thermal resistance
3.8°C/W (CBGA package)
Junction-to-lead (column) thermal resistance
4.0°C/W (CI_CGA package)
Figure 6 shows a simplified thermal network in which a C4/CBGA package is mounted
on a printed-circuit board.
Figure 6. C4/CBGA Package Mounted on a Printed Circuit Board
Radiation
External Resistance
Convection
Heat Sink
Thermal Interface Material
Die/Package
Die Junction
Package/Leads
Internal Resistance
Printed Circuit Board
External Resistance
Note:
Radiation
Convection
Internal package resistance differs from external package resistance.
21
2102C–HIREL–01/05
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 provided by the TSPC106.
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, system 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 awakened 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.
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 operating. 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 processor 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 concurrently.
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.
Suspend Mode
Suspend mode is activated through assertion of the SUSPEND signal. In suspend
mode, the TSPC106 may have its clock input and PLL shut down for additional power
savings. 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.
22
TSPC106A
2102C–HIREL–01/05
TSPC106A
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
Typical
1.2
2.2
W
Maximum
1.4
2.4
W
Typical
1.0
1.1
W
Maximum
1.2
1.4
W
Typical
1.0
1.1
W
Maximum
1.2
1.4
W
Typical
260
330
W
Maximum
360
450
W
Typical
140
220
W
Maximum
190
270
W
Full-on Mode
Doze Mode
Nap Mode
Sleep Mode
Suspend Mode
Notes:
Marking
1.
2.
3.
4.
Power consumption for common system configurations assuming 50 pF loads.
Suspend power saving mode assumes SYSCLK off and PLL in bypass mode.
Typical power is an average value measured at VDD = AVDD = 3.30 V and TA = 25 °C.
Maximum power is measured at VDD = AVDD = 3.45 V and TA = 25 °C.
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:
•
Manufacturer logo
•
Manufacturer part number
•
Class B identification (if applicable)
•
Date-code of inspection lot
•
ESD identifier (if available)
•
Country of manufacture
23
2102C–HIREL–01/05
Electrical Characteristics
Table 14. Recommended Operating Conditions
Characteristic
Symbol
Value
Unit
Supply voltage
VDD
3.3 ± 165 mv
V
PLL supply voltage
AVDD
Min 2.51 - Max 3.465
V
Input voltage
Vin
0 to 5.5
V
Die junction
temperature
Tj
-55 to +125
°C
General Requirements
Notes
All static and dynamic electrical characteristics specified for inspection purposes and the
relevant measurement conditions are given in Table 15.
Table 15. Clock DC Timing Specifications (VDD = 3.3V ± 5% dc, GND = 0V dc, -55°C ≤ Tj ≤ 125°C)
Symbol
Characteristic
Min
Max
Unit
VIH
Input high voltage (all inputs except SYSCLK)
2
5.5
V
VIL
Input low voltage (all inputs except SYSCLK)
GND
0.8
V
CVIH
SYSCLK input high voltage
2.4
5.5
V
CVIL
SYSCLK input low voltage
GND
0.4
V
Iin
Input leakage current, VIN = 3.3V(1)
15.0
µA
15.0
µA
(1)
ITSI
Hi-Z (off-state) leakage current, VIN = 3.3V
VOH
Output high voltage, IOH = -7 mA
VOL
Output low voltage, IOL = 7 mA
VOH
PCI 3.3V signaling output high voltage,
IOH = -0.5 mA
VOL
PCI 3.3V signaling output low voltage, IOL = 1.5 mA
Cin
Notes:
24
Capacitance, Vin = 0V, f = 1 MHz
2.4
V
0.5
(2)
2.7
V
V
0.3
V
7.0
pF
1. Excludes test signals (LSSD_MODE and JTAG signals).
2. This value represents worst case 40 Ohm drivers (default value for Processor/L2 control signals CI, WT, GBL, TBST, TSIZ[02], TT[0-4], TWE, and TV) only. Other signals have lower default driver impedance and will support larger IOH and IOL. All
drivers may optionally be programmed to different driver strengths.
3. Capacitance is periodically sampled rather than 100% tested.
TSPC106A
2102C–HIREL–01/05
TSPC106A
Dynamic Characteristics
This section provides the AC electrical characteristics for the TSPC106. After fabrication, parts are sorted by maximum 60x processor bus frequency as shown in Table 16
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] signals. All timings are specified relative to the rising
edge of SYSCLK.
Clock AC Specifications
Table 16 provides the clock AC timing specifications as defined in Figure 7.
Table 16. Clock AC Timing Specifications (VDD = 3.3V ± 5% dc, GND = 0V dc, -55°C ≤ Tj ≤ 125°C)
SYSCLK/Core 33/66 MHz
Ref
Characteristic
Min
Max
Min
Max
Unit
16.67
66
16.67
83.3
MHz
150
400
150
400
MHz
SYSCLK frequency
16.67
33.33
16.67
33.33
MHz
SYSCLK cycle time
30.0
60.0
30.0
60.0
ns
2.0
ns
60
%
±200
±200
ps
100
100
µs
60x processor bus (core) frequency(1)
VCO frequency(1)
(1)
1
2, 3
4
SYSCLK rise and fall time(2)
2.0
(3)
SYSCLK duty cycle measured at 1.4V
40
(4)
SYSCLK jitter
106 internal PLL relock time
Notes:
SYSCLK/Core 33/83.3 MHz
(3, 5)
60
40
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 33 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 during 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 7. SYSCLK Input Timing Diagram
1
2
4
3
4
CVIH
SYSCLK
VM
VM
VM
CVIL
Note:
VM = Midpoint Voltage (1.4V)
25
2102C–HIREL–01/05
Input AC Specifications
Table 17 provides the input AC timing specifications as shown in Figure 8 and Figure 9.
Table 17. Input AC Timing Specifications (VDD = 3.3V ± 5% dc, GND = 0V dc, CL = 50 pF, -55°C ≤ Tj ≤ 125°C)
66 MHz
83.3 MHz
Ref
Characteristic
Min
10a
Group I input signals valid to SYSCLK (input
setup)(1, 2, 3)
4.0
3.5
ns
10a
Group II input signals valid to SYSCLK (input
setup)(1, 2, 4)
3.5
3.5
ns
10a
Group III input signals valid to SYSCLK (input
setup)(1, 2, 5)
3.0
2.5
ns
10a
Group IV input signals valid to SYSCLK (input
setup)(1, 2, 6)
5.0
4.0
ns
10b
Group V input signals valid to SYSCLK (input
setup)(7, 8)
7.0
7.0
ns
10b
Group VI input signals valid to SYSCLK (input
setup)(7, 9)
7.0
7.0
ns
11a
60x Bus Clock to group I - IV inputs invalid (input
hold)(3, 4, 5, 6)
0
0
ns
-0.5
-0.5
ns
255 x tSYSCLK
+100 µs
255 x tSYSCLK
+100 µs
3 x tSYSCLK
3 x tSYSCLK
ns
1.0
1.0
ns
11b
SYSCLK to group V - VI inputs invalid (input hold)(8,
9)
HRST pulse width
10c
11c
Notes:
26
Mode select inputs valid to HRST (input setup)(10,
11, 12)
HRST to mode select input invalid (input hold)(10, 12)
Max
Min
Max
Unit
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 configured 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.
TSPC106A
2102C–HIREL–01/05
TSPC106A
Figure 8. Input Timing Diagram
VM
SYSCLK
10a
10b
11a
All Inputs
Note:
VM = Midpoint Voltage (1.4V)
Figure 9. Mode Select Input Timing Diagram
HRST
VM
10c
11c
MODE Pins
Note:
VM = Midpoint Voltage (1.4V)
27
2102C–HIREL–01/05
Output AC Specifications
Table 18 provides the output AC timing specifications as shown in Figure 10.
Table 18. Output AC Timing Specifications (VDD = 3.3V ± 5% dc, GND = 0V dc, CL = 50 pF, -55°C ≤ Tj ≤ 125°C)
66 MHz
Ref
Characteristic
Min
12
SYSCLK to output driven (output enable
time)(9)
2.0
13a
SYSCLK to output valid (for TS and
ARTRY)(1, 2, 3, 4)
7.0
6.0
ns
13b
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)
7.0
6.0
ns
14a
SYSCLK to output valid (for RAS[0:7] and
CAS[0:7])(1, 2, 3)
7.0
6.0
ns
14b
SYSCLK to output valid
(for PCI signals)(3, 6)
11.0
11.0
ns
15a
SYSCLK to output invalid for all non-PCI
signals (output hold)(7, 10)
1.0
1.0
ns
15b
SYSCLK to output valid for PCI signals
(output hold)(7)
1.0
1.0
ns
18
SYSCLK to ARTRY high impedance
before precharge (output hold)(9)
19
SYSCLK to ARTRY precharge enable(8, 9)
21
SYSCLK to ARTRY high impedance after
precharge(8, 9)
Notes:
28
83.3 MHz
Max
Min
Max
Unit
2.0
8.0
(0.4 x tSYSCLK)
+ 2.0
ns
8.0
ns
(0.4 x tSYSCLK)
+ 2.0
(1.5 x tSYSCLK)
+ 8.0
ns
(1.5 x tSYSCLK)
+ 8.0
ns
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 TSPC106 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 numbers 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 Freescale application note “Designing PCI 2.1-compliant MPC106 Systems” (order number AN1727/D).
TSPC106A
2102C–HIREL–01/05
TSPC106A
Figure 10. Output Timing Diagram
VM
VM
VM
SYSCLK
14
15
12
16
All Outputs
(except TS and ARTRY)
15
13
13
16
TS
21
20
19
18
ARTRY
Note:
VM = Midpoint Voltage (1.4V)
Table 19. JTAG AC Timing Specifications (Independent of SYSCLK) (VDD = 3.3V ± 5% dc, GND = 0V dc,CL = 50 pF,
-55°C ≤ Tj ≤ 125°C)
Ref
Characteristic
Min
Max
Unit
TCK frequency of operation
0
25
MHz
1
TCK cycle time
40
ns
2
TCK clock pulse width measured at 1.4 V
20
ns
3
TCK rise and fall times
4
TRST setup time to TCK rising edge(1)
10
ns
5
TRST assert time
10
ns
5
ns
ns
0
(2)
3
ns
6
Boundary-scan input data setup time
7
Boundary-scan input data hold time(2)
15
8
TCK to output data valid(3)
0
30
ns
30
ns
(3)
9
TCK to output high impedance
0
10
TMS, TDI data setup time
5
ns
11
TMS, TDI data hold time
15
ns
12
TCK to TDO data valid
0
15
ns
13
TCK to TDO high impedance
0
15
ns
Notes:
1.
2.
3.
4.
These values are guaranteed by design, and are not tested.
TRST is an asynchronous signal. The setup time is for test purposes only.
Non-test signal input timing with respect to TCK.
Non-test signal output timing with respect to TCK.
29
2102C–HIREL–01/05
Figure 11. JTAG Clock Input Timing Diagram
1
2
VM
2
VM
VM
TCK
3
3
Figure 12. TRST Timing Diagram
TCK
4
TRST
5
Figure 13. Boundary-scan Timing Diagram
TCK
6
7
Input Data Valid
Data Inputs
8
Data Outputs
Output Data Valid
9
Data Outputs
8
Data Outputs
30
Output Data Valid
TSPC106A
2102C–HIREL–01/05
TSPC106A
Figure 14. Test Access Port Timing Diagram
TCK
10
11
Input Data Valid
TDI, TMS
12
Output Data Valid
TDO
13
TDO
12
Output Data Valid
TDO
Architectural
Overview
60x Processor Interface
The TSPC106 supports a programmable interface to a variety 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 TSPC106 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.
Two signals on the TSPC106, 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).
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
synchronous with all timing relative to the rising edge of the 60x bus clock.
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 TSPC106
•
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.
31
2102C–HIREL–01/05
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 processors. The 60x processors share all 60x interface signals of the TSPC106,
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
TSPC106 to arbitrate between the external cache and the 60x processor(s).
Memory Interface
The memory interface controls processor and PCI interactions 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
1-Gbyte of DRAM or EDO DRAM, with 16M bytes of ROM or Flash ROM. The memory
controller of the TSPC106 supports the various memory sizes through software initialization of on-chip configuration registers. Parity or ECC is provided for error detection.
The TSPC106 controls the 64-bit data path to main memory (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 TSPC106 features configurable data/parity
buffer control logic to accommodate several buffer types.
The TSPC106 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
TSPC106 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).
The TSPC106 provides parity checking and generation in two forms, normal parity and
read-modify-write (RMW) parity. As an alternative to simple parity, the TSPC106tspc106
supports error checking and correction (ECC) for system memory. Using ECC, the
TSPC106 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).
For ROM/Flash support, the TSPC106 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 connected to the most-significant byte lane of the data bus (DH[0:7]).
The TSPC106 also supports a mixed ROM system configuration. 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/memory bus.
32
TSPC106A
2102C–HIREL–01/05
TSPC106A
PCI Interface
The TSPC106’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.
The PCI interface of the TSPC106 functions as both a master and target device. As a
master, the TSPC106 supports read 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. Mode selectable big-endian to little-endian
conversion is supplied at the PCI interface.
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
memory operation has one 32-byte read buffer and two 32-byte write buffers.
System Design
Information
PLL Configuration
This section provides electrical and thermal design recommendations for successful
application of the TSPC106.
The TSPC106 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 PLL is configured by the PLL[0:3] signals. For a given SYSCLK (PCI bus) frequency, the clock mode configuration signals (PLL[0:3]) set the core frequency (and the
frequency of the VCO controlling the PLL lock). The supported core and VCO frequencies and the corresponding PLL[0:3] settings are provided in Table 19.
33
2102C–HIREL–01/05
Table 20. Core/VCO Frequencies and PLL Settings
Core Frequency (VCO Frequency) in MHz
PLL[0:3](1)
Core/SYSCL
Ratio
VCO Multiplier
0010
1:1
x8
0101
2:1
x4
Notes:
0110
5:2
(2)
x2
0111
5:2(2)
x4
1000
3:1
x2
1001
3:1
x4
PCI Bus
16.6 MHz
PCI Bus
20 MHz
PCI Bus
25 MHz
PCI Bus
33.3 MHz
33.3 (266)
40 (160)
50 (200)
66.6 (266)
83.3 (166)
41.6 (166)
50 (200)
62,5 (250)
83,3 (333)
75 (150)
60 (240)
75 (300)
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
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 26 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. Therefore this configuration should not be used and 1:1 modes between 16 and 25 MHz are not supported.
PLL Power Supply
Filtering
The AVDD power signal is provided on the 106 to provide power to the clock generation
phase-locked loop. To ensure stability of the internal clock, the power supplied to the
AVDD input signal should be filtered using a circuit similar to the one shown in Figure 15.
The circuit should be placed as close as possible to the AVDD pin to ensure it filters out
as much noise as possible.
Figure 15. PLL Power Supply Filter Circuit
10Ω
VDD
(3.3V)
10 µF
0.1 µF
AVDD
GND
34
TSPC106A
2102C–HIREL–01/05
TSPC106A
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
power supply, especially while driving large capacitive loads. This noise must be prevented from reaching other components in the system, and the TSPC106 itself requires
a clean, tightly regulated source of power.
It is strongly recommended that the system design include six to eight 0.1 µF (ceramic)
and 10 µF (tantalum) decoupling 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 underside 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.
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 to
the power and ground planes through two vias to minimize inductance. Suggested bulk
capacitors are 100 µF (AVX TPS tantalum) or 330 µF (AVX TPS tantalum).
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.
Power and ground connections must be made to all external 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 maintain 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 21, the factory test signal, LSSD_MODE, must be pulled up for normal device operation.
During inactive periods on the bus, the address and transfer 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.
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.
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 21 summarizes the pull-up/pull-down recommendations for the
TSPC106.
35
2102C–HIREL–01/05
Table 21. 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
60x bus address/transfer attributes
A[0:31], TT[0:4], TBST, WT, CI, GBL
Pull-up to 3.3V dc
Cache control
ADS
Pull-up to 3.3V dc
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
Note: For closed systems not requiring 5V
power, these may be pulled up to 3.3 VDC.
JTAG
TRST
Pull-down to GND (during normal system
operation)
Factory test
LSSD_MODE
Pull-up to 3.3V dc
Preparation for Delivery
Packaging
Microcircuits are prepared for delivery in accordance with MIL-PRF-38535.
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.
Handling
36
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
minimize the effect of this static buildup. However, the following handling practices are
recommended:
•
Devices should be handled on benches with conductive and grounded surfaces.
•
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% if practical.
TSPC106A
2102C–HIREL–01/05
TSPC106A
Package Mechanical Data
CBGA Package Parameters
Table 22. CBGA Package Parameters
Parameter
Min
Max
Parameter
Min
Max
Package outline
21 mm x 25 mm
A
25.0 ± 0.2
Interconnects
303 (16 x 19 ball array minus one)
B
21.0 ± 0.2
Pitch
1.27 mm
C
2.3
3.16
Solder attach
63/37 Sn/Pb
D
0.82
0.93
Solder balls
10/90 Sn/Pb, 0.89 mm diameter
G
Maximum module height
3.16 mm
H
Co-planarity specification
0.15 mm
K
1.27 BASIC
0.79
0.99
0.635 BASIC
N
5.8
6.0
P
7.2
7.4
Figure 16. CBGA Mechanical Drawing
2X
0.200
–F–
Br
–E–
A1
*Not to scale
Top view
–T–
P
A
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
K
J
H
G
F
E
D
C
B
A
Bottom View
G
G
K
C
H
303X Æ D
Æ 0.300 S
T E S
Æ 0.150 S
T
F
S
37
2102C–HIREL–01/05
CI_CGA Package
Parameters
Table 23. CI-CGA Package Parameters
Parameter
Min
Max
Parameter
Min
Max
Package outline
21 mm x 25 mm
A
25.0 BASIC
Interconnects
303 (16 x 19 ball array minus one)
B
21.0 BASIC
Pitch
1.27 mm
C
3.84 BASIC
Solder attach
63/37 Sn/Pb
D
Solder balls
10/90 Sn/Pb, 0.89 mm diameter
G
Maximum module height
3.84 mm
H
Co-planarity specification
0.15 mm
K
0.635 BASIC
N
6.2 BASIC
P
7.6 BASIC
0.79
0.99
1.27 BASIC
1.545
1.695
Figure 17. CI_CGA Mechanical Drawing
2X
0.200
–F–
Br
–E–
A1
*Not to scale
Top view
–T–
P
A
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
K
J
H
G
F
E
D
C
B
A
Bottom View
G
G
38
K
C
H
303X ∅ D
∅ 0.300 S
T E S
∅ 0.150 S
T
F
S
TSPC106A
2102C–HIREL–01/05
TSPC106A
Ordering Information
TS
PC106A
M
GS
B/Q
66
CG
Manufacturer's
Prefix
Revision Level(1)
CG: Rev. 4.0
Type
Operating Frequency(1)
66: 66 MHz
83: 83.3 MHz
Temperature Range: Tj
M: -55 C, +125 C
V: -40 C, +110 C
Package
G: CBGA
GS: CI_CGA
Note:
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
For availability of different versions, contact your Atmel sales office.
Definitions
Datasheet Status
Validity
Objective specification
This datasheet contains target and goal specifications for
discussion with the customer and application validation
Before design phase
Target specification
This datasheet contains target or goal specifications for
product development
Valid during the design phase
Preliminary specification α site
This datasheet contains preliminary data. Additional data
may be published at a later date and could include
simulation results
Valid before characterization
phase
Preliminary specification β site
This datasheet also contains characterization results
Valid before the
industrialization phase
Product specification
This datasheet contains final product specifications
Valid for production purpose
Limiting Values
Limiting values given are in accordance with the Absolute Maximum Rating System (IEC 134). Stresses above one or more of the
limiting values may cause permanent damage to the device. These are stress ratings only and operation of the device at these or at
any other conditions above those given in the Characteristics sections of the specification is not implied. Exposure to limiting values
for extended periods may affect device reliability.
Application Information
Where application information is given, it is advisory and does not form part of the specification
39
2102C–HIREL–01/05
Life Support
Applications
These products are not designed for use in life support appliances, devices, or systems
where malfunction of these products can reasonably be expected to result in personal
injury. Atmel customers using or selling these products for use in such applications do
so at their own risk and agree to fully indemnify Atmel for any damages resulting from
such improper use or sale.
Document Revision History
Table 19 provides a revision history for this hardware specification.
Table 24. Revision History
40
Revision Number
Date
2102C
11/2004
Substantive Change(s)
Preliminary specification β site changed to product qualification
Motorola changed to Freescale
TSPC106A
2102C–HIREL–01/05
Atmel Corporation
2325 Orchard Parkway
San Jose, CA 95131, USA
Tel: 1(408) 441-0311
Fax: 1(408) 487-2600
Regional Headquarters
Europe
Atmel Sarl
Route des Arsenaux 41
Case Postale 80
CH-1705 Fribourg
Switzerland
Tel: (41) 26-426-5555
Fax: (41) 26-426-5500
Asia
Room 1219
Chinachem Golden Plaza
77 Mody Road Tsimshatsui
East Kowloon
Hong Kong
Tel: (852) 2721-9778
Fax: (852) 2722-1369
Japan
9F, Tonetsu Shinkawa Bldg.
1-24-8 Shinkawa
Chuo-ku, Tokyo 104-0033
Japan
Tel: (81) 3-3523-3551
Fax: (81) 3-3523-7581
Atmel Operations
Memory
2325 Orchard Parkway
San Jose, CA 95131, USA
Tel: 1(408) 441-0311
Fax: 1(408) 436-4314
Microcontrollers
2325 Orchard Parkway
San Jose, CA 95131, USA
Tel: 1(408) 441-0311
Fax: 1(408) 436-4314
La Chantrerie
BP 70602
44306 Nantes Cedex 3, France
Tel: (33) 2-40-18-18-18
Fax: (33) 2-40-18-19-60
ASIC/ASSP/Smart Cards
RF/Automotive
Theresienstrasse 2
Postfach 3535
74025 Heilbronn, Germany
Tel: (49) 71-31-67-0
Fax: (49) 71-31-67-2340
1150 East Cheyenne Mtn. Blvd.
Colorado Springs, CO 80906, USA
Tel: 1(719) 576-3300
Fax: 1(719) 540-1759
Biometrics/Imaging/Hi-Rel MPU/
High Speed Converters/RF Datacom
Avenue de Rochepleine
BP 123
38521 Saint-Egreve Cedex, France
Tel: (33) 4-76-58-30-00
Fax: (33) 4-76-58-34-80
Zone Industrielle
13106 Rousset Cedex, France
Tel: (33) 4-42-53-60-00
Fax: (33) 4-42-53-60-01
1150 East Cheyenne Mtn. Blvd.
Colorado Springs, CO 80906, USA
Tel: 1(719) 576-3300
Fax: 1(719) 540-1759
Scottish Enterprise Technology Park
Maxwell Building
East Kilbride G75 0QR, Scotland
Tel: (44) 1355-803-000
Fax: (44) 1355-242-743
Literature Requests
www.atmel.com/literature
Disclaimer: The information in this document is provided in connection with Atmel products. No license, express or implied, by estoppel or otherwise, to any
intellectual property right is granted by this document or in connection with the sale of Atmel products. EXCEPT AS SET FORTH IN ATMEL’S TERMS AND CONDITIONS OF SALE LOCATED ON ATMEL’S WEB SITE, ATMEL ASSUMES NO LIABILITY WHATSOEVER AND DISCLAIMS ANY EXPRESS, IMPLIED OR STATUTORY
WARRANTY RELATING TO ITS PRODUCTS INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR
PURPOSE, OR NON-INFRINGEMENT. IN NO EVENT SHALL ATMEL BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE, SPECIAL OR INCIDENTAL DAMAGES (INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS OF PROFITS, BUSINESS INTERRUPTION, OR LOSS OF INFORMATION) ARISING OUT
OF THE USE OR INABILITY TO USE THIS DOCUMENT, EVEN IF ATMEL HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Atmel makes no
representations or warranties with respect to the accuracy or completeness of the contents of this document and reserves the right to make changes to specifications
and product descriptions at any time without notice. Atmel does not make any commitment to update the information contained herein. Atmel’s products are not
intended, authorized, or warranted for use as components in applications intended to support or sustain life.
© Atmel Corporation 2005. All rights reserved. Atmel ® and combinations thereof are the registered trademarks of Atmel Corporation or its
subsidiaries. PowerPC ® is the registered trademark of IBM Corp. Other terms and product names may be the trademarks of others.
Printed on recycled paper.
2102C–HIREL–01/05