CYP15G0403DXB Independent Clock Quad HOTLink II Transceiver Datasheet.pdf

CYP15G0403DXB
Independent Clock Quad HOTLink II™
Transceiver
Features
■
Second-generation HOTLink® technology
■
Compliant to multiple standards
❐ ESCON, DVB-ASI, Fibre Channel and Gigabit Ethernet
(IEEE802.3z)
❐ CPRI™ compliant
❐ 8B/10B coded data or 10 bit uncoded data
■
Quad channel transceiver operates from 195 to 1500 MBaud
serial data rate
❐ Aggregate throughput of up to 12 Gbits/second
■
Second-generation HOTLink technology
■
Truly independent channels
❐ Each channel can operate at a different signaling rate
❐ Each channel can transport a different type of data
■
Selectable input/output clocking options
■
Internal phase-locked loops (PLLs) with no external PLL
components
■
Dual differential PECL-compatible serial inputs per channel
■
Internal DC-restoration
■
Dual differential PECL-compatible serial outputs per channel
❐ Source matched for 50-Ω transmission lines
❐ No external bias resistors required
❐ Signaling-rate controlled edge-rates
■
MultiFrame™ Receive Framer provides alignment options
❐ Bit and byte alignment
❐ Comma or Full K28.5 detect
❐ Single or Multi-byte Framer for byte alignment
❐ Low-latency option
■
Synchronous LVTTL parallel interface
■
JTAG boundary scan
■
Built-In Self-Test (BIST) for at-speed link testing
■
Compatible with
❐ Fiber-optic modules
❐ Copper cables
❐ Circuit board traces
■
Per-channel Link Quality Indicator
❐ Analog signal detect
❐ Digital signal detect
■
Low-power 3W at 3.3-V typical
■
Single 3.3-V supply
■
256-ball thermally enhanced BGA
■
Pb-free package option available
■
0.25 μ BiCMOS technology
Functional Description
The CYP15G0403DXB Independent Clock Quad
HOTLink II™ Transceiver is a point-to-point or point-to-multipoint communications building block enabling transfer of data
over a variety of high-speed serial links like optical fiber,
balanced, and unbalanced copper transmission lines. The
signaling rate can be anywhere in the range of 195 to
1500 MBaud per serial link. Each channel operates independently with its own reference clock allowing different rates.
Each transmit channel accepts parallel characters in an Input
Register, encodes each character for transport, and then
converts it to serial data. Each receive channel accepts serial
data and converts it to parallel data, decodes the data into
characters, and presents these characters to an Output
Register. Figure 1 on page 2 illustrates typical connections
between independent host systems and corresponding
CYP15G0403DXB chips
As a second-generation HOTLink device, the
CYP15G0403DXB extends the HOTLink family with enhanced
levels of integration and faster data rates, while maintaining
serial-link compatibility (data, command, and BIST) with other
HOTLink devices. The transmit (TX) section of the
CYP15G0403DXB Quad HOTLink II consists of four
independent byte-wide channels. Each channel can accept
either 8-bit data characters or preencoded 10-bit transmission
characters. Data characters may be passed from the Transmit
Input Register to an integrated 8B/10B Encoder to improve
their serial transmission characteristics. These encoded
characters are then serialized and output from dual Positive
ECL (PECL) compatible differential transmission-line drivers
at a bit-rate of either 10 or 20 times the input reference clock
for that channel.
.
Cypress Semiconductor Corporation
Document #: 38-02065 Rev. *J
•
198 Champion Court
•
San Jose, CA 95134-1709
•
408-943-2600
Revised June 9, 2014
CYP15G0403DXB
Figure 1. HOTLink II™ System Connections
10
Serial Links
10
10
10
System Host
Serial Links
10
10
Independent
Independent
CYP15G0403DXB
10
10
CYP15G0403DXB
Serial Links
10
10
Backplane or
Cabled
Connections
10
Serial Links
System Host
10
10
10
10
10
The receive (RX) section of the CYP15G0403DXB Quad
HOTLink II consists of four independent byte-wide channels.
Each channel accepts a serial bit-stream from one of two
PECL-compatible differential line receivers, and using a
completely integrated Clock and Data Recovery PLL, recovers
the timing information necessary for data reconstruction. Each
recovered bit-stream is deserialized and framed into characters,
8B/10B decoded, and checked for transmission errors.
Recovered decoded characters are then written to an internal
Elasticity Buffer, and presented to the destination host system.
The integrated 8B/10B encoder/decoder may be bypassed for
systems that present externally encoded or scrambled data at
the parallel interface.
tecture. In addition to clocking the transmit path with a local
reference clock, the receive interface may also be configured to
present data relative to a recovered clock or to a local reference
clock.
Each transmit and receive channel contains an independent
BIST pattern generator and checker. This BIST hardware allows
at-speed testing of the high-speed serial data paths in each
transmit and receive section, and across the interconnecting
links.
The CYP15G0403DXB is ideal for port applications where
different data rates and serial interface standards are necessary
for each channel. Some applications include multi-protocol
routers, aggregation equipment, and switches.
The parallel I/O interface may be configured for numerous forms
of clocking to provide the highest flexibility in system archi-
Document #: 38-02065 Rev. *J
Page 2 of 49
CYP15G0403DXB
Contents
CYP15G0403DXB Transceiver Logic Block Diagram.... 4
Transmit Path Block Diagram ......................................... 5
Receive Path Block Diagram ........................................... 6
Device Configuration and Control Block Diagram ........ 7
Pin Configuration (Top View) .......................................... 8
Pin Configuration (Bottom View) .................................... 9
Pin Descriptions
CYP15G0403DXB Quad HOTLink II Transceiver ......... 10
CYP15G0403DXB HOTLink II Operation ....................... 15
CYP15G0403DXB Transmit Data Path ..................... 15
Transmit Modes ......................................................... 16
Transmit BIST ................................................................. 16
Transmit PLL Clock Multiplier .................................... 16
Serial Output Drivers ...................................................... 17
CYP15G0403DXB Receive Data Path ............................ 17
Serial Line Receivers ................................................ 17
Signal Detect/Link Fault ............................................ 17
Clock/Data Recovery ................................................. 18
Deserializer/Framer ................................................... 18
10B/8B Decoder Block .............................................. 19
Receive BIST Operation ............................................ 19
Receive Elasticity Buffer ............................................ 20
Receive Modes .......................................................... 20
Power Control ............................................................ 20
Output Bus ...................................................................... 20
Device Configuration and Control Interface .............. 21
JTAG Support ................................................................. 26
Maximum Ratings ........................................................... 29
Power-up Requirements ............................................ 29
Operating Range ............................................................. 29
DC Electrical Characteristics ........................................ 29
AC Test Loads and Waveforms ..................................... 30
Document #: 38-02065 Rev. *J
AC Electrical Characteristics ........................................ 31
Transmitter LVTTL Switching Characteristics ........... 31
Receiver LVTTL Switching Characteristics ............... 31
REFCLKx Switching Characteristics ......................... 31
Bus Configuration Write Timing Characteristics ....... 32
JTAG Test Clock Characteristics ............................. 32
Device RESET Characteristics ................................. 32
Transmit Serial Outputs and TX PLL Characteristics 32
Receive Serial Inputs and CDR PLL Characteristics 33
Capacitance .................................................................... 33
HOTLink II Transmitter Switching Waveforms ............ 34
Switching Waveforms for the HOTLink II Receiver ..... 35
X3.230 Codes and Notation Conventions .................... 39
Notation Conventions ................................................ 39
8B/10B Transmission Code ....................................... 39
Transmission Order ................................................... 39
Valid and Invalid Transmission Characters ............... 39
Use of the Tables for Generating Transmission
Characters ................................................................. 40
Use of the Tables for Checking the Validity of Received
Transmission Characters .......................................... 40
Ordering Information ...................................................... 46
Ordering Code Definition ............................................... 46
Package Diagram ............................................................ 46
Document History Page ................................................. 47
Sales, Solutions, and Legal Information ...................... 48
Worldwide Sales and Design Support ....................... 48
Products .................................................................... 48
PSoC Solutions ......................................................... 48
Page 3 of 49
CYP15G0403DXB
TXDD[7:0]
TXCTD[1:0]
x11
x10
x11
Phase
Align
Buffer
Elasticity
Buffer
Phase
Align
Buffer
Elasticity
Buffer
Phase
Align
Buffer
Elasticity
Buffer
Phase
Align
Buffer
Elasticity
Buffer
Encoder
8B/10B
Decoder
8B/10B
Encoder
8B/10B
Decoder
8B/10B
Encoder
8B/10B
Decoder
8B/10B
Encoder
8B/10B
Decoder
8B/10B
TX
RX
TX
RX
TX
RX
INB1±
INB2±
OUTC1±
OUTC2±
INC1±
INC2±
Document #: 38-02065 Rev. *J
Serializer
TX
Deserializer
RX
IND1±
IND2±
Deserializer
OUTB1±
OUTB2±
Deserializer
Serializer
INA1±
INA2±
Serializer
Deserializer
OUTA1±
OUTA2±
Serializer
Framer
Framer
OUTD1±
OUTD2±
Framer
Framer
RXDD[7:0]
RXSTD[2:0]
RXDC[7:0]
RXSTC[2:0]
x10
REFCLKD±
TXDC[7:0]
TXCTC[1:0]
x11
REFCLKC±
RXDB[7:0]
RXSTB[2:0]
x10
REFCLKB±
TXDB[7:0]
TXCTB[1:0]
x11
REFCLKA±
x10
TXDA[7:0]
TXCTA[1:0]
RXDA[7:0]
RXSTA[2:0]
CYP15G0403DXB Transceiver Logic Block Diagram
Page 4 of 49
CYP15G0403DXB
Transmit Path Block Diagram
TXLB[A..D] are Internal Serial Loopback Signals
REFCLKA+
= Internal Signal
Bit-Rate Clock
REFCLKA–
Transmit
Transmit PLL
PLL
Clock Multiplier
A
Clock
Multiplier
TXRATEA
SPDSELA
TXCLKOA
OEA[2..1]
ENCBYPA
Character-Rate Clock A
TXERRA
TXCLKA
10
10
OUTA1+
OUTA1–
Shifter
2
TXCTA[1:0]
10
BIST
BIST LFSR
LFSR
Input
Register
8
TXDA[7:0]
Encoder
8B/10B
Encoder
1
Phase-Align
Phase-Align
Buffer
Buffer
0
TXCKSELA
OEA[2..1]
TXBISTA
PABRSTA
10
OUTA2+
OUTA2–
TXLBA
REFCLKB+
Bit-Rate Clock
REFCLKB–
Transmit PLL
Clock Multiplier B
TXRATEB
SPDSELB
TXCLKOB
OEB[2..1]
Character-Rate Clock B
ENCBYPB
TXERRB
0
10
10
OUTB1+
OUTB1–
Shifter
2
10
BIST
BIST LFSR
LFSR
TXDB[7:0]
10
Encoder
8B/10B
Encoder
Input
Register
8
OEB[2..1]
1
Phase-Align
Phase-Align
Buffer
Buffer
TXCKSELB
TXCTB[1:0]
TXBISTB
PABRSTB
TXCLKB
OUTB2+
OUTB2–
TXLBB
REFCLKC+
Bit-Rate Clock
REFCLKC–
Transmit PLL
Clock Multiplier C
TXRATEC
SPDSELC
TXCLKOC
OEC[2..1]
ENCBYPC
Character-Rate Clock C
TXERRC
10
10
10
OUTC1+
OUTC1–
Shifter
2
TXCTC[1:0]
10
BIST LFSR
Input
Register
8
TXDC[7:0]
8B/10B
Encoder
1
Phase-Align
Buffer
0
TXCKSELC
OEC[2..1]
TXBISTC
PABRSTC
TXCLKC
OUTC2+
OUTC2–
TXLBC
REFCLKD+
Bit-Rate Clock
REFCLKD–
Transmit PLL
Clock Multiplier D
TXRATED
OED[2..1]
SPDSELD
ENCBYPD
TXCLKOD
OED[2..1]
Character-Rate Clock D
TXERRD
Document #: 38-02065 Rev. *J
10
10
OUTD1+
OUTD1–
Shifter
10
BIST LFSR
2
10
8B/10B
Encoder
TXDD[7:0]
1
Phase-Align
Buffer
8
0
Input
Register
TXCKSELD
TXCTD[1:0]
TXBISTD
PABRSTD
TXCLKD
OUTD2+
OUTD2–
TXLBD
Page 5 of 49
CYP15G0403DXB
Receive Path Block Diagram
TXLB[A..D] are Internal Serial Loopback Signals
= Internal Signal
RESET
TRST
JTAG
Boundary
Scan
Controller
SPDSELA
RXPLLPDA
SDASELA[1:0]
TMS
TCLK
TDI
TDO
Receive
Signal
Monitor
Elasticity
Buffer
Output
Register
TXLBA
ULCA
Clock &
Data
Recovery
PLL
10B/8B
BIST
INA2+
INA2–
LFIA
Framer
INA1+
INA1–
Shifter
LPENA
INSELA
8
3
RXDA[7:0]
RXSTA[2:0]
SPDSELB
Clock
Select
RXPLLPDB
SDASELB[1:0]
Receive
Signal
Monitor
SPDSELC
Receive
Signal
Monitor
TXLBC
ULCC
SPDSELD
Clock &
Data
Recovery
PLL
Clock
Select
RXPLLPDD
SDASELD[1:0]
Receive
Signal
Monitor
IND2+
IND2–
TXLBD
ULCD
Clock &
Data
Recovery
PLL
LDTDEN
RFMODE[A..D][1:0]
RFEN[A..D]
FRAMCHAR[A..D]
DECMODE[A..D]
RXBIST[A..D]
RXCKSEL[A..D]
DECBYP[A..D]
RXRATE[A..D]
Document #: 38-02065 Rev. *J
Framer
IND1+
IND1–
Output
Register
RXDB[7:0]
RXSTB[2:0]
RXCLKB+
RXCLKB–
8
3
÷2
RXDC[7:0]
RXSTC[2:0]
RXCLKC+
RXCLKC–
LFID
Shifter
INSELD
10B/8B
BIST
LPEND
3
÷2
Output
Register
INC2+
INC2–
Framer
INC1+
INC1–
8
LFIC
Shifter
INSELC
10B/8B
BIST
LPENC
Elasticity
Buffer
Clock
Select
RXPLLPDC
SDASELC[1:0]
Clock
Select
Output
Register
ULCB
Elasticity
Buffer
TXLBB
Clock &
Data
Recovery
PLL
Elasticity
Buffer
INB2+
INB2–
10B/8B
BIST
INB1+
INB1–
RXCLKA+
RXCLKA–
LFIB
Framer
INSELB
Shifter
LPENB
÷2
÷2
8
3
RXDD[7:0]
RXSTD[2:0]
RXCLKD+
RXCLKD–
Page 6 of 49
CYP15G0403DXB
Device Configuration and Control Block Diagram
WREN
ADDR[3:0]
Device Configuration
and Control Interface
DATA[7:0]
Document #: 38-02065 Rev. *J
= Internal Signal
RFMODE[A..D][1:0]
RFEN[A..D]
FRAMCHAR[A..D]
DECMODE[A..D]
RXBIST[A..D]
RXCKSEL[A..D]
DECBYP[A..D]
RXRATE[A..D]
SDASEL[2..1][A..D][1:0]
RXPLLPD[A..D]
TXRATE[A..D]
TXCKSEL[A..D]
PABRST[A..D]
TXBIST[A..D]
OE[2..1][A..D]
ENCBYP[A..D]
GLEN[11..0]
FGLEN[2..0]
Page 7 of 49
CYP15G0403DXB
Pin Configuration (Top View)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
A
IN
C1–
OUT
C1–
IN
C2–
OUT
C2–
VCC
IN
D1–
OUT
D1–
GND
IN
D2–
OUT
D2–
IN
A1–
OUT
A1–
GND
IN
A2–
OUT
A2–
VCC
IN
B1–
OUT
B1–
IN
B2–
OUT
B2–
B
IN
C1+
OUT
C1+
IN
C2+
OUT
C2+
VCC
IN
D1+
OUT
D1+
GND
IN
D2+
OUT
D2+
IN
A1+
OUT
A1+
GND
IN
A2+
OUT
A2+
VCC
IN
B1+
OUT
B1+
IN
B2+
OUT
B2+
C
TDI
TMS
INSELC INSELB
VCC
ULCD
ULCC
GND
DATA
[7]
DATA
[5]
DATA
[3]
DATA
[1]
GND
NC
SPD
SELD
VCC
LDTD
EN
TRST
LPEND
TDO
D
TCLK
RESET INSELD INSELA
VCC
ULCA
SPD
SELC
GND
DATA
[6]
DATA
[4]
DATA
[2]
DATA
[0]
GND
LPENB
ULCB
VCC
LPENA LTEN1
SCAN
EN2
TMEN3
E
VCC
VCC
VCC
VCC
VCC
VCC
VCC
F
RX
DC[6]
RX
DC[7]
TX
DC[0]
NC
NC
G
TX
DC[7]
WREN
TX
DC[4]
TX
DC[1]
SPD
SELB
LP
ENC
SPD
SELA
RX
DB[1]
H
GND
GND
GND
GND
GND
GND
GND
GND
J
TX
CTC[1]
TX
DC[5]
TX
DC[2]
TX
DC[3]
RX
STB[2]
RX
DB[0]
RX
DB[5]
RX
DB[2]
K
RX
DC[2]
REF
TX
CLKC– CTC[0]
TX
CLKC
RX
DB[3]
RX
DB[4]
RX
DB[7]
LFIB
L
RX
DC[3]
REF
CLKC+
LFIC
TX
DC[6]
RX
DB[6]
M
RX
DC[4]
RX
DC[5]
NC
TX
ERRC
REF
REF
CLKB+ CLKB–
TX
ERRB
TX
CLKB
N
GND
GND
GND
GND
GND
GND
GND
GND
P
RX
DC[1]
RX
DC[0]
RX
STC[0]
RX
STC[1]
TX
DB[5]
TX
DB[4]
TX
DB[3]
TX
DB[2]
R
RX
TX
RX
RX
STC[2] CLKOC CLKC+ CLKC–
TX
DB[1]
TX
DB[0]
TX
CTB[1]
TX
DB[7]
VCC
VCC
VCC
VCC
T
VCC
VCC
VCC
VCC
U
TX
DD[0]
TX
DD[1]
TX
DD[2]
TX
CTD[1]
VCC
RX
DD[2]
RX
DD[1]
GND
TX
CTA[1]
ADDR
[0]
REF
CLKD–
V
TX
DD[3]
TX
DD[4]
TX
CTD[0]
RX
DD[6]
VCC
RX
DD[3]
RX
STD[0]
GND
RX
STD[2]
ADDR
[2]
W
TX
DD[5]
TX
DD[7]
LFID
RX
CLKD–
VCC
RX
DD[4]
RX
STD[1]
GND
ADDR
[3]
Y
TX
DD[6]
TX
CLKD
RX
DD[7]
RX
CLKD+
VCC
RX
DD[5]
RX
DD[0]
GND
TX
CLKOD
Document #: 38-02065 Rev. *J
VCC
RX
TX
RX
STB[1] CLKOB STB[0]
RX
RX
CLKB+ CLKB–
TX
DB[6]
GND
TX
DA[4]
TX
CTA[0]
VCC
RX
DA[2]
TX
CTB[0]
RX
STA[2]
RX
STA[1]
REF
TX
CLKD+ CLKOA
GND
TX
DA[3]
TX
DA[7]
VCC
RX
DA[7]
RX
DA[3]
RX
DA[0]
RX
STA[0]
ADDR
[1]
RX
CLKA+
TX
ERRA
GND
TX
DA[2]
TX
DA[6]
VCC
LFIA
REF
CLKA+
RX
DA[4]
RX
DA[1]
NC
TX
CLKA
RX
CLKA–
GND
TX
DA[0]
TX
DA[5]
VCC
TX
ERRD
REF
CLKA–
RX
DA[6]
RX
DA[5]
TX
DA[1]
Page 8 of 49
CYP15G0403DXB
Pin Configuration (Bottom View)
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
A
OUT
B2–
IN
B2–
OUT
B1–
IN
B1–
VCC
OUT
A2–
IN
A2–
GND
OUT
A1–
IN
A1–
OUT
D2–
IN
D2–
GND
OUT
D1–
IN
D1–
VCC
OUT
C2–
IN
C2–
OUT
C1–
IN
C1–
B
OUT
B2+
IN
B2+
OUT
B1+
IN
B1+
VCC
OUT
A2+
IN
A2+
GND
OUT
A1+
IN
A1+
OUT
D2+
IN
D2+
GND
OUT
D1+
IN
D1+
VCC
OUT
C2+
IN
2+
OUT
C1+
IN
C1+
C
TDO
LP
END
TRST
LDTD
EN
VCC
SPD
SELD
NC
GND
DATA
[1]
DATA
[3]
DATA
[5]
DATA
[7]
GND
ULCC
ULCD
VCC
IN
SELB
IN
SELC
TMS
TDI
D
TMEN3
SCAN
EN2
LTEN1
LP
ENA
VCC
ULCB
LP
ENB
GND
DATA
[0]
DATA
[2]
DATA
[4]
DATA
[6]
GND
SPD
SELC
ULCA
VCC
IN
SELA
IN
SELD
RESET
TCLK
E
VCC
VCC
VCC
VCC
VCC
VCC
VCC
VCC
NC
NC
TX
DC[0]
RX
DC[7]
Rx
DC[6]
F
RX
TX
RX
STB[0] CLKOB STB[1]
G
RX
DB[1]
SPD
SELA
LP
ENC
SPD
SELB
TX
DC[1]
TX
DC[4]
WREN
TX
DC[7]
H
GND
GND
GND
GND
GND
GND
GND
GND
J
RX
DB[2]
RX
DB[5]
RX
DB[0]
RX
STB[2]
TX
DC[3]
TX
DC[2]
TX
DC[5]
TX
CTC[1]
K
LFIB
RX
DB[7]
RX
DB[4]
RX
DB[3]
TX
CLKC
TX
REF
CTC[0] CLKC–
L
TX
DB[6]
RX
RX
CLKB– CLKB+
RX
DB[6]
TX
DC[6]
LFIC
REF
CLKC+
RX
DC[3]
M
TX
CLKB
TX
ERRB
REF
REF
CLKB– CLKB+
TX
ERRC
NC
RX
DC[5]
RX
DC[4]
N
GND
GND
GND
GND
GND
GND
GND
GND
P
TX
DB[2]
TX
DB[3]
TX
DB[4]
TX
DB[5]
RX
STC[1]
RX
STC[0]
RX
DC[0]
RX
DC[1]
R
TX
DB[7]
TX
CTB[1]
TX
DB[0]
TX
DB[1]
RX
RX
TX
RX
CLKC– CLKC+ CLKOC STC[2]
T
VCC
VCC
VCC
VCC
U
RX
STA[1]
RX
STA[2]
TX
CTB[0]
RX
DA[2]
VCC
TX
CTA[0]
TX
DA[4]
GND
V
RX
STA[0]
RX
DA[0]
RX
DA[3]
RX
DA[7]
VCC
TX
DA[7]
TX
DA[3]
GND
W
RX
DA[1]
RX
DA[4]
REF
CLKA+
LFIA
VCC
TX
DA[6]
TX
DA[2]
GND
TX
ERRA
Y
RX
DA[5]
RX
DA[6]
REF
CLKA–
TX
ERRD
VCC
TX
DA[5]
TX
DA[0]
GND
RX
CLKA–
Document #: 38-02065 Rev. *J
TX
DA[1]
RX
DC[2]
VCC
VCC
VCC
VCC
REF
CLKD–
ADDR
[0]
TXC
TA[1]
GND
RX
DD[1]
RX
DD[2]
VCC
TX
CTD[1]
TX
DD[2]
TX
DD[1]
TX
DD[0]
TX
REF
CLKOA CLKD+
ADDR
[2]
RX
STD[2]
GND
RX
STD[0]
RX
DD[3]
VCC
RX
DD[6]
TX
CTD[0]
TX
DD[4]
TX
DD[3]
RX
CLKA+
ADDR
[1]
ADDR
[3]
GND
RX
STD[1]
RX
DD[4]
VCC
RX
CLKD–
LFID
TX
DD[7]
TX
DD[5]
TX
CLKA
NC
TX
CLKOD
GND
RX
DD[0]
RX
DD[5]
VCC
RX
CLKD+
RX
DD[7]
TX
CLKD
TX
DD[6]
Page 9 of 49
CYP15G0403DXB
Pin Descriptions
CYP15G0403DXB Quad HOTLink II Transceiver
Name
I/O Characteristics
Signal Description
Transmit Path Data and Status Signals
TXDA[7:0]
TXDB[7:0]
TXDC[7:0]
TXDD[7:0]
LVTTL Input,
synchronous,
sampled by the
associated
TXCLKx↑ or
REFCLKx↑[1]
Transmit Data Inputs. TXDx[7:0] data inputs are captured on the rising edge of the
transmit interface clock. The transmit interface clock is selected by the TXCKSELx
latch via the device configuration interface, and passed to the encoder or Transmit
Shifter. When the Encoder is enabled, TXDx[7:0] specifies the specific data or
command character sent.
TXCTA[1:0]
TXCTB[1:0]
TXCTC[1:0]
TXCTD[1:0]
LVTTL Input,
synchronous,
sampled by the
associated
TXCLKx↑ or
REFCLKx↑ [1]
Transmit Control. TXCTx[1:0] inputs are captured on the rising edge of the transmit
interface clock. The transmit interface clock is selected by the TXCKSELx latch via
the device configuration interface, and passed to the Encoder or Transmit Shifter.
The TXCTA[1:0] inputs identify how the associated TXDx[7:0] characters are interpreted. When the Encoder is bypassed, these inputs are interpreted as data bits.
When the Encoder is enabled, these inputs determine if the TXDx[7:0] character is
encoded as Data, a Special Character code, or replaced with other Special
Character codes. See Table 3 on page 16 for details.
TXERRA
TXERRB
TXERRC
TXERRD
LVTTL Output,
synchronous to
REFCLKx↑ [2],
synchronous to
RXCLKx when
selected as
REFCLKx,
asynchronous to
transmit channel
enable / disable,
asynchronous to loss
or return of
REFCLKx±
Transmit Path Error. TXERRx is asserted HIGH to indicate detection of a transmit
Phase-Align Buffer underflow or overflow. If an underflow or overflow condition is
detected, TXERRx, for the channel in error, is asserted HIGH and remains asserted
until either a Word Sync Sequence is transmitted on that channel, or the transmit
Phase-Align Buffer is re-centered with the PABRSTx latch via the device configuration interface. When TXBISTx = 0, the BIST progress is presented on the
associated TXERRx output. The TXERRx signal pulses HIGH for one
transmit-character clock period to indicate a pass through the BIST sequence once
every 511 or 527 (depending on RXCKSELx) character times. If RXCKSELx = 1, a
one character pulse occurs every 527 character times. If RXCKSELx = 0, a one
character pulse occurs every 511 character times.
TXERRx is also asserted HIGH, when any of the following conditions is true:
■
The TXPLL for the associated channel is powered down. This occurs when OE2x
and OE1x for a given channel are both disabled by setting OE2x = 0 and OE1x = 0.
■
The absence of the REFCLKx± signal
Transmit Path Clock Signals
REFCLKA±
REFCLKB±
REFCLKC±
REFCLKD±
Differential LVPECL
or single-ended
LVTTL input clock
Reference Clock. REFCLKx± clock inputs are used as the timing references for the
transmit and receive PLLs. These input clocks may also be selected to clock the
transmit and receive parallel interfaces. When driven by a single-ended LVCMOS
or LVTTL clock source, connect the clock source to either the true or complement
REFCLKx input, and leave the alternate REFCLKx input open (floating). When
driven by an LVPECL clock source, the clock must be a differential clock, using both
inputs.
TXCLKA
TXCLKB
TXCLKC
TXCLKD
LVTTL Clock Input,
internal pull-down
Transmit Path Input Clock. When configuration latch TXCKSELx = 0, the
associated TXCLKx input is selected as the character-rate input clock for the
TXDx[7:0] and TXCTx[1:0] inputs. In this mode, the TXCLKx input must be
frequency-coherent to its associated TXCLKOx output clock, but may be offset in
phase by any amount. Once initialized, TXCLKx is allowed to drift in phase as much
as ±180 degrees. If the input phase of TXCLKx drifts beyond the handling capacity
of the Phase Align Buffer, TXERRx is asserted to indicate the loss of data, and
remains asserted until the Phase Align Buffer is initialized. The phase of the TXCLKx
input clock relative to its associated REFCLKx± is initialized when the configuration
latch PABRSTx is written as 0. When the associated TXERRx is deasserted, the
Phase Align Buffer is initialized and input characters are correctly captured.
Notes
1. When REFCLKx± is configured for half-rate operation, these inputs are sampled relative to both the rising and falling edges of the associated REFCLKx±.
2. When REFCLKx± is configured for half-rate operation, these outputs are presented relative to both the rising and falling edges of the associated REFCLKx±.
Document #: 38-02065 Rev. *J
Page 10 of 49
CYP15G0403DXB
Pin Descriptions (continued)
CYP15G0403DXB Quad HOTLink II Transceiver
Name
TXCLKOA
TXCLKOB
TXCLKOC
TXCLKOD
I/O Characteristics
LVTTL Output
Signal Description
Transmit Clock Output. TXCLKOx output clock is synthesized by each channel’s
transmit PLL and operates synchronous to the internal transmit character clock.
TXCLKOx operates at either the same frequency as REFCLKx± (TXRATEx = 0), or
at twice the frequency of REFCLKx± (TXRATEx = 1). The transmit clock outputs
have no fixed phase relationship to REFCLKx±.
Receive Path Data and Status Signals
RXDA[7:0]
RXDB[7:0]
RXDC[7:0]
RXDD[7:0]
LVTTL Output,
synchronous to the
selected RXCLK±
output or REFCLKx±
input
Parallel Data Output. RXDx[7:0] parallel data outputs change relative to the receive
interface clock. The receive interface clock is selected by the RXCKSELx latch. If
RXCLKx± is a full-rate clock, the RXCLKx± clock outputs are complementary clocks
operating at the character rate. The RXDx[7:0] outputs for the associated receive
channels follow rising edge of RXCLKx+ or falling edge of RXCLKx–. If RXCLKx±
is a half-rate clock, the RXCLKx± clock outputs are complementary clocks operating
at half the character rate. The RXDx[7:0] outputs for the associated receive
channels follow both the falling and rising edges of the associated RXCLKx± clock
outputs.
RXSTA[2:0]
RXSTB[2:0]
RXSTC[2:0]
RXSTD[2:0]
LVTTL Output,
synchronous to the
selected RXCLK±
output or REFCLKx±
input
Parallel Status Output. RXSTA[2:0] status outputs change relative to the receive
interface clock. The receive interface clock is selected by the RXCKSELx latch. If
RXCLKx± is a full-rate clock, the RXCLKx± clock outputs are complementary clocks
operating at the character rate. The RXSTAx[2:0] outputs for the associated receive
channels follow rising edge of RXCLKx+ or falling edge of RXCLKx–. If RXCLKx±
is a half-rate clock, the RXCLKx± clock outputs are complementary clocks operating
at half the character rate. The RXSTAx[2:0] outputs for the associated receive
channels follow both the falling and rising edges of the associated RXCLKx± clock
outputs. When the decoder is bypassed, RXSTx[1:0] become the two low-order bits
of the 10-bit received character. RXSTx[2] = HIGH indicates the presence of a
Comma character in the Output Register. When the decoder is enabled, RXSTx[2:0]
provide status of the received signal. See Table 11 on page 28 for a list of received
character status.
Receive Path Clock Signals
RXCLKA±
RXCLKB±
RXCLKC±
RXCLKD±
LVTTL Output Clock
Receive Clock Output. RXCLKx± is the receive interface clock used to control
timing of the RXDx[7:0] and RXSTA[2:0] parallel outputs. The source of the
RXCLKx± outputs is selected by the RXCKSELx latch via the device configuration
interface. These true and complement clocks are used to control timing of data
output transfers. These clocks are output continuously at either the dual-character
rate (1/20th the serial bit-rate) or character rate (1/10th the serial bit-rate) of the data
being received, as selected by RXRATEx. When configured such that the output
data path is clocked by the REFCLKx± instead of a recovered clock, the RXCLKx±
output drivers present a buffered or divided form (depending on RXRATEx) of the
associated REFCLKx± that are delayed in phase to align with the data. This phase
difference allows the user to select the optimal clock (REFCLKx± or RXCLK±) for
setup/hold timing for their specific system.
When REFCLKx± is a full-rate clock, the RXCLKx± rate depends on the value of
RXRATEx.
When REFCLKx± is a half-rate clock and RXCKSELx = 0, the RXCLKx± rate
depends on the value of RXRATEx.
When REFCLKx± is a half-rate clock and RXCKSELx=1, the RXCLKx± rate does
not depend on the value of RXRATEx and operates at the same rate as REFCLKx±.
Document #: 38-02065 Rev. *J
Page 11 of 49
CYP15G0403DXB
Pin Descriptions (continued)
CYP15G0403DXB Quad HOTLink II Transceiver
Name
I/O Characteristics
Signal Description
Device Control Signals
RESET
LVTTL Input,
asynchronous,
internal pull-up
Asynchronous Device Reset. RESET initializes all state machines, counters, and
configuration latches in the device to a known state. RESET must be asserted LOW
for a minimum pulse width. When the reset is removed, all state machines, counters
and configuration latches are at an initial state. As per the JTAG specifications the
device RESET cannot reset the JTAG controller. Therefore, the JTAG controller has
to be reset separately. Refer to “JTAG Support” on page 27 for the methods to reset
the JTAG state machine. See Table 9 on page 22 for the initialize values of the
device configuration latches.
LDTDEN
LVTTL Input,
internal pull-up
Level Detect Transition Density Enable. When LDTDEN is HIGH, the Signal Level
Detector, Range Controller, and Transition Density Detector are all enabled to
determine if the RXPLL tracks REFCLKx± or the selected input serial data stream.
If the Signal Level Detector, Range Controller, or Transition Density Detector are out
of their respective limits while LDTDEN is HIGH, the RXPLL locks to REFCLK± until
such a time they become valid. The (SDASEL[A..D][1:0]) are used to configure the
trip level of the Signal Level Detector. The Transition Density Detector limit is one
transition in every 60 consecutive bits. When LDTDEN is LOW, only the Range
Controller is used to determine if the RXPLL tracks REFCLKx± or the selected input
serial data stream. For the cases when RXCKSELx = 0 (recovered clock), it is
recommended to set LDTDEN = HIGH.
ULCA
ULCB
ULCC
ULCD
LVTTL Input,
internal pull-up
Use Local Clock. When ULCx is LOW, the RXPLL is forced to lock to REFCLKx±
instead of the received serial data stream. While ULCx is LOW, the LFIx for the
associated channel is LOW indicating a link fault.
SPDSELA
SPDSELB
SPDSELC
SPDSELD
3-Level Select[3]
static control input
When ULCx is HIGH, the RXPLL performs Clock and Data Recovery functions on
the input data streams. This function is used in applications in which a stable
RXCLKx± is needed. In cases when there is an absence of valid data transitions for
a long period of time, or the high-gain differential serial inputs (INx±) are left floating,
there may be brief frequency excursions of the RXCLKx± outputs from REFCLKx±.
Serial Rate Select. The SPDSELx inputs specify the operating signaling-rate range
of each channel’s transmit and receive PLL.
LOW = 195 – 400 MBaud
MID = 400 – 800 MBaud
HIGH = 800 – 1500 MBaud
INSELA
INSELB
INSELC
INSELD
LVTTL Input,
asynchronous
Receive Input Selector. The INSELx input determines which external serial bit
stream is passed to the receiver’s Clock and Data Recovery circuit. When INSELx
is HIGH, the Primary Differential Serial Data Input, INx1±, is selected for the
associated receive channel. When INSELx is LOW, the Secondary Differential
Serial Data Input, INx2±, is selected for the associated receive channel.
LPENA
LPENB
LPENC
LPEND
LVTTL Input,
asynchronous,
internal pull-down
Loop-Back-Enable. The LPENx input enables the internal serial loop-back for the
associated channel. When LPENx is HIGH, the transmit serial data from the
associated channel is internally routed to the associated receive Clock and Data
Recovery (CDR) circuit. All enabled serial drivers on the channel are forced to differential logic-1, and the serial data inputs are ignored. When LPENx is LOW, the
internal serial loop-back function is disabled.
Note
3. 3-Level Select inputs are used for static configuration. These are ternary inputs that make use of logic levels of LOW, MID, and HIGH. The LOW level is usually
implemented by direct connection to VSS (ground). The HIGH level is usually implemented by direct connection to VCC (power). The MID level is usually
implemented by not connecting the input (left floating), which allows it to self bias to the proper level.
Document #: 38-02065 Rev. *J
Page 12 of 49
CYP15G0403DXB
Pin Descriptions (continued)
CYP15G0403DXB Quad HOTLink II Transceiver
Name
LFIA
LFIB
LFIC
LFID
I/O Characteristics
LVTTL Output,
asynchronous
Signal Description
Link Fault Indication Output. LFIx is an output status indicator signal. LFIx is the
logical OR of six internal conditions. LFIx is asserted LOW when any of the following
conditions is true:
■ Received serial data rate outside expected range
■ Analog amplitude below expected levels
■ Transition density lower than expected
■ Receive channel disabled
■ ULCx is LOW
■ Absence of REFCLKx±.
Device Configuration and Control Bus Signals
WREN
LVTTL input,
asynchronous,
internal pull-up
Control Write Enable. The WREN input writes the values of the DATA[7:0] bus into
the latch specified by the address location on the ADDR[3:0] bus.[4]
ADDR[3:0]
LVTTL input
asynchronous,
internal pull-up
Control Addressing Bus. The ADDR[3:0] bus is the input address bus used to
configure the device. The WREN input writes the values of the DATA[7:0] bus into
the latch specified by the address location on the ADDR[3:0] bus.[4] Table 9 on page
22 lists the configuration latches within the device, and the initialization value of the
latches upon the assertion of RESET. Table 10 on page 27 shows how the latches
are mapped in the device.
DATA[7:0]
LVTTL input
asynchronous,
internal pull-up
Control Data Bus. The DATA[7:0] bus is the input data bus used to configure the
device. The WREN input writes the values of the DATA[7:0] bus into the latch
specified by address location on the ADDR[3:0] bus.[4 ] Table 9 lists the configuration
latches within the device, and the initialization value of the latches upon the
assertion of RESET. Table 10 shows how the latches are mapped in the device.
Internal Device Configuration Latches
RFMODE[A..D][1:0] Internal Latch[5]
Reframe Mode Select.
Internal
Latch[5]
Framing Character Select.
Internal
Latch[5]
Receiver Decoder Mode Select.
Internal
Latch[5]
Receiver Decoder Bypass.
Internal
Latch[5]
Receive Clock Select.
Internal
Latch[5]
Receive Clock Rate Select.
SDASEL[A..D][1:0] Internal
Latch[5]
Signal Detect Amplitude Select.
ENCBYP[A..D]
Latch[5]
Transmit Encoder Bypassed.
FRAMCHAR[A..D]
DECMODE[A..D]
DECBYP[A..D]
RXCKSEL[A..D]
RXRATE[A..D]
TXCKSEL[A..D]
TXRATE[A..D]
RFEN[A..D]
RXPLLPD[A..D]
RXBIST[A..D]
TXBIST[A..D]
OE2[A..D]
OE1[A..D]
Internal
Internal Latch
[5]
Transmit Clock Select.
Internal
Latch[5]
Transmit PLL Clock Rate Select.
Internal
Latch[5]
Reframe Enable.
Internal Latch
[5]
Receive Channel Power Control.
Internal
Latch[5]
Receive Bist Disabled.
Internal
Latch[5]
Transmit Bist Disabled.
Internal
Latch[5]
Differential Serial Output Driver 2 Enable.
Internal
Latch[5]
Differential Serial Output Driver 1 Enable.
Notes
4. See “Device Configuration and Control Interface” on page 22 for detailed information on the operation of the Configuration Interface.
5. See “Device Configuration and Control Interface” on page 22 for detailed information on the internal latches.
Document #: 38-02065 Rev. *J
Page 13 of 49
CYP15G0403DXB
Pin Descriptions (continued)
CYP15G0403DXB Quad HOTLink II Transceiver
Name
PABRST[A..D]
GLEN[11..0]
FGLEN[2..0]
I/O Characteristics
Signal Description
Internal Latch
[5]
Transmit Clock Phase Alignment Buffer Reset.
Internal Latch
[5]
Global Latch Enable.
Internal Latch
[5]
Force Global Latch Enable.
Factory Test Modes
LTEN1
LVTTL input,
internal pull-down
Factory Test 1. LTEN1 input is for factory testing only. This input may be left as a
NO CONNECT, or GND only.
SCANEN2
LVTTL input,
internal pull-down
Factory Test 2. SCANEN2 input is for factory testing only. This input may be left as
a NO CONNECT, or GND only.
TMEN3
LVTTL input,
internal pull-down
Factory Test 3. TMEN3 input is for factory testing only. This input may be left as a
NO CONNECT, or GND only.
OUTA1±
OUTB1±
OUTC1±
OUTD1±
CML Differential
Output
Primary Differential Serial Data Output. The OUTx1± PECL-compatible CML
outputs (+3.3V referenced) are capable of driving terminated transmission lines or
standard fiber-optic transmitter modules, and must be AC-coupled for
PECL-compatible connections.
OUTA2±
OUTB2±
OUTC2±
OUTD2±
CML Differential
Output
Secondary Differential Serial Data Output. The OUTx2± PECL-compatible CML
outputs (+3.3V referenced) are capable of driving terminated transmission lines or
standard fiber-optic transmitter modules, and must be AC-coupled for PECL-compatible
connections.
INA1±
INB1±
INC1±
IND1±
Differential Input
Primary Differential Serial Data Input. The INx1± input accepts the serial data
stream for deserialization and decoding. The INx1± serial stream is passed to the
receive CDR circuit to extract the data content when INSELx = HIGH.
INA2±
INB2±
INC2±
IND2±
Differential Input
Secondary Differential Serial Data Input. The INx2± input accepts the serial data
stream for deserialization and decoding. The INx2± serial stream is passed to the
receiver CDR circuit to extract the data content when INSELx = LOW.
TMS
LVTTL Input,
internal pull-up
Test Mode Select. Used to control access to the JTAG Test Modes. If maintained
high for ≥5 TCLK cycles, the JTAG test controller is reset.
TCLK
LVTTL Input,
internal pull-down
JTAG Test Clock.
TDO
3-State LVTTL Output Test Data Out. JTAG data output buffer. High-Z while JTAG test mode is not
selected.
TDI
LVTTL Input,
internal pull-up
Test Data In. JTAG data input port.
TRST
LVTTL Input,
internal pull-up
JTAG reset signal. When asserted (LOW), this input asynchronously resets the
JTAG test access port controller.
Analog I/O
JTAG Interface
Power
VCC
+3.3V Power.
GND
Signal and Power Ground for all internal circuits.
Document #: 38-02065 Rev. *J
Page 14 of 49
CYP15G0403DXB
CYP15G0403DXB HOTLink II Operation
The CYP15G0403DXB is a highly configurable, independent
clocking, quad-channel transceiver designed to support reliable
transfer of large quantities of data, using high-speed serial links
from multiple sources to multiple destinations. This device
supports four single-byte channels.
CYP15G0403DXB Transmit Data Path
Input Register
The bits in the Input Register for each channel support different
assignments, based on if the input data is encoded or
unencoded. These assignments are shown in Table 1.
When the ENCODER is enabled, each input register captures
eight data bits and two control bits on each input clock cycle.
When the Encoder is bypassed, the control bits are part of the
pre-encoded 10-bit character.
When the Encoder is enabled, the TXCTx[1:0] bits are interpreted along with the associated TXDx[7:0] character to
generate a specific 10-bit transmission character.
Phase-Align Buffer
Data from each Input Register is passed to the associated
Phase-Align Buffer, when the TXDx[7:0] and TXCTx[1:0] input
registers are clocked using TXCLKx¦ (TXCKSELx = 0 and
TXRATEx = 0). When the TXDx[7:0] and TXCTx[1:0] input
registers are clocked using REFCLKx± (TXCKSELx = 1) and
REFCLKx± is a full-rate clock, the associated Phase Alignment
Buffer in the transmit path is bypassed. These buffers are used
to absorb clock phase differences between the TXCLKx input
clock and the internal character clock for that channel.
Once initialized, TXCLKx is allowed to drift in phase as much as
±180 degrees. If the input phase of TXCLKx drifts beyond the
handling capacity of the Phase Align Buffer, TXERRx is asserted
to indicate the loss of data, and remains asserted until the Phase
Align Buffer is initialized. The phase of the TXCLKx relative to its
associated internal character rate clock is initialized when the
configuration latch PABRSTx is written as 0. When the
associated TXERRx is deasserted, the Phase Align Buffer is
initialized and input characters are correctly captured.
Table 1. Input Register Bit Assignments[6]
If the phase offset, between the initialized location of the input
clock and REFCLKx¦, exceeds the skew handling capabilities of
the Phase-Align Buffer, an error is reported on that channel’s
TXERRx output. This output indicates an error continuously until
the Phase-Align Buffer for that channel is reset. While the error
remains active, the transmitter for that channel outputs a
continuous C0.7 character to indicate to the remote receiver that
an error condition is present in the link.
Each Phase-Align Buffer may be individually reset with minimal
disruption of the serial data stream. When a Phase-Align Buffer
error is present, the transmission of a Word Sync Sequence
re-centers the Phase-Align Buffer and clears the error indication.
Note. K28.5 characters may be added or removed from the data
stream during the Phase Align Buffer reset operation. When
used with non-Cypress devices that require a complete
16-character Word Sync Sequence for proper receive Elasticity
Buffer Operation, it is recommend that the Phase Alignment
Buffer reset be followed by a Word Sync Sequence to ensure
proper operation.
Encoder
Each character received from the Input Register or Phase-Align
Buffer is passed to the Encoder logic. This block interprets each
character and any associated control bits, and outputs a 10-bit
transmission character.
Depending on the operational mode, the generated transmission
character may be
■
the 10-bit pre-encoded character accepted in the Input
Register.
■
the 10-bit equivalent of the 8-bit Data character accepted in the
Input Register.
■
the 10-bit equivalent of the 8-bit Special Character code
accepted in the Input Register.
■
the 10-bit equivalent of the C0.7 violation character if a
Phase-Align Buffer overflow or underflow error is present.
■
a character that is part of the 511-character BIST sequence.
■
a K28.5 character generated as an individual character or as
part of the 16-character Word Sync Sequence.
Data Encoding
Signal Name
Unencoded
Encoded
TXDx[0] (LSB)
DINx[0]
TXDx[0]
TXDx[1]
DINx[1]
TXDx[1]
TXDx[2]
DINx[2]
TXDx[2]
TXDx[3]
DINx[3]
TXDx[3]
TXDx[4]
DINx[4]
TXDx[4]
TXDx[5]
DINx[5]
TXDx[5]
TXDx[6]
DINx[6]
TXDx[6]
TXDx[7]
DINx[7]
TXDx[7]
TXCTx[0]
DINx[8]
TXCTx[0]
TXCTx[1] (MSB)
DINx[9]
TXCTx[1]
Raw data, as received directly from the Transmit Input Register,
is seldom in a form suitable for transmission across a serial link.
The characters must usually be processed or transformed to
guarantee
■
a minimum transition density (to allow the receive PLL to extract
a clock from the serial data stream).
■
a DC-balance in the signaling (to prevent baseline wander).
■
run-length limits in the serial data (to limit the bandwidth requirements of the serial link).
■
the remote receiver a way of determining the correct character
boundaries (framing).
Note
6. LSB shifted out first.
Document #: 38-02065 Rev. *J
Page 15 of 49
CYP15G0403DXB
When the Encoder is enabled (ENCBYPx = 1), the characters
transmitted are converted from Data or Special Character codes
to 10-bit transmission characters, using an integrated 8B/10B
encoder. When directed to encode the character as a Special
Character code, the encoder uses the Special Character
encoding rules listed in Table 16 on page 46. When directed to
encode the character as a Data character, it is encoded using the
Data Character encoding rules in Table 15 on page 42.
The 8B/10B encoder is standards compliant with ANSI/NCITS
ASC X3.230-1994 Fibre Channel, IEEE 802.3z Gigabit Ethernet,
the IBM® ESCON® and FICON™ channels, ETSI DVB-ASI, and
ATM Forum standards for data transport.
Many of the Special Character codes listed in Table 16 may be
generated by more than one input character. The
CYP15G0403DXB is designed to support two independent (but
non-overlapping) Special Character code tables. This allows the
CYP15G0403DXB to operate in mixed environments with other
Cypress HOTLink devices using the enhanced Cypress
command code set, and the reduced command sets of other
non-Cypress devices. Even when used in an environment that
normally uses non-Cypress Special Character codes, the
selective use of Cypress command codes can permit operation
where running disparity and error handling must be managed.
Following conversion of each input character from eight bits to a
10-bit transmission character, it is passed to the Transmit Shifter
and is shifted out LSB first, as required by ANSI and IEEE
standards for 8B/10B coded serial data streams.
Transmit Modes
Encoder Bypass
When the Encoder is bypassed, the character captured from the
TXDx[7:0] and TXCTx[1:0] input register is passed directly to the
transmit shifter without modification. With the encoder bypassed,
the TXCTx[1:0] inputs are considered part of the data character
and do not perform a control function that would otherwise
modify the interpretation of the TXDx[7:0] bits. The bit usage and
mapping of these control bits when the Encoder is bypassed is
shown in Table 2.
Table 2. Encoder Bypass Mode
Signal Name
Bus Weight
10B Name
TXDx[0] (LSB)
20
a[6]
TXDx[1]
21
b
TXDx[2]
22
c
TXDx[3]
23
d
TXDx[4]
24
e
TXDx[5]
25
i
TXDx[6]
26
f
TXDx[7]
27
g
TXCTx[0]
28
h
TXCTx[1] (MSB)
29
j
Document #: 38-02065 Rev. *J
When the encoder is enabled, the TXCTx[1:0] data control bits
control the interpretation of the TXDx[7:0] bits and the characters
generated by them. These bits are interpreted as listed in
Table 3.
Table 3. Transmit Modes
TXCTx[1]
TXCTx[0]
0
0
Encoded data character
Characters Generated
0
1
K28.5 fill character
1
0
Special character code
1
1
16-character Word Sync Sequence
Word Sync Sequence
When TXCTx[1:0] = 11, a 16-character sequence of K28.5
characters, known as a Word Sync Sequence, is generated on
the associated channel. This sequence of K28.5 characters may
start with either a positive or negative disparity K28.5 (as determined by the current running disparity and the 8B/10B coding
rules). The disparity of the second and third K28.5 characters in
this sequence are reversed from what normal 8B/10B coding
rules would generate. The remaining K28.5 characters in the
sequence follow all 8B/10B coding rules. The disparity of the
generated K28.5 characters in this sequence follow a pattern of
either ++––+–+–+–+–+–+– or ––++–+–+–+–+–+–+.
The generation of this sequence, once started, cannot be
stopped until all 16 characters have been sent. The content of
the associated input registers are ignored for the duration of this
sequence. At the end of this sequence, if the TXCTx[1:0] = 11
condition is sampled again, the sequence restarts and remains
uninterruptible for the following 15 character clocks.
Transmit BIST
Each transmit channel contains an internal pattern generator that
can be used to validate both the link and device operation. These
generators are enabled by the associated TXBISTx latch via the
device configuration interface. When enabled, a register in the
associated transmit channel becomes a signature pattern
generator by logically converting to a Linear Feedback Shift
Register (LFSR). This LFSR generates a 511-character (or
526-character) sequence that includes all Data and Special
Character codes, including the explicit violation symbols. This
provides a predictable yet pseudo-random sequence that can be
matched to an identical LFSR in the attached Receiver(s).
A device reset (RESET sampled LOW) presets the BIST Enable
Latches to disable BIST on all channels.
All data and data-control information present at the associated
TXDx[7:0] and TXCTx[1:0] inputs are ignored when BIST is
active on that channel. If the receive channels are configured for
reference clock operation, each pass is preceded by a
16-character Word Sync Sequence to allow Elasticity Buffer
alignment and management of clock-frequency variations.
Transmit PLL Clock Multiplier
Each Transmit PLL Clock Multiplier accepts a character-rate or
half-character-rate external clock at the associated REFCLKx±
input, and that clock is multiplied by 10 or 20 (as selected by
TXRATEx) to generate a bit-rate clock for use by the transmit
shifter. It also provides a character-rate clock used by the
Page 16 of 49
CYP15G0403DXB
transmit paths, and outputs this character rate clock as
TXCLKOx.
internal logic for that channel is also powered down. A device
reset (RESET sampled LOW) disables all output drivers.
Each clock multiplier PLL can accept a REFCLKx± input
between 19.5 MHz and 150 MHz, however, this clock range is
limited by the operating mode of the CYP15G0403DXB clock
multiplier (TXRATEx) and by the level on the associated
SPDSELx input.
Note. When a disabled transmit channel (i.e., both outputs
disabled) is re-enabled:
■
data on the serial outputs may not meet all timing specifications
for up to 250 μs
SPDSELx are 3-level select[3] inputs that select one of three
operating ranges for the serial data outputs and inputs of the
associated channel. The operating serial signaling-rate and
allowable range of REFCLKx± frequencies are listed in Table 4.
■
the state of the phase-align buffer cannot be guaranteed, and
a phase-align reset is required if the phase-align buffer is used
Table 4. Operating Speed Settings
Serial Line Receivers
SPDSELx
TXRATE
REFCLKx±
Frequency
(MHz)
Signaling Rate
(MBaud)
LOW
1
reserved
195–400
0
19.5–40
1
20–40
0
40–80
MID (Open)
HIGH
1
40–75
0
80–150
400–800
800–1500
The REFCLKx± inputs are differential inputs with each input
internally biased to 1.4V. If the REFCLKx+ input is connected to
a TTL, LVTTL, or LVCMOS clock source, the input signal is
recognized when it passes through the internally biased
reference point. When driven by a single-ended TTL, LVTTL, or
LVCMOS clock source, connect the clock source to either the
true or complement REFCLKx input, and leave the alternate
REFCLKx input open (floating).
When both the REFCLKx+ and REFCLKx– inputs are
connected, the clock source must be a differential clock. This can
either be a differential LVPECL clock that is DC-or AC-coupled
or a differential LVTTL or LVCMOS clock.
By connecting the REFCLKx– input to an external voltage
source, it is possible to adjust the reference point of the
REFCLKx+ input for alternate logic levels. When doing so it is
necessary to ensure that the input differential crossing point
remains within the parametric range supported by the input.
Serial Output Drivers
The serial output interface drivers use differential Current Mode
Logic (CML) drivers to provide source-matched drivers for transmission lines. These drivers accept data from the Transmit
Shifters. These drivers have signal swings equivalent to that of
standard PECL drivers, and are capable of driving AC-coupled
optical modules or transmission lines. When configured for local
loopback (LPENx = HIGH), all enabled serial drivers are
configured to drive a static differential logic.
Transmit Channels Enabled
Each driver can be enabled or disabled separately via the device
configuration interface.
When a driver is disabled via the configuration interface, it is
internally powered down to reduce device power. If both serial
drivers for a channel are in this disabled state, the associated
Document #: 38-02065 Rev. *J
CYP15G0403DXB Receive Data Path
Two differential Line Receivers, INx1± and INx2±, are available
on each channel for accepting serial data streams. The active
Serial Line Receiver on a channel is selected using the
associated INSELx input. The Serial Line Receiver inputs are
differential, and can accommodate wire interconnect and filtering
losses or transmission line attenuation greater than 16 dB. For
normal operation, these inputs should receive a signal of at least
VIDIFF > 100 mV, or 200 mV peak-to-peak differential. Each Line
Receiver can be DC- or AC-coupled to +3.3V powered fiber-optic
interface modules (any ECL/PECL family, not limited to 100K
PECL) or AC-coupled to +5V powered optical modules. The
common-mode tolerance of these line receivers accommodates
a wide range of signal termination voltages. Each receiver
provides internal DC-restoration, to the center of the receiver’s
common mode range, for AC-coupled signals.
The local internal loopback (LPENx) allows the serial transmit
data outputs to be routed internally back to the Clock and Data
Recovery circuit associated with each channel. When configured
for local loopback, the associated transmit serial driver outputs
are forced to output a differential logic-1. This prevents local
diagnostic patterns from being broadcast to attached remote
receivers.
Signal Detect/Link Fault
Each selected Line Receiver (i.e., that routed to the clock and
data recovery PLL) is simultaneously monitored for
■
analog amplitude above amplitude level selected by SDASELx
■
transition density above the specified limit
■
range controls report the received data stream inside normal
frequency range (±1500 ppm[27])
■
receive channel enabled
■
Presence of reference clock
■
ULCx is not asserted.
All of these conditions must be valid for the Signal Detect block
to indicate a valid signal is present. This status is presented on
the LFIx (Link Fault Indicator) output associated with each
receive channel, which changes synchronous to the selected
receive interface clock.
Analog Amplitude
While most signal monitors are based on fixed constants, the
analog amplitude level detection is adjustable to allow operation
with highly attenuated signals, or in high-noise environments.
The analog amplitude level detection is set by the SDASELx
Page 17 of 49
CYP15G0403DXB
latch via device configuration interface. The SDASELx latch sets
the trip point for the detection of a valid signal at one of three
levels, as listed in Table 5. This control input affects the analog
monitors for all receive channels.
Table 5. Analog Amplitude Detect Valid Signal Levels[7]
SDASEL Typical Signal with Peak Amplitudes Above
00
Analog Signal Detector is disabled
01
140 mV p-p differential
10
280 mV p-p differential
11
420 mV p-p differential
The Analog Signal Detect monitors are active for the Line
Receiver as selected by the associated INSELx input. When
configured for local loopback, no input receivers are selected,
and the LFIx output for each channel reports only the receive
VCO frequency out-of-range and transition density status of the
associated transmit signal. When local loopback is active, the
associated Analog Signal Detect Monitor is disabled.
Transition Density
The Transition Detection logic checks for the absence of transitions spanning greater than six transmission characters (60 bits).
If no transitions are present in the data received, the Detection
logic for that channel asserts LFIx.
Range Controls
The CDR circuit includes logic to monitor the frequency of the
PLL Voltage Controlled Oscillator (VCO) used to sample the
incoming data stream. This logic ensures that the VCO operates
at, or near the rate of the incoming data stream for two primary
cases:
■
when the incoming data stream resumes after a time in which
it has been “missing.”
■
when the incoming data stream is outside the acceptable
signaling rate range.
To perform this function, the frequency of the RXPLL VCO is
periodically compared to the frequency of the REFCLKx± input.
If the VCO is running at a frequency beyond ±1500 ppm[27] as
defined by the REFCLKx± frequency, it is periodically forced to
the correct frequency (as defined by REFCLKx±, SPDSELx, and
TXRATEx) and then released in an attempt to lock to the input
data stream.
The sampling and relock period of the Range Control is calculated as follows: RANGE_CONTROL_ SAMPLING_PERIOD =
(RECOVERED BYTE CLOCK PERIOD) * (4096).
During the time that the Range Control forces the RXPLL VCO
to track REFCLKx±, the LFIx output is asserted LOW. After a
valid serial data stream is applied, it may take up to one RANGE
CONTROL SAMPLING PERIOD before the PLL locks to the
input data stream, after which LFIx should be HIGH.
Receive Channel Enabled
The CYP15G0403DXB contains four receive channels that can
be independently enabled and disabled. Each channel can be
enabled or disabled separately through the RXPLLPDx input
latch as controlled by the device configuration interface. When
the RXPLLPDx latch = 0, the associated PLL and analog circuitry
of the channel is disabled. Any disabled channel indicates a
constant link fault condition on the LFIx output. When
RXPLLPDx = 1, the associated PLL and receive channel is
enabled to receive and decode a serial stream.
Note. When a disabled receive channel is reenabled, the status
of the associated LFIx output and data on the parallel outputs for
the associated channel may be indeterminate for up to 2 ms.
Clock/Data Recovery
The extraction of a bit-rate clock and recovery of bits from each
received serial stream is performed by a separate CDR block
within each receive channel. The clock extraction function is
performed by an integrated PLL that tracks the frequency of the
transitions in the incoming bit stream and align the phase of the
internal bit-rate clock to the transitions in the selected serial data
stream.
Each CDR accepts a character-rate (bit-rate ÷ 10) or
half-character-rate (bit-rate ÷ 20) reference clock from the
associated REFCLKx± input. This REFCLKx± input is used to
■
ensure that the VCO (within the CDR) is operating at the correct
frequency (rather than a harmonic of the bit-rate)
■
reduce PLL acquisition time
■
limit unlocked frequency excursions of the CDR VCO when
there is no input data present at the selected Serial Line
Receiver.
Regardless of the type of signal present, the CDR attempts to
recover a data stream from it. If the signalling rate of the
recovered data stream is outside the limits set by the range
control monitors, the CDR tracks REFCLKx± instead of the data
stream. Once the CDR output (RXCLK±) frequency returns back
close to REFCLKx± frequency, the CDR input is switched back
to the input data stream. If no data is present at the selected line
receiver, this switching behavior may result in brief RXCLK±
frequency excursions from REFCLKx±. However, the validity of
the input data stream is indicated by the LFIx output. The
frequency of REFCLKx± is required to be within ±1500 ppm[27]
of the frequency of the clock that drives the REFCLKx± input of
the remote transmitter to ensure a lock to the incoming data
stream.
For systems using multiple or redundant connections, the LFIx
output can be used to select an alternate data stream. When an
LFIx indication is detected, external logic can toggle selection of
the associated INx1± and INx2± input through the associated
INSELx input. When a port switch takes place, it is necessary for
the receive PLL for that channel to reacquire the new serial
stream and frame to the incoming character boundaries.
Note
7. The peak amplitudes listed in this table are for typical waveforms that have generally 3–4 transitions for every ten bits. In a worse case environment the signals
may have a sine-wave appearance (highest transition density with repeating 0101...). Signal peak amplitudes levels within this environment type could increase
the values in the table above by approximately 100 mV.
Document #: 38-02065 Rev. *J
Page 18 of 49
CYP15G0403DXB
Deserializer/Framer
Each CDR circuit extracts bits from the associated serial data
stream and clocks these bits into the Shifter/Framer at the
bit-clock rate. When enabled, the Framer examines the data
stream looking for one or more COMMA or K28.5 characters at
all possible bit positions. The location of this character in the data
stream is used to determine the character boundaries of all
following characters.
Framing Character
The CYP15G0403DXB allows selection of different framing
characters on each channel. Two combinations of framing
characters are supported to meet the requirements of different
interfaces. The selection of the framing character is made
through the FRAMCHARx latches via the configuration interface.
makes the associated link much more robust to incorrect framing
due to aliased SYNC characters in the data stream. In this mode,
the framer does not adjust the character clock boundary, but
instead aligns the character to the already recovered character
clock. This ensures that the recovered clock does not contain
any significant phase changes or hops during normal operation
or framing, and allows the recovered clock to be replicated and
distributed to other external circuits or components using
PLL-based clock distribution elements. In this framing mode the
character boundaries are only adjusted if the selected framing
character is detected at least twice within a span of 50 bits, with
both instances on identical 10-bit character boundaries.
When RFMODEx[1:0] = 01, the Alternate-mode Multi-Byte
Framer is enabled. Like the Cypress-mode Multi-Byte Framer,
multiple framing characters must be detected before the
character boundary is adjusted. In this mode, the data stream
must contain a minimum of four of the selected framing
characters, received as consecutive characters, on identical
10-bit boundaries, before character framing is adjusted.
The specific bit combinations of these framing characters are
listed in Table 6. When the specific bit combination of the
selected framing character is detected by the framer, the boundaries of the characters present in the received data stream are
known.
10B/8B Decoder Block
Table 6. Framing Character Selector
The decoder logic block performs two primary functions:
FRAMCHARx
Bits detected in framer
Character Name
Bits Detected
0
COMMA+
COMMA–
00111110XX[8]
or 11000001XX
1
–K28.5
+K28.5
0011111010 or
1100000101
Framer
The framer on each channel operates in one of three different
modes. Each framer may be enabled or disabled using the
RFENx latches via the configuration interface. When the framer
is disabled (RFENx = 0), no combination of received bits alters
the frame information.
When the Low-Latency framer is selected (RFMODEx[1:0] = 00),
the framer operates by stretching the recovered character clock
until it aligns with the received character boundaries. In this
mode the framer starts its alignment process on the first
detection of the selected framing character. To reduce the impact
on external circuits that use the recovered clock, the clock period
is not stretched by more than two bit-periods in any one clock
cycle. When operated with a character-rate output clock, the
output of properly framed characters may be delayed by up to
nine character-clock cycles from the detection of the selected
framing character. When operated with a half-character-rate
output clock, the output of properly framed characters may be
delayed by up to 14 character-clock cycles from the detection of
the framing character.
Note. When Receive BIST is enabled on a channel, the
Low-Latency Framer must not be enabled. The BIST sequence
contains an aliased K28.5 framing character, which causes the
Receiver to update its character boundaries incorrectly.
When RFMODEx[1:0] = 10, the Cypress-Mode Multi-Byte framer
is selected. The required detection of multiple framing characters
■
decoding the received transmission characters to Data and
Special Character codes
■
comparing generated BIST patterns with received characters
to permit at-speed link and device testing.
The framed parallel output of each deserializer shifter is passed
to its associated 10B/8B Decoder where, if the decoder is
enabled, the input data is transformed from a 10-bit transmission
character back to the original Data or Special Character code.
This block uses the 10B/8B decoder patterns in Table 15 on page
42 and Table 16 on page 46. Received Special Code characters
are decoded using Table 16. Valid data characters are indicated
by a 000b bit-combination on the associated RXSTx[2:0] status
bits, and Special Character codes are indicated by a 001b
bit-combination of these status outputs. Framing characters,
Invalid patterns, disparity errors, and synchronization status are
presented as alternate combinations of these status bits.
When DECBYPx = 0, the 10B/8B decoder is bypassed via the
configuration interface. When bypassed, raw 10-bit characters
are passed through the receiver and presented at the RXDx[7:0]
and the RXSTA[1:0] outputs as 10-bit wide characters.
When the decoder is enabled by setting DECBYPx = 1 via the
configuration interface, the 10-bit transmission characters are
decoded using Table 15 and Table 16. Received Special
characters are decoded using Table 16. The columns used in
Table 16 are determined by the DECMODEx latch via the device
configuration interface. When DECMODEx = 0 the ALTERNATE
table is used and when DECMODEx = 1 the CYPRESS table is
used.
Receive BIST Operation
The receiver channel contains an internal pattern checker that
can be used to validate both device and link operation. These
pattern checkers are enabled by the associated RXBISTx latch
via the device configuration interface. When enabled, a register
Note
8. The standard definition of a Comma contains only seven bits. However, since all valid Comma characters within the 8B/10B character set also have the eighth
bit as an inversion of the seventh bit, the compare pattern is extended to a full eight bits to reduce the possibility of a framing error.
Document #: 38-02065 Rev. *J
Page 19 of 49
CYP15G0403DXB
in the associated receive channel becomes a signature pattern
generator and checker by logically converting to a Linear
Feedback Shift Register (LFSR). This LFSR generates a
511-character or 526-character sequence that includes all Data
and Special Character codes, including the explicit violation
symbols. This provides a predictable yet pseudo-random
sequence that can be matched to an identical LFSR in the
attached Transmitter(s). When synchronized with the received
data stream, the associated Receiver checks each character in
the Decoder with each character generated by the LFSR and
indicates compare errors and BIST status at the RXSTx[2:0] bits
of the Output Register.
When BIST is first recognized as being enabled in the Receiver,
the LFSR is preset to the BIST-loop start-code of D0.0. This code
D0.0 is sent only once per BIST loop. The status of the BIST
progress and any character mismatches are presented on the
RXSTx[2:0] status outputs.
Code rule violations or running disparity errors that occur as part
of the BIST loop do not cause an error indication. RXSTx[2:0]
indicates 010b or 100b for one character period per BIST loop to
indicate loop completion. This status can be used to check test
pattern progress. These same status values are presented when
the decoder is bypassed and BIST is enabled on a receive
channel.
The specific status reported by the BIST state machine are listed
in Table 11 on page 28. These same codes are reported on the
receive status outputs.
The specific patterns checked by each receiver are described in
detail in the Cypress application note “HOTLink Built-In
Self-Test.” The sequence compared by the CYP15G0403DXB is
identical to that in the CY7B933, CY7C924DX, and
CYP15G0401DXB, allowing interoperable systems to be built
when used at compatible serial signaling rates.
If the number of invalid characters received ever exceeds the
number of valid characters by 16, the receive BIST state
machine aborts the compare operations and resets the LFSR to
the D0.0 state to look for the start of the BIST sequence again.
When the receive paths are configured for REFCLKx± operation,
each pass must be preceded by a 16-character Word Sync
Sequence to allow management of clock frequency variations.
The receive BIST state machine requires the characters to be
correctly framed for it to detect the BIST sequence. If the Low
Latency Framer is enabled, the Framer misaligns to an aliased
SYNC character within the BIST sequence. If the Alternate
Multi-Byte Framer is enabled and the Receiver outputs are
clocked relative to a recovered clock, it is generally necessary to
frame the receiver before BIST is enabled. If the receive outputs
are clocked relative to REFCLKx±, the transmitter precedes
every 511 character BIST sequence with a 16
character-character Word Sync Sequence.
A device reset (RESET sampled LOW) presets the BIST Enable
Latches to disable BIST on all channels.
Receive Elasticity Buffer
Each receive channel contains an Elasticity Buffer that is
designed to support multiple clocking modes. These buffers
allow data to be read using a clock that is asynchronous in both
Document #: 38-02065 Rev. *J
frequency and phase from the Elasticity Buffer write clock, or to
be read using a clock that is frequency coherent but with uncontrolled phase relative to the Elasticity Buffer write clock.
If the chip is configured for operation with a recovered clock, the
Elasticity Buffer is bypassed.
Each Elasticity Buffer is 10 characters deep, and supports and
an 11 bit wide data path. It is capable of supporting a decoded
character and three status bits for each character present in the
buffer. The write clock for these buffers is always the recovered
clock for the associated read channel.
Receive Modes
When the receive channel is clocked by REFCLKx±, the
RXCLKx± outputs present a buffered or divided (depending on
RXRATEx) and delayed form of REFCLKx±. In this mode, the
receive Elasticity Buffers are enabled. For REFCLKx± clocking,
the Elasticity Buffers must be able to insert K28.5 characters and
delete framing characters as appropriate.
The insertion of a K28.5 or deletion of a framing character can
occur at any time on any channel, however, the actual timing of
these insertions and deletions is controlled in part by how the
transmitter sends its data. Insertion of a K28.5 character can only
occur when the receiver has a framing character in the Elasticity
Buffer. Likewise, to delete a framing character, one must also be
in the Elasticity Buffer. To prevent a buffer overflow or underflow
on a receive channel, a minimum density of framing characters
must be present in the received data streams.
When the receive channel Output Register is clocked by a
recovered clock, no characters are added or deleted and the
receiver Elasticity Buffer is bypassed.
Power Control
The CYP15G0403DXB supports user control of the powered up
or down state of each transmit and receive channel. The receive
channels are controlled by the RXPLLPDx latch via the device
configuration interface. When RXPLLPDx = 0, the associated
PLL and analog circuitry of the channel is disabled. The transmit
channels are controlled by the OE1x and the OE2x latches via
the device configuration interface. When a driver is disabled via
the configuration interface, it is internally powered down to
reduce device power. If both serial drivers for a channel are in
this disabled state, the associated internal logic for that channel
is also powered down.
Device Reset State
When the CYP15G0403DXB is reset by assertion of RESET, all
state machines, counters, and configuration latches in the device
are initialized to a reset state, and the Elasticity Buffer pointers
are set to a nominal offset. Additionally, the JTAG controller must
also be reset to ensure valid operation (even if JTAG testing is
not performed). See “JTAG Support” on page 27 for JTAG state
machine initialization. See Table 9 on page 22 for the initialize
values of the configuration latches.
Following a device reset, it is necessary to enable the transmit
and receive channels used for normal operation. This can be
done by sequencing the appropriate values on the device configuration interface.[4]
Page 20 of 49
CYP15G0403DXB
Output Bus
proper character boundary, and LOW for all other bit combinations.
Each receive channel presents an 11-signal output bus
consisting of
When the low-latency framer and half-rate receive port clocking
are also enabled, the framer stretches the recovered clock to the
nearest 20-bit boundary such that the rising edge of RXCLKx+
occurs when COMDETx is present on the associated output bus.
■
an 8-bit data bus
■
a 3-bit status bus.
The signals present on this output bus are modified by the
present operating mode of the CYP15G0403DXB as selected by
the DECBYPx configuration latch. This mapping is shown in
Table 7.
Table 7. Output Register Bit Assignments
When the Cypress or Alternate Mode Framer is enabled and
half-rate receive port clocking is also enabled, the output clock
is not modified when framing is detected, but a single pipeline
stage may be added or subtracted from the data stream by the
framer logic such that the rising edge of RXCLKx+ occurs when
COMDETx is present on the associated output bus.
This adjustment only occurs when the framer is enabled. When
the framer is disabled, the clock boundaries are not adjusted,
and COMDETx may be asserted during the rising edge of
RXCLKx– (if an odd number of characters were received
following the initial framing).
Signal Name
BYPASS ACTIVE
(DECBYPx = 0)
DECODER
(DECBYP = 1)
RXSTx[2] (LSB)
COMDETx
RXSTx[2]
RXSTx[1]
DOUTx[0]
RXSTx[1]
RXSTx[0]
DOUTx[1]
RXSTx[0]
Receive Status Bits
RXDx[0]
DOUTx[2]
RXDx[0]
RXDx[1]
DOUTx[3]
RXDx[1]
When the 10B/8B decoder is enabled, each character presented
at the Output Register includes three associated status bits.
These bits are used to identify
RXDx[2]
DOUTx[4]
RXDx[2]
RXDx[3]
DOUTx[5]
RXDx[3]
RXDx[4]
DOUTx[6]
RXDx[4]
RXDx[5]
DOUTx[7]
RXDx[6]
RXDx[7] (MSB)
■
if the contents of the data bus are valid,
■
the type of character present,
RXDx[5]
■
the state of receive BIST operations,
DOUTx[8]
RXDx[6]
■
character violations.
DOUTx[9]
RXDx[7]
These conditions often overlap; e.g. a valid data character
received with incorrect running disparity is not reported as a valid
data character. It is instead reported as a decoder violation of
some specific type. This implies a hierarchy or priority level to the
various status bit combinations. The hierarchy and value of each
status are listed in Table 11.
When the 10B/8B decoder is bypassed, the framed 10-bit value
is presented to the associated Output Register, along with a
status output signal indicating if the character in the Output
Register is one of the selected framing characters. The bit usage
and mapping of the external signals to the raw 10B transmission
character is shown in Table 8.
Table 8. Decoder Bypass Mode
Signal Name
Bus Weight
10 Bit Name
RXSTx[2] (LSB)
COMDETx
RXSTx[1]
20
a
RXSTx[0]
21
b
RXDx[0]
22
c
RXDx[1]
23
d
RXDx[2]
24
e
RXDx[3]
25
i
RXDx[4]
26
f
RXDx[5]
27
g
RXDx[6]
28
h
RXDx[7] (MSB)
29
j
The COMDETx status output operates the same regardless of
the bit combination selected for character framing by the
FRAMCHARx latch. COMDETx is HIGH when the character in
the output register contains the selected framing character at the
Document #: 38-02065 Rev. *J
A second status mapping, listed in Table 11, is used when the
receive channel is configured for BIST operation. This status is
used to report receive BIST status and progress.
BIST Status State Machine
When a receive path is enabled to look for and compare the
received data stream with the BIST pattern, the RXSTx[2:0] bits
identify the present state of the BIST compare operation.
The BIST state machine has multiple states, as shown in
Figure 2 and Table 11. When the receive PLL detects an
out-of-lock condition, the BIST state is forced to the Start-of-BIST
state, regardless of the present state of the BIST state machine.
If the number of detected errors ever exceeds the number of
valid matches by greater than 16, the state machine is forced to
the WAIT_FOR_BIST state where it monitors the receive path for
the first character of the next BIST sequence (D0.0). Also, if the
Elasticity Buffer ever hits an overflow/underflow condition, the
status is forced to the BIST_START until the buffer is re-centered
(approximately nine character periods).
To ensure compatibility between the source and destination
systems when operating in BIST modes, the sending and
receiving ends of the link must use the same receive clock
configuration.
Page 21 of 49
CYP15G0403DXB
Device Configuration and Control Interface
The CYP15G0403DXB is highly configurable via the configuration interface. The configuration interface allows the device to
be configured globally or allows each channel to be configured
independently. Table 9 lists the configuration latches within the
device including the initialization value of the latches upon the
assertion of RESET. Table 10 on page 27 shows how the latches
are mapped in the device. Each row in the Table 10 maps to a
8-bit latch bank. There are 16 such write-only latch banks. When
WREN = 0, the logic value in the DATA[7:0] is latched to the latch
bank specified by the values in ADDR[3:0]. The second column
of Table 10 specifies the channels associated with the corresponding latch bank. For example, the first three latch banks (0,1
and 2) consist of configuration bits for channel A. The latch banks
12, 13 and 14 consist of Global configuration bits and the last
latch bank (15) is the Mask latch bank that can be configured to
perform bit-by-bit configuration.
Global Enable Function
The global enable function, controlled by the GLENx bits, is a
feature that can be used to reduce the number of write operations needed to setup the latch banks. This function is beneficial
in systems that use a common configuration in multiple
channels. The GLENx bit is present in bit 0 of latch banks 0
through 11 only. Its default value (1) enables the global update
of the latch bank's contents. Setting the GLENx bit to 0 disables
this functionality.
Latch Banks 12, 13, and 14 are used to load values in the related
latch banks in a global manner. A write operation to latch bank
12 could do a global write to latch banks 0, 3, 6, and 9 depending
on the value of GLENx in these latch banks; latch bank 13 could
do a global write to latch banks 1, 4, 7 and 10; and latch banks
14 could do a global write to latch banks 2, 5, 8 and 11. The
GLENx bit cannot be modified by a global write operation.
for the associated global channel, FGLENx forces the global
update of the target latch banks.
Mask Function
An additional latch bank (15) is used as a global mask vector to
control the update of the configuration latch banks on a bit-by-bit
basis. A logic 1 in a bit location allows for the update of that same
location of the target latch bank(s), whereas a logic 0 disables it.
The reset value of this latch bank is FFh, thereby making its use
optional by default. The mask latch bank is not maskable. The
FGLEN functionality is not affected by the bit 0 value of the mask
latch bank.
Latch Types
There are two types of latch banks: static (S) and dynamic (D).
Each channel is configured by 2 static and 1 dynamic latch
banks. The S type contain those settings that normally do not
change for a given application, whereas the D type controls the
settings that could change dynamically during the application's
lifetime.The first row of latches for each channel (address
numbers 0, 3, 7, and 10) are the static receiver control latches.
The second row of latches for each channel (address numbers
1, 4, 8, and 11) are the static transmitter control latches. The third
row of latches for each channel (address numbers 2, 5, 9, and
12) are the dynamic control latches that are associated with
enabling dynamic functions within the device.
Latch Bank 14 is also useful for those users that do not need the
latch-based programmable feature of the device. This latch bank
could be used in those applications that do not need to modify
the default value of the static latch banks, and that can afford a
global (i.e., not independent) control of the dynamic signals. In
this case, this feature becomes available when ADDR[3:0] is left
unchanged with a value of “1110” and WREN is left asserted. The
signals present in DATA[7:0] effectively become global control
pins, and for the latch banks 2, 5, 8 and 11.
Force Global Enable Function
FGLENx forces the global update of the target latch banks, but
does not change the contents of the GLENx bits. If FGLENx = 1
Table 9. Device Configuration and Control Latch Descriptions
Name
Signal Description
RFMODEA[1:0]
RFMODEB[1:0]
RFMODEC[1:0]
RFMODED[1:0]
Reframe Mode Select. The initialization value of the RFMODEx [1:0] latches = 10. RFMODEx is used to
select the operating mode of the framer. When RFMODEx[1:0] = 00, the low-latency framer is selected. This
frames on each occurrence of the selected framing character(s) in the received data stream. This mode of
framing stretches the recovered clock for one or multiple cycles to align that clock with the recovered data.
When RFMODEx[1:0] = 01, the alternate mode Multi-Byte parallel framer is selected. This requires detection
of the selected framing character(s) in the received serial bit stream, on identical 10-bit boundaries, on four
directly adjacent characters. The recovered character clock remains in the same phasing regardless of
character offset. When RFMODEx[1:0] =10, the Cypress-mode Multi-Byte parallel framer is selected. This
requires a pair of the selected framing character(s), on identical 10-bit boundaries, within a span of 50 bits,
before the character boundaries are adjusted. The recovered character clock remains in the same phasing
regardless of character offset. RFMODEx[1:0] = 11 is reserved for test.
FRAMCHARA
FRAMCHARB
FRAMCHARC
FRAMCHARD
Framing Character Select. The initialization value of the FRAMCHARx latch = 1. FRAMCHARx is used to
select the character or portion of a character used for framing of each channel’s received data stream. When
FRAMCHARx = 1, the framer looks for either disparity of the K28.5 character. When FRAMCHARx = 0, the
framer looks for either disparity of the 8-bit Comma characters. The specific bit combinations of these framing
characters are listed in Table 6 on page 19.
Document #: 38-02065 Rev. *J
Page 22 of 49
CYP15G0403DXB
Table 9. Device Configuration and Control Latch Descriptions (continued)
Name
Signal Description
DECMODEA
DECMODEB
DECMODEC
DECMODED
Receiver Decoder Mode Select. The initialization value of the DECMODEx latch = 1. DECMODEx selects
the Decoder Mode used for the associated channel. When DECMODEx = 1 and decoder is enabled, the
Cypress Decoding Mode is used. When DECMODEx = 0 and decoder is enabled, the Alternate Decoding
mode is used. When the decoder is enabled (DECBYPx = 1), the 10-bit transmission characters are decoded
using Table 15 on page 42 and Table 16 on page 46. The column used in the Special Characters Table 16 is
determined by the DECMODEx latch.
DECBYPA
DECBYPB
DECBYPC
DECBYPD
Receiver Decoder Bypass. The initialization value of the DECBYPx latch = 1. DECBYPx selects if the
Receiver Decoder is enabled or bypassed. When DECBYPx = 1, the decoder is enabled and the Decoder
Mode is selected by DECMODEx. When DECBYPx = 0, the decoder is bypassed and raw 10-bit characters
are passed through the receiver.
RXCKSELA
RXCKSELB
RXCKSELC
RXCKSELD
Receive Clock Select. The initialization value of the RXCKSELx latch = 1. RXCKSELx selects the receive
clock source used to transfer data to the Output Registers and the clock source for the RXCLK± output. When
RXCKSELx = 1, the associated Output Registers, are clocked by REFCLKx± at the associated RXCLKx±
output buffer. When RXCKSELx = 0, the associated Output Registers, are clocked by the Recovered Byte
clock at the associated RXCLKx± output buffer. These output clocks may operate at the character-rate or half
the character-rate as selected by RXRATEx.
RXRATEA
RXRATEB
RXRATEC
RXRATED
Receive Clock Rate Select. The initialization value of the RXRATEx latch = 1. RXRATEx is used to select
the rate of the RXCLKx± clock output.
When RXRATEx = 1 and RXCKSELx = 0, the RXCLKx± clock outputs are complementary clocks that follow
the recovered clock operating at half the character rate. Data for the associated receive channels should be
latched alternately on the rising edge of RXCLKx+ and RXCLKx–.
When RXRATEx = 0 and RXCKSELx = 0, the RXCLKx± clock outputs are complementary clocks that follow
the recovered clock operating at the character rate. Data for the associated receive channels should be latched
on the rising edge of RXCLKx+ or falling edge of RXCLKx–.
When RXRATEx = 1 with RXCKSELx = 1 and REFCLKx± is a full-rate clock, the RXCLKx± clock outputs are
complementary clocks that follow the reference clock operating at half the character rate. Data for the
associated receive channels should be latched alternately on the rising edge of RXCLKx+ and RXCLKx–.
When RXRATEx = 0 with RXCKSELx = 1 and REFCLKx± is a full-rate clock, the RXCLKx± clock outputs are
complementary clocks that follow the reference clock operating at the character rate. Data for the associated
receive channels should be latched on the rising edge of RXCLKx+ or falling edge of RXCLKx–.
When RXCKSELx = 1 and REFCLKx± is a half-rate clock, the value of RXRATEx is not interpreted and the
RXCLKx± clock outputs are complementary clocks that follow the reference clock operating at half the
character rate. Data for the associated receive channels should be latched alternately on the rising edge of
RXCLKx+ and RXCLKx–.
SDASEL1A[1:0]
SDASEL1B[1:0]
SDASEL1C[1:0]
SDASEL1D[1:0]
Primary Serial Data Input Signal Detector Amplitude Select. The initialization value of the SDASEL1x[1:0]
latch = 10. SDASEL1x[1:0] selects the trip point for the detection of a valid signal for the INx1± Primary
Differential Serial Data Inputs.
When SDASEL1x[1:0] = 00, the Analog Signal Detector is disabled.
When SDASEL1x[1:0] = 01, the typical p-p differential voltage threshold level is 140 mV.
When SDASEL1x[1:0] = 10, the typical p-p differential voltage threshold level is 280 mV.
When SDASEL1x[1:0] = 11, the typical p-p differential voltage threshold level is 420 mV.
SDASEL2A[1:0]
SDASEL2B[1:0]
SDASEL2C[1:0]
SDASEL2D[1:0]
Secondary Serial Data Input Signal Detector Amplitude Select. The initialization value of the
SDASEL2x[1:0] latch = 10. SDASEL2x[1:0] selects the trip point for the detection of a valid signal for the INx2±
Secondary Differential Serial Data Inputs.
When SDASEL2x[1:0] = 00, the Analog Signal Detector is disabled
When SDASEL2x[1:0] = 01, the typical p-p differential voltage threshold level is 140 mV.
When SDASEL2x[1:0] = 10, the typical p-p differential voltage threshold level is 280 mV.
When SDASEL2x[1:0] = 11, the typical p-p differential voltage threshold level is 420 mV.
ENCBYPA
ENCBYPB
ENCBYPC
ENCBYPD
Transmit Encoder Bypassed. The initialization value of the ENCBYPx latch = 1. ENCBYPx selects if the
Transmit Encoder is enabled or bypassed. When ENCBYPx = 1, the Transmit encoder is enabled. When
ENCBYPx = 0, the Transmit Encoder is bypassed and raw 10-bit characters are transmitted.
Document #: 38-02065 Rev. *J
Page 23 of 49
CYP15G0403DXB
Table 9. Device Configuration and Control Latch Descriptions (continued)
Name
Signal Description
TXCKSELA
TXCKSELB
TXCKSELC
TXCKSELD
Transmit Clock Select. The initialization value of the TXCKSELx latch = 1. TXCKSELx selects the clock
source used to write data into the Transmit Input Register. When TXCKSELx = 1, the associated input register,
TXDx[7:0] and TXCTx[1:0], is clocked by REFCLKx↑. In this mode, the phase alignment buffer in the transmit
path is bypassed. When TXCKSELx = 0, the associated TXCLKx↑ is used to clock in the input registers,
TXDx[7:0] and TXCTx[1:0].
TXRATEA
TXRATEB
TXRATEC
TXRATED
Transmit PLL Clock Rate Select. The initialization value of the TXRATEx latch = 0. TXRATEx is used to
select the clock multiplier for the Transmit PLL. When TXRATEx = 0, each transmit PLL multiples the
associated REFCLKx± input by 10 to generate the serial bit-rate clock. When TXRATEx = 0, the TXCLKOx
output clocks are full-rate clocks and follow the frequency and duty cycle of the associated REFCLKx± input.
When TXRATEx = 1, each Transmit PLL multiplies the associated REFCLKx± input by 20 to generate the
serial bit-rate clock. When TXRATEx = 1, the TXCLKOx output clocks are twice the frequency rate of the
REFCLKx± input. When TXCKSELx = 1 and TXRATEx = 1, the Transmit Data Inputs are captured using both
the rising and falling edges of REFCLKx. TXRATEx = 1 and SPDSELx is LOW, is an invalid state and this
combination is reserved.
RFENA
RFENB
RFENC
RFEND
Reframe Enable. The initialization value of the RFENx latch = 1. RFENx selects if the receiver framer is
enabled or disabled. When RFENx = 1, the associated channel’s framer is enabled to frame per the presently
enabled framing mode and selected framing character. When RFENx = 0, the associated channel’s framer is
disabled, and no received bits alters the frame offset.
RXPLLPDA
RXPLLPDB
RXPLLPDC
RXPLLPDD
Receive Channel Enable. The initialization value of the RXPLLPDx latch = 0. RXPLLPDx selects if the
associated receive channel is enabled or powered-down. When RXPLLPDx = 0, the associated PLL and
analog circuitry is powered-down. When RXPLLPDx = 1, the associated PLL and analog circuitry is enabled.
RXBISTA
RXBISTB
RXBISTC
RXBISTD
Receive Bist Disabled. The initialization value of the RXBISTx latch = 1. RXBISTx selects if receive BIST is
disabled or enabled. When RXBISTx = 1, the receiver BIST function is disabled. When RXBISTx = 0, the
receive BIST function is enabled.
TXBISTA
TXBISTB
TXBISTC
TXBISTD
Transmit Bist Disabled. The initialization value of the TXBISTx latch = 1. TXBISTx selects if the transmit
BIST is disabled or enabled. When TXBISTx = 1, the transmit BIST function is disabled. When TXBISTx = 0,
the transmit BIST function is enabled.
OE2A
OE2B
OE2C
OE2D
Secondary Differential Serial Data Output Driver Enable. The initialization value of the OE2x latch = 0.
OE2x selects if the OUT2± secondary differential output drivers are enabled or disabled. When OE2x = 1, the
associated serial data output driver is enabled allowing data to be transmitted from the transmit shifter. When
OE2x = 0, the associated serial data output driver is disabled. When a driver is disabled via the configuration
interface, it is internally powered down to reduce device power. If both serial drivers for a channel are in this
disabled state, the associated internal logic for that channel is also powered down. A device reset (RESET
sampled LOW) disables all output drivers.
OE1A
OE1B
OE1C
OE1D
Primary Differential Serial Data Output Driver Enable. The initialization value of the OE1x latch = 0. OE1x
selects if the OUT1± primary differential output drivers are enabled or disabled. When OE1x = 1, the associated
serial data output driver is enabled allowing data to be transmitted from the transmit shifter. When OE1x = 0,
the associated serial data output driver is disabled. When a driver is disabled via the configuration interface,
it is internally powered down to reduce device power. If both serial drivers for a channel are in this disabled
state, the associated internal logic for that channel is also powered down. A device reset (RESET sampled
LOW) disables all output drivers.
PABRSTA
PABRSTB
PABRSTC
PABRSTD
Transmit Clock Phase Alignment Buffer Reset. The initialization value of the PABRSTx latch = 1. The
PABRSTx is used to re-center the Transmit Phase Align Buffer. When the configuration latch PABRSTx is
written as a 0, the phase of the TXCLKx input clock relative to its associated REFCLKx+/- is initialized.
PABRST is an asynchronous input, but is sampled by each TXCLKx↑ to synchronize it to the internal clock
domain. PABRSTx is a self clearing latch. This eliminates the requirement of writing a 1 to complete the
initialization of the Phase Alignment Buffer.
GLEN[11..0]
Global Enable. The initialization value of the GLENx latch = 1. The GLENx is used to reconfigure several
channels simultaneously in applications where several channels may have the same configuration. When
GLENx = 1 for a given address, that address is allowed to participate in a global configuration. When GLENx
= 0 for a given address, that address is disabled from participating in a global configuration.
Document #: 38-02065 Rev. *J
Page 24 of 49
CYP15G0403DXB
Table 9. Device Configuration and Control Latch Descriptions (continued)
Name
FGLEN[2..0]
Signal Description
Force Global Enable. The initialization value of the FGLENx latch is NA. The FGLENx latch forces a GLobal
ENable no matter what the setting is on the GLENx latch. If FGLENx = 1 for the associated Global channel,
FGLEN forces the global update of the target latch banks.
Document #: 38-02065 Rev. *J
Page 25 of 49
CYP15G0403DXB
Device Configuration Strategy
The following is a series of ordered events needed to load the
configuration latches on a per channel basis:
1. Pulse RESET Low after device power-up. This operation
resets all four channels. Initialize the JTAG state machine to
its reset state as detailed in “JTAG Support” on page 27.
2. Set the static receiver latch bank for the target channel. May
be performed using a global operation, if the application
permits it. [Optional step if the default settings match the
desired configuration.]
3. Set the static transmitter latch bank for the target channel.
May be performed using a global operation, if the application
permits it. [Optional step if the default settings match the
desired configuration.]
Document #: 38-02065 Rev. *J
4. Set the dynamic bank of latches for the target channel. Enable
the Receive PLLs and transmit channels. May be performed
using a global operation, if the application permits it.
[Required step.]
5. Reset the Phase Alignment Buffer for the target channel. May
be performed using a global operation, if the application
permits it. [Optional if phase align buffer is bypassed.]
When a receive channel is configured with the decoder
bypassed and the receive clock selected as recovered clock in
half-rate mode (DECBYPx = 0, RXRATEx = 1, RXCKSELx = 0),
the channel cannot be dynamically reconfigured to enable the
decoder with RXCLKx selected as the REFCLKx (DECBYPx =
1, RXCKSELx = 1). If such a change is desired, a global reset
should be performed and all channels should be reconfigured to
the desired settings.
Page 26 of 49
CYP15G0403DXB
Table 10. Device Control Latch Configuration Table
ADDR Channel Type
DATA7
DATA6
DATA5
DATA4
DATA3
DATA2
DATA1
DATA0
Reset
Value
0
(0000b)
A
S
RFMODEA[1]
RFMODEA[0]
FRAMCHARA
DECMODEA
DECBYPA
RXCKSELA
RXRATEA
GLEN0
10111111
1
(0001b)
A
S
SDASEL2A[1]
SDASEL2A[0]
SDASEL1A[1]
SDASEL1A[0]
ENCBYPA
TXCKSELA
TXRATEA
GLEN1
10101101
2
(0010b)
A
D
RFENA
RXPLLPDA
RXBISTA
TXBISTA
OE2A
OE1A
PABRSTA
GLEN2
10110011
3
(0011b)
B
S
RFMODEB[1]
RFMODEB[0]
FRAMCHARB
DECMODEB
DECBYPB
RXCKSELB
RXRATEB
GLEN3
10111111
4
(0100b)
B
S
SDASEL2B[1]
SDASEL2B[0]
SDASEL1B[1]
SDASEL1B[0]
ENCBYPB
TXCKSELB
TXRATEB
GLEN4
10101101
5
(0101b)
B
D
RFENB
RXPLLPDB
RXBISTB
TXBISTB
OE2B
OE1B
PABRSTB
GLEN5
10110011
6
(0110b)
C
S
RFMODEC[1]
RFMODEC[0]
FRAMCHARC
DECMODEC
DECBYPC
RXCKSELC
RXRATEC
GLEN6
10111111
7
(0111b)
C
S
SDASEL2C[1]
SDASEL2C[0]
SDASEL1C[1]
SDASEL1C[0]
ENCBYPC
TXCKSELC
TXRATEC
GLEN7
10101101
8
(1000b)
C
D
RFENC
RXPLLPDC
RXBISTC
TXBISTC
OE2C
OE1C
PABRSTC
GLEN8
10110011
9
(1001b)
D
S
RFMODED[1]
RFMODED[0]
FRAMCHARD
DECMODED
DECBYPD
RXCKSELD
RXRATE D
GLEN9
10111111
10
(1010b)
D
S
SDASEL2D[1]
SDASEL2D[0]
SDASEL1D[1]
SDASEL1D[0]
ENCBYPD
TXCKSELD
TXRATED
GLEN10
10101101
11
(1011b)
D
D
RFEND
RXPLLPDD
RXBISTD
TXBISTD
OE2D
OE1D
PABRSTD
GLEN11
10110011
12
GLOBAL
(1100b)
S
RFMODEGL[1]
RFMODE
GL[0]
FRAMCHARGL DECMODEGL DECBYPGL RXCKSELGL RXRATEG
L
FGLEN0
N/A
13
GLOBAL
(1101b)
S
SDASEL2GL[1] SDASEL2GL[ SDASEL1GL[1] SDASEL1GL[0 ENCBPGL
0]
]
14
GLOBAL
(1110b)
D
RFENGL
RXPLLPDGL
RXBISTGL
TXBISTGL
15
(1111b)
D
D7
D6
D5
D4
MASK
JTAG Support
The CYP15G0403DXB contains a JTAG port to allow system
level diagnosis of device interconnect. Of the available JTAG
modes, boundary scan, and bypass are supported. This
capability is present only on the LVTTL inputs and outputs and
the REFCLKx± clock input. The high-speed serial inputs and
outputs are not part of the JTAG test chain.
To ensure valid device operation after power-up (including
non-JTAG operation), the JTAG state machine should also be
initialized to a reset state. This should be done in addition to the
device reset (using RESET). The JTAG state machine can be
initialized using TRST (asserting it LOW and de-asserting it or
leaving it asserted), or by asserting TMS HIGH for at least 5
consecutive TCLK cycles. This is necessary in order to ensure
Document #: 38-02065 Rev. *J
TXCKSELGL
TXRATEG
L
FGLEN1
N/A
OE2GL
OE1GL
PABRSTG
L
FGLEN2
N/A
D3
D2
D1
D0
11111111
that the JTAG controller does not enter any of the test modes
after device power-up. In this JTAG reset state, the rest of the
device will be in normal operation.
Note. The order of device reset (using RESET) and JTAG initialization does not matter.
3-Level Select Inputs
Each 3-Level select inputs reports as two bits in the scan register.
These bits report the LOW, MID, and HIGH state of the
associated input as 00, 10, and 11 respectively
JTAG ID
The JTAG device ID for the CYP15G0403DXB is ‘0C810069’x.
Page 27 of 49
CYP15G0403DXB
Table 11. Receive Character Status Bits
Description
RXSTx[2:0] Priority
Normal Status
Receive BIST Status
(Receive BIST = Enabled)
000
7
Normal character received. The valid Data
character on the output bus meets all the
formatting requirements of Data characters
listed in Table 15 on page 42.
001
7
Special code detected. The valid special
BIST Command Compare. Character compared
character on the output bus meets all the
correctly.
formatting requirements of Special Code
characters listed in Table 16 on page 46, but is
not the presently selected framing character or a
decoder violation indication.
010
2
Receive Elasticity buffer underrun/overrun
error. The receive buffer was not able to
add/drop a K28.5 or framing character
011
5
Framing character detected. This indicates
that a character matching the patterns identified
as a framing character (as selected by
FRAMCHARx) was detected. The decoded
value of this character is present in the
associated output bus.
100
4
Codeword violation. The character on the
BIST Last Bad. Last Character of BIST sequence
output bus is a C0.7. This indicates that the
detected invalid.
received character cannot be decoded into any
valid character.
101
1
Loss of sync. This indicates a PLL Out of Lock BIST Start. Receive BIST is enabled on this channel,
condition
but character compares have not yet commenced. This
also indicates a PLL Out of Lock condition, and
Elasticity Buffer overflow/underflow conditions.
110
6
Running disparity error. The character on the BIST Error. While comparing characters, a mismatch
output bus is a C4.7, C1.7, or C2.7.
was found in one or more of the decoded character bits.
111
3
Reserved
Document #: 38-02065 Rev. *J
BIST Data Compare. Character compared correctly.
BIST Last Good. Last Character of BIST sequence
detected and valid.
BIST Wait. The receiver is comparing characters. but
has not yet found the start of BIST character to enable
the LFSR.
Page 28 of 49
CYP15G0403DXB
Figure 2. Receive BIST State Machine
Monitor Data
Received
RX PLL
Out of Lock
RXSTx =
BIST_START (101)
RXSTx =
BIST_WAIT (111)
Elasticity
Buffer Error
Yes
No
Receive BIST
Detected LOW
RXSTx =
BIST_START (101)
Start of
BIST Detected
No
Yes, RXSTx =
BIST_DATA_COMPARE (000) / BIST_COMMAND_COMPARE (001)
Compare
Next Character
RXSTx =
Match BIST_COMMAND_COMPARE (001)
Mismatch
Yes
Command
Auto-Abort
Condition
Data or
Command
No
Data
End-of-BIST
State
End-of-BIST
State
Yes, RXSTx =
BIST_LAST_BAD (100)
Yes, RXSTx =
BIST_LAST_GOOD (010)
RXSTx =
BIST_DATA_COMPARE (000)
No
No, RXSTx =
BIST_ERROR (110)
Document #: 38-02065 Rev. *J
Page 29 of 49
CYP15G0403DXB
Maximum Ratings
Static Discharge Voltage.......................................... > 2000 V
(per MIL-STD-883, Method 3015)
Above which the useful life may be impaired. User guidelines
only, not tested
Latch-up Current..................................................... > 200 mA
Storage Temperature .................................. –65°C to +150°C
Power-up Requirements
Ambient Temperature with
Power Applied ............................................ –55°C to +125°C
The CYP15G0403DXB requires one power-supply. The Voltage
on any input or I/O pin cannot exceed the power pin during
power-up.
Supply Voltage to Ground Potential................–0.5V to +3.8V
DC Voltage Applied to LVTTL Outputs
in High-Z State....................................... –0.5V to VCC + 0.5V
Output Current into LVTTL Outputs (LOW) ................. 60 mA
Operating Range
Range
Commercial
Ambient Temperature
0°C to +70°C
VCC
+3.3V ±5%
DC Input Voltage ................................... –0.5V to VCC + 0.5V
DC Electrical Characteristics
Parameter
Description
Test Conditions
Min.
Max.
Unit
LVTTL-compatible Outputs
VOHT
Output HIGH Voltage
IOH = − 4 mA, VCC = Min.
VOLT
Output LOW Voltage
IOL = 4 mA, VCC = Min.
0.4
V
IOST
Output Short Circuit Current
VOUT = 0V[9], VCC = 3.3V
–20
–100
mA
IOZL
High-Z Output Leakage Current
VOUT = 0V, VCC
–20
20
µA
2.4
V
LVTTL-compatible Inputs
VIHT
Input HIGH Voltage
2.0
VCC + 0.3
V
VILT
Input LOW Voltage
–0.5
0.8
V
IIHT
Input HIGH Current
REFCLKx Input, VIN = VCC
1.5
mA
Other Inputs, VIN = VCC
+40
µA
REFCLKx Input, VIN = 0.0V
–1.5
mA
Other Inputs, VIN = 0.0V
–40
µA
IILT
Input LOW Current
IIHPDT
Input HIGH Current with internal pull-down
VIN = VCC
+200
µA
IILPUT
Input LOW Current with internal pull-up
VIN = 0.0V
–200
µA
VCC
mV
LVDIFF Inputs: REFCLKx±
VDIFF[10]
Input Differential Voltage
400
VIHHP
Highest Input HIGH Voltage
1.2
VCC
V
VILLP
Lowest Input LOW voltage
0.0
VCC/2
V
Common Mode Range
1.0
VCC – 1.2V
V
VCOMREF
[11]
3-Level Inputs
VIHH
Three-Level Input HIGH Voltage
Min. ≤ VCC ≤ Max.
0.87 * VCC
VCC
V
VIMM
Three-Level Input MID Voltage
Min. ≤ VCC ≤ Max.
0.47 * VCC
0.53 * VCC
V
VILL
Three-Level Input LOW Voltage
Min. ≤ VCC ≤ Max.
0.0
0.13 * VCC
V
IIHH
Input HIGH Current
VIN = VCC
200
µA
IIMM
Input MID current
VIN = VCC/2
50
µA
IILL
Input LOW current
VIN = GND
–200
µA
–50
Notes
9. Tested one output at a time, output shorted for less than one second, less than 10% duty cycle.
10. This is the minimum difference in voltage between the true and complement inputs required to ensure detection of a logic-1 or logic-0. A logic-1 exists when the
true (+) input is more positive than the complement (−) input. A logic-0 exists when the complement (−) input is more positive than true (+) input.
11. The common mode range defines the allowable range of REFCLKx+ and REFCLKx− when REFCLKx+ = REFCLKx−. This marks the zero-crossing between
the true and complement inputs as the signal switches between a logic-1 and a logic-0.
Document #: 38-02065 Rev. *J
Page 30 of 49
CYP15G0403DXB
DC Electrical Characteristics
Parameter
(continued)
Description
Test Conditions
Min.
Max.
Unit
Differential CML Serial Outputs: OUTA1±, OUTA2±, OUTB1±, OUTB2±, OUTC1±, OUTC2±, OUTD1±, OUTD2±
VOHC
Output HIGH Voltage
(Vcc Referenced)
100Ω differential load
VCC – 0.5
VCC – 0.2
V
150Ω differential load
VCC – 0.5
VCC – 0.2
V
VOLC
Output LOW Voltage
(VCC Referenced)
100Ω differential load
VCC – 1.4
VCC – 0.7
V
150Ω differential load
VCC – 1.4
VCC – 0.7
V
VODIF
Output Differential Voltage
|(OUT+) − (OUT−)|
100Ω differential load
450
900
mV
150Ω differential load
560
1000
mV
1200
mV
VCC
V
1350
μA
Differential Serial Line Receiver Inputs: INA1±, INA2±, INB1±, INB2±, INC1±, INC2±, IND1±, IND2±
VDIFFs[10]
Input Differential Voltage |(IN+) − (IN−)|
VIHE
Highest Input HIGH Voltage
VILE
Lowest Input LOW Voltage
IIHE
Input HIGH Current
VIN = VIHE Max.
IILE
Input LOW Current
VIN = VILE Min.
–700
VICOM[12]
Common Mode input range
((VCC – 2.0V)+0.5)min,
(VCC – 0.5V) max.
+1.25
+3.1
Typ.
Max.
100
VCC – 2.0
Power Supply
ICC
[13, 14]
ICC [13, 14]
V
μA
V
Max Power Supply Current
REFCLKx = Commercial
MAX
910
1270
mA
Typical Power Supply Current
REFCLKx = Commercial
125 MHz
900
1270
mA
AC Test Loads and Waveforms
3.3V
RL = 100Ω
R1
R1 = 590Ω
R2 = 435Ω
CL
CL ≤ 7 pF
(Includes fixture and
probe capacitance)
(Includes fixture and
probe capacitance)
R2
(b) CML Output Test Load
(a) LVTTL Output Test Load
GND
2.0V
2.0V
0.8V
0.8V
[15]
[15]
3.0V
Vth = 1.4V
RL
VIHE
VIHE
Vth = 1.4V
VILE
≤ 1 ns
≤ 1 ns
[16]
(c) LVTTL Input Test Waveform
80%
80%
20%
≤ 270 ps
20%
VILE
≤ 270 ps
(d) CML/LVPECL Input Test Waveform
Notes
12. The common mode range defines the allowable range of INPUT+ and INPUT− when INPUT+ = INPUT−. This marks the zero-crossing between the true and
complement inputs as the signal switches between a logic-1 and a logic-0.
13. Maximum ICC is measured with VCC = MAX, RFENx = 0, TA = 25°C, with all channels and Serial Line Drivers enabled, sending a continuous alternating 01
pattern, and outputs unloaded.
14. Typical ICC is measured under similar conditions except with VCC = 3.3V, TA = 25°C, RFENx = 0, with all channels enabled and one Serial Line Driver per transmit
channel sending a continuous alternating 01 pattern. The redundant outputs on each channel are powered down and the parallel outputs are unloaded.
15. Cypress uses constant current (ATE) load configurations and forcing functions. This figure is for reference only.
16. The LVTTL switching threshold is 1.4V. All timing references are made relative to where the signal edges cross the threshold voltage.
Document #: 38-02065 Rev. *J
Page 31 of 49
CYP15G0403DXB
AC Electrical Characteristics
Parameter
Description
Min.
Max
Unit
Transmitter LVTTL Switching Characteristics Over the Operating Range
fTS
TXCLKx Clock Cycle Frequency
19.5
150
MHz
tTXCLK
TXCLKx Period=1/fTS
6.66
51.28
ns
tTXCLKH[17]
tTXCLKL[17]
tTXCLKR [17, 18, 19, 20]
tTXCLKF [17, 18, 19, 20]
TXCLKx HIGH Time
2.2
TXCLKx LOW Time
2.2
TXCLKx Rise Time
0.2
1.7
ns
TXCLKx Fall Time
0.2
1.7
ns
tTXDS
Transmit Data Set-up Time to TXCLKx↑ (TXCKSELx ≠ 0)
2.2
ns
tTXDH
Transmit Data Hold Time from TXCLKx↑ (TXCKSELx ≠ 0)
1.0
ns
fTOS
TXCLKOx Clock Frequency = 1x or 2x REFCLKx Frequency
19.5
150
MHz
tTXCLKO
TXCLKOx Period=1/fTOS
6.66
51.28
ns
tTXCLKOD
TXCLKO Duty Cycle centered at 60% HIGH time
–1.9
0
ns
ns
ns
Receiver LVTTL Switching Characteristics Over the Operating Range
fRS
RXCLKx± Clock Output Frequency
9.75
150
MHz
tRXCLKP
RXCLKx± Period = 1/fRS
6.66
102.56
ns
tRXCLKD
RXCLKx± Duty Cycle Centered at 50% (Full Rate and Half Rate when
RXCKSELx = 0)
–1.0
+1.0
ns
tRXCLKR [17]
RXCLKx± Rise Time
0.3
1.2
ns
tRXCLKF [17]
RXCLKx± Fall Time
0.3
1.2
ns
tRXDv–
[21]
tRXDv+[21]
Status and Data Valid Time to RXCLKx± (RXRATEx = 0, RXCKSELx =
0) (Full Rate)
5UI –
2.0[22]
ns
Status and Data Valid Time to RXCLKx± (RXRATEx = 1, RXCKSELx =
0) (Half Rate)
5UI – 1.3[22]
ns
Status and Data Valid Time to RXCLKx± (RXRATEx = 0, RXCKSELx =
0) (Full Rate)
5UI–1.8[22]
ns
Status and Data Valid Time to RXCLKx± (RXRATEx = 1, RXCKSELx =0)
(Half Rate)
5UI – 2.6[22]
ns
REFCLKx Switching Characteristics Over the Operating Range
19.5
150[17]
MHz
6.66
51.28
ns
fREF
REFCLKx Clock Frequency
tREFCLK
REFCLKx Period = 1/fREF
tREFH
REFCLKx HIGH Time (TXRATEx = 1)(Half Rate)
5.9[23]
ns
REFCLKx HIGH Time (TXRATEx = 0)(Full Rate)
2.9[17]
ns
REFCLKx LOW Time (TXRATEx = 1)(Half Rate)
5.9
[23]
ns
REFCLKx LOW Time (TXRATEx = 0)(Full Rate)
2.9[17]
ns
tREFL
tREFD[24]
tREFR [17, 18, 19, 20]
REFCLKx Duty Cycle
REFCLKx Rise Time (20%–80%)
30
70
%
2
ns
Notes
17. Tested initially and after any design or process changes that may affect these parameters, but not 100% tested.
18. The ratio of rise time to falling time must not vary by greater than 2:1.
19. For a given operating frequency, neither rise or fall specification can be greater than 20% of the clock-cycle period or the data sheet maximum time.
20. All transmit AC timing parameters measured with 1 ns typical rise time and fall time.
21. Parallel data output specifications are only valid if all outputs are loaded with similar DC and AC loads.
22. Receiver UI (Unit Interval) is calculated as 1/(fREF * 20) (when TXRATEx = 1) or 1/(fREF * 10) (when TXRATEx = 0). In an operating link this is equivalent to tB.
23. If REFCLKx is selected as receive interface clock (RXCKSELx=1), then this parameter has to be greater than or equal to 6.3 ns.
24. The duty cycle specification is a simultaneous condition with the tREFH and tREFL parameters. This means that at faster character rates the REFCLKx± duty
cycle cannot be as large as 30%–70%.
Document #: 38-02065 Rev. *J
Page 32 of 49
CYP15G0403DXB
AC Electrical Characteristics
(continued)
Parameter
Description
Min.
Max
Unit
2
ns
tREFF[17, 18, 19, 20]
REFCLKx Fall Time (20%–80%)
tTREFDS
Transmit Data Set-up Time to REFCLKx - Full Rate
(TXRATEx = 0, TXCKSELx = 1)
2.4
ns
Transmit Data Set-up Time to REFCLKx - Half Rate
(TXRATEx = 1, TXCKSELx = 1)
2.3
ns
Transmit Data Hold Time from REFCLKx - Full Rate
(TXRATEx = 0, TXCKSELx = 1)
1.0
ns
Transmit Data Hold Time from REFCLKx - Half Rate
(TXRATEx = 1, TXCKSELx = 1)
1.6
ns
tTREFDH
tRREFDA
Receive Data Access Time to REFCLKx (RXCKSELx = 1)
tRREFDW
Receive Data Valid Time Window (RXCKSELx = 1)
tREFxDV–
Received Data Valid Time to RXCLK when RXCKSELx = 1
(TXRATEx = 0, RXRATEx = 0)
10UI[22]
– 6.16
ns
Received Data Valid Time to RXCLK when RXCKSELx = 1
(TXRATEx = 0, RXRATEx = 1)
5UI – 2.53[26]
ns
Received Data Valid Time to RXCLK when RXCKSELx = 1
(TXRATEx = 1)
10UI – 5.86[26]
ns
Received Data Valid Time from RXCLK when RXCKSELx = 1
(TXRATEx = 0, RXRATEx = 0)
1.4
ns
Received Data Valid Time from RXCLK when RXCKSELx = 1
(TXRATEx = 0, RXRATEx = 1)
5UI – 1.83[26]
ns
Received Data Valid Time from RXCLK when RXCKSELx = 1
(TXRATEx = 1)
1.0[26]
ns
REFCLKx Frequency Referenced to Received Clock Period
–0.15
tREFxDV+
tREFRX[27]
9.7[25]
10UI – 5.8
ns
ns
+0.15
%
Bus Configuration Write Timing Characteristics Over the Operating Range
tDATAH
Bus Configuration Data Hold
0
ns
tDATAS
Bus Configuration Data Set-up
10
ns
tWRENP
Bus Configuration WREN Pulse Width
10
ns
JTAG Test Clock Characteristics Over the Operating Range
fTCLK
JTAG Test Clock Frequency
tTCLK
JTAG Test Clock Period
20
MHz
50
ns
30
ns
Device RESET Characteristics Over the Operating Range
tRST
Device RESET Pulse Width
Transmit Serial Outputs and TX PLL Characteristics Over the Operating Range
Parameter
tB
Description
Bit Time
Condition
Min.
Max.
Unit
5128
666
ps
Notes
25. Since this timing parameter is greater than the minimum time period of REFCLK it sets an upper limit to the frequency in which REFCLKx can be used to clock
the receive data out of the output register. For predictable timing, users can use this parameter only if REFCLK period is greater than sum of tRREFDA and set-up
time of the upstream device. When this condition is not true, RXCLKx± (a buffered or divided version of REFCLK when RXCKSELx = 1) could be used to clock
the receive data out of the device.
26. Measured using a 50% duty cycle reference clock.
27. REFCLKx has no phase or frequency relationship with the recovered clock and only acts as a centering reference to reduce clock synchronization time. REFCLKx
must be within ±1500 ppm (±0.15%) of the remote transmitter’s PLL reference (REFCLKx) frequency. Although transmitting to a HOTLink II receiver channel
necessitates the frequency difference between the transmitter and receiver reference clocks to be within ±1500 ppm, the stability of the crystal needs to be
within the limits specified by the appropriate standard when transmitting to a remote receiver that is compliant to that standard. For example, to be IEEE 802.3z
Gigabit Ethernet compliant, the frequency stability of the crystal needs to be within ±100 ppm.l.
Document #: 38-02065 Rev. *J
Page 33 of 49
CYP15G0403DXB
AC Electrical Characteristics
(continued)
Parameter
tRISE[17]
tFALL[17]
Description
CML Output Rise Time 20−80% (CML Test Load)
CML Output Fall Time 80−20% (CML Test Load)
Min.
Max
Unit
SPDSELx = HIGH
60
270
ps
SPDSELx = MID
100
500
ps
SPDSELx =LOW
180
1000
ps
SPDSELx = HIGH
60
270
ps
SPDSELx = MID
100
500
ps
SPDSELx =LOW
180
1000
ps
Deterministic Jitter (peak-peak)[31]
IEEE 802.3z
27
ps
tRJ[17, 29, 30]
tREFJ[17]
Random Jitter (σ)
IEEE 802.3z
11
ps
REFCLKx jitter tolerance / Phase noise limits
TBD
tTXLOCK
Transmit PLLx lock to REFCLKx±
200
μs
Receive PLL lock to input data stream (cold start)
376k
UI
Receive PLL lock to input data stream
376k
UI
46
UI
tDJ [17, 28, 30]
[31]
Receive Serial Inputs and CDR PLL Characteristics Over the Operating Range
tRXLOCK
tRXUNLOCK
tJTOL
[17]
tDJTOL
[17]
Receive PLL Unlock Rate
Total Jitter
Tolerance[31]
Deterministic Jitter
Tolerance[31]
IEEE 802.3z
600
ps
IEEE 802.3z
370
ps
Capacitance[17]
Max.
Unit
CINTTL
Parameter
TTL Input Capacitance
Description
TA = 25°C, f0 = 1 MHz, VCC = 3.3V
Test Conditions
7
pF
CINPECL
PECL input Capacitance
TA = 25°C, f0 = 1 MHz, VCC = 3.3V
4
pF
Notes
28. While sending continuous K28.5s, outputs loaded to a balanced 100Ω load, measured at the cross point of differential outputs, over the operating range.
29. While sending continuous K28.7s, after 100,000 samples measured at the cross point of differential outputs, time referenced to REFCLKx± input, over the operating
range.
30. Total jitter is calculated at an assumed BER of 1E −12. Hence: Total Jitter (tJ) = (tRJ * 14) + tDJ.
31. Also meets all Jitter Generation and Jitter Tolerance requirements as specified by ESCON, FICON, Fibre Channel, and DVB-ASI.
Document #: 38-02065 Rev. *J
Page 34 of 49
CYP15G0403DXB
HOTLink II Transmitter Switching Waveforms
tTXCLK
Transmit Interface
Write Timing
TXCLKx selected
tTXCLKH
tTXCLKL
TXCLKx
tTXDS
tTXDH
TXDx[7:0],
TXCTx[1:0],
Transmit Interface
Write Timing
REFCLKx selected
TXRATEx = 0
tREFCLK
tREFH
tREFL
REFCLKx
tTREFDS
tTREFDH
TXDx[7:0],
TXCTx[1:0],
Transmit Interface
Write Timing
REFCLKx selected
TXRATEx = 1
tREFCLK
tREFH
tREFL
REFCLKx
Note 32
tTREFDS
TXDx[7:0],
TXCTx[1:0],
Transmit Interface
TXCLKOx Timing
tTREFDH
tTREFDS
tTREFDH
tREFCLK
tREFH
TXRATEx = 1
REFCLKx
tREFL
Note 33
tTXCLKO
Note 34
TXCLKOx
(internal)
Notes
32. When REFCLKx± is configured for half-rate operation (TXRATE = 1) and data is captured using REFCLKx instead of a TXCLKx clock. Data is captured using
both the rising and falling edges of REFCLKx.
33. The TXCLKOx output remains at the character rate regardless of the state of TXRATE and does not follow the duty cycle of REFCLKx±.
34. The rising edge of TXCLKOx output has no direct phase relationship to the REFCLKx± input.
Document #: 38-02065 Rev. *J
Page 35 of 49
CYP15G0403DXB
HOTLink II Transmitter Switching Waveforms
Transmit Interface
TXCLKOx Timing
(continued)
tREFCLK
tREFH
TXRATEx = 0
tREFL
Note33
REFCLKx
tTXCLKO
Note34
tTXOH
tTXOL
TXCLKOx
Switching Waveforms for the HOTLink II Receiver
Receive Interface
Read Timing
REFCLKx Selected
Full-rate RXCLKx±
tREFCLK
tREFH
tREFL
REFCLKx
tRREFDA
tRREFDW
tRREFDW
RXDx[7:0],
RXSTx[2:0],
[36]
TXERRx
tREFxDV+
tREFxDV–
RXCLKx
Receive Interface
Read Timing
REFCLKx Selected
Half-rate RXCLKx±
tREFCLK
tREFH
tREFL
REFCLKx
tRREFDA
tRREFDW
tRREFDA
tRREFDW
RXDx[7:0],
RXSTx[2:0],
TXERRx
[36]
tREFxDV+
RXCLKx
tREFxDV–
Note 35
Notes
35. When operated with a half-rate REFCLKx±, the set-up and hold specifications for data relative to RXCLKx are relative to both rising and falling edges of the
respective clock output
36. TXERRx is synchronous to RXCLKx only when RXCLKx is selected as REFCLK.
Document #: 38-02065 Rev. *J
Page 36 of 49
CYP15G0403DXB
Switching Waveforms for the HOTLink II Receiver
Receive Interface
Read Timing
Recovered Clock selected
RXRATEx = 0
tRXCLKP
RXCLKx+
RXCLKx–
RXDx[7:0],
RXSTx[2:0],
tRXDV–
tRXDV+
Receive Interface
Read Timing
Recovered Clock selected
RXRATEx = 1
tRXCLKP
RXCLKx+
RXCLKx–
tRXDV–
RXDx[7:0],
RXSTx[2:0]
tRXDV+
Bus Configuration
Write Timing
ADDR[3:0]
DATA[7:0]
tWRENP
WREN
tDATAS
tDATAH
Document #: 38-02065 Rev. *J
Page 37 of 49
CYP15G0403DXB
Table 12. Package Coordinate Signal Allocation
Ball
ID
A01
A02
A03
A04
A05
A06
A07
A08
A09
A10
A11
A12
A13
A14
A15
A16
A17
A18
A19
A20
B01
B02
B03
B04
B05
B06
B07
B08
B09
B10
B11
B12
B13
B14
B15
B16
B17
B18
B19
B20
C01
C02
C03
C04
C05
C06
Signal Name
Signal Type
INC1–
OUTC1–
INC2–
OUTC2–
VCC
IND1–
OUTD1–
GND
IND2–
OUTD2–
INA1–
OUTA1–
GND
INA2–
OUTA2–
VCC
INB1–
OUTB1–
INB2–
OUTB2–
INC1+
OUTC1+
INC2+
OUTC2+
VCC
IND1+
OUTD1+
GND
IND2+
OUTD2+
INA1+
OUTA1+
GND
INA2+
OUTA2+
VCC
INB1+
OUTB1+
INB2+
OUTB2+
TDI
TMS
INSELC
INSELB
VCC
ULCD
CML IN
CML OUT
CML IN
CML OUT
POWER
CML IN
CML OUT
GROUND
CML IN
CML OUT
CML IN
CML OUT
GROUND
CML IN
CML OUT
POWER
CML IN
CML OUT
CML IN
CML OUT
CML IN
CML OUT
CML IN
CML OUT
POWER
CML IN
CML OUT
GROUND
CML IN
CML OUT
CML IN
CML OUT
GROUND
CML IN
CML OUT
POWER
CML IN
CML OUT
CML IN
CML OUT
LVTTL IN PU
LVTTL IN PU
LVTTL IN
LVTTL IN
POWER
LVTTL IN PU
Document #: 38-02065 Rev. *J
Ball
ID
C07
C08
C09
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
D01
D02
D03
D04
D05
D06
D07
D08
D09
D10
D11
D12
D13
D14
D15
D16
D17
D18
D19
D20
E01
E02
E03
E04
E17
E18
E19
E20
F01
F02
F03
F04
Signal Name
Signal Type
ULCC
GND
DATA[7]
DATA[5]
DATA[3]
DATA[1]
GND
NC
SPDSELD
VCC
LDTDEN
TRST
LPEND
TDO
TCLK
RESET
INSELD
INSELA
VCC
ULCA
SPDSELC
GND
DATA[6]
DATA[4]
DATA[2]
DATA[0]
GND
LPENB
ULCB
VCC
LPENA
LTEN1
SCANEN2
TMEN3
VCC
VCC
VCC
VCC
VCC
VCC
VCC
VCC
RXDC[6]
RXDC[7]
TXDC[0]
NC
LVTTL IN PU
GROUND
LVTTL IN PU
LVTTL IN PU
LVTTL IN PU
LVTTL IN PU
GROUND
NO CONNECT
3-LEVEL SEL
POWER
LVTTL IN PU
LVTTL IN PU
LVTTL IN PD
LVTTL 3-S OUT
LVTTL IN PD
LVTTL IN PU
LVTTL IN
LVTTL IN
POWER
LVTTL IN PU
3-LEVEL SEL
GROUND
LVTTL IN PU
LVTTL IN PU
LVTTL IN PU
LVTTL IN PU
GROUND
LVTTL IN PD
LVTTL IN PU
POWER
LVTTL IN PD
LVTTL IN PD
LVTTL IN PD
LVTTL IN PD
POWER
POWER
POWER
POWER
POWER
POWER
POWER
POWER
LVTTL OUT
LVTTL OUT
LVTTL IN
NO CONNECT
Ball
ID
F17
F18
F19
F20
G01
G02
G03
G04
G17
G18
G19
G20
H01
H02
H03
H04
H17
H18
H19
H20
J01
J02
J03
J04
J17
J18
J19
J20
K01
K02
K03
K04
K17
K18
K19
K20
L01
L02
L03
L04
L17
L18
L19
L20
M01
M02
Signal Name
Signal Type
NC
RXSTB[1]
TXCLKOB
RXSTB[0]
TXDC[7]
WREN
TXDC[4]
TXDC[1]
SPDSELB
LPENC
SPDSELA
RXDB[1]
GND
GND
GND
GND
GND
GND
GND
GND
TXCTC[1]
TXDC[5]
TXDC[2]
TXDC[3]
RXSTB[2]
RXDB[0]
RXDB[5]
RXDB[2]
RXDC[2]
REFCLKC–
TXCTC[0]
TXCLKC
RXDB[3]
RXDB[4]
RXDB[7]
LFIB
RXDC[3]
REFCLKC+
LFIC
TXDC[6]
RXDB[6]
RXCLKB+
RXCLKB–
TXDB[6]
RXDC[4]
RXDC[5]
NO CONNECT
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL IN
LVTTL IN PU
LVTTL IN
LVTTL IN
3-LEVEL SEL
LVTTL IN PD
3-LEVEL SEL
LVTTL OUT
GROUND
GROUND
GROUND
GROUND
GROUND
GROUND
GROUND
GROUND
LVTTL IN
LVTTL IN
LVTTL IN
LVTTL IN
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL OUT
PECL IN
LVTTL IN
LVTTL IN PD
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL OUT
PECL IN
LVTTL OUT
LVTTL IN
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL IN
LVTTL OUT
LVTTL OUT
Page 38 of 49
CYP15G0403DXB
Table 12. Package Coordinate Signal Allocation (continued)
Ball
ID
M03
M04
M17
M18
M19
M20
N01
N02
N03
N04
N17
N18
N19
N20
P01
P02
P03
P04
P17
P18
P19
P20
R01
R02
R03
R04
R17
R18
R19
R20
T01
T02
T03
T04
T17
T18
T19
T20
U01
U02
Signal Name
Signal Type
NC
TXERRC
REFCLKB+
REFCLKB–
TXERRB
TXCLKB
GND
GND
GND
GND
GND
GND
GND
GND
RXDC[1]
RXDC[0]
RXSTC[0]
RXSTC[1]
TXDB[5]
TXDB[4]
TXDB[3]
TXDB[2]
RXSTC[2]
TXCLKOC
RXCLKC+
RXCLKC–
TXDB[1]
TXDB[0]
TXCTB[1]
TXDB[7]
VCC
VCC
VCC
VCC
VCC
VCC
VCC
VCC
TXDD[0]
TXDD[1]
NO CONNECT
LVTTL OUT
PECL IN
PECL IN
LVTTL OUT
LVTTL IN PD
GROUND
GROUND
GROUND
GROUND
GROUND
GROUND
GROUND
GROUND
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL IN
LVTTL IN
LVTTL IN
LVTTL IN
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL IN
LVTTL IN
LVTTL IN
LVTTL IN
POWER
POWER
POWER
POWER
POWER
POWER
POWER
POWER
LVTTL IN
LVTTL IN
Document #: 38-02065 Rev. *J
Ball Signal Name
ID
U03
TXDD[2]
U04
TXCTD[1]
U05
VCC
U06
RXDD[2]
U07
RXDD[1]
U08
GND
U09
TXCTA[1]
U10
ADDR [0]
U11
REFCLKD–
U12
TXDA[1]
U13
GND
U14
TXDA[4]
U15
TXCTA[0]
U16
VCC
U17
RXDA[2]
U18
TXCTB[0]
U19
RXSTA[2]
U20
RXSTA[1]
V01
TXDD[3]
V02
TXDD[4]
V03
TXCTD[0]
V04
RXDD[6]
V05
VCC
V06
RXDD[3]
V07
RXSTD[0]
V08
GND
V09
RXSTD[2]
V10
ADDR [2]
V11
REFCLKD+
V12
TXCLKOA
V13
GND
V14
TXDA[3]
V15
TXDA[7]
V16
VCC
V17
RXDA[7]
V18
RXDA[3]
V19
RXDA[0]
V20
RXSTA[0]
W01
TXDD[5]
W02
TXDD[7]
Signal Type
LVTTL IN
LVTTL IN
POWER
LVTTL OUT
LVTTL OUT
GROUND
LVTTL IN
LVTTL IN PU
PECL IN
LVTTL IN
GROUND
LVTTL IN
LVTTL IN
POWER
LVTTL OUT
LVTTL IN
LVTTL OUT
LVTTL OUT
LVTTL IN
LVTTL IN
LVTTL IN
LVTTL OUT
POWER
LVTTL OUT
LVTTL OUT
GROUND
LVTTL OUT
LVTTL IN PU
PECL IN
LVTTL OUT
GROUND
LVTTL IN
LVTTL IN
POWER
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL OUT
LVTTL IN
LVTTL IN
Ball Signal Name
ID
W03
LFID
W04
RXCLKD–
W05
VCC
W06
RXDD[4]
W07
RXSTD[1]
W08
GND
W09
ADDR [3]
W10
ADDR [1]
W11
RXCLKA+
W12
TXERRA
W13
GND
W14
TXDA[2]
W15
TXDA[6]
W16
VCC
W17
LFIA
W18 REFCLKA+
W19
RXDA[4]
W20
RXDA[1]
Y01
TXDD[6]
Y02
TXCLKD
Y03
RXDD[7]
Y04
RXCLKD+
Y05
VCC
Y06
RXDD[5]
Y07
RXDD[0]
Y08
GND
Y09
TXCLKOD
Y10
NC
Y11
TXCLKA
Y12
RXCLKA–
Y13
GND
Y14
TXDA[0]
Y15
TXDA[5]
Y16
VCC
Y17
TXERRD
Y18
REFCLKA–
Y19
RXDA[6]
Y20
RXDA[5]
Signal Type
LVTTL OUT
LVTTL OUT
POWER
LVTTL OUT
LVTTL OUT
GROUND
LVTTL IN PU
LVTTL IN PU
LVTTL OUT
LVTTL OUT
GROUND
LVTTL IN
LVTTL IN
POWER
LVTTL OUT
PECL IN
LVTTL OUT
LVTTL OUT
LVTTL IN
LVTTL IN PD
LVTTL OUT
LVTTL OUT
POWER
LVTTL OUT
LVTTL OUT
GROUND
LVTTL OUT
NO CONNECT
LVTTL IN PD
LVTTL OUT
GROUND
LVTTL IN
LVTTL IN
POWER
LVTTL OUT
PECL IN
LVTTL OUT
LVTTL OUT
Page 39 of 49
CYP15G0403DXB
X3.230 Codes and Notation Conventions
Information transmitted over a serial link is encoded eight bits at
a time into a 10-bit Transmission Character and then sent
serially, bit by bit. Information received over a serial link is
collected ten bits at a time, and those Transmission Characters
that are used for data characters are decoded into the correct
eight-bit codes. The 10-bit Transmission Code supports all 256
8-bit combinations. Some of the remaining Transmission
Characters (Special Characters) are used for functions other
than data transmission.
The primary use of a Transmission Code is to improve the transmission characteristics of a serial link. The encoding defined by
the Transmission Code ensures that sufficient transitions are
present in the serial bit stream to make clock recovery possible
at the Receiver. Such encoding also greatly increases the
likelihood of detecting any single or multiple bit errors that may
occur during transmission and reception of information. In
addition, some Special Characters of the Transmission Code
selected by Fibre Channel Standard contain a distinct and easily
recognizable bit pattern that assists the receiver in achieving
character alignment on the incoming bit stream.
Notation Conventions
The documentation for the 8B/10B Transmission Code uses
letter notation for the bits in an 8-bit byte. Fibre Channel Standard
notation uses a bit notation of A, B, C, D, E, F, G, H for the 8-bit
byte for the raw 8-bit data, and the letters a, b, c, d, e, i, f, g, h, j
for encoded 10-bit data. There is a correspondence between bit
A and bit a, B and b, C and c, D and d, E and e, F and f, G and
g, and H and h. Bits i and j are derived, respectively, from
(A,B,C,D,E) and (F,G,H).
The bit labeled A in the description of the 8B/10B Transmission
Code corresponds to bit 0 in the numbering scheme of the FC-2
specification, B corresponds to bit 1, as shown below.
FC-2 bit designation—76543210
HOTLink D/Q designation—76543210
8B/10B bit designation—HGFEDCBA
To clarify this correspondence, the following example shows the
conversion from an FC-2 Valid Data Byte to a Transmission
Character.
FC-2 45H
Bits: 7654 3210
0100 0101
Converted to 8B/10B notation, note that the order of bits has
been reversed):
Data Byte Name D5.2
Bits: ABCDE FGH
10100 010
Translated to a transmission Character in the 8B/10B Transmission Code:
Bits: abcdei fghj
101001 0101
Each valid Transmission Character of the 8B/10B Transmission
Code has been given a name using the following convention:
cxx.y, where c is used to show whether the Transmission
Character is a Data Character (c is set to D, and SC/D = LOW)
or a Special Character (c is set to K, and SC/D = HIGH). When
Document #: 38-02065 Rev. *J
c is set to D, xx is the decimal value of the binary number
composed of the bits E, D, C, B, and A in that order, and the y is
the decimal value of the binary number composed of the bits H,
G, and F in that order. When c is set to K, xx and y are derived
by comparing the encoded bit patterns of the Special Character
to those patterns derived from encoded Valid Data bytes and
selecting the names of the patterns most similar to the encoded
bit patterns of the Special Character.
Under the above conventions, the Transmission Character used
for the examples above, is referred to by the name D5.2. The
Special Character K29.7 is so named because the first six bits
(abcdei) of this character make up a bit pattern similar to that
resulting from the encoding of the unencoded 11101 pattern (29),
and because the second four bits (fghj) make up a bit pattern
similar to that resulting from the encoding of the unencoded 111
pattern (7).
Note. This definition of the 10-bit Transmission Code is based
on the following references, which describe the same 10-bit
transmission code.
A.X. Widmer and P.A. Franaszek. “A DC-Balanced, Partitioned-Block, 8B/10B Transmission Code” IBM Journal of
Research and Development, 27, No. 5: 440-451 (September, 1983).
U.S. Patent 4,486,739. Peter A. Franaszek and Albert X.
Widmer. “Byte-Oriented DC Balanced (0.4) 8B/10B Partitioned
Block Transmission Code” (December 4, 1984).
Fibre Channel Physical and Signaling Interface (ANS
X3.230-1994 ANSI FC-PH Standard).
IBM Enterprise Systems Architecture/390 ESCON I/O Interface
(document number SA22-7202).
8B/10B Transmission Code
The following information describes how the tables are used for
both generating valid Transmission Characters (encoding) and
checking the validity of received Transmission Characters
(decoding). It also specifies the ordering rules followed when
transmitting the bits within a character and the characters within
any higher-level constructs specified by a standard.
Transmission Order
Within the definition of the 8B/10B Transmission Code, the bit
positions of the Transmission Characters are labeled a, b, c, d,
e, i, f, g, h, j. Bit “a” is transmitted first followed by bits b, c, d, e,
i, f, g, h, and j in that order.
Note that bit i is transmitted between bit e and bit f, rather than
in alphabetical order.
Valid and Invalid Transmission Characters
The following tables define the valid Data Characters and valid
Special Characters (K characters), respectively. The tables are
used for both generating valid Transmission Characters and
checking the validity of received Transmission Characters. In the
tables, each Valid-Data-byte or Special-Character-code entry
has two columns that represent two Transmission Characters.
The two columns correspond to the current value of the running
disparity. Running disparity is a binary parameter with either a
negative (–) or positive (+) value.
After powering on, the Transmitter may assume either a positive
or negative value for its initial running disparity. Upon transmission of any Transmission Character, the transmitter selects
Page 40 of 49
CYP15G0403DXB
Character transmitted, a new value of the running disparity is
calculated. This new value is used as the Transmitter’s current
running disparity for the next Valid Data byte or Special
Character byte encoded and transmitted. Table 13 shows
naming notations and examples of valid transmission
characters.
the proper version of the Transmission Character based on the
current running disparity value, and the Transmitter calculates a
new value for its running disparity based on the contents of the
transmitted character. Special Character codes C1.7 and C2.7
can be used to force the transmission of a specific Special
Character with a specific running disparity as required for some
special sequences in X3.230.
Use of the Tables for Checking the Validity of Received
Transmission Characters
After powering on, the Receiver may assume either a positive or
negative value for its initial running disparity. Upon reception of
any Transmission Character, the Receiver decides whether the
Transmission Character is valid or invalid according to the
following rules and tables and calculates a new value for its
Running Disparity based on the contents of the received
character.
The column corresponding to the current value of the Receiver’s
running disparity is searched for the received Transmission
Character. If the received Transmission Character is found in the
proper column, then the Transmission Character is valid and the
associated Data byte or Special Character code is determined
(decoded). If the received Transmission Character is not found
in that column, then the Transmission Character is invalid. This
is called a code violation. Independent of the Transmission
Character’s validity, the received Transmission Character is
used to calculate a new value of running disparity. The new value
is used as the Receiver’s current running disparity for the next
received Transmission Character.
The following rules for running disparity are used to calculate the
new running-disparity value for Transmission Characters that
have been transmitted and received.
Running disparity for a Transmission Character is calculated
from sub-blocks, where the first six bits (abcdei) form one
sub-block and the second four bits (fghj) form the other
sub-block. Running disparity at the beginning of the 6-bit
sub-block is the running disparity at the end of the previous
Transmission Character. Running disparity at the beginning of
the 4-bit sub-block is the running disparity at the end of the 6-bit
sub-block. Running disparity at the end of the Transmission
Character is the running disparity at the end of the 4-bit
sub-block.
Table 13. Valid Transmission Characters
Data
Byte Name
Running disparity for the sub-blocks is calculated as follows:
1. Running disparity at the end of any sub-block is positive if the
sub-block contains more ones than zeros. It is also positive at
the end of the 6-bit sub-block if the 6-bit sub-block is 000111,
and it is positive at the end of the 4-bit sub-block if the 4-bit
sub-block is 0011.
2. Running disparity at the end of any sub-block is negative if the
sub-block contains more zeros than ones. It is also negative
at the end of the 6-bit sub-block if the 6-bit sub-block is
111000, and it is negative at the end of the 4-bit sub-block if
the 4-bit sub-block is 1100.
3. Otherwise, running disparity at the end of the sub-block is the
same as at the beginning of the sub-block.
Use of the Tables for Generating Transmission
Characters
DIN or QOUT
Hex Value
765
43210
D0.0
000
00000
00
D1.0
000
00001
01
D2.0
000
00010
02
.
.
.
.
.
.
.
.
D5.2
010
00101
45
.
.
.
.
.
.
.
.
D30.7
111
11110
FE
D31.7
111
11111
FF
Detection of a code violation does not necessarily show that the
Transmission Character in which the code violation was detected
is in error. Code violations may result from a prior error that
altered the running disparity of the bit stream which did not result
in a detectable error at the Transmission Character in which the
error occurred. Table 14 shows an example of this behavior.
The appropriate entry in Table 15 for the Valid Data byte or
Table 16 for Special Character byte identify which Transmission
Character is generated. The current value of the Transmitter’s
running disparity is used to select the Transmission Character
from its corresponding column. For each Transmission
Table 14. Code Violations Resulting from Prior Errors
RD
Character
RD
Character
RD
Character
RD
Transmitted data character
–
D21.1
–
D10.2
–
D23.5
+
Transmitted bit stream
–
101010 1001
–
010101 0101
–
111010 1010
+
Bit stream after error
–
101010 1011
+
010101 0101
+
111010 1010
+
Decoded data character
–
D21.0
+
D10.2
+
Code Violation
+
Document #: 38-02065 Rev. *J
Page 41 of 49
CYP15G0403DXB
Table 15. Valid Data Characters (TXCTx[0] = 0, RXSTx[2:0] = 000)
Data
Byte
Name
Bits
Current RD−
Current RD+
Bits
Current RD−
Current RD+
abcdei fghj
Data
Byte
Name
HGF EDCBA
abcdei fghj
HGF EDCBA
abcdei fghj
abcdei fghj
D0.0
000 00000
100111 0100
011000 1011
D0.1
001 00000
100111 1001
011000 1001
D1.0
000 00001
011101 0100
100010 1011
D1.1
001 00001
011101 1001
100010 1001
D2.0
000 00010
101101 0100
010010 1011
D2.1
001 00010
101101 1001
010010 1001
D3.0
000 00011
110001 1011
110001 0100
D3.1
001 00011
110001 1001
110001 1001
D4.0
000 00100
110101 0100
001010 1011
D4.1
001 00100
110101 1001
001010 1001
D5.0
000 00101
101001 1011
101001 0100
D5.1
001 00101
101001 1001
101001 1001
D6.0
000 00110
011001 1011
011001 0100
D6.1
001 00110
011001 1001
011001 1001
D7.0
000 00111
111000 1011
000111 0100
D7.1
001 00111
111000 1001
000111 1001
D8.0
000 01000
111001 0100
000110 1011
D8.1
001 01000
111001 1001
000110 1001
D9.0
000 01001
100101 1011
100101 0100
D9.1
001 01001
100101 1001
100101 1001
D10.0
000 01010
010101 1011
010101 0100
D10.1
001 01010
010101 1001
010101 1001
D11.0
000 01011
110100 1011
110100 0100
D11.1
001 01011
110100 1001
110100 1001
D12.0
000 01100
001101 1011
001101 0100
D12.1
001 01100
001101 1001
001101 1001
D13.0
000 01101
101100 1011
101100 0100
D13.1
001 01101
101100 1001
101100 1001
D14.0
000 01110
011100 1011
011100 0100
D14.1
001 01110
011100 1001
011100 1001
D15.0
000 01111
010111 0100
101000 1011
D15.1
001 01111
010111 1001
101000 1001
D16.0
000 10000
011011 0100
100100 1011
D16.1
001 10000
011011 1001
100100 1001
D17.0
000 10001
100011 1011
100011 0100
D17.1
001 10001
100011 1001
100011 1001
D18.0
000 10010
010011 1011
010011 0100
D18.1
001 10010
010011 1001
010011 1001
D19.0
000 10011
110010 1011
110010 0100
D19.1
001 10011
110010 1001
110010 1001
D20.0
000 10100
001011 1011
001011 0100
D20.1
001 10100
001011 1001
001011 1001
D21.0
000 10101
101010 1011
101010 0100
D21.1
001 10101
101010 1001
101010 1001
D22.0
000 10110
011010 1011
011010 0100
D22.1
001 10110
011010 1001
011010 1001
D23.0
000 10111
111010 0100
000101 1011
D23.1
001 10111
111010 1001
000101 1001
D24.0
000 11000
110011 0100
001100 1011
D24.1
001 11000
110011 1001
001100 1001
D25.0
000 11001
100110 1011
100110 0100
D25.1
001 11001
100110 1001
100110 1001
D26.0
000 11010
010110 1011
010110 0100
D26.1
001 11010
010110 1001
010110 1001
D27.0
000 11011
110110 0100
001001 1011
D27.1
001 11011
110110 1001
001001 1001
D28.0
000 11100
001110 1011
001110 0100
D28.1
001 11100
001110 1001
001110 1001
D29.0
000 11101
101110 0100
010001 1011
D29.1
001 11101
101110 1001
010001 1001
D30.0
000 11110
011110 0100
100001 1011
D30.1
001 11110
011110 1001
100001 1001
D31.0
000 11111
101011 0100
010100 1011
D31.1
001 11111
101011 1001
010100 1001
Document #: 38-02065 Rev. *J
Page 42 of 49
CYP15G0403DXB
Table 15. Valid Data Characters (TXCTx[0] = 0, RXSTx[2:0] = 000) (continued)
Data
Byte
Name
Bits
Current RD−
Current RD+
Bits
Current RD−
Current RD+
abcdei fghj
Data
Byte
Name
HGF EDCBA
abcdei fghj
HGF EDCBA
abcdei fghj
abcdei fghj
D0.2
010 00000
100111 0101
011000 0101
D0.3
011 00000
100111 0011
011000 1100
D1.2
010 00001
011101 0101
100010 0101
D1.3
011 00001
011101 0011
100010 1100
D2.2
010 00010
101101 0101
010010 0101
D2.3
011 00010
101101 0011
010010 1100
D3.2
010 00011
110001 0101
110001 0101
D3.3
011 00011
110001 1100
110001 0011
D4.2
010 00100
110101 0101
001010 0101
D4.3
011 00100
110101 0011
001010 1100
D5.2
010 00101
101001 0101
101001 0101
D5.3
011 00101
101001 1100
101001 0011
D6.2
010 00110
011001 0101
011001 0101
D6.3
011 00110
011001 1100
011001 0011
D7.2
010 00111
111000 0101
000111 0101
D7.3
011 00111
111000 1100
000111 0011
D8.2
010 01000
111001 0101
000110 0101
D8.3
011 01000
111001 0011
000110 1100
D9.2
010 01001
100101 0101
100101 0101
D9.3
011 01001
100101 1100
100101 0011
D10.2
010 01010
010101 0101
010101 0101
D10.3
011 01010
010101 1100
010101 0011
D11.2
010 01011
110100 0101
110100 0101
D11.3
011 01011
110100 1100
110100 0011
D12.2
010 01100
001101 0101
001101 0101
D12.3
011 01100
001101 1100
001101 0011
D13.2
010 01101
101100 0101
101100 0101
D13.3
011 01101
101100 1100
101100 0011
D14.2
010 01110
011100 0101
011100 0101
D14.3
011 01110
011100 1100
011100 0011
D15.2
010 01111
010111 0101
101000 0101
D15.3
011 01111
010111 0011
101000 1100
D16.2
010 10000
011011 0101
100100 0101
D16.3
011 10000
011011 0011
100100 1100
D17.2
010 10001
100011 0101
100011 0101
D17.3
011 10001
100011 1100
100011 0011
D18.2
010 10010
010011 0101
010011 0101
D18.3
011 10010
010011 1100
010011 0011
D19.2
010 10011
110010 0101
110010 0101
D19.3
011 10011
110010 1100
110010 0011
D20.2
010 10100
001011 0101
001011 0101
D20.3
011 10100
001011 1100
001011 0011
D21.2
010 10101
101010 0101
101010 0101
D21.3
011 10101
101010 1100
101010 0011
D22.2
010 10110
011010 0101
011010 0101
D22.3
011 10110
011010 1100
011010 0011
D23.2
010 10111
111010 0101
000101 0101
D23.3
011 10111
111010 0011
000101 1100
D24.2
010 11000
110011 0101
001100 0101
D24.3
011 11000
110011 0011
001100 1100
D25.2
010 11001
100110 0101
100110 0101
D25.3
011 11001
100110 1100
100110 0011
D26.2
010 11010
010110 0101
010110 0101
D26.3
011 11010
010110 1100
010110 0011
D27.2
010 11011
110110 0101
001001 0101
D27.3
011 11011
110110 0011
001001 1100
D28.2
010 11100
001110 0101
001110 0101
D28.3
011 11100
001110 1100
001110 0011
D29.2
010 11101
101110 0101
010001 0101
D29.3
011 11101
101110 0011
010001 1100
D30.2
010 11110
011110 0101
100001 0101
D30.3
011 11110
011110 0011
100001 1100
D31.2
010 11111
101011 0101
010100 0101
D31.3
011 11111
101011 0011
010100 1100
Document #: 38-02065 Rev. *J
Page 43 of 49
CYP15G0403DXB
Table 15. Valid Data Characters (TXCTx[0] = 0, RXSTx[2:0] = 000) (continued)
Data
Byte
Name
Bits
Current RD−
Current RD+
Bits
Current RD−
Current RD+
abcdei fghj
Data
Byte
Name
HGF EDCBA
abcdei fghj
HGF EDCBA
abcdei fghj
abcdei fghj
D0.4
100 00000
100111 0010
011000 1101
D0.5
101 00000
100111 1010
011000 1010
D1.4
100 00001
011101 0010
100010 1101
D1.5
101 00001
011101 1010
100010 1010
D2.4
100 00010
101101 0010
010010 1101
D2.5
101 00010
101101 1010
010010 1010
D3.4
100 00011
110001 1101
110001 0010
D3.5
101 00011
110001 1010
110001 1010
D4.4
100 00100
110101 0010
001010 1101
D4.5
101 00100
110101 1010
001010 1010
D5.4
100 00101
101001 1101
101001 0010
D5.5
101 00101
101001 1010
101001 1010
D6.4
100 00110
011001 1101
011001 0010
D6.5
101 00110
011001 1010
011001 1010
D7.4
100 00111
111000 1101
000111 0010
D7.5
101 00111
111000 1010
000111 1010
D8.4
100 01000
111001 0010
000110 1101
D8.5
101 01000
111001 1010
000110 1010
D9.4
100 01001
100101 1101
100101 0010
D9.5
101 01001
100101 1010
100101 1010
D10.4
100 01010
010101 1101
010101 0010
D10.5
101 01010
010101 1010
010101 1010
D11.4
100 01011
110100 1101
110100 0010
D11.5
101 01011
110100 1010
110100 1010
D12.4
100 01100
001101 1101
001101 0010
D12.5
101 01100
001101 1010
001101 1010
D13.4
100 01101
101100 1101
101100 0010
D13.5
101 01101
101100 1010
101100 1010
D14.4
100 01110
011100 1101
011100 0010
D14.5
101 01110
011100 1010
011100 1010
D15.4
100 01111
010111 0010
101000 1101
D15.5
101 01111
010111 1010
101000 1010
D16.4
100 10000
011011 0010
100100 1101
D16.5
101 10000
011011 1010
100100 1010
D17.4
100 10001
100011 1101
100011 0010
D17.5
101 10001
100011 1010
100011 1010
D18.4
100 10010
010011 1101
010011 0010
D18.5
101 10010
010011 1010
010011 1010
D19.4
100 10011
110010 1101
110010 0010
D19.5
101 10011
110010 1010
110010 1010
D20.4
100 10100
001011 1101
001011 0010
D20.5
101 10100
001011 1010
001011 1010
D21.4
100 10101
101010 1101
101010 0010
D21.5
101 10101
101010 1010
101010 1010
D22.4
100 10110
011010 1101
011010 0010
D22.5
101 10110
011010 1010
011010 1010
D23.4
100 10111
111010 0010
000101 1101
D23.5
101 10111
111010 1010
000101 1010
D24.4
100 11000
110011 0010
001100 1101
D24.5
101 11000
110011 1010
001100 1010
D25.4
100 11001
100110 1101
100110 0010
D25.5
101 11001
100110 1010
100110 1010
D26.4
100 11010
010110 1101
010110 0010
D26.5
101 11010
010110 1010
010110 1010
D27.4
100 11011
110110 0010
001001 1101
D27.5
101 11011
110110 1010
001001 1010
D28.4
100 11100
001110 1101
001110 0010
D28.5
101 11100
001110 1010
001110 1010
D29.4
100 11101
101110 0010
010001 1101
D29.5
101 11101
101110 1010
010001 1010
D30.4
100 11110
011110 0010
100001 1101
D30.5
101 11110
011110 1010
100001 1010
D31.4
100 11111
101011 0010
010100 1101
D31.5
101 11111
101011 1010
010100 1010
Document #: 38-02065 Rev. *J
Page 44 of 49
CYP15G0403DXB
Table 15. Valid Data Characters (TXCTx[0] = 0, RXSTx[2:0] = 000) (continued)
Data
Byte
Name
Bits
Current RD−
Current RD+
Bits
Current RD−
Current RD+
abcdei fghj
Data
Byte
Name
HGF EDCBA
abcdei fghj
HGF EDCBA
abcdei fghj
abcdei fghj
D0.6
110 00000
100111 0110
011000 0110
D0.7
111 00000
100111 0001
011000 1110
D1.6
110 00001
011101 0110
100010 0110
D1.7
111 00001
011101 0001
100010 1110
D2.6
110 00010
101101 0110
010010 0110
D2.7
111 00010
101101 0001
010010 1110
D3.6
110 00011
110001 0110
110001 0110
D3.7
111 00011
110001 1110
110001 0001
D4.6
110 00100
110101 0110
001010 0110
D4.7
111 00100
110101 0001
001010 1110
D5.6
110 00101
101001 0110
101001 0110
D5.7
111 00101
101001 1110
101001 0001
D6.6
110 00110
011001 0110
011001 0110
D6.7
111 00110
011001 1110
011001 0001
D7.6
110 00111
111000 0110
000111 0110
D7.7
111 00111
111000 1110
000111 0001
D8.6
110 01000
111001 0110
000110 0110
D8.7
111 01000
111001 0001
000110 1110
D9.6
110 01001
100101 0110
100101 0110
D9.7
111 01001
100101 1110
100101 0001
D10.6
110 01010
010101 0110
010101 0110
D10.7
111 01010
010101 1110
010101 0001
D11.6
110 01011
110100 0110
110100 0110
D11.7
111 01011
110100 1110
110100 1000
D12.6
110 01100
001101 0110
001101 0110
D12.7
111 01100
001101 1110
001101 0001
D13.6
110 01101
101100 0110
101100 0110
D13.7
111 01101
101100 1110
101100 1000
D14.6
110 01110
011100 0110
011100 0110
D14.7
111 01110
011100 1110
011100 1000
D15.6
110 01111
010111 0110
101000 0110
D15.7
111 01111
010111 0001
101000 1110
D16.6
110 10000
011011 0110
100100 0110
D16.7
111 10000
011011 0001
100100 1110
D17.6
110 10001
100011 0110
100011 0110
D17.7
111 10001
100011 0111
100011 0001
D18.6
110 10010
010011 0110
010011 0110
D18.7
111 10010
010011 0111
010011 0001
D19.6
110 10011
110010 0110
110010 0110
D19.7
111 10011
110010 1110
110010 0001
D20.6
110 10100
001011 0110
001011 0110
D20.7
111 10100
001011 0111
001011 0001
D21.6
110 10101
101010 0110
101010 0110
D21.7
111 10101
101010 1110
101010 0001
D22.6
110 10110
011010 0110
011010 0110
D22.7
111 10110
011010 1110
011010 0001
D23.6
110 10111
111010 0110
000101 0110
D23.7
111 10111
111010 0001
000101 1110
D24.6
110 11000
110011 0110
001100 0110
D24.7
111 11000
110011 0001
001100 1110
D25.6
110 11001
100110 0110
100110 0110
D25.7
111 11001
100110 1110
100110 0001
D26.6
110 11010
010110 0110
010110 0110
D26.7
111 11010
010110 1110
010110 0001
D27.6
110 11011
110110 0110
001001 0110
D27.7
111 11011
110110 0001
001001 1110
D28.6
110 11100
001110 0110
001110 0110
D28.7
111 11100
001110 1110
001110 0001
D29.6
110 11101
101110 0110
010001 0110
D29.7
111 11101
101110 0001
010001 1110
D30.6
110 11110
011110 0110
100001 0110
D30.7
111 11110
011110 0001
100001 1110
D31.6
110 11111
101011 0110
010100 0110
D31.7
111 11111
101011 0001
010100 1110
Document #: 38-02065 Rev. *J
Page 45 of 49
CYP15G0403DXB
Table 16. Valid Special Character Codes and Sequences (TXCTx = special character code or RXSTx[2:0] = 001)[37, 38]
S.C. Byte Name
Cypress
S.C. Code Name
S.C. Byte Name
[39]
Alternate
Bits
S.C. Byte Name
[39]
HGF EDCBA
Bits
HGF EDCBA
Current RD−
abcdei fghj
Current RD+
abcdei fghj
K28.0
C0.0
(C00)
000 00000
C28.0
(C1C)
000 11100
001111 0100
110000 1011
K28.1[40]
C1.0
(C01)
000 00001
C28.1
(C3C)
001 11100
001111 1001
110000 0110
[40]
C2.0
(C02)
000 00010
C28.2
(C5C)
010 11100
001111 0101
110000 1010
C3.0
(C03)
000 00011
C28.3
(C7C)
011 11100
001111 0011
110000 1100
K28.2
K28.3
K28.4
[40]
C4.0
(C04)
000 00100
C28.4
(C9C)
100 11100
001111 0010
110000 1101
K28.5[40, 41]
C5.0
(C05)
000 00101
C28.5
(CBC)
101 11100
001111 1010
110000 0101
K28.6[40]
C6.0
(C06)
000 00110
C28.6
(CDC)
110 11100
001111 0110
110000 1001
K28.7[40, 42]
C7.0
(C07)
000 00111
C28.7
(CFC)
111 11100
001111 1000
110000 0111
K23.7
C8.0
(C08)
000 01000
C23.7
(CF7)
111 10111
111010 1000
000101 0111
K27.7
C9.0
(C09)
000 01001
C27.7
(CFB)
111 11011
110110 1000
001001 0111
K29.7
C10.0
(C0A)
000 01010
C29.7
(CFD)
111 11101
101110 1000
010001 0111
K30.7
C11.0
(C0B)
000 01011
C30.7
(CFE)
111 11110
011110 1000
100001 0111
(C22)
001 00010
C2.1
(C22)
001 00010
End of Frame Sequence
EOFxx
C2.1
−K28.5,Dn.xxx0[43] +K28.5,Dn.xxx1[43]
Code Rule Violation and SVS Tx Pattern
Exception[42, 44]
C0.7
(CE0)
111 00000
C0.7
(CE0)
111 00000
−K28.5[45]
C1.7
(CE1)
111 00001
C1.7
(CE1)
111 00001
001111 1010
001111 1010
+K28.5[46]
C2.7
(CE2)
111 00010
C2.7
(CE2)
111 00010
110000 0101
110000 0101
111 00100
C4.7
(CE4)
111 00100
110111 0101
001000 1010
100111 1000
011000 0111
Running Disparity Violation Pattern
Exception[47]
C4.7
(CE4)
Notes
37. All codes not shown are reserved.
38. Notation for Special Character Code Name is consistent with Fibre Channel and ESCON naming conventions. Special Character Code Name is intended to
describe binary information present on I/O pins. Common usage for the name can either be in the form used for describing Data patterns (i.e., C0.0 through
C31.7), or in hex notation (i.e., Cnn where nn = the specified value between 00 and FF).
39. Both the Cypress and alternate encodings may be used for data transmission to generate specific Special Character Codes. The decoding process for received
characters generates Cypress codes or Alternate codes as selected by the BOE[7:0] configuration inputs.
40. These characters are used for control of ESCON interfaces. They can be sent as embedded commands or other markers when not operating using ESCON
protocols.
41. The K28.5 character is used for framing operations by the receiver. It is also the pad or fill character transmitted to maintain the serial link when no user data is
available.
42. Care must be taken when using this Special Character code. When a C7.0 or a C0.7 is followed by a D11.x or D20.x, an alias K28.5 sync character is created.
These sequences can cause erroneous framing and should be avoided while RFENx = 1.
43. C2.1 = Transmit either –K28.5+ or +K28.5– as determined by Current RD and modify the Transmission Character that follows, by setting its least significant bit
to 1 or 0. If Current RD at the start of the following character is plus (+) the LSB is set to 0, and if Current RD is minus (–) the LSB becomes 1. This modification
allows construction of X3.230 “EOF” frame delimiters wherein the second data byte is determined by the Current RD.
For example, to send “EOFdt” the controller could issue the sequence C2.1–D21.4– D21.4–D21.4, and the HOTLink Transmitter sends either
K28.5–D21.4–D21.4–D21.4 or K28.5–D21.5– D21.4–D21.4 based on Current RD. Likewise to send “EOFdti” the controller could issue the sequence
C2.1–D10.4–D21.4–D21.4, and the HOTLink Transmitter sends either K28.5–D10.4–D21.4– D21.4 or K28.5–D10.5–D21.4–D21.4 based on Current RD.
The receiver never outputs this Special Character, since K28.5 is decoded as C5.0, C1.7, or C2.7, and the subsequent bytes are decoded as data.
44. C0.7 = Transmit a deliberate code rule violation. The code chosen for this function follows the normal Running Disparity rules. Transmission of this Special
Character has the same effect as asserting TXSVS = HIGH. The receiver only outputs this Special Character if the Transmission Character being decoded is
not found in the tables.
45. C1.7 = Transmit Negative K28.5 (−K28.5+) disregarding Current RD. The receiver only outputs this Special Character if K28.5 is received with the wrong running
disparity. The receiver outputs C1.7 if −K28.5 is received with RD+, otherwise K28.5 is decoded as C5.0 or C2.7.
46. C2.7 = Transmit Positive K28.5 (+K28.5−) disregarding Current RD. The receiver only outputs this Special Character if K28.5 is received with the wrong running
disparity. The receiver outputs C2.7 if +K28.5 is received with RD−, otherwise K28.5 is decoded as C5.0 or C1.7.
47. C4.7 = Transmit a deliberate code rule violation to indicate a Running Disparity violation. The receiver only outputs this Special Character if the Transmission
Character being decoded is found in the tables, but Running Disparity does not match. This might indicate that an error occurred in a prior byte.
Document #: 38-02065 Rev. *J
Page 46 of 49
CYP15G0403DXB
Ordering Information
Speed
Standard
Ordering Code
Package
Name
CYP15G0403DXB-BGXC
BL256
Package Type
Pb-free 256-Ball Thermally Enhanced Ball Grid Array
Operating
Range
Commercial
Ordering Code Definition
CY P 15G 04 03 DXB - BG X C
C - Temperature Grade: commercial
X - Pb-free
BG - Package Type, 256 Ball Grid Array
DXB - Silicon revision
03 - Independent Clocking for each channel available.
04 - No. of Channels
15G - Serial Data rate.
P - Not compliant to SMPTE
CY - Company ID: CY = Cypress.
Package Diagram
Figure 3. 256-Lead L2 Ball Grid Array (27 x 27 x 1.57 mm) BL256
51-85123 *I
Document #: 38-02065 Rev. *J
Page 47 of 49
CYP15G0403DXB
Document History Page
Document Title: CYP15G0403DXB Independent Clock Quad HOTLink II™ Transceiver
Document Number: 38-02065
REV.
ECN NO.
ISSUE
DATE
ORIG. OF
CHANGE
DESCRIPTION OF CHANGE
**
118422
09/24/02
LNM
New Data Sheet
*A
125289
04/04/03
CGX
Revised entire data sheet. Redefined device.
*B
128692
08/14/03
PDS
Provided AC timing information for TXERRx
Added additional information regarding the availability of half-rate RXCLKx±
when REFCLKx is a full-rate clock with RXCKSELx = 1
Added influence of ULCx input on LFIx status
Added influence of DECMODEx on decoder bypass
Revised the text for “Device Configuration and Control Interface” for better
clarity. Removed the timing parameter tRREFDV and added the timing
parameter tRREFDW instead. This change was done to provide a more
meaningful timing parameter. Revised tRREFDA from 9.5 ns to 9.7 ns
Added additional information to “Device Configuration Strategy”
*C
234390
See ECN
PDS
Removed dependence of DECMODEx on decoder bypass.
Revised AC timing parameters (AC Electrical Characteristics) to match final
device characterization.
Expanded the CDR Range Controller’s permissible frequency offset between
incoming serial signalling rate and Reference clock from ±200-PPM to
±1500-PPM (changed parameter tREFRX).
*D
338721
See ECN
SUA
Added CYW15G0403DXB part number for OBSAI RP3 compliance to support
operating data rate up to 1540 MBaud. Made changes to reflect OBSAI RP3
and CPR compliance. Added Pb-Free Package option for all parts listed in the
data sheet. Modified Timing Parameters
Changed MBd to MBaud in SPDSEL pin description
*E
384307
See ECN
AGT
Revised setup and hold time parameters (tTXDH, tTREFDS, tTREFDH, tRXDv–,
tRXDv+, tRXDV+, tREFxDV–, tREFxDV+)
*F
1034001
See ECN
UKK
Added clarification for the necessity of JTAG controller reset and the methods
to implement it.
*G
2897246
03/22/10
CGX
Removed inactive parts from Ordering Information.Updated package diagram.
*H
2952699
06/15/10
CGX
Removed references to inactive part numbers in the datasheet.
*I
3277398
06/10/11
SAAC/
CGX
*J
4401100
06/09/2014
YLIU
Document #: 38-02065 Rev. *J
Updated Template. Removed existence of CYV and SMPTE from datasheet
as CYV15G0403DXB part is obsolete.
Updated package diagram to current revision.
Updated links in Sales, Solutions, and Legal Information per template.
Page 48 of 49
CYP15G0403DXB
Sales, Solutions, and Legal Information
Worldwide Sales and Design Support
Cypress maintains a worldwide network of offices, solution centers, manufacturer’s representatives, and distributors. To find the office
closest to you, visit us at Cypress Locations.
PSoC® Solutions
Products
Automotive
Clocks & Buffers
Interface
Lighting & Power Control
cypress.com/go/automotive
cypress.com/go/clocks
cypress.com/go/interface
cypress.com/go/powerpsoc
psoc.cypress.com/solutions
PSoC 1 | PSoC 3 | PSoC 4 | PSoC 5LP
Cypress Developer Community
Community | Forums | Blogs | Video | Training
cypress.com/go/plc
Memory
PSoC
Touch Sensing
USB Controllers
Wireless/RF
cypress.com/go/memory
cypress.com/go/psoc
Technical Support
cypress.com/go/support
cypress.com/go/touch
cypress.com/go/USB
cypress.com/go/wireless
© Cypress Semiconductor Corporation, 2002-2014. The information contained herein is subject to change without notice. Cypress Semiconductor Corporation assumes no responsibility for the use of
any circuitry other than circuitry embodied in a Cypress product. Nor does it convey or imply any license under patent or other rights. Cypress products are not warranted nor intended to be used for
medical, life support, life saving, critical control or safety applications, unless pursuant to an express written agreement with Cypress. Furthermore, Cypress does not authorize its products for use as
critical components in life-support systems where a malfunction or failure may reasonably be expected to result in significant injury to the user. The inclusion of Cypress products in life-support systems
application implies that the manufacturer assumes all risk of such use and in doing so indemnifies Cypress against all charges.
Any Source Code (software and/or firmware) is owned by Cypress Semiconductor Corporation (Cypress) and is protected by and subject to worldwide patent protection (United States and foreign),
United States copyright laws and international treaty provisions. Cypress hereby grants to licensee a personal, non-exclusive, non-transferable license to copy, use, modify, create derivative works of,
and compile the Cypress Source Code and derivative works for the sole purpose of creating custom software and or firmware in support of licensee product to be used only in conjunction with a Cypress
integrated circuit as specified in the applicable agreement. Any reproduction, modification, translation, compilation, or representation of this Source Code except as specified above is prohibited without
the express written permission of Cypress.
Disclaimer: CYPRESS MAKES NO WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, WITH REGARD TO THIS MATERIAL, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. Cypress reserves the right to make changes without further notice to the materials described herein. Cypress does not
assume any liability arising out of the application or use of any product or circuit described herein. Cypress does not authorize its products for use as critical components in life-support systems where
a malfunction or failure may reasonably be expected to result in significant injury to the user. The inclusion of Cypress’ product in a life-support systems application implies that the manufacturer
assumes all risk of such use and in doing so indemnifies Cypress against all charges.
Use may be limited by and subject to the applicable Cypress software license agreement.
Document #: 38-02065 Rev. *J
Revised June 9, 2014
Page 49 of 49
HOTLink is a registered trademark and HOTLink II and MultiFrame are trademarks of Cypress Semiconductor. CPRI is a trademark of Siemens AG. IBM and ESCON are registered trademarks, and
FICON is a trademark, of International Business Machines. All products and company names mentioned in this document may be the trademarks of their respective holders.