Cypress CYV15G0404DXB User Manual

CYV15G0404DXB
Independent Clock Quad HOTLink II™
Transceiver with Reclocker

Features

Video Coprocessor
Serial Links
10
10
10
10
10
10
10
10
Video Coprocessor
10
10
10
10
10
10
10
10
Serial Links
Serial Links
Serial Links
Cable
Connections
Independent
CYV15G0404DXB
Independent
Reclocker
Reclocker
Channel
CYV15G0404DXB
Channel
Quad channel transceiver for 195 to 1500 MBaud serial
Aggregate throughput of up to 12 Gbits/second
Second-generation HOTLink
Compliant to multiple standardsSMPTE-292M, SMPTE-259M, DVB-ASI, Fibre Channel, ES-
CON, and Gigabit Ethernet (IEEE802.3z)
10 bit uncoded data or 8B/10B coded data
Truly independent channelsEach channel is able to:
• Perform reclocker function
• Operate at a different signaling rate
• Transport a different data format
Internal phase-locked loops (PLLs) with no external PLL
components
Selectable differential PECL compatible serial inputs per
channel
Internal DC restoration
Redundant differential PECL compatible serial outputs per
channel
No external bias resistors requiredSignaling rate controlled edge ratesSource matched for 50Ω transmission lines
MultiFrame™ Receive Framer provides alignment optionsComma or full K28.5 detectSingle or multibyte Framer for byte alignment
Low latency option
Selectable input and output clocking options
®
technology
Figure 1. HOTLink II™ System Connections
Synchronous LVTTL parallel interface
JTAG boundary scan
Built In Self Test (BIST) for at-speed link testing
Link quality indicator by channelAnalog signal detectDigital signal detect
Low power 3W at 3.3V typical
Single 3.3V supply
256 ball thermally enhanced BGA
0.25μ BiCMOS technology
JTAG device ID ‘0C811069’x

Functional Description

The CYV15G0404DXB Independent Clock Quad HOTLink II™ Transceiver is a point-to-point or point-to-multipoint communica­tions building block enabling the transfer of data over a variety of high speed serial links including SMPTE 292, SMPTE 259, and DVB-ASI video applications. The signaling rate can be anywhere in the range of 195 to 1500 MBaud for each 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. The received serial data can also be reclocked and retransmitted through the serial outputs. Figure 1 illustrates typical connections between independent video coprocessors and corresponding CYV15G0404DXB chips.
Cypress Semiconductor Corporation 198 Champion Court San Jose, CA 95134-1709 408-943-2600 Document #: 38-02097 Rev. *B Revised December 14, 2007
[+] Feedback [+] Feedback
CYV15G0404DXB
The CYV15G0404DXB satisfies the SMPTE-259M and

CYV15G0404DXB T ransceiver Logic Block Diagram

x10
Serializer
Phase
Encoder
8B/10B
Decoder
8B/10B
x11
Framer
Deserializer
TX
RX
x10
Serializer
Encoder
8B/10B
Decoder
8B/10B
x11
Framer
Deserializer
TX
RX
x10
Serializer
Encoder
8B/10B
Decoder
8B/10B
x11
Framer
Deserializer
TX
RX
x10
Serializer
Encoder
8B/10B
Decoder
8B/10B
x11
Framer
Deserializer
TX
RX
TXDA[7:0]
RXDA[7:0]
TXDB[7:0]
RXDB[7:0]
TXDC[7:0]
RXDC[7:0]
TXDD[7:0]
RXDD[7:0]
OUTA1±
OUTA2±
INA1±
INA2±
OUTB1±
OUTB2±
INB1±
INB2±
OUTC1±
OUTC2±
INC1±
INC2±
OUTD1±
OUTD2±
IND1±
IND2±
Align
Buffer
Phase
Align
Buffer
Phase
Align
Buffer
Phase
Align
Buffer
Elasticity
Buffer
Elasticity
Buffer
Elasticity
Buffer
Elasticity
Buffer
TXCTA[1:0]
RXSTA[2:0]
TXCTB[1:0]
RXSTB[2:0]
TXCTC[1:0]
RXSTC[2:0]
TXCTD[1:0]
RXSTD[2:0]
REFCLKA±
REFCLKB±
REFCLKC±
REFCLKD±
SMPTE-292M compliance according to SMPTE EG34-1999 Pathological Test Requirements.
As a second generation HOTLink device, the CYV15G0404DXB 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 CYV15G0404DXB Quad HOTLink II consists of four independent byte-wide channels. Each channel accepts 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.
The receive (RX) section of the CYV15G0404DXB 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 or decoder may be bypassed for systems that present externally encoded or scrambled data at the parallel interface.
The parallel IO interface may be configured for numerous forms of clocking to provide the highest flexibility in system archi­tecture. 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 CYV15G0404DXB is ideal for port applications where different data rates and serial interface standards are necessary for each channel. Some applications include multi-format routers and switchers.
Document #: 38-02097 Rev. *B Page 2 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Shifter
TXLBA
TXLBC

Transmit Path Block Diagram

TXRATEA
Input
Register
Phase-Align
Buffer
Encoder
BIST LFSR
SPDSELA
REFCLKA+ REFCLKA–
Transmit PLL
Clock Multiplier
TXCLKA
Bit-Rate Clock
Character-Rate Clock A
OUTA1+ OUTA1–
OUTA2+ OUTA2–
8
TXRATEB
Input
Register
Phase-Align
Buffer
Encoder
BIST LFSR
Shifter
SPDSELB
REFCLKB+
REFCLKB–
Bit-Rate Clock
Character-Rate Clock B
OUTB1+ OUTB1–
OUTB2+ OUTB2–
Input
Register
Phase-Align
Buffer
8B/10B
BIST LFSR
Transmit PLL
Clock Multiplier A
Input
Register
Phase-Align
Buffer
8B/10B
BIST LFSR
Shifter
TXCLKB
TXRATEC
Input
Register
Phase-Align
Buffer
8B/10B
BIST LFSR
SPDSELC
REFCLKC+ REFCLKC–
TXCLKC
Bit-Rate Clock
Character-Rate Clock C
TXRATED
Input
Register
Phase-Align
Buffer
8B/10B
BIST LFSR
Shifter
SPDSELD
REFCLKD+
REFCLKD–
Transmit PLL
Clock Multiplier D
TXCLKD
Bit-Rate Clock
Character-Rate Clock D
OUTD1+ OUTD1–
OUTD2+ OUTD2–
OUTC1+ OUTC1–
OUTC2+ OUTC2–
TXCTA[1:0]
TXDD[7:0]
OEA[2..1]
TXBIST
ENCBYPA
TXCKSELA
= Internal Signal
TXERRA
TXERRB
TXERRD
TXERRC
TXCLKOA
TXCLKOB
TXCLKOC
TXCLKOD
TXDA[7:0]
2
TXDB[7:0]
8
2
TXCTB[1:0]
8
2
TXDC[7:0]
TXCTC[1:0]
8
2
TXCTD[1:0]
10
10
10
10
10 10 10 10
10
10
10 10
10 10
10
10
A
ENCBYPB
ENCBYPC
ENCBYPD
TXBIST
B
TXBIST
C
TXBISTD
OEB[2..1]
OEC[2..1]
OED[2..1]
PABRSTA
PABRSTB
PABRSTC
PABRSTD
OEA[2..1]
OEB[2..1]
OEC[2..1]
OED[2..1]
TXLBD
Shifter
TXLBB
Transmit PLL
Clock Multiplier B
Transmit PLL
Clock Multiplier C
10
TXCKSELB
0
TXCKSELC
10
TXCKSELD
10
RECLCK[A..D] are Internal Reclocker Signals
Encoder
Encoder
Encoder
Encoder
1
RECLCKA
RECLCKB
RECLCKC
RECLCKD
TXLB[A..D] are Internal Serial Loopback Signals
Document #: 38-02097 Rev. *B Page 3 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
INA1+ INA1–
INA2+ INA2–
INSELA
INB1+ INB1–
INB2+ INB2–
INSELB
INC1+ INC1–
INC2+ INC2–
INSELC
IND1+ IND1–
IND2+ IND2–
INSELD
Clock &
Data
Recovery
PLL
Shifter
Clock &
Data
Recovery
PLL
Shifter
Clock &
Data
Recovery
PLL
Shifter
Clock &
Data
Recovery
PLL
Shifter
LFID
LFIC
LFIB
LFIA
8
RXSTC[2:0]
RXDC[7:0]
3
8
RXSTB[2:0]
RXDB[7:0]
3
8
RXSTD[2:0]
RXDD[7:0]
3
8
RXSTA[2:0]
RXDA[7:0]
3
Receive
Signal
Monitor
Receive
Signal
Monitor
Receive
Signal
Monitor
Receive
Signal
Monitor
Output
Register
Output
Register
Output
Register
Output
Register
Elasticity
Buffer
Framer
RXCLKD+ RXCLKD–
10B/8B
BIST
Elasticity
Buffer
10B/8B
BIST
Framer
Elasticity
Buffer
10B/8B
BIST
Framer
Elasticity
Buffer
10B/8B
BIST
Framer
÷2
RXCLKC+ RXCLKC–
÷2
RXCLKB+ RXCLKB–
÷2
RXCLKA+ RXCLKA–
÷2
RXRATE[A..D]
FRAMCHAR[A..D]
RFEN[A..D]
JTAG
Boundary
Scan
Controller
TDO
TMS TCLK TDI
Clock
Select
Clock
Select
Clock
Select
Clock
Select
RXCKSEL[A..D]
RESET
Receive Path Block
= Internal Signal
RXPLLPDA
RFMODE[A..D][1:0]
LPENA
RXBIST[A..D]
DECMODE[A..D]
LPENB
LPENC
LPEND
TRST
RXPLLPDB
RXPLLPDC
RCLKEND
RXPLLPDD
DECBYP[A..D]
SPDSELA
SPDSELB
SPDSELC
SPDSELD
ULCB
ULCA
ULCC
ULCD
LDTDEN
TXLBD
TXLBC
TXLBB
TXLBA
RECLCK[A..D] are Internal Reclocker Signals
SDASEL[A..D][1:0]
RCLKENC
RCLKENB
RCLKENA
RECLCKD
RECLCKC
RECLCKB
RECLCKA
TXLB[A..D] are Internal Serial Loopback Signals
Document #: 38-02097 Rev. *B Page 4 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
WREN ADDR[3:0] DATA[7:0]
Device Configuration and Control Block
= Internal Signal
RXRATE[A..D]
FRAMCHAR[A..D]
RFEN[A..D]
RXCKSEL[A..D]
RFMODE[A..D][1:0]
RXBIST[A..D]
DECMODE[A..D]
DECBYP[A..D]
SDASEL[A..D][1:0] RXPLLPD[A..D] TXRATE[A..D] TXCKSEL[A..D]
TXBIST[A..D] OE[A..D][2..1]
PABRST[A..D]
ENCBYP[A..D] GLEN[1 1..0] FLEN[2..0]
Device Configura-
tion and Control
Interface
Document #: 38-02097 Rev. *B Page 5 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

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
OUT
C1–
C1–INC2–
B
IN
OUT
C1+
C1+INC2+
C
TDI TMS INSELC INSELB
D
TCLK INSELD INSELA
E
VCCVCCVCCV
F
RX
DC[6]RXDC[7]TXDC[0]
G
TX
WREN TX
DC[7]
H
GND GND GND GND GNDGNDGNDGND
J
TX
CTC[1]TXDC[5]TXDC[2]TXDC[3]
K
RX
REF
DC[2]
CLKC–TXCTC[0]TXCLKC
L
RX
REF
DC[3]
CLKC+
M
RX
DC[4]RXDC[5]
N
GND GND GND GND GNDGNDGNDGND
OUT
C2–
OUT
C2+
CC
RCLK
END
DC[4]TXDC[1]
LFIC TX
DC[6]
RCLK
ENCTXERRC
V
CC
V
CC
V
CC
V
CC
IN
OUT
D1–
D1–
IN
OUT
D1+
D1+
ULCD ULCC
ULCA SPD
SELC
GND
GND
GND
GND
IN
D2–
IN
D2+
DATA
[7]
DATA
[6]
OUT
D2–INA1–
OUT
D2+INA1+
DATA
DATA
[5]
DATA
DATA
[4]
OUT
A1–
OUT A1+
DATA
[3]
[1]
DATA
[2]
[0]
GND
GND
GND
GND
IN
OUT
A2–
A2–
IN
OUT
A2+
A2+
RCLK
SPD
ENB
SELD
LPENB ULCB
IN
V
CC
V
CC
V
CC
V
CC
OUT
B1–
B1–INB2–
IN
OUT
B1+
B1+INB2+
LDTD ENTRST LPEND TDO
LPENA VCC SCAN
VCCVCCVCCV
RCLK
ENARXSTB[1]TXCLKOBRXSTB[0]
SPD
SELBLPENC
RX
STB[2]RXDB[0]RXDB[5]RXDB[2]
RX
DB[3]RXDB[4]RXDB[7]
RX
DB[6]RXCLKB+RXCLKB–TXDB[6]
REF
REF
CLKB+
CLKB–TXERRBTXCLKB
OUT
B2–
OUT
B2+
TMEN3
EN2
CC
SPD
SELARXDB[1]
LFIB
P
RX
DC[1]RXDC[0]RXSTC[0]RXSTC[1]
R
RX
STC[2]TXCLKOCRXCLKC+RXCLKC–
T
VCCVCCVCCV
U
TX
DD[0]TXDD[1]TXDD[2]TXCTD[1]
V
TX
DD[3]TXDD[4]TXCTD[0]RXDD[6]
W
TX
DD[5]TXDD[7]
Y
TX
DD[6]TXCLKDRXDD[7]
LFID RX
CLKD–
RX
CLKD+
CC
V
CC
V
CC
V
CC
V
CC
RX
DD[2]RXDD[1]
RX
DD[3]RXSTD[0]
RX
DD[4]RXSTD[1]
RX
DD[5]RXDD[0]
GND
GND
GND
GND
TX
CTA[1]
RX
STD[2]
ADDR
[3]
TX
CLKOD
ADDR
REF
[0]
CLKD–TXDA[1]
ADDR
REF
[2]
CLKD+TXCLKOA
ADDR
[1]RXCLKA+TXERRA
[1]
NC
TX
CLKARXCLKA–
GND
GND
GND
GND
TX
DA[4]TXCTA[0]
TX
DA[3]TXDA[7]
TX
DA[2]TXDA[6]
TX
DA[0]TXDA[5]
TX
DB[5]TXDB[4]TXDB[3]TXDB[2]
TX
DB[1]TXDB[0]TXCTB[1]TXDB[7]
VCCVCCVCCV
RX
V
CC
DA[2]TXCTB[0]RXSTA[2]RXSTA[1]
RX
V
CC
DA[7]RXDA[3]RXDA[0]RXSTA[0]
LFIA REF
V
CC
V
CC
CLKA+RXDA[4]RXDA[1]
TX
REF
ERRD
CLKA–RXDA[6]RXDA[5]
CC
Document #: 38-02097 Rev. *B Page 6 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

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–INB2–
OUT
B1–INB1–
V
CC
OUT
A2–INA2–
GND
OUT
A1–INA1–
OUT
D2–IND2–
GND
OUT D1–IND1–
V
CC
OUT
C2–INC2–
OUT
C1–INC1–
OUT
B
B2+INB2+
TDO LP
C
TMEN3 SCAN
D
VCCVCCVCCV
E
RX
F
STB[0]TXCLKOBRXSTB[1]
RX
G
DB[1]
GND GND GND GND GNDGNDGNDGND
H
RX
J
DB[2]RXDB[5]RXDB[0]RXSTB[2]
LFIB RX
K
TX
L
DB[6]RXCLKB–RXCLKB+RXDB[6]
TX
M
CLKBTXERRB
OUT
B1+INB1+
TRST LDTD
END
EN2
SPD
SELALPENC
DB[7]RXDB[4]RXDB[3]
VCC LP
REF
CLKB–
EN
ENA
CC
RCLK
ENA
SPD
SELB
REF
CLKB+
V
CC
V
CC
V
CC
OUT
A2+INA2+
SPD
RCLK
SELD
ENB
ULCB LP
ENB
GND
GND
GND
OUT
A1+INA1+
DATA
DATA
[1]
DATA
DATA
[0]
[3]
[2]
OUT
D2+IND2+
DATA
[5]
DATA
[4]
DATA
[7]
DATA
[6]
GND
GND
GND
OUT D1+IND1+
ULCC ULCD
SPD
ULCA
SELC
OUT
V
CC
C2+IN2+
IN
V
CC
SELBINSELC
IN
V
CC
SELAINSELD
VCCVCCVCCV
RCLK
ENDTXDC[0]RXDC[7]RxDC[6]
TX
DC[1]TXDC[4]
TX
DC[3]TXDC[2]TXDC[5]TXCTC[1]
TX
CLKCTXCTC[0]
TX
LFIC REF
DC[6]
TX
RCLK
ERRC
ENCRXDC[5]RXDC[4]
OUT
TMS TDI
RESET TCLK
WREN TX
REF
CLKC–RXDC[2]
CLKC+RXDC[3]
C1+INC1+
CC
DC[7]
GND GND GND GND GNDGNDGNDGND
N
TX
P
DB[2]TXDB[3]TXDB[4]TXDB[5]
TX
R
DB[7]TXCTB[1]TXDB[0]TXDB[1]
VCCVCCVCCV
T
RX
U
STA[1]RXSTA[2]TXCTB[0]RXDA[2]
RX
V
STA[0]RXDA[0]RXDA[3]RXDA[7]
RX
W
DA[1]RXDA[4]
RX
Y
DA[5]RXDA[6]
CC
REF
LFIA
CLKA+
REF
CLKA–TXERRD
V
CC
V
CC
V
CC
V
CC
TX
CTA[0]TXDA[4]
TX
DA[7]TXDA[3]
TX
DA[6]TXDA[2]
TX
DA[5]TXDA[0]
GND
GND
GND
GND
TX
REF
DA[1]
CLKD–
TX
REF
CLKOA
CLKD+
TX
ERRARXCLKA+
RX
CLKA–TXCLKA
ADDR
[0]
ADDR
[2]RXSTD[2]
ADDR
[1]
[1]
NC
CLKOD
TXC
TA[1]
ADDR
[3]
TX
GND
GND
GND
GND
RX
DD[1]RXDD[2]
RX
STD[0]RXDD[3]
RX
STD[1]RXDD[4]
RX
DD[0]RXDD[5]
RX
STC[1]RXSTC[0]RXDC[0]RXDC[1]
RX
CLKC–RXCLKC+TXCLKOCRXSTC[2]
VCCVCCVCCV
TX
V
CC
CTD[1]TXDD[2]TXDD[1]TXDD[0]
RX
V
CC
DD[6]TXCTD[0]TXDD[4]TXDD[3]
RX
V
CC
V
CC
LFID TX
CLKD–
RX
CLKD+RXDD[7]TXCLKDTXDD[6]
CC
DD[7]TXDD[5]
Note
1. NC=Do Not Connect
Document #: 38-02097 Rev. *B Page 7 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

Pin Definitions CYV15G0404DXB 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]
TXCTA[1:0] TXCTB[1:0] TXCTC[1:0] TXCTD[1:0]
TXERRA TXERRB TXERRC TXERRD
Transmit Path Clock Signals
REFCLKA± REFCLKB± REFCLKC± REFCLKD±
TXCLKA TXCLKB TXCLKC TXCLKD
Notes
2. When REFCLKx± is configured for half rate operation, these inputs are sampled relative to both the ri sing and falling edges of the associated REFCLKx±.
3. When REFCLKx± is configured for half rate operation, these outputs are presented relative to both the risin g and falling edges of the associated REFCLKx±.
LVTTL Input, synchronous, sampled by the associated TXCLKx or REFCLKx
[2]
LVTTL Input, synchronous, sampled by the associated TXCLKx or REFCLKx
[2]
LVTTL Output, synchronous to REFCLKx
[3]
, synchronous to RXCLKx when selected as REFCLKx, asynchronous to transmit channel enable/disable, asynchronous to loss or return of REFCLKx±
Differential LVPECL or single ended LVTTL input clock
LVTTL Clock Input, internal pull down
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.
Transm it Contr ol. 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 through the device configuration interface, and passed to the encoder or transmit shifter. The TXCTA[1:0] inputs identify how the associated TXDx[ 7:0 ] charac ters 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 for details.
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 recentered with the PABRSTx latch through 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 these conditions is true:
The TXPLL for the associated channel is powered down. This occurs when OE2x
and OE1x for a given channel are simultaneous disabled by setting OE2x = 0 and OE1x = 0.
The absence of the REFCLKx± signal.
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 L VPECL clock source, the clock must be a differential clock, using both inputs.
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 i nput 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.
Document #: 38-02097 Rev. *B Page 8 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Pin Definitions (continued) CYV15G0404DXB Quad HOTLink II Transceiver
Name I/O Characteristics Signal Description
TXCLKOA TXCLKOB TXCLKOC TXCLKOD
Receive Path Data and Status Signals
RXDA[7:0] RXDB[7:0] RXDC[7:0] RXDD[7:0]
RXSTA[2:0] RXSTB[2:0] RXSTC[2:0] RXSTD[2:0]
Receive Path Clock Signals
RXCLKA± RXCLKB± RXCLKC± RXCLKD±
Device Control Signals
RESET
LV TTL Output 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± (TXRATE = 0), or at twice the frequency of REFCLKx± (TXRATE = 1). The transmit clock outputs have no fixed phase relationship to REFCLKx±.
LVTTL Output, synchronous to the selected RXCLK± output or REFCLKx± input
LVTTL Output, synchronous to the selected RXCLK± output or REFCLKx± input
LVTTL Output Clock Receive Clock Output. RXCLKx± is the receive interface clock used to control timing
LVTTL Input, asynchronous, internal pull up
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.
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 for a list of received character status.
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 (1/20 serial bit-rate) or character rate (1/10 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 or 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±.
Asynchronous Device Reset. RESET configuration latches in the device to a known state. RESET 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 to be reset separately. Refer to JTAG Support on page 23 for the methods to reset the JTAG state machine. See Table 9 for the initialize values of the device configu- ration latches.
cannot reset the JTAG controller. Therefore, the JTAG controller has
th
th
the serial bit-rate) of the data being received,
initializes all state machines, counters, and
must be asserted LOW
the
Document #: 38-02097 Rev. *B Page 9 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Pin Definitions (continued) CYV15G0404DXB Quad HOTLink II Transceiver
Name I/O Characteristics Signal Description
LDTDEN LVTTL Input,
internal pull up
RCLKENA RCLKENB
LVTTL Input, internal pull down
RCLKENC RCLKEND
ULCA ULCB
LVTTL Input, internal pull up
ULCC ULCD
SPDSELA SPDSELB
3-Level Select static control input
[4]
SPDSELC SPDSELD
INSELA INSELB
LVTTL Input, asynchronous
INSELC INSELD
LPENA LPENB LPENC
LVTTL Input, asynchronous, internal pull down
LPEND
Notes
4. 3-Level Select inputs are used for static configuration. These are ternary inpu ts that make use of logic levels of LOW, MID, and HIGH. The LOW level is usually implemented by direct connection to V implemented by not connecting the input (left floating), which allows it to self bias to the proper level.
5. See Device Configuration and Control Interface for detailed information on the operation of the Configuration Interface.
SS
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]) 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 determines 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.
Reclocker Enable. When RCLKENx is HIGH, the RXPLL performs clock and data recovery functions on the input serial data stream and routes the deserialized data to the RXDx[7:0] and RXSTA[2:0] parallel data outputs as configured by DECBYPx. It also presents the reclocked serial data to the enabled differential serial outputs.
When RCLKENx is LOW, the receive reclocker is disabled and the TXDx[7:0] parallel data inputs and TXCTx[1:0] inputs are interpreted (as configured by ENCBYPx) to generate appropriate 10-bit characters that are presented to the differential serial outputs.
The reclocker feature is optimized to be used for SMPTE video applications. 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. 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 MBd MID = 400 – 800 MBd HIGH = 800 – 1500 MBd. 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.
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 an d 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.
(ground). The HIGH level is usually implemented by direct connection to VCC (power). The MID level is usually
Document #: 38-02097 Rev. *B Page 10 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Pin Definitions (continued) CYV15G0404DXB Quad HOTLink II Transceiver
Name I/O Characteristics Signal Description
LFIA LFIB LFIC LFID
Device Configuration and Control Bus Signals
WREN
ADDR[3:0] LVTTL input
DATA[7:0] LVTTL input
Internal Device Configuration Latches
RFMODE[A..D][1:0] In ternal Latch FRAMCHAR[A..D] Internal Latch DECMODE[A..D] Internal Latch DECBYP[A..D] Internal Latch RXCKSEL[A..D] Internal Latch RXRATE[A..D] Internal Latch SDASEL[A..D][1:0] Internal Latch ENCBYP[A..D] Internal Latch TXCKSEL[A..D] Internal Latch TXRATE[A..D] Internal Latch RFEN[A..D] Internal Latch RXPLLPD[A..D] Internal Latch RXBIST[A..D] Internal Latch TXBIST[A..D] Internal Latch OE2[A..D] Internal Latch OE1[A..D] Internal Latch PABRST[A..D] Internal Latch GLEN[11..0] Internal Latch FGLEN[2..0] Internal Latch
Note
6. See Device Configuration and Control Interface for detailed information on the internal latches.
LVTTL Output, asynchronous
LVTTL input, asynchronous, internal pull up
asynchronous, internal pull up
asynchronous, internal pull up
Link Fault Indication Output. LFIx is an output status indicator signal. LFIx is the logical OR of five internal conditions. LFIx tions are true:
Received serial data rate outside expected range
Analog amplitude below expected levels
Transition density lower than expected
Receive channel disabled
ULCx is LOW
No REFCLKx±.
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.
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 DAT A[7:0] bus into the latch specified by the address location on the AD DR[3:0] bus. 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.
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. 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.
[6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6] [6]
Reframe Mode Select. Framing Character Select. Receiver Decoder Mode Select. Receiver Decoder Bypass. Receive Clock Select. Receive Clock Rate Select. Signal Detect Amplitude Select. Transmit Encoder Bypass. Transmit Clock Select. Transmit PLL Clock Rate Select. Reframe Enable. Receive Channel Power Control. Receive Bist Disabled. Transmit Bist Disabled. Differential Serial Output Driver 2 Enable. Differential Serial Output Driver 1 Enable. Transmit Clock Phase Alignment Buffer Reset. Global Latch Enable. Force Global Latch Enable.
is asserted LOW when any of these condi -
[5]
[5]
Table 9 lists the
[5]
Table 9 lists the configuration
Document #: 38-02097 Rev. *B Page 11 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Pin Definitions (continued) CYV15G0404DXB Quad HOTLink II Transceiver
Name I/O Characteristics Signal Description
Factory Test Modes
SCANEN2 LVTTL input,
internal pull down
TMEN3 LVTTL input,
internal pull down
Analog I/O
OUTA1± OUTB1± OUTC1± OUTD1±
OUTA2± OUTB2± OUTC2± OUTD2±
INA1± INB1± INC1± IND1±
INA2± INB2± INC2± IND2±
JTAG Interface
TMS LVTTL Input,
TCLK LVTTL Input,
TDO 3-State LVTTL Output Test Data Out. JTAG data output buf fer . High-Z while JTAG test mode is not selected. TDI LVTTL Input,
TRST
Power
V
CC
GND Signal and Power Ground for all internal circuits.
CML Differentia l Output
CML Differentia l Output
Differential Input Primary Differential Serial Data Input. The INx1± input accepts the serial data
Differential Input Secondary Differentia l Serial Data Input. The INx2± input accepts the serial data
internal pull up
internal pull down
internal pull up LVTTL Input,
internal pull up
Factory Test 2. SCANEN2 input is for factory testing only . Leave this input as a NO CONNECT or GND only.
Factory Test 3. TMEN3 input is for factory testing only. Leave this input as a NO CONNECT or GND only.
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.
Secondary Differential Serial Data Output. The OUTx2± PECL-comp atible 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 comp atible connections.
stream for deserialization and decoding. The INx1± serial stream is passed to the receive CDR circuit to extract the data content when INSELx = HIGH.
stream for deserialization and decoding. The INx2± serial stream is passed to the receiver CDR circuit to extract the data content when INSELx = LOW.
Te s t M o d e S e lect. Used to control access to the JTAG Test Modes. If maintained high for 5 TCLK cycles, the JTAG test controller is reset.
JTAG Test Clock.
Test Data In. JTAG data input port.
JTAG reset signal. When asserted (LOW), this input asynchronously resets the
JT AG test access port controller.
+3.3V Power.

CYV15G0404DXB HOTLink II Operation

The CYV15G0404DXB 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.

CYV15G0404DXB 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 capture s eight data bits and two control bits on each input clock cycle.
Document #: 38-02097 Rev. *B Page 12 of 44
When the encoder is bypassed, the control bits are part of the preencoded 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 TXRA TEx = 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.
[+] Feedback [+] Feedback
CYV15G0404DXB
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
[7]
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]
Note
7. LSB shifted out first.
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 w ith minimal disruption of the serial data stream. When a phase align buffer error is present, the transmission of a word sync sequence recenters 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, follow the phase alignment buffer reset 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 preencoded 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

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)
When the encoder is enabled (ENCBYPx = 1), th e 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 15. When directed to encode the character as a data character, it is encoded using the data character encoding rules in Table 14.
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 st andards for data transport.
Many of the special character codes listed in Table 15 may be generated by more than one input character. The CYV15G0404DXB is designed to support two independent (but non-overlapping) special character code tables. This allows the CYV15G0404DXB 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.
Document #: 38-02097 Rev. *B Page 13 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

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) 2
TXDx[1] 2 TXDx[2] 2 TXDx[3] 2 TXDx[4] 2 TXDx[5] 2 TXDx[6] 2 TXDx[7] 2
TXCTx[0] 2
TXCTx[1] (MSB) 2
0 1 2 3 4 5 6 7 8 9
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] Characters Generated
0 0 Encoded data character 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
[7]
a
b c d e
i
f g h
j
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 through 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 transmit paths, and outputs this character rate clock as TXCLKOx.
Each clock multiplier PLL is able to accept a REFCLKx± input between 19.5 MHz and 150 MHz, however, this clock range is limited by the operating mode of the CYV15G0404DXB clock multiplier (TXRATEx) and by the level on the associated SPDSELx input.
SPDSELx are 3-level select 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.
Table 4. Operating Speed Settings
SPDSELx TXRATE
LOW
1 reserved 195 – 400 0 19.5 – 40
MID (Open) 1 20 – 40 400 – 800
0 40 – 80
HIGH 1 40 – 75 800 – 1500
0 80 – 150
[4]
inputs that select one of three
REFCLKx± Frequency
(MHz)
Signaling
Rate (MBaud)
Document #: 38-02097 Rev. *B Page 14 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
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
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.
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, 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. They 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 1.

Transmit Channels Enabled

Each driver can be enabled or disabled separately using the device configuration interface.
When a driver is disabled throug h 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.
[8]

CYV15G0404DXB Receive Data Path

Serial Line Receivers

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

Signal Detect/Link Fault

Each selected line receiver (that is 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)
Receive channel enabled
The presence of a 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 noi se environments. The analog amplitude level detection is set by the SDASELx latch via device configuration interface. The SDASELx latch sets the trip point for the detection of a va lid 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
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
[9]
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 VI 100 mV , or 200 mV peak-to-peak differential. Each Line Receiver
DIFF
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
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.
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.
Notes
8. When a disabled transmit channel (i.e., both outputs disabled) is re-enabled, the dat a on the serial outputs may not meet all timing specifications for up to 250 ms.
9. The peak amplitudes listed in this table are for ty pical waveforms t hat h ave g enerally 3 – 4 t ransi tio ns for every t en bits. In a worse case environment the signals may have a sign-wave appearance (highest transition density with repeat ing 0101...). Signal peak amplit udes levels within this en vironmen t type could increase the value s in the table above by approximately 100 mV.
Document #: 38-02097 Rev. *B Page 15 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

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 fu nction, 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, 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 in the following manner: 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 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

Receive Channel Enabled

The CYV15G0404DXB 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 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 the associated channel may be indeterminate for up to 2 ms.
output is asserted LOW. After a valid
output and data on the parallel outputs for
.
should be HIGH.
output. When

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 half-character rate (bit-rate 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 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 frequency of REFCLKx± is required to be within ±1500 ppm 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 can be output 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.
÷ 20) reference clock from the
÷ 10) or
output. The

Reclocker

The CYV15G0404DXB contains a reclocker mode on each receive channel that can be independently enabled and disabled. When the reclocker mode is enabled by RCLKENx, the received serial data is reclocked and transmitted through the enabled differential serial outputs of the selected channel. In the reclocker mode, the RXPLL performs clock and data recovery functions on the input serial data stream and the reclocked serial data is routed to the enabled differential serial outputs. The serial data is also routed to the deserializer and the deserialized data is presented to the RXDx[7:0] and RXSTA[2:0] parallel data outputs as configured by DECBYPx. When the reclocker is enabled, the data on the TXDx[7:0] and TXCT[1:0] is ignored and not transmitted through the enabled serial outputs.

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 determines the character boundaries of all following characters.

Framing Character

The CYV15G0404DXB 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
Document #: 38-02097 Rev. *B Page 16 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
through the FRAMCHARx latches through the configuration
Note
10.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 p attern is extended to a full eight bits to reduce the possibility of a framing error.
interface. 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.
Table 6. Framing Character Selector
FRAMCHARx
0 COMMA+
1–K28.5
Bits Detected in Framer
Character Name Bits Detected
[10]
COMMA–
00111110XX or 11000001XX
0011111010 or
+K28.5
1100000101

Framer

The framer on each channel operates in one of three different modes. Each framer is enabled or disabled using the RFENx latches using 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. T o 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.
When RFMODEx[1:0] = 10, the Cypress-Mode Multi-Byte framer is selected. The required detection of multiple framing characters 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.

10B/8B Decoder Block

The decoder logic block performs two primary functions:
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 14 and
Table 15. Received special code characters are decoded using Table 15. 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 through 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 through the configuration interface, the 10-bit transmission characters are decoded using Table 14 and Table 15. Received Special characters are decoded using Table 15. The columns used in
Table 15 are determined by the DECMODEx latch through 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 using the device configuration interface. When enabled, a register in the associated receive channel becomes a si gnature 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 transmitters. When synchronized with the received data stream, the associated Receiver checks each character in the Decoder with each character generate d 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.
Document #: 38-02097 Rev. *B Page 17 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Code rule violations or running disparity errors that occur as part
Note
11.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.
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. 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-T est.” The sequence compared by the CYV15G0404DXB is identical to that in the CY7B933, CY7C924DX, and CYP(V)15G0401DXB, 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 Receive BIST is enabled on a channel, do not enable the low latency framer. The BIST sequence contains an aliased K28.5 framing character, which causes the receiver to update its character boundaries incorrectly.
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 word sync sequence.
[11]
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 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 buffe r w r it e cl ock.
If the chip is configured for operation with a recovered clock, the elasticity buffer is byp assed.
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 cha racter 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 CYV15G0404DXB 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 through the device configuration interface. When RXPLLPDx = 0, the associated PLL and analog circuitry of the channel is di sabled. The transmit channels are controlled by the OE1x and the OE2x latches through the device configuration interface. When a driver is disabled through 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 powered down as well.

Device Reset State

When the CYV15G0404DXB 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 the JTAG Support section for JTAG state machine initialization. See Table 9 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 is done by sequencing the appropriate values on the device configuration interface.
[5]

Output Bus

Each receive channel presents an 11-signal output bus consisting of
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 CYV15G0404DXB as selected by the DECBYPx configuration latch. This mapping is shown in
Table 7.
Document #: 38-02097 Rev. *B Page 18 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 7. Output Register Bit Assignments
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]
RXDx[0] DOUTx[2] RXDx[0] RXDx[1] DOUTx[3] RXDx[1] RXDx[2] DOUTx[4] RXDx[2] RXDx[3] DOUTx[5] RXDx[3] RXDx[4] DOUTx[6] RXDx[4] RXDx[5] DOUTx[7] RXDx[5] RXDx[6] DOUTx[8] RXDx[6]
RXDx[7] (MSB) DOUTx[9] RXDx[7]
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] 2 RXSTx[0] 2
RXDx[0] 2 RXDx[1] 2 RXDx[2] 2 RXDx[3] 2 RXDx[4] 2 RXDx[5] 2 RXDx[6] 2
RXDx[7] (MSB) 2
0 1 2 3 4 5 6 7 8 9
a b c d e
i
f g h
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 proper character boundary, and LOW for all other bit combina­tions.
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.
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).

Receive Status Bits

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
If the contents of the data bus are valid
The type of character present
The state of receive BIST operations
Character violations
These conditions often overlap; for example, 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.
A second status mapping, listed in Table 11, is u sed 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 S tart-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 buf fer 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.

Device Configuration and Control Interface

The CYP(V)15G0404DX is highly configurable through 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 latche s upon the assertion of RESET 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 DAT A[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
. Table 10 shows how the latches
Document #: 38-02097 Rev. *B Page 19 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
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 is 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 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.

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 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
Table 9. Device Configuration and Control Latch Descriptions
Name Signal Description
RFMODEA[1:0] RFMODEB[1:0] RFMODEC[1:0] RFMODED[1:0]
FRAMCHARA FRAMCHARB FRAMCHARC FRAMCHARD
DECMODEA DECMODEB DECMODEC DECMODED
DECBYPA DECBYPB DECBYPC DECBYPD
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 altern ate mo de Mult i-B yte 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.
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 o f the 8-bit Comma characters. The specific bit combinations of these framing characters are listed in Table 6.
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 14 and Table 15. The column used in the Special Characters Table 15 is determined by the DECMODEx latch.
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.
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 two static and one dynamic latch bank. The S type contain those settings that normally do not change for a given application, while the D type controls the settings that could change dynamically during the ap plication'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 (that is, 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 “11 10” and WREN The signals present in DATA[7:0] effectively become global control pins, and for the latch banks 2, 5, 8, and 11.
is left asserted.
Document #: 38-02097 Rev. *B Page 20 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 9. Device Configuration and Control Latch Descriptions (continued)
Name Signal Description
RXCKSELA RXCKSELB RXCKSELC RXCKSELD
RXRATEA RXRATEB RXRATEC RXRATED
SDASEL1A[1:0] SDASEL1B[1:0] SDASEL1C[1:0] SDASEL1D[1:0]
SDASEL2A[1:0] SDASEL2B[1:0] SDASEL2C[1:0] SDASEL2D[1:0]
ENCBYPA ENCBYPB ENCBYPC ENCBYPD
TXCKSELA TXCKSELB TXCKSELC TXCKSELD
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.
Receive Clock Rate Select. The initialization value of the RXRATEx latch = 1. RXRA TEx 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–.
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.
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.
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.
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].
Document #: 38-02097 Rev. *B Page 21 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 9. Device Configuration and Control Latch Descriptions (continued)
Name Signal Description
TXRATEA TXRATEB TXRATEC TXRATED
RFENA RFENB RFENC RFEND
RXPLLPDA RXPLLPDB RXPLLPDC RXPLLPDD
RXBISTA RXBISTB RXBISTC RXBISTD
TXBISTA TXBISTB TXBISTC TXBISTD
OE2A OE2B OE2C OE2D
OE1A OE1B OE1C OE1D
PABRSTA PABRSTB PABRSTC PABRSTD
GLEN[11..0] Global Enable. The initialization value of the GLENx la tch = 1 . Th e GLENx is used to reconfigure several
FGLEN[2..0] Force Global Enable. The initialization value of the FGLENx latch is NA. The FGLENx latch forces a GLobal
Transmit PLL Clock Rate Select. The initialization value of the TXRATEx latch = 0. TXRA TEx 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 TXRA TEx = 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.
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.
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.
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.
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.
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.
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.
Transmit Clock Phase Alignment Buffer Reset. The initialization value of the PABRSTx latch = 1. The PA BRSTx 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. PA BRSTx is a self clearing latch. This eliminates the requirement of writing a 1 to complete the initialization of the Phase Alignment Buffer.
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.
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 #: 38-02097 Rev. *B Page 22 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

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 the JTAG Support section.
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.]
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 = 0, 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.
Table 10. Device Control Latch Configuration Table
ADDR Channel Type DATA7 DATA6 DATA5 DATA4 DATA3 DATA2 DATA1 DATA0
0
(0000b)
1
(0001b)
2
(0010b)
3
(0011b)
4
(0100b)
5
(0101b)
6
(0110b)
7
(0111b)
8
(1000b)
9
(1001b)
10
(1010b)
11
(1011b)
12
(1100b)
13
(1101b)
14
(1110b)
15
(1111b)
A S RFMODEA[1] RFMODEA[0] FRAMCHARA DECMODEA DECBYPA RXCKSELA RXRATEA GLEN0 10111111
A S SDASEL2A[1] SDASEL2A[0] SDASEL1A[1] SDASEL1A[0] ENCBYPA TXCKSELA TXRATEA GLEN1 10101101
A D RFENA RXPLLPDA RXBISTA TXBISTA OE2A OE1A PABRSTA GLEN2 10110011
B S RFMODEB[1] RFMODEB[0] FRAMCHARB DECMODEB DECBYPB RXCKSELB RXRATEB GLEN3 10111111
B S SDASEL2B[1] SDASEL2B[0] SDASEL1B[1] SDASEL1B[0] ENCBYPB TXCKSELB TXRATEB GLEN4 10101101
B D RFENB RXPLLPDB RXBISTB TXBISTB OE2B OE1B PABRSTB GLEN5 10110011
C S RFMODEC[1] RFMODEC[0] FRAMCHARC DECMODEC DECBYPC RXCKSELC RXRATEC GLEN6 10111111
C S SDASEL2C[1] SDASEL2C[0] SDASEL1C[1] SDASEL1C[0] ENCBYPC TXCKSELC TXRATEC GLEN7 10101101
C D RFENC RXPLLPDC RXBISTC TXBISTC OE2C OE1C PABRSTC GLEN8 10110011
D S RFMODED[1] RFMODED[0] FRAMCHARD DECMODED DECBYPD RXCKSELD RXRATE D GLEN9 10111111
D S SDASEL2D[1] SDASEL2D[0] SDASEL1D[1] SDASEL1D[0] ENCBYPD TXCKSELD TXRATED GLEN10 10101101
D D RFEND RXPLLPDD RXBISTD TXBISTD OE2D OE1D PABRSTD GLEN11 10110011
GLOBAL S RFMODEGL[1] RFMODE
GLOBAL S SDASEL2GL[1] SDASEL2GL[0]SDASEL1GL[1] SDASEL1GL[0]ENCBPGL TXCKSELGL TXRATEGLFGLEN1 N/A
GLOBAL D RFENGL RXPLLPDGL RXBISTGL TXBISTGL OE2GL OE1GL PABRSTGLFGLEN2 N/A
MASK D D7 D6 D5 D4 D3 D2 D1 D0 11111111
GL[0]
FRAMCHARGL DECMODEGL DECBYPGL RXCKSELGL RXRATEGLFGLEN0 N/A
Reset Value

JTAG Support

The CYV15G0404DXB contains a JTAG port to allow system level diagnosis of device interconnect. Of th e 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 must also be
reset (using RESET) TRST
(asserting it LOW and de-asserting it or leaving it asserted), or by asserting TMS HIGH for at least five consecutive TCLK cycles. This is necessary to ensure 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 is in normal operation.
Note. The order of device reset (using RESET ization does not matter.
. The JT AG state machine is initialized using
) and JT AG initial-
initialized to a reset state. This is done in addition to the device
Document #: 38-02097 Rev. *B Page 23 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

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 CYV15G0404DXB is ‘0C811069’x .
Receive Character Status Bits
Description
RXSTx[2:0] Priority
000 7 Normal character received. The valid Data
character on the output bus meets all the formatting requirements of Data characters listed in Table 14.
001 7 Special code detected. The valid special
character on the output bus meets all the formatting requirements of Special Code characters listed in Table 15, 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
output bus is a C0.7. This indicates that the received character cannot be decoded into any valid character.
101 1 Loss of sync. This indicates a PLL Out of Lock
condition
110 6 Running disparity error. The character on the
output bus is a C4.7, C1.7, or C2.7.
111 3 Reserved BIST Wait. The receiver is comparing characters. but
Normal Status
BIST Data Compare. Character compared correctly.
BIST Command Compare. Character compared
correctly.
BIST Last Good. Last Character of BIST sequence detected and valid.
BIST Last Bad. Last Character of BIST sequence detected invalid.
BIST Start. Receive BIST is enabled on this channel, but character compares have not yet commenced. This also indicates a PLL Out of Lock condition, and Elasticity Buffer overflow/underflow conditions.
BIST Error. While comparing characters, a mismatch was found in one or more of the decoded character bits.
has not yet found the start of BIST character to enable the LFSR.
Receive BIST Status
(Receive BIST = Enabled)
Document #: 38-02097 Rev. *B Page 24 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Figure 2. Receive BIST State Machine
Receive BIST
Detected LOW
Monitor Data
Received
RXSTx =
BIST_START (101)
No
RX PLL
Out of Lock
Yes, RXSTx =
BIST_DATA_COMPARE (000) / BIST_COMMAND_COMPARE (001)
Compare
Next Character
Auto-Abort
Condition
Mismatch
End-of-BIST
State
Yes, RXSTx =
BIST_LAST_BAD (100)
Yes
No
No, RXSTx =
BIST_ERROR (110)
Data or
Command
Match
Command
RXSTx =
BIST_COMMAND_COMPARE (001)
End-of-BIST
State
Data
Yes, RXSTx =
BIST_LAST_GOOD (010)
No
RXSTx =
BIST_DATA_COMPARE (000)
Elasticity
Buffer Error
Start of
BIST Detected
RXSTx =
BIST_WAIT (111)
Yes
RXSTx =
BIST_START (101)
No
Document #: 38-02097 Rev. *B Page 25 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

Maximum Ratings

Exceeding maximum ratings may impair the useful life of device. These user guidelines are not tested.
Storage Temperature.................................. –65°C to +150°C
Ambient Temperature with
Power Applied ............................................–55°C to +125°C
Supply Voltage to Ground Potential................–0.5V to +3.8V
DC Voltage Applied to LVTTL Outputs
in High-Z State.......................................–0.5V to V
Output Current into LVTTL Outputs (LOW).................60 mA
DC Input Voltage...................................–0.5V to V
CC
CC
+ 0.5V
+ 0.5V
Static Discharge Voltage..........................................> 2000 V
(according to MIL-STD-883, Method 3015)
Latch-up Current.....................................................> 200 mA

Power Up Requirements

The CYP(V)15G0404DXB requires one power supply. The Voltage on any input or IO pin cannot exceed the power pin during power up.

Operating Range

Range Ambient Temperature V
Commercial 0°C to +70°C +3.3V ±5% Industrial –40°C to +85°C +3.3V ±5%

CYV15G0404DXB DC Electrical Characteristics

Parameter Description Test Conditions Min. Max. Unit LVTTL-compatible Outputs
V
OHT
V
OLT
I
OST
I
OZL
LVTTL-compatible Inputs
V
IHT
V
ILT
I
IHT
I
ILT
I
IHPDT
I
ILPUT
LVDIFF Inputs: REFCLKx±
[13]
V
DIFF
V
IHHP
V
ILLP
V
COMREF
3-Level Inputs
V
IHH
V
IMM
V
ILL
I
IHH
I
IMM
I
ILL
Notes
12.Tested one output at a time, output shorted for less than one second, less than 10% duty cycle.
13.This is the minimum difference in voltage between the true and complement input s required t o ensure detectio n 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 (-) inp ut is more positive than true (+) input.
14.The common mode range defines the allowable range of REFCLKx+ and REFCLKx- when REFCLKx+ = REFCLKx-. This marks the zero-cro ssing between the true and complement inputs as the signal switches between a logic-1 and a logic-0.
Output HIGH Voltage IOH = 4 mA, VCC = Min. 2.4 V Output LOW Voltage IOL = 4 mA, VCC = Min. 0.4 V Output Short Circuit Current V High-Z Output Leakage Current V
OUT OUT
[12]
= 0V = 0V , V
, VCC = 3.3V –20 –100 mA
CC
–20 20 µA
Input HIGH Voltage 2.0 VCC + 0.3 V Input LOW Voltage –0.5 0.8 V Input HIGH Current REFCLKx Input, VIN = V
Other Inputs, VIN = V
CC
CC
1.5 mA
+40 µA
Input LOW Current REFCLKx Input, VIN = 0.0V –1.5 mA
Other Inputs, V
Input HIGH Current with internal pull down VIN = V
CC
= 0.0V –40 µA
IN
+200 µA
Input LOW Current with internal pull up VIN = 0.0V –200 µA
Input Differential Voltage 400 V Highest Input HIGH Voltage 1.2 V
CC CC
Lowest Input LOW voltage 0.0 VCC/2 V
[14]
Common Mode Range 1.0 VCC – 1.2V V
Three-Level Input HIGH Voltage Min. ≤ VCC Max. 0.87 * V Three-Level Input MID Voltage Min. ≤ VCC Max. 0.47 * V
CC CC
V
CC
0.53 * V Three-Level Input LOW Voltage Min. ≤ VCC Max. 0.0 0.13 * V Input HIGH Current VIN = V
CC
200 µA Input MID current VIN = VCC/2 –50 50 µA Input LOW current VIN = GND –200 µA
CC CC
CC
mV
V
V V V
Document #: 38-02097 Rev. *B Page 26 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
CYV15G0404DXB DC Electrical Characteristics (continued)
2.0V
0.8V
GND
2.0V
0.8V
80%
20%
80%
20%
R
L
(Includes fixture and probe capacitance)
3.0V
Vth=1.4V
270 ps
270 ps
[19]
Vth=1.4V
3.3V R1
R2
R1 = 590Ω R2 = 435Ω
(Includes fixture and probe capacitance)
C
L
7 pF
(a) LVTTL Output Test Load
RL= 100Ω
(b) CML Output Test Load
C
L
(c) LVTTL Input Test Waveform
(d) CML/LVPECL Input Test Waveform
1ns
1 ns
V
IHE
V
ILE
V
IHE
V
ILE
[18]
[18]
Parameter Description Test Conditions Min. Max. Unit Differential CML Serial Outputs: OUTA1±, OUTA2±, OUTB1±, OUTB2±, OUTC1±, OUTC2±, OUTD1±, OUTD2±
V
OHC
V
OLC
V
ODIF
Differential Serial Line Receiver Inputs: INA1±, INA2±, INB1±, INB2±, INC1±, INC2±, IND1±, IND2±
[13]
V
DIFFs
V
IHE
V
ILE
I
IHE
I
ILE
[15]
VI
COM
Power Supply T yp. Max.
[16, 17]
I
CC
[16, 17]
I
CC
Output HIGH Voltage (V
Referenced)
cc
Output LOW Voltage (V
Referenced)
CC
Output Differential Voltage |(OUT+) (OUT)|
100Ω differential load V 150Ω differential load VCC – 0.5 VCC – 0.2 V 100Ω differential load V 150Ω differential load V 100Ω differential load 450 900 mV 150Ω differential load 560 1000 mV
– 0.5 V
CC
– 1.4 VCC – 0.7 V
CC
– 1.4 VCC – 0.7 V
CC
– 0.2 V
CC
Input Differential Voltage |(IN+) (IN−)| 100 1200 mV Highest Input HIGH Voltage V
CC
Lowest Input LOW Voltage VCC – 2.0 V Input HIGH Current VIN = V Input LOW Current VIN = V Common Mode input range ((VCC – 2.0V) + 0.5)min,
(V
CC
Max Power Supply Current REFCLKx =
MAX
Typical Power Supply Current REFCLKx =
125 MHz
Max. 1350 μA
IHE
Min. –700 μA
ILE
+1.25 +3.1 V
– 0.5V) max.
Commercial 910 1270 mA Industrial 1320 mA Commercial 900 1270 mA Industrial 1320 mA
V

AC Test Loads and Waveforms

Notes
15.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.
16.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.
17.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 patte rn. The redu ndant outpu t s on each channel are power ed down and t he p arallel out put s are unl oaded.
18.Cypress uses constant current (ATE) load configurations and forcing functions. This figure is for reference only.
19.The LVTTL switching threshold is 1.4V. All timing references are made relative to where the signal edges cross the threshold voltage.
Document #: 38-02097 Rev. *B Page 27 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

CYV15G0404DXB AC Electrical Characteristics

Parameter Description Min. Max Unit

CYV15G0404DXB Transmitter LVTTL Switching Characteristics Over the Operating Range

f
TS
t
TXCLK
[20]
t
TXCLKH
[20]
t
TXCLKL
[20, 21, 22]
t
TXCLKR
[20, 21, 22]
t
TXCLKF
t
TXDS
t
TXDH
f
TOS
t
TXCLKO
t
TXCLKOD

CYV15G0404DXB Receiver LVTTL Switching Characteristics Over the Operating Range

f
RS
t
RXCLKP
t
RXCLKD
[20]
t
RXCLKR
[20]
t
RXCLKF
[23]
t
RXDv–
[23]
t
RXDv+

CYV15G0404DXB REFCLKx Switching Characteristics Over the Operating Range

f
REF
t
REFCLK
t
REFH
t
REFL
[25]
t
REFD
[20, 21, 22]
t
REFR
[20, 21, 22]
t
REFF
Notes
20.Tested initially and after any design or process changes that may affect these parameters, but not 100% tested.
21.The ratio of rise time to falling time must not vary by greater than 2:1.
22.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.
23. Parallel data output specifications are only valid if all outputs are lo aded with similar DC an d AC loads.
24.Receiver UI (Unit Interval) is c a l c u l a t e d a s 1 / ( f
25.The duty cycle specification is a simultaneous condition with the t cannot be as large as 30%–70%.
TXCLKx Clock Cycle Frequency 19.5 150 MHz TXCLKx Period=1/f
TS
6.66 51.28 ns TXCLKx HIGH Time 2.2 ns TXCLKx LOW Time 2.2 ns TXCLKx Rise Time 0.2 1.7 ns TXCLKx Fall Time 0.2 1.7 ns Transmit Data Set-up Time to TXCLKx↑ (TXCKSELx 0) 2.2 ns Transmit Data Hold Time from TXCLKx↑ (TXCKSELx ≠ 0) 1.0 ns TXCLKOx Clock Frequency = 1x or 2x REFCLKx Frequency 19.5 150 MHz TXCLKOx Period=1/f
TOS
6.66 51.28 ns TXCLKO Duty Cycle centered at 60% HIGH time –1.9 0 ns
RXCLKx± Clock Output Frequency 9.75 150 MHz RXCLKx± Period = 1/f
RS
RXCLKx± Duty Cycle Centered at 50% (Full Rate and Half Rate when
6.66 102.56 ns
–1.0 +1.0 ns
RXCKSELx = 0) RXCLKx± Rise Time 0.3 1.2 ns RXCLKx± Fall Time 0.3 1.2 ns Status and Data Valid Time to RXCLKx± (RXRATEx = 0, RXCKSELx = 0)
5UI–2.0
[24]
(Full Rate) Status and Data Valid Time to RXCLKx± (RXRATEx = 1, RXCKSELx = 0)
5UI–1.3
[24]
(Half Rate) Status and Data Valid Time to RXCLKx± (RXRATEx = 0, RXCKSELx = 0)
5UI–1.8
[24]
(Full Rate) Status and Data Valid Time to RXCLKx± (RXRATEx = 1, RXCKSELx =0)
5UI–2.6
[24]
(Half Rate)
REFCLKx Clock Frequency 19.5 150 MHz REFCLKx Period = 1/f
REF
6.6 51.28 ns REFCLKx HIGH Time (TXRATEx = 1)(Half Rate) 5.9 ns REFCLKx HIGH Time (TXRATEx = 0)(Full Rate) 2.9
[20]
REFCLKx LOW Time (TXRATEx = 1)(Half Rate) 5.9 ns REFCLKx LOW Time (TXRATEx = 0)(Full Rate) 2.9
[20]
REFCLKx Duty Cycle 30 70 % REFCLKx Rise Time (2 0% – 8 0% ) 2 ns REFCLKx Fall Time (20%–80%) 2 ns
* 20) (when TXRATEx = 1) or 1/(f
REF
REFH
and t
* 10) (when TXRATEx = 0). In an operating link this is equivalent to tB.
REF
parameters. This means that at faster character rates the REFCLKx± duty cycle
REFL
ns
ns
ns
ns
ns
ns
Document #: 38-02097 Rev. *B Page 28 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
CYV15G0404DXB AC Electrical Characteristics (continued)
Parameter Description Min. Max Unit
t
TREFDS
t
TREFDH
t
RREFDA
t
RREFDW
t
REFxDV–
t
REFxDV+
[28]
t
REFRX

CYV15G0404DXB Bus Configuration Write Timing Characteristics Over the Operating Range

t
DATAH
t
DATAS
t
WRENP

CYV15G0404DXB JTAG Test Clock Characteristics Over the Operating Range

f
TCLK
t
TCLK

CYV15G0404DXB Device RESET Characteristics Over the Operating Range

t
RST

CYV15G0404DXB Transmit Serial Outputs and TX PLL Characteristics Over the Operating Range

Parameter Description Condition Min. Max. Unit
t
B
Notes
26.Since this timing parameter is greater than the minimum time period of REFCLK it se t s an u pper li mit to the f requency in whi ch REFCLKx can be used t o 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 t time of the upstream device. When t his condit ion is n ot tru e, RXCLKx± (a b uf fered or divided ve rsion of REFCLK when RXCK SELx = 1) could b e used to clock the receive data out of the device.
27.Measured using a 50% duty cycle reference clock
28.REFCLKx± has no phase or frequency relationship with the recovered clock(s) and only acts as a centering reference to reduce clock synchronization time. REFCLKx± must be within ±1500 PPM (±0.15%) of the transmit ter PLL reference (REFCLKx±) frequency. Although transmitting to a HOTLink II receiver neces­sitates the frequency difference between the transmitter and receiver ref erence clocks to be within ±1500-PPM, the st ability of the cr ystal needs to be within t he limits specified by the appropriate standard when transmitti ng 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.
29.While sending continuous K28.5s, outputs loaded to a balanced 100Ω load, measured at the cross point of differential outputs, over the operating range.
30.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.
31.Total jitter is calculated at an assumed BER of 1E -12. Hence: Total Jitter (tJ) = (tRJ * 14) + tDJ.
32.Also meets all Jitter Generation and Jitter Tolerance requirements as specified by SMPTE 259, SMPTE 292, ESCON, FICON, Fibre Channel, and DVB-ASI.
Transmit Data Set-up Time to REFCLKx - Full Rate
2.4 ns
(TXRATEx = 0, TXCKSELx = 1) Transmit Data Set-up Time to REFCLKx - Half Rate
2.3 ns
(TXRATEx = 1, TXCKSELx = 1) Transmit Data Hold Time from REFCLKx - Full Rate
1.0 ns
(TXRATEx = 0, TXCKSELx = 1) Transmit Data Hold Time from REFCLKx - Half Rate
1.6 ns
(TXRATEx = 1, TXCKSELx = 1) Receive Data Access Time to REFCLKx (RXCKSELx = 1) 9.7
[26]
Receive Data Valid Time Window (RXCKSELx = 1) 10UI – 5.8 ns Received Data Valid Time to RXCLK when RXCKSELx = 1
10UI
[24 ]
– 6.16 ns
(TXRATEx = 0, RXRATEx = 0) Received Data Valid Time to RXCLK when RXCKSELx = 1
5UI – 2.53
[27]
(TXRATEx = 0, RXRATEx = 1)
[27]
Received Data Valid T ime to RXCLK when RXCKSELx = 1 (TXRATEx = 1) 10UI – 5.86 Received Data Valid Time from RXCLK when RXCKSELx = 1
1.4 ns
ns
(TXRATEx = 0, RXRATEx = 0) Received Data Valid Time from RXCLK when RXCKSELx = 1
5UI – 1.83
[27]
(TXRATEx = 0, RXRATEx = 1) Received Data Valid Time from RXCLK when RXCKSELx = 1
1.0
[27]
(TXRATEx = 1) REFCLKx Frequency Referenced to Received Clock Period –0.15 +0.15 %
Bus Configuration Data Hold 0 ns Bus Configuration Data Setup 10 ns Bus Configuration WREN Pulse Width 10 ns
JTAG Test Clock Frequency 20 MHz JTAG Test Clock Period 50 ns
Device RESET Pulse Width 30 ns
Bit Time 5128 666 ps
and set-up
RREFDA
ns
ns
ns
ns
Document #: 38-02097 Rev. *B Page 29 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
CYV15G0404DXB AC Electrical Characteristics (continued)
TXCLKx
TXDx[7:0],
TXCTx[1:0],
t
TXDH
t
TXDS
t
TXCLK
t
TXCLKH
t
TXCLKL
Transmit Interface Write Timing TXCLKx selected
REFCLKx
Transmit Interface
t
REFCLK
t
REFH
t
REFL
t
TREFDS
t
TREFDH
Write Timing TXRATEx = 0
TXDx[7:0],
TXCTx[1:0],
REFCLKx selected
Parameter Description Min. Max Unit
[20]
t
RISE
[20]
t
FALL
[20, 29, 31]
t
DJ
[20, 30, 31]
Z
RJ
[20]
t
REFJ
t
TXLOCK

CYV15G0404DXB Receive Serial Inputs and CDR PLL Characteristics Over the Operating Range

t
RXLOCK
t
RXUNLOCK
[20]
t
JTOL
[20]
t
DJTOL
CML Output Rise Time 2080% (CML Test Load) SPDSELx = HIGH 60 270 ps
SPDSELx = MID 100 500 ps SPDSELx =LOW 180 1000 ps
CML Output Fall Time 8020% (CML Test Load) SPDSELx = HIGH 60 270 ps
SPDSELx = MID 100 500 ps
SPDSELx =LOW 180 1000 ps Deterministic Jitter (peak-peak) Random Jitter (σ)
[32]
[32]
IEEE 802.3z 27 ps
IEEE 802.3z 11 ps REFCLKx jitter tolerance / Phase noise limits TBD Transmit PLLx lock to REFCLKx± 200 μs
Receive PLL lock to input data stream (cold start) 376k UI Receive PLL lock to input data strea m 376k UI Receive PLL Unlock Rate 46 UI Total Jitter Tolerance Deterministic Jitter Tolerance
[32]
[32]
IEEE 802.3z 600 ps
IEEE 802.3z 370 ps
Capacitance
[20]
Parameter Description Test Conditions Max. Unit
C
INTTL
C
INPECL
TTL Input Capacitance TA = 25°C, f0 = 1 MHz, VCC = 3.3V 7 pF PECL input Capacitance TA = 25°C, f0 = 1 MHz, VCC = 3.3V 4 pF

CYV15G0404DXB HOTLink II Transmitter Switching Waveforms

Document #: 38-02097 Rev. *B Page 30 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
CYV15G0404DXB HOTLink II Transmitter Switching Waveforms (continued)
t
TREFDH
Transmit Interface Write Timing
TXRATEx = 1
REFCLKx
t
REFCLK
t
REFL
t
REFH
Note 33
TXDx[7:0],
TXCTx[1:0],
t
TREFDS
t
TREFDS
t
TREFDH
REFCLKx selected
TXCLKOx
t
TXCLKO
Transmit Interface TXCLKOx Timing
TXRATEx = 1
(internal)
REFCLKx
t
REFCLK
t
REFL
t
REFH
Note 34
Note 35
TXCLKOx
t
TXCLKO
t
TXOH
t
TXOL
Transmit Interface TXCLKOx Timing
REFCLKx
Note 34
Note 35
t
REFCLK
t
REFH
t
REFL
TXRATEx = 0
Note
33.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.
Document #: 38-02097 Rev. *B Page 31 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

Switching Waveforms for the CYV15G0404DXB HOTLink II Receiver

REFCLKx
RXDx[7:0],
RXSTx[2:0],
t
REFCLK
t
REFH
t
REFL
Receive Interface Read Timing
full rate RXCLKx±
RXCLKx
t
REFxDV+
t
REFxDV
t
RREFDA
REFCLKx Selected
t
RREFDW
t
RREFDW
TXERRx
[36]
REFCLKx
RXDx[7:0],
RXSTx[2:0],
t
REFCLK
t
REFH
t
REFL
Receive Interface Read Timing
half rate RXCLKx±
t
RREFDA
RXCLKx
t
REFxDV+
t
REFxDV
Note 37
t
RREFDA
REFCLKx Selected
t
RREFDW
t
RREFDW
TXERRx
[36]
RXCLKx+
RXDx[7:0],
RXSTx[2:0],
t
RXDV+
t
RXCLKP
Receive Interface Read Timing
RXRATEx = 0
RXCLKx-
t
RXDV
Recovered Clock selected
Notes
34.The TXCLKOx output remains at the character rate regardless of the state of TXRATE and does not follow the duty cycle of REFCLKx±.
35.The rising edge of TXCLKOx output has no direct phase relationship to the REFCLKx± input.
36.TXERRx is synchronous to RXCLKx only when RXCLKx is selected as REFCLK.
Document #: 38-02097 Rev. *B Page 32 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Switching Waveforms for the CYV15G0404DXB HOTLink II Receiver
RXCLKx+
RXDx[7:0],
RXSTx[2:0]
t
RXDV+
t
RXDV
t
RXCLKP
Receive Interface Read Timing
RXRATEx = 1
RXCLKx-
Recovered Clock selected
ADDR[3:0]
t
DATAS
Bus Configuration Write Timing
DATA[7:0]
WREN
t
DATAH
t
WRENP
Note
37.When operated with a half rate REFCLKx±, the setup and hold specifications for data relative to RXCLKx are relative to both rising and falling edges of th e respective clock output
Document #: 38-02097 Rev. *B Page 33 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 11. Package Coordinate Signal Allocation
Ball
Signal Name Signal Type
ID
A01 INC1– CML IN C07 ULCC LVTTL IN PU F17 RCLKENA LVTTL IN PD A02 OUTC1– CML OUT C08 GND GROUND F18 RXSTB[1] LVTTL OUT A03 INC2– CML IN C09 DA TA[7] LVTTL IN PU F19 TXCLKOB LVTTL OUT A04 OUTC2– CML OUT C10 DA TA[5] LVTTL IN PU F20 RXSTB[0] LVTTL OUT A05 VCC POWER C11 DATA[3] LVTTL IN PU G01 TXDC[7] LVTTL IN A06 IND1– CML IN C12 DATA[1] LVTTL IN PU G02 WREN LVTTL IN PU A07 OUTD1– CML OUT C13 GND GROUND G03 TXDC[4] LVTTL IN A08 GND GROUND C14 RCLKENB LVTTL IN PD G04 TXDC[1] LVTTL IN A09 IND2– CML IN C15 SPDSELD 3-LEVEL SEL G17 SPDSELB 3-LEVEL SEL A10 OUTD2– CML OUT C16 VCC POWER G18 LPENC LVTTL IN PD
A1 1 INA1– CML IN C17 LDTDEN LVTTL IN PU G19 SPDSELA 3-LEVEL SEL A12 OUTA1– CML OUT C18 TRST LVTTL IN PU G20 RXDB[1] LVTTL OUT A13 GND GROUND C19 LPEND LVTTL IN PD H0 1 G ND GROUND A14 I NA2– CML IN C20 TDO LVTTL 3-S OUT H02 GND GROUND A15 OUTA2– CML OUT D01 TCLK LVTTL IN PD H03 GND GROUND A16 VCC POWER D02 RESET A17 INB1– CML IN D03 INSELD LVTTL IN H17 GND GROUND A18 OUTB1– CML OUT D04 INSELA LVTTL IN H18 GND GROUND A19 INB2– CML IN D05 VCC POWER H19 GND GROUND A20 OUTB2– CML OUT D06 ULCA LVTTL IN PU H20 GND GROUND B01 INC1+ CML IN D07 SPDSELC 3-LEVEL SEL J01 TXCTC[1] LVTTL IN B02 OUTC1+ CML OUT D08 GND GROUND J02 TXDC[5] LVTTL IN B03 INC2+ CML IN D09 DATA[6] LVTTL IN PU J03 TXDC[2] LVTTL IN B04 OUTC2+ CML OUT D10 DATA[4] LVTTL IN PU J04 TXDC[3] LVTTL IN B05 VCC POWER D11 DATA[2] LVTTL IN PU J17 RXSTB[2] LVTTL OUT B06 IND1+ CML IN D12 DATA[0] LVTTL IN PU J18 RXDB[0] LVTTL OUT B07 OUTD1+ CML OUT D13 GND GROUND J19 RXDB[5] LVTTL OUT B08 GND GROUND D14 LPENB LVTTL IN PD J20 RXDB[2] LVTTL OUT B09 IND2+ CML IN D15 ULCB LVTTL IN PU K01 RXDC[2] LVTTL OUT B10 OUTD2+ CML OUT D16 VCC POWER K02 REFCLKC– PECL IN
B1 1 INA1+ CML IN D17 LPENA LVTTL IN PD K03 TXCTC[0] LVTTL IN B12 OUTA1+ CML OUT D18 VCC POWER K04 TXCLKC LVTTL IN PD B13 GND GROUND D19 SCANEN2 LVTTL IN PD K17 RXDB[3] LVTTL OUT B14 INA2+ CML IN D20 TMEN3 LVTTL IN PD K18 RXDB[4] LVTTL OUT B15 OUTA2+ CML OUT E01 VCC POWER K19 RXDB[7] LVTTL OUT B16 VCC POWER E02 VCC POWER K20 LFIB LVTTL OUT B17 INB1+ CML IN E03 VCC POWER L01 RXDC[3] LVTTL OUT B18 OUTB1+ CML OUT E04 VCC POWER L02 REFCLKC+ PECL IN B19 INB2+ CML IN E17 VCC POWER L03 LFIC LVTTL OUT B20 OUTB2+ CML OUT E18 VCC POWER L04 TXDC[6] LVTTL IN C01 TDI LVTTL IN PU E19 VCC POWER L17 RXDB[6] LVTTL OUT C02 TMS LVTTL IN PU E20 VCC POWER L18 RXCLKB+ LVTTL OUT C03 INSELC LVTTL IN F01 RXDC[6] LVTTL OUT L19 RXCLKB– LVTTL OUT
Ball
Signal Name Signal Type
ID
Ball
Signal Name Signal Type
ID
LVTTL IN PU H04 GND GROUND
Document #: 38-02097 Rev. *B Page 34 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 11. Package Coordinate Signal Allocation (continued)
Ball
Signal Name Signal Type
ID
C04 INSELB LVTTL IN F02 RXDC[7] LVTTL OUT L20 TXDB[6] LVTTL IN C05 VCC POWER F03 TXDC[0] LVTTL IN M01 RXDC[4] LVTTL OUT C06 ULCD LVTTL IN PU F04 RCLKEND LVTTL IN PD M02 RXDC[5] LVTTL OUT
M03 RCLKENC LVTTL IN PD U03 TXDD[2] LVTTL IN W03 LFID M04 TXERRC LVTTL OUT U04 TXCTD[1] LVTTL IN W04 RXCLKD– LVTTL OUT M17 REFCLKB+ PECL IN U05 VCC POWER W05 VCC POWER M18 REFCLKB– PECL IN U06 RXDD[2] LVTTL OUT W06 RXDD[4] LVTTL OUT M19 TXERRB LVTTL OUT U07 RXDD[1] LVTTL OUT W07 RXSTD[1] LVTTL OUT M20 TXCLKB LVTTL IN PD U08 GND GROUND W08 GND GROUND
N01 GND GROUND U09 TXCTA[1] LVTTL IN W09 ADDR [3] LVTTL IN PU N02 GND GROUND U10 ADDR [0] LVTTL IN PU W10 ADDR [1] LVTTL IN PU N03 GND GROUND U11 REFCLKD– PECL IN W11 RXCLKA+ LVTTL OUT N04 GND GROUND U12 TXDA[1] LVTTL IN W12 TXERRA LVTTL OUT N17 GND GROUND U13 GND GROUND W13 GND GROUND N18 GND GROUND U14 TXDA[4] LVTTL IN W14 TXDA[2] LVTTL IN N19 GND GROUND U15 TXCTA[0] LVTTL IN W15 TXDA[6] LVTTL IN N20 GND GROUND U16 VCC POWER W16 VCC POWER P01 RXDC[1] LVTTL OUT U17 RXDA[2] LVTTL OUT W17 LFIA LVTTL OUT P02 RXDC[0] L VTTL OUT U18 TXCTB[0] LVTTL IN W18 REFCLKA+ PECL IN P03 RXSTC[0] LVTTL OUT U19 RXSTA[2] LVTTL OUT W19 RXDA[4] LVTTL OUT P04 RXSTC[1] LVTTL OUT U20 RXSTA[1] LVTTL OUT W20 RXDA[1] LVTTL OUT P17 TXDB[5] LVTTL IN V01 TXDD[3] LVTTL IN Y01 TXDD[6] LVTTL IN P18 TXDB[4] LVTTL IN V02 TXDD[4] LVTTL IN Y02 TXCLKD LVTTL IN PD P19 TXDB[3] LVTTL IN V03 TXCTD[0] LVTTL IN Y03 RXDD[7] LVTTL OUT P20 TXDB[2] LVTTL IN V04 RXDD[6] LVTTL OUT Y04 RXCLKD+ LVTTL OUT R01 RXSTC[2] LVTTL OUT V05 VCC POWER Y05 VCC POWER R02 TXCLKOC LVTTL OUT V06 RXDD[3] LVTTL OUT Y06 RXDD[5] LVTTL OUT R03 RXCLKC+ LVTTL OUT V07 RXSTD[0] LVTTL OUT Y07 RXDD[0] LVTTL OUT R04 RXCLKC– LVTTL OUT V08 GND GROUND Y08 GND GROUND R17 TXDB[1] LVTTL IN V09 RXSTD[2] LVTTL OUT Y09 TXCLKOD LVTTL OUT R18 TXDB[0] LVTTL IN V10 ADDR [2] LVTTL IN PU Y10 NC NO CONNECT R19 TXCTB[1] LVTTL IN V11 REFCLKD+ PECL IN Y11 TXCLKA LVTTL IN PD R20 TXDB[7] LVTTL IN V12 TXCLKOA LVTTL OUT Y12 RXCLKA– LVTTL OUT
T01 VCC POWER V13 GND GROUND Y13 GND GROUND
T02 VCC POWER V14 TXDA[3] LVTTL IN Y14 TXDA[0] LVTTL IN
T03 VCC POWER V15 TXDA[7] LVTTL IN Y15 TXDA[5] LVTTL IN
T04 VCC POWER V16 VCC POWER Y16 VCC POWER
T17 VCC POWER V17 RXDA[7] LVTTL OUT Y17 TXERRD LVTTL OUT
T18 VCC POWER V18 RXDA[3] LVTTL OUT Y18 REFCLKA– PECL IN
T19 VCC POWER V19 RXDA[0] LVTTL OUT Y19 RXDA[6] LVTTL OUT
T20 VCC POWER V20 RXSTA[0] LVTTL OUT Y20 RXDA[5] LVTTL OUT U01 TXDD[0] LVTTL IN W01 TXDD[5] LVTTL IN U02 TXDD[1] LVTTL IN W02 TXDD[7] LVTTL IN
Ball
Signal Name Signal Type
ID
Ball
Signal Name Signal Type
ID
LVTTL OUT
Document #: 38-02097 Rev. *B Page 35 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

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 8-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 b y 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 bi t stre a m .

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
Converted to 8B/10B notation, note that the order of bits has been reversed):
Data Byte Name D5.2
Bits: ABCDE
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 c
3210
0100 0101
FGH
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.
Using these 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 (fg hj) 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-bi t trans­mission code.
A.X. Widmer and P.A. Franaszek. “A DC-Balanced, Parti-
tioned-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 trans-
Document #: 38-02097 Rev. *B Page 36 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
mission of any transmission character, the transmitter selects 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 cod es 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.
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 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 subblocks, where the first six bits (abcdei) form one subblock and the second four bits (fghj) form the other subblock. Running disparity at the beginning of the 6-bit subblock is the running disparity at the end of the previous transmission character. running disparity at the beginning of the 4-bit subblock is the running disparity at the end of the 6-bit subblock. Running disparity at the end of the transmission character is the running disparity at the end of the 4-bit subblock.
Running disparity for the subblocks is calculated as follows:
1. Running disparity at the end of any subblock is positive if the subblock contains more ones than zeros. It is also positive at the end of the 6-bit subblock if the 6-bit subblock is 000111, and it is positive at the end of the 4-bit subblock if the 4-bit subblock is 0011.
2. Running disparity at the end of any subblock is negative if the subblock contains more zeros than ones. It is also negative at the end of the 6-bit subblock if the 6-bit subblock is 11 1000, and it is negative at the end of the 4-bit subblock if the 4-bit subblock is 1100.
3. Otherwise, running disparity at the end of the subblock is the same as at the beginning of the subblock.
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 12 shows naming notations and examples of valid transmission characters.

Use of the T ables for Checking the V alidity of Received Transmission Characters

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 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.
Table 12. Valid Transmission Characters
Data
Byte Name
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
DIN or Q
765 43210
. .
. .
OUT
. .
. .
Hex Value
. .
. .

Use of the Tables for Generating Transmission Characters

The appropriate entry in Table 14 for the valid data byte or
Table 15 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 character
Table 13. 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 +
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 12 shows an example of this behavior.
Document #: 38-02097 Rev. *B Page 37 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 14. Valid Data Characters (TXCTx [0] = 0, RXSTx[2:0] = 000)
Data Byte
Name
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
Bits Current RD Current RD+
HGF EDCBA abcdei fghj abcdei fghj HGF EDCBA abcdei fghj abcdei fghj
Data
Byte
Name
Bits Current RD Current RD+
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 #: 38-02097 Rev. *B Page 38 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 14. Valid Data Characters (TXCTx [0] = 0, RXSTx[2:0] = 000) (continued)
Data Byte
Name
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
Bits Current RD Current RD+
HGF EDCBA abcdei fghj abcdei fghj HGF EDCBA abcdei fghj abcdei fghj
Data
Byte
Name
Bits Current RD Current RD+
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 #: 38-02097 Rev. *B Page 39 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 14. Valid Data Characters (TXCTx [0] = 0, RXSTx[2:0] = 000) (continued)
Data Byte
Name
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
Bits Current RD Current RD+
HGF EDCBA abcdei fghj abcdei fghj HGF EDCBA abcdei fghj abcdei fghj
Data
Byte
Name
Bits Current RD Current RD+
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 #: 38-02097 Rev. *B Page 40 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 14. Valid Data Characters (TXCTx [0] = 0, RXSTx[2:0] = 000) (continued)
Data Byte
Name
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
Bits Current RD Current RD+
HGF EDCBA abcdei fghj abcdei fghj HGF EDCBA abcdei fghj abcdei fghj
Data
Byte
Name
Bits Current RD Current RD+
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 #: 38-02097 Rev. *B Page 41 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Table 15. Valid Special Char acter Codes and Sequences (TXCTx = special character code or RXSTx[2:0] = 001)
[38, 39]
S.C. Byte Name
S.C. Code Name
Cypress Alternate
S.C. Byte
[40]
Name
Bits
HGF EDCBA
S.C. Byte
[40]
Name
Bits
HGF EDCBA
Current RD
abcdei fghj
Current RD+
abcdei fghj
K28.0 C0.0 (C00) 000 00000 C28.0 (C1C) 000 1 1100 001111 0100 110000 1011
[41]
K28.1 K28.2
[41]
C1.0 (C01) 000 00001 C28.1 (C3C) 001 11100 0011 11 1001 110000 0110 C2.0 (C02) 000 00010 C28.2 (C5C) 010 11100 0011 11 0101 110000 1010
K 28.3 C3.0 (C03) 000 00011 C28.3 (C7C) 011 11100 001111 0011 110000 1100
[41]
K28.4 K28.5 K28.6 K28.7
[41, 42] [41] [41, 43]
C4.0 (C04) 000 00100 C28.4 (C9C) 100 11100 0011 11 0010 110000 1101 C5.0 (C05) 000 00101 C28.5 (CBC) 101 11100 0011111010 110000 0101 C6.0 (C06) 000 00110 C28.6 (CDC) 110 11100 001111 0110 110000 1001
C7.0 (C07) 000 00111 C28.7 (CFC) 111 11100 001111 1000 110000 0111 K23.7 C8.0 (C08) 000 01000 C23.7 (CF7) 11 1 10111 111010 1000 000101 0111 K27.7 C9.0 (C09) 000 01001 C27.7 (CFB) 111 11011 1 10110 1000 001001 0111 K29.7 C10.0 (C0A) 000 01010 C29.7 (CFD) 11 1 11101 101110 1000 010001 0111 K30.7 C11.0 (C0B) 000 01011 C30.7 (CFE) 111 11110 0111101000 100001 0111
End of Frame Sequence
EOFxx C2.1 (C22) 001 00010 C2.1 (C22) 001 00010 K28.5,Dn.xxx0
[44]
+K28.5,Dn.xxx1
Code Rule Violation and SVS Tx Pattern
[46]
[47]
[43, 45 ]
C0.7 (CE0) 111 00000 C0.7 (CE0) 111 00000 100111 1000 011000 0111
C1.7 (CE1) 111 00001 C1.7 (CE1) 111 00001 0011111010 00111 11010
C2.7 (CE2) 111 00010 C2.7 (CE2) 111 00010 110000 0101 110000 0101
Exception
K28.5 +K28.5
Running Disparity Violation Pattern
Exception
Notes
38.All codes not shown are reserved.
39.Notation for Special Character Code Name is consistent with Fibre Channel and ESCON naming conventions. Special Chara cter Code Name is intended to
40.Both the Cypress and alternate encodings may be used for data transmission to generate specific Special Character Codes. The decoding process for received
41.These characters are used for control of ESCON interfaces. They can be sent as embedded commands or other markers when not operating using ESCON
42.The K28.5 character is used for framing operations by the receiver. It is also the pad or fill cha racter transmitte d to main t ain the se rial link when no u ser data is
43.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.
44.C2.1 = Transmit either –K28.5+ or +K28.5– as determined by Current RD and modi fy t he Transmission Character that follows, by setting its least significant bit
45.C0.7 = Transmit a deliberate code rule violation. The code chosen for this function follows the normal Running Disparity rules. Transmission of this Special
46.C1.7 = Transmit Negative K28.5 (K28.5+) disregarding Current RD. The receiver only outputs t his Special Character if K28.5 is received with the wrong running
47.C2.7 = Transmit Positive K28.5 (+K28.5) disregarding Current RD. The receiver on ly output s this Special Character if K28.5 is received with the wrong running
48.C4.7 = Transmit a deliberate code rule violation to indicate a Running Disparity violation. The receiver only outputs this Special Character if the Transmission
[48]
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).
characters generates Cypress codes or Alternate codes as selected by the BOE[7:0] configur ation inputs. protocols. available. These sequences can cause erroneous framing and should be avoided while RFENx = 1. to 1 or 0. If Current RD at the start of t he following character is plus (+) t he LSB is set to 0, a nd if Curren t RD is minus ( –) t he LSB become s 1. This mod ification
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– D2 1.4–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.
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.
disparity. The receiver outputs C1.7 if K28.5 is received with RD+, otherwise K28.5 is decoded as C5.0 or C2.7. disparity. The receiver outputs C2.7 if +K28.5 is received with RD, otherwise K28.5 is decoded as C5.0 or C1.7. 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.
C4.7 (CE4) 111 00100 C4.7 (CE4) 111 00100 110111 0101 001000 1010
[44]
Document #: 38-02097 Rev. *B Page 42 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB

Ordering Information

256-Lead L2 Ball Grid Array (27 x 27 x 1.57 mm) BL256
51-85123-*E
Speed Ordering Code Package Name Package Type
Standard CYV15G0404DXB-BGC BL256 256-Ball Thermally Enhanced Ball Grid Array Commercial Standard CYV15G0404DXB-BGI BL256 256-Ball Ther mally Enhanced Ball Grid Array Industrial
Operating
Range

Package Diagram

Document #: 38-02097 Rev. *B Page 43 of 44
[+] Feedback [+] Feedback
CYV15G0404DXB
Document History Page
Document Title: CYV15G0404DXB Independent Clock Quad HOTLink II™ Transceiver with Reclocker Document Number: 38-02097
REV. ECN NO.
ISSUE
DATE
** 231494 See ECN BCD New Data Sheet
*A 384307 See ECN AGT Revised setup and hold times (t
*B 1845306 See ECN UKK/VED
ORIG. OF CHANGE
DESCRIPTION OF CHANGE
, t
TXDH
t
RXDv–
, t
RXDv+, tTREFDS
, t
REFxDV–, tREFxDV+
TXDS, tTREFDH
, t
RST, tRISE, tFALL, tDJ
Added clarification for the necessity of JTAG controller reset and the methods to implement it.
, t
RXDv+, tTXCLKOD,
)
© Cypress Semiconductor Corporation, 2005- 2007. The infor mation cont ain ed herein is subj ect to change wi thout notice. C ypress 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 fo r medical, life support, life saving, critica l contr o l or safety applications, unless pursuant to an exp re ss wr itte n agreement with Cypress. Furthermore, Cypress does not auth or iz e it s pr o ducts for use as critical components in life-support systems where a malfunction or fa ilure may reasonably be expe cted to result in significa nt injury to the us er . The inclu sion of Cypress p roducts 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 gr ant s to l icense e a pers onal, no n-exclu sive , non-tr ansfer able 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 w ith a Cypress integrated circuit as specified in the ap plicable agreem ent. Any reprod uction, modificatio n, translation, co mpilation, or repr esentation of this Source Co de except as speci fied above is pro hibited with out 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 n ot assume any liability arising out of the applic ation or use o f any pr oduct or circ uit de scribed herein . Cypr ess does n ot author ize its p roducts fo r use as critical compon ents in life-su pport systems whe re a malfunction or failure may reason ably 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 #: 38-02097 Rev. *B Revised December 14, 2007 Page 44 of 44
IBM and ESCON are registered trademarks, and FICON is a trademark, of International Business Machines. HOTLink is a registered trademark and HOTLink II and MultiFrame are trademarks of Cypress Semiconductor. All product and company names mentioned in this document may be the trademarks of their respective holders.
[+] Feedback [+] Feedback
Loading...