CYPRESS CYV15G0404DXB

CYV15G0404DXB
PRELIMINARY
Independent Clock Quad HOTLink II™ Transceiver
with Reclocker
Features
• Quad channel transceiver for 195- to 1500-MBaud serial
signaling rate
— Aggregate throughput of up to 12 Gbits/second
• Second-generation HOTLink® technology
• Compliant to multiple standards
— SMPTE-292M, SMPTE-259M, DVB-ASI, Fibre
Channel, ESCON and Gigabit Ethernet (IEEE802.3z)
— 10 bit uncoded data or 8B/10B coded data
• Truly independent channels
— Each channel can perform reclocker function
— Each channel can operate at a different signaling
rate
— Each channel can 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 required
— Signaling-rate controlled edge-rates
— Source matched for 50Ω transmission lines
• MultiFrame™ Receive Framer provides alignment
options
— Comma or Full K28.5 detect
— Single or Multi-byte Framer for byte alignment
— Low-latency option
•
•
•
•
•
•
•
•
•
•
Selectable input/output clocking options
Synchronous LVTTL parallel interface
JTAG boundary scan
Built-In Self-Test (BIST) for at-speed link testing
Per-channel Link Quality Indicator
— Analog signal detect
— Digital signal detect
Low-power 3W @ 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 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 per serial link.
Each channel operates independently with its own reference
clock allowing different rates. Each transmit channel accepts
parallel characters in an Input Register, encodes each
character for transport, and then converts it to serial data.
Each receive channel accepts serial data and converts it to
parallel data, decodes the data into characters, and presents
these characters to an Output Register. The received serial
data can also be reclocked and retransmitted through the
serial outputs. Figure 1 illustrates typical connections between
independent video co-processors and corresponding
CYV15G0404DXB chips.
10
10
Serial Links
10
Serial Links
10
10
Video Coprocessor
10
10
Independent
Channel
CYV15G0404DXB
Reclocker
10
Serial Links
10
Serial Links
Independent
Channel
CYV15G0404DXB
Reclocker
10
10
Video Coprocessor
10
10
10
10
10
Cable
Connections
Figure 1. HOTLink II™ System Connections
Cypress Semiconductor Corporation
Document #: 38-02097 Rev. **
•
3901 North First Street
•
San Jose, CA 95134
•
408-943-2600
Revised June 04, 2004
CYV15G0404DXB
PRELIMINARY
The CYV15G0404DXB satisfies the SMPTE-259M and
SMPTE-292M compliance as per 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
can accept either 8-bit data characters or preencoded 10-bit
transmission characters. Data characters may be passed from
the Transmit Input Register to an integrated 8B/10B Encoder
to improve their serial transmission characteristics. These
encoded characters are then serialized and output from dual
Positive ECL (PECL) compatible differential transmission-line
drivers at a bit-rate of either 10- or 20-times the input reference
clock for that channel.
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/decoder may be bypassed for
systems that present externally encoded or scrambled data at
the parallel interface.
The parallel I/O 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 multiformat routers and switchers.
TXDD[7:0]
TXCTD[1:0]
x11
x10
x11
Phase
Align
Buffer
Elasticity
Buffer
Phase
Align
Buffer
Elasticity
Buffer
Phase
Align
Buffer
Elasticity
Buffer
Phase
Align
Buffer
Elasticity
Buffer
Encoder
8B/10B
Decoder
8B/10B
Encoder
8B/10B
Decoder
8B/10B
Encoder
8B/10B
Decoder
8B/10B
Encoder
8B/10B
Decoder
8B/10B
TX
RX
TX
RX
TX
RX
INB1±
INB2±
OUTC1±
OUTC2±
INC1±
INC2±
Document #: 38-02097 Rev. **
Serializer
TX
Deserializer
RX
IND1±
IND2±
Deserializer
OUTB1±
OUTB2±
Deserializer
Serializer
INA1±
INA2±
Serializer
Deserializer
OUTA1±
OUTA2±
Serializer
Framer
Framer
OUTD1±
OUTD2±
Framer
Framer
RXDD[7:0]
RXSTD[2:0]
RXDC[7:0]
RXSTC[2:0]
x10
REFCLKD±
TXDC[7:0]
TXCTC[1:0]
x11
REFCLKC±
RXDB[7:0]
RXSTB[2:0]
x10
REFCLKB±
TXDB[7:0]
TXCTB[1:0]
x11
REFCLKA±
x10
TXDA[7:0]
TXCTA[1:0]
RXDA[7:0]
RXSTA[2:0]
CYV15G0404DXB Transceiver Logic Block Diagram
Page 2 of 43
CYV15G0404DXB
PRELIMINARY
Transmit Path Block Diagram
REFCLKA+
RECLCK[A..D] are Internal Reclocker Signals
TXLB[A..D] are Internal Serial Loopback Signals
Bit-Rate Clock
REFCLKA–
Transmit
Transmit PLL
PLL
Clock Multiplier
A
Clock
Multiplier
TXRATEA
SPDSELA
TXCLKOA
ENCBYPA
Character-Rate Clock A
TXERRA
TXCLKA
= Internal Signal
OEA[2..1]
RECLCKA
2
TXCTA[1:0]
10
10
10
OUTA1+
OUTA1–
Shifter
Input
Register
10
Encoder
8B/10B
Encoder
8
TXDA[7:0]
BIST
BIST LFSR
LFSR
1
Phase-Align
Phase-Align
Buffer
Buffer
0
TXCKSELA
OEA[2..1]
TXBISTA
PABRSTA
OUTA2+
OUTA2–
TXLBA
REFCLKB+
Bit-Rate Clock
REFCLKB–
Transmit PLL
Clock Multiplier B
TXRATEB
SPDSELB
TXCLKOB
OEB[2..1]
Character-Rate Clock B
ENCBYPB
RECLCKB
TXERRB
0
10
10
OUTB1+
OUTB1–
Shifter
2
BIST
BIST LFSR
LFSR
TXDB[7:0]
10
Encoder
8B/10B
Encoder
10
Input
Register
8
OEB[2..1]
1
Phase-Align
Phase-Align
Buffer
Buffer
TXCKSELB
TXCTB[1:0]
TXBISTB
PABRSTB
TXCLKB
OUTB2+
OUTB2–
TXLBB
REFCLKC+
Bit-Rate Clock
REFCLKC–
Transmit PLL
Clock Multiplier C
TXRATEC
SPDSELC
TXCLKOC
OEC[2..1]
RECLCKC
ENCBYPC
Character-Rate Clock C
TXERRC
1
BIST LFSR
2
10
8B/10B
Encoder
Input
Register
TXCTC[1:0]
10
Phase-Align
Buffer
8
TXDC[7:0]
10
10
OUTC1+
OUTC1–
Shifter
0
TXCKSELC
OEC[2..1]
TXBISTC
PABRSTC
TXCLKC
OUTC2+
OUTC2–
TXLBC
REFCLKD+
Bit-Rate Clock
REFCLKD–
Transmit PLL
Clock Multiplier D
TXRATED
SPDSELD
ENCBYPD
TXCLKOD
OED[2..1]
Character-Rate Clock D
TXERRD
TXBISTD
PABRSTD
TXCLKD
Document #: 38-02097 Rev. **
10
OUTD1+
OUTD1–
Shifter
10
10
BIST LFSR
2
10
8B/10B
Encoder
TXDD[7:0]
1
Phase-Align
Buffer
8
0
Input
Register
TXCKSELD
TXCTD[1:0]
RECLCKD
OED[2..1]
OUTD2+
OUTD2–
TXLBD
Page 3 of 43
CYV15G0404DXB
PRELIMINARY
Receive Path Block Diagram
RECLCK[A..D] are Internal Reclocker Signals
TXLB[A..D] are Internal Serial Loopback Signals
= Internal Signal
RESET
SPDSELA
RXPLLPDA
RCLKENA
TRST
JTAG
Boundary
Scan
Controller
RECLCKA
TMS
TCLK
TDI
TDO
Receive
Signal
Monitor
Elasticity
Buffer
Output
Register
TXLBA
ULCA
Clock &
Data
Recovery
PLL
10B/8B
BIST
INA2+
INA2–
LFIA
Framer
INA1+
INA1–
Shifter
LPENA
INSELA
8
3
RXDA[7:0]
RXSTA[2:0]
SPDSELB
Receive
Signal
Monitor
SPDSELC
Receive
Signal
Monitor
TXLBC
ULCC
SPDSELD
RXPLLPDD
LPEND
Receive
Signal
Monitor
IND1+
IND1–
IND2+
IND2–
TXLBD
ULCD
LDTDEN
Clock &
Data
Recovery
PLL
SDASEL[A..D][1:0]
RFMODE[A..D][1:0]
RFEN[A..D]
FRAMCHAR[A..D]
DECMODE[A..D]
RXBIST[A..D]
RXCKSEL[A..D]
DECBYP[A..D]
RXRATE[A..D]
Document #: 38-02097 Rev. **
Elasticity
Buffer
Output
Register
3
RXDB[7:0]
RXSTB[2:0]
RXCLKB+
RXCLKB–
÷2
8
3
RXDC[7:0]
RXSTC[2:0]
RXCLKC+
RXCLKC–
÷2
LFID
Framer
INSELD
Clock
Select
RECLCKD
Shifter
RCLKEND
Clock &
Data
Recovery
PLL
10B/8B
BIST
INC2+
INC2–
Framer
INC1+
INC1–
8
LFIC
Shifter
INSELC
10B/8B
BIST
LPENC
Clock
Select
RECLCKC
Output
Register
RXPLLPDC
RCLKENC
Clock
Select
Output
Register
ULCB
Elasticity
Buffer
TXLBB
Clock &
Data
Recovery
PLL
10B/8B
BIST
INB2+
INB2–
Framer
INB1+
INB1–
RXCLKA+
RXCLKA–
÷2
LFIB
Shifter
LPENB
INSELB
Clock
Select
RECLCKB
Elasticity
Buffer
RXPLLPDB
RCLKENB
÷2
8
3
RXDD[7:0]
RXSTD[2:0]
RXCLKD+
RXCLKD–
Page 4 of 43
PRELIMINARY
Device Configuration and Control Block Diagram
WREN
ADDR[3:0]
DATA[7:0]
Device Configuration
and Control Interface
Document #: 38-02097 Rev. **
CYV15G0404DXB
= Internal Signal
RFMODE[A..D][1:0]
RFEN[A..D]
FRAMCHAR[A..D]
DECMODE[A..D]
RXBIST[A..D]
RXCKSEL[A..D]
DECBYP[A..D]
RXRATE[A..D]
SDASEL[A..D][1:0]
RXPLLPD[A..D]
TXRATE[A..D]
TXCLKSEL[A..D]
PABRST[A..D]
TXBIST[A..D]
OE[A..D][2..1]
ENCBYP[A..D]
GLEN[11..0]
FLEN[2..0]
Page 5 of 43
CYV15G0404DXB
PRELIMINARY
Pin Configuration (Top View)
A
B
C
D
E
F
G
H
J
K
L
M
N
P
R
T
U
V
W
Y
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
IN
C1–
OUT
C1–
IN
C2–
OUT
C2–
VCC
IN
D1–
OUT
D1–
GND
IN
D2–
OUT
D2–
IN
A1–
OUT
A1–
GND
IN
A2–
OUT
A2–
VCC
IN
B1–
OUT
B1–
IN
B2–
OUT
B2–
IN
C1+
OUT
C1+
IN
C2+
OUT
C2+
VCC
IN
D1+
OUT
D1+
GND
IN
D2+
OUT
D2+
IN
A1+
OUT
A1+
GND
IN
A2+
OUT
A2+
VCC
IN
B1+
OUT
B1+
IN
B2+
OUT
B2+
TDI
TMS
INSELC INSELB
VCC
ULCD
ULCC
GND
DATA
[7]
DATA
[5]
DATA
[3]
DATA
[1]
GND
RCLK
ENB
SPD
SELD
VCC
LDTD
EN
TRST
LPEND
TDO
RESET INSELD INSELA
VCC
ULCA
SPD
SELC
GND
DATA
[6]
DATA
[4]
DATA
[2]
DATA
[0]
GND LPENB ULCB
VCC
LPENA
VCC
SCAN TMEN3
EN2
VCC
TCLK
VCC
VCC
VCC
VCC
VCC
RX
DC[6]
RX
DC[7]
TX
DC[0]
RCLK
END
RCLK
ENA
TX
DC[7]
WREN
TX
DC[4]
TX
DC[1]
SPD
SELB
LP
ENC
PD
SELA
RX
DB[1]
GND
GND
GND
GND
GND
GND
GND
GND
TX
CTC[1]
TX
DC[5]
TX
DC[2]
TX
DC[3]
RX
STB[2]
RX
DB[0]
RX
DB[5]
RX
DB[2]
RX
DB[4]
RX
DB[7]
LFIB
RX
DC[2]
REF
TX
CLKC– CTC[0]
TX
CLKC
RX
DB[3]
RX
DC[3]
REF
CLKC+
LFIC
TX
DC[6]
RX
DB[6]
RX
DC[4]
RX
DC[5]
RCLK
ENC
TX
ERRC
GND
GND
GND
GND
GND
RX
DC[1]
RX
DC[0]
RX
RX
STC[0] STC[1]
RX
TX
RX
RX
STC[2] CLKOC CLKC+ CLKC–
VCC
VCC
VCC
TX
DD[0]
TX
DD[1]
TX
DD[2]
TX
CTD[1]
VCC
RX
DD[2]
RX
DD[1]
GND
TX
CTA[1]
ADDR
REF
[0]
CLKD–
TX
DD[3]
TX
DD[4]
TX
CTD[0]
RX
DD[6]
VCC
RX
DD[3]
RX
STD[0]
GND
RX
STD[2]
TX
DD[5]
TX
DD[7]
LFID
RX
CLKD–
VCC
RX
DD[4]
RX
STD[1]
GND
TX
DD[6]
TX
CLKD
RX
DD[7]
RX
CLKD+
VCC
RX
DD[5]
RX
DD[0]
GND
Document #: 38-02097 Rev. **
TX
DA[1]
VCC
RX
TX
RX
STB[1] CLKOB STB[0]
RX
RX
CLKB+ CLKB–
REF
REF
CLKB+ CLKB–
VCC
VCC
TX
DB[6]
TX
ERRB
TX
CLKB
GND
GND
GND
TX
DB[5]
TX
DB[4]
TX
DB[3]
TX
DB[2]
TX
DB[1]
TX
DB[0]
TX
CTB[1]
TX
DB[7]
VCC
VCC
VCC
VCC
GND
TX
DA[4]
TX
CTA[0]
VCC
RX
DA[2]
TX
RX
CTB[0] STA[2]
RX
STA[1]
ADDR
REF
TX
[2]
CLKD+ CLKOA
GND
TX
DA[3]
TX
DA[7]
VCC
RX
DA[7]
RX
DA[3]
RX
DA[0]
RX
STA[0]
ADDR
[3]
ADDR
[1]
RX
CLKA+
TX
ERRA
GND
TX
DA[2]
TX
DA[6]
VCC
LFIA
REF
CLKA+
RX
DA[4]
RX
DA[1]
TX
CLKOD
NC[1]
TX
CLKA
RX
CLKA–
GND
TX
DA[0]
TX
DA[5]
VCC
TX
ERRD
REF
CLKA–
RX
DA[6]
RX
DA[5]
Page 6 of 43
CYV15G0404DXB
PRELIMINARY
Pin Configuration (Bottom View)
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
A
OUT
B2–
IN
B2–
OUT
B1–
IN
B1–
VCC
OUT
A2–
IN
A2–
GND
OUT
A1–
IN
A1–
OUT
D2–
IN
D2–
GND
OUT
D1–
IN
D1–
VCC
OUT
C2–
IN
C2–
OUT
C1–
IN
C1–
B
OUT
B2+
IN
B2+
OUT
B1+
IN
B1+
VCC
OUT
A2+
IN
A2+
GND
OUT
A1+
IN
A1+
OUT
D2+
IN
D2+
GND
OUT
D1+
IN
D1+
VCC
OUT
C2+
IN
2+
OUT
C1+
IN
C1+
C
TDO
LP
END
TRST
LDTD
EN
VCC
SPD
SELD
RCLK
ENB
GND
DATA
[1]
DATA
[3]
DATA
[5]
DATA
[7]
GND
ULCC
ULCD
VCC
IN
SELB
IN
SELC
TMS
TDI
VCC
LP
ENA
VCC
ULCB
LP
ENB
GND
DATA
[0]
DATA
[2]
DATA
[4]
DATA
[6]
GND
SPD
SELC
ULCA
VCC
IN
SELA
IN
SELD
RESET
TCLK
VCC
VCC
VCC
VCC
VCC
VCC
D
E
TMEN3 SCAN
EN2
VCC
VCC
F
RX
TX
RX
STB[0] CLKOB STB[1]
RCLK
ENA
RCLK
END
TX
DC[0]
RX
DC[7]
Rx
DC[6]
G
RX
DB[1]
SPD
SELA
LP
ENC
SPD
SELB
TX
DC[1]
TX
DC[4]
WREN
TX
DC[7]
H
GND
GND
GND
GND
GND
GND
GND
GND
J
RX
DB[2]
RX
DB[5]
RX
DB[0]
RX
STB[2]
TX
DC[3]
TX
DC[2]
TX
DC[5]
TX
CTC[1]
K
LFIB
RX
DB[7]
RX
DB[4]
RX
DB[3]
TX
CLKC
L
TX
DB[6]
RX
RX
CLKB– CLKB+
RX
DB[6]
TX
DC[6]
LFIC
REF
CLKC+
RX
DC[3]
M
TX
CLKB
TX
ERRB
REF
REF
CLKB– CLKB+
TX
ERRC
RCLK
ENC
RX
DC[5]
RX
DC[4]
N
GND
GND
GND
GND
GND
GND
GND
GND
P
TX
DB[2]
TX
DB[3]
TX
DB[4]
TX
DB[5]
RX
RX
STC[1] STC[0]
RX
DC[0]
RX
DC[1]
R
TX
DB[7]
TX
CTB[1]
TX
DB[0]
TX
DB[1]
RX
RX
TX
RX
CLKC– CLKC+ CLKOC STC[2]
T
VCC
VCC
VCC
VCC
U
RX
STA[1]
RX
TX
STA[2] CTB[0]
RX
DA[2]
VCC
TX
CTA[0]
TX
DA[4]
GND
V
RX
STA[0]
RX
DA[0]
RX
DA[3]
RX
DA[7]
VCC
TX
DA[7]
TX
DA[3]
GND
W
RX
DA[1]
RX
DA[4]
REF
CLKA+
LFIA
VCC
TX
DA[6]
TX
DA[2]
GND
TX
ERRA
RX
CLKA+
Y
RX
DA[5]
RX
DA[6]
REF
CLKA–
TX
ERRD
VCC
TX
DA[5]
TX
DA[0]
GND
RX
CLKA–
TX
CLKA
TX
DA[1]
REF
ADDR
CLKD–
[0]
TX
REF
CTC[0] CLKC–
RX
DC[2]
VCC
VCC
VCC
VCC
TXC
TA[1]
GND
RX
DD[1]
RX
DD[2]
VCC
TX
CTD[1]
TX
DD[2]
TX
DD[1]
TX
DD[0]
RX
STD[2]
GND
RX
STD[0]
RX
DD[3]
VCC
RX
DD[6]
TX
CTD[0]
TX
DD[4]
TX
DD[3]
ADDR
[1]
ADDR
[3]
GND
RX
STD[1]
RX
DD[4]
VCC
RX
CLKD–
LFID
TX
DD[7]
TX
DD[5]
NC[1]
TX
CLKOD
GND
RX
DD[0]
RX
DD[5]
VCC
RX
CLKD+
RX
DD[7]
TX
CLKD
TX
DD[6]
TX
REF
ADDR
CLKOA CLKD+
[2]
Note:
1. NC=Do Not Connect
Document #: 38-02097 Rev. **
Page 7 of 43
PRELIMINARY
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]
LVTTL Input,
synchronous,
sampled by the
associated
TXCLKx↑ or
REFCLKx↑[2]
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
TXCLKSELx latch via the device configuration interface, and passed to the
encoder or Transmit Shifter. When the Encoder is enabled, TXDx[7:0] specifies
the specific data or command character sent.
TXCTA[1:0]
TXCTB[1:0]
TXCTC[1:0]
TXCTD[1:0]
LVTTL Input,
synchronous,
sampled by the
associated
TXCLKx↑ or
REFCLKx↑ [2]
Transmit Control. TXCTx[1:0] inputs are captured on the rising edge of the
transmit interface clock. The transmit interface clock is selected by the
TXCLKSELx latch via the device configuration interface, and passed to the
Encoder or Transmit Shifter. The TXCTA[1:0] inputs identify how the associated
TXDx[7:0] characters are interpreted. When the Encoder is bypassed, these
inputs are interpreted as data bits. When the Encoder is enabled, these inputs
determine if the TXDx[7:0] character is encoded as Data, a Special Character
code, or replaced with other Special Character codes. See Table 3 for details.
TXERRA
TXERRB
TXERRC
TXERRD
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±
Transmit Path Error. TXERRx is asserted HIGH to indicate detection of a
transmit Phase-Align Buffer underflow or overflow. If an underflow or overflow
condition is detected, TXERRx, for the channel in error, is asserted HIGH and
remains asserted until either a Word Sync Sequence is transmitted on that
channel, or the transmit Phase-Align Buffer is re-centered with the PABRSTx latch
via the device configuration interface. When TXBISTx = 0, the BIST progress is
presented on the associated TXERRx output. The TXERRx signal pulses HIGH
for one transmit-character clock period to indicate a pass through the BIST
sequence once every 511 or 527 (depending on RXCKSELx) character times. If
RXCKSELx = 1, a one character pulse occurs every 527 character times. If
RXCKSELx = 0, a one character pulse occurs every 511 character times.
TXERRx is also asserted HIGH, when any of the following conditions is true:
• The TXPLL for the associated channel is powered down. This occurs when
OE2x and OE1x for a given channel are simultaneous disabled by setting OE2x
= 0 and OE1x = 0.
• The absence of the REFCLKx± signal.
Transmit Path Clock Signals
REFCLKA±
REFCLKB±
REFCLKC±
REFCLKD±
Differential LVPECL
or single-ended
LVTTL input clock
Reference Clock. REFCLKx± clock inputs are used as the timing references for
the transmit and receive PLLs. These input clocks may also be selected to clock
the transmit and receive parallel interfaces. When driven by a single-ended
LVCMOS or LVTTL clock source, connect the clock source to either the true or
complement REFCLKx input, and leave the alternate REFCLKx input open
(floating). When driven by an LVPECL clock source, the clock must be a differential clock, using both inputs.
TXCLKA
TXCLKB
TXCLKC
TXCLKD
LVTTL Clock Input,
internal pull-down
Transmit Path Input Clock. When configuration latch TXCKSELx = 0, the
associated TXCLKx input is selected as the character-rate input clock for the
TXDx[7:0] and TXCTx[1:0] inputs. In this mode, the TXCLKx input must be
frequency-coherent to its associated TXCLKOx output clock, but may be offset in
phase by any amount. Once initialized, TXCLKx is allowed to drift in phase as
much as ±180 degrees. If the input phase of TXCLKx drifts beyond the handling
capacity of the Phase Align Buffer, TXERRx is asserted to indicate the loss of data,
and remains asserted until the Phase Align Buffer is initialized. The phase of the
TXCLKx input clock relative to its associated REFCLKx± is initialized when the
configuration latch PABRSTx is written as 0. When the associated TXERRx is
deasserted, the Phase Align Buffer is initialized and input characters are correctly
captured.
Notes:
2. When REFCLKx± is configured for half-rate operation, these inputs are sampled relative to both the rising and falling edges of the associated REFCLKx±.
3. When REFCLKx± is configured for half-rate operation, these outputs are presented relative to both the rising and falling edges of the associated REFCLKx±.
Document #: 38-02097 Rev. **
Page 8 of 43
PRELIMINARY
CYV15G0404DXB
Pin Definitions (continued)
CYV15G0404DXB Quad HOTLink II Transceiver
Name
TXCLKOA
TXCLKOB
TXCLKOC
TXCLKOD
I/O Characteristics
LVTTL Output
Signal Description
Transmit Clock Output. TXCLKOx output clock is synthesized by each channel’s
transmit PLL and operates synchronous to the internal transmit character clock.
TXCLKOx operates at either the same frequency as REFCLKx± (TXRATE = 0),
or at twice the frequency of REFCLKx± (TXRATE = 1). The transmit clock outputs
have no fixed phase relationship to REFCLKx±.
Receive Path Data and Status Signals
RXDA[7:0]
RXDB[7:0]
RXDC[7:0]
RXDD[7:0]
LVTTL Output,
synchronous to the
selected RXCLK±
output or REFCLKx±
input
Parallel Data Output. RXDx[7:0] parallel data outputs change relative to the
receive interface clock. The receive interface clock is selected by the RXCKSELx
latch. If RXCLKx± is a full-rate clock, the RXCLKx± clock outputs are complementary clocks operating at the character rate. The RXDx[7:0] outputs for the
associated receive channels follow rising edge of RXCLKx+ or falling edge of
RXCLKx–. If RXCLKx± is a half-rate clock, the RXCLKx± clock outputs are
complementary clocks operating at half the character rate. The RXDx[7:0] outputs
for the associated receive channels follow both the falling and rising edges of the
associated RXCLKx± clock outputs.
RXSTA[2:0]
RXSTB[2:0]
RXSTC[2:0]
RXSTD[2:0]
LVTTL Output,
synchronous to the
selected RXCLK±
output or REFCLKx±
input
Parallel Status Output. RXSTA[2:0] status outputs change relative to the receive
interface clock. The receive interface clock is selected by the RXCKSELx latch. If
RXCLKx± is a full-rate clock, the RXCLKx± clock outputs are complementary
clocks operating at the character rate. The RXSTAx[2:0] outputs for the
associated receive channels follow rising edge of RXCLKx+ or falling edge of
RXCLKx–. If RXCLKx± is a half-rate clock, the RXCLKx± clock outputs are
complementary clocks operating at half the character rate. The RXSTAx[2:0]
outputs for the associated receive channels follow both the falling and rising edges
of the associated RXCLKx± clock outputs. When the decoder is bypassed,
RXSTx[1:0] become the two low-order bits of the 10-bit received character.
RXSTx[2] = HIGH indicates the presence of a Comma character in the Output
Register. When the decoder is enabled, RXSTx[2:0] provide status of the received
signal. See Table 11 for a list of received character status.
Receive Path Clock Signals
RXCLKA±
RXCLKB±
RXCLKC±
RXCLKD±
LVTTL Output Clock
Receive Clock Output. RXCLKx± is the receive interface clock used to control
timing of the RXDx[7:0] and RXSTA[2:0] parallel outputs. The source of the
RXCLKx± outputs is selected by the RXCKSELx latch via the device configuration
interface. These true and complement clocks are used to control timing of data
output transfers. These clocks are output continuously at either the dual-character
rate (1/20th the serial bit-rate) or character rate (1/10th the serial bit-rate) of the
data being received, as selected by RXRATEx. When configured such that the
output data path is clocked by the REFCLKx± instead of a recovered clock, the
RXCLKx± output drivers present a buffered or divided form (depending on
RXRATEx) of the associated REFCLKx± that are delayed in phase to align with
the data. This phase difference allows the user to select the optimal clock
(REFCLKx± or RXCLK±) for setup/hold timing for their specific system.
When REFCLKx± is a full-rate clock, the RXCLKx± rate depends on the value of
RXRATEx.
When REFCLKx± is a half-rate clock and RXCKSELx = 0, the RXCLKx± rate
depends on the value of RXRATEx.
When REFCLKx± is a half-rate clock and RXCKSELx=1, the RXCLKx± rate does
not depend on the value of RXRATEx and operates at the same rate as
REFCLKx±.
Device Control Signals
RESET
LVTTL Input,
asynchronous,
internal pull-up
Document #: 38-02097 Rev. **
Asynchronous Device Reset. RESET initializes all state machines, counters,
and configuration latches in the device to a known state. RESET must be asserted
LOW for a minimum pulse width. When the reset is removed, all state machines,
counters and configuration latches are at an initial state. See Table 9 for the
initialize values of the device configuration latches.
Page 9 of 43
PRELIMINARY
CYV15G0404DXB
Pin Definitions (continued)
CYV15G0404DXB Quad HOTLink II Transceiver
Name
I/O Characteristics
Signal Description
LDTDEN
LVTTL Input,
internal pull-up
Level Detect Transition Density Enable. When LDTDEN is HIGH, the Signal
Level Detector, Range Controller, and Transition Density Detector are all enabled
to determine if the RXPLL tracks REFCLKx± or the selected input serial data
stream. If the Signal Level Detector, Range Controller, or Transition Density
Detector are out of their respective limits while LDTDEN is HIGH, the RXPLL locks
to REFCLK± until such a time they become valid. The (SDASEL[A..D][1:0]) are
used to configure the trip level of the Signal Level Detector. The Transition Density
Detector limit is one transition in every 60 consecutive bits. When LDTDEN is
LOW, only the Range Controller is used to determine if the RXPLL tracks
REFCLKx± or the selected input serial data stream. For the cases when
RXCKSELx = 0 (recovered clock), it is recommended to set LDTDEN = HIGH.
RCLKENA
RCLKENB
RCLKENC
RCLKEND
LVTTL Input,
internal pull-down
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, as well as presenting 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.
ULCA
ULCB
ULCC
ULCD
LVTTL Input,
internal pull-up
SPDSELA
SPDSELB
SPDSELC
SPDSELD
3-Level Select[4]
static control input
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.
INSELA
INSELB
INSELC
INSELD
LVTTL Input,
asynchronous
Receive Input Selector. The INSELx input determines which external serial bit
stream is passed to the receiver’s Clock and Data Recovery circuit. When INSELx
is HIGH, the Primary Differential Serial Data Input, INx1±, is selected for the
associated receive channel. When INSELx is LOW, the Secondary Differential
Serial Data Input, INx2±, is selected for the associated receive channel.
LPENA
LPENB
LPENC
LPEND
LVTTL Input,
asynchronous,
internal pull-down
Loop-Back-Enable. The LPENx input enables the internal serial loop-back for the
associated channel. When LPENx is HIGH, the transmit serial data from the
associated channel is internally routed to the associated receive Clock and Data
Recovery (CDR) circuit. All enabled serial drivers on the channel are forced to
differential logic-1, and the serial data inputs are ignored. When LPENx is LOW,
the internal serial loop-back function is disabled.
Notes:
4. 3-Level Select inputs are used for static configuration. These are ternary inputs that make use of logic levels of LOW, MID, and HIGH. The LOW level is usually
implemented by direct connection to VSS (ground). The HIGH level is usually implemented by direct connection to VCC (power). The MID level is usually
implemented by not connecting the input (left floating), which allows it to self bias to the proper level.
5. See Device Configuration and Control Interface for detailed information on the operation of the Configuration Interface.
Document #: 38-02097 Rev. **
Page 10 of 43
PRELIMINARY
CYV15G0404DXB
Pin Definitions (continued)
CYV15G0404DXB Quad HOTLink II Transceiver
Name
LFIA
LFIB
LFIC
LFID
I/O Characteristics
LVTTL Output,
asynchronous
Signal Description
Link Fault Indication Output. LFIx is an output status indicator signal. LFIx is the
logical OR of five internal conditions. LFIx is asserted LOW when any of the
following conditions is true:
• Received serial data rate outside expected range
• Analog amplitude below expected levels
• Transition density lower than expected
• Receive channel disabled
• ULCx is LOW
• Absence of REFCLKx±.
Device Configuration and Control Bus Signals
WREN
LVTTL input,
asynchronous,
internal pull-up
Control Write Enable. The WREN input writes the values of the DATA[7:0] bus
into the latch specified by the address location on the ADDR[3:0] bus.[5]
ADDR[3:0]
LVTTL input
asynchronous,
internal pull-up
Control Addressing Bus. The ADDR[3:0] bus is the input address bus used to
configure the device. The WREN input writes the values of the DATA[7:0] bus into
the latch specified by the address location on the ADDR[3:0] bus.[5] Table 9 lists
the configuration latches within the device, and the initialization value of the
latches upon the assertion of RESET. Table 10 shows how the latches are
mapped in the device.
DATA[7:0]
LVTTL input
asynchronous,
internal pull-up
Control Data Bus. The DATA[7:0] bus is the input data bus used to configure the
device. The WREN input writes the values of the DATA[7:0] bus into the latch
specified by address location on the ADDR[3:0] bus.[5 ] Table 9 lists the configuration latches within the device, and the initialization value of the latches upon the
assertion of RESET. Table 10 shows how the latches are mapped in the device.
Internal Device Configuration Latches
RFMODE[A..D][1:0] Internal Latch[6]
Latch[6]
FRAMCHAR[A..D]
Internal
DECMODE[A..D]
Internal Latch[6]
Latch[6]
DECBYP[A..D]
Internal
RXCKSEL[A..D]
Internal Latch[6]
Latch[6]
RXRATE[A..D]
Internal
SDASEL[A..D][1:0]
Internal Latch[6]
Latch[6]
ENCBYP[A..D]
Internal
TXCKSEL[A..D]
Internal Latch[6]
Latch[6]
TXRATE[A..D]
Internal
RFEN[A..D]
Internal Latch[6]
Latch[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 Bypassed.
Transmit Clock Select.
Transmit PLL Clock Rate Select.
Reframe Enable.
RXPLLPD[A..D]
Internal
RXBIST[A..D]
Internal Latch[6]
TXBIST[A..D]
Internal
Latch[6]
OE2[A..D]
Internal Latch[6]
Differential Serial Output Driver 2 Enable.
OE1[A..D]
Internal
Latch[6]
Differential Serial Output Driver 1 Enable.
PABRST[A..D]
Internal Latch[6]
Latch[6]
GLEN[11..0]
Internal
FGLEN[2..0]
Internal Latch[6]
Receive Channel Power Control.
Receive Bist Disabled.
Transmit Bist Disabled.
Transmit Clock Phase Alignment Buffer Reset.
Global Latch Enable.
Force Global Latch Enable.
Factory Test Modes
Note:
6. See Device Configuration and Control Interface for detailed information on the internal latches.
Document #: 38-02097 Rev. **
Page 11 of 43
PRELIMINARY
CYV15G0404DXB
Pin Definitions (continued)
CYV15G0404DXB Quad HOTLink II Transceiver
Name
I/O Characteristics
Signal Description
SCANEN2
LVTTL input,
internal pull-down
Factory Test 2. SCANEN2 input is for factory testing only. This input may be left
as a NO CONNECT, or GND only.
TMEN3
LVTTL input,
internal pull-down
Factory Test 3. TMEN3 input is for factory testing only. This input may be left as
a NO CONNECT, or GND only.
OUTA1±
OUTB1±
OUTC1±
OUTD1±
CML Differential
Output
Primary Differential Serial Data Output. The OUTx1± PECL-compatible CML
outputs (+3.3V referenced) are capable of driving terminated transmission lines
or standard fiber-optic transmitter modules, and must be AC-coupled for PECLcompatible connections.
OUTA2±
OUTB2±
OUTC2±
OUTD2±
CML Differential
Output
Secondary Differential Serial Data Output. The OUTx2± PECL-compatible CML
outputs (+3.3V referenced) are capable of driving terminated transmission lines or
standard fiber-optic transmitter modules, and must be AC-coupled for PECL-compatible
connections.
INA1±
INB1±
INC1±
IND1±
Differential Input
Primary Differential Serial Data Input. The INx1± input accepts the serial data
stream for deserialization and decoding. The INx1± serial stream is passed to the
receive CDR circuit to extract the data content when INSELx = HIGH.
INA2±
INB2±
INC2±
IND2±
Differential Input
Secondary Differential Serial Data Input. The INx2± input accepts the serial
data stream for deserialization and decoding. The INx2± serial stream is passed
to the receiver CDR circuit to extract the data content when INSELx = LOW.
TMS
LVTTL Input,
internal pull-up
Test Mode Select. Used to control access to the JTAG Test Modes. If maintained
high for ≥5 TCLK cycles, the JTAG test controller is reset.
TCLK
LVTTL Input,
internal pull-down
JTAG Test Clock.
TDO
3-State LVTTL Output Test Data Out. JTAG data output buffer. High-Z while JTAG test mode is not
selected.
TDI
LVTTL Input,
internal pull-up
Test Data In. JTAG data input port.
TRST
LVTTL Input,
internal pull-up
JTAG reset signal. When asserted (LOW), this input asynchronously resets the
JTAG test access port controller.
Analog I/O
JTAG Interface
Power
VCC
+3.3V Power.
GND
Signal and Power Ground for all internal circuits.
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 captures
eight data bits and two control bits on each input clock cycle.
When the Encoder is bypassed, the control bits are part of the
pre-encoded 10-bit character.
Document #: 38-02097 Rev. **
When the Encoder is enabled, the TXCTx[1:0] bits are interpreted along with the associated TXDx[7:0] character to
generate a specific 10-bit transmission character.
Phase-Align Buffer
Data from each Input Register is passed to the associated
Phase-Align Buffer, when the TXDx[7:0] and TXCTx[1:0] input
registers are clocked using TXCLKx¦ (TXCKSELx = 0 and
TXRATEx = 0). When the TXDx[7:0] and TXCTx[1:0] input
registers are clocked using REFCLKx± (TXCKSELx = 1) and
REFCLKx± is a full-rate clock, the associated Phase
Alignment Buffer in the transmit path is bypassed. These
buffers are used to absorb clock phase differences between
the TXCLKx input clock and the internal character clock for
that channel.
Page 12 of 43
PRELIMINARY
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]
If the phase offset, between the initialized location of the input
clock and REFCLKx, exceeds the skew handling capabilities
of the Phase-Align Buffer, an error is reported on that
channel’s TXERRx output. This output indicates an error
continuously until the Phase-Align Buffer for that channel is
reset. While the error remains active, the transmitter for that
channel outputs a continuous C0.7 character to indicate to the
remote receiver that an error condition is present in the link.
Each Phase-Align Buffer may be individually reset with
minimal disruption of the serial data stream. When a PhaseAlign Buffer error is present, the transmission of a Word Sync
Sequence re-centers the Phase-Align Buffer and clears the
error indication.
Note. K28.5 characters may be added or removed from the
data stream during the Phase Align Buffer reset operation.
When used with non-Cypress devices that require a complete
16-character Word Sync Sequence for proper receive
Elasticity Buffer Operation, it is recommend that the Phase
Alignment Buffer reset be followed by a Word Sync Sequence
to ensure proper operation.
Encoder
Each character received from the Input Register or PhaseAlign Buffer is passed to the Encoder logic. This block interprets each character and any associated control bits, and
outputs a 10-bit transmission character.
Depending on the operational mode, the generated transmission character may be
• the 10-bit pre-encoded character accepted in the Input
Register.
• the 10-bit equivalent of the 8-bit Data character accepted in
the Input Register.
CYV15G0404DXB
• 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), the characters
transmitted are converted from Data or Special Character
codes to 10-bit transmission characters, using an integrated
8B/10B encoder. When directed to encode the character as a
Special Character code, the encoder uses the Special
Character encoding rules listed in Table 16. When directed to
encode the character as a Data character, it is encoded using
the Data Character encoding rules in Table 15.
The 8B/10B encoder is standards compliant with ANSI/NCITS
ASC X3.230-1994 Fibre Channel, IEEE 802.3z Gigabit
Ethernet, the IBM® ESCON® and FICON™ channels, ETSI
DVB-ASI, and ATM Forum standards for data transport.
Many of the Special Character codes listed in Table 16 may be
generated by more than one input character. The
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.
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.
Note:
7. LSB shifted out first.
Document #: 38-02097 Rev. **
Page 13 of 43
CYV15G0404DXB
PRELIMINARY
Table 2. Encoder Bypass Mode
Signal Name
Bus Weight
10B Name
TXDx[0] (LSB)
20
a[7]
TXDx[1]
2
1
b
TXDx[2]
22
c
TXDx[3]
23
d
TXDx[4]
24
e
TXDx[5]
25
i
TXDx[6]
26
f
TXDx[7]
27
g
TXCTx[0]
28
h
TXCTx[1] (MSB)
29
j
When the encoder is enabled, the TXCTx[1:0] data control bits
control the interpretation of the TXDx[7:0] bits and the
characters generated by them. These bits are interpreted as
listed in Table 3.
Table 3. Transmit Modes
TXCTx[1]
TXCTx[0]
0
0
Encoded data character
0
1
K28.5 fill character
1
0
Special character code
1
1
16-character Word Sync Sequence
Characters Generated
Word Sync Sequence
When TXCTx[1:0] = 11, a 16-character sequence of K28.5
characters, known as a Word Sync Sequence, is generated on
the associated channel. This sequence of K28.5 characters
may start with either a positive or negative disparity K28.5 (as
determined by the current running disparity and the 8B/10B
coding rules). The disparity of the second and third K28.5
characters in this sequence are reversed from what normal
8B/10B coding rules would generate. The remaining K28.5
characters in the sequence follow all 8B/10B coding rules. The
disparity of the generated K28.5 characters in this sequence
follow a pattern of either ++––+–+–+–+–+–+– or
––++–+–+–+–+–+–+.
The generation of this sequence, once started, cannot be
stopped until all 16 characters have been sent. The content of
the associated input registers are ignored for the duration of
this sequence. At the end of this sequence, if the TXCTx[1:0]
= 11 condition is sampled again, the sequence restarts and
remains uninterruptible for the following 15 character clocks.
Transmit BIST
Each transmit channel contains an internal pattern generator
that can be used to validate both the link and device operation.
These generators are enabled by the associated TXBISTx
latch via the device configuration interface. When enabled, a
register in the associated transmit channel becomes a
signature pattern generator by logically converting to a Linear
Feedback Shift Register (LFSR). This LFSR generates a 511character (or 526-character) sequence that includes all Data
and Special Character codes, including the explicit violation
Document #: 38-02097 Rev. **
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 16character 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 can 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[4] inputs that select one of three
operating ranges for the serial data outputs and inputs of the
associated channel. The operating serial signaling-rate and
allowable range of REFCLKx± frequencies are listed in
Table 4.
Table 4. Operating Speed Settings
REFCLKx±
Frequency
(MHz)
Signaling
Rate
(MBaud)
195 – 400
SPDSELx
TXRATE
LOW
1
reserved
0
19.5 – 40
1
20 – 40
0
40 – 80
MID (Open)
HIGH
1
40 – 75
0
80 – 150
400 – 800
800 – 1500
The REFCLKx± inputs are differential inputs with each input
internally biased to 1.4V. If the REFCLKx+ input is connected
to a TTL, LVTTL, or LVCMOS clock source, the input signal is
recognized when it passes through the internally biased
reference point. When driven by a single-ended TTL, LVTTL,
or LVCMOS clock source, connect the clock source to either
the true or complement REFCLKx input, and leave the
alternate REFCLKx input open (floating).
When both the REFCLKx+ and REFCLKx– inputs are
connected, the clock source must be a differential clock. This
can either be a differential LVPECL clock that is DC-or
AC-coupled or a differential LVTTL or LVCMOS clock.
By connecting the REFCLKx– input to an external voltage
source, it is possible to adjust the reference point of the
REFCLKx+ input for alternate logic levels. When doing so it is
necessary to ensure that the input differential crossing point
remains within the parametric range supported by the input.
Page 14 of 43
CYV15G0404DXB
PRELIMINARY
Serial Output Drivers
The serial output interface drivers use differential Current
Mode Logic (CML) drivers to provide source-matched drivers
for transmission lines. These drivers accept data from the
Transmit Shifters. These drivers have signal swings equivalent
to that of standard PECL drivers, and are capable of driving
AC-coupled optical modules or transmission lines. When
configured for local loopback (LPENx = HIGH), all enabled
serial drivers are configured to drive a static differential logic 1.
Transmit Channels Enabled
Each driver can be enabled or disabled separately via the
device configuration interface.
When a driver is disabled via the configuration interface, it is
internally powered down to reduce device power. If both serial
drivers for a channel are in this disabled state, the associated
internal logic for that channel is also powered down. A device
reset (RESET sampled LOW) disables all output drivers.
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 can accommodate wire interconnect and
filtering losses or transmission line attenuation greater than
16 dB. For normal operation, these inputs should receive a
signal of at least VIDIFF > 100 mV, or 200 mV peak-to-peak
differential. Each Line Receiver can be DC- or AC-coupled to
+3.3V powered fiber-optic interface modules (any ECL/PECL
family, not limited to 100K PECL) or AC-coupled to +5V
powered optical modules. The common-mode tolerance of
these line receivers accommodates a wide range of signal
termination voltages. Each receiver provides internal DCrestoration, to the center of the receiver’s common mode
range, for AC-coupled signals.
The local internal loopback (LPENx) allows the serial transmit
data outputs to be routed internally back to the Clock and Data
Recovery circuit associated with each channel. When
configured for local loopback, the associated transmit serial
driver outputs are forced to output a differential logic-1. This
prevents local diagnostic patterns from being broadcast to
attached remote receivers.
Signal Detect/Link Fault
Each selected Line Receiver (i.e., that routed to the clock and
data recovery PLL) is simultaneously monitored for
• analog amplitude above amplitude level selected by
SDASELx
• transition density above the specified limit
• range controls report the received data stream inside
normal frequency range (±1500 ppm)
• receive channel enabled
• Presence of reference clock
• ULCx is not asserted.
All of these conditions must be valid for the Signal Detect block
to indicate a valid signal is present. This status is presented on
the LFIx (Link Fault Indicator) output associated with each
receive channel, which changes synchronous to the selected
receive interface clock.
Analog Amplitude
While most signal monitors are based on fixed constants, the
analog amplitude level detection is adjustable to allow
operation with highly attenuated signals, or in high-noise
environments. The analog amplitude level detection is set by
the SDASELx latch via device configuration interface. The
SDASELx latch sets the trip point for the detection of a valid
signal at one of three levels, as listed in Table 5. This control
input affects the analog monitors for all receive channels.
Table 5. Analog Amplitude Detect Valid Signal Levels[9]
SDASEL Typical Signal with Peak Amplitudes Above
00
Analog Signal Detector is disabled
01
140 mV p-p differential
10
280 mV p-p differential
11
420 mV p-p differential
The Analog Signal Detect monitors are active for the Line
Receiver as selected by the associated INSELx input. When
configured for local loopback, no input receivers are selected,
and the LFIx output for each channel reports only the receive
VCO frequency out-of-range and transition density status of
the associated transmit signal. When local loopback is active,
the associated Analog Signal Detect Monitor is disabled.
Transition Density
The Transition Detection logic checks for the absence of
transitions spanning greater than six transmission characters
(60 bits). If no transitions are present in the data received, the
Detection logic for that channel asserts LFIx.
Range Controls
The CDR circuit includes logic to monitor the frequency of the
PLL Voltage Controlled Oscillator (VCO) used to sample the
incoming data stream. This logic ensures that the VCO
operates at, or near the rate of the incoming data stream for
two primary cases:
• when the incoming data stream resumes after a time in
which it has been “missing.”
• when the incoming data stream is outside the acceptable
signaling rate range.
To perform this function, the frequency of the RXPLL VCO is
periodically compared to the frequency of the REFCLKx±
input.
If the VCO is running at a frequency beyond ±1500ppm 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.
Note:
8. When a disabled transmit channel (i.e., both outputs disabled) is re-enabled, the data 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 typical waveforms that have generally 3 – 4 transitions for every ten bits. In a worse case environment the signals
may have a sign-wave appearance (highest transition density with repeating 0101...). Signal peak amplitudes levels within this environment type could increase
the values in the table above by approximately 100 mV.
Document #: 38-02097 Rev. **
Page 15 of 43
CYV15G0404DXB
PRELIMINARY
The sampling and relock period of the Range Control is calculated as follows: RANGE_CONTROL_ SAMPLING_PERIOD
= (RECOVERED BYTE CLOCK PERIOD) * (4096).
During the time that the Range Control forces the RXPLL VCO
to track REFCLKx±, the LFIx output is asserted LOW. After a
valid serial data stream is applied, it may take up to one
RANGE CONTROL SAMPLING PERIOD before the PLL
locks to the input data stream, after which LFIx should be
HIGH.
Receive Channel Enabled
The 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
output. When RXPLLPDx = 1, the associated PLL and receive
channel is enabled to receive and decode a serial stream.
Note. When a disabled receive channel is reenabled, the
status of the associated LFIx output and data on the parallel
outputs for the associated channel may be indeterminate for
up to 2 ms.
Clock/Data Recovery
The extraction of a bit-rate clock and recovery of bits from each
received serial stream is performed by a separate CDR block
within each receive channel. The clock extraction function is
performed by an integrated PLL that tracks the frequency of
the transitions in the incoming bit stream and align the phase
of the internal bit-rate clock to the transitions in the selected
serial data stream.
Each CDR accepts a character-rate (bit-rate ÷ 10) or halfcharacter-rate (bit-rate ÷ 20) reference clock from the
associated REFCLKx± input. This REFCLKx± input is used to
• ensure that the VCO (within the CDR) is operating at the
correct frequency (rather than a harmonic of the bit-rate)
• reduce PLL acquisition time
• limit unlocked frequency excursions of the CDR VCO when
there is no input data present at the selected Serial Line
Receiver.
Regardless of the type of signal present, the CDR attempts to
recover a data stream from it. If the signalling rate of the
recovered data stream is outside the limits set by the range
control monitors, the CDR tracks REFCLKx± instead of the
data stream. Once the CDR output (RXCLK±) frequency
returns back close to REFCLKx± frequency, the CDR input is
switched back to the input data stream. If no data is present at
the selected line receiver, this switching behavior may result
in brief RXCLK± frequency excursions from REFCLKx±.
However, the validity of the input data stream is indicated by
the LFIx output. The frequency of REFCLKx± is required to be
within ±1500 ppm of the frequency of the clock that drives the
REFCLKx± input of the remote transmitter to ensure a lock to
the incoming data stream.
For systems using multiple or redundant connections, the LFIx
output can be used to select an alternate data stream. When
an LFIx indication is detected, external logic can toggle
selection of the associated INx1± and INx2± input through the
associated INSELx input. When a port switch takes place, it is
necessary for the receive PLL for that channel to reacquire the
new serial stream and frame to the incoming character boundaries.
Reclocker
The CYV15G0404DXB contains a reclocker mode on each
receive channel which 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 bitclock rate. When enabled, the Framer examines the data
stream looking for one or more COMMA or K28.5 characters
at all possible bit positions. The location of this character in the
data stream is used to determine the character boundaries of
all following characters.
Framing Character
The 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
through the FRAMCHARx latches via the configuration
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
Bits detected in framer
FRAMCHARx
Character Name
Bits Detected
0
COMMA+
COMMA–
00111110XX[10]
or 11000001XX
1
-K28.5
+K28.5
0011111010 or
1100000101
Framer
The framer on each channel operates in one of three different
modes. Each framer may be enabled or disabled using the
RFENx latches via the configuration interface. When the
framer is disabled (RFENx = 0), no combination of received
bits alters the frame information.
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 pattern is extended to a full eight bits to reduce the possibility of a framing error.
Document #: 38-02097 Rev. **
Page 16 of 43
PRELIMINARY
When the Low-Latency framer is selected (RFMODEx[1:0] =
00), the framer operates by stretching the recovered character
clock until it aligns with the received character boundaries. In
this mode the framer starts its alignment process on the first
detection of the selected framing character. To reduce the
impact on external circuits that use the recovered clock, the
clock period is not stretched by more than two bit-periods in
any one clock cycle. When operated with a character-rate
output clock, the output of properly framed characters may be
delayed by up to nine character-clock cycles from the
detection of the selected framing character. When operated
with a half-character-rate output clock, the output of properly
framed characters may be delayed by up to 14 character-clock
cycles from the detection of the framing character.
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 15 and Table 16. Received Special Code characters
are decoded using Table 16. Valid data characters are
indicated by a 000b bit-combination on the associated
RXSTx[2:0] status bits, and Special Character codes are
indicated by a 001b bit-combination of these status outputs.
Framing characters, Invalid patterns, disparity errors, and
synchronization status are presented as alternate combinations of these status bits.
When DECBYPx = 0 and DECMODEx = 1, the 10B/8B
decoder is bypassed via the configuration interface. When
bypassed, raw 10-bit characters are passed through the
CYV15G0404DXB
receiver and presented at the RXDx[7:0] and the RXSTA[1:0]
outputs as 10-bit wide characters.
When the decoder is enabled by setting DECBYPx = 1 via the
configuration interface, the 10-bit transmission characters are
decoded using Table 15 and Table 16. Received Special
characters are decoded using Table 16. The columns used in
Table 16 are determined by the DECMODEx latch via the
device configuration interface. When DECMODEx = 0 the
ALTERNATE table is used and when DECMODEx = 1 the
CYPRESS table is used.
Receive BIST Operation
The receiver channel contains an internal pattern checker that
can be used to validate both device and link operation. These
pattern checkers are enabled by the associated RXBISTx
latch via the device configuration interface. When enabled, a
register in the associated receive channel becomes a
signature pattern generator and checker by logically
converting to a Linear Feedback Shift Register (LFSR). This
LFSR generates a 511-character or 526-character sequence
that includes all Data and Special Character codes, including
the explicit violation symbols. This provides a predictable yet
pseudo-random sequence that can be matched to an identical
LFSR in the attached Transmitter(s). When synchronized with
the received data stream, the associated Receiver checks
each character in the Decoder with each character generated
by the LFSR and indicates compare errors and BIST status at
the RXSTx[2:0] bits of the Output Register.
When BIST is first recognized as being enabled in the
Receiver, the LFSR is preset to the BIST-loop start-code of
D0.0. This code D0.0 is sent only once per BIST loop. The
status of the BIST progress and any character mismatches are
presented on the RXSTx[2:0] status outputs.
Code rule violations or running disparity errors that occur as
part of the BIST loop do not cause an error indication.
RXSTx[2:0] indicates 010b or 100b for one character period
per BIST loop to indicate loop completion. This status can be
used to check test pattern progress. These same status values
are presented when the decoder is bypassed and BIST is
enabled on a receive channel.
The specific status reported by the BIST state machine are
listed in Table 11. 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 SelfTest.” 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, the Low-Latency
Framer must not be enabled. The BIST sequence contains an
aliased K28.5 framing character, which causes the Receiver
to update its character boundaries incorrectly.
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.
Document #: 38-02097 Rev. **
Page 17 of 43
CYV15G0404DXB
PRELIMINARY
The receive BIST state machine requires the characters to be
correctly framed for it to detect the BIST sequence. If the Low
Latency Framer is enabled, the Framer misaligns to an aliased
SYNC character within the BIST sequence. If the Alternate
Multi-Byte Framer is enabled and the Receiver outputs are
clocked relative to a recovered clock, it is generally necessary
to frame the receiver before BIST is enabled. If the receive
outputs are clocked relative to REFCLKx±, the transmitter
precedes every 511 character BIST sequence with a 16
character-character Word Sync Sequence.
A device reset (RESET sampled LOW) presets the BIST
Enable Latches to disable BIST on all channels.
Receive Elasticity Buffer
Each receive channel contains an Elasticity Buffer that is
designed to support multiple clocking modes. These buffers
allow data to be read using a clock that is asynchronous in both
frequency and phase from the Elasticity Buffer write clock, or
to be read using a clock that is frequency coherent but with
uncontrolled phase relative to the Elasticity Buffer write clock.
If the chip is configured for operation with a recovered clock,
the Elasticity Buffer is bypassed.
Each Elasticity Buffer is 10 characters deep, and supports and
an 11 bit wide data path. It is capable of supporting a decoded
character and three status bits for each character present in
the buffer. The write clock for these buffers is always the
recovered clock for the associated read channel.
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. 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 can be
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 and DECMODEx configuration latch. This
mapping is shown in Table 7.
Table 7. Output Register Bit Assignments
Signal Name
BYPASS ACTIVE
(DECBYPx = 0 and
DECMODEx = 1)
DECODER
(DECBYP = 1)
RXSTx[2] (LSB)
COMDETx
RXSTx[2]
RXSTx[1]
DOUTx[0]
RXSTx[1]
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.
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]
The insertion of a K28.5 or deletion of a framing character can
occur at any time on any channel, however, the actual timing
of these insertions and deletions is controlled in part by how
the transmitter sends its data. Insertion of a K28.5 character
can only occur when the receiver has a framing character in
the Elasticity Buffer. Likewise, to delete a framing character,
one must also be in the Elasticity Buffer. To prevent a buffer
overflow or underflow on a receive channel, a minimum
density of framing characters must be present in the received
data streams.
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]
Receive Modes
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 via the
device configuration interface. When RXPLLPDx = 0, the
associated PLL and analog circuitry of the channel is disabled.
The transmit channels are controlled by the OE1x and the
OE2x latches via the device configuration interface. When a
driver is disabled via the configuration interface, it is internally
powered down to reduce device power. If both serial drivers
for a channel are in this disabled state, the associated internal
logic for that channel is also powered down.
Document #: 38-02097 Rev. **
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.
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
Page 18 of 43
PRELIMINARY
Table 8. Decoder Bypass Mode
Signal Name
Bus Weight
10 Bit Name
RXSTx[2] (LSB)
COMDETx
RXSTx[1]
20
a
RXSTx[0]
21
b
RXDx[0]
22
c
RXDx[1]
23
d
RXDx[2]
24
e
RXDx[3]
25
i
RXDx[4]
26
f
RXDx[5]
27
g
RXDx[6]
28
h
RXDx[7] (MSB)
29
j
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; e.g. a valid data character
received with incorrect running disparity is not reported as a
valid data character. It is instead reported as a decoder
violation of some specific type. This implies a hierarchy or
priority level to the various status bit combinations. The
hierarchy and value of each status are listed in Table 11.
A second status mapping, listed in Table 11, is used when the
receive channel is configured for BIST operation. This status
is used to report receive BIST status and progress.
BIST Status State Machine
When a receive path is enabled to look for and compare the
received data stream with the BIST pattern, the RXSTx[2:0]
bits identify the present state of the BIST compare operation.
The BIST state machine has multiple states, as shown in
Figure 2 and Table 11. When the receive PLL detects an outof-lock condition, the BIST state is forced to the Start-of-BIST
state, regardless of the present state of the BIST state
machine. If the number of detected errors ever exceeds the
number of valid matches by greater than 16, the state machine
is forced to the WAIT_FOR_BIST state where it monitors the
receive path for the first character of the next BIST sequence
(D0.0). Also, if the Elasticity Buffer ever hits an
overflow/underflow condition, the status is forced to the
BIST_START until the buffer is re-centered (approximately
nine character periods).
Document #: 38-02097 Rev. **
CYV15G0404DXB
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 via the configuration interface. The configuration interface allows the device
to be configured globally or allows each channel to be
configured independently. Table 9 lists the configuration
latches within the device including the initialization value of the
latches upon the assertion of RESET. Table 10 shows how the
latches are mapped in the device. Each row in the Table 10
maps to a 8-bit latch bank. There are 16 such write-only latch
banks. When WREN = 0, the logic value in the DATA[7:0] is
latched to the latch bank specified by the values in ADDR[3:0].
The second column of Table 10 specifies the channels
associated with the corresponding latch bank. For example,
the first three latch banks (0,1 and 2) consist of configuration
bits for channel A. The latch banks 12, 13 and 14 consist of
Global configuration bits and the last latch bank (15) is the
Mask latch bank that can be configured to perform bit-by-bit
configuration.
Global Enable Function
The global enable function, controlled by the GLENx bits, is a
feature that can be used to reduce the number of write operations needed to setup the latch banks. This function is
beneficial in systems that use a common configuration in
multiple channels. The GLENx bit is present in bit 0 of latch
banks 0 through 11 only. Its default value (1) enables the global
update of the latch bank's contents. Setting the GLENx bit to
0 disables this functionality.
Latch Banks 12, 13, and 14 are used to load values in the
related latch banks in a global manner. A write operation to
latch bank 12 could do a global write to latch banks 0, 3, 6, and
9 depending on the value of GLENx in these latch banks; latch
bank 13 could do a global write to latch banks 1, 4, 7 and 10;
and latch banks 14 could do a global write to latch banks 2, 5,
8 and 11. The GLENx bit cannot be modified by a global write
operation.
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 bitby-bit basis. A logic 1 in a bit location allows for the update of
that same location of the target latch bank(s), whereas a logic
0 disables it. The reset value of this latch bank is FFh, thereby
making its use optional by default. The mask latch bank is not
maskable. The FGLEN functionality is not affected by the bit 0
value of the mask latch bank.
Latch Types
There are two types of latch banks: static (S) and dynamic (D).
Each channel is configured by 2 static and 1 dynamic latch
banks. The S type contain those settings that normally do not
change for a given application, whereas the D type controls
the settings that could change dynamically during the appliPage 19 of 43
PRELIMINARY
cation'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.
CYV15G0404DXB
Latch Bank 14 is also useful for those users that do not need
the latch-based programmable feature of the device. This
latch bank could be used in those applications that do not need
to modify the default value of the static latch banks, and that
can afford a global (i.e., not independent) control of the
dynamic signals. In this case, this feature becomes available
when ADDR[3:0] is left unchanged with a value of “1110” and
WREN is left asserted. The signals present in DATA[7:0] effectively become global control pins, and for the latch banks 2, 5,
8 and 11.
Table 9. Device Configuration and Control Latch Descriptions
Name
Signal Description
RFMODEA[1:0]
RFMODEB[1:0]
RFMODEC[1:0]
RFMODED[1:0]
Reframe Mode Select. The initialization value of the RFMODEx [1:0] latches = 10. RFMODEx is used
to select the operating mode of the framer. When RFMODEx[1:0] = 00, the low-latency framer is
selected. This frames on each occurrence of the selected framing character(s) in the received data
stream. This mode of framing stretches the recovered clock for one or multiple cycles to align that clock
with the recovered data. When RFMODEx[1:0] = 01, the alternate mode Multi-Byte parallel framer is
selected. This requires detection of the selected framing character(s) in the received serial bit stream,
on identical 10-bit boundaries, on four directly adjacent characters. The recovered character clock
remains in the same phasing regardless of character offset. When RFMODEx[1:0] =10, the Cypressmode Multi-Byte parallel framer is selected. This requires a pair of the selected framing character(s), on
identical 10-bit boundaries, within a span of 50 bits, before the character boundaries are adjusted. The
recovered character clock remains in the same phasing regardless of character offset. RFMODEx[1:0]
= 11 is reserved for test.
FRAMCHARA
FRAMCHARB
FRAMCHARC
FRAMCHARD
Framing Character Select. The initialization value of the FRAMCHARx latch = 1. FRAMCHARx is used
to select the character or portion of a character used for framing of each channel’s received data stream.
When FRAMCHARx = 1, the framer looks for either disparity of the K28.5 character. When FRAMCHARx
= 0, the framer looks for either disparity of the 8-bit Comma characters. The specific bit combinations of
these framing characters are listed in Table 6.
DECMODEA
DECMODEB
DECMODEC
DECMODED
Receiver Decoder Mode Select. The initialization value of the DECMODEx latch = 1. DECMODEx
selects the Decoder Mode used for the associated channel. When DECMODEx = 1 and decoder is
enabled, the Cypress Decoding Mode is used. When DECMODEx = 0 and decoder is enabled, the
Alternate Decoding mode is used. When the decoder is enabled (DECBYPx = 1), the 10-bit transmission
characters are decoded using Table 15 and Table 16. The column used in the Special Characters
Table 16 is determined by the DECMODEx latch.
DECBYPA
DECBYPB
DECBYPC
DECBYPD
Receiver Decoder Bypass. The initialization value of the DECBYPx latch = 1. DECBYPx selects if the
Receiver Decoder is enabled or bypassed. When DECBYPx = 1, the decoder is enabled and the Decoder
Mode is selected by DECMODEx. When DECBYPx = 0 and DECMODEx = 1, the decoder is bypassed
and raw 10-bit characters are passed through the receiver.
RXCKSELA
RXCKSELB
RXCKSELC
RXCKSELD
Receive Clock Select. The initialization value of the RXCKSELx latch = 1. RXCKSELx selects the
receive clock source used to transfer data to the Output Registers and the clock source for the RXCLK±
output. When RXCKSELx = 1, the associated Output Registers, are clocked by REFCLKx± at the
associated RXCLKx± output buffer. When RXCKSELx = 0, the associated Output Registers, are clocked
by the Recovered Byte clock at the associated RXCLKx± output buffer. These output clocks may operate
at the character-rate or half the character-rate as selected by RXRATEx.
Document #: 38-02097 Rev. **
Page 20 of 43
PRELIMINARY
CYV15G0404DXB
Table 9. Device Configuration and Control Latch Descriptions (continued)
RXRATEA
RXRATEB
RXRATEC
RXRATED
Receive Clock Rate Select. The initialization value of the RXRATEx latch = 1. RXRATEx is used to
select the rate of the RXCLKx± clock output.
When RXRATEx = 1 and RXCKSELx = 0, the RXCLKx± clock outputs are complementary clocks that
follow the recovered clock operating at half the character rate. Data for the associated receive channels
should be latched alternately on the rising edge of RXCLKx+ and RXCLKx–.
When RXRATEx = 0 and RXCKSELx = 0, the RXCLKx± clock outputs are complementary clocks that
follow the recovered clock operating at the character rate. Data for the associated receive channels
should be latched on the rising edge of RXCLKx+ or falling edge of RXCLKx–.
When RXRATEx = 1 with RXCKSELx = 1 and REFCLKx± is a full-rate clock, the RXCLKx± clock outputs
are complementary clocks that follow the reference clock operating at half the character rate. Data for
the associated receive channels should be latched alternately on the rising edge of RXCLKx+ and
RXCLKx–.
When RXRATEx = 0 with RXCKSELx = 1 and REFCLKx± is a full-rate clock, the RXCLKx± clock outputs
are complementary clocks that follow the reference clock operating at the character rate. Data for the
associated receive channels should be latched on the rising edge of RXCLKx+ or falling edge of
RXCLKx–.
When RXCKSELx = 1 and REFCLKx± is a half-rate clock, the value of RXRATEx is not interpreted and
the RXCLKx± clock outputs are complementary clocks that follow the reference clock operating at half
the character rate. Data for the associated receive channels should be latched alternately on the rising
edge of RXCLKx+ and RXCLKx–.
SDASEL1A[1:0]
SDASEL1B[1:0]
SDASEL1C[1:0]
SDASEL1D[1:0]
Primary Serial Data Input Signal Detector Amplitude Select. The initialization value of the
SDASEL1x[1:0] latch = 10. SDASEL1x[1:0] selects the trip point for the detection of a valid signal for the
INx1± Primary Differential Serial Data Inputs.
When SDASEL1x[1:0] = 00, the Analog Signal Detector is disabled.
When SDASEL1x[1:0] = 01, the typical p-p differential voltage threshold level is 140mV.
When SDASEL1x[1:0] = 10, the typical p-p differential voltage threshold level is 280mV.
When SDASEL1x[1:0] = 11, the typical p-p differential voltage threshold level is 420mV.
SDASEL2A[1:0]
SDASEL2B[1:0]
SDASEL2C[1:0]
SDASEL2D[1:0]
Secondary Serial Data Input Signal Detector Amplitude Select. The initialization value of the
SDASEL2x[1:0] latch = 10. SDASEL2x[1:0] selects the trip point for the detection of a valid signal for the
INx2± Secondary Differential Serial Data Inputs.
When SDASEL2x[1:0] = 00, the Analog Signal Detector is disabled
When SDASEL2x[1:0] = 01, the typical p-p differential voltage threshold level is 140mV.
When SDASEL2x[1:0] = 10, the typical p-p differential voltage threshold level is 280mV.
When SDASEL2x[1:0] = 11, the typical p-p differential voltage threshold level is 420mV.
ENCBYPA
ENCBYPB
ENCBYPC
ENCBYPD
Transmit Encoder Bypassed. The initialization value of the ENCBYPx latch = 1. ENCBYPx selects if
the Transmit Encoder is enabled or bypassed. When ENCBYPx = 1, the Transmit encoder is enabled.
When ENCBYPx = 0, the Transmit Encoder is bypassed and raw 10-bit characters are transmitted.
TXCKSELA
TXCKSELB
TXCKSELC
TXCKSELD
Transmit Clock Select. The initialization value of the TXCKSELx latch = 1. TXCKSELx selects the clock
source used to write data into the Transmit Input Register. When TXCKSELx = 1, the associated input
register, TXDx[7:0] and TXCTx[1:0], is clocked by REFCLKx↑. In this mode, the phase alignment buffer
in the transmit path is bypassed. When TXCKSELx = 0, the associated TXCLKx↑ is used to clock in the
input registers, TXDx[7:0] and TXCTx[1:0].
TXRATEA
TXRATEB
TXRATEC
TXRATED
Transmit PLL Clock Rate Select. The initialization value of the TXRATEx latch = 0. TXRATEx is used
to select the clock multiplier for the Transmit PLL. When TXRATEx = 0, each transmit PLL multiples the
associated REFCLKx± input by 10 to generate the serial bit-rate clock. When TXRATEx = 0, the
TXCLKOx output clocks are full-rate clocks and follow the frequency and duty cycle of the associated
REFCLKx± input. When TXRATEx = 1, each Transmit PLL multiplies the associated REFCLKx± input
by 20 to generate the serial bit-rate clock. When TXRATEx = 1, the TXCLKOx output clocks are twice
the frequency rate of the REFCLKx± input. When TXCLKSELx = 1 and TXRATEx = 1, the Transmit Data
Inputs are captured using both the rising and falling edges of REFCLKx. TXRATEx = 1 and SPDSELx
is LOW, is an invalid state and this combination is reserved.
RFENA
RFENB
RFENC
RFEND
Reframe Enable. The initialization value of the RFENx latch = 1. RFENx selects if the receiver framer
is enabled or disabled. When RFENx = 1, the associated channel’s framer is enabled to frame per the
presently enabled framing mode and selected framing character. When RFENx = 0, the associated
channel’s framer is disabled, and no received bits alters the frame offset.
Document #: 38-02097 Rev. **
Page 21 of 43
PRELIMINARY
CYV15G0404DXB
Table 9. Device Configuration and Control Latch Descriptions (continued)
RXPLLPDA
RXPLLPDB
RXPLLPDC
RXPLLPDD
Receive Channel Enable. The initialization value of the RXPLLPDx latch = 0. RXPLLPDx selects if the
associated receive channel is enabled or powered-down. When RXPLLPDx = 0, the associated PLL and
analog circuitry is powered-down. When RXPLLPDx = 1, the associated PLL and analog circuitry is
enabled.
RXBISTA
RXBISTB
RXBISTC
RXBISTD
Receive Bist Disabled. The initialization value of the RXBISTx latch = 1. RXBISTx selects if receive
BIST is disabled or enabled. When RXBISTx = 1, the receiver BIST function is disabled. When RXBISTx
= 0, the receive BIST function is enabled.
TXBISTA
TXBISTB
TXBISTC
TXBISTD
Transmit Bist Disabled. The initialization value of the TXBISTx latch = 1. TXBISTx selects if the transmit
BIST is disabled or enabled. When TXBISTx = 1, the transmit BIST function is disabled. When
TXBISTx = 0, the transmit BIST function is enabled.
OE2A
OE2B
OE2C
OE2D
Secondary Differential Serial Data Output Driver Enable. The initialization value of the OE2x latch =
0. OE2x selects if the OUT2± secondary differential output drivers are enabled or disabled. When OE2x
= 1, the associated serial data output driver is enabled allowing data to be transmitted from the transmit
shifter. When OE2x = 0, the associated serial data output driver is disabled. When a driver is disabled
via the configuration interface, it is internally powered down to reduce device power. If both serial drivers
for a channel are in this disabled state, the associated internal logic for that channel is also powered
down. A device reset (RESET sampled LOW) disables all output drivers.
OE1A
OE1B
OE1C
OE1D
Primary Differential Serial Data Output Driver Enable. The initialization value of the OE1x latch = 0.
OE1x selects if the OUT1± primary differential output drivers are enabled or disabled. When OE1x = 1,
the associated serial data output driver is enabled allowing data to be transmitted from the transmit
shifter. When OE1x = 0, the associated serial data output driver is disabled. When a driver is disabled
via the configuration interface, it is internally powered down to reduce device power. If both serial drivers
for a channel are in this disabled state, the associated internal logic for that channel is also powered
down. A device reset (RESET sampled LOW) disables all output drivers.
PABRSTA
PABRSTB
PABRSTC
PABRSTD
Transmit Clock Phase Alignment Buffer Reset. The initialization value of the PABRSTx latch = 1. The
PABRSTx is used to re-center the Transmit Phase Align Buffer. When the configuration latch PABRSTx
is written as a 0, the phase of the TXCLKx input clock relative to its associated REFCLKx+/- is initialized.
PABRST is an asynchronous input, but is sampled by each TXCLKx↑ to synchronize it to the internal
clock domain. PABRSTx is a self clearing latch. This eliminates the requirement of writing a 1 to complete
the initialization of the Phase Alignment Buffer.
GLEN[11..0]
Global Enable. The initialization value of the GLENx latch = 1. The GLENx is used to reconfigure several
channels simultaneously in applications where several channels may have the same configuration.
When GLENx = 1 for a given address, that address is allowed to participate in a global configuration.
When GLENx = 0 for a given address, that address is disabled from participating in a global configuration.
FGLEN[2..0]
Force Global Enable. The initialization value of the FGLENx latch is NA. The FGLENx latch forces a
GLobal ENable no matter what the setting is on the GLENx latch. If FGLENx = 1 for the associated
Global channel, FGLEN forces the global update of the target latch banks.
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.
2. Set the static receiver latch bank for the target channel. May
be performed using a global operation, if the application
permits it. [Optional step if the default settings match the
desired configuration.]
3. Set the static transmitter latch bank for the target channel.
May be performed using a global operation, if the application permits it. [Optional step if the default settings match
the desired configuration.]
Document #: 38-02097 Rev. **
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, DECMODEx = 1, 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.
Page 22 of 43
CYV15G0404DXB
PRELIMINARY
Table 10. Device Control Latch Configuration Table
ADDR Channel Type
DATA7
DATA6
DATA5
DATA4
DATA3
DATA2
DATA1
DATA0
Reset
Value
0
(0000b)
A
S
RFMODEA[1]
RFMODEA[0]
FRAMCHARA
DECMODEA
DECBYPA
RXCKSELA
RXRATEA
GLEN0
10111111
1
(0001b)
A
S
SDASEL2A[1]
SDASEL2A[0]
SDASEL1A[1]
SDASEL1A[0]
ENCBYPA
TXCKSELA
TXRATEA
GLEN1
10101101
2
(0010b)
A
D
RFENA
RXPLLPDA
RXBISTA
TXBISTA
OE2A
OE1A
PABRSTA
GLEN2
10110011
3
(0011b)
B
S
RFMODEB[1]
RFMODEB[0]
FRAMCHARB
DECMODEB
DECBYPB
RXCKSELB
RXRATEB
GLEN3
10111111
4
(0100b)
B
S
SDASEL2B[1]
SDASEL2B[0]
SDASEL1B[1]
SDASEL1B[0]
ENCBYPB
TXCKSELB
TXRATEB
GLEN4
10101101
5
(0101b)
B
D
RFENB
RXPLLPDB
RXBISTB
TXBISTB
OE2B
OE1B
PABRSTB
GLEN5
10110011
6
(0110b)
C
S
RFMODEC[1]
RFMODEC[0]
FRAMCHARC
DECMODEC
DECBYPC
RXCKSELC
RXRATEC
GLEN6
10111111
7
(0111b)
C
S
SDASEL2C[1]
SDASEL2C[0]
SDASEL1C[1]
SDASEL1C[0]
ENCBYPC
TXCKSELC
TXRATEC
GLEN7
10101101
8
(1000b)
C
D
RFENC
RXPLLPDC
RXBISTC
TXBISTC
OE2C
OE1C
PABRSTC
GLEN8
10110011
9
(1001b)
D
S
RFMODED[1]
RFMODED[0]
FRAMCHARD
DECMODED
DECBYPD
RXCKSELD
RXRATE D
GLEN9
10111111
10
(1010b)
D
S
SDASEL2D[1]
SDASEL2D[0]
SDASEL1D[1]
SDASEL1D[0]
ENCBYPD
TXCKSELD
TXRATED
GLEN10
10101101
11
(1011b)
D
D
RFEND
RXPLLPDD
RXBISTD
TXBISTD
OE2D
OE1D
PABRSTD
GLEN11
10110011
12
GLOBAL
(1100b)
S
RFMODEGL[1]
RFMODE
GL[0]
FRAMCHARGL DECMODEGL DECBYPGL RXCKSELGL RXRATEG
L
FGLEN0
N/A
13
GLOBAL
(1101b)
S
SDASEL2GL[1] SDASEL2GL[ SDASEL1GL[1] SDASEL1GL[0 ENCBPGL
0]
]
14
GLOBAL
(1110b)
D
RFENGL
RXPLLPDGL
RXBISTGL
TXBISTGL
15
(1111b)
D
D7
D6
D5
D4
MASK
TXCKSELGL
TXRATEG
L
FGLEN1
N/A
OE2GL
OE1GL
PABRSTG
L
FGLEN2
N/A
D3
D2
D1
D0
11111111
JTAG Support
3-Level Select Inputs
The CYV15G0404DXB contains a JTAG port to allow system
level diagnosis of device interconnect. Of the available JTAG
modes, boundary scan, and bypass are supported. This
capability is present only on the LVTTL inputs and outputs and
the REFCLKx± clock input. The high-speed serial inputs and
outputs are not part of the JTAG test chain.
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
Document #: 38-02097 Rev. **
JTAG ID
The JTAG device ID for the CYV15G0404DXB is ‘0C811069’x.
Page 23 of 43
PRELIMINARY
CYV15G0404DXB
Table 11. Receive Character Status Bits
Description
RXSTx[2:0]
Priority
000
7
Normal character received. The valid Data BIST Data Compare. Character compared correctly.
character on the output bus meets all the
formatting requirements of Data characters
listed in Table 15.
001
7
Special code detected. The valid special BIST Command Compare. Character compared
character on the output bus meets all the correctly.
formatting requirements of Special Code
characters listed in Table 16, but is not the
presently selected framing character or a
decoder violation indication.
010
2
Receive Elasticity buffer underrun/overrun BIST Last Good. Last Character of BIST sequence
error. The receive buffer was not able to detected and valid.
add/drop a K28.5 or framing character
011
5
Framing character detected. This indicates
that a character matching the patterns
identified as a framing character (as selected
by FRAMCHARx) was detected. The decoded
value of this character is present in the
associated output bus.
100
4
Codeword violation. The character on the BIST Last Bad. Last Character of BIST sequence
output bus is a C0.7. This indicates that the detected invalid.
received character cannot be decoded into any
valid character.
101
1
Loss of sync. This indicates a PLL Out of Lock BIST Start. Receive BIST is enabled on this channel,
condition
but character compares have not yet commenced.
This also indicates a PLL Out of Lock condition, and
Elasticity Buffer overflow/underflow conditions.
110
6
Running disparity error. The character on the BIST Error. While comparing characters, a mismatch
output bus is a C4.7, C1.7, or C2.7.
was found in one or more of the decoded character
bits.
111
3
Reserved
Normal Status
Document #: 38-02097 Rev. **
Receive BIST Status
(Receive BIST = Enabled)
BIST Wait. The receiver is comparing characters. but
has not yet found the start of BIST character to enable
the LFSR.
Page 24 of 43
CYV15G0404DXB
PRELIMINARY
Monitor Data
Received
RX PLL
Out of Lock
RXSTx =
BIST_START (101)
RXSTx =
BIST_WAIT (111)
Elasticity
Buffer Error
Yes
No
RXSTx =
BIST_START (101)
Receive BIST
Detected LOW
Start of
BIST Detected
No
Yes, RXSTx =
BIST_DATA_COMPARE (000) / BIST_COMMAND_COMPARE (001)
Compare
Next Character
RXSTx =
Match BIST_COMMAND_COMPARE (001)
Mismatch
Yes
Command
Auto-Abort
Condition
Data or
Command
No
Data
End-of-BIST
State
End-of-BIST
State
Yes, RXSTx =
BIST_LAST_BAD (100)
Yes, RXSTx =
BIST_LAST_GOOD (010)
RXSTx =
BIST_DATA_COMPARE (000)
No
No, RXSTx =
BIST_ERROR (110)
Figure 2. Receive BIST State Machine
Document #: 38-02097 Rev. **
Page 25 of 43
CYV15G0404DXB
PRELIMINARY
Maximum Ratings
Above which the useful life may be impaired. User guidelines
only, not tested
Static Discharge Voltage.......................................... > 2000 V
(per MIL-STD-883, Method 3015)
Latch-up Current..................................................... > 200 mA
Storage Temperature .................................. –65°C to +150°C
Power-up Requirements
Ambient Temperature with
Power Applied............................................. –55°C to +125°C
The CYP(V)15G0404DXB requires one power-supply. The
Voltage on any input or I/O pin cannot exceed the power pin
during power-up.
Supply Voltage to Ground Potential ............... –0.5V to +3.8V
Operating Range
DC Voltage Applied to LVTTL Outputs
in High-Z State .......................................–0.5V to VCC + 0.5V
Range
Commercial
Industrial
Output Current into LVTTL Outputs (LOW)..................60 mA
DC Input Voltage....................................–0.5V to VCC + 0.5V
Ambient Temperature
0°C to +70°C
–40°C to +85°C
VCC
+3.3V ±5%
+3.3V ±5%
CYV15G0404DXB DC Electrical Characteristics
Parameter
Description
Test Conditions
Min.
Max.
Unit
LVTTL-compatible Outputs
VOHT
Output HIGH Voltage
IOH = − 4 mA, VCC = Min.
VOLT
Output LOW Voltage
IOL = 4 mA, VCC = Min.
IOST
Output Short Circuit Current
VOUT = 0V[12], VCC
IOZL
High-Z Output Leakage Current
VOUT = 0V, VCC
= 3.3V
2.4
V
0.4
V
–20
–100
mA
–20
20
µA
2.0
VCC + 0.3
V
–0.5
0.8
V
1.5
mA
LVTTL-compatible Inputs
VIHT
Input HIGH Voltage
VILT
Input LOW Voltage
IIHT
Input HIGH Current
Other Inputs, VIN = VCC
+40
µA
IILT
Input LOW Current
REFCLKx Input, VIN = 0.0V
–1.5
mA
Other Inputs, VIN = 0.0V
–40
µA
IIHPDT
Input HIGH Current with internal pull-down
VIN = VCC
+200
µA
IILPUT
Input LOW Current with internal pull-up
VIN = 0.0V
–200
µA
REFCLKx Input, VIN = VCC
LVDIFF Inputs: REFCLKx±
VDIFF[13]
Input Differential Voltage
400
VCC
mV
VIHHP
Highest Input HIGH Voltage
1.2
VCC
V
VILLP
Lowest Input LOW voltage
VCOMREF[14] Common Mode Range
0.0
VCC/2
V
1.0
VCC – 1.2V
V
3-Level Inputs
VIHH
Three-Level Input HIGH Voltage
Min. ≤ VCC ≤ Max.
0.87 * VCC
VCC
V
VIMM
Three-Level Input MID Voltage
Min. ≤ VCC ≤ Max.
0.47 * VCC
0.53 * VCC
V
VILL
Three-Level Input LOW Voltage
Min. ≤ VCC ≤ Max.
0.0
0.13 * VCC
V
200
µA
50
µA
–200
µA
IIHH
Input HIGH Current
VIN = VCC
IIMM
Input MID current
VIN = VCC/2
IILL
Input LOW current
VIN = GND
–50
Differential CML Serial Outputs: OUTA1±, OUTA2±, OUTB1±, OUTB2±, OUTC1±, OUTC2±, OUTD1±, OUTD2±
VOHC
Output HIGH Voltage
(Vcc Referenced)
100Ω differential load
VCC – 0.5
VCC – 0.2
V
150Ω differential load
VCC – 0.5
VCC – 0.2
V
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 inputs required to ensure detection of a logic-1 or logic-0. A logic-1 exists when
the true (+) input is more positive than the complement (−) input. A logic-0 exists when the complement (−) input is more positive than true (+) input.
14. The common mode range defines the allowable range of REFCLKx+ and REFCLKx− when REFCLKx+ = REFCLKx−. This marks the zero-crossing between
the true and complement inputs as the signal switches between a logic-1 and a logic-0.
Document #: 38-02097 Rev. **
Page 26 of 43
CYV15G0404DXB
PRELIMINARY
CYV15G0404DXB DC Electrical Characteristics (continued)
Parameter
Description
Test Conditions
Min.
Max.
VCC – 0.7
Unit
VOLC
Output LOW Voltage
(VCC Referenced)
100Ω differential load
VCC – 1.4
V
150Ω differential load
VCC – 1.4
VCC – 0.7
V
VODIF
Output Differential Voltage
|(OUT+) − (OUT−)|
100Ω differential load
450
900
mV
150Ω differential load
560
1000
mV
1200
mV
VCC
V
1350
µA
Differential Serial Line Receiver Inputs: INA1±, INA2±, INB1±, INB2±, INC1±, INC2±, IND1±, IND2±
VDIFFs[13]
Input Differential Voltage |(IN+) − (IN−)|
VIHE
Highest Input HIGH Voltage
100
VILE
Lowest Input LOW Voltage
IIHE
Input HIGH Current
VCC – 2.0
IILE
Input LOW Current
VIN = VILE Min.
–700
VICOM[15]
Common Mode input range
((VCC – 2.0V)+0.5)min,
(VCC – 0.5V) max.
+1.25
+3.1
Typ.
Max.
Power Supply
ICC
[16,17]
Max Power Supply Current
ICC [16,17]
V
VIN = VIHE Max.
REFCLKx
= MAX
Typical Power Supply Current
Commercial
µA
910
Industrial
REFCLKx Commercial
= 125 MHz Industrial
900
V
1270
mA
1320
mA
1270
mA
1320
mA
AC Test Loads and Waveforms
3.3V
RL = 100Ω
R1
R1 = 590Ω
R2 = 435Ω
CL
CL ≤ 7 pF
(Includes fixture and
probe capacitance)
(Includes fixture and
probe capacitance)
R2
(a) LVTTL Output Test Load
(b) CML Output Test Load
GND
2.0V
2.0V
0.8V
0.8V
VIHE
VIHE
Vth = 1.4V
≤ 1 ns
VILE
≤ 1 ns
(c) LVTTL Input Test Waveform
Document #: 38-02097 Rev. **
Note 18
Note 18
3.0V
Vth = 1.4V
RL
Note 19
80%
80%
20%
≤ 270 ps
20%
VILE
≤ 270 ps
(d) CML/LVPECL Input Test Waveform
Page 27 of 43
CYV15G0404DXB
PRELIMINARY
CYV15G0404DXB AC Electrical Characteristics
Parameter
Description
Min.
Max
Unit
CYV15G0404DXB Transmitter LVTTL Switching Characteristics Over the Operating Range
fTS
TXCLKx Clock Cycle Frequency
19.5
150
MHz
tTXCLK
TXCLKx Period=1/fTS
6.66
51.28
ns
tTXCLKH[20]
tTXCLKL[20]
tTXCLKR [20, 21, 22]
tTXCLKF [20, 21, 22]
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
tTXDS
Transmit Data Set-up Time to TXCLKx↑ (TXCKSELx ≠ 0)
1.7
ns
tTXDH
Transmit Data Hold Time from TXCLKx↑ (TXCKSELx ≠ 0)
0.8
ns
fTOS
TXCLKOx Clock Frequency = 1x or 2x REFCLKx Frequency
19.5
150
MHz
tTXCLKO
TXCLKOx Period=1/fTOS
6.66
51.28
ns
tTXCLKOD
TXCLKO Duty Cycle centered at 60% HIGH time
-1
0
ns
CYV15G0404DXB Receiver LVTTL Switching Characteristics Over the Operating Range
fRS
RXCLKx± Clock Output Frequency
9.75
150
MHz
tRXCLKP
RXCLKx± Period = 1/fRS
6.66
102.56
ns
tRXCLKD
RXCLKx± Duty Cycle Centered at 50% (Full Rate and Half Rate when
RXCKSELx = 0)
–1.0
+1.0
ns
tRXCLKR [20]
RXCLKx± Rise Time
0.3
1.2
ns
tRXCLKF [20]
RXCLKx± Fall Time
0.3
1.2
ns
tRXDv–
[23]
tRXDv+[23]
Status and Data Valid Time to RXCLKx± (RXRATEx = 0, RXCKSELx = 0)
(Full Rate)
5UI–1.5[24]
ns
Status and Data Valid Time to RXCLKx± (RXRATEx = 1, RXCKSELx = 0)
(Half Rate)
5UI–1.0[24]
ns
Status and Data Valid Time to RXCLKx± (RXRATEx = 0, RXCKSELx = 0)
5UI–1.8[24]
ns
Status and Data Valid Time to RXCLKx± (RXRATEx = 1, RXCKSELx =0)
5UI–2.3[24]
ns
CYV15G0404DXB REFCLKx Switching Characteristics Over the Operating Range
fREF
REFCLKx Clock Frequency
19.5
150
MHz
tREFCLK
REFCLKx Period = 1/fREF
6.6
51.28
ns
tREFH
REFCLKx HIGH Time (TXRATEx = 1)(Half Rate)
5.9
ns
REFCLKx HIGH Time (TXRATEx = 0)(Full Rate)
2.9[20]
ns
REFCLKx LOW Time (TXRATEx = 1)(Half Rate)
5.9
ns
REFCLKx LOW Time (TXRATEx = 0)(Full Rate)
2.9[20]
tREFL
tREFD[25]
tREFR
[20, 21, 22]
REFCLKx Duty Cycle
30
ns
70
%
REFCLKx Rise Time (20%–80%)
2
ns
tREFF[20, 21, 22]
REFCLKx Fall Time (20%–80%)
2
ns
tTREFDS
Transmit Data Set-up Time to REFCLKx (TXCKSELx = 1)
1.7
ns
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 pattern. The redundant outputs on each channel are powered down and the parallel outputs are unloaded.
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.
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 loaded with similar DC and AC loads.
24. Receiver UI (Unit Interval) is calculated as 1/(fREF * 20) (when TXRATEx = 1) or 1/(fREF * 10) (when TXRATEx = 0). In an operating link this is equivalent to tB.
25. The duty cycle specification is a simultaneous condition with the tREFH and tREFL parameters. This means that at faster character rates the REFCLKx± duty
cycle cannot be as large as 30%–70%.
Document #: 38-02097 Rev. **
Page 28 of 43
CYV15G0404DXB
PRELIMINARY
CYV15G0404DXB AC Electrical Characteristics (continued)
Parameter
tTREFDH
Description
Min.
Transmit Data Hold Time from REFCLKx (TXCKSELx = 1)
tRREFDA
Receive Data Access Time to REFCLKx (RXCKSELx = 1)
tRREFDW
Receive Data Valid Time Window (RXCKSELx = 1)
Max
0.8
ns
[26]
9.7
10UI-5.8
Received Data Valid Time to RXCLK (RXCKSELx = 1)
tREFxDV+
Received Data valid Time from RXCLK (RXCKSELx = 1)
tREFRX[27]
REFCLKx Frequency Referenced to Received Clock Period
10UI –
5.2
ns
1.5
–0.15
ns
ns
[24]
tREFxDV–
Unit
ns
+0.15
%
CYV15G0404DXB Bus Configuration Write Timing Characteristics Over the Operating Range
tDATAH
Bus Configuration Data Hold
0
ns
tDATAS
Bus Configuration Data Setup
10
ns
tWRENP
Bus Configuration WREN Pulse Width
10
ns
CYV15G0404DXB JTAG Test Clock Characteristics Over the Operating Range
fTCLK
JTAG Test Clock Frequency
tTCLK
JTAG Test Clock Period
20
MHz
50
ns
15
ns
CYV15G0404DXB Device RESET Characteristics Over the Operating Range
tRST
Device RESET Pulse Width
CYV15G0404DXB Transmit Serial Outputs and TX PLL Characteristics Over the Operating Range
Parameter
Description
tB
Bit Time
tRISE[20]
CML Output Rise Time 20−80% (CML Test Load)
tFALL[20]
tDJ
[20, 28, 30]
tRJ[20, 29, 30]
[20]
CML Output Fall Time 80−20% (CML Test Load)
Deterministic Jitter
(peak-peak)[31]
Random Jitter (σ)[31]
Condition
Min.
Max.
Unit
5128
660
ps
SPDSELx = HIGH
50
270
ps
SPDSELx = MID
100
500
ps
SPDSELx =LOW
180
1000
ps
SPDSELx = HIGH
50
270
ps
SPDSELx = MID
100
500
ps
SPDSELx =LOW
180
1000
ps
IEEE 802.3z
25
ps
IEEE 802.3z
11
ps
tREFJ
REFCLKx jitter tolerance / Phase noise limits
TBD
tTXLOCK
Transmit PLLx lock to REFCLKx±
200
µs
Receive PLL lock to input data stream (cold start)
376k
UI
Receive PLL lock to input data stream
376k
UI
46
UI
CYV15G0404DXB Receive Serial Inputs and CDR PLL Characteristics Over the Operating Range
tRXLOCK
tRXUNLOCK
Receive PLL Unlock Rate
tJTOL[20]
Total Jitter Tolerance[31]
IEEE 802.3z
600
ps
tDJTOL[20]
Deterministic Jitter Tolerance[31]
IEEE 802.3z
370
ps
Notes:
26. Since this timing parameter is greater than the minimum time period of REFCLK it sets an upper limit to the frequency in which REFCLKx can be used to clock
the receive data out of the output register. For predictable timing, users can use this parameter only if REFCLK period is greater than sum of tRREFDA and setup time of the upstream device. When this condition is not true, RXCLKx± (a buffered or divided version of REFCLK when RXCKSELx = 1) could be used to
clock the receive data out of the device.
27. 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 transmitter PLL reference (REFCLKx±) frequency. Although transmitting to a HOTLink II receiver
necessitates the frequency difference between the transmitter and receiver reference clocks to be within ±1500-PPM, the stability of the crystal needs to be
within the limits specified by the appropriate standard when transmitting to a remote receiver that is compliant to that standard. For example, to be IEEE 802.3z
Gigabit Ethernet compliant, the frequency stability of the crystal needs to be within ±100 PPM.l.
28. While sending continuous K28.5s, outputs loaded to a balanced 100Ω load, measured at the cross point of differential outputs, over the operating range.
29. While sending continuous K28.7s, after 100,000 samples measured at the cross point of differential outputs, time referenced to REFCLKx± input, over the
operating range.
30. Total jitter is calculated at an assumed BER of 1E −12. Hence: Total Jitter (tJ) = (tRJ * 14) + tDJ.
31. Also meets all Jitter Generation and Jitter Tolerance requirements as specified by SMPTE 259, SMPTE 292, ESCON, FICON, Fibre Channel, and DVB-ASI.
Document #: 38-02097 Rev. **
Page 29 of 43
CYV15G0404DXB
PRELIMINARY
Capacitance [20]
Max.
Unit
CINTTL
Parameter
TTL Input Capacitance
Description
TA = 25°C, f0 = 1 MHz, VCC = 3.3V
Test Conditions
7
pF
CINPECL
PECL input Capacitance
TA = 25°C, f0 = 1 MHz, VCC = 3.3V
4
pF
CYV15G0404DXB HOTLink II Transmitter Switching Waveforms
tTXCLK
Transmit Interface
Write Timing
TXCLKx selected
tTXCLKH
tTXCLKL
TXCLKx
tTXDS
TXDx[7:0],
TXCTx[1:0],
Transmit Interface
Write Timing
REFCLKx selected
TXRATEx = 0
tREFH
tREFCLK
tTXDH
tREFL
REFCLKx
tTREFDS
TXDx[7:0],
TXCTx[1:0],
Transmit Interface
Write Timing
REFCLKx selected
TXRATEx = 1
tTREFDH
tREFCLK
tREFH
tREFL
REFCLKx
Note 32
tTREFDS
TXDx[7:0],
TXCTx[1:0],
Transmit Interface
TXCLKOx Timing
tTREFDH
tTREFDS
tTREFDH
tREFCLK
tREFH
TXRATEx = 1
REFCLKx
tREFL
Note 33
tTXCLKO
Note 34
TXCLKOx
(internal)
Notes:
32. When REFCLKx± is configured for half-rate operation (TXRATE = 1) and data is captured using REFCLKx instead of a TXCLKx clock. Data is captured using
both the rising and falling edges of REFCLKx.
33. The TXCLKOx output remains at the character rate regardless of the state of TXRATE and does not follow the duty cycle of REFCLKx±.
34. The rising edge of TXCLKOx output has no direct phase relationship to the REFCLKx± input.
Document #: 38-02097 Rev. **
Page 30 of 43
CYV15G0404DXB
PRELIMINARY
CYV15G0404DXB HOTLink II Transmitter Switching Waveforms (continued)
Transmit Interface
TXCLKOx Timing
tREFCLK
tREFH
TXRATEx = 0
tREFL
Note33
REFCLKx
tTXCLKO
Note34
tTXOH
tTXOL
TXCLKOx
Switching Waveforms for the CYV15G0404DXB HOTLink II Receiver
Receive Interface
Read Timing
REFCLKx Selected
Full-rate RXCLKx±
tREFCLK
tREFH
tREFL
REFCLKx
tRREFDA
tRREFDW
tRREFDW
RXDx[7:0],
RXSTx[2:0],
[36]
TXERRx
tREFxDV+
tREFxDV–
RXCLKx
Receive Interface
Read Timing
REFCLKx Selected
Half-rate RXCLKx±
tREFCLK
tREFH
tREFL
REFCLKx
tRREFDA
tRREFDW
tRREFDA
tRREFDW
RXDx[7:0],
RXSTx[2:0],
TXERRx
[36]
tREFxDV+
RXCLKx
tREFxDV–
Note 35
Note:
35. 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 the
respective clock output
36. TXERRx is synchronous to RXCLKx only when RXCLKx is selected as REFCLK.
Document #: 38-02097 Rev. **
Page 31 of 43
CYV15G0404DXB
PRELIMINARY
Switching Waveforms for the CYV15G0404DXB HOTLink II Receiver
Receive Interface
Read Timing
Recovered Clock selected
RXRATEx = 0
tRXCLKP
RXCLKx+
RXCLKxRXDx[7:0],
RXSTx[2:0],
tRXDV–
tRXDV+
Receive Interface
Read Timing
Recovered Clock selected
RXRATEx = 1
tRXCLKP
RXCLKx+
RXCLKx-
tRXDV–
RXDx[7:0],
RXSTx[2:0]
tRXDV+
Bus Configuration
Write Timing
ADDR[3:0]
DATA[7:0]
tWRENP
WREN
Document #: 38-02097 Rev. **
tDATAS
tDATAH
Page 32 of 43
CYV15G0404DXB
PRELIMINARY
Table 12. Package Coordinate Signal Allocation
Ball
ID
Signal Name
Signal Type
Ball
ID
Signal Name
Signal Type
Ball
ID
Signal Name
Signal Type
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
DATA[7]
LVTTL IN PU
F19
TXCLKOB
LVTTL OUT
A04
OUTC2–
CML OUT
C10
DATA[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
A11
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
H01
GND
GROUND
A14
INA2–
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
LVTTL IN PU
H04
GND
GROUND
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
B11
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
Document #: 38-02097 Rev. **
Page 33 of 43
CYV15G0404DXB
PRELIMINARY
Table 12. Package Coordinate Signal Allocation (continued)
Ball
ID
Signal Name
Signal Type
Ball
ID
Signal Name
Signal Type
Ball
ID
Signal Name
Signal Type
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
LVTTL OUT
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]
LVTTL 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
Document #: 38-02097 Rev. **
Page 34 of 43
PRELIMINARY
X3.230 Codes and Notation Conventions
Information transmitted over a serial link is encoded eight bits
at a time into a 10-bit Transmission Character and then sent
serially, bit by bit. Information received over a serial link is
collected ten bits at a time, and those Transmission
Characters that are used for data characters are decoded into
the correct eight-bit codes. The 10-bit Transmission Code
supports all 256 8-bit combinations. Some of the remaining
Transmission Characters (Special Characters) are used for
functions other than data transmission.
The primary use of a Transmission Code is to improve the
transmission characteristics of a serial link. The encoding
defined by the Transmission Code ensures that sufficient
transitions are present in the serial bit stream to make clock
recovery possible at the Receiver. Such encoding also greatly
increases the likelihood of detecting any single or multiple bit
errors that may occur during transmission and reception of
information. In addition, some Special Characters of the Transmission Code selected by Fibre Channel Standard contain a
distinct and easily recognizable bit pattern that assists the
receiver in achieving character alignment on the incoming bit
stream.
Notation Conventions
The documentation for the 8B/10B Transmission Code uses
letter notation for the bits in an 8-bit byte. Fibre Channel
Standard notation uses a bit notation of A, B, C, D, E, F, G, H
for the 8-bit byte for the raw 8-bit data, and the letters a, b, c,
d, e, i, f, g, h, j for encoded 10-bit data. There is a correspondence between bit A and bit a, B and b, C and c, D and d, E
and e, F and f, G and g, and H and h. Bits i and j are derived,
respectively, from (A,B,C,D,E) and (F,G,H).
The bit labeled A in the description of the 8B/10B Transmission
Code corresponds to bit 0 in the numbering scheme of the FC2 specification, B corresponds to bit 1, as shown below.
FC-2 bit designation—
7 6 5 4 3 2 1 0
HOTLink D/Q designation— 7 6 5 4 3 2 1 0
8B/10B bit designation—
H G F E D C B A
To clarify this correspondence, the following example shows
the conversion from an FC-2 Valid Data Byte to a Transmission
Character.
FC-2 45H
Bits: 7654 3210
0100 0101
Converted to 8B/10B notation, note that the order of bits has
been reversed):
Data Byte Name D5.2
Bits: ABCDE FGH
10100 010
Translated to a transmission Character in the 8B/10B Transmission Code:
Bits: abcdei fghj
101001 0101
Each valid Transmission Character of the 8B/10B Transmission Code has been given a name using the following
convention: cxx.y, where c is used to show whether the Transmission Character is a Data Character (c is set to D, and SC/D
= LOW) or a Special Character (c is set to K, and SC/D =
HIGH). When c is set to D, xx is the decimal value of the binary
Document #: 38-02097 Rev. **
CYV15G0404DXB
number composed of the bits E, D, C, B, and A in that order,
and the y is the decimal value of the binary number composed
of the bits H, G, and F in that order. When c is set to K, xx and
y are derived by comparing the encoded bit patterns of the
Special Character to those patterns derived from encoded
Valid Data bytes and selecting the names of the patterns most
similar to the encoded bit patterns of the Special Character.
Under the above conventions, the Transmission Character
used for the examples above, is referred to by the name D5.2.
The Special Character K29.7 is so named because the first six
bits (abcdei) of this character make up a bit pattern similar to
that resulting from the encoding of the unencoded 11101
pattern (29), and because the second four bits (fghj) make up
a bit pattern similar to that resulting from the encoding of the
unencoded 111 pattern (7).
Note. This definition of the 10-bit Transmission Code is based
on the following references, which describe the same 10-bit
transmission code.
A.X. Widmer and P.A. Franaszek. “A DC-Balanced, Partitioned-Block, 8B/10B Transmission Code” IBM Journal of
Research and Development, 27, No. 5: 440-451 (September,
1983).
U.S. Patent 4,486,739. Peter A. Franaszek and Albert X.
Widmer. “Byte-Oriented DC Balanced (0.4) 8B/10B Partitioned Block Transmission Code” (December 4, 1984).
Fibre Channel Physical and Signaling Interface (ANS X3.2301994 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 SpecialCharacter-code entry has two columns that represent two
Transmission Characters. The two columns correspond to the
current value of the running disparity. Running disparity is a
binary parameter with either a negative (–) or positive (+)
value.
After powering on, the Transmitter may assume either a
positive or negative value for its initial running disparity. Upon
transmission of any Transmission Character, the transmitter
selects the proper version of the Transmission Character
Page 35 of 43
CYV15G0404DXB
PRELIMINARY
based on the current running disparity value, and the Transmitter calculates a new value for its running disparity based on
the contents of the transmitted character. Special Character
codes C1.7 and C2.7 can be used to force the transmission of
a specific Special Character with a specific running disparity
as required for some special sequences in X3.230.
byte or Special Character byte encoded and transmitted.
Table 13 shows naming notations and examples of valid transmission characters.
After powering on, the Receiver may assume either a positive
or negative value for its initial running disparity. Upon reception
of any Transmission Character, the Receiver decides whether
the Transmission Character is valid or invalid according to the
following rules and tables and calculates a new value for its
Running Disparity based on the contents of the received
character.
The column corresponding to the current value of the
Receiver’s running disparity is searched for the received
Transmission Character. If the received Transmission
Character is found in the proper column, then the Transmission Character is valid and the associated Data byte or
Special Character code is determined (decoded). If the
received Transmission Character is not found in that column,
then the Transmission Character is invalid. This is called a
code violation. Independent of the Transmission Character’s
validity, the received Transmission Character is used to
calculate a new value of running disparity. The new value is
used as the Receiver’s current running disparity for the next
received Transmission Character.
Use of the Tables for Checking the Validity of Received
Transmission Characters
The following rules for running disparity are used to calculate
the new running-disparity value for Transmission Characters
that have been transmitted and received.
Running disparity for a Transmission Character is calculated
from sub-blocks, where the first six bits (abcdei) form one subblock and the second four bits (fghj) form the other sub-block.
Running disparity at the beginning of the 6-bit sub-block is the
running disparity at the end of the previous Transmission
Character. Running disparity at the beginning of the 4-bit subblock is the running disparity at the end of the 6-bit sub-block.
Running disparity at the end of the Transmission Character is
the running disparity at the end of the 4-bit sub-block.
Table 13. Valid Transmission Characters
Data
DIN or QOUT
Running disparity for the sub-blocks is calculated as follows:
1. Running disparity at the end of any sub-block is positive if
the sub-block contains more ones than zeros. It is also
positive at the end of the 6-bit sub-block if the 6-bit subblock is 000111, and it is positive at the end of the 4-bit subblock if the 4-bit sub-block is 0011.
2. Running disparity at the end of any sub-block is negative if
the sub-block contains more zeros than ones. It is also
negative at the end of the 6-bit sub-block if the 6-bit subblock is 111000, and it is negative at the end of the 4-bit subblock if the 4-bit sub-block is 1100.
3. Otherwise, running disparity at the end of the sub-block is
the same as at the beginning of the sub-block.
Use of the Tables for Generating Transmission Characters
Byte Name
765
43210
Hex Value
D0.0
000
00000
00
D1.0
000
00001
01
D2.0
000
00010
02
.
.
.
.
.
.
.
.
D5.2
010
00101
45
.
.
.
.
.
.
.
.
D30.7
111
11110
FE
D31.7
111
11111
FF
Detection of a code violation does not necessarily show that
the Transmission Character in which the code violation was
detected is in error. Code violations may result from a prior
error that altered the running disparity of the bit stream which
did not result in a detectable error at the Transmission
Character in which the error occurred. Table 14 shows an
example of this behavior.
The appropriate entry in Table 15 for the Valid Data byte or
Table 16 for Special Character byte identify which Transmission Character is generated. The current value of the
Transmitter’s running disparity is used to select the Transmission Character from its corresponding column. For each
Transmission Character transmitted, a new value of the
running disparity is calculated. This new value is used as the
Transmitter’s current running disparity for the next Valid Data
Table 14. Code Violations Resulting from Prior Errors
RD
Character
RD
Character
RD
Character
RD
Transmitted data character
–
D21.1
–
D10.2
–
D23.5
+
Transmitted bit stream
–
101010 1001
–
010101 0101
–
111010 1010
+
Bit stream after error
–
101010 1011
+
010101 0101
+
111010 1010
+
Decoded data character
–
D21.0
+
D10.2
+
Code Violation
+
Document #: 38-02097 Rev. **
Page 36 of 43
CYV15G0404DXB
PRELIMINARY
Table 15. Valid Data Characters (TXCTx[0] = 0, RXSTx[2:0] = 000)
Data
Byte
Name
Bits
Current RD−
Current RD+
Bits
Current RD−
Current RD+
abcdei fghj
Data
Byte
Name
HGF EDCBA
abcdei fghj
HGF EDCBA
abcdei fghj
abcdei fghj
D0.0
000 00000
100111 0100
011000 1011
D0.1
001 00000
100111 1001
011000 1001
D1.0
000 00001
011101 0100
100010 1011
D1.1
001 00001
011101 1001
100010 1001
D2.0
000 00010
101101 0100
010010 1011
D2.1
001 00010
101101 1001
010010 1001
D3.0
000 00011
110001 1011
110001 0100
D3.1
001 00011
110001 1001
110001 1001
D4.0
000 00100
110101 0100
001010 1011
D4.1
001 00100
110101 1001
001010 1001
D5.0
000 00101
101001 1011
101001 0100
D5.1
001 00101
101001 1001
101001 1001
D6.0
000 00110
011001 1011
011001 0100
D6.1
001 00110
011001 1001
011001 1001
D7.0
000 00111
111000 1011
000111 0100
D7.1
001 00111
111000 1001
000111 1001
D8.0
000 01000
111001 0100
000110 1011
D8.1
001 01000
111001 1001
000110 1001
D9.0
000 01001
100101 1011
100101 0100
D9.1
001 01001
100101 1001
100101 1001
D10.0
000 01010
010101 1011
010101 0100
D10.1
001 01010
010101 1001
010101 1001
D11.0
000 01011
110100 1011
110100 0100
D11.1
001 01011
110100 1001
110100 1001
D12.0
000 01100
001101 1011
001101 0100
D12.1
001 01100
001101 1001
001101 1001
D13.0
000 01101
101100 1011
101100 0100
D13.1
001 01101
101100 1001
101100 1001
D14.0
000 01110
011100 1011
011100 0100
D14.1
001 01110
011100 1001
011100 1001
D15.0
000 01111
010111 0100
101000 1011
D15.1
001 01111
010111 1001
101000 1001
D16.0
000 10000
011011 0100
100100 1011
D16.1
001 10000
011011 1001
100100 1001
D17.0
000 10001
100011 1011
100011 0100
D17.1
001 10001
100011 1001
100011 1001
D18.0
000 10010
010011 1011
010011 0100
D18.1
001 10010
010011 1001
010011 1001
D19.0
000 10011
110010 1011
110010 0100
D19.1
001 10011
110010 1001
110010 1001
D20.0
000 10100
001011 1011
001011 0100
D20.1
001 10100
001011 1001
001011 1001
D21.0
000 10101
101010 1011
101010 0100
D21.1
001 10101
101010 1001
101010 1001
D22.0
000 10110
011010 1011
011010 0100
D22.1
001 10110
011010 1001
011010 1001
D23.0
000 10111
111010 0100
000101 1011
D23.1
001 10111
111010 1001
000101 1001
D24.0
000 11000
110011 0100
001100 1011
D24.1
001 11000
110011 1001
001100 1001
D25.0
000 11001
100110 1011
100110 0100
D25.1
001 11001
100110 1001
100110 1001
D26.0
000 11010
010110 1011
010110 0100
D26.1
001 11010
010110 1001
010110 1001
D27.0
000 11011
110110 0100
001001 1011
D27.1
001 11011
110110 1001
001001 1001
D28.0
000 11100
001110 1011
001110 0100
D28.1
001 11100
001110 1001
001110 1001
D29.0
000 11101
101110 0100
010001 1011
D29.1
001 11101
101110 1001
010001 1001
D30.0
000 11110
011110 0100
100001 1011
D30.1
001 11110
011110 1001
100001 1001
D31.0
000 11111
101011 0100
010100 1011
D31.1
001 11111
101011 1001
010100 1001
Document #: 38-02097 Rev. **
Page 37 of 43
CYV15G0404DXB
PRELIMINARY
Table 15. Valid Data Characters (TXCTx[0] = 0, RXSTx[2:0] = 000) (continued)
Data
Byte
Name
Bits
Current RD−
Current RD+
Bits
Current RD−
Current RD+
abcdei fghj
Data
Byte
Name
HGF EDCBA
abcdei fghj
HGF EDCBA
abcdei fghj
abcdei fghj
D0.2
010 00000
100111 0101
011000 0101
D0.3
011 00000
100111 0011
011000 1100
D1.2
010 00001
011101 0101
100010 0101
D1.3
011 00001
011101 0011
100010 1100
D2.2
010 00010
101101 0101
010010 0101
D2.3
011 00010
101101 0011
010010 1100
D3.2
010 00011
110001 0101
110001 0101
D3.3
011 00011
110001 1100
110001 0011
D4.2
010 00100
110101 0101
001010 0101
D4.3
011 00100
110101 0011
001010 1100
D5.2
010 00101
101001 0101
101001 0101
D5.3
011 00101
101001 1100
101001 0011
D6.2
010 00110
011001 0101
011001 0101
D6.3
011 00110
011001 1100
011001 0011
D7.2
010 00111
111000 0101
000111 0101
D7.3
011 00111
111000 1100
000111 0011
D8.2
010 01000
111001 0101
000110 0101
D8.3
011 01000
111001 0011
000110 1100
D9.2
010 01001
100101 0101
100101 0101
D9.3
011 01001
100101 1100
100101 0011
D10.2
010 01010
010101 0101
010101 0101
D10.3
011 01010
010101 1100
010101 0011
D11.2
010 01011
110100 0101
110100 0101
D11.3
011 01011
110100 1100
110100 0011
D12.2
010 01100
001101 0101
001101 0101
D12.3
011 01100
001101 1100
001101 0011
D13.2
010 01101
101100 0101
101100 0101
D13.3
011 01101
101100 1100
101100 0011
D14.2
010 01110
011100 0101
011100 0101
D14.3
011 01110
011100 1100
011100 0011
D15.2
010 01111
010111 0101
101000 0101
D15.3
011 01111
010111 0011
101000 1100
D16.2
010 10000
011011 0101
100100 0101
D16.3
011 10000
011011 0011
100100 1100
D17.2
010 10001
100011 0101
100011 0101
D17.3
011 10001
100011 1100
100011 0011
D18.2
010 10010
010011 0101
010011 0101
D18.3
011 10010
010011 1100
010011 0011
D19.2
010 10011
110010 0101
110010 0101
D19.3
011 10011
110010 1100
110010 0011
D20.2
010 10100
001011 0101
001011 0101
D20.3
011 10100
001011 1100
001011 0011
D21.2
010 10101
101010 0101
101010 0101
D21.3
011 10101
101010 1100
101010 0011
D22.2
010 10110
011010 0101
011010 0101
D22.3
011 10110
011010 1100
011010 0011
D23.2
010 10111
111010 0101
000101 0101
D23.3
011 10111
111010 0011
000101 1100
D24.2
010 11000
110011 0101
001100 0101
D24.3
011 11000
110011 0011
001100 1100
D25.2
010 11001
100110 0101
100110 0101
D25.3
011 11001
100110 1100
100110 0011
D26.2
010 11010
010110 0101
010110 0101
D26.3
011 11010
010110 1100
010110 0011
D27.2
010 11011
110110 0101
001001 0101
D27.3
011 11011
110110 0011
001001 1100
D28.2
010 11100
001110 0101
001110 0101
D28.3
011 11100
001110 1100
001110 0011
D29.2
010 11101
101110 0101
010001 0101
D29.3
011 11101
101110 0011
010001 1100
D30.2
010 11110
011110 0101
100001 0101
D30.3
011 11110
011110 0011
100001 1100
D31.2
010 11111
101011 0101
010100 0101
D31.3
011 11111
101011 0011
010100 1100
Document #: 38-02097 Rev. **
Page 38 of 43
CYV15G0404DXB
PRELIMINARY
Table 15. Valid Data Characters (TXCTx[0] = 0, RXSTx[2:0] = 000) (continued)
Data
Byte
Name
Bits
Current RD−
Current RD+
Bits
Current RD−
Current RD+
abcdei fghj
Data
Byte
Name
HGF EDCBA
abcdei fghj
HGF EDCBA
abcdei fghj
abcdei fghj
D0.4
100 00000
100111 0010
011000 1101
D0.5
101 00000
100111 1010
011000 1010
D1.4
100 00001
011101 0010
100010 1101
D1.5
101 00001
011101 1010
100010 1010
D2.4
100 00010
101101 0010
010010 1101
D2.5
101 00010
101101 1010
010010 1010
D3.4
100 00011
110001 1101
110001 0010
D3.5
101 00011
110001 1010
110001 1010
D4.4
100 00100
110101 0010
001010 1101
D4.5
101 00100
110101 1010
001010 1010
D5.4
100 00101
101001 1101
101001 0010
D5.5
101 00101
101001 1010
101001 1010
D6.4
100 00110
011001 1101
011001 0010
D6.5
101 00110
011001 1010
011001 1010
D7.4
100 00111
111000 1101
000111 0010
D7.5
101 00111
111000 1010
000111 1010
D8.4
100 01000
111001 0010
000110 1101
D8.5
101 01000
111001 1010
000110 1010
D9.4
100 01001
100101 1101
100101 0010
D9.5
101 01001
100101 1010
100101 1010
D10.4
100 01010
010101 1101
010101 0010
D10.5
101 01010
010101 1010
010101 1010
D11.4
100 01011
110100 1101
110100 0010
D11.5
101 01011
110100 1010
110100 1010
D12.4
100 01100
001101 1101
001101 0010
D12.5
101 01100
001101 1010
001101 1010
D13.4
100 01101
101100 1101
101100 0010
D13.5
101 01101
101100 1010
101100 1010
D14.4
100 01110
011100 1101
011100 0010
D14.5
101 01110
011100 1010
011100 1010
D15.4
100 01111
010111 0010
101000 1101
D15.5
101 01111
010111 1010
101000 1010
D16.4
100 10000
011011 0010
100100 1101
D16.5
101 10000
011011 1010
100100 1010
D17.4
100 10001
100011 1101
100011 0010
D17.5
101 10001
100011 1010
100011 1010
D18.4
100 10010
010011 1101
010011 0010
D18.5
101 10010
010011 1010
010011 1010
D19.4
100 10011
110010 1101
110010 0010
D19.5
101 10011
110010 1010
110010 1010
D20.4
100 10100
001011 1101
001011 0010
D20.5
101 10100
001011 1010
001011 1010
D21.4
100 10101
101010 1101
101010 0010
D21.5
101 10101
101010 1010
101010 1010
D22.4
100 10110
011010 1101
011010 0010
D22.5
101 10110
011010 1010
011010 1010
D23.4
100 10111
111010 0010
000101 1101
D23.5
101 10111
111010 1010
000101 1010
D24.4
100 11000
110011 0010
001100 1101
D24.5
101 11000
110011 1010
001100 1010
D25.4
100 11001
100110 1101
100110 0010
D25.5
101 11001
100110 1010
100110 1010
D26.4
100 11010
010110 1101
010110 0010
D26.5
101 11010
010110 1010
010110 1010
D27.4
100 11011
110110 0010
001001 1101
D27.5
101 11011
110110 1010
001001 1010
D28.4
100 11100
001110 1101
001110 0010
D28.5
101 11100
001110 1010
001110 1010
D29.4
100 11101
101110 0010
010001 1101
D29.5
101 11101
101110 1010
010001 1010
D30.4
100 11110
011110 0010
100001 1101
D30.5
101 11110
011110 1010
100001 1010
D31.4
100 11111
101011 0010
010100 1101
D31.5
101 11111
101011 1010
010100 1010
Document #: 38-02097 Rev. **
Page 39 of 43
CYV15G0404DXB
PRELIMINARY
Table 15. Valid Data Characters (TXCTx[0] = 0, RXSTx[2:0] = 000) (continued)
Data
Byte
Name
Bits
Current RD−
Current RD+
Bits
Current RD−
Current RD+
abcdei fghj
Data
Byte
Name
HGF EDCBA
abcdei fghj
HGF EDCBA
abcdei fghj
abcdei fghj
D0.6
110 00000
100111 0110
011000 0110
D0.7
111 00000
100111 0001
011000 1110
D1.6
110 00001
011101 0110
100010 0110
D1.7
111 00001
011101 0001
100010 1110
D2.6
110 00010
101101 0110
010010 0110
D2.7
111 00010
101101 0001
010010 1110
D3.6
110 00011
110001 0110
110001 0110
D3.7
111 00011
110001 1110
110001 0001
D4.6
110 00100
110101 0110
001010 0110
D4.7
111 00100
110101 0001
001010 1110
D5.6
110 00101
101001 0110
101001 0110
D5.7
111 00101
101001 1110
101001 0001
D6.6
110 00110
011001 0110
011001 0110
D6.7
111 00110
011001 1110
011001 0001
D7.6
110 00111
111000 0110
000111 0110
D7.7
111 00111
111000 1110
000111 0001
D8.6
110 01000
111001 0110
000110 0110
D8.7
111 01000
111001 0001
000110 1110
D9.6
110 01001
100101 0110
100101 0110
D9.7
111 01001
100101 1110
100101 0001
D10.6
110 01010
010101 0110
010101 0110
D10.7
111 01010
010101 1110
010101 0001
D11.6
110 01011
110100 0110
110100 0110
D11.7
111 01011
110100 1110
110100 1000
D12.6
110 01100
001101 0110
001101 0110
D12.7
111 01100
001101 1110
001101 0001
D13.6
110 01101
101100 0110
101100 0110
D13.7
111 01101
101100 1110
101100 1000
D14.6
110 01110
011100 0110
011100 0110
D14.7
111 01110
011100 1110
011100 1000
D15.6
110 01111
010111 0110
101000 0110
D15.7
111 01111
010111 0001
101000 1110
D16.6
110 10000
011011 0110
100100 0110
D16.7
111 10000
011011 0001
100100 1110
D17.6
110 10001
100011 0110
100011 0110
D17.7
111 10001
100011 0111
100011 0001
D18.6
110 10010
010011 0110
010011 0110
D18.7
111 10010
010011 0111
010011 0001
D19.6
110 10011
110010 0110
110010 0110
D19.7
111 10011
110010 1110
110010 0001
D20.6
110 10100
001011 0110
001011 0110
D20.7
111 10100
001011 0111
001011 0001
D21.6
110 10101
101010 0110
101010 0110
D21.7
111 10101
101010 1110
101010 0001
D22.6
110 10110
011010 0110
011010 0110
D22.7
111 10110
011010 1110
011010 0001
D23.6
110 10111
111010 0110
000101 0110
D23.7
111 10111
111010 0001
000101 1110
D24.6
110 11000
110011 0110
001100 0110
D24.7
111 11000
110011 0001
001100 1110
D25.6
110 11001
100110 0110
100110 0110
D25.7
111 11001
100110 1110
100110 0001
D26.6
110 11010
010110 0110
010110 0110
D26.7
111 11010
010110 1110
010110 0001
D27.6
110 11011
110110 0110
001001 0110
D27.7
111 11011
110110 0001
001001 1110
D28.6
110 11100
001110 0110
001110 0110
D28.7
111 11100
001110 1110
001110 0001
D29.6
110 11101
101110 0110
010001 0110
D29.7
111 11101
101110 0001
010001 1110
D30.6
110 11110
011110 0110
100001 0110
D30.7
111 11110
011110 0001
100001 1110
D31.6
110 11111
101011 0110
010100 0110
D31.7
111 11111
101011 0001
010100 1110
Document #: 38-02097 Rev. **
Page 40 of 43
PRELIMINARY
CYV15G0404DXB
Table 16. Valid Special Character Codes and Sequences (TXCTx = special character code or RXSTx[2:0] = 001)[37, 38]
S.C. Byte Name
Cypress
S.C. Code
Name
S.C. Byte Name
[39]
Bits
HGF EDCB
A
Alternate
S.C. Byte Name
[39]
Bits
HGF EDCBA
Current RD−
abcdei fghj
Current RD+
abcdei fghj
K28.0
C0.0
(C00)
000 00000
C28.0
(C1C)
000 11100
001111 0100
110000 1011
K28.1[40]
C1.0
(C01)
000 00001
C28.1
(C3C)
001 11100
001111 1001
110000 0110
K28.2
C2.0
(C02)
000 00010
C28.2
(C5C)
010 11100
001111 0101
110000 1010
K28.3
C3.0
(C03)
000 00011
C28.3
(C7C)
011 11100
001111 0011
110000 1100
K28.4[40]
C4.0
(C04)
000 00100
C28.4
(C9C)
100 11100
001111 0010
110000 1101
K28.5[40, 41]
C5.0
(C05)
000 00101
C28.5
(CBC)
101 11100
001111 1010
110000 0101
K28.6[40]
C6.0
(C06)
000 00110
C28.6
(CDC)
110 11100
001111 0110
110000 1001
K28.7[40, 42]
C7.0
(C07)
000 00111
C28.7
(CFC)
111 11100
001111 1000
110000 0111
K23.7
C8.0
(C08)
000 01000
C23.7
(CF7)
111 10111
111010 1000
000101 0111
K27.7
C9.0
(C09)
000 01001
C27.7
(CFB)
111 11011
110110 1000
001001 0111
K29.7
C10.0
(C0A)
000 01010
C29.7
(CFD)
111 11101
101110 1000
010001 0111
K30.7
C11.0
(C0B)
000 01011
C30.7
(CFE)
111 11110
011110 1000
100001 0111
(C22)
001 00010
C2.1
(C22)
001 00010
−
K28.5,Dn.xxx0[43]
+K28.5,Dn.xxx1[4
[40]
End of Frame Sequence
EOFxx
C2.1
3]
Code Rule Violation and SVS Tx Pattern
Exception[42, 44]
C0.7
(CE0)
111 00000
C0.7
(CE0)
111 00000
100111 1000
011000 0111
−K28.5[45]
C1.7
(CE1)
111 00001
C1.7
(CE1)
111 00001
001111 1010
001111 1010
+K28.5[46]
C2.7
(CE2)
111 00010
C2.7
(CE2)
111 00010
110000 0101
110000 0101
C4.7
(CE4)
111 00100
110111 0101
001000 1010
Running Disparity Violation Pattern
Exception[47]
C4.7
(CE4)
111 00100
Notes:
37. All codes not shown are reserved.
38. Notation for Special Character Code Name is consistent with Fibre Channel and ESCON naming conventions. Special Character Code Name is intended to
describe binary information present on I/O pins. Common usage for the name can either be in the form used for describing Data patterns (i.e., C0.0 through
C31.7), or in hex notation (i.e., Cnn where nn = the specified value between 00 and FF).
39. Both the Cypress and alternate encodings may be used for data transmission to generate specific Special Character Codes. The decoding process for received
characters generates Cypress codes or Alternate codes as selected by the BOE[7:0] configuration inputs.
40. These characters are used for control of ESCON interfaces. They can be sent as embedded commands or other markers when not operating using ESCON
protocols.
41. The K28.5 character is used for framing operations by the receiver. It is also the pad or fill character transmitted to maintain the serial link when no user data
is available.
42. Care must be taken when using this Special Character code. When a C7.0 or a C0.7 is followed by a D11.x or D20.x, an alias K28.5 sync character is created.
These sequences can cause erroneous framing and should be avoided while RFENx = 1.
43. C2.1 = Transmit either –K28.5+ or +K28.5– as determined by Current RD and modify the Transmission Character that follows, by setting its least significant
bit to 1 or 0. If Current RD at the start of the following character is plus (+) the LSB is set to 0, and if Current RD is minus (–) the LSB becomes 1. This
modification allows construction of X3.230 “EOF” frame delimiters wherein the second data byte is determined by the Current RD.
For example, to send “EOFdt” the controller could issue the sequence C2.1–D21.4– D21.4–D21.4, and the HOTLink Transmitter sends either
K28.5–D21.4–D21.4–D21.4 or K28.5–D21.5– D21.4–D21.4 based on Current RD. Likewise to send “EOFdti” the controller could issue the sequence
C2.1–D10.4–D21.4–D21.4, and the HOTLink Transmitter sends either K28.5–D10.4–D21.4– D21.4 or K28.5–D10.5–D21.4–D21.4 based on Current RD.
The receiver never outputs this Special Character, since K28.5 is decoded as C5.0, C1.7, or C2.7, and the subsequent bytes are decoded as data.
44. C0.7 = Transmit a deliberate code rule violation. The code chosen for this function follows the normal Running Disparity rules. Transmission of this Special
Character has the same effect as asserting TXSVS = HIGH. The receiver only outputs this Special Character if the Transmission Character being decoded is
not found in the tables.
45. C1.7 = Transmit Negative K28.5 (−K28.5+) disregarding Current RD. The receiver only outputs this Special Character if K28.5 is received with the wrong
running disparity. The receiver outputs C1.7 if −K28.5 is received with RD+, otherwise K28.5 is decoded as C5.0 or C2.7.
46. C2.7 = Transmit Positive K28.5 (+K28.5−) disregarding Current RD. The receiver only outputs this Special Character if K28.5 is received with the wrong running
disparity. The receiver outputs C2.7 if +K28.5 is received with RD−, otherwise K28.5 is decoded as C5.0 or C1.7.
47. C4.7 = Transmit a deliberate code rule violation to indicate a Running Disparity violation. The receiver only outputs this Special Character if the Transmission
Character being decoded is found in the tables, but Running Disparity does not match. This might indicate that an error occurred in a prior byte.
Document #: 38-02097 Rev. **
Page 41 of 43
CYV15G0404DXB
PRELIMINARY
Ordering Information
Speed
Ordering Code
Package Name
Package Type
Operating
Range
Standard
CYV15G0404DXB-BGC
BL256
256-Ball Thermally Enhanced Ball Grid Array
Commercial
Standard
CYV15G0404DXB-BGI
BL256
256-Ball Thermally Enhanced Ball Grid Array
Industrial
Package Diagram
256-Lead L2 Ball Grid Array (27 x 27 x 1.57 mm) BL256
51-85123-*E
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.
Document #: 38-02097 Rev. **
Page 42 of 43
© Cypress Semiconductor Corporation, 2004. The information contained herein is subject to change without notice. Cypress Semiconductor Corporation assumes no responsibility for the use
of any circuitry other than circuitry embodied in a Cypress product. Nor does it convey or imply any license under patent or other rights. Cypress products are not warranted nor intended to be
used for medical, life support, life saving, critical control or safety applications, unless pursuant to an express written agreement with Cypress. Furthermore, Cypress does not authorize its
products for use as critical components in life-support systems where a malfunction or failure may reasonably be expected to result in significant injury to the user. The inclusion of Cypress
products in life-support systems application implies that the manufacturer assumes all risk of such use and in doing so indemnifies Cypress against all charges.
CYV15G0404DXB
PRELIMINARY
Document History Page
Document Title: CYV15G0404DXB Independent Clock Quad HOTLink II™ Transceiver
Document Number: 38-02097
REV.
ECN NO.
ISSUE
DATE
ORIG. OF
CHANGE
**
231494
SEE ECN
BCD
Document #: 38-02097 Rev. **
DESCRIPTION OF CHANGE
New Data Sheet
Page 43 of 43