CYPRESS CYP15G0101DXB-BBXI

CYP15G0101DXB
CYV15G0101DXB
Single-channel HOTLink II™ Transceiver
Single-channel HOTLink II™ Transceiver
Features
■
Second-generation HOTLink® technology
■
Compliant to multiple standards
®
❐ ESCON , DVB-ASI, fibre channel and gigabit ethernet
(IEEE802.3z)
❐ CPRI™ compliant
❐ CYV15G0101DXB compliant to SMPTE 259M and SMPTE
292M
❐ 8B/10B encoded or 10-bit unencoded data
■
Single-channel transceiver operates from 195 to 1500 MBaud
serial data rate
■
Selectable parity check/generate
■
Selectable input clocking options
■
Selectable output clocking options
■
MultiFrame™ Receive Framer
❐ Bit and byte alignment
❐ Comma or full K28.5 detect
❐ Single- or multi-byte framer for byte alignment
❐ Low-latency option
Synchronous LVTTL parallel input and parallel output interface
■
Internal phase-locked loops (PLLs) with no external PLL
components
Dual differential PECL-compatible serial inputs
❐ Internal DC-restoration
■
Dual differential PECL-compatible serial outputs
❐ Source matched for driving 50 Ω transmission lines
❐ No external bias resistors required
❐ Signaling-rate controlled edge-rates
■
Optional elasticity buffer in receive path
■
Optional phase align buffer in transmit path
Compatible with
❐ Fiber-optic modules
❐ Copper cables
❐ Circuit board traces
■
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 1.25 W at 3.3 V typical
■
Single 3.3 V supply
■
100-ball BGA
■
Pb-free package option available
■
0.25 µ BiCMOS technology
Functional Description
■
■
■
The CYP15G0101DXB[1] single-channel HOTLink II™
transceiver is a point-to-point communications building block
allowing the transfer of data over a high-speed serial link (optical
fiber, balanced, and unbalanced copper transmission lines) at
signaling speeds ranging from 195 to 1500 MBaud.
The transmit channel accepts parallel characters in an input
register, encodes each character for transport, and converts it to
serial data. The receive channel accepts serial data and converts
it to parallel data, frames the data to character boundaries,
decodes the framed characters into data and special characters,
and presents these characters to an output register. Figure 1
illustrates typical connections between independent host
systems and corresponding CYP(V)15G0101DXB parts. As a
second-generation HOTLink device, the CYP(V)15G0101DXB
extends the HOTLink II family with enhanced levels of
integration and faster data rates, while maintaining serial-link
compatibility (data, command, and BIST) with other HOTLink
devices.
Serial Link
Backplane or Cabled
Connections
10
10
System Host
10
CYP(V)15G0101DXB
10
CYP(V)15G0101DXB
System Host
Figure 1. HOTLink II System Connections
Note
1. CYV15G0101DXB refers to SMPTE 259M and SMPTE 292M compliant devices. CYP15G0101DXB refers to devices not compliant to SMPTE 259M and SMPTE
292M pathological test requirements. CYP(V)15G0101DXB refers both devices.
Cypress Semiconductor Corporation
Document Number: 38-02031 Rev. *M
•
198 Champion Court
•
San Jose, CA 95134-1709
•
408-943-2600
Revised April 29, 2011
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
The CYV15G0101DXB satisfies the SMPTE 259M and SMPTE
292M compliance as per the EG34-1999 pathological test
requirements. The
transmit
(TX)
section
of
the
CYP(V)15G0101DXB single-channel HOTLink II consists of a
byte-wide channel. The channel can accept either eight-bit data
characters or pre-encoded 10-bit transmission characters. Data
characters are passed from the transmit input register to an
embedded 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.
HOTLink II devices are ideal for a variety of applications where
parallel interfaces can be replaced with high-speed,
point-to-point serial links. Some applications include
interconnecting backplanes on switches, routers, base-stations,
servers and video transmission systems.
The CYV15G0101DXB is verified by testing to be compliant to
all the pathological test patterns documented in SMPTE
EG34-1999, for both the SMPTE 259M and 292M signaling
rates. The tests ensure that the receiver recovers data with no
errors for the following patterns:
1. Repetitions of 20 ones and 20 zeros.
2. Single burst of 44 ones or 44 zeros.
3. Repetitions of 19 ones followed by 1 zero or 19 zeros followed
by 1 one.
RXD[7:0]
RXST[2:0]
Transceiver Logic Block Diagram
The transmit and the receive channels contain BIST pattern
generators and checkers, respectively. This BIST hardware
allows at-speed testing of the high-speed serial data paths in
both transmit and receive sections, as well as across the
interconnecting links.
TXD[7:0]
TXCT[1:0]
The receive (RX) section of the CYP(V)15G0101DXB
single-channel HOTLink II consists of a byte-wide channel. The
channel accepts a serial bit-stream from one of two
PECL-compatible differential line receivers and, using a
completely integrated PLL clock synchronizer, recovers the
timing information necessary for data reconstruction. The
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 interfaces
from one or multiple sources, the receive interface may be
configured to present data relative to a recovered clock or to a
local reference clock.
x10
x11
Phase
Align
Buffer
Elasticity
Buffer
Encoder
8B/10B
Decoder
8B/10B
Framer
Document Number: 38-02031 Rev. *M
TX
RX
IN1±
IN2±
Deserializer
OUT1±
OUT2±
Serializer
Page 2 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
= Internal Signal
Logic Block Diagram
TRSTZ
REFCLK+
REFCLK–
TXRATE
Character-Rate Clock
Transmit PLL
Clock Multiplier
Bit-Rate Clock
SPDSEL
Character-Rate Clock
TXCLKO+
TXCLKO–
2
TXMODE[1:0]
Transmit
Mode
TXCKSEL
12
12
10
OUT2+
OUT2–
TXLB
H M L
TXCLK
TXRST
2
4
Output
Enable
Latch
PARCTL
BOE[1:0]
RXLE
OUT1+
OUT1–
Shifter
2
12
BIST LFSR
8B/10B
TXOP
TXCT[1:0]
8
Parity
Check
TXD[7:0]
Input
Register
SCSEL
Phase-Align
Buffer
TXPER
OELE
BIST Enable
Latch
RX PLL Enable
Latch
BISTLE
Character-Rate Clock
SDASEL
FRAMCHAR
RFEN
RFMODE
Clock
Select
Output
Register
Elasticity
Buffer
10B/8B
BIST
Clock &
Data
Recovery
PLL
LFI
Framer
IN1+
IN1–
IN2+
IN2–
TXLB
Receive
Signal
Monitor
Shifter
LPEN
INSEL
8
3
RXD[7:0]
RXOP
RXST[2:0]
RXCLK+
RXCLK–
÷2
Delay
DECMODE
RXRATE
RXCLKC+
RXMODE
RXCKSEL
JTAG
Boundary
Scan
Controller
Document Number: 38-02031 Rev. *M
TMS
TCLK
TDI
TDO
Page 3 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Contents
Pin Configuration ............................................................ 5
Pin Descriptions ............................................................... 6
CYP(V)15G0101DXB HOTLink II Operation .................. 12
CYP(V)15G0101DXB Transmit Data Path ................ 12
Transmit Modes ......................................................... 13
Transmit BIST ........................................................... 15
Serial Output Drivers ................................................. 16
Transmit PLL Clock Multiplier .................................... 16
CYP(V)15G0101DXB Receive Data Path ....................... 16
Serial Line Receivers ................................................ 16
Signal Detect/Link Fault ............................................ 17
Clock/Data Recovery ................................................. 18
Deserializer/Framer ................................................... 18
10B/8B Decoder Block .............................................. 19
Receive BIST Operation ............................................ 19
Receive Elasticity Buffer ............................................ 20
Receive Modes .......................................................... 20
Power Control ............................................................ 20
Output Bus ................................................................ 21
Parity Generation ...................................................... 22
JTAG Support ............................................................ 23
Maximum Ratings ........................................................... 25
Power-up Requirements ............................................ 25
Document Number: 38-02031 Rev. *M
DC Electrical Characteristics ........................................ 25
AC Test Loads and Waveforms ..................................... 26
CYP(V)15G0101DXB AC Characteristics ...................... 27
Switching Waveforms for the HOTLink II Transmitter 29
X3.230 Codes and Notation Conventions .................... 33
Notation Conventions ................................................ 33
8B/10B Transmission Code ....................................... 33
Transmission Order ................................................... 33
Valid and Invalid Transmission Characters ............... 33
Use of the Tables for Generating
Transmission Characters .......................................... 34
Use of the Tables for Checking the Validity
of Received Transmission Characters ...................... 34
Ordering Information ...................................................... 40
Ordering Code Definitions ......................................... 40
Package Diagram ............................................................ 41
Acronyms ........................................................................ 42
Document Conventions ................................................ 42
Document History Page ................................................. 43
Sales, Solutions, and Legal Information ...................... 44
Worldwide Sales and Design Support ....................... 44
Products .................................................................... 44
PSoC Solutions ......................................................... 44
Page 4 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Pin Configuration
Top View
1
2
3
4
A
VCC
IN2+
VCC
OUT2–
B
VCC
IN2–
TDO
OUT2+
7
8
9
10
RXMODE TXMODE[1]
IN1+
VCC
OUT1–
VCC
TXRATE TXMODE[0]
IN1–
#NC[2]
OUT1+
VCC
RFEN
LPEN
RXLE
SDASEL
SPDSEL
PARCTL
RFMODE
INSEL
BOE[0]
BOE[1]
FRAMCHAR
GND
GND
GND
GND
TMS
TRSTZ
TDI
BISTLE
DECMODE
OELE
GND
GND
GND
GND
TCLK
RXCKSEL TXCKSEL
RXST[2]
RXST[1]
RXST[0]
GND
GND
GND
GND
TXPER
REFCLK– REFCLK+
RXOP
RXD[1]
RXD[5]
GND
GND
GND
GND
TXOP
TXCLKO+ TXCLKO–
RXD[0]
RXD[2]
RXD[6]
LFI
TXCT[1]
TXD[6]
TXD[3]
TXCLK
TXRST
#NC[2]
J
VCC
RXD[3]
RXD[7]
RXCLK–
TXCT[0]
TXD[5]
TXD[2]
TXD[0]
#NC[2]
VCC
K
VCC
RXD[4]
VCC
RXCLK+
TXD[7]
TXD[4]
TXD[1]
VCC
SCSEL
VCC
4
3
2
1
C
D
E
F
G
H
5
RXCLKC+ RXRATE
6
Bottom View
10
9
8
7
6
5
VCC
OUT1–
VCC
IN1+
TXMODE[1] RXMODE
OUT2–
VCC
IN2+
VCC
A
VCC
OUT1+
#NC[2]
IN1–
TXMODE[0] TXRATE
OUT2+
TDO
IN2–
VCC
B
INSEL
RFMODE
PARCTL
SPDSEL
SDASEL
RXLE
LPEN
RFEN
TDI
TRSTZ
TMS
GND
GND
GND
GND
FRAMCHAR
BOE[1]
BOE[0]
TXCKSEL RXCKSEL
TCLK
GND
GND
GND
GND
OELE
DECMODE
BISTLE
REFCLK+ REFCLK–
TXPER
GND
GND
GND
GND
RXST[0]
RXST[1]
RXST[2]
TXCLKO– TXCLKO+
TXOP
GND
GND
GND
GND
RXD[5]
RXD[1]
RXOP
RXRATE RXCLKC+
C
D
E
F
G
#NC[2]
TXRST
TXCLK
TXD[3]
TXD[6]
TXCT[1]
LFI
RXD[6]
RXD[2]
RXD[0]
VCC
#NC[2]
TXD[0]
TXD[2]
TXD[5]
TXCT[0]
RXCLK–
RXD[7]
RXD[3]
VCC
J
VCC
SCSEL
VCC
TXD[1]
TXD[4]
TXD[7]
RXCLK+
VCC
RXD[4]
VCC
K
H
Note
2. #NC = Do Not Connect.
Document Number: 38-02031 Rev. *M
Page 5 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Pin Descriptions
CYP(V)15G0101DXB single-channel HOTLink II
Pin Name
I/O Characteristics Signal Description
Transmit Path Data Signals
TXPER
LVTTL output,
changes relative to
REFCLK↑[3]
Transmit path parity error. Active HIGH. Asserted (HIGH) if parity checking is enabled
(PARCTL ≠ LOW) and a parity error is detected at the encoder. This output is HIGH for one
transmit character-clock period to indicate detection of a parity error in the character
presented to the encoder.
If a parity error is detected, the character in error is replaced with a C0.7 character to force a
corresponding bad-character detection at the remote end of the link. This replacement takes
place regardless of the encoded/un-encoded state of the interface.
When BIST is enabled for the specific transmit channel, BIST progress is presented on this
output. Once every 511 character times (plus a 16-character Word Sync Sequence when the
receive channel is clocked by REFCLK, i.e., RXCKSEL = LOW), the TXPER signal pulses
HIGH for one transmit-character clock period (if RXCKSEL = MID) or seventeen
transmit-character clock periods (if RXCKSEL = LOW or HIGH) to indicate a complete pass
through the BIST sequence. For RXCKSEL = LOW or HIGH, If TXMODE[1:0] = LL, then no
Word Sync Sequence is sent in BIST, and TXPER pulses HIGH for one transmit-character
clock period.
This output also provides an indication of a phase-align buffer underflow/overflow condition.
When the phase-align buffer is enabled (TXCKSEL ≠ LOW, or TXCKSEL = LOW and
TXRATE = HIGH), and an underflow/overflow condition is detected, TXPER is asserted and
remains asserted until either an atomic Word Sync Sequence is transmitted or TXRST is
sampled LOW to recenter the phase-align buffer.
TXCT[1:0]
LVTTL input,
synchronous,
sampled by
TXCLK↑ or
REFCLK↑[3]
Transmit control. These inputs are captured on the rising edge of the transmit interface clock
as selected by TXCKSEL, and are passed to the encoder or transmit shifter. They identify
how the TXD[7:0] characters are interpreted. When the encoder is enabled, these inputs
determine if the TXD[7:0] character is encoded as data, a special character code, a K28.5 fill
character or a Word Sync Sequence. When the encoder is bypassed, these inputs are
interpreted as data bits. See Table 1 for details.
TXD[7:0]
LVTTL input,
synchronous,
sampled by
TXCLK↑ or
REFCLK↑[3]
Transmit data inputs. These inputs are captured on the rising edge of the transmit interface
clock as selected by TXCKSEL, and passed to the encoder or transmit shifter.
TXOP
LVTTL input,
synchronous,
internal pull-up,
sampled by
TXCLK↑ or
REFCLK↑[3]
Transmit path odd parity. When parity checking is enabled (PARCTL ≠ LOW), the parity
captured at this input is XORed with the data on the TXD bus (and sometimes TXCT[1:0]) to
verify the integrity of the captured character. See Table 2 for details.
SCSEL
LVTTL input,
synchronous,
internal pull-down,
sampled by
TXCLK↑ or
REFCLK↑[3]
Special character select. Used in some transmit modes along with TXCTx[1:0] to encode
special characters or to initiate a Word Sync Sequence. When the transmit path is configured
to select TXCLK to clock the input register (TXCKSEL = MID or HIGH), SCSEL is captured
relative to TXCLK↑.
When the encoder is enabled (TXMODE[1] ≠ LOW), TXD[7:0] specify the specific data or
command character to be sent. When the encoder is bypassed, these inputs are interpreted
as data bits of the 10-bit input character. See Table 1 for details.
Note
3. When REFCLK is configured for half-rate operation (TXRATE
REFCLK.
Document Number: 38-02031 Rev. *M
= HIGH), this input is sampled (or the outputs change) relative to both the rising and falling edges of
Page 6 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Pin Descriptions
CYP(V)15G0101DXB single-channel HOTLink II (continued)
Pin Name
I/O Characteristics Signal Description
TXRST
LVTTL input,
asynchronous,
internal pull-up,
sampled by
REFCLK↑[4]
Transmit clock phase reset. Active LOW. When sampled LOW, the transmit phase-align
buffer is allowed to adjust its data-transfer timing (relative to the selected input clock) to allow
clean transfer of data from the input register to the encoder or transmit shifter. When TXRST
is sampled HIGH, the internal phase relationship between the TXCLK and the internal
character-rate clock is fixed and the device operates normally.
When configured for half-rate REFCLK sampling of the transmit character stream (TXCKSEL
= LOW and TXRATE = HIGH), assertion of TXRST is only used to clear phase-align buffer
faults caused by highly asymmetric reference clock periods or reference clocks with excessive
cycle-to-cycle jitter. During this alignment period, one or more characters may be added to or
lost from all the associated transmit paths as the transmit phase-align buffers are adjusted.
TXRST must be sampled LOW by a minimum of two consecutive rising edges of REFCLK to
ensure the reset operation is initiated correctly on all channels. This input is ignored when
both TXCKSEL and TXRATE are LOW, since the phase align buffer is bypassed. In all other
configurations, TXRST should be asserted during device initialization to ensure proper
operation of the phase-align buffer. TXRST should be asserted after the assertion and
deassertion of TRSTZ, after the presence of a valid TXCLK and after allowing enough time
for the TXPLL to lock to the reference clock (as specified by parameter tTXLOCK).
Transmit Path Clock and Clock Control
TXCKSEL
3-level select static
control input[5]
Transmit clock select. Selects the clock source used to write data into the transmit input
register of the transmit channel. When LOW, the input register is clocked by REFCLK↑.[4]
When HIGH or MID, TXCLK↑ is the input register clock for TXD[7:0] and TXCT[1:0].
When TXRATE=HIGH, configuring TXCKSEL = HIGH or MID is an invalid mode of operation.
TXCLKO±
LVTTL output
Transmit clock output. This true and complement output clock is synthesized by the transmit
PLL and is synchronous to the internal transmit character clock. It has the same frequency
as REFCLK (when TXRATE = LOW), or twice the frequency of REFCLK (when
TXRATE = HIGH). This output clock has no direct phase relationship to REFCLK.
TXRATE
LVTTL input,
static control input,
internal pull-down
Transmit PLL clock rate select. When TXRATE = HIGH, the transmit PLL multiplies
REFCLK by 20 to generate the serial bit-rate clock.
When TXRATE = LOW, the transmit PLL multiplies REFCLK by 10 to generate the serial
bit-rate clock. See Table 9 for a list of operating serial rates.
When REFCLK is selected to clock the receive parallel interfaces (RXCKSEL = LOW), the
TXRATE input also determines if the clocks on the RXCLK± and RXCLKC+ outputs are full
or half-rate. When TXRATE = HIGH (REFCLK is half-rate), the RXCLK± and RXCLKC+ output
clocks are also half-rate clocks and follow the frequency and duty cycle of the REFCLK input.
When TXRATE = LOW (REFCLK is full-rate), the RXCLK± and RXCLKC+ output clocks are
also full-rate clocks and follow the frequency and duty cycle of the REFCLK input.
When TXRATE=HIGH, configuring TXCKSEL = HIGH or MID is an invalid mode of operation.
TXCLK
LVTTL clock input,
internal pull-down
Transmit path input clock. This clock must be frequency-coherent to TXCLKO±, but may
be offset in phase. The internal operating phase of the input clock (relative to REFLCK or
TXCLKO+) is adjusted when TXRST = LOW and locked when TXRST = HIGH.
Transmit Path Mode Control
TXMODE[1:0] 3-level select[5]
static control inputs
Transmit operating mode. These inputs are interpreted to select one of nine operating
modes of the transmit path. See Table 3 for a list of operating modes.
Notes
4. When REFCLK is configured for half-rate operation (TXRATE = HIGH), this input is sampled (or the outputs change) relative to both the rising and falling edges of
REFCLK.
5. 3-level select inputs are used for static configuration. They are ternary (not binary) inputs that make use of non-standard 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). When not
connected or allowed to float, a 3-level select input will self-bias to the MID level.
Document Number: 38-02031 Rev. *M
Page 7 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Pin Descriptions
CYP(V)15G0101DXB single-channel HOTLink II (continued)
Pin Name
I/O Characteristics Signal Description
Receive Path Data Signals
RXD[7:0]
RXST[2:0]
RXOP
LVTTL output,
synchronous to the
RXCLK↑ output (or
REFCLK↑ input[6]
when
RXCKSEL = LOW)
Parallel data output. These outputs change following the rising edge of the selected receive
interface clock.
When the decoder is enabled (DECMODE = HIGH or MID), these outputs represent either
received data or a special character. The status of the received data is represented by the
values of RXST[2:0].
When the decoder is bypassed (DECMODE = LOW), RXD[7:0] become the higher order bits
of the 10-bit received character. See Table 13 for details.
LVTTL output,
synchronous to the
RXCLK↑ output (or
REFCLK↑ input[6]
when
RXCKSEL = LOW)
Parallel status output. These outputs change following the rising edge of the selected
receive interface clock.
3-state,
LVTTL output,
synchronous to the
RXCLK↑ output (or
REFCLK↑ input[6]
when
RXCKSEL = LOW)
Receive path odd parity. When parity generation is enabled (PARCTL ≠ LOW), the parity
output is valid for the data on the RXD bus bits.
When the decoder is bypassed (DECMODE = LOW), RXST[1:0] become the two low-order
bits of the 10-bit received character, while RXST[2] = HIGH indicates the presence of a
comma character in the output register.
When the decoder is enabled (DECMODE = HIGH or MID), RXST[2:0] provide status of the
received signal. See Table 16 for a list of receive character status.
When parity generation is disabled (PARCTL = LOW), this output driver is disabled (high Z).
Receive Path Clock and Clock Control
RXCLK±
3-state,
LVTTL output clock
Receive character clock output. When configured such that the output data path is clocked
by the recovered clock (RXCKSEL = MID), these true and complement clocks are the receive
interface clocks which are used to control timing of output data (RXD[7:0], RXST[2:0] and
RXOP). This clock is 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
RXRATE.
When configured such that the output data path is clocked by REFCLK instead of recovered
clock (RXCKSEL = LOW), the RXCLK± output drivers present a buffered and delayed form
of REFCLK. In this mode, RXCLK± and RXCLKC+ are buffered forms of REFCLK that are
slightly different in phase, but follow the frequency and duty cycle of REFCLK. This phase
difference allows the user to select the optimal set-up/hold timing for their specific interface.
RXCLKC+
3-state, LVTTL
output
Delayed REFCLK+ when RXCKSEL = LOW. Delayed form of REFCLK+, used for transfer
of output data to a host system. This output is only enabled when the receive parallel interface
is configured to present data relative to REFCLK (RXCKSEL = LOW). When RXCKSEL =
LOW, the RXCLKC+ follows the frequency and duty cycle of REFCLK+.
RXRATE
LVTTL input
static control input,
internal pull-down
Receive clock rate select. When LOW, the RXCLK± recovered clock outputs are
complementary clocks operating at the recovered character rate. Data for the receive channel
should be latched on either the rising edge of RXCLK+ or falling edge of RXCLK–.
When HIGH, the RXCLK± recovered clock outputs are complementary clocks operating at
half the character rate. Data for the receive channel should be latched alternately on the rising
edge of RXCLK+ and RXCLK–.
When the output register is operated with REFCLK clocking (RXCKSEL = LOW), RXRATE is
not interpreted and RXCLK± follows the frequency and duty cycle of REFCLK.
RFEN
LVTTL input,
asynchronous,
internal pull-down
Reframe enable. Active HIGH. When HIGH, the Framer in the receive channel is enabled to
frame per the presently enabled framing mode and selected framing character.
Note
6. When REFCLK is configured for half-rate operation (TXRATE
REFCLK.
Document Number: 38-02031 Rev. *M
= HIGH), this input is sampled (or the outputs change) relative to both the rising and falling edges of
Page 8 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Pin Descriptions
CYP(V)15G0101DXB single-channel HOTLink II (continued)
Pin Name
I/O Characteristics Signal Description
RXMODE
3-level select[7]
static control input
Receive operating mode. This input selects one of two RXST channel status reporting
modes and is only interpreted when the decoder is enabled (DECMODE ≠ LOW). See
Table 19 for details.
FRAMCHAR
3-level select[7]
static control input
Framing character select. Used to select the character or portion of a character used for
character framing of the received data streams.
When MID, the framer looks for both positive and negative disparity versions of the eight-bit
comma character.
When HIGH, the framer looks for both positive and negative disparity versions of the K28.5
character.
Configuring FRAMCHAR = LOW is reserved for component test.
RFMODE
3-level select static
control input[7]
Reframe mode select. Used to select the type of character framing used to adjust the
character boundaries (based on detection of one or more framing characters in the data
stream. This signal operates in conjunction with the type of framing character selected.
When LOW, the low-latency framer is selected. This will frame on each occurrence of the
selected framing character(s) in the received data stream. This mode of framing stretches the
recovered character-rate clock for one or multiple cycles to align that clock with the recovered
data.
When MID, the Cypress-mode multi-byte parallel framer is selected. This requires a pair of
the selected framing character(s), on identical 10-bit boundaries, within a span of 50 bits (five
characters), before the character boundaries are adjusted. The recovered character clock
remains in the same phase regardless of character offset.
When HIGH, the alternate-mode multi-byte parallel framer is selected. This requires detection
of the selected framing character(s) in the received data stream, on identical 10-bit
boundaries, on four directly adjacent characters. The recovered character clock remains in
the same phase regardless of character offset.
PARCTL
3-level select static
control input[7]
Parity check/Generate control. Used to control the parity check and generate functions.
When LOW, parity checking is disabled, and the RXOP output is disabled (high Z).
When MID, and the 8B/10B encoder and decoder are enabled (TXMODE[1] ≠ LOW,
DECMODE ≠ LOW), TXD[7:0] inputs are checked (along with TXOP) for valid ODD parity,
and ODD parity is generated for the RXD[7:0] outputs and presented on RXOP. When the
8B/10B encoder and decoder are disabled (TXMODE[1] = LOW, DECMODE = LOW), the
TXD[7:0] and TXCT[1:0] inputs are checked (along with TXOP) for valid ODD parity, and ODD
parity is generated for the RXD[7:0] and RXST[1:0] outputs and presented on RXOP.
When HIGH, parity generation and checking are enabled. The TXD[7:0] and TXCT[1:0] inputs
are checked (along with TXOP) for valid ODD parity, and ODD parity is generated for the
RXD[7:0] and RXST[2:0] outputs and presented on RXOP.
See Table 2 and Table 15 for details.
DECMODE
3-level select static
control input[7]
Decoder mode select. When LOW, the decoder is bypassed and raw 10-bit characters are
passed to the output register. When the decoder is bypassed, RXCKSEL must be MID.
When MID, the Cypress Decoder table for special code characters is used. When HIGH, the
alternate Decoder table for special code characters is used. See Table 21 for a list of the
special codes supported in both encoded modes.
RXCKSEL
3-level select[7]
static control input
Receive clock mode. Selects the receive clock source used to transfer data to the output
registers and configures the elasticity buffer in the receive path.
When LOW, the output register is clocked by REFCLK. RXCLK± and RXCLKC+ present
buffered and delayed forms of REFCLK.
When MID, the RXCLK± output follows the recovered clock as selected by RXRATE and the
elasticity buffer is bypassed. When the 10B/8B decoder and elasticity buffer are bypassed
(DECMODE=LOW), RXCKSEL must be MID.
Configuring RXCKSEL = HIGH is an invalid mode of operation.
Note
7. 3-level select inputs are used for static configuration. They are ternary (not binary) inputs that make use of non-standard 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). When not
connected or allowed to float, a 3-level select input will self-bias to the MID level.
Document Number: 38-02031 Rev. *M
Page 9 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Pin Descriptions
CYP(V)15G0101DXB single-channel HOTLink II (continued)
Pin Name
I/O Characteristics Signal Description
Device Control Signals
SPDSEL
3-level select,[8]
static control input
REFCLK±
Differential LVPECL Reference clock. This clock input is used as the timing reference for the transmit PLL. It is
or single-ended
also used as the centering frequency of the range controller block of the receive CDR PLLs.
LVTTL input clock
This input clock may also be selected to clock the transmit and receive parallel interfaces.
Serial rate select. This input specifies the operating bit-rate range of both transmit and
receive PLLs. LOW = 195–400 MBaud, MID = 400–800 MBaud, HIGH = 800–1500 MBaud.
When SPDSEL=LOW, setting TXRATE=HIGH (half-rate reference clock) is invalid.
When driven by a single-ended LVCMOS or LVTTL clock source, the clock source may be
connected to either the true or complement REFCLK input, with the alternate REFCLK input
left open (floating). When driven by an LVPECL clock source, the clock must be a differential
clock, using both inputs. When TXCKSEL = LOW, REFCLK is also used as the clock for the
parallel transmit data (input) interface. When RXCKSEL = LOW and decoder is enabled, the
elasticity buffer is enabled and REFCLK is used as the clock source for the parallel receive
data (output) interface.
If the elasticity buffer is used, framing characters will be inserted or deleted to/from the data
stream to compensate for frequency differences between the reference clock and recovered
clock. When addition happens, a K28.5 will be appended immediately after a framing
character is detected in the elasticity buffer. When deletion happens, a framing character will
be removed from the data stream when detected in the elasticity buffer.
TRSTZ
LVTTL input,
internal pull-up
Device reset. Active LOW. Initializes all state machines and counters in the device.
When sampled LOW by the rising edge of REFLCK, this input resets the internal state
machines and sets the elasticity buffer pointers to a nominal offset. When the reset is removed
(TRSTZ sampled HIGH by REFCLK↑), the status and data outputs will become deterministic
in less than 16 REFCLK cycles. The BISTLE, OELE, and RXLE latches are reset by TRSTZ.
If the elasticity buffer or the phase-align buffer are used, TRSTZ should be applied after
power-up to initialize the internal pointers into these memory arrays.
Analog I/O and Control
OUT1±
CML differential
output
Primary differential serial data outputs. These PECL-compatible CML outputs (+3.3 V
referenced) are capable of driving terminated transmission lines or standard fiber-optic transmitter modules.
OUT2±
CML differential
output
Secondary differential serial data outputs. These PECL-compatible CML outputs (+3.3 V
referenced) are capable of driving terminated transmission lines or standard fiber-optic transmitter modules.
IN1±
LVPECL differential
Input, with internal
DC restoration
Primary differential serial data inputs. These inputs accept the serial data stream for
deserialization and decoding. The IN1± serial stream is passed to the receiver clock and data
recovery (CDR) circuit to extract the data content when INSEL = HIGH.
IN2±
LVPECL differential
input, with internal
DC restoration
Secondary differential serial data inputs. These inputs accept the serial data stream for
deserialization and decoding. The IN2± serial stream is passed to the receiver CDR circuit to
extract the data content when INSEL = LOW.
INSEL
LVTTL input,
asynchronous
Receive input selector. Determines which external serial bit stream is passed to the receiver
CDR. When HIGH, the IN1± input is selected. When LOW, the IN2± input is selected.
SDASEL
3-level select,[8]
static control input
Signal detect amplitude level select. Allows selection of one of three predefined amplitude
trip points for a valid signal indication, as listed in Table 10.
LPEN
LVTTL input,
asynchronous,
internal pull-down
Loop-back-enable. Active HIGH. When asserted (HIGH), the transmit serial data is internally
routed to the receiver CDR circuit.All enabled serial drivers are forced to differential logic “1.”
All serial data inputs are ignored.
8. 3-level select inputs are used for static configuration. They are ternary (not binary) inputs that make use of non-standard 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). When not
connected or allowed to float, a 3-level select input will self-bias to the MID level.
Document Number: 38-02031 Rev. *M
Page 10 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Pin Descriptions
CYP(V)15G0101DXB single-channel HOTLink II (continued)
Pin Name
I/O Characteristics Signal Description
OELE
LVTTL input,
asynchronous,
internal pull-up
Serial driver output enable latch enable. Active HIGH. When OELE = HIGH, the signals on
the BOE[1:0] inputs directly control the OUTx± differential drivers. When the BOE[x] input is
HIGH, the associated OUTx± differential driver is enabled. When the BOE[x] input is LOW,
the associated OUTx± differential driver is powered down. When OELE returns LOW, the last
values present on BOE[1:0] are captured in the internal output enable latch. The specific
mapping of BOE[1:0] signals to transmit output enables is listed in Table 14. If the device is
reset (TRSTZ is sampled LOW), the latch is reset to disable both outputs.
BISTLE
LVTTL input,
asynchronous,
internal pull-up
Transmit and receive BIST latch enable. Active HIGH. When BISTLE = HIGH, the signals
on the BOE[1:0] inputs directly control the transmit and receive BIST enables. When the
BOE[x] input is LOW, the associated transmit or receive channel is configured to generate or
compare the BIST sequence. When the BOE[x] input is HIGH, the associated transmit or
receive channel is configured for normal data transmission or reception. When BISTLE
returns LOW, the last values present on BOE[1:0] are captured in the internal BIST enable
latch. The specific mapping of BOE[1:0] signals to transmit and receive BIST enables is listed
in Table 14. When the latch is closed, if the device is reset (TRSTZ is sampled LOW), the latch
is reset to disable BIST on both the transmit and receive channels.
RXLE
LVTTL input,
asynchronous,
internal pull-up
Receive channel power-control latch enable. Active HIGH. When RXLE = HIGH, the
signal on the BOE[0] input directly controls the power enable for the receive PLL and analog
logic. When the BOE[0] input is HIGH, the receive channel PLL and analog logic are active.
When the BOE[0] input is LOW, the receive channel PLL and analog logic are placed in a
non-functional power saving mode. When RXLE returns LOW, the last value present on
BOE[0] is captured in the internal RX PLL enable latch. The specific mapping of BOE[1:0]
signals to the receive channel enable is listed in Table 14. When the latch is closed, if the
device is reset (TRSTZ is sampled LOW), the latch is reset to disable the receive channel.
BOE[1:0]
LVTTL input,
asynchronous,
internal pull-up
BIST, serial output, and receive channel enables. These inputs are passed to and through
the output enable latch when OELE = HIGH, and captured in this latch when OELE returns
LOW. These inputs are passed to and through the BIST enable latch when BISTLE = HIGH,
and captured in this latch when BISTLE returns LOW. These inputs are passed to and through
the receive channel enable latch when RXLE = HIGH, and captured in this latch when RXLE
returns LOW.
LVTTL output,
asynchronous
Link fault indication output. Active LOW. LFI is the logical OR of four internal conditions:
1. Received serial data frequency outside expected range
2. Analog amplitude below expected levels
3. Transition density lower than expected
4. Receive channel disabled.
LFI
JTAG Interface
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. The TAP controller is also reset automatically
upon application of power to the device.
TCLK
LVTTL input,
internal pull-down
JTAG test clock.
TDO
Three-state LVTTL
output
Test data out. JTAG data output buffer which is high Z while JTAG test mode is not selected.
TDI
LVTTL input,
internal pull-up
Test data in. JTAG data input port.
Power
VCC
+3.3 V power
GND
Signal and power ground for all internal circuits.
Document Number: 38-02031 Rev. *M
Page 11 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
CYP(V)15G0101DXB HOTLink II Operation
Phase-Align Buffer
The CYP(V)15G0101DXB is a highly configurable device
designed to support reliable transfer of large quantities of data
using high-speed serial links from a single source to one or more
destinations.
Data from the input register is passed either to the encoder or to
the phase-align buffer. When the transmit path is operated
synchronous
to
REFCLK↑
(TXCKSEL = LOW
and
TXRATE = LOW), the phase-align buffer is bypassed and data is
passed directly to the parity check and encoder block to reduce
latency.
CYP(V)15G0101DXB Transmit Data Path
Operating Modes
The transmit path of the CYP(V)15G0101DXB supports a single
character-wide data path. This data path is used in multiple
operating modes as controlled by the TXMODE[1:0] inputs.
Input Register
The bits in the input register support different assignments,
based on if the character is unencoded, encoded with two control
bits, or encoded with three control bits. These assignments are
shown in Table 1.
Table 1. Input Register Bit Assignments[9]
Signal Name
Unencoded
(Encoder
Bypassed)
TXD[0] (LSB)
TXD[1]
Encoded
(Encoder Enabled)
Two-bit
Control
Three-bit
Control
DIN[0]
TXD[0]
TXD[0]
DIN[1]
TXD[1]
TXD[1]
TXD[2]
DIN[2]
TXD[2]
TXD[2]
TXD[3]
DIN[3]
TXD[3]
TXD[3]
TXD[4]
DIN[4]
TXD[4]
TXD[4]
TXD5]
DIN[5]
TXD[5]
TXD[5]
TXD[6]
DIN[6]
TXD[6]
TXD[6]
TXD[7]
DIN[7]
TXD[7]
TXD[7]
TXCT[0]
DIN[8]
TXCT[0]
TXCT[0]
TXCT[1] (MSB)
DIN[9]
TXCT[1]
TXCT[1]
SCSEL
N/A
N/A
SCSEL
The input register captures a minimum of eight data bits and two
control bits on each input clock cycle. When the encoder is
bypassed, the TXCT[1:0] control bits are part of the pre-encoded
10-bit data character.
When the encoder is enabled (TXMODE[1] ≠ LOW), the
TXCT[1:0] bits are interpreted along with the TXD[7:0] character
to generate the specific 10-bit transmission character. When
TXMODE[0] ≠ HIGH, an additional special character select
(SCSEL) input is also captured and interpreted. This SCSEL
input is used to modify the encoding of the characters.
When an input register clock with an uncontrolled phase
relationship to REFCLK is selected (TXCKSEL ≠ LOW) or if data
is captured on both edges of REFCLK (TXRATE = HIGH), the
phase-align buffer is enabled. This buffer is used to absorb clock
phase differences between the presently selected input clock
and the internal character clock.
Initialization of the phase-align buffer takes place when the
TXRST input is sampled LOW by two consecutive rising edges
of REFCLK. When TXRST is returned HIGH, the present input
clock phase relative to REFCLK↑ is set. TXRST is an
asynchronous input, but is sampled internally to synchronize it to
the internal transmit path state machine.
Once set, the input clock is allowed to skew in time up to half a
character period in either direction relative to REFCLK↑; that
is ±180°. This time shift allows the delay path of the character
clock (relative to REFLCK↑) to change due to operating voltage
and temperature, while not affecting the design operation.
If the phase offset, between the initialized location of the input
clock and REFCLK↑, exceeds the skew handling capabilities of
the phase-align buffer, an error is reported on the TXPER output.
This output indicates a continuous error until the phase-align
buffer is reset. While the error remains active, the transmitter
outputs a continuous C0.7 character to indicate to the remote
receiver that an error condition is present in the link.
In specific transmit modes, it is also possible to reset the
phase-align buffer with minimal disruption of the serial data
stream. When the transmit interface is configured for generation
of atomic Word Sync Sequences (TXMODE[1] = MID) and a
phase-align buffer error is present, the transmission of a Word
Sync Sequence will recenter the phase-align buffer and clear the
error condition.[10]
Parity Support
In addition to the ten data and control bits that are captured at
the transmit input register, a TXOP input is also available. This
allows the CYP(V)15G0101DXB to support ODD parity
checking. Parity checking is available for all operating modes
(including encoder bypass). The specific mode of parity checking
is controlled by the PARCTL input, and operates per Table 2.
When PARCTL = MID (open) and the encoder is enabled
(TXMODE[1] ≠ LOW), only the TXD[7:0] data bits are checked
for ODD parity along with the TXOP bit. When PARCTL = HIGH
with the encoder enabled (or MID with the encoder bypassed),
the TXD[7:0] and TXCT[1:0] inputs are checked for ODD parity
along with the TXOP bit. When PARCTL = LOW, parity checking
is disabled.
Notes
9. The TXOP input is also captured in the input register, but its interpretation is under the separate control of PARCTL.
10. One or more K28.5 characters may be added or lost from the data stream during this reset operation. When used with non-Cypress devices that require a complete
16-character Word Sync Sequence for proper receive elasticity buffer alignment, it is recommend that the sequence be followed by a second Word Sync Sequence
to ensure proper operation.
Document Number: 38-02031 Rev. *M
Page 12 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
When parity checking and the encoder are both enabled
(TXMODE[1] ≠ LOW), the detection of a parity error causes a
C0.7 character of proper disparity to be passed to the transmit
shifter. When the encoder is bypassed (TXMODE[1] = LOW),
detection of a parity error causes a positive disparity version of
a C0.7 transmission character to be passed to the transmit
shifter.
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 serial receive PLL to
extract a clock from the data stream)
■
a DC-balance in the signaling (to prevent baseline wander)
■
run-length limits in the serial data (to limit the bandwidth of the
link)
■
the remote receiver a way of determining the correct character
boundaries (framing).
Table 2. Input Register Bits Checked for Parity[12]
Transmit Parity Check Mode (PARCTL)
Signal
Name
MID
LOW
TXMODE[1]
= LOW
TXMODE[1]
≠ LOW
HIGH
TXD[0]
X[11]
X
X
TXD[1]
X
X
X
TXD[2]
X
X
X
TXD[3]
X
X
X
TXD[4]
X
X
X
TXD[5]
X
X
X
TXD[6]
X
X
X
X
TXD[7]
X
TXCT[0]
X
TXCT[1]
X
TXOP
X
X
X
X
X
X
Encoder
The character, received from the input register or phase-align
buffer and parity check logic, is then passed to the encoder logic.
This block interprets each character and any control bits, and
outputs a 10-bit transmission character.
Depending on the configured operating mode, the generated
transmission character may be
■
the 10-bit pre-encoded character accepted in the input register
■
the 10-bit equivalent of the eight-bit data character accepted in
the input register
■
the 10-bit equivalent of the eight -bit special character code
accepted in the input register
■
the 10-bit equivalent of the C0.7 SVS character if parity
checking was enabled and a parity error was detected
■
the 10-bit equivalent of the C0.7 SVS 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.
The selection of the specific characters generated are controlled
by the TXMODE[1:0], SCSEL, TXCT[1:0], and TXD[7:0] inputs
for each character.
When the encoder is enabled (TXMODE[1] ≠ LOW), the
characters to be transmitted are converted from data or special
character codes to 10-bit transmission characters (as selected
by the TXCT[1:0] and SCSEL inputs), using an integrated
8B/10B encoder. When directed to encode the character as a
special character code, it is encoded using the special character
encoding rules listed in Table 21. When directed to encode the
character as a data character, it is encoded using the data
character encoding rules in Table 20.
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™, and digital video
broadcast (DVB-ASI) standards for data transport.
Many of the special character codes listed in Table 21 may be
generated by more than one input character. The
CYP(V)15G0101DXB is designed to support two independent
(but non-overlapping) special character code tables. This allows
the CYP(V)15G0101DXB 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
The operating mode of the transmit path is set through the
TXMODE[1:0] inputs. These 3-level select inputs allow one of
nine transmit modes to be selected. The transmit modes are
listed in Table 3.
The encoded modes (TX modes 3 through 8) support multiple
encoding tables. These encoding tables vary by the specific
combinations of SCSEL, TXCT[1], and TXCT[0] that are used to
control the generation of data and control characters. These
multiple encoding forms allow maximum flexibility in interfacing
to legacy applications, while also supporting numerous
extensions in capabilities.TX Mode 0—encoder bypass
Notes
11. Bits marked as X are XORed together. Result must be a logic-1 for parity to be valid.
12. Transmit path parity errors are reported on the TXPER output.
Document Number: 38-02031 Rev. *M
Page 13 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
When the encoder is bypassed, the character captured from the
TXD[7:0] and TXCT[1:0] inputs is passed directly to the transmit
shifter without modification. If parity checking is enabled
(PARCTL ≠ LOW) and a parity error is detected, the 10-bit
character is replaced with the 1001111000 pattern (+C0.7
character) regardless of the running disparity of the previous
character.
With the encoder bypassed, the TXCT[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 TXD[7:0]
bits. The bit usage and mapping of these control bits when the
encoder is bypassed is shown in Table 4.
In encoder bypass mode, the SCSEL input is ignored. All
clocking modes interpret the data in the same way.
TX Modes 1 and 2—Factory Test Modes
These modes enable specific factory test configurations. They
are not considered normal operating modes of the device. Entry
or configuration into these test modes will not damage the
device.
TX Mode 3—Atomic Word Sync and SCSEL Control of Special
Codes
When configured in TX Mode 3, the SCSEL input is captured
along with the TXCT[1:0] data control inputs. These bits combine
to control the interpretation of the TXD[7:0] bits and the
characters generated by them. These bits are interpreted as
listed in Table 5.
Table 5. TX Modes 3 and 6 Encoding
TXCT[1]
TXCT[0]
X
X
0
Encoded data character
Characters Generated
TXMODE
[1:0]
Operating Mode
Mode
Number
TX Mode
SCSEL
Table 3. Transmit Operating Modes
0
0
1
K28.5 fill character
0
LL
None
None
Encoder bypass
1
0
1
Special character code
1
LM
None
None
Reserved for test
X
1
1
16-character Word Sync Sequence
2
LH
None
None
Reserved for test
3
ML
Atomic
Special
Character
Encoder control
4
MM
Atomic
Word Sync
Encoder control
5
MH
Atomic
None
Encoder control
6
HL
Interruptible
Special
Character
Encoder control
7
HM
Interruptible
Word Sync
Encoder control
8
HH
Interruptible
None
Encoder control
Word Sync
Sequence
Support
SCSEL
Control
TXCT Function
Table 4. Encoder Bypass Mode (TXMODE[1:0] = LL)
Signal Name
Bus Weight
10B Name
20
a
TXD[1]
21
b
TXD[2]
22
c
TXD[3]
23
d
TXD[4]
24
e
TXD[5]
25
i
TXD[6]
26
f
TXD[7]
27
g
TXCT[0]
28
h
TXCT[1] (MSB)
29
j
TXD[0]
[13]
(LSB)
When TXCKSEL = MID or HIGH, the transmit channel captures
data into its input register using the TXCLK clock.
Word Sync Sequence
When TXCT[1:0] = 11, a 16-character sequence of K28.5
characters, known as a Word Sync Sequence, is generated on
the transmit 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
– – + + – + – + – + – + – + – +.
When TXMODE[1] = MID (open, TX modes 3, 4 and 5), the
generation of this character sequence is an atomic
(non-interruptible) operation. Once it has been successfully
started, it cannot be stopped until all 16 characters have been
generated. The content of the input register is ignored for the
duration of this 16-character sequence. At the end of this
sequence, if the TXCT[1:0] = 11 condition is sampled again, the
sequence restarts and remains uninterrupted for the following 15
character clocks.
If parity checking is enabled, the character used to start the Word
Sync Sequence must also have correct ODD parity. This is true
even though the contents of the TXD[7:0] bits do not directly
control the generation of characters during the Word Sync
Sequence. Once the sequence is started, parity is not checked
on the following 15 characters in the Word Sync Sequence.
Note
13. LSB is shifted out first.
Document Number: 38-02031 Rev. *M
Page 14 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
When TXCKSEL = LOW, the input register for the transmit
channel is clocked by REFCLK.[14] When TXCKSEL = HIGH or
MID, the input register for the transmit channel is clocked with
TXCLK↑.
TX Mode 4—Atomic Word Sync and SCSEL Control of
Word Sync Sequence Generation
When configured in TX Mode 4, the SCSEL input is captured
along with the TXCT[1:0] data control inputs. These bits combine
to control the interpretation of the TXD[7:0] bits and the
characters generated by them. These bits are interpreted as
listed in Table 6.
SCSEL
TXCT[1]
TXCT[0]
Table 6. TX Modes 4 and 7 Encoding
X
X
0
Encoded data character
0
0
1
K28.5 fill character
0
1
1
Special character code
1
X
1
16-character Word Sync Sequence
Characters Generated
TX Mode 4 also supports an Atomic Word Sync Sequence.
Unlike TX Mode 3, this sequence is started when both SCSEL
and TXCT[0] are sampled HIGH. With the exception of the
combination of control bits used to initiate the sequence, the
generation and operation of this Word Sync Sequence is the
same as that documented for TX Mode 3.
Note
14. When REFCLK is configured for half-rate operation (TXRATE
REFCLK.
Document Number: 38-02031 Rev. *M
When configured in TX Mode 5, the SCSEL signal is not used.
The TXCT[1:0] inputs control the characters generated by the
channel. The specific characters generated by these bits are
listed in Table 7.
TXCT[0]
Table 7. TX Modes 5 and 8 Encoding
TXCT[1]
When parity checking is enabled and TXMODE[1] = HIGH, all
characters (including those in the middle of a Word Sync
Sequence) must have correct parity. The detection of a character
with incorrect parity during a Word Sync Sequence (regardless
of the state of TXCT[1:0]) will interrupt that sequence and force
generation of a C0.7 SVS character. Any interruption of the Word
Sync Sequence causes the sequence to terminate.
TX Mode 5—Atomic Word Sync, No SCSEL
SCSEL
When TXMODE[1] = HIGH (TX modes 6, 7, and 8), the
generation of the Word Sync Sequence becomes an interruptible
operation. In TX Mode 6, this sequence is started as soon as the
TXCT[1:0] = 11 condition is detected on the channel. In order for
the sequence to continue, the TXCT[1:0] inputs must be sampled
as 00 for the remaining 15 characters of the sequence. If at any
time a sample period exists where TXCT[1:0] ≠ 00, the Word
Sync Sequence is terminated, and a character representing the
data and control bits is generated by the encoder. This resets the
Word Sync Sequence state machine such that it will start at the
beginning of the sequence at the next occurrence of
TXCT[1:0] = 11.
X
0
0
Encoded data character
X
0
1
K28.5 fill character
X
1
0
Special character code
X
1
1
16-character Word Sync Sequence
Characters Generated
TX Mode 5 also has the capability of generating an Atomic Word
Sync Sequence. For the sequence to be started, the TXCT[1:0]
inputs must both be sampled HIGH. The generation and
operation of this Word Sync Sequence is the same as that
documented for TX Mode 3.
Transmit BIST
The transmit channel contains an internal pattern generator that
can be used to validate both device and link operation. This
generator is enabled by the BOE[1] signal, as listed in Table 8
(when the BISTLE latch enable input is HIGH). When enabled,
a register in the transmit channel becomes a signature pattern
generator by logically converting to a linear feedback shift
register (LFSR). This LFSR generates a 511-character
sequence that includes all data and special character codes,
including the explicit violation symbols. This provides a
predictable yet pseudo-random sequence that can be matched
to an identical LFSR in the attached Receiver. If the receive
channel is configured for REFCLK clocking (RXCKSEL = LOW),
each pass is preceded by a 16-character Word Sync Sequence
to allow elasticity buffer alignment and management of
clock-frequency variations.
When the BISTLE signal is HIGH, if the BOE[1] input is LOW, the
BIST generator in the transmit channel is enabled (and if
BOE[0] = LOW the BIST checker in the receive channel is
enabled). When BISTLE returns LOW, the values of the
BOE[1:0] signals are captured in the BIST enable latch. These
values remain in the BIST enable latch until BISTLE is returned
high to open the latch again. A device reset (TRSTZ sampled
LOW), also presets the BIST enable latch to disable BIST on
both the transmit and receive channels.
All data and data-control information present at the TXD[7:0] and
TXCT[1:0] inputs are ignored when BIST is active on the transmit
channel.
= HIGH), this input is sampled (or the outputs change) relative to both the rising and falling edges of
Page 15 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Serial Output Drivers
The serial interface output drivers use high-performance
differential current mode logic (CML) to provide source-matched
drivers for the transmission lines. These serial drivers accept
data from the transmit shifter. These outputs have signal swings
equivalent to that of standard PECL drivers, and are capable of
driving AC-coupled optical modules or AC-coupled transmission
lines.
When configured for local loop-back (LPEN = HIGH), the
enabled serial drivers are configured to drive a static differential
logic-1.
Each serial driver can be enabled or disabled through the
BOE[1:0] inputs, as controlled by the OELE latch-enable signal.
When OELE = HIGH, the signals present on the BOE[1:0] inputs
are passed through the serial output enable latch to control the
serial driver. The BOE[1:0] input with OUT1± and OUT2± driver
is listed in Table 8.
Table 8. Output Enable, BIST, and Receive Channel Enable
Signal Map
BOE Input
Output
Controlled
(OELE)
BIST
Channel
Enable
(BISTLE)
Receive PLL
Channel
Enable
(RXLE)
BOE[1]
OUT2±
Transmit
X
BOE[0]
OUT1±
Receive
Receive
When OELE = HIGH and BOE[x] = HIGH, the associated serial
driver is enabled to drive any attached transmission line. When
OELE = HIGH and BOE[x] = LOW, the associated driver is
disabled and internally configured for minimum power
dissipation. If both serial drivers for the channel are disabled, the
internal logic for the transmit channel is also configured for
lowest power operation. When OELE returns LOW, the values
present on the BOE[1:0] inputs are latched in the output enable
latch, and remain there until OELE returns HIGH to open the
latch again. A device reset (TRSTZ sampled LOW) clears this
latch and disables both serial drivers.
Note. When both serial output drivers are disabled and a driver
is re-enabled, the data on the serial drivers may not meet all
timing specifications for up to 200 µs.
Transmit PLL Clock Multiplier
The transmit PLL clock multiplier accepts a character-rate or
half-character-rate external clock at the REFCLK input, and
multiples that clock by 10 or 20 (as selected by TXRATE) to
generate a bit-rate clock for use by the transmit shifter. It also
provides a character-rate clock used by the transmit path.
This clock multiplier PLL can accept a REFCLK input between
19.5 MHz and 150 MHz, however, this clock range is limited by
the operating mode of the CYP(V)15G0101DXB clock multiplier
(controlled by TXRATE) and by the level on the SPDSEL input.
Note
15. When REFCLK is configured for half-rate operation (TXRATE
REFCLK.
Document Number: 38-02031 Rev. *M
When TXRATE=HIGH, configuring TXCKSEL = HIGH or MID is
an invalid mode of operation.
SPDSEL is a 3-level select[15] (ternary) input that selects one of
three operating ranges for the serial data outputs and inputs. The
operating serial signaling-rate and allowable range of REFCLK
frequencies are listed in Table 9.
Table 9. Operating Speed Settings
SPDSEL
TXRATE
REFCLK
Frequency
(MHz)
Signaling Rate
(MBaud)
LOW
1
Reserved
195–400
0
19.5–40
MID (Open)
1
20–40
0
40–80
HIGH
1
40–75
0
80–150
400–800
800–1500
The REFCLK± input is a differential input with each input
internally biased to 1.4 V. If the REFCLK+ 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 both the REFCLK+ and REFCLK− inputs are connected,
the clock source must be a differential clock. This can be either
a differential LVPECL clock that is DC- or AC-coupled, or a
differential LVTTL or LVCMOS clock.
By connecting the REFCLK− input to an external voltage source
or resistive voltage divider, it is possible to adjust the reference
point of the REFCLK+ input for alternate logic levels. When doing
so, it is necessary to ensure that the 0 V-differential crossing
point remains within the parametric range supported by the input.
CYP(V)15G0101DXB Receive Data Path
Serial Line Receivers
Two differential line receivers, IN1± and IN2±, are available for
accepting serial data streams. The active serial line receiver is
selected using the INSEL input. Both serial line receivers have
differential inputs, 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 VDIFFS > 100 mV, or 200-mV peak-to-peak
differential. Each line receiver can be DC- or AC-coupled to
+3.3 V powered fiber-optic interface modules (any ECL/PECL
logic family, not limited to 100 K PECL) or AC-coupled to
+5 V-powered optical modules. The common-mode tolerance of
the receivers accommodates a wide range of signal termination
voltages. Each receiver provides internal DC-restoration, to the
center of the receiver’s common mode range, for AC-coupled
signals.
= HIGH), this input is sampled (or the outputs change) relative to both the rising and falling edges of
Page 16 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
The local loop-back input (LPEN) allows the serial transmit data
to be routed internally back to the clock and data recovery circuit.
When configured for local loop-back, the transmit serial driver
outputs are forced to output a differential logic-1. This prevents
local diagnostic patterns from being broadcast to attached
remote receivers.
Range Control
Signal Detect/Link Fault
■
when the incoming data stream resumes after a time in which
it has been “missing.”
■
when the incoming data stream is outside the acceptable
frequency range.
Each selected line receiver (i.e., that routed to the clock and data
recovery PLL) is simultaneously monitored for
■
analog amplitude above limit specified by SDASEL
■
transition density greater than specified limit
■
range controller reports the received data stream within normal
frequency range (±1500 ppm)[16]
■
receive channel enabled.
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 LFI (link fault indicator) output.
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.
This adjustment is made through the SDASEL signal, a 3-level
select[17] (ternary) input, which sets the trip point for the detection
of a valid signal at one of three levels, as listed in Table 10.
The analog signal detect monitor is active for the present line
receiver, as selected by the INSEL input. When configured for
local loop-back (LPEN = HIGH), the analog signal detect monitor
is disabled.
Transition Density
The transition detection logic checks for the absence of any
transitions spanning greater than six transmission characters
(60-bits). If no transitions are present in the data received (within
the referenced period), the transition detection logic asserts LFI.
The LFI output remains asserted until at least one transition is
detected in each of three adjacent received characters.
[18]
Table 10. Analog Amplitude Detect Valid Signal Levels
Typical Signal with Peak Amplitudes
Above
LOW
140-mV p-p differential
MID (Open) 280-mV p-p differential
HIGH
420-mV p-p differential
SDASEL
The clock/data recovery (CDR) circuit includes logic to monitor
the frequency of the phase-locked loop (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:
To perform this function, the frequency of the VCO is periodically
sampled and compared to the frequency of the REFCLK input. If
the VCO is running at a frequency beyond +1500 ppm[16] as
defined by the reference clock frequency, it is periodically forced
to the correct frequency (as defined by REFCLK, SPDSEL, and
TXRATE) and then released in an attempt to lock to the input
data stream. The sampling and relock period of the range control
is calculated as follows: RANGE CONTROL SAMPLING
PERIOD = (REFCLKPERIOD) × (16000).
During the time that the range control forces the PLL VCO to run
at REFCLK × 10 (or REFCLK × 20 when TXRATE = HIGH) rate,
the LFIx output will be asserted LOW. While the PLL is
attempting to re-lock to the incoming data stream, LFIx may be
either HIGH or LOW (depending on other factors such as
transition density and amplitude detection) and the recovered
byte clock (RXCLK) may run at an incorrect rate (depending on
the quality or existence of the input serial data stream). 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 CYP(V)15G0101DXB receive channel can be enabled and
disabled through the BOE[0] input, as controlled by the RXLE
latch-enable signal. When RXLE = HIGH, the signal present on
the BOE[0] input is passed through the receive channel enable
latch to control the PLL and logic of the receive channel. The
BOE[1:0] input functions are listed in Table 8.
When RXLE = HIGH and BOE[0] = HIGH, the receive channel is
enabled to receive and recover a serial stream from the line
receiver. When RXLE = HIGH and BOE[0] = LOW, the receive
channel is disabled and internally configured for minimum power
dissipation. When disabled, the channel indicates a constant LFI
output. When RXLE returns LOW, the values present on the
BOE[1:0] inputs are latched in the Receive Channel Enable
Latch, and remain there until RXLE returns HIGH to open the
latch again.[19]
Notes
16. REFCLK has no phase or frequency relationship with the recovered clock(s) and only acts as a centering reference to reduce clock synchronization time. REFCLK
must be within ±1500 PPM (±0.15%) of the remote transmitter’s PLL reference (REFCLK) 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.
17. When REFCLK is configured for half-rate operation (TXRATE = HIGH), this input is sampled (or the outputs change) relative to both the rising and falling edges of
REFCLK.
18. 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.
19. When a disabled receive channel is reenabled, the status of the LFI output and data on the parallel outputs may be indeterminate for up to 2 ms.
Document Number: 38-02031 Rev. *M
Page 17 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Clock/Data Recovery
Framing Character
The extraction of a bit-rate clock and recovery of bits from a
received serial stream is performed by a CDR block within the
receive channel. The clock extraction function is performed by a
high-performance embedded PLL that tracks the frequency of
the transitions in the incoming bit stream and aligns the phase of
the internal bit-rate clock to the transitions in the serial data
stream.
The CYP(V)15G0101DXB allows selection of two combinations
of framing characters to support requirements of different
interfaces. The selection of the framing character is made
through the FRAMCHAR input.
The CDR accepts a character-rate (bit-rate ÷ 10) or
half-character-rate (bit-rate ÷ 20) reference clock from the
REFCLK input. This REFCLK input is used to
■
ensure that the VCO (within the CDR) is operating at the correct
frequency
■
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 will attempt to
recover a data stream from it. If the frequency of the recovered
data stream is outside the limits of the range control monitor, the
CDR will switch to track REFCLK instead of the data stream.
Once the CDR output (RXCLK) frequency returns back close to
REFCLK frequency, the CDR input will be switched back to track
the input data stream. In case no data is present at the input, this
switching behavior may result in brief RXCLK frequency
excursions from REFCLK. However, the validity of the input data
stream is indicated by the LFIx output. The frequency of
REFCLK is required to be within ± 1500 ppm[20] of the frequency
of the clock that drives the REFCLK input of the remote
transmitter to ensure a lock to the incoming data stream.
For systems using multiple or redundant connections, the LFI
output can be used to select an alternate data stream. When an
LFI indication is detected, external logic can toggle selection of
the IN1± and IN2± inputs through the INSEL input. When a port
switch takes place, it is necessary for the receive PLL to
reacquire the new serial stream and frame to the incoming
character boundaries.
Deserializer/Framer
Each CDR circuit extracts bits from the serial data stream and
clocks these bits into the shifter/framer at the bit-clock rate.
When enabled, the framer examines the data stream, looking for
one or more comma or K28.5 characters at all possible bit
positions. The location of these characters in the data stream are
used to determine the character boundaries of all following
characters.
The specific bit combinations of these framing characters are
listed in Table 11. 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 11. Framing Character Selector
Bits Detected in Framer
FRAMCHAR
Character Name
LOW
Bits Detected
Reserved for test
MID (Open)
Comma+
Comma–
00111110XX[21]
or 11000001XX
HIGH
–K28.5
+K28.5
0011111010 or
1100000101
Framer
The framer operates in one of three different modes, as selected
by the RFMODE input. In addition, the framer itself may be
enabled or disabled through the RFEN input. When
RFEN = LOW, the framer is disabled, and no combination of bits
in a received data stream will alter the character boundaries.
When RFEN = HIGH, the framer-mode selected by RFMODE is
enabled.
When RFMODE = LOW, the low-latency framer is selected. This
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 make use of a 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
(RXRATE = LOW), 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 (RXRATE = HIGH), the output
of properly framed characters may be delayed by up to 14
character-clock cycles from the detection of the selected framing
character.[22]
Notes
20. REFCLK has no phase or frequency relationship with the recovered clock(s) and only acts as a centering reference to reduce clock synchronization time. REFCLK
must be within ±1500 ppm (±0.15%) of the remote transmitter’s PLL reference (REFCLK) 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.
21. 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 8th bit as
an inversion of the 7th bit, the compare pattern is extended to a full eight bits to reduce the possibility of a framing error.
22. 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
would cause the receiver to update its character boundaries incorrectly.
Document Number: 38-02031 Rev. *M
Page 18 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
When RFMODE = MID (open), the Cypress-mode multi-byte
framer is selected. The required detection of multiple framing
characters makes the link much more robust to incorrect framing
due to aliased framing 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 DECMODE = MID (or open), the 10-bit transmission
characters are decoded using Table 20 and Table 21. Received
special code characters are decoded using the Cypress column
of Table 21.
When DECMODE = HIGH, the 10-bit transmission characters
are decoded using Table 20 and Table 21. Received special
code characters are decoded using the alternate column of
Table 21.
Receive BIST Operation
When RFMODE = HIGH, 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 framer does not adjust
the character clock boundary, but instead aligns the character to
the already recovered character clock. In this mode, the data
stream must contain a minimum of four of the selected framing
characters, received as consecutive characters, on identical
10-bit boundaries, before character framing is adjusted.
The receiver interface contains an internal pattern generator that
can be used to validate both device and link operation. This
generator is enabled by the BOE[0] signal as listed in Table 8
(when the BISTLE latch enable input is HIGH). When enabled,
a register in the receive channel becomes a pattern generator
and checker by logically converting to a linear feedback shift
register (LFSR). This LFSR generates a 511-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. If the receive
channels
are
configured
for
REFCLK
clocking
(RXCKSEL = LOW), each pass is preceded by a 16-character
Word Sync Sequence.
Framing is enabled when RFEN = HIGH. If RFEN = LOW, the
framer is disabled. When the framer is disabled, no changes are
made to the recovered character boundary, regardless of the
presence of framing characters in the data stream.
When synchronized with the received data stream, the receiver
checks each character in the decoder with each character
generated by the LFSR and indicates compare errors and BIST
status at the RXST[2:0] bits of the output register.
10B/8B Decoder Block
When the BISTLE signal is HIGH, if the BOE[0] input is LOW the
BIST generator/checker in the receive channel is enabled (and
if BOE[1] = LOW the BIST generator in the transmit channel is
enabled). When BISTLE returns LOW, the values of the
BOE[1:0] signals are captured in the BIST enable latch. These
values remain in the BIST enable latch until BISTLE is returned
high to open the latch again. All captured signals in the BIST
enable latch are set HIGH (i.e., BIST is disabled) following a
device reset (TRSTZ is sampled LOW).
The decoder logic block performs three primary functions:
■
decoding the received transmission characters back into data
and special character codes
■
comparing generated BIST patterns with received characters
to permit at-speed link and device testing
■
generation of ODD parity on the decoded characters.
10B/8B Decoder
The framed parallel output of the deserializer shifter is passed to
the 10B/8B decoder where, if the decoder is enabled
(DECMODE ≠ LOW), it is transformed from a 10-bit transmission
character back to the original data and special character codes.
This block uses the 10B/8B decoder patterns in Table 20 and
Table 21 of this data sheet. Valid data characters are indicated
by a 000b bit-combination on the RXST[2:0] status bits, and
special character codes are indicated by a 001b bit-combination
on these same status outputs. Framing characters, invalid
patterns, disparity errors, and synchronization status are
presented as alternate combinations of these status bits.
The 10B/8B decoder operates in two normal modes, and can
also be bypassed. The operating mode for the decoder is
controlled by the DECMODE input.
When DECMODE = LOW, the decoder is bypassed and raw
10-bit characters are passed to the output register. In this mode,
the receive elasticity buffers are bypassed, and RXCKSEL must
be MID.
Document Number: 38-02031 Rev. *M
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 D0.0
character is sent only once per BIST loop. The status of the BIST
progress and any character mismatches is presented on the
RXST[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. RXST[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 the receive
channel.
The status reported on RXST[2:0] by the BIST state machine are
listed in Table 16. When receive BIST is enabled, the same
status is reported on the receive status outputs regardless of the
state of DECMODE.
The specific patterns checked by each receiver are described in
detail in the Cypress application note HOTLink Built-In Self-test.
The sequence compared by the CYP(V)15G0101DXB is
identical to that in the CY7B933 and CY7C924DX, allowing
interoperable systems to be built when used at compatible serial
signaling rates.
Page 19 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
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.
also be centered by a device reset operation initiated through the
TRSTZ input. However, following such an event, the
CYP(V)15G0101DXB will normally require a framing event
before it will correctly decode characters.
When the receive paths are configured for REFCLK clocking
(RXCKSEL = LOW), each pass must be preceded by a
16-character Word Sync Sequence to allow output buffer
alignment and management of clock frequency variations. This
is automatically generated by the transmitter when its local
RXCKSEL = LOW.
Receive Modes
The BIST state machine requires the characters to be correctly
framed for it to detect the BIST sequence. If the low-latency
framer is enabled (RFMODE = LOW), the framer will misalign to
an aliased framing character within the BIST sequence. If the
alternate-mode multi-byte framer is enabled (RFMODE = HIGH)
and the receiver outputs are clocked relative to a recovered clock
(RXCKSEL = MID), it is necessary to frame the receiver before
BIST is enabled. If the receiver outputs are clocked relative to
REFCLK (RXCKSEL = LOW), the transmitter precedes every
511 character BIST sequence with a 16-character Word Sync
Sequence.
Receive Elasticity Buffer
The receive channel contains an elasticity buffer that is designed
to support multiple clocking modes. This buffer allows data to be
read using an elasticity buffer read-clock that is asynchronous in
both frequency and phase from the elasticity buffer write clock,
or to use a read clock that is frequency coherent but with uncontrolled phase relative to the elasticity buffer write clock.
The elasticity buffer is 10 characters deep, and supports a
12-bit-wide data path. It is capable of supporting a decoded
character, three status bits, and a parity bit for each character
present in the buffer. The write clock for this buffer is always the
recovered clock for the read channel.
The read clock for the elasticity buffer can be set to
character-rate REFCLK (RXCKSEL = LOW and DECMODE ≠
LOW). The write clock for the elasticity buffer is always
recovered clock.
When RXCKSEL = LOW, the receive channel is clocked by
REFCLK. The RXCLK± and RXCLKC+ outputs present buffered
and delayed forms of REFCLK. In this mode, the receive
elasticity buffer is enabled. For REFCLK clocking, the elasticity
buffer must be able to insert K28.5 characters and delete framing
characters as appropriate. The elasticity buffer is bypassed
whenever the decoder is bypassed (DECMODE = LOW). When
the decoder and elasticity buffer are bypassed, RXCKSELx must
be set to MID. When RXCKSEL = MID (or open), the receive
channel output register is clocked by the recovered clock.
The insertion of a K28.5 or deletion of a framing character can
occur at any time. However, the actual timing on these insertions
and deletions is controlled in part by the 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 present
in the elasticity buffer. To prevent an elasticity buffer overflow or
underflow in the receive channel, a minimum density of framing
characters must be present in the received data stream.
Prior to reception of valid data, at least one Word Sync Sequence
(or at least four framing characters) must be received to allow the
receive elasticity buffer to be centered. The elasticity buffer may
Document Number: 38-02031 Rev. *M
The operating mode of the receive path is set through the
RXMODE input. The ‘Reserved for test’ setting (RXMODE = M)
is not allowed, even if the receiver is not being used, as it will
stop normal function of the device. When the decoder is
disabled, the RXMODE setting is ignored as long as it is not a
test mode. These modes determine the RXST status reporting.
The different receive modes are listed in Table 12.
Table 12. Receive Operating Modes
RX Mode
Mode
RXMODE
Number
0
L
1
M
2
H
RXST Status Reporting
Status A
Reserved for test
Status B
Power Control
The CYP(V)15G0101DXB supports user control of the powered
up or down state of the transmit and receive channel. The
receive channel is controlled by the RXLE signal and the values
present on the BOE[1:0] bus. The transmit channel is controlled
by the OELE signal and the values present on the BOE[1:0] bus.
If either the transmit or the receive channel is not used, then
powering down the unused channel will save power and reduce
system heat generation. Controlling system power dissipation
will improve the system performance.
Receive Channel
When RXLE = HIGH, the signal on the BOE[0] input directly
controls the power enable for the receive PLL and the analog
circuit. When BOE[0] = HIGH, the receive channel and its analog
circuits are active. When BOE[0] = LOW, the receive channel
and its analog circuits are powered down. When RXLE returns
LOW, the values present on the BOE[1:0] inputs are latched in
the receive channel enable latch. When a disabled receive
channel is re-enabled, the status of the LFI output and data on
the parallel outputs for the receive channel may be indeterminate
for up to 2 ms.
Transmit Channel
When OELE = HIGH, the signals on the BOE[1:0] inputs directly
control the power enables for the serial drivers. When a BOE[1:0]
input is HIGH, the associated serial driver is enabled. When a
BOE[1:0] input is LOW, the associated serial driver is disabled.
When both serial drivers are powered down, the logic in the
entire transmit channel is also powered down. When OELE
returns LOW, the values present on the BOE[1:0] inputs are
latched in the output enable latch.
Device Reset State
When the CYP(V)15G0101DXB is reset by assertion of TRSTZ,
both the transmit enable and receive enable latches are cleared,
and the BIST enable latch is preset. In this state, the transmit and
receive channels are disabled, and BIST is disabled.
Page 20 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Following a device reset, it is necessary to enable the transmit
and receive channels for normal operation. This can be done by
sequencing the appropriate values on the BOE[1:0] inputs while
the OELE and RXLE signals are raised and lowered. For
systems that do not require dynamic control of power, or want
the part to power-up in a fixed configuration, it is also possible to
strap the RXLE and OELE control signals HIGH to permanently
enable their associated latches. Connection of the associated
BOE[1:0] signals to a stable HIGH will then enable the transmit
and receive channels as soon as the TRSTZ signal is
deasserted.
Output Bus
The receive channel presents a 12-signal output bus consisting
of
■
an eight-bit data bus
■
a three-bit status bus
■
a parity bit.
The bit assignments of the Data and Status are dependent on the
setting of DECMODE. This mapping is shown in Table 13.
Table 13. Output Register Bit Assignments[23]
Signal Name
DECMODE = LOW
RXST[2] (LSB)
RXST[1]
RXST[0]
RXD[0]
RXD[1]
RXD[2]
RXD[3]
RXD[4]
RXD[5]
RXD[6]
RXD[7] (MSB)
COMDET
DOUT[0]
DOUT[1]
DOUT[2]
DOUT[3]
DOUT[4]
DOUT[5]
DOUT[6]
DOUT[7]
DOUT[8]
DOUT[9]
DECMODE = MID
or HIGH
RXST[2]
RXST[1]
RXST[0]
RXD[0]
RXD[1]
RXD[2]
RXD[3]
RXD[4]
RXD[5]
RXD[6]
RXD[7]
When the 10B/8B decoder is bypassed (DECMODE = LOW),
the framed 10-bit character is presented to the receiver output
register, along with a status output (COMDET) 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 14.
Table 14. Decoder Bypass Mode (DECMODE = LOW)
Signal Name
Bus Weight
RXST[2] (LSB)
COMDET
10B Name
RXST[1]
20
a
RXST[0]
21
b
RXD[0]
22
c
RXD[1]
23
d
RXD[2]
24
e
RXD[3]
25
i
RXD[4]
26
f
RXD[5]
27
g
RXD[6]
28
h
RXD[7] (MSB)
29
j
The COMDET output 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
is also enabled (RFMODE = LOW, RXRATE = HIGH, and
RXCKSEL = MID), the framer will stretch the recovered clock to
the nearest 20-bit boundary such that the rising edge of RXCLK+
occurs when COMDET = HIGH in the output register.
When the Cypress or alternate-mode framer is enabled and
half-rate receive port clocking is also enabled (RFMODE ≠ LOW
and RXRATE = HIGH), the output clock is not modified when
framing is detected, but a single pipeline stage may be added or
subtracted from the data stream by the framer logic such that the
rising edge of RXCLK+ occurs when COMDET = HIGH in the
output register. This adjustment only occurs when the framer is
enabled (RFEN = HIGH). When the framer is disabled, the clock
boundaries are not adjusted, and COMDET may be asserted
during the rising edge of RXCLK– (if an odd number of
characters were received following the initial framing).
Note
23. The RXOP output is also driven from the Output Register, but its interpretation is under the separate control of PARCTL.
Document Number: 38-02031 Rev. *M
Page 21 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Parity Generation
In addition to the eleven data and status bits that are presented,
an RXOP parity output is also available. This allows the
CYP(V)15G0101DXB to support ODD parity generation. To
handle a wide range of system environments, the
CYP(V)15G0101DXB supports different forms of parity generation (in addition to no parity). When the decoder is enabled
(DECMODE ≠ LOW), parity can be generated on
■
the RXD[7:0] character
■
the RXD[7:0] character and RXST[2:0] status.
When the decoder is bypassed (DECMODE = LOW), parity can
be generated on
■
the RXD[7:0] and RXST[1:0] bits
■
the RXD[7:0] and RXST[2:0] bits.
These modes differ in the number of bits which are included in
the parity calculation. For all cases, only ODD parity is provided
which ensures that at least one bit of the data bus is always a
logic-1. Those bits covered by parity generation are listed in
Table 15.
Parity generation is enabled through the 3-level select PARCTL
input. When PARCTL = LOW, parity checking is disabled, and
the RXOP output is disabled (high Z).
Table 15. Output Register Parity Generation
Receive Parity Generate Mode (PARCTL)
Signal
Name
MID
LOW[24]
DECMODE
= LOW
DECMODE
≠ LOW
HIGH
X[25]
RXST[2]
RXST[1]
X
X
RXST[0]
X
X
RXD[0]
X
X
X
RXD[1]
X
X
X
RXD[2]
X
X
X
RXD[3]
X
X
X
RXD[4]
X
X
X
RXD[5]
X
X
X
RXD[6]
X
X
X
RXD[7]
X
X
X
When PARCTL = MID (open) and the decoder is enabled
(DECMODE ≠ LOW), ODD parity is generated for the received
and decoded character in the RXD[7:0] signals and is presented
on the RXOP output.
When PARCTL = MID (open) and the decoder is bypassed
(DECMODE = LOW), ODD parity is generated for the received
and decoded character in the RXD[7:0] and RXST[1:0] bit
positions.
When PARCTL = HIGH, ODD parity is generated for the
TXD[7:0] and the RXST[2:0] status bits.
Receive Status Bits
When the 10B/8B decoder is enabled (DECMODE ≠ LOW),
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 (regardless of the state of
DECMODE)
■
character violations.
These conditions normally overlap; for example, a valid data
character received with incorrect running disparity is not reported
as a valid data character. It is instead reported as a decoder
violation of some specific type. This implies a hierarchy or priority
level to the various status bit combinations. The hierarchy and
value of each status is listed in Table 16.
Within these status decodes, there are three forms of status
reporting. The two normal or data status reporting modes
(Type A and Type B) are selectable through the RXMODE input.
These status types allow compatibility with legacy systems, while
allowing full reporting in new systems. The third status type is
used for reporting receive BIST status and progress.
BIST Status State Machine
When the receive path is enabled to look for and compare the
received data stream with the BIST pattern, the RXST[2:0] bits
identify the present state of the BIST compare operation.
The BIST state machine has multiple states, as shown in
Figure 2 on page 24 and Table 16. When the receive PLL detects
an out-of-lock condition, the BIST state is forced to the
Start-of-BIST state, regardless of the present state of the BIST
state machine. If the number of detected errors ever exceeds the
number of valid matches by greater than 16, the state machine
is forced to the WAIT_FOR_BIST state where it monitors the
interface for the first character (D0.0) of the next BIST sequence.
Also, if the elasticity buffer ever hits and overflow/underflow
condition, the status is forced to the BIST_START until the buffer
is re-centered (approximately nine character periods).
To ensure compatibility between the source and destination
systems when operating in BIST, the sending and receiving ends
of the BIST sequence must use the same clock setup
(RXCKSEL = MID or RXCKSEL = LOW).
Notes
24. Receive path parity output driver (RXOP) is disabled (high Z) when PARCTL = LOW.
25. When the decoder is bypassed (DECMODE = LOW) and BIST is not enabled (Receive BIST latch output is HIGH), RXST[2] is driven to a logic-0, except when the
character in the output buffer is a framing character.
Document Number: 38-02031 Rev. *M
Page 22 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
JTAG Support
JTAG ID
The CYP(V)15G0101DXB contains a JTAG port to allow system
level diagnosis of device interconnect. Of the available JTAG
modes, only boundary scan is supported. This capability is
present only on the LVTTL inputs, LVTTL outputs and the
REFCLK± clock input. The high-speed serial inputs and outputs
are not part of the JTAG test chain.
The JTAG device
“1C804069”x.
ID
for
the
CYP(V)15G0101DXB
is
3-Level Select Inputs
Each 3-level select input 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.
Table 16. Receive Character Status Bits
Description
RXST[2:0] Priority
Type-A Status
Type-B Status
Receive BIST Status
(Receive BIST = Enabled)
000
7
Normal character received. The valid data character on the output bus BIST data compare.
meets all the formatting requirements of data characters listed in Table 20. Character compared correctly
001
7
Special code detected. The valid special character on the output bus meets BIST command compare.
all the formatting requirements of the special code characters listed in Character compared correctly
Table 21, but is not the presently selected framing character or a decoder
violation indication.
010
2
Receive
elasticity
buffer RESERVED
Underrun/Overrun error.
The receive buffer was not able to
add/drop a K28.5 or framing
character.
011
5
Framing character detected. This indicates that a character matching the RESERVED
patterns identified as a framing character (as selected by FRAMCHAR) was
detected. The decoded value of this character is present on the output bus.
100
4
Codeword violation. The character on the output bus is a C0.7. This
indicates that the received character cannot be decoded into any valid
character.
BIST last bad.
Last
character
of
BIST
sequence detected invalid.
101
1
PLL out of lock. This indicates a PLL out of lock condition.
BIST start. Receive BIST is
enabled on this channel, but
character compares have not
yet commenced. This also
indicates a PLL out of lock
condition, and elasticity buffer
overflow/underflow conditions.
110
6
Running disparity error. The character on the output bus is a C4.7, C1.7, BIST error. While comparing
or C2.7.
characters, a mismatch was
found in one or more of the
decoded character bits.
111
3
RESERVED
Document Number: 38-02031 Rev. *M
BIST last good.
Last
character
of
BIST
sequence detected and valid.
BIST wait. The receiver is
comparing characters. but has
not yet found the start of BIST
character to enable the LFSR.
Page 23 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Figure 2. Receive BIST State Machine
Monitor Data
Received
RXST =
BIST_WAIT (111)
RX PLL
Out of Lock
RXST =
BIST_START (101)
Elasticity
Buffer Error
Yes
No
Receive BIST
Detected LOW
RXST =
BIST_START (101)
Start of
BIST Detected
No
Yes, RXST = BIST_DATA_COMPARE (000)
OR BIST_COMMAND_COMPARE(001)
Compare
Next Character
RXST =
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, RXST =
BIST_LAST_BAD (100)
Yes, RXST =
BIST_LAST_GOOD (010)
RXST =
BIST_DATA_COMPARE (000)
No
No, RXST =
BIST_ERROR (110)
Document Number: 38-02031 Rev. *M
Page 24 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Maximum Ratings
Static discharge voltage........................................... > 2000 V
(per MIL-STD-883, method 3015)
Exceeding maximum ratings may shorten the useful life of the
device. User guidelines are not tested.
Latch-up current ..................................................... > 200 mA
Storage temperature................................. –65 °C to +150 °C
Power-up Requirements
Ambient temperature with
power applied ........................................... –55 °C to +125 °C
The CYP(V)15G0101DXB 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 ...............–0.5 V to +3.8 V
DC voltage applied to LVTTL outputs
in high Z State...................................... –0.5 V to VCC + 0.5 V
Operating Range
Output current into LVTTL outputs (LOW) ................... 60 mA
Commercial
DC input voltage .................................. –0.5 V to VCC + 0.5 V
Industrial
Range
Ambient Temperature
VCC
0 °C to +70 °C
+3.3 V ± 5%
–40 °C to +85 °C
+3.3 V ± 5%
DC Electrical Characteristics
Over the Operating Range
Parameter
Description
LVTTL-compatible Outputs
VOHT
Output HIGH voltage
VOLT
Output LOW voltage
IOST
Output short circuit current
IOZL
High Z output leakage current
LVTTL-compatible Inputs
VIHT
Input HIGH voltage
VILT
Input LOW voltage
IIHT
Input HIGH current
IILT
Input LOW current
IIHPDT
Input HIGH current with internal pull-down
IILPUT
Input LOW current with internal pull-up
LVDIFF Inputs: REFCLK±
VDIFF[27]
Input differential voltage
VIHHP
Highest input HIGH voltage
VILLP
Lowest input LOW voltage
VCOMREF[28]
Test Conditions
Min
Max
Unit
2.4
0
–20
–20
VCC
0.4
–100
20
V
V
mA
µA
2.0
–0.5
–
–
–
VCC + 0.3
0.8
1.5
+40
–1.5
V
V
mA
µA
mA
–
–
–
-40
+200
–200
µA
µA
µA
400
1.2
0.0
VCC
VCC
VCC / 2
mV
V
V
1.0
VCC – 1.2
V
Min ≤ VCC ≤ Max
Min ≤ VCC ≤ Max
Min ≤ VCC ≤ Max
VIN = VCC
VIN = VCC/2
VIN = GND
0.87 × VCC
VCC
0.47 × VCC 0.53 × VCC
0.0
0.13 × VCC
–
200
–50
50
–
–200
V
V
V
µA
µA
µA
100 Ω differential load
150 Ω differential load
VCC – 0.5
VCC − 0.5
IOH = − 4 mA, VCC = Min
IOL = 4 mA, VCC = Min
VOUT = 0V[26]
REFCLK Input, VIN = VCC
Other Inputs, VIN = VCC
REFCLK Input, VIN = 0.0 V
Other Inputs, VIN = 0.0 V
VIN = VCC
VIN = 0.0 V
Common mode range
3-Level Inputs
VIHH
3-level input HIGH voltage
VIMM
3-level input MID voltage
VILL
3-level input LOW voltage
IIHH
Input HIGH current
IIMM
Input MID current
IILL
Input LOW current
Differential CML Serial Outputs: OUT1±, OUT2±
VOHC
Output HIGH voltage
(VCC referenced)
VCC − 0.2
VCC − 0.2
V
V
Notes
26. Tested one output at a time, output shorted for less than one second, less than 10% duty cycle.
27. 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.
28. The common mode range defines the allowable range of REFCLK+ and REFCLK− when REFCLK+ = REFCLK−. This marks the zero-crossing between the true
and complement inputs as the signal switches between a logic-1 and a logic-0.
Document Number: 38-02031 Rev. *M
Page 25 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
DC Electrical Characteristics
Over the Operating Range (continued)
Parameter
VOLC
Description
Output LOW voltage
(VCC referenced)
VODIF
Output differential voltage
|(OUT+) − (OUT−)|
Test Conditions
100 Ω differential load
150 Ω differential load
100 Ω differential load
150 Ω differential load
Differential Serial Line Receiver Inputs: IN1±, IN2±
VDIFFS [29]
Input differential voltage |(IN+) − (IN−)|
VIHE
Highest input HIGH voltage
VILE
Lowest input LOW voltage
IIHE
Input HIGH current
IILE
Input LOW current
VCOM [30, 31] Common mode input range
Power Supply
ICC
Power supply current
REFCLK= Max
ICC
Min
VCC − 1.4
VCC − 1.4
450
560
VIN = VIHE Max
VIN = VILE Min
Unit
V
V
mV
mV
100
1200
–
VCC
VCC – 2.0
–
–
1350
–700
–
VCC − 1.95 VCC − 0.05
mV
V
V
µA
µA
V
Typ [33]
390
Commercial
Industrial
Commercial
Industrial
Power supply current
REFCLK = 125 MHz
Max
VCC − 0.7
VCC − 0.7
900
1000
390
Max [32]
500
510
500
510
Unit
mA
mA
mA
mA
AC Test Loads and Waveforms
3.3 V
R1
R1 = 590 Ω
R2 = 435Ω
CL
CL ≤ 7 pF
(Includes fixture and
probe capacitance)
RL = 100 Ω
[34]
(b) CML Output Test Load
R2
(a) LVTTL Output Test Load
[34]
GND
2.0 V
2.0 V
0.8 V
0.8 V
VIHE
VIHE
3.0 V
Vth = 1.4 V
RL
Vth = 1.4 V
≤ 1 ns
≤ 1 ns
VILE
80%
80%
20%
≤ 270 ps
20%
VILE
≤ 270 ps
(d) CML/LVPECL Input Test Waveform
[35]
(c) LVTTL Input Test Waveform
Notes
29. 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.
30. 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.
31. Not applicable for AC-coupled interfaces. For AC-coupled interfaces, VDIFFS requirement still needs to be satisfied.
32. Maximum ICC is measured with VCC = MAX, with all Serial Drivers enabled, parallel outputs unloaded, sending a alternating 01 pattern to the Serial Input Receiver.
33. Typical ICC is measured under similar conditions except with VCC = 3.3V, TA = 25°C, parallel outputs unloaded, RXCKSEL = MID, and with one Serial Line
Driver sending a continuous alternating 01 pattern to the Serial Input Receiver.
34. Cypress uses constant current (ATE) load configurations and forcing functions. This figure is for reference only. 5pF differential load reflects tester capacitance,
and is recommended at low data rates only.
35. The LVTTL switching threshold is 1.4V. All timing references are made relative to the point where the signal edges crosses this threshold voltage.
Document Number: 38-02031 Rev. *M
Page 26 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
CYP(V)15G0101DXB AC Characteristics
Over the Operating Range
Parameter
Description
Transmitter LVTTL Switching Characteristics
fTS
TXCLK clock frequency
tTXCLK
TXCLK period
tTXCLKH[36]
TXCLK HIGH time
tTXCLKL[36]
TXCLK LOW time
[36, 37, 38]
tTXCLKR
TXCLK rise time
tTXCLKF [36, 37, 38] TXCLK fall time
tTXDS
Transmit data det-up time to TXCLK↑ (TXCKSEL ≠ LOW)
tTXDH
Transmit data hold time from TXCLK↑ (TXCKSEL ≠ LOW)
fTOS
TXCLKO clock frequency = 1x or 2x REFCLK frequency
tTXCLKO
TXCLKO period
tTXCLKOD+
TXCLKO+ duty cycle with 60% HIGH time
tTXCLKOD–
TXCLKO– duty cycle with 40% HIGH time
Receiver LVTTL Switching Characteristics
fRS
RXCLK clock output frequency
tRXCLKP
RXCLK period
tRXCLKH
RXCLK HIGH time (RXRATE = LOW)
RXCLK HIGH time (RXRATE = HIGH)
tRXCLKL
RXCLK LOW time (RXRATE = LOW)
RXCLK LOW time (RXRATE = HIGH)
tRXCLKD
RXCLK duty cycle centered at 50%
[36]
tRXCLKR
RXCLK rise time
tRXCLKF[36]
RXCLK fall time
tRXDV–[39]
Status and data valid time to RXCLK (RXCKSEL = MID)
Status and data valid time to RXCLK (HALF RATE RECOVERED CLOCK)
tRXDV+[39]
Status and data valid time from RXCLK (RXCKSEL = MID)
Status and data valid time from RXCLK (HALF RATE RECOVERED CLOCK)
REFCLK Switching Characteristics Over the Operating Range
fREF
REFCLK clock frequency
tREFCLK
REFCLK period
tREFH
REFCLK HIGH time (TXRATE = HIGH)
REFCLK HIGH time (TXRATE = LOW)
tREFL
REFCLK LOW time (TXRATE = HIGH)
REFCLK LOW time (TXRATE = LOW)
[40]
tREFD
REFCLK duty cycle
tREFR[36, 37, 38] REFCLK rise time (20% – 80%)
tREFF[36, 37, 38] REFCLK fall time (20% – 80%)
tTREFDS
Transmit data setup time to REFCLK (TXCKSEL = LOW)
tTREFDH
Transmit data hold time from REFCLK (TXCKSEL = LOW)
Min
Max
Unit
19.5
6.66
2.2
2.2
0.2
0.2
1.7
0.8
19.5
6.66
–1.0
–0.5
150
51.28
–
–
1.7
1.7
–
–
150
51.28
+0.5
+1.0
MHz
ns
ns
ns
ns
ns
ns
ns
MHz
ns
ns
ns
9.75
6.66
2.33 [36]
5.66
2.33 [36]
5.66
–1.0
0.3
0.3
5UI – 1.5
5UI – 1.0
150
102.56
26.64
52.28
26.64
52.28
+1.0
1.2
1.2
–
–
MHz
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
5UI – 1.8
5UI – 2.3
–
–
ns
ns
19.5
6.6
5.9
2.9[36]
5.9
2.9[36]
30
–
–
1.7
0.8
150
51.28
–
–
–
–
70
2
2
–
–
MHz
ns
ns
ns
ns
ns
%
ns
ns
ns
ns
Notes
36. Tested initially and after any design or process changes that may affect these parameters, but not 100% tested.
37. The ratio of rise time to falling time must not vary by greater than 2:1.
38. 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.
39. Parallel data output specifications are only valid if all inputs or outputs are loaded with similar DC and AC loads.
40. The duty cycle specification is a simultaneous condition with the tREFH and tREFL parameters. This means that at faster character rates the REFCLK duty cycle
cannot be as large as 30%–70%.
Document Number: 38-02031 Rev. *M
Page 27 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
CYP(V)15G0101DXB AC Characteristics
Over the Operating Range (continued)
Parameter
Description
[43]
tRREFDA
Receive data access time from REFCLK (RXCKSEL = LOW)
tRREFDV
Receive data valid time from REFCLK (RXCKSEL = LOW)
tREFDV–
Received data valid time to RXCLK (RXCKSEL = LOW)
tREFDV+
Received data valid time from RXCLK (RXCKSEL = LOW)
tREFCDV–
Received data valid time to RXCLKC (RXCKSEL = LOW)
tREFCDV+
Received data valid time from RXCLKC (RXCKSEL = LOW)
tREFRX[41, 42]
REFCLK frequency referenced to extracted received clock frequency
Transmit Serial Outputs and TX PLL Characteristics
tB
Bit time
tRISE[42]
CML output rise time 20%–80% (CML test load)
SPDSEL = HIGH
SPDSEL = MID
SPDSEL = LOW
tFALL[42]
CML output fall time 80%–20% (CML test load)
SPDSEL = HIGH
SPDSEL = MID
SPDSEL = LOW
tDJ[42, 44, 46]
Deterministic Jitter (peak-peak)
IEEE 802.3z[47]
[42, 45, 46]
tRJ
Random Jitter (σ)
IEEE 802.3z[47]
tTXLOCK
Transmit PLL lock to REFCLK
Receive Serial Inputs and CDR PLL Characteristics
tRXLOCK
Receive PLL lock to input data stream (cold start)
Receive PLL lock to input data stream
tRXUNLOCK
Receive PLL unlock rate
tJTOL[46]
Total jitter tolerance
IEEE 802.3z[47]
tDJTOL[46]
Deterministic jitter tolerance
IEEE 802.3z[47]
[42]
Capacitance
Parameter
CINTTL
CINPECL
Description
TTL input capacitance
PECL input capacitance
Min
–
2.5
10UI – 4.7
0.5
10UI – 4.3
–0.2
–1500
Max
9.5
–
–
–
–
–
+1500
Unit
ns
ns
ns
ns
ns
ns
ppm
5100
60
100
180
60
100
180
–
–
–
666
270
500
1000
270
500
1000
25
11
200
ps
ps
ps
ps
ps
ps
ps
ps
ps
us
–
–
–
600
370
376K
376K
46
–
–
UI[48]
UI
UI
ps
ps
Max
7
4
Unit
pF
pF
Test Conditions
TA = 25 °C, f0 = 1 MHz, VCC = 3.3 V
TA = 25 °C, f0 = 1 MHz, VCC = 3.3 V
Notes
41. REFCLK has no phase or frequency relationship with the recovered clock(s) and only acts as a centering reference to reduce clock synchronization time. REFCLK
must be within ±1500 ppm (±0.15%) of the remote transmitter’s PLL reference (REFCLK) 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.
42. Tested initially and after any design or process changes that may affect these parameters, but not 100% tested.
43. Since this timing parameter is greater than the minimum time period of REFCLK it sets an upper limit to the frequency in which REFCLKx can be used to clock
the receive data out of the output register. For predictable timing, users can use this parameter only if REFCLK period is greater than sum of tRREFDA and set-up
time of the upstream device. When this condition is not true, RXCLKC± or RXCLKA± (a buffered or delayed version of REFCLK when RXCKSELx = LOW) could
be used to clock the receive data out of the device.
44. While sending continuous K28.5s, outputs loaded to a balanced 100Ω load, measured at the cross point of the differential outputs over the operating range.
45. While sending continuous K28.7s, after 100,000 samples measured at the cross point of differential outputs, time referenced to REFCLK input, over the operating
range.
46. Total jitter is calculated at an assumed BER of 1E − 12. Hence: Total Jitter (tJ) = (tRJ * 14) + tDJ.
47. Also meets all Jitter Generation and Jitter Tolerance requirements as specified by SMPTE 259M, SMPTE 292M, CPRI, ESCON, FICON, Fibre Channel and
DVB-ASI.
48. Receiver UI (Unit Interval) is calculated as 1/(fREF * 20) (when RXRATE = HIGH) or 1/(fREF * 10) (when RXRATE = LOW) if no data is being received, or 1/(fREF * 20)(when
RXRATE = HIGH) or 1/(fREF * 10) (when RXRATE = LOW) of the remote transmitter if data is being received. In an operating link this is equivalent to tB
Document Number: 38-02031 Rev. *M
Page 28 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Switching Waveforms for the HOTLink II Transmitter
Transmit Interface
Write Timing
TXCKSEL ≠ LOW
tTXCLK
tTXCLKH
tTXCLKL
TXCLK
tTXDS
TXD[7:0],
TXCT[1:0],
TXOP,
SCSEL
tTXDH
Transmit Interface
Write Timing
TXCKSEL = LOW
TXRATE = LOW
tREFCLK
tREFH
tREFL
REFCLK
tTREFDS
TXD[7:0],
TXCT[1:0],
TXOP,
SCSEL
tTREFDH
Transmit Interface
Write Timing
TXCKSEL = LOW
TXRATE = HIGH
REFCLK
tREFCLK
tREFH
tREFL
Note 49
Note 49
tTREFDS
tTREFDS
TXD[7:0],
TXCT[1:0],
TXOP,
SCSEL
tTREFDH
tTREFDH
tREFCLK
Transmit Interface
TXCLKO Timing
TXCKSEL = LOW
TXRATE = HIGH
tREFH
tREFL
REFCLK
Note 51
tTXCLKO
tTXCLKOD+
tTXCLKOD–
Note 50
TXCLKO
Notes
49. When REFCLK is configured for half-rate operation (TXRATE = HIGH) and data is captured using REFCLK instead of TXCLK clock (TXCKSEL = LOW), data
is captured using both the rising and falling edges of REFCLK.
50. The TXCLKO output is at twice the rate of REFCLK when TXRATE = HIGH and same rate as REFCLK when TXRATE = LOW. TXCLKO does not follow the duty
cycle of REFCLK.
51. The rising edge of TXCLKO output has no direct phase relationship to the REFCLK input.
Document Number: 38-02031 Rev. *M
Page 29 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Transmit Interface
TXCLKO Timing
TXCKSEL = LOW
TXRATE = LOW
tREFCLK
tREFH
tREFL
Note 50
REFCLK
tTXCLKO
Note
52
tTXCLKOD+
tTXCLKOD–
TXCLKO
Receive Interface
Read Timing
RXCKSEL = LOW
TXRATE = LOW
tREFCLK
tREFH
tREFL
REFCLK
tRREFDV
tRREFDA
RXD[7:0],
RXST[2:0],
RXOP
tREFDV+
tREFCDV+
tREFDV–
tREFCDV–
Note 53
RXCLK
RXCLKC+
Receive Interface
Read Timing
RXCKSEL = LOW
TXRATE = HIGH
tREFCLK
tREFH
tREFL
REFCLK
tRREFDA
tRREFDA
tRREFDV
tRREFDV
RXD[7:0],
RXST[2:0],
RXOP
tREFDV+
tREFCDV+
RXCLK
RXCLKC+
Note 53
tREFDV–
tREFCDV–
Note 54
Notes
52. The rising edge of TXCLKO output has no direct phase relationship to the REFCLK input.
53. RXCLK and RXCLK+ are delayed versions of REFCLK when RXCKSEL = LOW, and are different in phase from each other.
54. When operated with a half-rate REFCLK, the setup and hold specifications for data relative to RXCLK are relative to both rising and falling edges of the clock output
Document Number: 38-02031 Rev. *M
Page 30 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Receive Interface
Read Timing
RXCKSEL = MID
RXRATE = LOW
tRXCLKP
tRXCLKH
tRXCLKL
RXCLK+
RXCLK–
tRXDV–
RXD[7:0],
RXST[2:0],
RXOP
tRXDV+
Receive Interface
Read Timing
RXCKSEL = MID
RXRATE = HIGH
tRXCLKP
tRXCLKH
tRXCLKL
RXCLK+
RXCLK–
tRXDV–
RXD[7:0],
RXST[2:0],
RXOP
tRXDV+
Document Number: 38-02031 Rev. *M
Page 31 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Table 17. Package Coordinate Signal Allocation
Ball
ID
Signal Name
Signal Type
Ball
ID
Signal Name
Signal Type
Ball
ID
Signal Name
Signal Type
A1
VCC
POWER
D5
GND
GROUND
G9
TXCLKO+
LVTTL OUT
A2
IN2+
CML IN
D6
GND
GROUND
G10
TXCLKO–
LVTTL OUT
A3
VCC
POWER
D7
GND
GROUND
H1
RXD[0]
LVTTL OUT
A4
OUT2–
CML OUT
D8
TMS
LVTTL IN PU
H2
RXD[2]
LVTTL OUT
A5
RXMODE
3-LEVEL SEL
D9
TRSTZ
LVTTL IN PU
H3
RXD[6]
LVTTL OUT
A6
TXMODE[1]
3-LEVEL SEL
D10
TDI
LVTTL IN PU
H4
LFI
LVTTL OUT
A7
IN1+
CML IN
E1
BISTLE
LVTTL IN PU
H5
TXCT[1]
LVTTL IN
A8
VCC
POWER
E2
DECMODE
3-LEVEL SEL
H6
TXD[6]
LVTTL IN
A9
OUT1–
CML OUT
E3
OELE
LVTTL IN PU
H7
TXD[3]
LVTTL IN
A10
VCC
POWER
E4
GND
GROUND
H8
TXCLK
LVTTL IN PD
B1
VCC
POWER
E5
GND
GROUND
H9
TXRST
LVTTL IN PU
B2
IN2–
CML IN
E6
GND
GROUND
H10
#NC
NO CONNECT
B3
TDO
LVTTL 3-S OUT
E7
GND
GROUND
J1
VCC
POWER
B4
OUT2+
CML OUT
E8
TCLK
LVTTL IN PD
J2
RXD[3]
LVTTL OUT
B5
TXRATE
LVTTL IN PD
E9
RXCKSEL
3-LEVEL SEL
J3
RXD[7]
LVTTL OUT
B6
TXMODE[0]
3-LEVEL SEL
E10
TXCKSEL
3-LEVEL SEL
J4
RXCLK–
LVTTL OUT
B7
IN1–
CML IN
F1
RXST[2]
LVTTL OUT
J5
TXCT[0]
LVTTL IN
B8
#NC
NO CONNECT
F2
RXST[1]
LVTTL OUT
J6
TXD[5]
LVTTL IN
B9
OUT1+
CML OUT
F3
RXST[0]
LVTTL OUT
J7
TXD[2]
LVTTL IN
B10
VCC
POWER
F4
GND
GROUND
J8
TXD[0]
LVTTL IN
C1
RFEN
LVTTL IN PD
F5
GND
GROUND
J9
#NC
NO CONNECT
C2
LPEN
LVTTL IN PD
F6
GND
GROUND
J10
VCC
POWER
C3
RXLE
LVTTL IN PU
F7
GND
GROUND
K1
VCC
POWER
C4
RXCLKC+
LVTTL 3-S OUT
F8
TXPER
LVTTL OUT
K2
RXD[4]
LVTTL OUT
C5
RXRATE
LVTTL IN PD
F9
REFCLK–
PECL IN
K3
VCC
POWER
C6
SDASEL
3-LEVEL SEL
F10
REFCLK+
PECL IN
K4
RXCLK+
LVTTL OUT
C7
SPDSEL
3-LEVEL SEL
G1
RXOP
LVTTL 3-S OUT
K5
TXD[7]
LVTTL IN
C8
PARCTL
3-LEVEL SEL
G2
RXD[1]
LVTTL OUT
K6
TXD[4]
LVTTL IN
C9
RFMODE
3-LEVEL SEL
G3
RXD[5]
LVTTL OUT
K7
TXD[1]
LVTTL IN
C10
INSEL
LVTTL IN
G4
GND
GROUND
K8
VCC
POWER
D1
BOE[0]
LVTTL IN PU
G5
GND
GROUND
K9
SCSEL
LVTTL IN PD
D2
BOE[1]
LVTTL IN PU
G6
GND
GROUND
K10
VCC
POWER
D3
FRAMCHAR
3-LEVEL SEL
G7
GND
GROUND
D4
GND
GROUND
G8
TXOP
LVTTL IN PU
Document Number: 38-02031 Rev. *M
Page 32 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
X3.230 Codes and Notation Conventions
Information to be 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
eight-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 eight-bit byte. Fibre Channel
Standard notation uses a bit notation of A, B, C, D, E, F, G, H for
the eight-bit byte for the raw eight-bit data, and the letters a, b,
c, d, e, i, f, g, h, j for encoded 10-bit data. There is a correspondence between bit A and bit a, B and b, C and c, D and d, E and
e, F and f, G and g, and H and h. Bits i and j are derived, respectively, from (A,B,C,D,E) and (F,G,H).
The bit labeled A in the description of the 8B/10B Transmission
Code corresponds to bit 0 in the numbering scheme of the FC-2
specification, B corresponds to bit 1, as shown below.
FC-2 bit designation—76543210
HOTLink D/Q designation—76543210
8B/10B bit designation—HGFEDCBA
c is set to D, xx is the decimal value of the binary number
composed of the bits E, D, C, B, and A in that order, and the y is
the decimal value of the binary number composed of the bits H,
G, and F in that order. When c is set to K, xx and y are derived
by comparing the encoded bit patterns of the Special Character
to those patterns derived from encoded Valid Data bytes and
selecting the names of the patterns most similar to the encoded
bit patterns of the Special Character.
Under the above conventions, the Transmission Character used
for the examples above, is referred to by the name D5.2. The
Special Character K29.7 is so named because the first six bits
(abcdei) of this character make up a bit pattern similar to that
resulting from the encoding of the unencoded 11101 pattern (29),
and because the second four bits (fghj) make up a bit pattern
similar to that resulting from the encoding of the unencoded 111
pattern (7). This definition of the 10-bit Transmission Code is
based on the following references.
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
X3.230-1994 ANSI FC-PH Standard).
Interface
(ANS
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 to be followed
when transmitting the bits within a character and the characters
within any higher-level constructs specified by a standard.
Transmission Order
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
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.
Converted to 8B/10B notation, note that the order of bits has
been reversed):
Data Byte Name D5.2
Bits: ABCDE FGH
10100 010
Valid and Invalid Transmission Characters
Note that bit i is transmitted between bit e and bit f, rather than
in alphabetical order.
Translated to a transmission Character in the 8B/10B Transmission Code:
Bits: abcdei fghj
101001 0101
The following tables define the valid Data Characters and valid
Special Characters (K characters), respectively. The tables are
used for both generating valid Transmission Characters and
checking the validity of received Transmission Characters. In the
tables, each Valid-Data-byte or Special-Character-code entry
has two columns that represent two Transmission Characters.
The two columns correspond to the current value of the running
disparity. Running disparity is a binary parameter with either a
negative (–) or positive (+) value.
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
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 will
select the proper version of the Transmission Character based
on the current running disparity value, and the Transmitter
Document Number: 38-02031 Rev. *M
Page 33 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
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.
disparity is calculated. This new value is used as the Transmitter’s current running disparity for the next Valid Data byte or
Special Character byte to be encoded and transmitted. Table 18
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.
Use of the Tables for Checking the Validity of Received
Transmission Characters
The column corresponding to the current value of the Receiver’s
running disparity is searched for the received Transmission
Character. If the received Transmission Character is found in the
proper column, then the Transmission Character is valid and the
Data byte or Special Character code is determined (decoded). If
the received Transmission Character is not found in that column,
then the Transmission Character is invalid. This is called a code
violation. Independent of the Transmission Character’s validity,
the received Transmission Character is used to calculate a new
value of running disparity. The new value is used as the
Receiver’s current running disparity for the next received Transmission Character.
The following rules for running disparity are used to calculate the
new running-disparity value for Transmission Characters that
have been transmitted (Transmitter’s running disparity) and that
have been received (Receiver’s running disparity).
Running disparity for a Transmission Character is calculated
from sub-blocks, where the first six bits (abcdei) form one
sub-block and the second four bits (fghj) form the other
sub-block. Running disparity at the beginning of the six-bit
sub-block is the running disparity at the end of the previous
Transmission Character. Running disparity at the beginning of
the four-bit sub-block is the running disparity at the end of the
six-bit sub-block. Running disparity at the end of the Transmission Character is the running disparity at the end of the
four-bit sub-block.
Table 18. Valid Transmission Characters
Data
Byte Name
Running disparity for the sub-blocks is calculated as follows:
1. Running disparity at the end of any sub-block is positive if the
sub-block contains more ones than zeros. It is also positive at
the end of the six-bit sub-block if the six-bit sub-block is
000111, and it is positive at the end of the four-bit sub-block if
the four-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 six-bit sub-block if the six-bit sub-block is
111000, and it is negative at the end of the four-bit sub-block
if the four-bit sub-block is 1100.
3. Otherwise, running disparity at the end of the sub-block is the
same as at the beginning of the sub-block.
Use of the Tables for Generating Transmission
Characters
DIN or QOUT
Hex Value
765
43210
D0.0
000
00000
00
D1.0
000
00001
01
D2.0
000
00010
02
.
.
.
.
.
.
.
.
D5.2
010
00101
45
.
.
.
.
.
.
.
.
D30.7
111
11110
FE
D31.7
111
11111
FF
Detection of a code violation does not necessarily show that the
Transmission Character in which the code violation was detected
is in error. Code violations may result from a prior error that
altered the running disparity of the bit stream which did not result
in a detectable error at the Transmission Character in which the
error occurred. Table 19 shows an example of this behavior.
The appropriate entry in the table is found for the Valid Data byte
or the Special Character byte for which a Transmission
Character is to be generated (encoded). 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
Table 19. 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 Number: 38-02031 Rev. *M
Page 34 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Table 20. 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
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
Document Number: 38-02031 Rev. *M
Page 35 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Table 20. 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
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
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
Document Number: 38-02031 Rev. *M
Page 36 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Table 20. 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
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
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
Document Number: 38-02031 Rev. *M
Page 37 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Table 20. 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
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 Number: 38-02031 Rev. *M
Page 38 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Table 21. Valid Special Character Codes and Sequences (TXCTx = Special Character Code or RXSTx[2:0] = 001)[55,56]
S.C. Byte Name
Cypress
S.C. Code Name
S.C. Byte
Name[57]
Bits
HGF EDCBA
Alternate
S.C. Byte
Name[57]
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[58]
C1.0
(C01)
000 00001
C28.1
(C3C)
001 11100
001111 1001
110000 0110
[58]
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
C4.0
(C04)
000 00100
C28.4
(C9C)
100 11100
001111 0010
110000 1101
K28.5[58, 59]
C5.0
(C05)
000 00101
C28.5
(CBC)
101 11100
001111 1010
110000 0101
K28.6[58]
C6.0
(C06)
000 00110
C28.6
(CDC)
110 11100
001111 0110
110000 1001
K28.7[58, 60]
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
001 00010
C2.1
(C22)
001 00010
−K28.5,Dn.xxx0
+K28.5,Dn.xxx1
C0.7
(CE0)
111 00000[66]
100111 1000
011000 0111
00001[66]
[58]
End of Frame Sequence
EOFxx[61]
C2.1
(C22)
Code Rule Violation and SVS Tx Pattern
Exception[60, 62] C0.7
(CE0)
111 00000
−K28.5[63]
C1.7
(CE1)
111 00001
C1.7
(CE1)
111
001111 1010
001111 1010
+K28.5[64]
C2.7
(CE2)
111 00010
C2.7
(CE2)
111 00010[66]
110000 0101
110000 0101
C4.7
(CE4)
111 00100[66]
110111 0101
001000 1010
Running Disparity Violation Pattern
Exception[65]
C4.7
(CE4)
111 00100
Notes
55. All codes not shown are reserved.
56. 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).
57. 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 DECMODE configuration input.
58. 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.
59. 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.
60. Care must be taken when using this Special Character code. When a K28.7(C7.0) or SVS(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 RFEN = HIGH.
61. 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 will send 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 will send either K28.5−D10.4−D21.4− D21.4 or K28.5−D10.5−D21.4− D21.4 based on Current RD.
The receiver will never output this Special Character, since K28.5 is decoded as C5.0, C1.7, or C2.7, and the subsequent bytes are decoded as data.
62. C0.7 = Transmit a deliberate code rule violation. The code chosen for this function follows the normal Running Disparity rules. The receiver will only output this
Special Character if the Transmission Character being decoded is not found in the tables.
63. C1.7 = Transmit Negative K28.5 (−K28.5+) disregarding Current RD. The receiver will only output this Special Character if K28.5 is received with the wrong
running disparity. The receiver will output C1.7 if −K28.5 is received with RD+, otherwise K28.5 is decoded as C5.0 or C2.7.
64. C2.7 = Transmit Positive K28.5 (+K28.5−) disregarding Current RD. The receiver will only output this Special Character if K28.5 is received with the wrong
running disparity. The receiver will output C2.7 if +K28.5 is received with RD−, otherwise K28.5 is decoded as C5.0 or C1.7.
65. C4.7 = Transmit a deliberate code rule violation to indicate a Running Disparity violation. The receiver will only output 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.
66. Supported only for data transmission. The receive status for these conditions will be reported by specific combinations of receive status bits.
Document Number: 38-02031 Rev. *M
Page 39 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Ordering Information
Speed
Ordering Code
Package Name
Package Type
Operating Range
Standard
CYP15G0101DXB-BBC
BB100
100-ball Grid Array
Commercial
Standard
CYP15G0101DXB-BBXC
BB100
Pb-free 100-ball Grid Array
Commercial
Standard
CYP15G0101DXB-BBXI
BB100
Pb-free 100-ball Grid Array
Industrial
Standard
CYV15G0101DXB-BBXC
BB100
Pb-free 100-ball Grid Array
Commercial
Ordering Code Definitions
CY X15G0101 DXB - BB X X
X = Temperature Grade = C or I (C = Commercial; I = Industrial)
X = Pb-free
BB = 100-ball BGA
Fixed Value
Part Identifier
Company ID: CY = Cypress
Document Number: 38-02031 Rev. *M
Page 40 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Package Diagram
Figure 3. 100-ball Thin Ball Grid Array (11 × 11 × 1.4 mm) BB100
51-85107 *C
Document Number: 38-02031 Rev. *M
Page 41 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Acronyms
Document Conventions
The following table lists the acronyms that are used in this
document.
Table 23. Units of Measure
Table 22. Acronyms Used in this Datasheet
Acronym
BGA
Description
ball grid array
Acronym
°C
Description
degree Celsius
kΩ
Kilo ohm
µA
micro Amperes
µs
micro second
milli Amperes
BIST
built-in self test
I/O
input/output
JTAG
joint test action group
mA
PLL
phase-locked loop
ms
milli second
TMS
test mode select
mV
milli Volts
TDO
test data out
nA
nano Amperes
TDI
test data in
Ω
ohm
pF
pico Farad
V
Volts
W
Watts
Document Number: 38-02031 Rev. *M
Page 42 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Document History Page
Document Title: CYP15G0101DXB/CYV15G0101DXB Single-channel HOTLink II™ Transceiver
Document Number: 38-02031
Revision
ECN
Submission
Date
Orig. of
Change
Description of Change
**
113123
05/20/02
TPS
New Data Sheet
*A
119704
10/30/02
LNM
Changed TXPER description
Changed TXCLKO description
Changed RXCKSEL to include RXCLKC+
Removed disparity reference from RFMODE
Removed the LOW setting for FRAMCHAR and related references
Removed references to ATM transport
Changed the IOST boundary values
Changed VODIF and VOLC for CML output
Changed the tTXCLKR and tTXCLKF min. values
Changed tTXDS, tTXDH, tTREFDS, and tTREFDH
Changed tREFDV–, tREFCDV–, and tREFCDV+
Changed the JTAG ID from 0C804069 to 1C804069
Added a section for characterization and standards compliance
Changed I/O type of RXCLKC in I/O coordinates table
*B
122209
12/28/02
RBI
Minor Change Document Control corrected Document History Page
*C
122546
02/13/03
CGX
Changed Minimum tRISE/tFALL for CML
Changed tRXLOCK
Changed tDJ, tRJ
Changed tJTOL
Changed tTXLOCK
Changed tRXCLKH, tRXCLKL
Changed tTXCLKOD+, tTXCLKODChanged Power Specs
Changed verbiage...Paragraph: Clock/Data Recovery
Changed verbiage...Paragraph: Range Control
Added Power-up Requirements
*D
124994
04/15/03
POT
Changed CYP15G0101DXB to CYP(V)15G0101DXB type corresponding to
the Video-compliant parts
Reduced the lower limit of the serial signaling rate from 200 Mbaud to
195 Mbaud and changed the associated specifications accordingly
*E
128366
7/3/03
PDS
Revised the value of tRREFDV, tREFADV+ and tREFCDV+
*F
128835
7/31/03
KKV
Minor change: corrections due to editorial error - old file used for *E revision
(reestablishing *D changes)
*G
131898
12/10/03
PDS
When TXCKSEL = MID or HIGH, TXRATE = HIGH is an invalid mode. Made
appropriate changes to reflect this invalid condition
Removed requirement of AC coupling for Serial I/Os for interfacing with
LVPECL I/Os
Changed LFI to Asynchronous output
Expanded the CDR Range Controller’s permissible frequency offset between
incoming serial signaling rate and Reference clock from ±200-PPM to
±1500-PPM (changed parameter tREFRX)
*H
211461
See ECN
KKV
Minor change: Package diagram isn’t legible in pdf
*I
230621
See ECN
LAR
Updated package information in features list to reflect correct package
*J
338721
See ECN
SUA
Added CYW15G0101DXB part number for OBSAI RP3 compliance to support
operating data rate upto 1540 MBaud. Made changes to reflect OBSAI RP3
and CPR compliance. Added Pb-Free Package option for all parts listed in the
datasheet.
Changed MBd to MBaud in SPDSEL pin description
*K
2898355
03/24/2010
CGX
Removed inactive parts from Ordering Information.
Updated Packaging Information.
Document Number: 38-02031 Rev. *M
Page 43 of 44
[+] Feedback
CYP15G0101DXB
CYV15G0101DXB
Document Title: CYP15G0101DXB/CYV15G0101DXB Single-channel HOTLink II™ Transceiver
Document Number: 38-02031
Revision
ECN
Submission
Date
Orig. of
Change
*L
3053045
10/08/2010
CGX
Updated Ordering Information and added Ordering Code Definitions.
Added Acronyms and Document Conventions.
Minor edits and updated in new template.
*M
3243269
04/28/2011
SAAC
Removed the following part numbers from the Ordering Information on page 40
since the parts are no longer active. (CDT 98703)
a) CYP15G0101DXB-BBI
b) CYV15G0101DXB-BBI
c) CYW15G0101DXB-BBXI
Removed all references to the part CYW15G0101DXB from the datasheet
(CDT 98703).
Description of Change
Sales, Solutions, and Legal Information
Worldwide Sales and Design Support
Cypress maintains a worldwide network of offices, solution centers, manufacturer’s representatives, and distributors. To find the office
closest to you, visit us at Cypress Locations.
Products
Automotive
Clocks & Buffers
Interface
Lighting & Power Control
PSoC Solutions
cypress.com/go/automotive
cypress.com/go/clocks
psoc.cypress.com/solutions
cypress.com/go/interface
PSoC 1 | PSoC 3 | PSoC 5
cypress.com/go/powerpsoc
cypress.com/go/plc
Memory
Optical & Image Sensing
cypress.com/go/memory
cypress.com/go/image
PSoC
cypress.com/go/psoc
Touch Sensing
cypress.com/go/touch
USB Controllers
Wireless/RF
cypress.com/go/USB
cypress.com/go/wireless
© Cypress Semiconductor Corporation, 2002-2011. The information contained herein is subject to change without notice. Cypress Semiconductor Corporation assumes no responsibility for the use of
any circuitry other than circuitry embodied in a Cypress product. Nor does it convey or imply any license under patent or other rights. Cypress products are not warranted nor intended to be used for
medical, life support, life saving, critical control or safety applications, unless pursuant to an express written agreement with Cypress. Furthermore, Cypress does not authorize its products for use as
critical components in life-support systems where a malfunction or failure may reasonably be expected to result in significant injury to the user. The inclusion of Cypress products in life-support systems
application implies that the manufacturer assumes all risk of such use and in doing so indemnifies Cypress against all charges.
Any Source Code (software and/or firmware) is owned by Cypress Semiconductor Corporation (Cypress) and is protected by and subject to worldwide patent protection (United States and foreign),
United States copyright laws and international treaty provisions. Cypress hereby grants to licensee a personal, non-exclusive, non-transferable license to copy, use, modify, create derivative works of,
and compile the Cypress Source Code and derivative works for the sole purpose of creating custom software and or firmware in support of licensee product to be used only in conjunction with a Cypress
integrated circuit as specified in the applicable agreement. Any reproduction, modification, translation, compilation, or representation of this Source Code except as specified above is prohibited without
the express written permission of Cypress.
Disclaimer: CYPRESS MAKES NO WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, WITH REGARD TO THIS MATERIAL, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. Cypress reserves the right to make changes without further notice to the materials described herein. Cypress does not
assume any liability arising out of the application or use of any product or circuit described herein. Cypress does not authorize its products for use as critical components in life-support systems where
a malfunction or failure may reasonably be expected to result in significant injury to the user. The inclusion of Cypress’ product in a life-support systems application implies that the manufacturer
assumes all risk of such use and in doing so indemnifies Cypress against all charges.
Use may be limited by and subject to the applicable Cypress software license agreement.
Document Number: 38-02031 Rev. *M
Revised April 29, 2011
Page 44 of 44
HOTLink is a registered trademark, and HOTLink II and MultiFrame are trademarks, of Cypress Semiconductor Corporation. CPRI is a trademark of Siemens AG. IBM and ESCON are registered
trademarks, and FICON is a trademark, of International Business Machines.
All products and company names mentioned in this document may be the trademarks of their respective holders.
[+] Feedback