View CYP15G0401RB-BGC_3290008.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

preliminary quad hotlink ii? receive r cyp15g0401rb cypress semiconductor corporation ? 3901 north first street ? san jose , ca 95134 ? 408-943-2600 document #: 38-02111 rev. ** revised february 14, 2005 features ? quad receiver for 195 to 1500 mbaud serial signaling rate ? aggregate throughput of 6 gbits/second ? second-generation hotlink ? technology ? compliant to multiple standards ? escon, dvb-asi, fibre channel and gigabit ethernet (ieee802.3z) ? 8b/10b encoded or 10-bit unencoded data ? selectable parity generate ? 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 interface ? optional elasticity buffer in receive path ? internal clock/data recovery (cdr) plls with no external pll components ? dual differential pecl-compatible serial inputs per channel ? internal dc-restoration ? 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 2.1w @ 3.3v typical ? single 3.3v supply ? 256-ball thermally enhanced bga ? pb free package available ? 0.25 bicmos technology functional description the cyp15g0401rb quad hotlink ii? receiver is a point-to-point or point-to-mult ipoint communications building block allowing the transfer of data over high-speed serial links (optical fiber, balanced, and unbalanced copper transmission lines) at signaling speeds ranging from 195-to-1500 mbaud per serial link. each receive channel accepts serial data and converts it to parallel data, decodes the data into characters, and presents these characters to an output register. figure 1 illustrates typical connections between independent host systems and corresponding cyp15g0401tb and cyp15g0401rb parts. figure 1. hotlink ii system connections system host serial link 10 10 10 10 system host 10 10 10 serial link serial link serial link backplane or cabled connections cyp15g0401rb cyp15g0401tb 10
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 2 of 35 as a second-generation hotlink device, the cyp15g0401rb extends the hotlink family with enhanced levels of integration and fast er data rates, while maintaining serial-link compatibility (data, command, and bist) with other hotlink devices. the receivers (rx) of the cyp15g0401rb quad hotlink ii consist of four byte-wide channels. each 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. each recovered serial 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 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 t he highest flexibility in system architecture. the receive interface may be configured to present data relative to a recovered clock or to a local training clock. each receive channel contains an independent bist pattern checker. this bist hardware allows at-speed testing of the high-speed serial data paths in each receive section, and across the interconnecting links. 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, servers and video transmission systems.
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 3 of 35 cyp15g0401rb receiver logic block diagram decoder 8b/10b framer deserializer rx decoder 8b/10b framer deserializer rx decoder 8b/10b framer deserializer rx decoder 8b/10b framer deserializer rx rxda[7:0] rxdb[7:0] rxdc[7:0] rxdd[7:0] ina1 ina2 inb1 inb2 inc1 inc2 ind1 ind2 elasticity buffer elasticity buffer elasticity buffer elasticity buffer rxsta[2:0] rxstb[2:0] rxstc[2:0] rxstd[2:0] x11 x11 x11 x11
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 4 of 35 ina1+ ina1? ina2+ ina2? insela inb1+ inb1? inb2+ inb2? inselb inc1+ inc1? inc2+ inc2? inselc ind1+ ind1? ind2+ ind2? inseld character-rate clock clock & data recovery pll shifter clock & data recovery pll shifter clock & data recovery pll shifter clock & data recovery pll shifter lfid lfic lfib lfia 8 rxstc[2:0] rxdc[7:0] rxopc 3 8 rxstb[2:0] rxdb[7:0] rxopb 3 8 rxstd[2:0] rxdd[7:0] rxopd 3 8 rxsta[2:0] rxda[7:0] rxopa 3 receive signal monitor receive signal monitor receive signal monitor receive signal monitor output register output register output register output register elasticity buffer framer rxclkd+ rxclkd? 10b/8b bist elasticity buffer 10b/8b bist framer elasticity buffer 10b/8b bist framer elasticity buffer 10b/8b bist framer 2 rxclkc+ rxclkc? 2 rxclkb+ rxclkb? 2 rxclka+ rxclka? 2 rxrate framchar rfmode rfen sdasel jtag boundary scan controller tdo tms tclk tdi clock select clock select clock select clock select rxcksel trstz decmode receive path block diagram = internal signal rx pll enable latch rxle bre[3:0] bist enable latch bre[3:0] bistle trgrate spdsel trgclk+ trgclk? clock multiplier parctl
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 5 of 35 pin configuration (top view) [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 a inc1- n/c inc2- n/c v cc ind1- n/c gnd ind2- n/c ina1- n/c gnd ina2- n/c v cc inb1- n/c inb2- n/c b inc1+ n/c inc2+ n/c v cc ind1+ n/c gnd ind2+ n/c ina1+ n/c gnd ina2+ n/c v cc inb1+ n/c inb2+ n/c c tdi tms inselc inselb v cc par ctl sda sel gnd n/c n/c n/c n/c gnd n/c gnd v cc trg rate rx rate gnd tdo d tclk trstz inseld insela v cc rf mode spd sel gnd bre[3] bre[2] bre[1] bre[0] gnd n/c gnd v cc v cc rxle rfen n/c e v cc v cc v cc v cc v cc v cc v cc v cc f n/c v cc v cc rxck sel bistle rxstb [1] rxopb rxstb [0] g gndgndgndgnd dec mode gnd fram char rxdb [1] h gndgndgndgnd gndgndgndgnd j gndgndgndgnd rxstb [2] rxdb [0] rxdb [5] rxdb [2] k rxdc [2] rxclk c? gnd lfic rxdb [3] rxdb [4] rxdb [7] rxclk b+ l rxdc [3] rxclk c+ gnd gnd rxdb [6] lfib rxclk b? gnd m rxdc [4] rxdc [5] rxdc [7] rxdc [6] gndgndgndgnd n gndgndgndgnd gndgndgndgnd p rxdc [1] rxdc [0] rxstc [0] rxstc [1] gndgndgndgnd r rxstc [2] rxop c n/c v cc v cc v cc v cc n/c t v cc v cc v cc v cc v cc v cc v cc v cc u v cc v cc v cc v cc v cc rxdd [2] rxdd [1] gnd rx opd n/c trg clk- gnd gnd gnd v cc v cc rxda [2] rxopa rxsta [2] rxsta [1] v v cc v cc v cc rxdd [6] v cc rxdd [3] rxstd [0] gnd rxstd [2] n/c trg clk+ n/c gnd gnd v cc v cc rxda [7] rxda [3] rxda [0] rxsta [0] w v cc v cc lfid rxclk d? v cc rxdd [4] rxstd [1] gnd n/c gnd gnd gnd gnd gnd v cc v cc lfia rxclk a- rxda [4] rxda [1] y v cc v cc rxdd [7] rxclk d+ v cc rxdd [5] rxdd [0] gnd n/c n/c gnd n/c gnd gnd v cc v cc v cc rxclk a+ rxda [6] rxda[ 5] note: 1. n/c = do not connect
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 6 of 35 pin configuration (bottom view) [1] 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 n/c inb2- n/c inb1- v cc n/c ina2- gnd n/c ina1- n/c ind2- gnd n/c ind1- v cc n/c inc2- n/c inc1- a n/c inb2+ n/c inb1+ v cc n/c ina2+ gnd n/c ina1+ n/c ind2+ gnd n/c ind1+ v cc n/c inc2+ n/c inc1+ b tdo gnd rx rate trg rate v cc gnd n/c gnd n/c n/c n/c n/c gnd sda sel par ctl v cc inselb inselc tms tdi c n/c rfen rxle v cc v cc gnd n/c gnd bre[0] bre[1] bre[2] bre[3] gnd spd sel rf mode v cc insela inseld trstz tclk d v cc v cc v cc v cc v cc v cc v cc v cc e rxstb [0] rxop b rxstb [1] bistle rxck sel v cc v cc n/c f rxdb [1] fram char gnd dec mode gnd gnd gnd gnd g gnd gnd gnd gnd gnd gnd gnd gnd h rxdb [2] rxdb [5] rxdb [0] rxstb [2] gnd gnd gnd gnd j rxclk b+ rxdb [7] rxdb [4] rxdb [3] lfic gnd rxclk c- rxdc [2] k gnd rxclk b- lfib rxdb [6] gnd gnd rxclk c+ rxdc [3] l gnd gnd gnd gnd rxdc [6] rxdc [7] rxdc [5] rxdc [4] m gnd gnd gnd gnd gnd gnd gnd gnd n gnd gnd gnd gnd rxstc [1] rxstc [0] rxdc [0] rxdc [1] p n/c v cc v cc v cc v cc n/c rxop c rxstc [2] r v cc v cc v cc v cc v cc v cc v cc v cc t rxsta [1] rxsta [2] rxopa rxda [2] v cc v cc gnd gnd gnd trg clk- n/c rxop d gnd rxdd [1] rxdd [2] v cc v cc v cc v cc v cc u rxsta [0] rxda [0] rxda [3] rxda [7] v cc v cc gnd gnd n/c trg clk+ n/c rxstd [2] gnd rxstd [0] rxdd [3] v cc rxdd [6] v cc v cc v cc v rxda [1] rxda [4] rxclk a- lfia v cc v cc gnd gnd gnd gnd gnd n/c gnd rxstd [1] rxdd [4] v cc rxclk d? lfid v cc v cc w rxda [5] rxda [6] rxclk a+ v cc v cc v cc gnd gnd n/c gnd n/c n/c gnd rxdd [0] rxdd [5] v cc rxclk d+ rxdd [7] v cc v cc y
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 7 of 35 pin descriptions cyp15g0401rb quad hotlink ii receiver pin name i/o characteristics signal description receive path data signals rxda[7:0] rxdb[7:0] rxdc[7:0] rxdd[7:0] lvttl output, synchronous to the selected rxclkx output (or trgclk input [2] 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 special characters. the status of the received data is represented by the values of rxstx[2:0]. when the decoder is bypassed (decmode = low), rxdx[7:0] become the higher order bits of the 10-bit received character. see table 7 for details. rxsta[2: 0] rxstb[2: 0] rxstc[2: 0] rxstd[2: 0] lvttl output, synchronous to the selected rxclkx output (or trgclk input [2] when rxcksel = low) parallel status output . these outputs change following the rising edge of the selected receive interface clock. when the decoder is bypassed (decmode = low), rxstx[1:0] become the two low-order bits of the 10-bit received char acter, while rxstx[2] = high indicates the presence of a comma character in the output register. see table 7 for details. when the decoder is enabled (decmode = hig h or mid), rxstx[2:0] provide status of the received signal. see ta ble 9 and table 10 for a list of receive character status. rxopa rxopb rxopc rxopd three-state, lvttl output, synchronous to the selected rxclkx output (or trgclk input [2] when rxcksel = low) receive path odd parity . when parity generation is enabled (parctl low), the parity output at these pins is valid for the data on the associated rxdx bus bits. when parity generation is disabled (parctl = low) these output drivers are disabled (high-z). receive path clock and clock control rxrate lvttl input, static control input, internal pull-down receive clock rate select . when low, the rxclkx recovered clock outputs are complementary clocks operating at the recovered character rate. data for the associated receive channels should be latched on the rising edge of rxclkx+ or falling edge of rxclkx?. when high, the rxclkx recovered clock outputs are complement ary clocks operating at half the character rate. data for the a ssociated receive channels should be latched alternately on the rising edge of rxclkx+ and rxclkx?. when trgclk is selected to clock the ou tput registers (rxckselx = low), rxratex is not interpreted. the rxclka and rxcl kc output clocks will follow the frequency and duty cycle of trgclk. trgrate lvttl input, static control input, internal pull-down training clock rate select . when trgclk is selected to clock the receive parallel interfaces (rxcksel = low), the trgrate in put also determines if the clocks on the rxclka and rxclkc outputs are full or half-rate. when trgrate = high (trgclk is half-rate) and rxcksel = low, the rxclka and rxclkc output clocks are also half-rate clocks and follow the freq uency and duty cycle of the trgclk input. when trgrate = low (trgclk is full-rate ) and rxcksel = low, the rxclka and rxclkc output clocks are full-rate clocks and follow the frequency and duty cycle of the trgclk input. framch ar three-level select [3] , static control input framing character select . used to select the character or portion of a character used for character framing of the received data st reams. 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 to low is reserved for component test. rfen lvttl input, asynchronous, internal pull-down reframe enable for all channels . active high. when high, the framers in all four channels are enabled to frame per the presen tly enabled framing mode as selected by rfmode and selected framing ch aracter as selected by framchar. notes: 2. when trgclk is configured for half-rate operation (trgrate = high), these inputs are sampled (or the outputs change) relative to both the rising and falling edges of trgclk. 3. three-level select inputs are used for static configuration. t hey are ternary (not binary) inputs that make use of non-standa rd logic levels of low, mid, and high. the low level is usually implemented by direct connection to v ss (ground). the high level is usually im plemented by direct connection to v cc . when not connected or allowed to float, a three-leve l select input will self-bias to the mid level.
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 8 of 35 rxclka rxclkb rxclkc rxclkd three-state, lvttl output clock or static control input receive character clock output or clock select input . when configured such that all output data paths are clocked by the recove red clock (rxcksel = mi d), these true and complement clocks are the receive interface clocks which are used to control timing of output data (rxdx[7:0], rxstx[2:0] and rx opx). these clocks are output continuously at either the dual-character rate (1/20 th the serial bit-rate) or character rate (1/10 th the serial bit-rate) of the data being received, as selected by rxrate. when configured such that all output data paths are clocked by trgclk instead of a recovered clock (rxcksel = low), the rxclka and rxclkc output drivers present a buffered and delayed form of trgclk. rxclka and rxclkc are buffered forms of trgclk that are slightly different in phas e. this phase difference allows the user to select the optimal setup/hold ti ming for their specific interface. rxcksel three-level select [3] , static control input receive clock mode . selects the receive clock source used to transfer data to the output registers. when low, all four output registers are clocked by trgclk. rxclkb and rxclkd outputs are disabled (high-z), and rxclka and rxclkc present buffered and delayed forms of trgclk. when mid, each rxclkx output follows the recovered clock for the respective channel, as selected by rxrate. when the 10b/8b decoder and elasticity buffer are bypassed (decmode = low), rxcksel must be mid. when high and the receive channels are operated in independent mode (rx modes 0 and 2), rxclka and rxclkc output the recovered clock from receive channel a, b, c, or d, as selected by rxclkb+ and rxclkd +. this output clock may operate at the character-rate or half the charac ter-rate as selected by rxrate. decmod e three-level select [3] , static control input decoder mode select . this input selects the behavior of the decoder block. 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 decoder is enabled and the cypress decoder table for special code characters is used. when high, the decoder is enabled and the alternate decoder table for special code characters is used. see table 15 for a list of the special codes supported in both encoded modes. rfmode three-level select [3] , static control input 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 received serial bit stream). this signal oper ates 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 re ceived 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 charac ter(s), on identical 10-bit boundaries, within a span of 50 bits, before the character boundaries are adjusted. the recovered character clock remains in the same 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 ) of the allowed disparities in the received serial bit stream, on identical 10-bit boundaries, on four directly adjacent characters. the recovered character clock remains in the sa me phase regardless of character offset. pin descriptions (continued) cyp15g0401rb quad hotlink ii receiver pin name i/o characteristics signal description
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 9 of 35 device control signals parctl three-level select [3] , static control input parity generate control . used to control the different parity generate functions. when low, parity checking is disabled, and the rxopx outputs are all disabled (high-z). when mid, and the 10b/8b decoder is enabled (decmode low), odd parity is generated for the rxdx[7:0] outputs and presented on rxopx. when the decoder is disabled (decmode = low), odd parity is generated for the rxdx[7:0] and rxstx[1:0] outputs and presented on rxopx. when high, parity generation is enabled. odd parity is generated for the rxdx[7:0] and rxstx[2:0] outputs and presented on rxopx. see table 8 for details. spdsel three-level select [3] static control input serial rate select . this input specifies the operating bit-rate range of the receive plls. low = 195?400 mbd, mid = 400?800 mbd, high = 800?1500 mbd. when spdsel is low, setting trgrate = high (half-rate training clock) is invalid. 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 trgclk , 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 trgclk ), the status and data outputs will become deterministic in less than 16 trgclk cycles . the bistle and rxle latches are reset by trstz . if the elasticity buffer is used, trstz should be applied after power up to initialize the internal pointers into these memory arrays. trgclk differential lvpecl or single-ended lvttl input clock training clock . this clock is used as the centering frequency of the range controller block of the receive cdr plls, via the clo ck multiplier. this in put clock may also be selected to clock the receive parallel interfaces. when driven by a single-ended lvcmos or lvttl clock source, connect the clock s ource to either the true or complement trgclk input, and leave the alternate trgclk input open (floating). when driven by an lvpecl clock source, the clock must be a differential clock, using both inputs. when rxcksel = low, the elasticity buffer is enabled and trgclk is used as the clock for the parallel receive da ta (output) interface. if the elasticity buffer is used, framing charac ters will be inserted or deleted to/from the data stream to compensate for frequency differences between the training clock and recovered clock. when an addition happens, a k28.5 will be appended immediately after a framing character is detected in the elasti city buffer. when deletion happens, a framing character will be removed from the data stre am when detected in the elasticity buffer. analog i/o and control ina1 inb1 inc1 ind1 lvpecl differential input primary differential serial data inputs . these inputs accept the serial data stream for deserialization and decoding. the inx1 serial streams are passed to the receiver clock and data recovery (cdr) circuits to extract the data content when inselx = high. ina2 inb2 inc2 ind2 lvpecl differential input secondary differential serial data inputs . these inputs accept the serial data stream for deserialization and decoding. the inx2 serial streams are passed to the receiver clock and data recovery (cdr) circuits to extract the data content when inselx = low. insela inselb inselc inseld lvttl input, asynchronous receive input selector . determines which external serial bit stream is passed to the receiver clock and data recovery circuit. when high, the inx1 input is selected. when low, the inx2 input is selected. sdasel three-level select [3] static configuration 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 1 . bistle lvttl input, asynchronous, internal pull-up receive bist latch enable . active high. when bistle = high, the signals on the bre[3:0] inputs directly control the receive bist enables. when the bre[x] input is low, the associated receive channel is configured to compare the bist sequence. when the bre[x] input is high, the associated receiv e channel is configured for normal data reception. the specific mapping of bre[3:0] signals to receiv e bist enables is listed in table 2 . when bistle returns low, the last va lues present on bre[3:0] are captured in the internal bist enable latch. when the la tch is closed, if the device is reset (trstz is sampled low), the latch is reset to disable bist on all receive channels. pin descriptions (continued) cyp15g0401rb quad hotlink ii receiver pin name i/o characteristics signal description
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 10 of 35 cyp15g0401rb hotlink ii operation the cyp15g0401rb is a highly configurable device designed to support reliable transfer of large quantities of data, using high-speed serial links, from one or multiple sources to one destination. this device supports four single-byte or single-character channels. cyp15g0401rb receive data path serial line receivers two differential line receivers, inx1 and inx2 , are available on each channel for accepting serial data streams. the active serial line receiver on a channel is selected using the associated inselx input. the serial line receiver inputs are differential, and can accommodate wire interconnect and filtering losses or transmissi on line attenuation greater than 16 db. for normal operation, these inputs should receive a signal of at least vi diff > 100 mv, or 200 mv peak-to-peak differential. each line receiver can be dc- or ac-coupled to +3.3v powered fiber-optic interface modules (any ecl/pecl family, not limited to 100k pecl) or ac-coupled to +5v powered optical modules. the common-mode tolerance of these line receivers accommodates a wide range of signal termination voltages. each receiver provides internal dc-restoration, to the center of the receiver?s common mode range, for ac-coupled signals. signal detect/link fault each selected line receiver (i.e., that routed to the clock and data recovery pll) is simultaneously monitored for ? analog amplitude above lim it specified by sdasel ? transition density greater than specified limit ? range controller reports the received data stream within normal frequency range (1500 ppm) [4] ? 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 lfix (link fault indicator) output associated with each receive channel. rxle lvttl input, asynchronous, internal pull-up receive channel power-control latch enable . active high. when rxle = high, the signals on the bre[3:0] inputs directly control the power enables for the receive plls and analog circuitry. when the bre[3:0] i nput is high, the associated receive channel a through d pll and analog circuitry are active. when the bre[3:0] input is low, the associated receive channel a through d pll and analog circuitry are powered down. the specific mapping of bre[3:0] signals to the associated receive channel enables is listed in table 2 . when rxle returns low, the last valu es present on bre[3:0] are captured in the internal rx pll enable latch. when the device is reset (trstz = low), the latch is reset to disable all receive channels. bre[3:0] lvttl input, asynchronous, internal pull-up bist and receive channel enables . these inputs are passed to and through the bist enable latch when bistle is high, and captured in this latch when bistle returns low. these inputs are passed to and through the receive channel enable latch when rxle is high, and captured in this latch when rxle returns low. lfia lfib lfic lfid lvttl output, asynchronous link fault indication output . active low. lfix is the logical or of four internal condi- tions: 1. received serial data frequency outside expected range 2. analog amplitude below expected levels 3. transition density lower than expected 4. receive channel disabled. 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 contro ller 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 v cc +3.3v power gnd signal and power ground for all internal circuits . pin descriptions (continued) cyp15g0401rb quad hotlink ii receiver pin name i/o characteristics signal description
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 11 of 35 analog amplitude while most signal monitors are based on fixed constants, the analog amplitude level detection is adjustable. this allows operation with highly attenuated signals, or in high-noise environments. this adjustment is made through the sdasel signal, a three-level select [3] input, which sets the trip point for the detection of a valid signal at one of three levels, as listed in table 1 . this control input affects the analog monitors for all receive channels. the analog signal detect monitors are active for the line receiver selected by the associated inselx input. transition density the transition detect ion 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 on a channel, the transition detection logic for that channel will assert lfix . the lfix output remains asserted until at least one transition is detected in each of three adjacent received characters. range controls 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: ? 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 to perform this function, the fr equency of the vco is periodi- cally sampled and compared to the frequency of the trgclk input. if the vco is running at a frequency beyond 1500 ppm [4] as defined by the training clock frequency, it is periodically forced to the correct frequency (as defined by trgclk, spdsel, and trgrate) 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 = (trgclk- period) * (16000). during the time that the range control forces the pll vco to run at trgclk*10 (or trgclk*20 when trgrate = 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 (dependi ng on other factors such as transition density and amplitude detection) and the recovered byte clock (rxclkx) 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 cyp15g0401rb contains four receive channels that can be independently enabled and disabled. each channel can be enabled or disabled separately through the bre[3:0] inputs, as controlled by the rxle latch-enable signal. when rxle is high, the signals present on the bre[3:0] inputs are passed through the receive channel enable latch to control the plls and logic of the associated receive channel. the bre[3:0] input associated with a specific receive channel is listed in table 2 . when rxle is high and bre[x] is high, the associated receive channel is enabled to receive and recover a serial stream. when rxle is high and bre[x] is low, the associated receive channel is disabled and powered down. any disabled channel indicates an asserted lfix output. when rxle returns low, the values present on the bre[3:0] inputs are latched in the receive channel enable latch, and remain there until rxle returns high to open the latch again. [6] clock multiplier the clock multiplier acce pts a character-rate or half-character-rate external cl ock at the trgclk input, to generate a character-rate clock for use by the clock/data recovery (cdr) blocks. this clock multiplier can accept a trgclk input between 20 mhz and 150 mhz (providing the user with the option to use a trgclk frequency at 1/10 or 1/20 the serial bit rate), however, this clock range is limited by the operating mode of the cyp15g0401rb clock multiplier (controlled by trgrate) and by the level on the spdsel input. spdsel is a static three-level select [3] (ternary) input that selects one of three operating ranges for the serial data inputs. the operating serial signaling-rate and allowable range of trgclk frequencies are listed in table 3 . notes: 4. trgclk has no phase or frequency relationship with the recovered clock(s) and only acts as a centering reference to reduce cl ock synchronization time. trgclk must be within 1500 ppm (0.15%) of the remote transmitter?s pll reference (refclk) frequency. although transmitting to a hotl ink ii receiver necessitates the frequency difference between the transmitter and receiver refer ence clocks to be within 1500- ppm, the stability of the cry stal 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, t o be ieee 802.3z gigabit ethernet compliant, the frequency stability of the crystal needs to be within 100 ppm. 5. the peak amplitudes listed in this table are for typical waveforms that have general ly 3?4 transitions for every ten bits. in a worse case environment the signals may have a sine-wave appearance (highest tran sition density with repeating 0101...). si gnal peak amplitudes levels within this environment type could increase the values in the table above by approximately 100 mv. 6. when a disabled receive channel is re-ena bled, the status of the associated lfix output and data on the parallel outputs for the associated channel may be indeterminate for up to 2 ms. table 1. analog amplitude detect valid signal levels [5] sdasel 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 table 2. bist and receive channel enable signal map bre input bist channel enable (bistle) receive pll channel enable (rxle) bre[3] receive d receive d bre[2] receive c receive c bre[1] receive b receive b bre[0] receive a receive a
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 12 of 35 the trgclk input is a differential input with each input inter- nally biased to 1.4v. if the trgclk+ input is connected to a ttl, lvttl, or lvcmos clock source, trgclk? can be left floating and the input signal is recognized when it passes through the internally biased reference point. when both the trgclk+ and trgclk? 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 trgclk? input to an external voltage source or resistive voltage divider, it is possible to adjust the reference point of the trgclk+ input for alternate logic levels. when doing so, it is nec essary to ensure that the input differential crossing point remains within the parametric range supported by the input. clock/data recovery the extraction of a bit-rate clock and recovery of bits from each received serial stream is performed by a separate clock/data recovery (cdr) block within each receive channel. the clock extraction function is performed by embedded phase-locked loops (plls) that track the freque ncy of the transitions in the incoming bit streams and align the phase of their internal bit-rate clocks to the transitions in the selected serial data streams. each cdr accepts a character-rate (bit-rate 10) or half-character-rate (bit-rate 20) training clock from the trgclk input. this trgclk input is used to ? ensure that the vco (within the cdr) is operating at the correct frequency. ? to reduce pll acquisition time ? and to limit unlocked frequency excursions of the cdr vco when there is no input data pres ent 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 trgclk instead of the data stream. once the cdr output (rxclkx) frequency returns back close to trgclk 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 rxclkx frequency excursions from trgclk. however, the validity of the in put data stream is indicated by the lfix output. the frequency of trgclk is required to be within 1500 ppm [4] of the frequency of the clock that drives the trgclk input of the remote transmitter to ensure a lock to the incoming data stream. for systems using multiple or redundant connections, the lfix output can be used to select an alternate data stream. when an lfix indication is detected, external logic can toggle selection of the associated inx1 and inx2 inputs through the associated inselx input. when a port switch takes place, it is necessary for the receive pll for that channel to reacquire the new serial stream and frame to the incoming character bound- aries. deserializer/framer each cdr circuit extracts bits from the associated serial data stream and clocks these bits in to the shifter/framer at the bit-clock rate. when enabled, the framer examines the data stream, looking for one or more comma or k28.5 characters at all possible bit positions. the location of this character in the data stream is used to determ ine the character boundaries of all following characters. framing character the cyp15g0401rb allows selection of two combinations of framing characters to support requirements of different inter- faces. the selection of the fr aming character is made through the framchar input. the specific bit combinations of these framing characters are listed in table 4 . 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. framer the framer on each channel 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 framers in all four receive paths are disabled, and no combination of bits in a received data stream will alter the character boundaries. when rfen = high, the framer selected by rfmode is enabled on all four channels. when rfmode = low, the low-latency framer is selected [8] . 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 table 3. operating speed settings spdsel trgrate trgclk frequency (mhz) signaling rate (mbaud) low 1 reserved 195?400 0 19.5?40 mid (open) 1 20?40 400?800 0 40?80 high 1 40?75 800?1500 0 80?150 table 4. framing character selector framchar bits detected in framer character name bits detected low reserved for test mid (open) comma+ or comma ? 00111110xx [7] or 11000001xx high ?k28.5 or +k28.5 0011111 010 or 1100000101 notes: 7. the standard definition of a comma contains only seven bits. howe ver, since all valid comma characters within the 8b/10b char acter set also have the eighth bit as an inversion of the seventh bit, the compare pattern is extended to a full eight bits to reduce the possibility of a fra ming error. 8. 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.
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 13 of 35 character. to reduce the impact on external circuits that make use of a recovered clock, the cl ock 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 fourteen character-clock cycles from the detection of the selected framing character. when rfmode = mid (open), the cypress-mode multi-byte framer is selected. the required detection of multiple framing characters makes the associated link much more robust to incorrect framing due to aliased 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 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. framing for all channels is enabled when rfen = high. if rfen = low, the framer for each channel is disabled. when the framers are disabled, no changes are made to the recovered character boundaries on any channel, regardless of the presence of framing charac ters in the data stream. 10b/8b decoder block 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 each 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 14 and table 15 of this data sheet. valid data characters are indicated by a 000b bit-combination on the associated rxstx[2:0] status bits, and special character codes are indicated by a 001b bit-combination 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. this clock mode generates separate rxclkx outputs for each receive channel. when decmode = mid (or open), the 10-bit transmission characters are decoded using ta ble 14 and table 15 . received special code characters are decoded using the cypress column of table 15 . when decmode = high, the 10-bit transmission characters are decoded using table 14 and table 15 . received special code characters are decoded using the alternate column of table 15 . receive bist operation the receiver interfaces contain internal pattern generators that can be used to validate both device and link operation. these generators are enabled by the associated bre[x] signals listed in table 2 (when the bistle latch enable input is high). when enabled, a register in the associated 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(s), the cyp15g0401tb for example. if the receive channels are configured for common clock operation (rxcksel mid) each pass must be preceded by a 16-character wo rd sync sequence. please note that bist cannot be used in a common clock configu- ration (rxcksel mid) when using the cyp15g0401tb device as the bist generator, as the 16-character word sync sequence will not be present in the bist pattern. when synchronized with the received data stream, the associated receiver checks each character in the decoder with each character generated by the lfsr and indicates compare errors and bist status at the rxstx[2:0] bits of the output register. see table 10 for details. when the bistle signal is high, any bre[x] input that is low enables the bist generator/checker in the associated receive channel. when bistle returns low, the values of all bre[x] signals are captured in the bist enable latch. these values remain in the bist enable latch until bistle is returned high. all captured signals in the bist enable latch are set high (i.e., bist is disabled) following a device reset (trstz is sampled low). 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 rxstx[2:0] status outputs. code rule violations or running disparity errors that occur as part of the bist loop do not cause an error indication. rxstx[2:0] indicates 010b or 100b for one character period per bist loop to indicate loop completion. this status can be used to check test pattern progress. these same status values
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 14 of 35 are presented when the decoder is bypassed and bist is enabled on a receive channel. the status reported on rxstx[2: 0] by the bist state machine are listed in table 10 . 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 cyp15g0401rb when rxcksel = mid is identical to that in the cy7b933 and cy7c924dx, allowing interopera ble systems to be built when used at compatible serial signaling rates. if the number of invalid characters received ever exceeds the number of valid characters by sixteen, the receive bist state machine aborts the compare operations and resets the lfsr to the d0.0 state to look for the start of the bist sequence again. when the receive paths are configured for common clock operation (rxcksel mid), each pass must be preceded by a 16-character word sync sequence to allow output buffer alignment and management of clock frequency variations (see cyp15g0401tb datasheet for details on how to send a 16-character word sync sequence from the remote trans- mitter). the bist state machine requires the characters to be correctly framed for it to detect the bi st 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 multi-byte framer is enabled (rfmode = high) and the receiver outputs are clocked relative to a recovered clock, it is necessary to frame the receiver before bist is enabled. receive elasticity buffer each receive channel contains an elasticity buffer that is designed to support multiple clocking modes. these buffers allow data to be read using 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. each elasticity buffer is 10-characters deep, and supports a twelve-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 these buffers is always the recovered clock for the associated read channel. the read clock for the elasticity buffers may come from one of three selectable sources. it may be a ? character-rate trgclk (rxcksel = low and decmode low) ? recovered clock from an alternate receive channel (rxcksel = high and decmode low). the elasticity buffers are bypassed whenever the decoders are bypassed (decmode = low). when the decoders and elasticity buffers are bypassed, rxckselx must be set to mid. receive normal data operation when rxcksel = low, all four receive channels are clocked by trgclk. rxclkb and rxclkd outputs are disabled (high-z), and the rxclka and rxclkc outputs present a buffered and delayed form of trgclk. in this mode, the receive elasticity buffers are enabled. for trgclk clocking, the elasticity buffers must be able to insert k28.5 characters and delete framing characters as appropriate. the insertion of a k28.5 or deletion of a framing character can occur at any time on any channel, however, the actual timing on these insertions and deletions is controlled in part by the how the attached remote 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 a receive buffer overflow or underflow on a receive channel, a minimum density of framing characters must be present in the received data streams. when rxcksel = mid (or open), each received channel output register is clocked by the recovered clock for that channel. since no characters may be added or deleted, the receiver elasticity buffer is bypassed. when rxcksel = high in independent channel mode, all channels are clocked by the selected recovered clock. this selection is made using the rxclkb+ and rxclkd+ signals as inputs per table 5 . this selected clock is always output on rxclka and rxclkc . in this mode the receive elasticity buffers are enabled. when data is output using a recovered clock (rxcksel = high), the receive channels are not allowed to insert and delete characters, except as necessary for elasticity buffer alignment. when the elasticity buffer is used, prior to reception of valid data, a word sync sequence (or at least four framing characters) must be received to center the elasticity buffers. the elasticity buffer may also be centered by a device reset operation initiated by trstz input. however, following such an event, the cyp15g0401rb also requires a framing event before it will correctly decode characters. when rxcksel = high, since the elasticity buffer is not allowed to insert or delete framing characters, the transmit clocks on all received channels must all be from a common source. power control the cyp15g0401rb supports user control of the powered up or down state of each receive channel. the receive channels are controlled by the rxle sig nal and the values present on the bre[3:0] bus. powering down unused channels will save power and reduce system heat generation. controlling system power dissipation will improve the system performance. receive channels when rxle is high, the signals on the bre[3:0] inputs directly control the power e nables for the receive plls and analog circuits. when a bre[3:0] input is high, the associated receive channel [a through d] pll and analog logic are active. when a bre[3:0] input is low, the table 5. independent recovered clock select rxclkb+ rxclkd+ rxclka /rxclkc clock source 0 0 rxclka 0 1 rxclkb 1 0 rxclkc 1 1 rxclkd
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 15 of 35 associated receive channel [a through d] pll and analog circuits are powered down. when rxle returns low, the last values present on the bre[3:0] inputs are captured in the receive channel enable latch. the specific bre[3:0] input signal associated with a receive channel is listed in table 2 . any disabled receive channel will indicate a constant lfix output. when a disabled receive channel is re-enabled, the status of the associated lfix output and data on the parallel outputs for the associated channel may be indeterminate for up to 2 ms. device reset state when the cyp15g0401rb is reset by assertion of trstz , the receive enable latches are both cleared, and the bist enable latch is preset. in this state, all receive channels are disabled, and bist is disabled on all channels. following a device reset, it is necessary to enable the receive channels used for normal operation. this can be done by sequencing the appropriate values on the bre[3:0] inputs while the rxle signals are raised and lowered. for systems that do not require dynamic control of power, or want the device to power up in a fixed configuration, it is also possible to strap the rxle control signa l high to permanently enable its associated latches. connection of the associated bre[3:0] signals to a stable high will then enable the respective receive channels as soon as the trstz signal is deasserted. output bus each 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. the bits are assigned as per table 6 . when the 10b/8b decoder is bypassed (decmode = low), the framed 10-bit character and a single status bit (comdet) are presented at the receiver output register. the status output indicates 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 7 . the comdetx outputs are high when the character in the output register for the associated channel contains the selected framing character at the proper character boundary, and low for all other bit combinations. when the low-latency framer and half-rate receive port clocking are also enabled (rfmode = low, rxrate = high, and rxcksel low), the framer will stretch the recovered clock to the nearest 20-bit boundary such that the rising edge of rxclkx+ occurs when comdetx is present on the associated output bus. notes: 9. the rxopx outputs are also driven from the associated output r egister, but their interpretation is under the separate control of parctl. table 6. output register bit assignments [9] signal name decmode = low decmode = mid or high rxstx[2] (lsb) comdetx rxstx[2] rxstx[1] doutx[0] rxstx[1] rxstx[0] doutx[1] rxstx[0] rxdx[0] doutx[2] rxdx[0] rxdx[1] doutx[3] rxdx[1] rxdx[2] doutx[4] rxdx[2] rxdx[3] doutx[5] rxdx[3] rxdx[4] doutx[6] rxdx[4] rxdx[5] doutx[7] rxdx[5] rxdx[6] doutx[8] rxdx[6] rxdx[7] (msb) doutx[9] rxdx[7]
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 16 of 35 . 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 rxclkx+ occurs when comdetx is present on the associated output bus. this adjustment only occurs when the framer is enabled (rfen = high). when the fram er is disabled, the clock boundaries are not adjusted, and comdetx may be asserted during the rising edge of rxclk? (if an odd number of characters were received following the initial framing). parity generation in addition to the eleven data and status bits that are presented by each channel, an rxopx parity output is also available on each channel. this allows the cyp15g0401rb to support odd parity generation for each channel. to handle a wide range of system environments, the cyp15g0401rb supports different forms of parity generation, including no parity. when the decoders are enabled (decmode low), parity can be generated on ? the rxdx[7:0] character ? the rxdx[7:0] character and rxstx[2:0] status. when the decoders are bypassed (decmode = low), parity can be generated on ? the rxdx[7:0] and rxstx[1:0] bits ? the rxdx[7:0] and rxstx[2:0] bits. these modes differ in the number of bits which are included in the parity calculation. 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 8 . parity generation is enabled through the three-level select parctl input. when parctl = low, parity checking is disabled, and the rxopx outputs are all disabled (high-z). when parctl = mid (open) and the decoders are enabled (decmode low), odd parity is generated for the received and decoded character in the rxdx[7:0] signals and is presented on the associ ated rxopx output. when parctl = mid and the decoders are bypassed (decmode = low), odd parity is generated for the received and decoded character in the rx dx[7:0] and rxstx[1:0] bit positions. when parctl = high, odd parity is generated for the rxdx[7:0] and the associated rxstx[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 oper ations (regardless of the state of decmode) ? character violations. these conditions normally overl ap; e.g., a valid data character received with incorrect running disparity is not reported as a valid data character. it is instead reported as a decoder violation of some specific type. this implies a hierarchy or priority level to the various status bit combinations. the hierarchy and value of each status is listed in table 9 . the receive status when normal data is received is shown in table 9 . the receive status when receive bist is enabled is shown in table 10 . notes: 10. receive path parity output drivers (rxopx) are disabled (high-z) when parctl = low. 11. when the decoder is bypassed (decmode = low) and bist is not enab led (receive bist latch output is high), rxstx[2] is driven to a logic-0, except when the character in the output buffer is a framing character. table 7. decoder bypass mode (decmode = low) signal name bus weight 10bit name rxstx[2] (lsb) comdetx rxstx[1] 2 0 a rxstx[0] 2 1 b rxdx[0] 2 2 c rxdx[1] 2 3 d rxdx[2] 2 4 e rxdx[3] 2 5 i rxdx[4] 2 6 f rxdx[5] 2 7 g rxdx[6] 2 8 h rxdx[7] (msb) 2 9 j table 8. output register parity generation signal name receive parity generate mode (parctl) low [10] mid high decmode = low decmode low rxstx[2] x [11] rxstx[1] x x rxstx[0] x x rxdx[0] x x x rxdx[1] x x x rxdx[2] x x x rxdx[3] x x x rxdx[4] x x x rxdx[5] x x x rxdx[6] x x x rxdx[7] x x x
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 17 of 35 table 9. receive character status when channels are operated to receive normal data rxstx[2:0] priority status 000 7 normal character received. the valid data character with the correct running disparity received 001 7 special code detected . special code other than the selected framing character or decoder violation received 010 2 receive elasticity buffer underrun/overrun error. the receive elasticity buffer was not able to add/drop a k28.5 or framing character. 011 5 framing character detected. this indicates that a character matching the patterns identified as a framing character was detected. the decoded value of this character is present on the associ- ated output bus. 100 4 codeword violation. the character on the output bus is a c0 .7. this indicates that the received character cannot be decoded into any valid character. 101 1 pll out of lock indication 110 6 running disparity error. the character on the output bus is a c4.7, c1.7 or c2.7 111 3 invalid table 10. receive character status when channels are operated to receive bist data rxstx[2:0] priority receive bist status (receive bist = enabled) 000 7 bist data compare . character compared correctly 001 7 bist command compare . character compared correctly 010 2 bist last good . last character of bist sequence detected and valid. 011 5 reserved for test 100 4 bist last bad . last character of bist sequence detected invalid. 101 1 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 bist error . while comparing characters, a mismatch was found in one or more of the decoded character bits. 111 3 bist wait . the receiver is comparing characters. but ha s not yet found the start of bist character to enable the lfsr.
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 18 of 35 bist status state machine when a receive path is enabled to look for and compare the received data stream with the bist pattern, the rxstx[2:0] bits identify the present state of the bist compare operation. the bist state machine has multiple states, as shown in figure 2 and ta ble 10 . 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 st ate of the bist state machine. if the number of detected errors ever exceeds the number of valid matches by greater than sixteen, 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 el asticity buffer ever hits an overflow/underflow condition, the status is forced to the bist_start until the buffer is recentered (approximately nine character periods). to ensure compatibility between the source and destination systems when operating in bist modes, the sending and receiving ends of the link must use the same receive clock setup, i.e. rxcksel = mid or rxcksel mid. this is appli- cable when interfacing to a cyp(v)15g0401dxb for example. when interfacing to transmitte r only hotlink ii devices such as the cyp15g0401tb it is necessary to have rxcksel = mid. jtag support the cyp15g0401rb 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 trgclk clock input. the high-speed serial inputs and outputs are not part of the jtag test chain. jtag id the jtag device id for the cyp15g0401rb is ?1c800069?x. three-level select inputs each three-level select input reports as two bits in the scan register. these bits report t he low, mid, and high state of the associated input as 00, 10, and 11, respectively.
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 19 of 35 receive bist detected low monitor data received rxstx = bist_start (101) no rx pll out of lock yes, rxstx = bist_command_compare (001) or bist_data_compare (000) compare next character auto-abort condition mismatch end-of-bist state yes, rxstx = bist_last_bad (100) yes no no, rxstx = bist_error (110) data or command match command rxstx = bist_command_compare (001) end-of-bist state data yes, rxstx = bist_last_good (010) no rxstx = bist_data_compare (000) figure 2. receive bist state machine elasticity buffer error start of bist detected rxstx = bist_wait (111) yes rxstx = bist_start (101) no
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 20 of 35 maximum ratings (above which the useful life may be impaired. user guidelines only, not tested.) storage temperature .............. .............. ...... ?65c to +150c ambient temperature with power applied....?55c to +125c supply voltage to ground potential ............... ?0.5v to +3.8v dc voltage applied to lvttl outputs in high-z state .......................................?0.5v to v cc + 0.5v output current into lvttl outputs (low)..................60 ma dc input voltage....................................?0.5v to v cc + 0.5v static discharge voltage......... .............. .............. ...... > 2000v (per mil-std-883, method 3015) latch-up current..................................................... > 200 ma power-up requirements the cyp15g0401rb requires one power-supply. the voltage on any input or i/o pin cannot exceed the power pin during power-up operating range range ambient temperature v cc commercial 0c to +70c +3.3v 5% industrial ?40c to +85c +3.3v 5% cyp15g0401rb dc electrical characteristics over the operating range parameter description test conditions min. max. unit lvttl-compatible outputs v oht output high voltage i oh = ? 4 ma, v cc = min. 2.4 v cc v v olt output low voltage i ol = 4 ma, v cc = min. 0 0.4 v i ost output short circuit current v out = 0v [12] ?20 ?100 ma i ozl high-z output leakage current ?20 20 a lvttl-compatible inputs v iht input high voltage 2.0 v cc + 0.3 v v ilt input low voltage ?0.5 0.8 v i iht input high current trgclk input, v in = v cc 1.5 ma other inputs, v in = v cc +40 a i ilt input low current trgclk input, v in = 0.0v ?1.5 ma other inputs, v in = 0.0v ?40 a i ihpdt input high current with internal pull-down v in = v cc +200 a i ilput input low current with internal pull-up v in = 0.0v ?200 a lvdiff inputs: trgclk v diff [13] input differential voltage 400 v cc mv v ihhp highest input high voltage 1.2 v cc v v illp lowest input low voltage 0.0 v cc/2 v v comref [14] common mode range 1.0 v cc ? 1.2v v three-level inputs v ihh three-level input high voltage min. v cc max. 0.87 * v cc v cc v v imm three-level input mid voltage min. v cc max. 0.47 * v cc 0.53 * v cc v v ill three-level input low voltage min. v cc max. 0.0 0.13 * v cc v i ihh input high current v in = v cc 200 a i imm input mid current v in = v cc /2 ?50 50 a i ill input low current v in = gnd ?200 a differential serial line receiver inputs: ina1 , ina2 , inb1 , inb2 , inc1 , inc2 , ind1 , ind2 v diffs [13] input differential voltage |(in+) ? (in ? )| 100 1200 mv v ihe highest input high voltage v cc v notes: 12. tested one output at a time, output shorted for less than one second, less than 10% duty cycle. 13. this is the minimum difference in voltage between the true and complement inputs required to ensure detection of a logic-1 o r logic-0. a logic-1 exists when the true (+) input is more positive than the complement ( ? ) input. a logic-0 exists when the complement ( ? ) input is more positive than true (+) input. 14. the common mode range defines the allowable range of trgclk+ and trgclk ? when trgclk+ = trgclk ? . this marks the zero-crossing between the true and complement inputs as the signal switches between a logic-1 and a logic-0.
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 21 of 35 test loads and waveforms v ile lowest input low voltage v cc ? 2.0 v i ihe input high current v in = v ihe max. 1350 a i ile input low current v in = v ile min. ?700 a v com [15, 16] common mode input range v cc ? 1.95 v cc ? 0.05 v power supply parameter description test conditions typ. [17] max. [18] unit i cc power supply current trgclk = max. commercial 660 690 ma industrial 740 ma i cc power supply current trgclk = 125 mhz commercial 640 650 ma industrial 700 ma cyp15g0401rb dc electrical characteristics over the operating range (continued) parameter description test conditions min. max. unit 2.0v 0.8v gnd 2.0v 0.8v 80% 20% 80% 20% 3.0v v th = 1.4v 270 ps 270 ps [20] v th = 1.4v 3.3v r1 r2 r1 = 590 ? r2 = 435 ? (includes fixture and probe capacitance) c l 7 pf (a) lvttl output test load c l (b) lvttl input test waveform (c) cml/lvpecl input test waveform 1 ns 1 ns v ihe v ile v ihe v ile [19] cyp15g0401rb ac characteristics over the operating range parameter description min. max. unit cyp15g0401rb receiver lvttl switching characteristics over the operating range f rs rxclkx clock output frequency 9.75 150 mhz t rxclkp rxclkx period 6.66 102.56 ns t rxclkh rxclkx high time (rxrate = low) 2.33 [21] 26.64 ns rxclkx high time (rxrate = high) 5.66 52.28 ns t rxclkl rxclkx low time (rxrate = low) 2.33 [21] 26.64 ns rxclkx low time (rxrate = high) 5.66 52.28 ns t rxclkd rxclkx duty cycle centered at 50% ?1.0 +1.0 ns t rxclkr [21] rxclkx rise time 0.3 1.2 ns t rxclkf [21] rxclkx fall time 0.3 1.2 ns notes: 15. the common mode range defines the allowable range of input+ and input ? when input+ = input ? . this marks the zero-crossing between the true and complement inputs as the signal switches between a logic-1 and a logic-0. 16. not applicable for ac-coupled inte rfaces. for ac-coupled interfaces, v diffs requirement still needs to be satisfied. 17. maximum i cc is measured with v cc = max, rxcksel = low, with all tx and rx channels and se rial line drivers enabled, send ing a continuous alternating 01 pattern to the associated receive channel, and outputs unloaded. 18. typical i cc is measured under similar conditions except with v cc = 3.3v, t a = 25c, rxcksel = low, with all rx channels enabled receiving a continuous alternating 01 pattern to the associated receive channel. the re dundant outputs on each channel are powered down and the parall el outputs are unloaded. 19. cypress uses constant current (ate) load configurations and fo rcing functions. this figure is for reference only. 5-pf diffe rential load reflects tester capacitance, and is recommended at low data rates only. 20. the lvttl switching threshold is 1.4v. all timing references ar e made relative to the point where the signal edges crosses t he threshold voltage. 21. tested initially and after any design or process changes that may affect these parameters, but not 100% tested.
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 22 of 35 t rxdv? [24] status and data valid time to rxclkx (rxcksel high or mid) 5ui ? 1.5 ns status and data valid time to rxclkx (half rate recovered clock) 5ui ? 1.0 ns t rxdv+ [24] status and data valid time from rxclkx (rxcksel high or mid) 5ui ? 1.8 ns status and data valid time from rxclkx (half rate recovered clock) 5ui ? 2.3 ns cyp15g0401rb trgclk switching characteristics over the operating range f trg trgclk clock frequency 19.5 150 mhz t trgclk trgclk period 6.66 51.28 ns t trgh trgclk high time (trgrate = high) 5.9 ns trgclk high time (trgrate = low) 2.9 [21] ns t trgl trgclk low time (trgrate = high) 5.9 ns trgclk low time (trgrate = low) 2.9 [21] ns t trgd [25] trgclk duty cycle 30 70 % t trgr [21, 22, 23] trgclk rise time (20% ? 80%) 2 ns t trgf [21, 22, 23] trgclk fall time (20% ? 80%) 2 ns t rtrgda [26] receive data access time from trgclk (rxcksel = low) 9.5 ns t rtrgdv receive data valid time from trgclk (rxcksel = low) 2.5 ns t trgadv? received data valid time to rxclka (rxcksel = low) 10ui ? 4.7 ns t trgadv+ received data valid time from rxclka (rxcksel = low) 0.5 ns t trgcdv? received data valid time to rxclkc (rxcksel = low) 10ui ? 4.3 ns t trgcdv+ received data valid time from rxclkc (rxcksel = low) ?0.2 ns t trgrx [4] trgclk frequency referenced to received clock period ?1500 +1500 ppm cyp15g0401rb receive serial inpu ts and cdr pll characteristics over the operating range t rxlock receive pll lock to input data stream (cold start) 376k ui [29] receive pll lock to input data stream 376k ui t rxunlock receive pll unlock rate 46 ui t jtol [27] to ta l j i t t e r to l e r a n c e ieee 802.3z [28] 600 ps t djtol [27 ] deterministic jitter tolerance ieee 802.3z [28] 370 ps cyp15g0401rb ac characteristics over the operating range (continued) parameter description min. max. unit capacitance [21] parameter description test conditions max. unit c inttl ttl input capacitance t a = 25c, f 0 = 1 mhz, v cc = 3.3v 7 pf c inpecl pecl input capacitance t a = 25c, f 0 = 1 mhz, v cc = 3.3v 4 pf notes: 22. the ratio of rise time to falling time must not vary by greater than 2:1. 23. for a given operating frequency, neither rise or fall specificat ion can be greater than 20% of the clock-cycle period or the data sheet maximum time. 24. parallel data output specifications are only valid if al l inputs or outputs are loaded with similar dc and ac loads. 25. the duty cycle specification is a simultaneous condition with the t trgh and t trgl parameters. this means that at faster character rates the trgclk duty cycle cannot be as large as 30% ? 70%. 26. since this timing parameter is greater than the minimum time period of trgclk it sets an upper limit to the frequency in whi ch trgclkx can be used to clock the receive data out of the output register. for predictable timi ng, users can use this parameter only if trgclk period is grea ter than sum of t rtrgda and set-up time of the upstream device. when this condition is not tr ue, rxclkc or rxclka (a buffered or delayed version of trgcl k when rxckselx = low) could be used to clock the receive data out of the device. 27. total jitter is calculated at an assumed ber of 1e ?12. hence: total jitter (t j ) = (t rj * 14) + t dj . 28. also meets all jitter tolerance requirements as specified by obsai rp3, cpri, escon, ficon, fibre channel and dvb-asi. 29. receiver ui (unit interval) is calculated as 1/(f trg * 20) (when rxrate = high) or 1/(f trg * 10) (when rxrate = low) if no data is being received, or 1/(f trg * 20) (when rxrate = high) or 1/(f trg * 10) (when rxrate = low) of the remote transmitter if da ta is being received. in an operating link this is equivalent to t b .
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 23 of 35 switching waveforms for the cyp15g0401rb hotlink ii receiver notes: 30. rxclka and rxclkc are delayed in phase from trgclk, and are different in phase from each other. 31. when operated with a half-rate trgclk, the set-up and hold specifications for data relative to rxclka and rxclkc are relativ e to both rising and falling edges of the respective clock output. trgclk rxdx[7:0], rxstx[2:0], t rtrgdv t trgclk t trgh t trgl receive interface read timing rxopx rxcksel = low trgrate = low rxclka rxclkc t trgadv+ t trgadv ? t trgcdv ? t trgcdv+ note 30 t rtrgda trgclk rxdx[7:0], rxstx[2:0], t rtrgdv t trgclk t trgh t trgl receive interface read timing rxopx rxcksel = low trgrate = high t rtrgda rxclka rxclkc t trgadv+ t trgadv ? t trgcdv ? t trgcdv+ note 30 note 31 t rtrgda t rtrgdv rxclkx+ rxdx[7:0], rxstx[2:0], t rxdv+ t rxclkp t rxclkh t rxclkl receive interface read timing rxopx rxcksel = high or mid rxrate = low rxclkx ? t rxdv ?
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 24 of 35 switching waveforms for the cyp15g0401rb hotlink ii receiver (continued) rxclkx+ rxdx[7:0], rxstx[2:0], t rxdv+ t rxclkp t rxclkh t rxclkl receive interface read timing rxopx rxcksel = high or mid rxrate = high rxclkx ? t rxdv ?
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 25 of 35 table 11.package coordinate signal allocation ball id signal name signal type ball id signal name signal type ball id signal name signal type a01 inc1? cml in c04 inselb lvttl in e19 vcc power a02 n/c no connect c05 vcc power e20 vcc power a03 inc2? cml in c06 parctl 3-level sel f01 n/c no connect a04 n/c no connect c07 sdasel 3-level sel f02 vcc power a05 vcc power c08 gnd ground f03 vcc power a06 ind1? cml in c09 n/c no connect f04 rxcksel 3-level sel a07 n/c no connect c10 n/c no connect f17 bistle lvttl in pu a08 gnd ground c11 n/c no connect f18 rxstb[1] lvttl out a09 ind2? cml in c12 n/c no connect f19 rxopb lvttl 3-s out a10 n/c no connect c13 gnd ground f20 rxstb[0] lvttl out a11 ina1? cml in c14 n/c no connect g01 gnd ground a12 n/c no connect c15 gnd ground g02 gnd ground a13 gnd ground c16 vcc power g03 gnd ground a14 ina2? cml in c17 trgrate lvttl in pd g04 gnd ground a15 n/c no connect c18 rxrate lvttl in pd g17 decmode 3-level sel a16 vcc power c19 gnd ground g18 gnd ground a17 inb1? cml in c20 tdo lvttl 3-s out g19 framchar 3-level sel a18 n/c no connect d01 tclk lvttl in pd g20 rxdb[1] lvttl out a19 inb2? cml in d02 trstz lvttl in pu h01 gnd ground a20 n/c no connect d03 inseld lvttl in h02 gnd ground b01 inc1+ cml in d04 insela lvttl in h03 gnd ground b02 n/c no connect d05 vcc power h04 gnd ground b03 inc2+ cml in d06 rfmode 3-level sel h17 gnd ground b04 n/c no connect d07 spdsel 3-level sel h18 gnd ground b05 vcc power d08 gnd ground h19 gnd ground b06 ind1+ cml in d09 bre[3] lvttl in pu h20 gnd ground b07 n/c no connect d10 bre[2] lvttl in pu j01 gnd ground b08 gnd ground d11 bre[1] lvt tl in pu j02 gnd ground b09 ind2+ cml in d12 bre[0] lvttl in pu j03 gnd ground b10 n/c no connect d13 gnd ground j04 gnd ground b11 ina1+ cml in d14 n/c no connect j17 rxstb[2] lvttl out b12 n/c no connect d15 gnd ground j18 rxdb[0] lvttl out b13 gnd ground d16 vcc power j19 rxdb[5] lvttl out b14 ina2+ cml in d17 vcc power j20 rxdb[2] lvttl out b15 n/c no connect d18 rxle lvttl in pu k01 rxdc[2] lvttl out b16 vcc power d19 rfen lvttl in pd k02 rxclkc? lvttl out b17 inb1+ cml in d20 n/c no connect k03 gnd ground b18 n/c no connect e01 vcc power k04 lfic lvttl out b19 inb2+ cml in e02 vcc power k17 rxdb[3] lvttl out b20 n/c no connect e03 vcc power k18 rxdb[4] lvttl out c01 tdi lvttl in pu e04 vcc power k19 rxdb[7] lvttl out c02 tms lvttl in pu e17 vcc power k20 rxclkb+ lvttl i/o pd c03 inselc lvttl in e18 vcc power l01 rxdc[3] lvttl out
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 26 of 35 l02 rxclkc+ lvttl i/o pd t17 vcc power v20 rxsta[0] lvttl out l03 gnd ground t18 vcc power w01 vcc power l04 gnd ground t19 vcc power w02 vcc power l17 rxdb[6] lvttl out t20 vcc power w03 lfid lvttl out l18 lfib lvttl out u01 vcc power w04 rxclkd? lvttl out l19 rxclkb? lvttl out u02 vcc power w05 vcc power l20 gnd ground u03 vcc power w06 rxdd[4] lvttl out m01 rxdc[4] lvttl out u04 vcc power w07 rxstd[1] lvttl out m02 rxdc[5] lvttl out u05 vcc power w08 gnd ground m03 rxdc[7] lvttl out u06 rxdd[2] lvttl out w09 n/c no connect m04 rxdc[6] lvttl out u07 rxdd[1] lvttl out w10 gnd ground m17 gnd ground u08 gnd ground w11 gnd ground m18 gnd ground u09 rxopd lvttl 3-s out w12 gnd ground m19 gnd ground u10 n/c no connect w13 gnd ground m20 gnd ground u11 trgclk? pecl in w14 gnd ground n01 gnd ground u12 gnd ground w15 vcc power n02 gnd ground u13 gnd ground w16 vcc power n03 gnd ground u14 gnd ground w17 lfia lvttl out n04 gnd ground u15 vcc power w18 rxclka? lvttl out n17 gnd ground u16 vcc power w19 rxda[4] lvttl out n18 gnd ground u17 rxda[2] lvttl out w20 rxda[1] lvttl out n19 gnd ground u18 rxopa lvttl out y01 vcc power n20 gnd ground u19 rxsta[2] lvttl out y02 vcc power p01 rxdc[1] lvttl out u20 rxsta[1] lvttl out y03 rxdd[7] lvttl out p02 rxdc[0] lvttl out v01 vcc power y04 rxclkd+ lvttl i/o pd p03 rxstc[0] lvttl out v02 vcc power y05 vcc power p04 rxstc[1] lvttl out v03 vcc power y06 rxdd[5] lvttl out p17 gnd ground v04 rxdd[6] lvttl out y07 rxdd[0] lvttl out p18 gnd ground v05 vcc power y08 gnd ground p19 gnd ground v06 rxdd[3] lvttl out y09 n/c no connect p20 gnd ground v07 rxstd[0] lvttl out y10 n/c no connect r01 rxstc[2] lvttl out v08 gnd ground y11 gnd ground r02 rxopc lvttl 3-s out v09 rxstd[2] lvttl out y12 n/c no connect r03 n/c no connect v10 n/c no connect y13 gnd ground r04 vcc power v11 trgclk+ pecl in y14 gnd ground r17 vcc power v12 n/c no connect y15 vcc power r18 vcc power v13 gnd ground y16 vcc power r19 vcc power v14 gnd ground y17 vcc power r20 n/c no connect v15 vcc power y18 rxclka+ lvttl i/o pd t01 vcc power v16 vcc power y19 rxda[6] lvttl out t02 vcc power v17 rxda[7] lvttl out y20 rxda[5] lvttl out t03 vcc power v18 rxda[3] lvttl out t04 vcc power v19 rxda[0] lvttl out table 11.package coordinate signal allocation (continued) ball id signal name signal type ball id signal name signal type ball id signal name signal type
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 27 of 35 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 trans- mission 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 fo r 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 correspo nds to bit 1, as shown below. fc-2 bit designation? 7 6 5 4 3 2 1 0 hotlink d/q designation?7 6 5 4 3 2 1 0 8b/10b bit designation? h g f e d c b a to clarify this correspondence, the following example shows the conversion from an fc-2 valid data byte to a transmission character. fc-2 45h bits: 7654 3210 0100 0101 converted to 8b/10b notation, not e that the order of bits has been reversed): data byte name d5.2 bits: abcde fgh 10100 010 translated to a transmission character in the 8b/10b trans- mission code: bits: abcdei fghj 101001 0101 each valid transmission character of the 8b/10b trans- mission code has been given a name using the following convention: cxx.y, where c is used to show whether the trans- mission character is a data character (c is set to d, and sc/d = low) or a special character (c is set to k, and sc/d = high). when c is set to d, xx is the decimal value of the binary number composed of the bits e, d, c, b, and a in that order, and the y is the decimal value of the bi nary 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 re sulting from the encoding of the unencoded 111 pattern (7). this definition of the 10-bit trans- mission code is based on the following references. a.x. widmer and p.a. franasze k. ?a dc-balanced, parti- tioned-block, 8b/10b transmission code? ibm journal of research and development, 27, no. 5: 440-451 (september, 1983). u.s. patent 4,486,739. peter a. franaszek and albert x. widmer. ?byte-oriented dc balanced (0.4) 8b/10b parti- tioned block transmission code? (december 4, 1984). fibre channel physical and signaling interface (ans x3.230-1994 ansi fc-ph standard). ibm enterprise systems architecture/390 escon i/o interface (document number sa22-7202). 8b/10b transmission code the following information describes how the tables are used for both generating valid transmission characters (encoding) and checking the validity of received transmission characters (decoding). it also specifies the ordering rules to be followed when transmitting the bits within a character and the characters within any higher-level constructs specified by a standard. transmission order within the definition of the 8b/10b transmission code, the bit positions of the transmission characters are labeled a, b, c, d, e, i, f, g, h, j. bit ?a? is tran smitted first followed by bits b, c, d, e, i, f, g, h, and j in that order. note that bit i is transmitted between bit e and bit f, rather than in alphabetical order. valid and invalid transmission characters the following tables define the valid data characters and valid special characters (k characters), respectively. the tables are used for both generating valid transmission characters and checking the validity of received transmission characters. in the tables, each valid-data-byte or special-character-code entry has two columns that represent two transmission characters. the two columns correspond to the current value of the running disparity. running disparity is a binary parameter with either a negative (?) or positive (+) value. after powering on, the transmitter may assume either a positive or negative value for its initial running disparity. upon transmission of any transmission character, the transmitter will select the proper version of the transmission character based on the current running disparity value, and the trans-
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 28 of 35 mitter 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. 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 va lid or invalid according to the following rules and tables and calculates a new value for its running disparity based on t he contents of the received character. the following rules for running disparity are used to calculate the new running-disparity value for transmission characters that have been transmitted and received. running disparity for a transmission character is calculated from 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 trans- mission character is the running disparity at the end of the four-bit sub-block. 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 pos- itive 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 six-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 the appropriate entry in ta ble 14 for the valid data byte or table 15 for special character byte identify which trans- mission character is to be generated. the current value of the transmitter?s running disparity is used to select the trans- mission character from its corresponding column. for each transmission character transmitted, a new value of the running disparity is calculated. this new value is used as the transmitter?s current running disparity for the next valid data byte or special character byte to be encoded and transmitted. table 12 shows naming notations and examples of valid trans- mission characters. 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 trans- mission character is valid and the associated data byte or special character code is determined (decoded). if the received transmission character is not found in that column, then the transmission character is invalid. this is called a code violation. independent of the transmission character?s validity, the received transmission character is used to calculate a new value of running disparity. the new value is used as the receiver?s current running disparity for the next received transmission character. detection of a code violation does not necessarily show that the transmission character in which the code violation was detected is in error. code violations may result from a prior error that altered the running disparity of the bit stream which did not result in a detectable error at the transmission character in which the error occurred. table 12 shows an example of this behavior. table 12.valid transmission characters data byte name d in or q out 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 table 13.code violations resulting from prior errors rd character rd character rd character rd transmitted data character ? d21.1 ? d10.2 ? d23.5 + transmitted bit stream ? 101010 1001 ? 010101 0101 ? 111010 1010 + bit stream after error ? 101010 1011 + 010101 0101 + 111010 1010 + decoded data character ? d21.0 + d10.2 + code violation +
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 29 of 35 table 14.valid data characters (rxstx[2:0] = 000) data byte name bits current rd ? current rd+ data byte name bits current rd ? current rd+ hgf edcba abcdei fghj 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
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 30 of 35 d0.2 010 00000 100111 0101 011000 0101 d0.3 011 00000 100111 0011 011000 1100 d1.2 010 00001 011101 0101 100010 0101 d1.3 011 00001 011101 0011 100010 1100 d2.2 010 00010 101101 0101 010010 0101 d2.3 011 00010 101101 0011 010010 1100 d3.2 010 00011 110001 0101 110001 0101 d3.3 011 00011 110001 1100 110001 0011 d4.2 010 00100 110101 0101 001010 0101 d4.3 011 00100 110101 0011 001010 1100 d5.2 010 00101 101001 0101 101001 0101 d5.3 011 00101 101001 1100 101001 0011 d6.2 010 00110 011001 0101 011001 0101 d6.3 011 00110 011001 1100 011001 0011 d7.2 010 00111 111000 0101 000111 0101 d7.3 011 00111 111000 1100 000111 0011 d8.2 010 01000 111001 0101 000110 0101 d8.3 011 01000 111001 0011 000110 1100 d9.2 010 01001 100101 0101 100101 0101 d9.3 011 01001 100101 1100 100101 0011 d10.2 010 01010 010101 0101 010101 0101 d10.3 011 01010 010101 1100 010101 0011 d11.2 010 01011 110100 0101 110100 0101 d11.3 011 01011 110100 1100 110100 0011 d12.2 010 01100 001101 0101 001101 0101 d12.3 011 01100 001101 1100 001101 0011 d13.2 010 01101 101100 0101 101100 0101 d13.3 011 01101 101100 1100 101100 0011 d14.2 010 01110 011100 0101 011100 0101 d14.3 011 01110 011100 1100 011100 0011 d15.2 010 01111 010111 0101 101000 0101 d15.3 011 01111 010111 0011 101000 1100 d16.2 010 10000 011011 0101 100100 0101 d16.3 011 10000 011011 0011 100100 1100 d17.2 010 10001 100011 0101 100011 0101 d17.3 011 10001 100011 1100 100011 0011 d18.2 010 10010 010011 0101 010011 0101 d18.3 011 10010 010011 1100 010011 0011 d19.2 010 10011 110010 0101 110010 0101 d19.3 011 10011 110010 1100 110010 0011 d20.2 010 10100 001011 0101 001011 0101 d20.3 011 10100 001011 1100 001011 0011 d21.2 010 10101 101010 0101 101010 0101 d21.3 011 10101 101010 1100 101010 0011 d22.2 010 10110 011010 0101 011010 0101 d22.3 011 10110 011010 1100 011010 0011 d23.2 010 10111 111010 0101 000101 0101 d23.3 011 10111 111010 0011 000101 1100 d24.2 010 11000 110011 0101 001100 0101 d24.3 011 11000 110011 0011 001100 1100 d25.2 010 11001 100110 0101 100110 0101 d25.3 011 11001 100110 1100 100110 0011 d26.2 010 11010 010110 0101 010110 0101 d26.3 011 11010 010110 1100 010110 0011 d27.2 010 11011 110110 0101 001001 0101 d27.3 011 11011 110110 0011 001001 1100 d28.2 010 11100 001110 0101 001110 0101 d28.3 011 11100 001110 1100 001110 0011 d29.2 010 11101 101110 0101 010001 0101 d29.3 011 11101 101110 0011 010001 1100 d30.2 010 11110 011110 0101 100001 0101 d30.3 011 11110 011110 0011 100001 1100 d31.2 010 11111 101011 0101 010100 0101 d31.3 011 11111 101011 0011 010100 1100 d0.4 100 00000 100111 0010 011000 1101 d0.5 101 00000 100111 1010 011000 1010 table 14.valid data characters (rxstx[2:0] = 000) (continued) data byte name bits current rd ? current rd+ data byte name bits current rd ? current rd+ hgf edcba abcdei fghj abcdei fghj hgf edcba abcdei fghj abcdei fghj
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 31 of 35 d1.4 100 00001 011101 0010 100010 1101 d1.5 101 00001 011101 1010 100010 1010 d2.4 100 00010 101101 0010 010010 1101 d2.5 101 00010 101101 1010 010010 1010 d3.4 100 00011 110001 1101 110001 0010 d3.5 101 00011 110001 1010 110001 1010 d4.4 100 00100 110101 0010 001010 1101 d4.5 101 00100 110101 1010 001010 1010 d5.4 100 00101 101001 1101 101001 0010 d5.5 101 00101 101001 1010 101001 1010 d6.4 100 00110 011001 1101 011001 0010 d6.5 101 00110 011001 1010 011001 1010 d7.4 100 00111 111000 1101 000111 0010 d7.5 101 00111 111000 1010 000111 1010 d8.4 100 01000 111001 0010 000110 1101 d8.5 101 01000 111001 1010 000110 1010 d9.4 100 01001 100101 1101 100101 0010 d9.5 101 01001 100101 1010 100101 1010 d10.4 100 01010 010101 1101 010101 0010 d10.5 101 01010 010101 1010 010101 1010 d11.4 100 01011 110100 1101 110100 0010 d11.5 101 01011 110100 1010 110100 1010 d12.4 100 01100 001101 1101 001101 0010 d12.5 101 01100 001101 1010 001101 1010 d13.4 100 01101 101100 1101 101100 0010 d13.5 101 01101 101100 1010 101100 1010 d14.4 100 01110 011100 1101 011100 0010 d14.5 101 01110 011100 1010 011100 1010 d15.4 100 01111 010111 0010 101000 1101 d15.5 101 01111 010111 1010 101000 1010 d16.4 100 10000 011011 0010 100100 1101 d16.5 101 10000 011011 1010 100100 1010 d17.4 100 10001 100011 1101 100011 0010 d17.5 101 10001 100011 1010 100011 1010 d18.4 100 10010 010011 1101 010011 0010 d18.5 101 10010 010011 1010 010011 1010 d19.4 100 10011 110010 1101 110010 0010 d19.5 101 10011 110010 1010 110010 1010 d20.4 100 10100 001011 1101 001011 0010 d20.5 101 10100 001011 1010 001011 1010 d21.4 100 10101 101010 1101 101010 0010 d21.5 101 10101 101010 1010 101010 1010 d22.4 100 10110 011010 1101 011010 0010 d22.5 101 10110 011010 1010 011010 1010 d23.4 100 10111 111010 0010 000101 1101 d23.5 101 10111 111010 1010 000101 1010 d24.4 100 11000 110011 0010 001100 1101 d24.5 101 11000 110011 1010 001100 1010 d25.4 100 11001 100110 1101 100110 0010 d25.5 101 11001 100110 1010 100110 1010 d26.4 100 11010 010110 1101 010110 0010 d26.5 101 11010 010110 1010 010110 1010 d27.4 100 11011 110110 0010 001001 1101 d27.5 101 11011 110110 1010 001001 1010 d28.4 100 11100 001110 1101 001110 0010 d28.5 101 11100 001110 1010 001110 1010 d29.4 100 11101 101110 0010 010001 1101 d29.5 101 11101 101110 1010 010001 1010 d30.4 100 11110 011110 0010 100001 1101 d30.5 101 11110 011110 1010 100001 1010 d31.4 100 11111 101011 0010 010100 1101 d31.5 101 11111 101011 1010 010100 1010 table 14.valid data characters (rxstx[2:0] = 000) (continued) data byte name bits current rd ? current rd+ data byte name bits current rd ? current rd+ hgf edcba abcdei fghj abcdei fghj hgf edcba abcdei fghj abcdei fghj
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 32 of 35 d0.6 110 00000 100111 0110 011000 0110 d0.7 111 00000 100111 0001 011000 1110 d1.6 110 00001 011101 0110 100010 0110 d1.7 111 00001 011101 0001 100010 1110 d2.6 110 00010 101101 0110 010010 0110 d2.7 111 00010 101101 0001 010010 1110 d3.6 110 00011 110001 0110 110001 0110 d3.7 111 00011 110001 1110 110001 0001 d4.6 110 00100 110101 0110 001010 0110 d4.7 111 00100 110101 0001 001010 1110 d5.6 110 00101 101001 0110 101001 0110 d5.7 111 00101 101001 1110 101001 0001 d6.6 110 00110 011001 0110 011001 0110 d6.7 111 00110 011001 1110 011001 0001 d7.6 110 00111 111000 0110 000111 0110 d7.7 111 00111 111000 1110 000111 0001 d8.6 110 01000 111001 0110 000110 0110 d8.7 111 01000 111001 0001 000110 1110 d9.6 110 01001 100101 0110 100101 0110 d9.7 111 01001 100101 1110 100101 0001 d10.6 110 01010 010101 0110 010101 0110 d10.7 111 01010 010101 1110 010101 0001 d11.6 110 01011 110100 0110 110100 0110 d11.7 111 01011 110100 1110 110100 1000 d12.6 110 01100 001101 0110 001101 0110 d12.7 111 01100 001101 1110 001101 0001 d13.6 110 01101 101100 0110 101100 0110 d13.7 111 01101 101100 1110 101100 1000 d14.6 110 01110 011100 0110 011100 0110 d14.7 111 01110 011100 1110 011100 1000 d15.6 110 01111 010111 0110 101000 0110 d15.7 111 01111 010111 0001 101000 1110 d16.6 110 10000 011011 0110 100100 0110 d16.7 111 10000 011011 0001 100100 1110 d17.6 110 10001 100011 0110 100011 0110 d17.7 111 10001 100011 0111 100011 0001 d18.6 110 10010 010011 0110 010011 0110 d18.7 111 10010 010011 0111 010011 0001 d19.6 110 10011 110010 0110 110010 0110 d19.7 111 10011 110010 1110 110010 0001 d20.6 110 10100 001011 0110 001011 0110 d20.7 111 10100 001011 0111 001011 0001 d21.6 110 10101 101010 0110 101010 0110 d21.7 111 10101 101010 1110 101010 0001 d22.6 110 10110 011010 0110 011010 0110 d22.7 111 10110 011010 1110 011010 0001 d23.6 110 10111 111010 0110 000101 0110 d23.7 111 10111 111010 0001 000101 1110 d24.6 110 11000 110011 0110 001100 0110 d24.7 111 11000 110011 0001 001100 1110 d25.6 110 11001 100110 0110 100110 0110 d25.7 111 11001 100110 1110 100110 0001 d26.6 110 11010 010110 0110 010110 0110 d26.7 111 11010 010110 1110 010110 0001 d27.6 110 11011 110110 0110 001001 0110 d27.7 111 11011 110110 0001 001001 1110 d28.6 110 11100 001110 0110 001110 0110 d28.7 111 11100 001110 1110 001110 0001 d29.6 110 11101 101110 0110 010001 0110 d29.7 111 11101 101110 0001 010001 1110 d30.6 110 11110 011110 0110 100001 0110 d30.7 111 11110 011110 0001 100001 1110 d31.6 110 11111 101011 0110 010100 0110 d31.7 111 11111 101011 0001 010100 1110 table 14.valid data characters (rxstx[2:0] = 000) (continued) data byte name bits current rd ? current rd+ data byte name bits current rd ? current rd+ hgf edcba abcdei fghj abcdei fghj hgf edcba abcdei fghj abcdei fghj
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 33 of 35 table 15.valid special character codes and sequences (rxstx[2:0] = 001) [32, 33] s.c. code name s.c. byte name current rd ? abcdei fghj current rd+ abcdei fghj cypress alternate s.c. byte name [34] bits hgf edcba s.c. byte name [34] bits hgf edcba k28.0 c0.0 (c00) 000 00000 c28.0 (c1c) 000 11100 001111 0100 110000 1011 k28.1 [35] c1.0 (c01) 000 00001 c28.1 (c3c) 001 11100 001111 1001 110000 0110 k28.2 [35] 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 [35] c4.0 (c04) 000 00100 c28.4 (c9c) 100 11100 001111 0010 110000 1101 k28.5 [35, 36] c5.0 (c05) 000 00101 c28.5 (cbc) 101 11100 001111 1010 110000 0101 k28.6 [35] c6.0 (c06) 000 00110 c28.6 (cdc) 110 11100 001111 0110 110000 1001 k28.7 [35, 37] 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 end of frame sequence eofxx [39] c2.1 (c22) 001 00010 c2.1 (c22) 001 00010 ?k28.5, dn.xxx0 +k28.5, dn.xxx1 code rule violatio n and svs tx pattern exception [37, 40] c0.7 (ce0) 111 00000 c0.7 (ce0) 111 00000 [43] 100111 1000 011000 0111 ? k28.5 [41] c1.7 (ce1) 111 00001 c1.7 (ce1) 111 00001 [43] 001111 1010 001111 1010 +k28.5 [42] c2.7 (ce2) 111 00010 c2.7 (ce2) 111 00010 [43] 110000 0101 110000 0101 running disparity violation pattern exception [42] c4.7 (ce4) 111 00100 c4.7 (ce4) 111 00100 [43] 110111 0101 001000 1010 notes: 32. all codes not shown are reserved. 33. notation for special character code name is consistent with fibre channel and escon naming conven tions. special character co de name is intended to describe binary information present on i/o pi ns. common usage for the name can either be in the form used for describing data p atterns (i.e., c0.0 through c31.7), or in hex notation (i.e., cnn where nn = the specified value between 00 and ff). 34. both the cypress and alternate encodings may be used for data tran smission to generate specific special character codes. the decoding process for received characters generates cypress codes or alternate c odes as selected by the decmode configuration input. 35. these characters are used for control of escon interfaces. th ey can be sent as embedded comma nds or other markers when not o perating using escon protocols. 36. the k28.5 character is used for framing op erations by the receiver. it is also the pad or fill character transmitted to main tain the serial link when no user data is available. 37. 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 c ause erroneous framing and should be avoided while rfen = high. 38. c2.1 = transmit either ? k28.5+ or +k28.5 ? as determined by current rd and modify the transmission character that follows, by sett ing 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 wherei n 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 ii 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 ii 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 k 28.5 is decoded as c5.0, c1.7, or c2.7, and the subsequent bytes are decoded as data. 39. c0.7 = transmit a deliberate code rule vi olation. the code chosen for this function follows the normal running disparity rul es. the receiver will only output this special character if the transmission characte r being decoded is not found in the tables. 40. c1.7 = transmit negative k28.5 ( ? k28.5+) disregarding current rd. the receiv er will only output this special characte r 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. 41. 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. 42. c4.7 = transmit a deliberate code rule violation to indicate a running disparity violation. the receiver will only output th is 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 occurre d in a prior byte. 43. supported only for data transmission. the receive status for thes e conditions will be reported by specific combinations of r eceive status bits.
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 34 of 35 ? cypress semiconductor corporation, 2005. the information contained herein is subject to change without notice. cypress semic onductor 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 ot her 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 agr eement 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 re sult in significant injury to the user. the inclusion of cypress products in life-support systems application implies that the manu facturer assumes all risk of such use and in doing so indemni fies cypress against all charges. package diagram hotlink is a registered trademark, and hotlink ii, and multif rame are trademarks, of cypress semiconductor. ibm and escon are registered trademarks, and ficon is a trademark, of inte rnational business machines. all product and company names mentioned in this document are the trademarks of their respective holders. ordering information speed ordering code package name package type operating range standard CYP15G0401RB-BGC bl256 256-ball thermally enhanced ball grid array commercial standard cyp15g0401rb-bgi bl256 256-ball thermally enhanced ball grid array industrial standard cyp15g0401rb-bgxc bl256 pb free 256-ball thermally enhanced ball grid array commercial standard cyp15g0401rb-bgxi bl256 pb free 256-ball thermally enhanced ball grid array industrial a b 0.15 c 0.15 c 0.97 ref. 0.600.10 1.570.175 c 0.20 min seating plane side view top view 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 a b c d e f g h j k l m n p r t u v w y a ?0.15 m c ?0.30 m c ?0.750.15(256x) b a 1.27 bottom view (ball side) a 0.20(4x) top of mold compound to top of balls 26 typ. a1 corner i.d. 0.50 min. 27.000.13 27.000.13 24.13 24.13 a1 corner i.d. r 2.5 max (4x) 12.065 256-lead l2 ball grid array (27 x 27 x 1.57 mm) bl256 51-85123-*e
preliminary cyp15g0401rb document #: 38-02111 rev. ** page 35 of 35 document history page document title: cyp15g0401rb quad hotlink ii? receiver document number: 38-02111 rev. ecn no. issue date orig. of change description of change ** 318023 see ecn rev new data sheet

▲Up To Search▲

Price & Availability of CYP15G0401RB-BGC

	To Download CYP15G0401RB-BGC Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .