View CY28551_4492368.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

universal clock generator for intel, via, and sis ? CY28551 cypress semiconductor corporation ? 198 champion court ? san jose , ca 95134-1709 ? 408-943-2600 document #: 001-05675 rev. *c revised july 26, 2006 features ? compliant to intel ? ck505 ? selectable cpu clock buffer type for intel p4 or k8 selection ? selectable cpu frequencies ? universal clock to support intel, sis and via platform ? 0.7v differential cpu clock for intel cpu ? 3.3v differential cpu clock for amd k8 ? 100 mhz differential src clocks ? 96 mhz differential dot clock ? 133 mhz link clock ? 48 mhz usb clock ? 33 mhz pci clocks ? dynamic frequency control ? dial-a-frequency ? ? watchdog timer ? two independent overclocking plls ? low-voltage frequency select input ?i 2 c support with readback capabilities ? ideal lexmark spread spectrum profile for maximum electromagnetic interference (emi) reduction ? 3.3v power supply ? 64-pin qfn package cpu src sata pci ref link dot96 24_48m 48m x 2 x 8 x1 x 7 x 3 x2 x 1 x1 x 1 pci5/*sel0 pci4/*selp4_k8 vddpci pc3/*fsb pci2/**fsa vsspci pci1/clkreq#a pci0/clkreq#b **doc1 reset_i#/ sreset# ref0/ **fsd ref1 /**fsc ref2/**mode vssref xin xout 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 pci6_f 1 48 vddref vdd48 247sclk **sel24_48 / 24_48m 3 46 sdata **sel1/48m 4 45 vttpwrg#/pd vss48 5 44 cput0 vdddot 6 43 cpuc0 link0/dot96t/satat 7 42 vddcpu link1/dot96c/satac 8 41 cput1 vssdot 9 40 cpuc1 vddsata 10 39 vsscpu satat/pciext0 11 38 **doc2 satac/pciexc0 12 37 vssa vsssata 13 36 vdda pciext1 14 35 pciext8/cpu_stp# pciexc1 15 34 pciexc8/pci_stp# vsspcie 16 33 vddpcie 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 pciext2 pciexc2 vddpcie pciext3 pciexc3 vsspcie pciext4 pciexc4 vddpcie pciexc5 pciext5 pciexc6 pciext6 vsspcie pciexc7 pciext7 CY28551 block diagram pin configuration * indicates internal pull up ** indicates internal pull down vdd_ref vdd_cpu vdd_pci ex t pwr_gd#/ pd pll reference fs[d:a] pll1 cpu vdd_dot pll3 sata divider divider pll4 fixed divider ref[ 2: 0] cput[ 1: 0] cpuc[1:0] pciet [8:1] pciec [8:1] dot96t/satat/link0 dot96c/ satac/link1 i2c logic sdata sclk 14.318mhz crystal xi n xout vdd_48 48m vdd_pci pci[ 6:0] doc[2:1] pll2 pciex vdd_sata pci et0 /satat pciec0 /satac wdt sreset# 24_48m sel[1:0] sel_p4_k8 sel24_48 multiplexer controller di vi der reset_i#
CY28551 document #: 001-05675 rev. *c page 2 of 30 pin description pin no. name type description 1 pci6_f o free running 33 mhz clock output. intel type-3a output buffer 2 vdd48 pwr 3.3v power supply for outputs. 3 **sel24_48#/24_4 8m i/o, pd 3.3v tolerant input for 24 mhz, 48 mhz selection/24_48mhz clock output. internal 150k pull down 1 = 24 mhz, 0 = 48 mhz intel type-3a output buffer 4 **sel1/48mhz i/o, pd 3.3v tolerant input for ou tput selection/48mhz clock output. refer to table 1 for selection options internal 150k pull down 5 vss48 gnd ground for outputs 6 vdddot pwr 3.3v power supply for outputs 7,8 link0/dot96t/sa tat link1/dot96c/sa tac o, se/dif link output for via and sis, differential 96 mhz clock output and 100 mhz differential clock. the output is selected by sel[1:0] 9 vssdot gnd ground for outputs 10 vddsata pwr 3.3v power supply for outputs 11,12 pciex0[t/c]/sata [t/c] o, dif differential src clock output /differential sata src clock output intel type-sr output buffer 13 vsssata gnd ground for outputs 14,15 pciex[t/c]1 o, di f 100 mhz differential serial reference clock. intel type-sr output buffer 16 vsspcie gnd ground for outputs 17,18 pciex[t/c]2 o, di f 100 mhz differential serial reference clock. intel type-sr output buffer 19 vddpcie pwr 3.3v power supply for outputs. 20,21 pciex[t/c]3 o, di f 100 mhz differential serial reference clock. intel type-sr output buffer 22 vsspcie gnd ground for outputs 23,24 pciex[t/c]4 o, di f 100 mhz differential serial reference clock. intel type-sr output buffer 25 vddpcie pwr 3.3v power supply for outputs 26,27,28,29 pciex[t/c][5 :6] o, dif 100 mhz differential serial refe rence clock. intel type-sr output buffer 30 vsspcie gnd ground for outputs 31,32 pciex[t/c]7 o, di f 100 mhz differential serial reference clock. intel type-sr output buffer 33 vddpcie pwr 3.3v power supply for outputs 34,35 pciext8/cpu_st op# pciexc8/pci_st op# i/o, dif 3.3v-tolerant input for stopping pci and src outputs/3.3v-tolerant input for stopping cpu outputs/100-mhz diff erential serial reference clocks. the two multifunction pins are se lected by mode. default pciex8 intel type-sr output buffer 36 vdda pwr 3.3v power supply for pll. 37 vssa gnd ground for pll. 38 **doc2 i, pd dynamic over clocking pin 0 = normal, 1 = frequency will be changed depend on doc register. internal 150k pull-down. 39 vsscpu gnd ground for outputs. 40,41 cpu[t/c]1 o, dif differential cpu cloc k output. intel type-sr output buffer. 42 vddcpu pwr 3.3v power supply for outputs+ 43, 44 cpu[t/c]0 o, dif differential cpu cl ock output. intel type-sr output buffer.
CY28551 document #: 001-05675 rev. *c page 3 of 30 45 vtt_pwrgd#/pd i 3.3v lvttl input. this pin is a le vel-sensitive strobe used to latch the hw strapping pin inputs. after asserting vtt_pwrgd# (active low), this pin becomes a real-time input for asserting power-down (active high). 46 sdata i/o smbus compatible sdata 47 sclk i smbus compatible sclock . 48 vddref pwr 3.3v power supply for outputs 49 xout o 14.318 mhz crystal output 50 xin i 14.318 mhz crystal input 51 vssref gnd ground for outputs 52 ref2 o, se 14.318 mhz ref clock output. intel type-5 output buffer 53 **fsc/ref1 i/o,pd, se 3.3v tolerant input for cpu frequency selection/14.318 mhz ref clock output internal 150k pull down intel type-5 output buffer refer to dc electrical specifications table for vil_fs and vih_fs specifications 54 **fsd/ref0 i/o,pd, se 3.3v tolerant input for cpu frequency selection/14.318 mhz ref clock output internal 150k pull down intel type-5 output buffer refer to dc electrical specifications table for vil_fs and vih_fs specifications 55 reset_i#/srese t# i/o, od 3.3v tolerant input for rese t all of registers to default setting 3.3v lvttl output for watchdog reset signal 56 **doc1 i, pd dynamic over clocking pin 0 = normal; 1 = frequency will be changed depend on doc register. internal 150k pull-down 57 pci0/**clkreq#b i/o,se, pd 33 mhz clock output/output enable cont rol for pciex4; 5 via i2c register default is pci0 0 = selected pciexs are enabled, 1 = selected pciexs are disabled. internal 150k pull down intel type-3a output buffer 58 pci1/**clkreq#a i/o,se, pd 33 mhz clock output/output enab le control for pciex6, 7via i2c register. default is pci1 0 = selected pciexs are enabled, 1 = selected pciexs are disabled. internal 150k pull down intel type-3a output buffer 59 vsspci gnd grou nd for outputs. 60 **fsa/pci2 i/o, pd 3.3v tolerant input for cpu frequency selection/33 mhz clock output. internal 150k pull down intel type-3a output buffer refer to dc electrical specifications table for vil_fs and vih_fs specifications 61 *fsb/pci3 i/o, pu 3.3v tolerant input for cpu frequ ency selection/33 mhz clock output. internal 150k pull up intel type-3a output buffer refer to dc electrical specifications table for vil_fs and vih_fs specifications 62 vddpci pwr 3.3v power supply for outputs. 63 *selp4_k8/pci3 i/o, pu 3.3v tolerant input for cpu clock output buffer type selection/33 mhz clock output. internal 150k pull up intel type-3a output buffer refer to dc electrical specifications table for vil_fs and vih_fs specifications 0 = k8 cpu buffer type, 1 = p4 cpu buffer type. 64 *sel0/pci5 i/o, pu 3.3v tolerant input for ou tput selection/33 mhz clock output. refer to table 1 for selection options. internal 150k pull up pin description (continued) pin no. name type description
CY28551 document #: 001-05675 rev. *c page 4 of 30 frequency select pins (fs[d:a]) to achieve host clock frequency selection, apply the appro- priate logic levels to fs_a, fs_b, fs_c, and fs_d inputs prior to vtt_pwrgd# assertion (as seen by the clock synthe- sizer). when vtt_pwrgd# is sampled low by the clock chip (indicating processor vtt voltage is stable), the clock chip samples the fs_a, fs_b, fs_c, and fs_d input values. for all logic levels of fs_a, fs_b, fs_c, fs_d, and fs_e, vtt_pwrgd# employs a one-shot functionality, in that once a valid low on vtt_pwrgd# has been sampled, all further vtt_pwrgd#, fs_a, fs_b, fs_c, and fs_d transitions will be ignored, except in test mode. serial data interface to enhance the flexibility and function of the clock synthesizer, a two-signal serial interface is provided. through the serial data interface, various device functions, such as individual clock output buffers, can be individually enabled or disabled. the registers associated with the serial data interface initialize to their default setting upon power-up, and therefore use of this interface is optional. clock device register changes are normally made upon system initialization, if any are required. the inte rface cannot be used during system operation for power management functions. data protocol the clock driver serial protocol accepts byte write, byte read, block write, and block read operations from the controller. for block write/read operation, the bytes must be accessed in sequential order from lowest to highest byte (most significant bit first) with the ability to st op after any complete byte has been transferred. for byte writ e and byte read operations, the system controller can access indi vidually indexed bytes. the offset of the indexed byte is encoded in the command code, as described in table 2 . the block write and block read protocol is outlined in table 3 , while table 4 outlines the correspondi ng byte write and byte read protocol. the slave receiver address is 11010010 (d2h). table 1. frequency select table fsd fsc fsb fsa frequency table (rom) fsel3 fsel2 fsel1 fsel0 cpu0 cpu1 src link pci cpu vco cpu pll gear constant (g) cpu m cpu n pcie vco src pll gear constant pcie m pcie n 0 0 0 0 266.6666667 266.6666667 100 66.6667 33.3333 800 80 60 200 800 30 60 200 0 0 0 1 133.3333333 133.3333333 100 66.6667 33.3333 800 40 60 200 800 30 60 200 0 0 1 0 200 200 100 66.6667 33.3333 800 60 60 200 800 30 60 200 0 0 1 1 166.6666667 166.6666667 100 66.6667 33.3333 666.6666667 60 63 175 800 30 60 200 0 1 0 0 333.3333333 333.3333333 100 66.6667 33.3333 666.6666667 120 63 175 800 30 60 200 0 1 0 1 100 100 100 66.6667 33.3333 800 30 60 200 800 30 60 200 0 1 1 0 400 400 100 66.6667 33.3333 800 120 60 200 800 30 60 200 0 1 1 1 200 250 100 66.6667 33.3333 1000 60 60 250 800 30 60 200 1 0 0 0 266.6666667 266.6666667 100 133.3333 33.3333 800 80 60 200 800 30 60 200 1 0 0 1 133.3333333 133.3333333 100 133.3333 33.3333 800 40 60 200 800 30 60 200 1 0 1 0 200 200 100 133.3333 33.3333 800 60 60 200 800 30 60 200 1 0 1 1 166.6666667 166.6666667 100 133.3333 33.3333 666.6666667 60 63 175 800 30 60 200 1 1 0 0 333.3333333 333.3333333 100 133.3333 33.3333 666.6666667 120 63 175 800 30 60 200 1 1 0 1 100 100 100 133.3333 33.3333 800 30 60 200 800 30 60 200 1 1 1 0 400 400 100 133.3333 33.3333 800 120 60 200 800 30 60 200 1 1 1 1 200 250 100 133.3333 33.3333 1000 60 60 250 800 30 60 200 table 2. command code definition bit description 7 0 = block read or block write operation, 1 = byte read or byte write operation (6:0) byte offset for byte read or byte write operation. fo r block read or block write operations, these bits should be '0000000'
CY28551 document #: 001-05675 rev. *c page 5 of 30 table 3. block read and block write protocol block write protocol block read protocol bit description bit description 1 start 1 start 8:2 slave address ? 7 bits 8:2 slave address ? 7 bits 9 write 9 write 10 acknowledge from slave 10 acknowledge from slave 18:11 command code ? 8 bits 18:11 command code ? 8 bits 19 acknowledge from slave 19 acknowledge from slave 27:20 byte count ? 8 bits (skip this step if i 2 c_en bit set) 20 repeat start 28 acknowledge from slave 27:21 slave address ? 7 bits 36:29 data byte 1 ? 8 bits 28 read = 1 37 acknowledge from slave 29 acknowledge from slave 45:38 data byte 2 ? 8 bits 37:30 byte count from slave ? 8 bits 46 acknowledge from slave 38 acknowledge .... data byte/slave acknowledges 46:39 data byte 1 from slave ? 8 bits .... data byte n ? 8 bits 47 acknowledge .... acknowledge from slave 55:48 data byte 2 from slave ? 8 bits .... stop 56 acknowledge .... data bytes from slave/acknowledge .... data byte n from slave ? 8 bits .... not acknowledge .... stop table 4. byte read and byte write protocol byte write protocol byte read protocol bit description bit description 1start 1start 8:2 slave address ? 7 bits 8:2 slave address ? 7 bits 9write 9write 10 acknowledge from slave 10 acknowledge from slave 18:11 command code ? 8 bits 18:11 command code ? 8 bits 19 acknowledge from slave 19 acknowledge from slave 27:20 data byte ? 8 bits 20 repeated start 28 acknowledge from slave 27:21 slave address ? 7 bits 29 stop 28 read 29 acknowledge from slave 37:30 data from slave ? 8 bits 38 not acknowledge 39 stop
CY28551 document #: 001-05675 rev. *c page 6 of 30 control registers byte 0: control register 0 bit @pup type name description 7 1 r/w pciex[t/c]7 pciex[ t/c]7 output enable 0 = disable (tri-state), 1 = enable 6 1 r/w pciex[t/c]6 pciex[ t/c]6 output enable 0 = disable (tri-state), 1 = enable 5 1 r/w pciex[t/c]5 pciex[ t/c]5 output enable 0 = disable (tri-state), 1 = enable 4 1 r/w pciex[t/c]4 pciex[ t/c]4 output enable 0 = disable (tri-state), 1 = enable 3 1 r/w pciex[t/c]3 pciex[ t/c]3 output enable 0 = disable (tri-state), 1 = enable 2 1 r/w pciex[t/c]2 pciex[ t/c]2 output enable 0 = disable (tri-state), 1 = enable 1 1 r/w pciex[t/c]1 pciex[ t/c]1 output enable 0 = disable (tri-state), 1 = enable 0 1 r/w sata/pciex[t/c]0 sata/ pciex[t/c]0 output enable 0 = disable (tri-state), 1 = enable byte 1: control register 1 bit @pup type name description 7 1 r/w sata/dot96] sata/dot96output enable 0 = disable (tri-state), 1 = enable 6 1 r/w 24_48m 24_48m output enable 0 = disable, 1 = enable 5 1 r/w 48m 48m output enable 0 = disable, 1 = enable 4 1 r/w ref2 ref2 output enable 0 = disable, 1 = enable 3 1 r/w ref1 ref1 output enable 0 = disable, 1 = enable 2 1 r/w ref0 ref0 output enable 0 = disable, 1 = enable 1 1 r/w cpu[t/c]1 cpu[t/c]1 output enable 0 = disable (tri-state), 1 = enable 0 1 r/w cpu[t/c]0 cpu[t/c]0 output enable 0 = disable (tri-state), 1 = enable byte 2: control register 2 bit @pup type name description 7 1 r/w reserved reserved 6 1 r/w pci6_f pci6_f output enable 0 = disable, 1 = enable 5 1 r/w pci5 pci5 output enable 0 = disable, 1 = enable 4 1 r/w pci4 pci4 output enable 0 = disable, 1 = enable
CY28551 document #: 001-05675 rev. *c page 7 of 30 3 1 r/w pci3 pci3 output enable 0 = disable, 1 = enable 2 1 r/w pci2 pci2 output enable 0 = disable, 1 = enable 1 1 r/w pci1 pci1 output enable 0 = disable, 1 = enable 0 1 r/w pci0 pci0 output enable 0 = disable, 1 = enable byte 2: control register 2 (continued) bit @pup type name description byte 3: control register 3 bit @pup type name description 7 1 r/w link1 link1 output enable 0 = disable, 1 = enable 6 1 r/w link0 linki0 output enable 0 = disable, 1 = enable 5 1 r/w pciex[t/c]8 pciex[ t/c]8 output enable 0 = disable (tri-state), 1 = enable 4 1 r/w reserved reserved 3 0 r/w reserved reserved 2 1 r/w pci 33 mhz output drive strength 0 = 2x, 1 = 1x 1 1 r/w ref ref output drive strength 0 = 2x, 1 = 1x 0 0 r/w 48m, 24_48m 48 mhz and 24_48m output drive strength 0 = 2x, 1 = 1x byte 4: control register 4 bit @pup type name description 7 0 r/w cpu1 allow control of cpu1 with assertion of cpu_stp# 0 = free running 1 = stopped with cpu_stp# 6 0 r/w cpu0 allow control of cpu0 with assertion of cpu_stp# 0 = free running 1 = stopped with cpu_stp# 5 0 r/w pci6_f allow control of pci6_f with assertion of pci_stp# 0 = free running 1 = stopped with pci_stp# 4 0 r/w pciex allow control of pc iex with assert ion of pci_stp# 0 = free running 1 = stopped with pci_stp# 3 0 r/w fsel_d sw frequency selection bits. see table 1 . 2 0 r/w fsel_c 1 0 r/w fsel_b 0 0 r/w fsel_a
CY28551 document #: 001-05675 rev. *c page 8 of 30 byte 5: control register 5 bit @pup type name description 7 0 r/w cpu_ss1 cpu (pll1) spread spectrum selection 00: ?0.5% (peak to peak) 01: 0.25% (peak to peak) 10: ?1.0% (peak to peak) 11: 0.5% (peak to peak) 60 r/w cpu_ss0 5 0 r/w cpu_ss_off pll1 (cpupll) spread spectrum enable 0 = spread off, 1 = spread on 4 0 r/w pcie_ss0 pll2 (pciepll) spread spectrum selection 0: ?0.5% (peak to peak) 0: ?1.0% (peak to peak) 3 0 r/w pcie_ss_off pll2 (pciepll) spread spectrum enable 0 = src spread off, 1 = src spread on 2 0 r/w sata_ss_off pll3 (satapl l) spread spectrum enable 0 = spread off, 1 = spread on 1 hw r/w sel24_48 24m/48 mhz output selection 0 = 48 mhz, 1 = 24 mhz 0 1 r/w reserved reserved byte 6: control register 6 bit @pup type name description 7 0 r/w sw_reset software reset. when set, the device asserts a reset signal on sreset# upon completion of the block/word/byte wr ite that set it. after asserting and deasserting the sreset# this bit will self clear (set to 0). 6 0 r/w reserved reserved 5 0 r/w fix_link_pci link and pci clock source selection 0 = pll2(srcpll), 1 = pll (satapll) 4 hw r fsd fsd reflects the value of the fsd pin sampled on power up. 0 = fsd was low during vtt_pwrgd# assertion. 3 hw r fsc fsc reflects the value of the fsc pin sampled on power up. 0 = fsc was low during vtt_pwrgd# assertion. 2 hw r fsb fsb reflects the value of the fsb pin sampled on power up. 0 = fsb was low during vtt_pwrgd# assertion 1 hw r fsa fsa reflects the value of the fsa pin sampled on power up. 0 = fsa was low during vtt_pwrgd# assertion 0 hw r powergood power status bit: 0 = internal power or internal resets are not valid 1 = internal power and internal resets are valid read only bit 7 sets to 0 when bit 7 = 0 byte 7: vendor id bit @pup type name description 7 0 r revision code bit 3 revision code bit 3 6 0 r revision code bit 2 revision code bit 2 5 1 r revision code bit 1 revision code bit 1 4 0 r revision code bit 0 revision code bit 0 3 1 r vendor id bit 3 vendor id bit 3 2 0 r vendor id bit 2 vendor id bit 2
CY28551 document #: 001-05675 rev. *c page 9 of 30 1 0 r vendor id bit 1 vendor id bit 1 0 0 r vendor id bit 0 vendor id bit 0 byte 8: control register 8 bit @pup type name description 7 0 r/w cr1_pciex7 pciex[t/ c75 clkreq#a control 1 = pciex [t/c]5 stoppable by clkreq#b pin 0 = free running 6 0 r/w cr1_pciex6 pciex[t/ c]6 clkreq#a control 1 = pciex [t/c]4 stoppable by clkreq#b pin 0 = free running 5 0 r/w cr1_pciex5 pciex[t/ c]5 clkreq#b control 1 = pciex [t/c]5 stoppable by clkreq#b pin 0 = free running 4 0 r/w cr1_pciex4 pciex[t/ c]4 clkreq#b control 1 = pciex [t/c]4 stoppable by clkreq#b pin 0 = free running 3 0 r/w reserved reserved, set = 0 2 0 r/w reserved reserved, set = 0 1 0 r/w reserved reserved, set = 0 0 0 r/w reserved reserved, set = 0 byte 7: vendor id (continued) bit @pup type name description byte 9: control register 9 bit @pup type name description 7 0 r/w df3_n8 the df3_n[8:0] configures cpu frequency for dynamic frequency. doc[1:2] =11 6 0 r/w df2_n8 the df2_n[8:0] configures cpu frequency for dynamic frequency. doc[1:2] =10 5 0 r/w df1_n8 the df1_n[8:0] configures cpu frequency for dynamic frequency. doc[1:2] =01 4 0 r/w reserved reserved, set = 0 3 0 r/w reserved reserved, set = 0 2 1 r/w smsw_bypass smooth switch bypass 0 = activate smsw block 1 = bypass and deactivate smsw block. 1 0 r/w smsw_sel smooth switch select 0 = select cpu_pll 1 = select src_pll 0 0 r/w reserved reserved, set = 0
CY28551 document #: 001-05675 rev. *c page 10 of 30 byte 10: control register 10 bit @pup type name description 7 0 r/w df1_n7 the df1_n[8:0] configures cpu frequency for dynamic frequency. doc[1:2] =01. 6 0 r/w df1_n6 5 0 r/w df1_n5 4 0 r/w df1_n4 3 0 r/w df1_n3 2 0 r/w df1_n2 1 0 r/w df1_n1 0 0 r/w df1_n0 byte 11: contro l register 11 bit @pup type name description 7 0 r/w df2_n7 the df2_n[8:0] configures cpu frequency for dynamic frequency. doc[1:2] =10 6 0 r/w df2_n6 5 0 r/w df2_n5 4 0 r/w df2_n4 3 0 r/w df2_n3 2 0 r/w df2_n2 1 0 r/w df2_n1 0 0 r/w df2_n0 byte 12: control register 12 bit @pup type name description 7 0 r/w df3_n7 the df3_n[8:0] configures cpu frequency for dynamic frequency. doc[1:2] =11 60 r/w df3_n6 50 r/w df3_n5 40 r/w df3_n4 30 r/w df3_n3 20 r/w df3_n2 10 r/w df3_n1 00 r/w df3_n0 byte 13: control register 13 bit @pup type name description 7 0 r/w recovery_frequency this bit allows selection of the frequency setting to which the clock will be restored once the system is rebooted 0 = use hw settings 1 = recovery n[8:0] 6 0 r/w timer_sel timer_sel selects the wd re set function at sreset pin when wd times out. 0 = reset and reload recovery_frequency 1 = only reset
CY28551 document #: 001-05675 rev. *c page 11 of 30 5 1 r/w time_scale time_scale allows selection of wd time scale 0 = 294 ms, 1 = 2.34 s 4 0 r/w wd_alarm wd_alarm is set to ?1? when the watchdog times out. it is reset to ?0? when the system clears the wd_timer time stamp 3 0 r/w wd_timer2 watchdog timer time stamp selection 000: reserved (test mode) 001: 1 * time_scale 010: 2 * time_scale 011: 3 * time_scale 100: 4 * time_scale 101: 5 * time_scale 110: 6 * time_scale 111: 7 * time_scale 2 0 r/w wd_timer1 1 0 r/w wd_timer0 0 0 r/w wd_en watchdog timer enable. when the bit is asserted, watchdog timer is triggered and time stamp of wd_timer is loaded 0 = disable, 1 = enable byte 13: control register 13 bit @pup type name description byte 14: control register 14 bit @pup type name description 7 0 r/w cpu_daf_n7 if prog_cpu_en is set, the values programmed in cpu_daf_n[8:0] and cpu_daf_m[6:0] will be used to de termine the cpu output frequency. the setting of the fs_override bit determines the frequency ratio for cpu and other output clocks. when it is cleared, the same frequency ratio stated in the latched fs[e:a] register will be used. when it is set, the frequency ratio stated in the fs el[3:0] register will be used 6 0 r/w cpu_daf_n6 5 0 r/w cpu_daf_n5 4 0 r/w cpu_daf_n4 3 0 r/w cpu_daf_n3 2 0 r/w cpu_daf_n2 1 0 r/w cpu_daf_n1 0 0 r/w cpu_daf_n0 byte 15: control register 15 bit @pup type name description 7 0 r/w cpu_daf_n8 if prog_cpu_en is set, the values programmed in cpu_daf_n[8:0] and cpu_daf_m[6:0] will be used to de termine the cpu output frequency. the setting of the fs_override bit determines the frequency ratio for cpu and other output clocks. when it is cleared, the same frequency ratio stated in the latched fs[e:a] register will be used. when it is set, the frequency ratio stated in the fs el[3:0] register will be used. 6 0 r/w cpu_daf_m6 5 0 r/w cpu_daf_m5 4 0 r/w cpu_daf_m4 3 0 r/w cpu_daf_m3 2 0 r/w cpu_daf_m2 1 0 r/w cpu_daf_m1 0 0 r/w cpu_daf_m0
CY28551 document #: 001-05675 rev. *c page 12 of 30 byte 16: control register 16 bit @pup type name description 7 0 r/w pcie_daf_n7 the pcie_daf_n[8:0] configures the pcie frequency for dial-a-frequency 6 0 r/w pcie_daf_n6 5 0 r/w pcie_daf_n5 4 0 r/w pcie_daf_n4 3 0 r/w pcie_daf_n3 2 0 r/w pcie_daf_n2 1 0 r/w pcie_daf_n1 0 0 r/w pcie_daf_n0 byte 17: control register 17 bit @pup type name description 7 0 r/w recovery n7 watchdog recovery bit 6 0 r/w recovery n6 watchdog recovery bit 5 0 r/w recovery n5 watchdog recovery bit 4 0 r/w recovery n4 watchdog recovery bit 3 0 r/w recovery n3 watchdog recovery bit 2 0 r/w recovery n2 watchdog recovery bit 1 0 r/w recovery n1 watchdog recovery bit 0 0 r/w recovery n0 watchdog recovery bit byte 18: control register 18 bit @pup type name description 7 0 r/w pcie_n8 pci-e dial -a-frequency bit n8 6 0 r/w fs[d:a] fs_override 0 = select operating frequency by fs(d:a) input pins 1 = select operating frequency by fsel_(3:0) settings 5 0 r/w df_en dynamic frequency for cpu frequency enable 0 = disable, 1 = enable 4 0 r/w reset_i_en reset_i# enable 0 = disable, 1 = enable 3 0 r/w prog_pcie_en programmable src frequency enable 0 = disable, 1 = enabled 2 0 r/w prog_cpu_en programmable cpu frequency enable 0 = disable, 1 = enable 1 0 r/w watchdog autorecovery watchdog autorecovery mode 0 = disable (manual), 1= enable (auto) 0 0 r/w recovery n8 watchdog recovery bit table 5. crystal recommendations frequency (fund) cut loading load cap drive (max.) shunt cap (max.) motional (max.) tolerance (max.) stability (max.) aging (max.) 14.31818 mhz at parallel 20 pf 0.1 mw 5 pf 0.016 pf 35 ppm 30 ppm 5 ppm
CY28551 document #: 001-05675 rev. *c page 13 of 30 crystal recommendations the CY28551 requires a parallel resonance crystal. substi- tuting a series resonance crystal will cause the CY28551 to operate at the wrong frequency and violate the ppm specifi- cation. for most applications there is a 300-ppm frequency shift between series and parallel crystals due to incorrect loading. crystal loading crystal loading plays a critical role in achieving low ppm perfor- mance. to realize low ppm performance, the total capacitance the crystal will see must be considered to calculate the appro- priate capacitive loading (cl). figure 1 shows a typical crystal configuration using the two trim capacitors. an important clarification for the following discussion is that the trim capa citors are in series with the crystal not parallel. it is a common misconception that load capacitors are in parallel with the crystal and should be approximately equal to the load capacitance of the crystal. this is not true. figure 1. crystal capacitive clarification calculating load capacitors in addition to the standard external trim capacitors, trace capacitance and pin capacitance must also be considered to correctly calculate crystal load ing. as mentioned previously, the capacitance on each side of the crystal is in series with the crystal. this means the total capacitance on each side of the crystal must be twice the specified crystal load capacitance (cl). while the capacitance on each side of the crystal is in series with the crystal, trim capacitors (ce1,ce2) should be calculated to provide equal capacitive loading on both sides. figure 2. crystal loading example use the following formulas to calculate the trim capacitor values for ce1 and ce2. cl ........................................... .........crystal load capacitance cle ............. .............. .............. actual loading seen by crystal using standard value trim capacitors ce ..................................................... external trim capacitors cs ........................................ ......stray capaci tance (terraced) ci .......................................................... internal capacitance (lead frame, bond wires etc.) multifunction pin selection in the CY28551, some of the pins can provide different types of frequency, depending on the sel[1:0] hw strapping pin setting, to support different chipset vendors. the configuration is shown as follows: sel[1:0] link/dot/sa ta sata/pcie platform 00 link sata sis 01 dot sata intel w/gfx 10 link pciex via 11 sata pciex intel xtal ce2 ce1 cs1 cs2 x1 x2 ci1 ci2 clock chip trace 2.8 pf trim 33 pf pin 3 to 6p load capacitance (each side) total capacitance (as seen by the crystal) ce = 2 * cl ? (cs + ci) ce1 + cs1 + ci1 1 + ce2 + cs2 + ci2 1 ( ) 1 = cle
CY28551 document #: 001-05675 rev. *c page 14 of 30 dynamic frequency dynamic frequency ? dynamic frequency (df) is a technique used to increase cpu frequency or src frequency dynami- cally from any starting value. the user selects the starting point, either by hw, fsel, or daf, then enables df. after that, df will dynamically change as determined by df-n registers and the m value of frequency table. df pin ? there are two pins to be used on dynamic frequency (df). when used as df, these two pins will map to four df-n registers that correspond to different ?n? values for dynamic frequency. any time there is a change in df, it should load the new value. df_en bit ? this bit enables the df mode. by default, it is not set. when set, the operating frequency is determined by df[2:0] pins. default = 0, (no df) dial-a-frequency (cpu & pciex) this feature allows users to overclock their systems by slowly stepping up the cpu or src frequency. when the program- mable output frequency feature is enabled, the cpu and src frequencies are determined by the following equation: fcpu = g * n/m or fcpu = g2 * n, where g2 = g/m. ?n? and ?m? are the values programmed in programmable frequency select n-value register and m-value register, respectively. ?g? stands for the pll gear constant, which is determined by the programmed value of fs[e:a]. see ta ble 1 for the gear constant for each frequency selection. the pci express only allows user control of the n register; the m value is fixed and documented in table 1 . in this mode, the user writes the desired n and m value into the daf i2c registers. the user cannot change only the m value and must change both the m and the n values at the same time, if they require a change to the m value. the user may change only the n value if required. associated register bits cpu_daf enable ? this bit enables cpu daf mode. by default, it is not set. when set, the operating frequency is determined by the values entered into the cpu_daf_n register. note: the cpu_daf_n and m register must contain valid values before cpu_daf is set. default = 0, (no daf). cpu_daf_n ? there are nine bits (for 512 values) to linearly change the cpu frequency (limited by vco range). default = 0, (0000). the allowable values for n are detailed in the frequency select table ( table 1 ). cpu_daf_m ? there are 7 bits (for 128 values) to linearly change the cpu frequency (limited by vco range). default = 0. the allowable values for m are detailed in the frequency select table ( ta ble 1 ). src_daf enable ? this bit enables src daf mode. by default, it is not set. when set, the operating frequency is determined by the values entered into the src_daf_n register. note: the src_daf_n register must contain valid values before src_daf is se t. default = 0, (no daf). src_daf_n ? there are nine bits (for 512 values) to linearly change the cpu frequency (limited by vco range). default = 0, (0000). the allowable values for n are detailed in the frequency select table ( table 1 ). recovery ? the recovery mechanism during cpu daf, when the system locks up and the watchdog timer is enabled, is determined by the ?watchdog recovery mode? and ?watchdog autorecovery enable? bits. the possible recovery methods are: (a) auto, (b) manual (by recovery n), (c) hw, and (d) no recovery, ju st send reset signal. there is no recovery mode for src dial-a-frequency. software frequency select this mode allows the user to select the cpu output frequencies using the software frequency select bits in the smbus register. fsel ? there are four bits (for 16 combinations) to select predetermined cpu frequencies from a table. the table selec- tions are detailed in table 1 . fs_override ? this bit allows the cpu frequency to be selected from hw or fsel settings. by default, this bit is not set and the cpu frequency is selected by hw. when this bit is set, the cpu frequency is sele cted by the fsel bits. default = 0. recovery ? the recovery mech anism during fsel when the system locks up is determined by the ?watchdog recovery mode? and ?watchdog autorecovery enable? bits. the only possible recovery method is to use hardware settings. auto recovery or manual recovery can cause a wrong output frequency because the output divider may have changed with the selected cpu frequency and these recovery methods will not recover the original output divider setting. smooth switching the device contains one smooth switch circuit, which is shared by the cpu pll and src pll. the smooth switch circuit ensures that when the ou tput frequency changes by overclocking, the transition from the old frequency to the new frequency is a slow, smooth transi tion containing no glitches. the rate of change of output frequency when using the smooth switch circuit is le ss than 1 mhz/0.667 s. the frequency overshoot and undershoot will be less than 2%. the smooth switch circuit can be assigned auto or manual mode. in auto mode, the clo ck generator will assign smooth switch automatically when the pll will perform overclocking. for manual mode, the smooth switch circuit can be assigned to either pll via smbus. by default the smooth switch circuit is set to auto mode. either p ll can still be overclocked when it does not have control of the smooth switch circuit, but it is not guaranteed to transition to the new frequency without large frequency glitches. doc[2:1] doc n register 00 original frequency 01 df1_n 10 df2_n 11 df3_n
CY28551 document #: 001-05675 rev. *c page 15 of 30 do not enable overclocking and change the n values of both plls in the same smbus block write and use smooth switch mechanism on spread spectrum on/off. watchdog timer the watchdog timer is used in the system in conjunction with overclocking. it is used to prov ide a reset to a system that has hung up due to overclocking the cpu and the front side bus. the watchdog is enabled by the user and if the system completes its checkpoints, the system will clear the timer. however, when the timer runs out, there will be a reset pulse generated on the sreset# pin for 20 ms that is used to reset the system. when the watchdog is enabled (wd_en = 1) the watchdog timer will start counting down from a value of watchdog_timer * time scale. if the watchdog timer reaches 0 before the wd_en bit is cleared then it will assert the sreset# signal and set the watchdog alarm bit to ?1?. to use the watchdog, the sr eset# pin must be enabled by sampling sreset_en pin low by vttpwrgd# assertion during system boot up. if at any point during the watchdog timer countdown the time stamp or watchdog timer bits are changed, the timer will reset and start counting down from the new value. after the reset pulse, the wa tchdog will stay inactive until either: 1. a new time stamp or watchdog timer value is loaded. 2. the wd_en bit is cleared and then set again. watchdog register bits the following register bits are associated with the watchdog timer: watchdog enable ? this bit (by default) is not set, which disables the watchdog. when set, the watchdog is enabled. also, when there is a transition from low to high, the timer reloads. default = 0, disable watchdog timer ? there are three bits (for seven combina- tions) to select the timer value. default = 000, the value '000' is a reserved test mode. watchdog alarm ? this bit is a flag and when it is set, it indicates that the timer has expired. this bit is not set by default. when the bit is set, the user is allowed to clear. default = 0. watchdog time scale ? this bit selects the multiplier. when this bit is not set, the multiplier will be 250 ms. when set (by default), the multiplier will be 3s. default = 1 watchdog reset mode ? this selects the watchdog reset mode. when this bit is not set (by default), the watchdog will send a reset pulse and reload the recovery frequency depending on the watchdog recovery mode setting. when set, it sends a reset pulse. default = 0, reset & recover frequency. watchdog recovery mode ? this bit selects the location to recover from. one option is to recover from the hw settings (already stored in smbus registers for readback capability) and the second is to recover from a register called ?recovery n?. default = 0 (recover from the hw setting) watchdog autorecovery enable ? this bit is set by default and the recovered values are automatically written into the ?watchdog recovery register? and reloaded by the watchdog function. when this bit is not set, the user is allowed to write to the ?watchdog recovery register?. the value stored in the ?watchdog recovery register? will be used for recovery. default = 1, autorecovery. watchdog recovery register ? this is a nine-bit register to store the watchdog n recovery value. this value can be written by the autorecovery or user depending on the state of the ?watchdog autorecovery enable bit?. watchdog recovery modes there are three operating modes that require watchdog recovery. the modes are dial-a-frequency (daf), dynamic clocking (df), or frequency select. there are four different recovery modes; the following sections list the operating mode and the recovery mode associated with it. recover to hardware m, n, o when this recovery mode is selected, in the event of a watchdog timeout, the original m, n, and o values that were latched by the hw fsel pins at chip boot-up will be reloaded. autorecovery when this recovery mode is selected, in the event of a watchdog timeout, the m and n values stored in the recovery m and n registers will be reloaded. the current values of m and n will be latched into the internal recovery m and n registers by the wd_en bit being set. manual recovery when this recovery mode is selected, in the event of a watchdog timeout, the n value as programmed by the user in the n recovery register, and the m value that is stored in the recovery m register (not accessible by the user), will be restored. the current m value will be latched to m recovery register by the wd_en bit being set. no recovery if no recovery mode is selected, in the event of a watchdog time out, the devic e will assert the sr eset# and keep the current values of m and n software reset software reset is a reset function that is used to send out a pulse from the sreset# pin. it is controlled by the sw_reset enable register bit. upon completion of the byte/word/block write in which the sw_reset bit was set, the device will send a reset pulse on the sreset# pin. the duration of the sreset# pulse will be the same as the duration of the sreset# pulse after a watchdog timer time out. after the sreset# pulse is asserted the sw_reset bit will be automatically cleared by the device. pd clarification the vtt_pwrgd#/pd pin is a dual -function pin. during initial power up, the pin functions as vtt_pwrgd#. once
CY28551 document #: 001-05675 rev. *c page 16 of 30 vtt_pwrgd# has been sampled low by the clock chip, the pin assumes pd functionality. the pd pin is an asynchronous active high input used to shut off all clocks cleanly prior to shutting off power to the device. this signal must be synchro- nized internal to the device prior to powering down the clock synthesizer. pd is also an asynchronous input for powering up the system. when pd is asse rted high, all clocks must be driven to a low value and held prior to turning off the vcos and the crystal oscillator pd assertion when pd is sampled high by two consecutive rising edges of cpuc, all single-ended outputs must be held low on their next high-to-low transition and differential clocks must be held high or tri-stated (dependi ng on the state of the control register drive mode bit) on the next ?diff clock#? high-to-low transition within 4 clock periods. when the smbus pd drive mode bit corresponding to the differential (cpu, src, and dot) clock output of interest is programmed to '0', the clock output must be held with ?diff clo ck? pin driven high at 2 x iref, and ?diff clock#? tri-state. if the control register pd drive mode bit corresponding to the output of interest is programmed to ?1?, then both the ?diff clock? and the ?diff clock#? are tri-state. note figure 3 shows cput = 133 mhz and pd drive mode = '1' for all differential outputs. this diagram and description is appli- cable to valid cpu frequencies 100, 133, 166, and 200 mhz. in the event that pd mode is desired as the initial power-on state, pd must be asserted high in less than 10 s after asserting vtt_pwrgd#. pd deassertion the power-up latency must be less than 1.8 ms. this is the time from the deassertion of th e pd pin or the ramping of the power supply until the time that stable clocks are output from the clock chip. all differential outputs stopped in a tri-state condition resulting from power down must be driven high in less than 300 s of pd deassertion to a voltage greater than 200 mv. after the clock chip's internal pll is powered up and locked, all outputs are to be e nabled within a few clock cycles of each other. figure 4 is an example showing the relationship of clocks coming up. unfortunat ely, we can not show all possible combinations; designers need to ensure that from the first active clock output to the last takes no more than two full pci clock cycles. figure 3. pd assertion timing waveform do t96c pd cpuc , 133 m hz cput, 133 m hz srcc 100 m hz usb, 48 mhz dot96t srct 100 m hz pci, 33 mhz ref link
CY28551 document #: 001-05675 rev. *c page 17 of 30 figure 4. pd deassertion timing waveform cpu_stp# clarification the cpu_stp# signal is an active low input used for cleanly stopping and starting the cpu outputs while the rest of the clock generator continues to function. note that the assertion and deassertion of this signal is absolutely asynchronous. cpu_stp# assertion the cpu_stp# signal is an active low input used for synchronous stopping and starting of the cpu output clocks while the rest of the clock generator continues to function. when the cpu_stp# pin is asserted, all cpu outputs that are set with the smbus configuration to be stoppable via assertion of cpu_stp# will be stopped after being sampled by 2 to 6 rising edges of the internal cpuc clock. the final state of the stopped cpu clock is low due to tri-state; both cput and cpuc outputs will not be driven. figure 5. cpu_stp# assertion timing waveform figure 6. cpu_stp# deassertion cpu_stp# deassertion the deassertion of the cpu_stp# signal will cause all cpu outputs that were stopped to resume normal operation in a synchronous manner, synchronous manner meaning that no dot96c pd cpuc, 133 mhz cput, 133 mhz srcc 100 mhz usb, 48 mhz dot96t srct 100 mhz tstable <1.8 m s p c i, 3 3 m h z ref tdrive_pwrdn# <300 s > 200 m v link cpu_stp# cput cpuc cpu_stp# cput cpuc cput internal tdrive_cpu_stp#,10ns>200mv cpuc internal
CY28551 document #: 001-05675 rev. *c page 18 of 30 short or stretched clock pulses will be produced when the clock resumes. the maximum la tency from the deassertion to active outputs is between 2 and 6 cpu clock periods (2 clocks are shown). if the control regist er tri-state bit corresponding to the output of interest is progr ammed to '1', then the stopped cpu outputs will be driven hi gh within 10 ns of cpu_stop# deassertion to a voltage greater than 200 mv. pci_stp# clarification the pci_stp# signal is an active low input used for cleanly stopping and starting the pci and pciex outputs while the rest of the clock generator continues to function. the pcif and pciex clocks are special in that they can be programmed to ignore pci_stp# by setting the register bit corresponding to the output of interest to free running. outputs set to free running will ignore the pci_stp# pin. pci_stp# assertion the impact of asserting the pci_stp# signal is as follows. the clock chip is to sample the pc i_stp# signal on a rising edge of pcif clock. after detectin g the pci_stp# assertion low, all pci and stoppable pcif clocks will latch low on their next high-to-low transi tion. after the pci clocks are latched low, the stoppable pciex cl ocks will latch to low due to tri-state, as shown in figure 7 . the one pci clock latency shown is critical to system func tionality; any violation of this may result in system failure. the tsu_pci_stp# is the setup time required by the clock generator to correctly sample the pci_stp# assertion. this time is 10 ns minimum. pci_stp# deassertion the deassertion of the pci_stp# signal functions as follows. the deassertion of the pci_stp# signal is to be sampled on the rising edge of the pcif free running clock domain. after detecting pci_stp# deassertion, all pci, stoppable pcif and stoppable pciex clocks will resume in a glitch-free manner. the pci and pciex clock resume latency should exactly match the 1 pci clock latency required for pci_stp# entry. the stoppable pciex clocks must be driven high within 15 ns of pci_stp# deassertion. figure 8 shows the appro- priate relationship. the tsu_cpu_stp# is the setup time required by the clock generator to correctly sample the pci_stp# deassertion. this time is 10 ns minimum. figure 7. pci_ stp# assertion figure 8. pci_stp# deassertion clkreq# clarification the clkreq# signals are active low inputs used to cleanly stop and start selected src outputs. the outputs controlled by clkreq# are determined by th e settings in register bytes 10 and 11. the clkreq# signal is a debounced signal in that its state must remain unchanged during two consecutive rising edges of difc to be recognized as a valid assertion or deassertion. (the assertion and deassertion of this signal is absolutely asynchronous.) tsu_pci_stp# > 10ns pci_stp# pci_f pci pciex 100mhz pci_stp# pci_f pci pciex 100mhz tdrive_pciex <15 ns
CY28551 document #: 001-05675 rev. *c page 19 of 30 clkreq# assertion all differential outputs that were stopped will resume normal operation in a glitch-free manner. the maximum latency from the deassertion to active outputs is between 2 and 6 pciex clock periods (2 clocks are show n) with all cl kreq# outputs resuming simultaneously. if the clkreq# drive mode is tri-state, all stopped pciex outputs must be driven high within 10 ns of clkreq# deasse rtion to a voltage greater than 200 mv. clkreq# deassertion the impact of asserting the cl kreq# pins is that all dif outputs that are set in the cont rol registers to stoppable via assertion of clkreq# are to be stopped after their next transition. when the control r egister clkreq# drive mode bit is programmed to '0', the final state of all stopped pciex signals is pciext clock = high and pciexc = low. there will be no change to the output drive current values. srct will be driven high with a current value equal 6 x iref. when the control register clkreq# drive mode bit is programmed to '1', the final state of all stopped dif signals is low; both pciext clock and pciexc clock outputs will not be driven. figure 9. clkreq# deassertion figure 10. vtt_pwrgd# timing diagram pciexc(stoppable) pciext(stoppable) pciexc(free running) pciext(free running) pe_req# tdrive_pe_req# < 10 ns vtt_pwrgd# = low delay > 0.25 ms s1 power off s0 vdd_a = 2.0v sample inputs straps s2 normal operation wait for <1.8 ms enable outputs s3 vtt_pwrgd# = toggle vdd_a = off
CY28551 document #: 001-05675 rev. *c page 20 of 30 figure 11. vtt_pwrgd# timing diagram fs_[d:a] vtt_pwrgd# pwrgd_vrm vdd clock gen clock state clock outputs clock vco 0.2-0.3 ms delay state 0 state 2 state 3 wait for vtt_pwrgd# sample sels off off on on state 1 device is not affected, vtt_pwrgd# is ignored
CY28551 document #: 001-05675 rev. *c page 21 of 30 absolute maximum conditions parameter description condition min. max. unit v dd core supply voltage ?0.5 4.6 v v dd_a analog supply voltage ?0.5 4.6 v v in input voltage relative to v ss ?0.5 v dd + 0.5 vdc t s temperature, storage non-functional ?65 150 c t a temperature, operating ambient functional 0 70 c t j temperature, junction functional ? 150 c ? jc dissipation, junction to case mil-std-883e method 1012.1 ? 20 c/w ? ja dissipation, junction to ambient jedec (jesd 51) ? 60 c/w esd hbm esd protection (human body model) mil-std-883, method 3015 2000 ? v ul-94 flammability rating at 1/8 in. v?0 msl moisture sensitivity level 1 multiple supplies : the voltage on any input or i/o pin cannot exceed the power pin during power-up. power supply sequencing is not required. dc electrical specifications parameter description condition min. max. unit all v dd s 3.3v operating voltage 3.3 5% 3.135 3.465 v v ili2c input low voltage sdata, sclk ? 1.0 v v ihi2c input high voltage sdata, sclk 2.2 ? v v il_fs fs_[a:d] input low voltage v ss ? 0.3 0.35 v v ih_fs fs_[a:d] input high voltage 0.7 v dd + 0.5 v v il 3.3v input low voltage v ss ? 0.3 0.8 v v ih 3.3v input high voltage 2.0 v dd + 0.3 v i il input low leakage current except internal pull-up resistors, 0 < v in < v dd ?5 ? a i ih input high leakage current except internal pull-down resistors, 0 < v in < v dd ?5 a v ol 3.3v output low voltage i ol = 1 ma ? 0.4 v v oh 3.3v output high voltage i oh = ?1 ma 2.4 ? v i oz high-impedance output current ?10 10 a c in input pin capacitance 3 5 pf c out output pin capacitance 3 5 pf l in pin inductance ? 7 nh v xih xin high voltage 0.7v dd v dd v v xil xin low voltage 0 0.3v dd v i dd3.3v dynamic supply current at max. load and freq. per figure 14 ?500ma i pt3.3v power-down supply current pd asserted, outputs tri-state ? 12 ma
CY28551 document #: 001-05675 rev. *c page 22 of 30 ac electrical specifications parameter description condition min. max. unit crystal t dc xin duty cycle the device will operate reliably with input duty cycles up to 30/70 but the ref clock duty cycle will not be within specification 47.5 52.5 % t period xin period when xin is dr iven from an external clock source 69.841 71.0 ns t r /t f xin rise and fall times measured between 0.3v dd and 0.7v dd ?10.0ns t ccj xin cycle to cycle jitter as an average over 1- s duration ? 500 ps l acc long-term accuracy over 150 ms ? 30 ppm cpu at 0.7v (ssc refers to ?0.5% spread spectrum) t dc cput and cpuc duty cycle measured at crossing point v ox @ 0.1s 45 55 % t period 100 mhz cput and cpuc period measured at crossing point v ox @ 0.1s 9.99900 10.0100 ns t period 133 mhz cput and cpuc period measured at crossing point v ox @ 0.1s 7.49925 7.50075 ns t period 166 mhz cput and cpuc period measured at crossing point v ox @ 0.1s 5.99940 6.00060 ns t period 200 mhz cput and cpuc period measured at crossing point v ox @ 0.1s 4.99950 5.00050 ns t period 266 mhz cput and cpuc period measured at crossing point v ox @ 0.1s 3.74963 3.75038 ns t period 333 mhz cput and cpuc period measured at crossing point v ox @ 0.1s 2.99970 3.00030 ns t period 400 mhz cput and cpuc period measured at crossing point v ox @ 0.1s 2.49975 2.50025 ns t periodss 100 mhz cput and cpuc period, ssc measured at crossing point v ox @ 0.1s 9.99900 10.0100 ns t periodss 133 mhz cput and cpuc period, ssc measured at crossing point v ox @ 0.1s 7.49925 7.50075 ns t periodss 166 mhz cput and cpuc period, ssc measured at crossing point v ox @ 0.1s 5.99940 6.00060 ns t periodss 200 mhz cput and cpuc period, ssc measured at crossing point v ox @ 0.1s 4.99950 5.00050 ns t periodss 266 mhz cput and cpuc period, ssc measured at crossing point v ox @ 0.1s 3.74963 3.75038 ns t periodss 333 mhz cput and cpuc period, ssc measured at crossing point v ox @ 0.1s 2.99970 3.00030 ns t periodss 400 mhz cput and cpuc period, ssc measured at crossing point v ox @ 0.1s 2.49975 2.50025 ns t periodabs 100 mhz cput and cpuc absolute period measured at crossing point v ox @ 1 clock 9.91400 10.0860 ns t periodabs 133 mhz cput and cpuc absolute period measured at crossing point v ox @ 1 clock 7.41425 7.58575 ns t periodabs 166 mhz cput and cpuc absolute period measured at crossing point v ox @ 1 clock 5.91440 6.08560 ns t periodabs 200 mhz cput and cpuc absolute period measured at crossing point v ox @ 1 clock 4.91450 5.08550 ns t periodabs 266 mhz cput and cpuc absolute period measured at crossing point v ox @ 1 clock 3.66463 3.83538 ns t periodabs 333 mhz cput and cpuc absolute period measured at crossing point v ox @ 1 clock 2.91470 3.08530 ns t periodabs 400 mhz cput and cpuc absolute period measured at crossing point v ox @ 1 clock 2.41475 2.58525 ns t periodssabs 100- mhz cput and cpuc absolute period, ssc measured at crossing point v ox @ 1 clock 9.91400 10.1363 ns t periodssabs 133 mhz cput and cpuc absolute period, ssc measured at crossing point v ox @ 1 clock 7.41425 7.62345 ns t periodssabs 166 mhz cput and cpuc absolute period, ssc measured at crossing point v ox @ 1 clock 5.91440 6.11576 ns
CY28551 document #: 001-05675 rev. *c page 23 of 30 t periodssabs 200 mhz cput and cpuc absolute period, ssc measured at crossing point v ox @ 1 clock 4.91450 5.11063 ns t periodssabs 266 mhz cput and cpuc absolute period, ssc measured at crossing point v ox @ 1 clock 3.66463 3.85422 ns t periodssabs 333 mhz cput and cpuc absolute period, ssc measured at crossing point v ox @ 1 clock 2.91470 3.10038 ns t periodssabs 400 mhz cput and cpuc absolute period, ssc measured at crossing point v ox @ 1 clock 2.41475 2.59782 ns t skew cpu0 to cpu1 measured at crossing point v ox ? 100 ps t ccj cput/c cycle to cycle measured at crossing point v ox ?85ps l acc long term accuracy measured at crossing point v ox ? 100 ppm t r /t f cput and cpuc rise and fall times measured differentially from 150 mv 2.5 8 v/ns t rfm rise/fall matching measured single-endedly from 75 mv ? 20 % v _max max output voltage math averages figure 14 ?1.15v v _min min output voltage math averages figure 14 ?0.3 ? v v ox crossing point voltage at 0.7v swing 300 550 mv cpu at 3.3v (ssc refers to ?0.5% spread spectrum) t r output rise edge rate measured @ k8 test load using vocm 400 mv, 0.850v to 1.650v 27v/ns t f output fall edge rate measured @ k8test load using vocm 400 mv, 1.650v to 0.850v 27v/ns v diff differential voltage measured @ k8 test load (single-ended) 0.4 2.3 v ? diff change in vdiff_dc magnitude measured @ k8 test load (single-ended) ?150 150 mv v cm common mode voltage measured @ k8 test load (single-ended) 1.05 1.45 v ? v cm change in vcm measured @ k8 test load (single-ended) ?200 200 mv t dc duty cycle measured at v ox 45 53 % t cyc jitter, cycle to cy cle measured at v ox 0 200 ps t accum jitter, accumulated measured at v ox ?1000 1000 ps pciex t dc pciext and pciexc duty cycle measured at crossing point v ox 45 55 % t period 100 mhz pciext and pciexc period measured at crossing point v ox @ 0.1s 9.99900 10.0010 ns t periodss 100 mhz pciext and pciexc period, ssc measured at crossing point v ox @ 0.1s 9.99900 10.0010 ns t periodabs 100 mhz pciext and pciexc absolute period measured at crossing point v ox @ 1 clock 9.87400 10.1260 ns t periodssabs 100 mhz pciext and pciexc absolute period, ssc measured at crossing point v ox @ 1 clock 9.87400 10.1763 ns t skew any pciext/c to pciext/c clock skew measured at crossing point v ox ? 250 ps t ccj pciext/c cycle to cycle jitter measured at crossing point v ox ? 125 ps l acc pciext/c long term accuracy measured at crossing point v ox ? 100 ppm t r /t f pciext and pciexc rise and fall times measured differentially from 150 mv 2.5 8 v/ns t rfm rise/fall matching measured single-endedly from 75 mv ? 20 % v _max max output voltage math averages, see figure 14 ?1.15v v _min min output voltage math averages, see figure 14 ?0.3 ? v ac electrical specifications (continued) parameter description condition min. max. unit
CY28551 document #: 001-05675 rev. *c page 24 of 30 v ox crossing point voltage at 0.7v swing 300 550 mv dot t dc dot96t and dot96c duty cycle measured at crossing point v ox 45 55 % t period dot96t and dot96c period measured at crossing point v ox @ 0.1s 10.4156 10.4177 ns t periodabs dot96t and dot96c absolute period measured at crossing point v ox @ 0.1s 10.1656 10.6677 ns t ccj dot96t/c cycle to cycle jitter measured at crossing point v ox ? 250 ps l acc dot96t/c long term accuracy measured at crossing point v ox ? 300 ppm t ltj long term jitter measurement taken from cross point v ox @ 10 s ? 700 ps t r /t f dot96t and dot96c rise and fall time measured differentially from 150 mv 2.5 8 v/ns t rfm rise/fall matching measured single-endedly from 75 mv ? 20 % v _max max output voltage math averages , see figure 14 ?1.15v v _min min output voltage math averages, see figure 14 ?0.3 ? v v ox crossing point voltage at 0.7v swing 300 550 mv link - 133 mhz t dc link (133 mhz) duty cycle measurement at 1.5v 45 55 % t period spread disabled link (133 mhz) pe riod measurement at 1.5v 7.50 7.56 ns t periodss spread enabled link (133 mhz) period, ssc measurement at 1.5v 7.50 7.56 ns t r /t f link (133 mhz) rising and falling edge rate measured between 0.8v and 2.0v 1.0 4.0 v/ns t ccj link (133 mhz) cycle-to-cycle ji tter measurement at 1.5v ? 250 ps t skew any link clock skew measurement at 1.5v ? 175 ps l acc link (133 mhz) long term accuracy measured at crossing point v ox ? 300 ppm link - 66 mhz t dc link (66 mhz) duty cycle measurement at 1.5v 45 55 % t period spread disabled link (66 mhz) peri od measurement at 1.5v 14.9955 15.0045 ns t periodss spread enabled link (66 mhz) period, ssc measurement at 1.5v 14.9955 15.0045 ns t r /t f link (66 mhz) rising and falling edge rate measured between 0.8v and 2.0v 1.0 4.0 v/ns t ccj link (66 mhz) cycle-to-cycle ji tter measurement at 1.5v ? 250 ps t skew any link clock skew measurement at 1.5v ? 175 ps l acc link (66 mhz) long term accuracy measured at crossing point v ox ? 300 ppm pci t dc pci duty cycle measurement at 1.5v 45 55 % t period spread disabled pcif/pci period meas urement at 1.5v 29.99100 30.00900 ns t periodss spread enabled pcif/pci period, ssc measurement at 1. 5v 29.9910 30.15980 ns t periodabs spread disabled pcif/pci period meas urement at 1.5v 29.49100 30.50900 ns t periodssabs spread enabled pcif/pci period, ssc measurement at 1.5v 29.49100 30.65980 ns t high pcif and pci high time measurement at 2.4v 12.0 ? ns t low pcif and pci low time measurement at 0.4v 12.0 ? ns t r /t f pcif and pci rising and falling edge rate measured between 0.8v and 2.0v 1.0 4.0 v/ns ac electrical specifications (continued) parameter description condition min. max. unit
CY28551 document #: 001-05675 rev. *c page 25 of 30 t skew any pci clock to any pci clock skew measurement at 1.5v ? 250 ps t ccj pcif and pci cycle to cycle jitter measurement at 1.5v ? 500 ps usb t dc duty cycle measurement at 1.5v in high drive mode 45 55 % t period period measurement at 1.5v 20.83125 20.83542 ns t periodabs absolute period measurement at 1.5v 20.48125 21.18542 ns t high usb high time measurement at 2.4v 8.094 10.5 ns t low usb low time measurement at 0.4v 7.694 10.5 ns t r /t f rising and falling edge rate measured between 0.8v and 2.0v 1.0 2.0 v/ns t ccj cycle-to-cycle jitter measu rement at 1.5v ? 350 ps l acc 48m long term accuracy measured at crossing point v ox ? 100 ppm t ltj long term jitter measurement taken from cross point v ox @ 1 s ?1.0ns 24m t dc duty cycle measurement at 1.5v in high drive mode 45 55 % t period period measurement at 1.5v 41.6646 41.6688 ns t high usb high time measurement at 2.4v 18.8323 18.8323 ns t low usb low time measurement at 0.4v 18.8323 18.8323 ns t r /t f rising and falling edge rate measured between 0.8v and 2.0v 1.0 4.0 v/ns t ccj cycle-to-cycle jitter measu rement at 1.5v ? 350 ps l acc 48m long term accuracy measured at crossing point v ox ? 100 ppm t ltj long term jitter measurement taken from cross point v ox @ 1 s ?1.0ns ref t dc ref duty cycle measurement at 1.5v 45 55 % t period ref period measurement at 1.5v 69.8203 69.8622 ns t periodabs ref absolute period measurement at 1.5v 68.82033 70.86224 ns t r /t f ref rise and fall times edge rate meas ured between 0.8v and 2.0v 1.0 4.0 v/ns t ccj ref cycle-to-cycle jitter mea surement at 1.5v ? 1000 ps t skew ref clock to ref clock measurement at 1.5v ? 500 ps l acc long term accuracy measurement at 1.5v ? 300 ppm enable/disable and set-up t stable clock stabilization from power-up ? 1.8 ms ac electrical specifications (continued) parameter description condition min. max. unit
CY28551 document #: 001-05675 rev. *c page 26 of 30 test and measurement set-up for pci/usb and 24m single-ended signals and reference figure 12 and figure 13 show the test load configurations for the single-ended pci, usb, 24m, and ref output signals figure 12. single-ended load configuration. figure 13. single-ended load configuration high drive option pci/usb ref 22 ? measurement point 5 pf 50 ? 12 ? measurement point 5 pf 12 ? measurement point 5 pf 50 ? 50 ? pci/usb ref 12 ? measurement point 5 pf 12 ? measurement point 5 pf 12 ? measurement point 5 pf 12 ? measurement point 5 pf 12 ? measurement point 5 pf 50 ? 50 ? 50 ? 50 ? 50 ?
CY28551 document #: 001-05675 rev. *c page 27 of 30 the following diagrams show the test load configuration for the differential cpu and pciex outputs. figure 14. differential load configuration for 0.7v push pull clock figure 15. differential load configuration for 0.7 push pull clock figure 16. differential measuremen t for differential output signal s (for ac parameters measurement) 22 ? measurement point 2 pf 50 ? 22 ? measurement point 2 pf 50 ? l 1 l 2 l 1 l 2 l1 = 0.5", l2 = 7" out+ out- cput_k8 t pcb cpuc_k8 15 o hm 15 o hm 3900 pf l1 l2 l3 l1 l2 l3 t pcb 3900 pf 169 o hm m easurem ent point m easurem ent point 5pf 5pf 125 o hm 1.25v 1.25v 125 o hm
CY28551 document #: 001-05675 rev. *c page 28 of 30 ? cypress semiconductor corporation, 2006. 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. figure 17. single-ended measurement for differentia l output signals (for ac parameters measurement) figure 18. single-ended output signals (for ac parameters measurement) ordering information part number package type product flow lead-free CY28551lfxc 64-pin qfn commercial, 0 to 85 c CY28551lfxct 64-pin qfn ? tape and reel commercial, 0 to 85 c
CY28551 document #: 001-05675 rev. *c page 29 of 30 package diagram figure 19. 64-lead qfn 9 x 9 mm (punch version) lf64a purchase of i 2 c components from cypress or one of its sublicensed a ssociated companies conveys a license under the philips i 2 c patent rights to use these components in an i 2 c system, provided that the system conforms to the i 2 c standard specification as defined by philips. intel and pentium are registered trademar ks of intel corporation. sis is a registered trademark of silic on integrated systems. dial-a-frequency is a registered trademark of cypr ess semiconductor corporation. all product and company names mentioned in this document are the trademarks of their respective holders. 2 1 pin1 id 0.80 dia. 3 2 c plane seating n a 0.20[0.008] ref. 0.05[0.002] max. 0.80[0.031] max. 1.00[0.039] max. c 0.08[0.003] 1 n 0.20[0.008] r. 9.10[0.358] 8.90[0.350] 8.80[0.346] 8.70[0.342] 8.80[0.346] 8.70[0.342] 9.10[0.358] 8.90[0.350] 0-12 0.28[0.011] 0.18[0.007] 0.30[0.012] 0.50[0.020] 0.50[0.020] 7.55[0.297] 7.45[0.293] 0.45[0.018] 0.60[0.024] 0.24[0.009] by device type) (pad size vary e-pad top view side view bottom view dimensions in mm[inches] min. max. reference jedec mo-220 weight: 0.2 grams 51-85215-**
CY28551 document #: 001-05675 rev. *c page 30 of 30 document history page document title: CY28551 universal clock generator for intel, via, and sis ? document number: 001-05675 rev. ecn no. issue date orig. of change description of change ** 409135 see ecn hgs new data sheet *a 417501 see ecn hgs register alignment changed: -byte 9 contains df3_n8,df2_n8,df1_n8 -byte 10 contains df1_n<7:0> -byte 11 contains df2_n<7:0> -byte 12 contains df3_n<7:0> add powergood status bit at byte6 [0] add cpu_stp#, pci_stp# and clkreq# description *b 460105 see ecn rgl minor change ? to post on web *c 491596 see ecn hgs 1. change 48m/24_48m driving strength to high by default 2. change revision id (byte7[7:4]) to 0010

▲Up To Search▲

Price & Availability of CY28551

	To Download CY28551 Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .