View AD9557_4619553.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

dual input multiservice line card adaptive clock translator data sheet AD9557 rev. a information furnished by analog devices is believed to be accurate and reliable. however, no responsibility is assumed by analog devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. specifications subject to change without notice. no license is granted by implication or otherwise under any patent or patent rights of analog devices. trademarks and registered trademarks are the property of their respective owners. one technology way, p.o. box 9106, norwood, ma 02062-9106, u.s.a. tel: 781.329.4700 www.analog.com fax: 781.461.3113 ?2011C2012 analog devices, inc. all rights reserved. features supports gr-1244 stratum 3 stability in holdover mode supports smooth reference switchover with virtually no disturbance on output phase supports telcordia gr-253 jitter generation, transfer, and tolerance for sonet/sdh up to oc-192 systems supports itu-t g.8262 synchronous ethernet slave clocks supports itu-t g.823, g.824, g.825, and g.8261 auto/manual holdover and reference switchover 2 reference inputs (single-ended or differential) input reference frequencies: 2 khz to 1250 mhz reference validation and frequency monitoring (1 ppm) programmable input reference switchover priority 20-bit programmable input reference divider 2 pairs of clock output pins, with each pair configurable as a single differential lvds/hstl output or as 2 single-ended cmos outputs output frequencies: 360 khz to 1250 mhz programmable 17-bit integer and 24-bit fractional feedback divider in digital pll programmable digital loop filter covering loop bandwidths from 0.1 hz to 5 khz (2 khz maximum for <0.1 db of peaking) low noise system clock multiplier frame sync support adaptive clocking optional crystal resonator for system clock input on-chip eeprom to store multiple power-up profiles pin program function for easy frequency translation configuration software controlled power-down 40-lead, 6 mm 6 mm, lfcsp package applications network synchronization, including synchronous ethernet and sdh to otn mapping/demapping cleanup of reference clock jitter sonet/sdh clocks up to oc-192, including fec stratum 3 holdover, jitter cleanup, and phase transient control wireless base station controllers cable infrastructure data communications general description the AD9557 is a low loop bandwidth clock multiplier that provides jitter cleanup and synchronization for many systems, including synchronous optical networks (sonet/sdh). the AD9557 generates an output clock synchronized to up to four external input references. the digital pll allows for reduction of input time jitter or phase noise associated with the external references. the digitally controlled loop and holdover circuitry of the AD9557 continuously generates a low jitter output clock even when all reference inputs have failed. the AD9557 operates over an industrial temperature range of ?40c to +85c. if more inputs/outputs are needed, refer to the ad9558 for the four-input/six-output version of the same part. functional block diagram reference input and monitor mux status and control pins serial interface (spi or i 2 c) eeprom digital pll clock multiplier stable source AD9557 analog pll channel 0 divider channel 1 divider 3 to 11 hf divider 0 3 to 11 hf divider 1 09197-001 figure 1.
AD9557 data sheet rev. a | page 2 of 92 table of contents features .............................................................................................. 1 ? applications....................................................................................... 1 ? general description ......................................................................... 1 ? functional block diagram .............................................................. 1 ? revision history ............................................................................... 3 ? specifications..................................................................................... 4 ? supply voltage............................................................................... 4 ? supply current.............................................................................. 4 ? power dissipation......................................................................... 5 ? logic inputs ( reset , sync , pincontrol, m3 to m0 ) .... 5 ? logic outputs (m3 to m0, irq) ................................................ 6 ? system clock inputs (xoa, xob) ............................................. 6 ? reference inputs ........................................................................... 7 ? reference mo nitors ...................................................................... 8 ? reference switchover specifications.......................................... 8 ? distribution clock outputs ........................................................ 9 ? time duration of digital functions ........................................ 10 ? digital pll .................................................................................. 11 ? digital pll lock detection ...................................................... 11 ? holdover specifications............................................................. 11 ? serial port specificationsspi mode...................................... 12 ? serial port specificationsi 2 c mode ...................................... 13 ? jitter generation ......................................................................... 13 ? absolute maximum ratings.......................................................... 16 ? esd caution................................................................................ 16 ? pin configuration and function descriptions........................... 17 ? typical performance characteristics ........................................... 19 ? input/output termination recommendations .......................... 24 ? getting started ................................................................................ 25 ? chip power monitor and startup............................................. 25 ? multifunction pins at reset/power-up ................................... 25 ? device register programming using a register setup file.. 25 ? register programming overview............................................. 25 ? theory of operation ...................................................................... 28 ? overview...................................................................................... 28 ? reference clock inputs.............................................................. 29 ? reference monitors .................................................................... 29 ? reference profiles ....................................................................... 29 ? reference switchover ................................................................. 29 ? digital pll (dpll) core .......................................................... 30 ? loop control state machine..................................................... 32 ? system clock (sysclk)................................................................ 33 ? system clock inputs................................................................... 33 ? sys tem clock multiplier ......... . .................................................. 33 ? output pll (apll) ....................................................................... 35 ? clock distribution.......................................................................... 36 ? clock dividers ............................................................................ 36 ? output power-down ................................................................. 36 ? output enable............................................................................. 36 ? output mode .............................................................................. 36 ? clock distribution synchronization........................................ 36 ? status and control.......................................................................... 37 ? multifunction pins (m3 to m0) ............................................... 37 ? irq pin ........................................................................................ 37 ? watchdog timer......................................................................... 38 ? eeprom ..................................................................................... 38 ? serial control port ......................................................................... 44 ? spi/i2c port selection................................................................ 44 ? spi serial port operation.......................................................... 44 ? i 2 c serial port operation .......................................................... 48 ? programming the i/o registers ................................................... 51 ? buffered/active registers.......................................................... 51 ? autoclear registers..................................................................... 51 ? register access restrictions...................................................... 51 ? thermal performance.................................................................... 52 ? power supply partitions................................................................. 53 ? recommended configuration for 3.3 v switching supply .. 53 ? configuration for 1.8 v supply ................................................ 53 ? pin program function description ............................................. 54 ? overview of on-chip rom features ..................................... 54 ? hard pin programming mode.................................................. 55 ? soft pin programming mode overview ................................. 55 ? register map ................................................................................... 56 ? register map bit descriptions ...................................................... 65 ? serial port configuration (register 0x0000 to register 0x0005) ......................................................................... 65 ? silicon revision (register 0x000a) ......................................... 65 ? clock part serial id (register 0x000c to register 0x000d) ........................................................................ 65 ? system clock (register 0x0100 to register 0x0108) ............. 66 ?
data sheet AD9557 rev. a | page 3 of 92 general configuration (register 0x0200 to register 0x0214)..........................................................................67 ? irq mask (register 0x020a to register 0x020f)...................68 ? dpll configuration (register 0x0300 to register 0x032e) .69 ? output pll configuration (register 0x0400 to register 0x0408)..........................................................................72 ? output clock distribution (register 0x0500 to register 0x0515)..........................................................................74 ? reference inputs (register 0x0600 to register 0x0602) ........76 ? dpll profile registers (register 0x0700 to register 0x0766)..........................................................................77 ? operational controls (register 0x0a00 to register 0x0a0d)........................................................................79 ? quick in/out frequency soft pin configuration (register 0x0c00 to register 0x0c08) .....................................82 ? status readback (register 0x0d00 to register 0x0d14) .......83 ? eeprom control (register 0x0e00 to register 0x0e3c) ....86 ? eeprom storage sequence (register 0x0e10 to register 0x0e3c).........................................................................86 ? outline dimensions........................................................................92 ? ordering guide ...........................................................................92 ? revision history 3/12rev. 0 to rev. a change to output frequency range parameter, table 6 ............. 6 changes to test conditions/comments column, table 9 .......... 8 changed name of pin 21 in figure 2............................................ 17 changes to table 20 ........................................................................ 18 changes to chip power monitor and startup, device register programming using a register setup file, and registers that differ from the defaults for optimal performance sections .... 25 changes to initialize and calibrate the output pll (apll) section .............................................................................................. 26 changes to program the reference profiles section; changed lock the digital pll section name to generate the reference acquisition; changes to generate the reference acquisition section .............................................................................................. 27 changes to figure 35; changed 225 mhz to 200 mhz and 3.45 ghz to 3.35 ghz in overview section................................ 28 changed 180 mhz to 175 mhz in dpll overview section .... 30 changed dpll output frequency to dco frequency throughout; changes to programmable digital loop filter section .............................................................................................. 31 changes to system clock inputs section..................................... 33 changed vco2 lower frequency to 3.35 ghz in figure 39; changes to output pll (apll) section...................................... 35 changed 1024 to 1023 in clock dividers section; changes to divider synchronization section.............................. 36 changes to the multifunction pins (m0 to m3) section ........... 37 added the programming the eeprom to configure an m pin to control synchronization of the clock distribution section..... 42 changes to the power supply partitions section ........................ 53 changed 89.5 to 88.5 in dpll phase margin section ............ 54 changes to register 0x000a, table 35 ......................................... 56 changes to register 0x0304, table 35 .......................................... 57 change to default value in register 0x0400 and register 0x0403; changes to register 0x0405, table 35 .......................................... 58 change to bit 0, register 0x070e, table 35 ................................. 59 change to bit 6, register 0x0d01, table 35................................. 63 added address 0x0e3d to address 0xe45, table 35 ................. 64 changes to description, register 0x0005, table 38; added table 40, renumbered sequentially; changes to descriptions, register 0x000c and register 0x000d, table 41 ... 65 changes to summary text, register 0x0200 to register 0x0209, table 46 and table 47........................................ 67 changes to register 0x0304, table 54; change to bits[7:6], table 55............................................................................................. 69 changes to table title, table 63; changes to description, register 0x0400 and register 0x0403, table 64 .......................... 72 changes to register 0x0405, table 64 .......................................... 73 changes to description column, register 0x0500, table 67; changes to description column, register 0x0501, bits[6:4] and bit 0, table 68 ........................................................................... 74 change to description column, register 0x0505, bits[6:4], table 70............................................................................................. 75 change to register 0x0600, bits[7:2], table 72 ........................... 76 changes to register 0x0707; change to register 0x070a, bits[3:0], table 76............................................................................ 77 changes to register 0x0a01, table 87 ......................................... 79 changes to table 96 ........................................................................ 81 changes to register 0x0d01, bit 6 and bit 1, table 99 .............. 83 added table 123.............................................................................. 89 changes to table 124 ...................................................................... 90 changes to table 125 ...................................................................... 91 10/11revision 0: initial version
AD9557 data sheet rev. a | page 4 of 92 specifications minimum (min) and maximum (max) values apply for the full range of supply voltage and operating temperature variations. typical (typ) values apply for avdd3 = dvdd_i/o = 3.3 v; avdd = dvdd = 1.8 v; t a = 25c, unless otherwise noted. supply voltage table 1. parameter min typ max unit test conditions/comments supply voltage dvdd3 3.135 3.30 3.465 v dvdd 1.71 1.80 1.89 v avdd3 3.135 3.30 3.465 v avdd 1.71 1.80 1.89 v supply current the test conditions for the maximum (max) supply current are the same as the test conditions for the all blocks running paramet er of table 3 . the test conditions for the typical (typ) supply current are the same as the test conditions for the typical configuration para meter of table 3 . table 2. parameter min typ max unit test conditions/comments supply current for typical configuration typical numbers are for the typical configuration listed in table 3 i dvdd3 12 18 26 ma pin 30, pin 31, pin 40 i dvdd 13 20 28 ma pin 6, pin 34, pin 35 i avdd3 35 49 63 ma pin 14, pin 19 i avdd 112 162 215 ma pin 7, pin 10, pin 11, pin 17, pin 18, pin 22, pin 23, pin 24 supply current for the all blocks running configuration maximum numbers are for all blocks running configuration in table 3 i dvdd3 12 18 33 ma pin 30, pin 31, pin 40 i dvdd 10 19 30 ma pin 6, pin 34, pin 35 i avdd3 47 68 89 ma pin 14, pin 19 i avdd 113 163 215 ma pin 7, pin 10, pin 11, pin 17, pin 18, pin 22, pin 23, pin 24
data sheet AD9557 rev. a | page 5 of 92 power dissipation table 3. parameter min typ max unit test conditions/comments power dissipation typical configuration 0.36 0.55 0.76 w system clock: 49.152 mhz crystal; dpll active; both 19.44 mhz input references in differential mode; one hstl driver at 644.53125 mhz; one 3.3 v cmos driver at 161.1328125 mhz and 80 pf capacitive load on cmos output all blocks running 0.39 0.61 0.85 w system clock: 49.152 mhz crystal; dpll active; both input references in differential mode; one hstl driver at 750 mhz; two 3.3 v cmos drivers at 250 mhz and 80 pf capacitive load on cmos outputs full power-down 44 125 mw typical configuration with no external pull-up or pull- down resistors; about 2/3 of this power is on avdd3 incremental power dissipation conditions = typical configuration; table values show the change in power due to the indicated operation input reference on/off differential without divide-by-2 20 25 32 mw addition al current draw is in the dvdd3 domain only differential with divide-by-2 26 32 40 mw addition al current draw is in the dvdd3 domain only single-ended without divide-by-2 5 7 9 mw addition al current draw is in the dvdd3 domain only output distribution driver on/off lvds (at 750 mhz) 12 17 22 mw additional cu rrent draw is in the avdd domain only hstl (at 750 mhz) 14 21 28 mw additional cu rrent draw is in the avdd domain only 1.8 v cmos (at 250 mhz) 14 21 28 mw a single 1.8 v cmos output with an 80 pf load 3.3 v cmos (at 250 mhz) 18 27 36 mw a single 3.3 v cmos output with an 80 pf load other blocks on/off second rf divider 36 51 64 mw additional cu rrent draw is in the avdd domain only channel divider bypassed 10 17 23 mw additional current draw is in the avdd domain only logic inputs ( reset , sync , pincontrol, m3 to m0) table 4. parameter min typ max unit test conditions/comments logic inputs ( reset , sync , pincontrol ) input high voltage (v ih ) 2.1 v input low voltage (v il ) 0.8 v input current (i inh , i inl ) 50 100 a input capacitance (c in ) 3 pf logic inputs (m3 to m0) input high voltage (v ih ) 2.5 v input ? level voltage (v im ) 1.0 2.2 v input low voltage (v il ) 0.6 v input current (i inh , i inl ) 60 100 a input capacitance (c in ) 3 pf
AD9557 data sheet rev. a | page 6 of 92 logic outputs (m3 to m0, irq) table 5. parameter min typ max unit test conditions/comments logic outputs (m3 to m0, irq) output high voltage (v oh ) dvdd3 ? 0.4 v i oh = 1 ma output low voltage (v ol ) 0.4 v i ol = 1 ma irq leakage current open-drain mode active low output mode ?200 a v oh = 3.3 v active high output mode 100 a v ol = 0 v system clock inputs (xoa, xob) table 6. parameter min typ max unit test conditions/comments system clock multiplier output frequency range 750 805 mhz the vco range may place limitations on nonstandard system clock input frequencies phase frequency detector (pfd) rate 150 mhz frequency multiplication range 2 255 a ssumes valid system clock and pfd rates system clock reference input path input frequency range 10 400 mhz minimum input slew rate 20 v/s minimum limit imposed for jitter performance common-mode voltage 1.05 1.16 1.25 v internally generated differential input voltage sensitivity 250 mv p-p minimum voltage across pins required to ensure switching between logic states; the instantaneous voltage on either pin must not exceed the supply rails; can accommodate single-ended input by ac grounding of complementary input; 1 v p-p recommended for optimal jitter performance system clock input doubler duty cycle this is the amount of duty cycle variation that can be tolerated on the system clock input to use the doubler system clock input = 50 mhz 45 50 55 % system clock input = 20 mhz 46 50 54 % system clock input = 16 mhz to 20 mhz 47 50 53 % input capacitance 3 pf single-ended, each pin input resistance 4.2 k crystal resonator path crystal resonator frequency range 10 50 mhz fundamental mode, at cut crystal maximum crystal motional resistance 100
data sheet AD9557 rev. a | page 7 of 92 reference inputs table 7. parameter min typ max unit test conditions/comments differential operation frequency range sinusoidal input 10 750 mhz lvpecl input 0.002 1250 mhz the reference input divide-by-2 block must be engaged for f in > 705 mhz lvds input 0.002 750 mhz the reference input divide-by-2 block must be engaged for f in > 705 mhz minimum input slew rate 40 v/s min imum limit imposed for jitter performance common-mode input voltage ac-coupled 1.9 2 2.1 v internally generated dc-coupled 1.0 2.4 v differential input voltage sensitivity mv minimum differential voltage across pins is required to ensure switching between logic levels; instantaneous voltage on either pin must not exceed the supply rails f in < 800 mhz 240 mv f in = 800 to 1050 mhz 320 mv f in = 1050 to 1250 mhz 400 mv differential input voltage hysteresis 58 100 mv input resistance 21 k input capacitance 3 pf minimum pulse width high lvpecl 390 ps lvds 640 ps minimum pulse width low lvpecl 390 ps lvds 640 ps single-ended operation frequency range (cmos) 0.002 300 mhz minimum input slew rate 40 v/s min imum limit imposed for jitter performance input voltage high (v ih ) 1.2 v to 1.5 v threshold setting 1.0 v 1.8 v to 2.5 v threshold setting 1.4 v 3.0 v to 3.3 v threshold setting 2.0 v input voltage low (v il ) 1.2 v to 1.5 v threshold setting 0.35 v 1.8 v to 2.5 v threshold setting 0.5 v 3.0 v to 3.3 v threshold setting 1.0 v input resistance 47 k input capacitance 3 pf minimum pulse width high 1.5 ns minimum pulse width low 1.5 ns
AD9557 data sheet rev. a | page 8 of 92 reference monitors table 8. parameter min typ max unit test conditions/comments reference monitors reference monitor loss of reference detection time 1.1 dpll pfd period nominal phase detector period = r/f ref 1 frequency out-of range limits <2 10 5 f/f ref (ppm) programmable (lower bound is subject to quality of the system clock (sysclk)); sysclk accuracy must be better than the lower bound validation timer 0.001 65.535 sec programmable in 1 ms increments 1 f ref is the frequency of the active reference; r is the frequency division factor determined by the r divider. reference switchover specifications table 9. parameter min typ max unit test conditions/comments reference switchover specifications maximum output phase perturbation (phase build-out switchover) assumes a jitter-free reference; satisfies telcordia gr-1244-core requirements; select high pm base loop filter bi t (register 0x070e, bit 0) is set to 1 for all active references 50 hz dpll loop bandwidth valid for automatic and manual reference switching peak 0 100 ps steady state 0 100 ps 2 khz dpll loop bandwidth valid for automatic and manual reference switching peak 0 250 ps steady state 0 100 ps time required to switch to a new reference phase build-out switchover 1.1 dpll pfd period calculated using the nominal phase detector period (npdp = r/f ref ); the total time required is equal to the time plus the reference validation time and the time required to lock to the new reference
data sheet AD9557 rev. a | page 9 of 92 distribution clock outputs table 10. parameter min typ max unit test conditions/comments hstl mode output frequency 0.36 1250 mhz rise/fall time (20% to 80%) 1 140 250 ps 100 termination across output pins duty cycle up to f out = 700 mhz 45 48 52 % up to f out = 750 mhz 42 48 53 % up to f out = 1250 mhz 43 % differential output voltage swing 700 950 1200 mv magnitude of voltage across pins; output driver static common-mode output voltage 700 870 960 mv output driver static lvds mode output frequency 0.36 1250 mhz rise/fall time (20% to 80%) 1 185 280 ps 100 termination across the output pair duty cycle up to f out = 750 mhz 44 48 53 % up to f out = 800 mhz 43 47 53 % up to f out = 1250 mhz 43 % differential output voltage swing balanced, v od 247 454 mv voltage swing between output pins; output driver static unbalanced, v od 50 mv absolute difference between voltage swing of normal pin and inverted pin; output driver static offset voltage common mode, v os 1.125 1.26 1.375 v output driver static common-mode difference, v os 50 mv voltage difference between pins; output driver static short-circuit output current 13 24 ma output driver static cmos mode output frequency 1.8 v supply 0.36 150 mhz 10 pf load 3.3 v supply (out0) strong drive strength setting 0.36 250 mhz 10 pf load weak drive strength setting 0.36 25 mhz 10 pf load rise/fall time(20% to 80%) 1 1.8 v supply 1.5 3 ns 10 pf load 3.3 v supply strong drive strength setting 0.4 0.6 ns 10 pf load weak drive strength setting 8 ns 10 pf load duty cycle 1.8 v mode 50 % 10 pf load 3.3 v strong mode 47 % 10 pf load 3.3 v weak mode 51 % 10 pf load output voltage high (v oh ) output driver static; strong drive strength avdd3 = 3.3 v, i oh = 10 ma avdd3 ? 0.3 v avdd3 = 3.3 v, i oh = 1 ma avdd3 ? 0.1 v avdd3 = 1.8 v, i oh = 1 ma avdd ? 0.2 v output voltage low (v ol ) output driver static; strong drive strength avdd3 = 3.3 v, i ol = 10 ma 0.3 v avdd3 = 3.3 v, i ol = 1 ma 0.1 v avdd3 = 1.8 v, i ol = 1 ma 0.1 v
AD9557 data sheet rev. a | page 10 of 92 parameter min typ max unit test conditions/comments output timing skew 10 pf load between out0 and out1 10 70 ps hstl mode on both drivers; rising edge only; any divide value additional delay on one driver by changing its logic type hstl to lvds ?5 +1 +5 ps positive value indicates that the lvds edge is delayed relative to hstl hstl to 1.8 v cmos ?5 0 +5 ps positive value indicates that the cmos edge is delayed relative to hstl out1 hstl to out0 3.3 v cmos, strong mode 3.53 3.59 ns the cmos edge is delayed relative to hstl 1 the listed values are for the slower edge (rise or fall). time duration of digital functions table 11. parameter min typ max unit test conditions/comments time duration of digital functions eeprom-to-register download time 13 20 ms using default eeprom storage sequence (see register 0x0e10 to register 0x0e3f) register-to-eeprom upload time 138 145 ms using default eeprom storage sequence (see register 0x0e10 to register 0x0e3f minimum power-down exit time 1 ms time from power-down exit to system clock lock detect
data sheet AD9557 rev. a | page 11 of 92 digital pll table 12. parameter min typ max unit test conditions/comments digital pll phase-frequency detector (pfd) input frequency range 2 100 khz loop bandwidth 0.1 2000 hz programmable design parameter phase margin 30 89 degrees programmable design parameter closed-loop peaking <0.1 db programmable design parameter; part can be programmed for <0.1 db peaking in accordance with telcordia gr-253 jitter transfer reference input (r) division factor 1 2 20 1, 2, , 1,048,576 integer feedback (n1) division factor 180 2 17 180, 181, , 131,072 fractional feedback divide ratio 0 0.999 maximum value: 16,777,215/16,777,216 digital pll lock detection table 13. parameter min typ max unit test conditions/comments phase lock detector threshold programming range 0.001 65.5 ns threshold resolution 1 ps frequency lock detector threshold programming range 0.001 16,700 ns reference-to-feedback period difference threshold resolution 1 ps holdover specifications table 14. parameter min typ max unit test conditions/comments holdover specifications initial frequency accuracy <0.01 ppm excludes frequency drift of sysclk source; excludes frequency drift of input reference prior to entering holdover; compliant with gr-1244 stratum 3
AD9557 data sheet rev. a | page 12 of 92 serial port specificationsspi mode table 15. parameter min typ max unit test conditions/comments cs input logic 1 voltage 2.2 v input logic 0 voltage 1.2 v input logic 1 current 44 a input logic 0 current 88 a input capacitance 2 pf sclk internal 30 k pull-down resistor input logic 1 voltage 2.2 v input logic 0 voltage 0.8 1.2 v input logic 1 current 200 a input logic 0 current 1 a input capacitance 2 pf sdio as an input input logic 1 voltage 2.2 v input logic 0 voltage 1.2 v input logic 1 current 1 a input logic 0 current 1 a input capacitance 2 pf as an output output logic 1 voltage dvdd3 ? 0.6 v 1 ma load current output logic 0 voltage 0.4 v 1 ma load current sdo output logic 1 voltage dvdd3 ? 0.6 v 1 ma load current output logic 0 voltage 0.4 v 1 ma load current timing sclk clock rate, 1/t clk 40 mhz pulse width high, t high 10 ns pulse width low, t low 13 ns sdio to sclk setup, t ds 3 ns sclk to sdio hold, t dh 6 ns sclk to valid sdio and sdo, t dv 10 ns cs to sclk setup (t s ) 10 ns cs to sclk hold (t c ) 0 ns cs minimum pulse width high 6 ns
data sheet AD9557 rev. a | page 13 of 92 serial port specificationsi 2 c mode table 16. parameter min typ max unit test conditions/comments sda, scl (as input) input logic 1 voltage 0.7 dvdd3 v input logic 0 voltage 0.3 dvdd3 v input current ?10 +10 a for v in = 10% to 90% dvdd3 hysteresis of schmitt trigger inputs 0.015 dvdd3 pulse width of spikes that must be suppressed by the input filter, t sp 50 ns sda (as output) output logic 0 voltage 0.4 v i o = 3 ma output fall time from v ihmin to v ilmax 20 + 0.1 c b 1 250 ns 10 pf c b 400 pf 1 timing scl clock rate 400 khz bus-free time between a stop and start condition, t buf 1.3 s repeated start condition setup time, t su; sta 0.6 s repeated hold time start condition, t hd; sta 0.6 s after this period, the first clock pulse is generated stop condition setup time, t su; sto 0.6 s low period of the scl clock, t low 1.3 s high period of the scl clock, t high 0.6 s scl/sda rise time, t r 20 + 0.1 c b 1 300 ns scl/sda fall time, t f 20 + 0.1 c b 1 300 ns data setup time, t su; dat 100 ns data hold time, t hd; dat 100 ns capacitive load for each bus line, c b 1 400 pf 1 c b is the capacitance (pf) of a single bus line. jitter generation jitter generation (random jitter) uses 49.152 mhz crystal for system clock input. table 17. parameter min typ max unit test conditions/comments jitter generation system clock doubler enabled; high phase margin mode enabled; register 0x0405 = 0x20; register 0x0403 = 0x07; register 0x0400 = 0x81; in cases where multiple driver types are listed, both driver types were tested at those conditions, and the one with higher jitter is quoted, although there is usually not a significant jitter difference between the driver types f ref = 19.44 mhz; f out = 622.08 mhz; f loop = 50 hz hstl driver bandwidth: 5 khz to 20 mhz 304 fs rms bandwidth: 12 khz to 20 mhz 296 fs rms bandwidth: 20 khz to 80 mhz 300 fs rms bandwidth: 50 khz to 80 mhz 266 fs rms bandwidth: 16 mhz to 320 mhz 185 fs rms
AD9557 data sheet rev. a | page 14 of 92 parameter min typ max unit test conditions/comments f ref = 19.44 mhz; f out = 644.53 mhz; f loop = 50 hz hstl and/or lvds driver bandwidth: 5 khz to 20 mhz 334 fs rms bandwidth: 12 khz to 20 mhz 321 fs rms bandwidth: 20 khz to 80 mhz 319 fs rms bandwidth: 50 khz to 80 mhz 277 fs rms bandwidth: 16 mhz to 320 mhz 185 fs rms f ref = 19.44 mhz; f out = 693.48 mhz; f loop = 50 hz hstl driver bandwidth: 5 khz to 20 mhz 298 fs rms bandwidth: 12 khz to 20 mhz 285 fs rms bandwidth: 20 khz to 80 mhz 286 fs rms bandwidth: 50 khz to 80 mhz 252 fs rms bandwidth: 16 mhz to 320 mhz 183 fs rms f ref = 19.44 mhz; f out = 174.703 mhz; f loop = 1 khz hstl driver bandwidth: 5 khz to 20 mhz 354 fs rms bandwidth: 12 khz to 20 mhz 301 fs rms bandwidth: 20 khz to 80 mhz 321 fs rms bandwidth: 50 khz to 80 mhz 290 fs rms bandwidth: 4 mhz to 80 mhz 177 fs rms f ref = 19.44 mhz; f out = 174.703 mhz; f loop = 100 hz lvds and/or 3.3 v cmos driver bandwidth: 5 khz to 20 mhz 306 fs rms bandwidth: 12 khz to 20 mhz 293 fs rms bandwidth: 20 khz to 80 mhz 313 fs rms bandwidth: 50 khz to 80 mhz 283 fs rms bandwidth: 4 mhz to 80 mhz 166 fs rms f ref = 25 mhz; f out = 161.1328 mhz; f loop = 100 hz hstl driver bandwidth: 5 khz to 20 mhz 316 fs rms bandwidth: 12 khz to 20 mhz 302 fs rms bandwidth: 20 khz to 80 mhz 324 fs rms bandwidth: 50 khz to 80 mhz 292 fs rms bandwidth: 4 mhz to 80 mhz 171 fs rms f ref = 2 khz; f out = 70.656 mhz; f loop = 100 hz; hstl and/or 3.3 v cmos driver bandwidth: 10 hz to 30 mhz 3.22 ps rms bandwidth: 5 khz to 20 mhz 338 fs rms bandwidth: 12 khz to 20 mhz 324 fs rms bandwidth: 10 khz to 400 khz 278 fs rms bandwidth: 100 khz to 10 mhz 210 fs rms f ref = 25 mhz; f out = 1 ghz; f loop = 500 hz hstl driver bandwidth: 100 hz to 500 mhz (broadband) 1.71 ps rms bandwidth: 12 khz to 20 mhz 343 fs rms bandwidth: 20 khz to 80 mhz 338 fs rms
data sheet AD9557 rev. a | page 15 of 92 jitter generation (random jitter) uses 19.2 mhz tcxo for system clock input. table 18. parameter min typ max unit test conditions/comments jitter generation system clock doubler enabled; high phase margin mode enabled; register 0x0405 = 0x20; register 0x0403 = 0x07; register 0x0400 = 0x81; in cases where multiple driver types are listed, both driver types were tested at those conditions, and the one with higher jitter is quoted, although there is usually not a significant jitter difference between the driver types f ref = 19.44 mhz; f out = 644.53 mhz; f loop = 0.1 hz hstl driver bandwidth: 5 khz to 20 mhz 402 fs rms bandwidth: 12 khz to 20 mhz 393 fs rms bandwidth: 20 khz to 80 mhz 391 fs rms bandwidth: 50 khz to 80 mhz 347 fs rms bandwidth: 16 mhz to 320 mhz 179 fs rms f ref = 19.44 mhz; f out = 693.48 mhz; f loop = 0.1 hz hstl driver bandwidth: 5 khz to 20 mhz 379 fs rms bandwidth: 12 khz to 20 mhz 371 fs rms bandwidth: 20 khz to 80 mhz 371 fs rms bandwidth: 50 khz to 80 mhz 335 fs rms bandwidth: 16 mhz to 320 mhz 175 fs rms f ref = 19.44 mhz; f out = 312.5 mhz; f loop = 0.1 hz hstl driver bandwidth: 5 khz to 20 mhz 413 fs rms bandwidth: 12 khz to 20 mhz 404 fs rms bandwidth: 20 khz to 80 mhz 407 fs rms bandwidth: 50 khz to 80 mhz 358 fs rms bandwidth: 4 mhz to 80 mhz 142 fs rms f ref = 25 mhz; f out = 161.1328 mhz; f loop = 0.1 hz hstl driver bandwidth: 5 khz to 20 mhz 399 fs rms bandwidth: 12 khz to 20 mhz 391 fs rms bandwidth: 20 khz to 80 mhz 414 fs rms bandwidth: 50 khz to 80 mhz 376 fs rms bandwidth: 4 mhz to 80 mhz 190 fs rms f ref = 2 khz; f out = 70.656 mhz; f loop = 0.1 hz hstl and/or 3.3 v cmos driver bandwidth: 10 hz to 30 mhz 970 fs rms bandwidth: 12 khz to 20 mhz 404 fs rms bandwidth: 10 khz to 400 khz 374 fs rms bandwidth: 100 khz to 10 mhz 281 fs rms
AD9557 data sheet rev. a | page 16 of 92 absolute maximum ratings table 19. parameter rating analog supply voltage (avdd) 2 v digital supply voltage (dvdd) 2 v digital i/o supply voltage (dvdd3) 3.6 v analog supply voltage (avdd3) 3.6 v maximum digital input voltage ?0.5 v to dvdd3 + 0.5 v storage temperature range ?65c to +150c operating temperature range ?40c to +85c lead temperature (soldering 10 sec) 300c junction temperature 150c stresses above those listed under absolute maximum ratings may cause permanent damage to the device. this is a stress rating only; functional operation of the device at these or any other conditions above those indicated in the operational section of this specification is not implied. exposure to absolute maximum rating conditions for extended periods may affect device reliability. esd caution
data sheet AD9557 rev. a | page 17 of 92 pin configuration and fu nction descriptions pin 1 indicator notes 1. the exposed pad must be connected to ground (vss). 1 irq 2 sclk/scl 3 sdio/sda 4 sdo 5 cs 6 dvdd 7 avdd 8 xoa 9 xob 10 avdd 23 avdd 24 avdd 25 reset 26 pincontrol 27 sync 28 refa 29 refa 30 dvdd3 22 avdd 21 lf_vco2 11 a vdd 12 out1 13 out1 15 out0 17 avdd 16 out0 18 avdd 19 avdd3 20 ldo_v co2 14 avdd3 33 refb 34 dvdd 35 dvdd 36 m0 37 m1 38 m2 39 m3 40 dvdd3 32 ref b 31 dvdd3 top view (not to scale) AD9557 09197-002 figure 2. pin configuration table 20. pin function descriptions pin no. mnemonic input/ output pin type description 1 irq o 3.3 v cmos interrupt request line. 2 sclk/scl i 3.3 v cmos serial programming clock (sclk) in spi mo de. data clock for serial programming. serial clock pin (scl) in i 2 c mode. 3 sdio/sda i/o 3.3 v cmos serial data input/output (sdio) in spi mo de. when the device is in 4-wire spi mode, data is written via this pin. in 3-wire mode, both data reads and writes occur on this pin. there is no internal pull-up/pull-down resistor on this pin. serial data pin (sda) in i 2 c mode. 4 sdo o 3.3 v cmos serial data output. use this pin to read da ta in 4-wire mode. there is no internal pull-up/pull-down resistor on this pin. this pin is high impedance in the default 3-wire mode. 5 cs i 3.3 v cmos chip select (spi), active low. when programming a device, this pin must be held low. in systems where more than one AD9557 is present, this pin enables individual programming of each AD9557 . this pin has an internal 10 k pull-up resistor. 6, 34, 35 dvdd i power 1.8 v digital supply. 7, 10, 22, 23, 24 avdd i power 1.8 v analog power supply. 8 xoa i differential input system clock input. xoa contains internal dc biasing and should be ac-coupled with a 0.01 f capacitor, except when using a crystal, in which case connect the crystal across xoa and xob. single-ended 1.8 v cmos is also an option but can introduce a spur if the duty cycle is no t 50%. when using xoa as a single-ended input, connect a 0.01 f capacitor from xob to ground. 9 xob i differential input complementary system clock input. complementary signal to xoa. xob contains internal dc biasing and should be ac-coupled with a 0.01 f capacitor, except when using a crystal, in which case connect the crystal across xoa and xob. 11, 17, 18 avdd i power 1.8 v analog (output divider and drivers) power supply. 12 out1 o hstl, lvds, or 1.8 v cmos complementary output 1. this output can be configured as hstl, lvds, or single-ended 1.8 v cmos. 13 out1 o hstl, lvds, or 1.8 v cmos output 1. this output can be configured as hstl, lvds, or single-ended 1.8 v cmos. lvpecl levels can be achieved by ac coupling and using the thevenin-equivalent termination as described in the input/output termination recommendations section. 14, 19 avdd3 i power 3.3 v analog power supply.
AD9557 data sheet rev. a | page 18 of 92 pin no. mnemonic input/ output pin type description 15 out0 o hstl, lvds, 1.8 v cmos, 3.3 v cmos complementary output 0. this output can be configured as hstl, lvds, or single-ended 1.8 v or 3.3 v cmos. 16 out0 o hstl, lvds, 1.8 v cmos, 3.3 v cmos output 0. this output can be configured as hstl, lvds, or single-ended 1.8 v or 3.3 v cmos. lvpecl levels can be achieved by ac coupling and using the thevenin-equivalent termination as described in the input/output termination recommendations section. 20 ldo_vco2 i ldo bypass output pll loop filter voltage regulator. connect a 0.47 f capacitor from this pin to ground. this pin is also the ac ground reference for the integrated output pll external loop filter. 21 lf_vco2 i/o loop filter loop filter node for the output pll. connect an external 6.8 nf capacitor from this pin to pin 20 (ldo_vco2). 25 reset i 3.3 v cmos chip reset. when this active low pin is asserted, the chip goes into reset. this pin has an internal 50 k pull-up resistor. 26 pincontrol i 3.3 v cmos pin program mode enable pin. when pulled high during startup, this pin enables pin programming of the AD9557 configuration during startup. if this pin is low during startup, the user must program the part via the serial port or use values that are stored in the eeprom. 27 sync i 3.3 v cmos clock distribution synchronization pin. when this pin is activated, output drivers are held static and then synchronized on a low-to-high transition of this pin. this pin has an internal 60 k pull-up resistor. 28 refa i differential input reference a input. this internally biased input is typically ac-coupled and, when configured as such, can accept any differen tial signal with single-ended swing up to 3.3 v. if dc-coupled, input can be lvpecl, lvds, or single-ended cmos. 29 refa i differential input complementary reference a input. this pin is the complementary input to pin 28. 30, 31, 40 dvdd3 i power 3.3 v digital power supply. 32 refb i differential input reference b input. this internally biased input is typically ac-coupled and, when configured as such, can accept any differen tial signal with single-ended swing up to 3.3 v. if dc-coupled, input can be lvpecl, lvds, or single-ended cmos. 33 refb i differential input complementary reference b input. this pin is the complementary input to pin 32. 36, 37, 38, 39 m0, m1, m2, m3 i/o 3.3 v cmos (3-level logic at startup) configurable i/o pins. these pins are 3-le vel logic at startup and are used for pin strapping the input and o utput frequency configuration at startup. setting register 0x0200[0] = 1 changes these pins to 2-level logic and allows these pins to be used for status and control of the AD9557 . these pins have both a 30 k pull-up resistor and a 30 k pull-down resistor. ep vss o exposed pad the exposed pad mu st be connected to ground (vss).
data sheet AD9557 rev. a | page 19 of 92 typical performance characteristics f r = input reference clock frequency; f o = output clock frequency; f sys = sysclk input frequency; f s = internal system clock frequency; lf = sysclk pll internal loop filter used. avdd, avdd3, and dvdd at nominal supply voltage; f s = 786.432 mhz, unless otherwise noted. ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ?70 ? 60 1k 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 296fs 09197-003 ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ?70 ? 60 1k 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 285fs 09197-005 figure 3. absolute phase noise (output driver = hstl), f r = 19.44 mhz, f o = 622.08 mhz, dpll loop bw = 50 hz, f sys = 49.152 mhz crystal figure 5. absolute phase noise (output driver = hstl), f r = 19.44 mhz, f o = 693.482991 mhz, dpll loop bw = 50 hz, f sys = 49.152 mhz crystal ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ?70 ? 60 1k 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 320fs 09197-004 ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ? 70 1k 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 301fs 09197-006 figure 4. absolute phase noise (output driver = hstl), f r = 19.44 mhz, f o = 644.53125 mhz, dpll loop bw = 50 hz, f sys = 49.152 mhz crystal figure 6. absolute phase noise (output driver = hstl), f r = 19.44 mhz, f o = 174.703 mhz, dpll loop bw = 1 khz, f sys = 49.152 mhz crystal
AD9557 data sheet rev. a | page 20 of 92 ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ? 80 1k 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 302fs 09197-007 figure 7. absolute phase noise (output driver = 3.3.v cmos), f r = 19.44 mhz, f o = 161.1328125 mhz, dpll loop bw = 100 hz, f sys = 49.152 mhz crystal ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ? 70 1k 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 308fs 09197-008 figure 8. absolute phase noise (output driver = hstl), f r = 2 khz, f o = 125 mhz, dpll loop bw = 100 hz, f sys = 49.152 mhz crystal ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ?70 ? 60 1k 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 343fs 09197-009 figure 9. absolute phase noise (output driver = hstl), f r = 25 mhz, f o = 1 ghz, dpll loop bw = 500 hz, f sys = 49.152 mhz crystal ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ?70 ? 60 1k 10 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 393fs 09197-010 figure 10. absolute phase noise (output driver = hstl), f r = 19.44 mhz, f o = 644.53 mhz, dpll loop bw = 0.1 hz, f sys = 19.2 mhz tcxo ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ?70 ? 60 phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 371s 1k 10 100 10k 100k 1m 10m 100m 09197-011 figure 11. absolute phase noise (output driver = hstl), f r = 19.44 mhz, f o = 693.482991 mhz, dpll loop bw = 0.1 hz, f sys = 19.2 mhz tcxo ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ? 70 1k 10 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 404fs 09197-012 figure 12. absolute phase noise (output driver = hstl), f r = 19.44 mhz, f o = 312.5 mhz, dpll loop bw = 0.1 hz, f sys = 19.2 mhz tcxo
data sheet AD9557 rev. a | page 21 of 92 ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ? 70 1k 10 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 391fs 09197-013 figure 13. absolute phase noise (output driver = 3.3 v cmos), f r = 19.44 mhz, f o =161.1328125 mhz, dpll loop bw = 0.1 hz, f sys = 19.2 mhz tcxo ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ? 70 1k 10 100 10k 100k 1m 10m 100m phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 395fs 09197-014 figure 14. absolute phase noise (output driver = 1.8 v cmos), f r = 2 khz, f o = 70.656 mhz, dpll loop bw = 0.1 hz, f sys = 19.2 mhz tcxo 1k 10 100 10k 100k 1m 10m 100m ?160 ?150 ?140 ?130 ?120 ?110 ?100 ?90 ?80 ?70 ? 60 phase noise (dbc/hz) frequency offset (hz) integrated rms jitter (12khz to 20mhz): 388fs 09197-016 figure 15. absolute phase noise (output driver = hstl), f r = 19.44 mhz, f o = 644.53 mhz, f sys = 19.2 mhz tcxo, holdover mode 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 differenti a l peak-to-peak amplitude (v) 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 frequency (mhz) 09197-116 figure 16. amplitude vs. toggle rate, hstl mode (lvpecl-compatible mode) 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 100 200 300 400 500 600 700 800 differenti a l peak-to-peak amplitude (v) frequency (mhz) lvds boost mode lvds default 09197-117 figure 17. amplitude vs. toggle rate, lvds 1.0 1.5 2.0 2.5 3.0 3.5 0 50 100 150 200 250 300 peak-to-peak amplitude (v) frequency (mhz) 3.3v cmos 1.8v cmos 09197-118 figure 18. amplitude vs. toggl e rate with 10 pf load, 3.3 v (strong mode) and 1.8 v cmos
AD9557 data sheet rev. a | page 22 of 92 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 1020304050607080 peak-to-peak amplitude (v) frequency (mhz) 09197-119 figure 19. amplitude vs. toggl e rate with 10 pf load, 3.3 v (weak mode) cmos 09197-120 30 35 40 45 50 55 60 65 70 75 0 250 500 750 1000 1250 power (mw) frequency (mhz) figure 20. power consum ption vs. frequency, hstl mode on output driver power supply only (pin 11 and pin 17) 09197-121 0 5 10 15 20 25 30 35 40 45 50 55 60 65 0 100 200 300 400 500 600 700 800 power (mw) frequency (mhz) figure 21. power consumption vs. freque ncy, lvds mode on output driver power supply only (pin 11 and pin 17) 0 10 20 30 40 50 60 70 0 50 100 150 200 power (mw) frequency (mhz) 1.8v cmos mode 3.3v cmos strong mode 3.3v cmos weak mode 09197-122 figure 22. power consumption vs. freque ncy, cmos mode on output driver power supply only (pin 11 and pin 17) for 1.8 v cmos mode or on pin 19 for 3.3 v cmos mode, one cmos driver ?1.0 ?0.8 ?0.6 ?0.4 ?0.2 0 0.2 0.4 0.6 0.8 1.0 ?1012345 differenti a l amplitude (v) time (ns) 09197-123 figure 23. output waveform, hstl (400 mhz) ?0.4 ?0.3 ?0.2 ?0.1 0 0.1 0.2 0.3 0.4 ? 101234 differential amplitude (v) time (ns) 09197-124 figure 24. output waveform, lvds (400 mhz)
data sheet AD9557 rev. a | page 23 of 92 ?30 ?27 ?24 ?21 ?18 ?15 ?12 ?9 ?6 ?3 0 3 10 100 1k 10k 100k frequency offset (hz) lo o p gain (db) 09197-129 loop bw = 100hz; high phase margin; peaking: 0.06db; ?3db: 69hz loop bw = 2khz; high phase margin; peaking: 0.097db; ?3db: 1.23khz loop bw = 5khz; high phase margin; peaking: 0.14db; ?3db: 4.27khz ?0.2 0.2 0.6 1.0 1.4 1.8 2.2 2.6 3.0 3.4 ?10123456789101112131415 amplitude (v) time (ns) 10pf load 2pf load 09197-126 figure 28. closed-loop transfer function for 100 hz, 2 khz, and 5 khz loop bandwidth settings; high phase margin loop filter setting (this is compliant with telcordia gr-253 jitter transfer test for loop bandwidths < 2 khz.) figure 25. output waveform, 3.3 v cmos (100 mhz, strong mode) ?10123456789101112131415 amplitude (v) time (ns) ?0.1 0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 10pf load 2pf load 09197-127 ?30 ?27 ?24 ?21 ?18 ?15 ?12 ?9 ?6 ?3 0 3 10 100 1k frequency offset (hz) lo o p gain (db) 10k 100k 09197-230 loop bw = 100hz; normal phase margin; peaking: 0.09db; ?3db: 117hz loop bw = 2khz; normal phase margin; peaking: 1.6db; ?3db: 2.69khz figure 29. closed-loop transfer function for 100 hz and 2 khz loop bandwidth settings; normal phas e margin loop filter setting figure 26. output waveform, 1.8 v cmos (100 mhz) 0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 ?5 5 152535455565758595 amplitude (v) time (ns) 10pf load 2pf load 09197-128 figure 27. output waveform, 3. 3 v cmos (20 mhz, weak mode)
AD9557 data sheet rev. a | page 24 of 92 input/output termination recommendations AD9557 ad9558 hstl or lvds downstream device with high impedance input and internal dc-bias 0.1f 0.1f 100? 09197-130 figure 30. ac-coupled lvds or hstl output driver (100 resistor can go on either side of decoupling capacitors and should be as close as possible to the destination receiver.) AD9557 ad9558 hstl or lvds z 0 = 50 ? z 0 = 50 ? single-ended (not coupled) lvds or 1.8v hstl high-impedance differential receiver 100 ? 09197-131 figure 31. dc-coupled lvds or hstl output driver single-ended (not coupled) v s = 3.3 v 3.3v lvpecl 82 ? 82? 127? 127? 0.1f 0.1f AD9557 ad9558 1.8v hstl z 0 = 50 ? z 0 = 50 ? 09197-132 figure 32. interfacing the hstl driver to a 3.3 v lvpecl input (this method incorporates impedance matching and dc biasing for bipolar lvpecl receivers. if the receiver is self-biased, the termination scheme shown in figure 30 is recommended.) xoa xob AD9557/ ad9558 10mhz to 50mhz fundamental at-cut crystal with 10pf load capacitance 10pf 10pf 09197-133 figure 33. system clock input (xoa, xob) in crystal mode (the recommended c load = 10 pf is shown. the values of the 10 pf shunt capacitors shown here should equal the c load of the crystal.) xoa 300 ? 150 ? 0.1f xob AD9557/ ad9558 3.3v cmos tcxo 0.1f 09197-134 figure 34. system clock input (xoa, xob) when using a tcxo/ocxo with 3.3 v cmos output
data sheet AD9557 rev. a | page 25 of 92 getting started chip power monitor and startup the AD9557 monitors the voltage on the power supplies at power-up. when dvdd3 is greater than 2.35 v 0.1 v and dvdd and avdd are greater than 1.4 v 0.05 v, the device generates a 20 ms reset pulse. the power-up reset pulse is internal and independent of the reset pin. this internal power-up reset sequence eliminates the need for the user to provide external power supply sequencing. within 45 ns after the leading edge of the internal reset pulse, the m3 to m0 multifunction pins behave as high impedance digital inputs and continue to do so until programmed otherwise. during a device reset (either via the power-up reset pulse or the reset pin), the multifunction pins (m3 to m0) behave as high impedance inputs; but upon removal of the reset condition, level-sensitive latches capture the logic pattern present on the multifunction pins. multifunction pins at reset/power-up the AD9557 requires the user to supply the desired logic state to the pincontrol pin, as well as the m3 to m0 pins. if pincontrol is high, the part is in hard pin programming mode. see the pin program function description section for details on hard pin programming. at startup, there are three choices for the m3 to m0 pins: pull-up, pull-down, and floating. if the pincontrol pin is low, the m3 to m0 pins determine the following configurations: ? following a reset, the m1 and m0 pins determine whether the serial port interface behaves according to the spi or i 2 c protocol. specifically, 0x00 selects the spi interface, and any other value selects the i 2 c port. the 3-level logic of m1 and m0 allows the user to select eight possible i 2 c addresses (see table 24 for details). ? the m3 and m2 pins select which of the eight possible eeprom profiles are loaded, or if the eeprom loading is bypassed. leaving m3 and m2 floating at startup bypasses the eeprom loading, and the factory defaults are used instead (see table 22 for details). device register programming using a register setup file t he evaluation software contains a programming wizard and a convenient graphical user interface that assists the user in determining the optimal configuration for the dpll, apll, and sysclk based on the desired input and output frequencies. it generates a register setup file with a .stp extension that is easily readable using a text editor. a fter using the evaluation software to create the setup file, use the following sequence to program the AD9557 once: 1. register 0x0a01 = 0x20 (set user free run mode). 2. register 0x0a02 = 0x02 (hold outputs in static sync). (skip this step if using sync on dpll phase lock or sync on dpll frequency lock. see register 0x0500[1:0].) 3. register 0x0405 = 0x20 (clear apll vco calibration). 4. write the register values in the stp file from address 0x0000 to address 0x032e. 5. register 0x0005 = 0x01 (update all registers). 6. write the rest of the registers in the stp file, starting at address 0x0400. 7. register 0x0405 = 0x21 (calibrate apllon next i/o update). 8. register 0x0403 = 0x07 (configure apll). 9. register 0x0400 = 0x81 (configure apll). 10. register 0x0005 = 0x01 (update all registers). 11. register 0x0a01[5] = 0b (clear user free run mode). 12. register 0x0005 = 0x01 (update all registers). register programming overview this section provides an overview of the register blocks in the AD9557 , describing what they do and why they are important. reg isters differing from defaults for optimal performance ens ure that the following registers are programmed to the listed values for optimal performance: ? register 0x0405[7:4] = 0x2 ? register 0x0403 = 0x07 ? register 0x0400 = 0x81 if the silicon revision (registe r 0x000a) equals 0x21 or higher, the values listed here are already the default values.
AD9557 data sheet rev. a | page 26 of 92 p rogram the system clock an d free run tuning word t he system clock multiplier (sysclk) parameters are at register 0x0100 to register 0x0108, and the free run tuning word is at register 0x0300 to register 0x0303. use the following steps for optimal performance: 1. set the system clock pll input type and divider values. 2. set the system clock period. it is essential to program the system clock period because many of the AD9557 subsystems rely on this value. 3. set the system clock stability timer. it is highly recommended that the system clock stability timer be programmed. this is especially important when using the system clock multiplier and also applies when using an external system clock source, especially if the external source is not expected to be completely stable when power is applied to the AD9557 . the system clock stability timer specifies the amount of time that the system clock pll must be locked before the part declares that the system clock is stable. the default value is 50 ms. 4. program the free run tuning word. the free run frequency of the digital pll (dpll) determines the frequency appearing at the apll input when free run mode is selected. the free run tuning word is at register 0x0300 to register 0x0303. the correct free run frequency is required for the apll to calibrate and lock correctly. 5. set user free run mode (register 0x0a01[5] = 1b). in itialize and calibrate the output pll (apll) th e registers controlling the apll are at register 0x0400 to register 0x0408. this low noise, integer-n pll multiplies the dpll output (which is usually 175 mhz to 200 mhz) to a frequency in the 3.35 ghz to 4.05 ghz range. after the system clock is configured and the free run tuning word is set in register 0x0300 to register 0x0303, the user can set the manual apll vco calibration bit (register 0x0405[0]) and issue an i/o update (register 0x0005[0]). this process performs the apll vco calibration. vco calibration ensures that, at the time of calibration, the dc control voltage of the apll vco is centered in the middle of its operating range. it is important to remember the following points when calibrating the apll vco: ? t he system clock must be stable. ? the apll vco must have the correct frequency from the 30-bit dco (digitally controlled oscillator) during calibration. ? the apll vco must be recalibrated any time the apll frequency changes. ? apll vco calibration occurs on the low-to-high transition of the manual apll vco calibration bit, and this bit is not autoclearing. therefore, this bit must be cleared (and an i/o update issued) before another apll calibration is started. ? the best way to monitor successful apll calibration is to monitor bit 2 in register 0x0d01 (apll lock). p rogram the clock distribution outputs t he apll output goes to the clock distribution block. the clock distribution parameters reside in register 0x0500 to register 0x0509. they include the following: ? output power-down control ? output enable (disabled by default) ? output synchronization ? output mode control ? output divider functionality see the clock distribution section for more information. generate the output clock if register 0x0500[1:0] is programmed for automatic clock distribution synchronization via the dpll phase or frequency lock, the synthesized output signal appears at the clock distribution outputs. otherwise, set and then clear the soft sync clock distribution bit (register 0x0a02, bit 1) or use a multifunction pin input (if programmed for use) to generate a clock distribution sync pulse, which causes the synthesized output signal to appear at the clock distribution outputs. program the multifunction pins (optional) this step is required only if the user intends to use any of the multifunction pins for status or control. the multifunction pin parameters are at register 0x0200 to register 0x0208. program the irq functionality (optional) this step is required only if the user intends to use the irq feature. the irq monitor registers are at register 0x0d02 to register 0x0d09. if the desired bits in the irq mask registers at register 0x020a to register 0x020f are set high, the appropriate irq monitor bit at register 0x0d02 to register 0x0d07 is set high when the indicated event occurs. individual irq events are cleared by using the irq clearing registers at register 0x0a04 to register 0x0a09 or by setting the clear all irqs bit (register 0x0a03[1]) to 1b. the default values of the irq mask registers are such that interrupts are not generated. the irq pin mode default is open- drain nmos. program the watchdog timer (optional) this step is required only if the user intends to use the watchdog timer. the watchdog timer control is in register 0x0210 and register 0x0211 and is disabled by default. the watchdog timer is useful for generating an irq after a fixed amount of time. the timer is reset by setting the clear watchdog timer bit (register 0x0a03[0]) to 1b.
data sheet AD9557 rev. a | page 27 of 92 p rogram the digital phas e-locked loop (dpll) t he dpll parameters reside in register 0x0300 to register 0x032e. they include the following: ? free run frequency ? dpll pull-in range limits ? dpll closed-loop phase offset ? phase slew control (for hitless reference switching) ? tuning word history control (for holdover operation) p rogram the reference inputs t he reference input parameters reside in register 0x0600 to register 0x0602. see the reference clock input section for details on programming these functions. they include the following: ? reference power-down ? reference logic family ? reference priority p rogram the reference profiles t he reference profile parameters reside in register 0x0700 to register 0x0766. the AD9557 evaluation software contains a wizard that calculates these values based on the users input frequency. see the reference profiles section for details on programming these functions. they include the following: ? reference period ? reference period tolerance ? reference validation timer ? selection of high phase margin, loop filter coefficients ? dpll loop bandwidth ? reference prescaler (r divider) ? feedback dividers (n1, n2, n3, frac1, and mod1) ? phase and frequency lock detector controls generate the reference acquisition after the registers are programmed, the user can clear the user freerun bit (register 0x0a01[5]) and issue an i/o update, using register 0x0005[0] to invoke all of the register settings that are programmed up to this point. after the registers are programmed, the dpll locks to the first available reference that has the highest priority.
AD9557 data sheet rev. a | page 28 of 92 theory of operation spi/i 2 c serial port eeprom ref monitoring automatic switching n1 n2 n3 2 m0 out0 out0 out1 out1 10-bit integer dividers max 1.25ghz m1 2 2 lf pfd/cp rf divider 1 3 to 11 x o or xt a l xo frequencies 10mhz to 180mhz xtal: 10mhz to 50mhz rf divider 2 3 to 11 f out = 360khz to 1.25ghz integer divider output pll (apll) frac1/ mod1 17-bit integer 24b/24b resolution digital pll (dpll) 2 register space 2khz to 1.25ghz r divider (20-bit) sync reset pincontrol m0 m1 m2 m3 irq spi/i 2 c digital loop filter tuning word clamp and history free run tw lf_vco2 pfd/cp lf vco2 3.35ghz to 4.05ghz dpfd 30-bit nco rom and fsm multifunction i/o pins (control and status readback) system clock pll 2 AD9557 refa refa refb refb 09197-135 figure 35. detailed block diagram overview the AD9557 provides clocking outputs that are directly related in phase and frequency to the selected (active) reference, but with jitter characteristics that are governed by the system clock, the dco, and the output pll (apll). the AD9557 supports up to two reference inputs and input frequencies ranging from 2 khz to 1250 mhz. the core of this product is a digital phase-locked loop (dpll). the dpll has a programmable digital loop filter that greatly reduces jitter that is transferred from the active reference to the output. the AD9557 supports both manual and automatic holdover. while in holdover, the AD9557 continues to provide an output as long as the system clock is present. the holdover output frequency is a time average of the output frequency history just prior to the transition to the holdover condition. the device offers manual and automatic reference switchover capability if the active reference is degraded or fails completely. the AD9557 also has adaptive clocking capability that allows the dpll divider ratios to be changed while the dpll is locked. the AD9557 has a system clock multiplier, a digital pll (dpll), and an analog pll (apll). the input signal goes first to the dpll, which performs the jitter cleaning and most of the frequency translation. the dpll features a 30-bit digitally controlled oscillator (dco) output that generates a signal in the 175 mhz to 200 mhz range. the dpll output goes to an analog integer-n pll (apll), which multiplies the signal up to the 3.35 ghz to 4.05 ghz range. that signal is then sent to the clock distribution section, which has two divide-by-3 to divide-by-11 rf dividers that are cascaded with 10-bit in teger (divide-by-1 to divide-by- 1024) channel dividers. the xoa and xob inputs provide the input for the system clock. these pins accept a reference clock in the 10 mhz to 600 mhz range, or a 10 mhz to 50 mhz crystal connected directly across the xoa and xob inputs. the system clock provides the clocks to the frequency monitors, the dpll, and internal switching logic. the AD9557 has two differential output drivers. each driver has a dedicated 10-bit programmable post divider. each differential driver is programmable either as a single differential or dual single-ended cmos output. the clock distribution section operates at up to 1250 mhz. in differential mode, the output drivers run on a 1.8 v power supply to offer very high performance with minimal power consumption. there are two differential modes: lvds and 1.8 v hstl. in 1.8 v hstl mode, the voltage swing is compatible with lvpecl. if lvpecl signal levels are required, the designer can ac-couple the AD9557 output and use thevenin-equivalent termination at the destination to drive the lvpecl inputs. in single-ended mode, each differential output driver can operate as two single-ended cmos outputs. out0 supports either 1.8 v or 3.3 v cmos operation. out1 supports only 1.8 v operation.
data sheet AD9557 rev. a | page 29 of 92 reference clock inputs two pairs of pins provide access to the reference clock receivers. to accommodate input signals with slow rising and falling edges, both the differential and single-ended input receivers employ hysteresis. hysteresis also ensures that a disconnected or floating input does not cause the receiver to oscillate. when configured for differential operation, the input receivers accommodate either ac- or dc-coupled input signals. the input receivers are capable of accepting dc-coupled lvds and 2.5 v and 3.3 v lvpecl signals. the receiver is internally dc biased to handle ac-coupled operation, but there is no internal 50 or 100 termination. when configured for single-ended operation, the input receivers exhibit a pull-down load of 45 k (typical). three user-programmable threshold voltage ranges are available for each single-ended receiver. reference monitors the accuracy of the input reference monitors depends on a known and accurate system clock period. therefore, the functioning of the reference monitors is not operable until the system clock is stable. reference period monitor each reference input has a dedicated monitor that repeatedly measures the reference period. the AD9557 uses the reference period measurements to determine the validity of the reference based on a set of user-provided parameters in the profile register area of the register map. the monitor works by comparing the measured period of a particular reference input with the parameters stored in the profile register assigned to that same reference input. the parameters include the reference period, an inner tolerance, and an outer tolerance. a 40-bit number defines the reference period in units of femtoseconds (fs). the 40-bit range allows for a reference period entry of up to 1.1 ms. a 20-bit number defines the inner and outer tolerances. the value stored in the register is the reciprocal of the tolerance specification. for example, a tolerance specification of 50 ppm yields a register value of 1/(50 ppm) = 1/0.000050 = 20,000 (0x04e20). the use of two tolerance values provides hysteresis for the monitor decision logic. the inner tolerance applies to a previously faulted reference and specifies the largest period tolerance that a previously faulted reference can exhibit before it qualifies as nonfaulted. the outer tolerance applies to an already nonfaulted reference. it specifies the largest period tolerance that a nonfaulted reference can exhibit before being faulted. to produce decision hysteresis, the inner tolerance must be less than the outer tolerance. that is, a faulted reference must meet tighter requirements to become nonfaulted than a nonfaulted reference must meet to become faulted. reference validation timer each reference input has a dedicated validation timer. the validation timer establishes the amount of time that a previously faulted reference must remain unfaulted before the AD9557 declares it valid. the timeout period of the validation timer is programmable via a 16-bit register. the 16-bit number stored in the validation register represents units of milliseconds (ms), which yields a maximum timeout period of 65,535 ms. it is possible to disable the validation timer by programming the validation timer to 0b. with the validation timer disabled, the user must validate a reference manually via the manual reference validation override controls register (address 0x0a0b). reference validation override control the user also has the ability to override the reference validation logic and can either force an invalid reference to be treated as valid, or force a valid reference to be treated as an invalid reference. these controls are in register 0x0a0b to register 0x0a0d. reference profiles the AD9557 has an independent profile for each reference input. a profile consists of a set of device parameters such as the r divider and n divider, among others. the profiles allow the user to prescribe the specific device functionality that should take effect when one of the input references becomes the active reference. the AD9557 evaluation software includes a frequency planning wizard that can configure the profile parameters, given the input and output frequencies. the user should not change a profile that is currently in use because unpredictable behavior may result. the user can either select free run or holdover mode, or invalidate the reference input prior to changing it. reference switchover a n attractive feature of the AD9557 is its versatile reference switchover capability. the flexibility of the reference switchover functionality resides in a sophisticated prioritization algorithm that is coupled with register-based controls. this scheme provides the user with maximum control over the state machine that handles reference switchover. t he main reference switchover control resides in the loop mode register (address 0x0a01). the ref switchover mode bits (register 0x0a01, bits[4:2]) allow the user to select one of the five operating modes of the reference switchover state machine, as follows: ? automatic revertive mode ? automatic non-revertive mode ? manual with automatic fallback mode ? manual with holdover mode ? full manual mode (without auto-holdover)
AD9557 data sheet rev. a | page 30 of 92 i n the automatic modes, a fully automatic priority-based algorithm selects which reference is the active reference. when programmed for an automatic mode, the device chooses the highest priority valid reference. when both references have the same priority, refa gets preference over refb. however, the reference position is used only as a tie-breaker and does not initiate a reference switch. t he following list gives an overview of the five operating modes: ? automatic revertive mode. the device selects the highest priority valid reference and switches to a higher priority reference if it becomes available, even if the reference in use is still valid. in this mode, the user reference is ignored. ? automatic non-revertive mode. the device stays with the currently selected reference as long as it is valid, even if a higher priority reference becomes available. the user reference is ignored in this mode. ? manual with automatic fallback mode. the device uses the user reference for as long as it is valid. if it becomes invalid, the reference input with the highest priority is chosen in accordance with the priority-based algorithm. ? manual with holdover mode. the user reference is the active reference until it becomes invalid. at that point, the device automatically goes into holdover. ? manual mode without holdover. the user reference is the active reference, regardless of whether or not it is valid. the user also has the option to force the device directly into holdover or free run operation via the user holdover and user freerun bits. in free run mode, the free run frequency tuning word register defines the free run output frequency. in holdover mode, the output frequency depends on the holdover control settings (see the holdover section). phase build-out reference switching the AD9557 supports phase build-out reference switching, which is the term given to a reference switchover that completely masks any phase difference between the previous reference and the new reference. that is, there is virtually no phase change detectable at the output when a phase build-out switchover occurs. digital pll (dpll) core dpll overview a diagram of the dpll core of the AD9557 appears in figure 36 . the phase/frequency detector, feedback path, lock detectors, phase offset, and phase slew rate limiting that comprise this second generation dpll are all digital implementations. the start of the dpll signal chain is the reference signal, f r , which is the frequency of the reference input. a reference prescaler reduces the frequency of this signal by an integer factor, r + 1, where r is the 20-bit value stored in the appropriate profile register and 0 r 1,048,575. therefore, the frequency at the output of the r divider (or the input to the time-to-digital converter (tdc)) is 1 + = r f f r tdc digital loop filter n1 r divider (20-bit) 24-bit/24-bit resolution frac1/ mod1 17-bit integer tuning word clamp and history 2 free run tw + 30-bit nco dpfd system clock from apll from ref input mux to apll 09197-136 f igure 36. digital pll core a tdc samples the output of the r divider. the tdc/pfd produces a time series of digital words and delivers them to the digital loop filter. the digital loop filter offers the following advantages: ? determination of the filter response by numeric coefficients rather than by discrete component values ? the absence of analog components (r/l/c), which eliminates tolerance variations due to aging ? the absence of thermal noise associated with analog components ? the absence of control node leakage current associated with analog components (a source of reference feed- through spurs in the output spectrum of a traditional analog pll) the digital loop filter produces a time series of digital words at its output and delivers them to the frequency tuning input of a sigma-delta (-) modulator (sdm). the digital words from the loop filter steer the dco frequency toward frequency and phase lock with the input signal (f tdc ). the dpll includes a feedback divider that causes the digital loop to operate at an integer-plus-fractional multiple. the output of the dpll is ? ? ? ? ? ? ++= mod1 frac1 n1f f tdc dpll out )1( _ where n1 is the 17-bit value stored in the appropriate profile registers (register 0x0715 to register 0x0717 for refa). frac1 and mod1 are the 24-bit numerators and denominators of the fractional feedback divider block. the fractional portion of the feedback divider can be bypassed by setting frac1 to 0, but mod1 should never be 0. the dpll output frequency is usually 175 mhz to 200 mhz for optimal performance. tdc/pfd the phase-frequency detector (pfd) is an all-digital block. it compares the digital output from the tdc (which relates to the active reference edge) with the digital word from the feedback block. it uses a digital code pump and digital integrator (rather than a conventional charge pump and capacitor) to generate the error signal that steers the dco frequency toward phase lock.
data sheet AD9557 rev. a | page 31 of 92 programmable digital loop filter the AD9557 loop filter is a third-order digital iir filter that is analogous to the third-order analog loop shown in figure 37 . c 3 c 2 c 1 r 2 r 3 09197-015 figure 37. third order analog loop filter the AD9557 loop filter block features a simplified architecture in which the user enters the desired loop characteristics directly into the profile registers. this architecture makes the calculation of individual coefficients unnecessary in most cases, while still offering complete flexibility. the AD9557 has two preset digital loop filters: high (88.5) phase margin and normal (70) phase margin. the loop filter coefficients are stored in register 0x0317 to register 0x0322 for high phase margin and register 0x0323 to register 0x032e for normal phase margin. the high phase margin loop filter is intended for applications in which the closed-loop transfer function must not have greater than 0.1 db of peaking. bit 0 of register 0x070e selects which filter is used for profile a, and bit 0 of 0x074e selects the filter for profile b. the loop bandwidth for profile a is set in register 0x070f to register 0x0711, and the loop bandwidth for profile b is set in register 0x074f to register 0x0751. the two preset conditions should cover all of the intended applications for the AD9557 . for special cases where these conditions must be modified, the tools for calculating these coefficients are available by contacting analog devices directly. dpll digitally controlled osc illator free run frequency the AD9557 uses a - modulator (sdm) as a digitally controlled oscillator (dco). the dco free run frequency can be calculated by 30 _ 2 0 8 2 ftw f f sys freerun dco + = where ftw0 is the value in register 0x0300 to register 0x0303, and f sys is the system clock frequency. see the system clock section for information on calculating the system clock frequency. adaptive clocking the AD9557 can support adaptive clocking applications such as asynchronous mapping and demapping. in these applications, the output frequency can be dynamically adjusted by up to 100 ppm from the nominal output frequency without manually breaking the dpll loop and reprogramming the part. this function is supported for refa only, not refb. t he following registers are used in this function: ? register 0x0717 (dpll n1 divider) ? register 0x0718 to register 0x071a (dpll frac1 divider) ? register 0x071b to register 0x071d (dpll mod1 divider) writing to these registers requires an i/o update by writing 0x01 to register 0x0005 before the new values take effect. to make small adjustments to the output frequency, the user can vary the frac1 and issue an i/o update. the advantage to using only frac1 to adjust the output frequency is that the dpll does not briefly enter holdover. therefore, the frac1 bit can be updated as fast as the phase detector frequency of the dpll. writing to the n1 and mod1 dividers allows for larger changes to the output frequency. when the AD9557 detects that the n1 or mod1 values have changed, it automatically enters and exits holdover for a brief instant without any disturbance in the output frequency. this limits how quickly the output frequency can be adapted. it is important to realize that the amount of frequency adjustment is limited to 100 ppm before the output pll (apll) needs a recalibration. variations that are larger than 100 ppm are possible, but the ability of the AD9557 to maintain lock over temperature extremes may be compromised. it is also important to remember that the rate of change in output frequency depends on the dpll loop bandwidth. dpll phase lock detector the dpll contains an all-digital phase lock detector. the user controls the threshold sensitivity and hysteresis of the phase detector via the profile registers. the phase lock detector behaves in a manner analogous to water in a tub (see figure 38 ). the total capacity of the tub is 4096 units with ?2048 denoting empty, 0 denoting the 50% point, and +2048 denoting full. the tub also has a safeguard to prevent overflow. furthermore, the tub has a low water mark at ?1024 and a high water mark at +1024. to change the water level, the user adds water with a fill bucket or removes water with a drain bucket. the user specifies the size of the fill and drain buckets via the 8-bit fill rate and drain rate values in the profile registers. 0 2048 ?2048 1024 ?1024 lock level unlock level locked unlocked previous state fill rate drain rate 09197-017 figure 38. lock detector diagram the water level in the tub is what the lock detector uses to deter- mine the lock and unlock conditions. when the water level is below the low water mark (?1024), the detector indicates an unlock condition. conversely, whenever the water level is above the high water mark (+1024), the detect or indicates a lock condition. when the water level is between the marks, the detector holds its last condition. this concept appears graphically in figure 38 , with an overlay of an example of the instantaneous water level (vertical) vs. time (horizontal) and the resulting lock/unlock states.
AD9557 data sheet rev. a | page 32 of 92 during any given pfd cycle, the detector either adds water with the fill bucket or removes water with the drain bucket (one or the other but not both). the decision of whether to add or remove water depends on the threshold level specified by the user. the phase lock threshold value is a 16-bit number stored in the profile registers and is expressed in picoseconds (ps). thus, the phase lock threshold extends from 0 ns to 65.535 ns and represents the magnitude of the phase error at the output of the pfd. the phase lock detector compares each phase error sample at the output of the pfd to the programmed phase threshold value. if the absolute value of the phase error sample is less than or equal to the programmed phase threshold value, then the detector control logic dumps one fill bucket into the tub. otherwise, it removes one drain bucket from the tub. note that it is not the polarity of the phase error sample, but its magnitude relative to the phase threshold value, that determines whether to fill or drain. if more filling is taking place than draining, the water level in the tub eventually rises above the high water mark (+1024), which causes the phase lock detector to indicate lock. if more draining is taking place than filling, then the water level in the tub eventually falls below the low water mark (?1024), which causes the phase lock detector to indicate unlock. the ability to specify the threshold level, fill rate, and drain rate enables the user to tailor the operation of the phase lock detector to the statistics of the timing jitter associated with the input reference signal. note that whenever the AD9557 enters the free run or holdover mode, the dpll phase lock detector indicates an unlocked state. however, when the AD9557 performs a reference switch, the lock detector state prior to the switch is preserved during the transition period. dpll frequency lo ck detector the operation of the frequency lock detector is identical to that of the phase lock detector. the only difference is that the fill or drain decision is based on the period deviation between the reference and feedback signals of the dpll instead of the phase error at the output of the pfd. the frequency lock detector uses a 24-bit frequency threshold register specified in units of picoseconds (ps). thus, the frequency threshold value extends from 0 s to 16.777215 s. it represents the magnitude of the difference in period between the reference and feedback signals at the input to the dpll. for example, if the reference signal is 1.25 mhz and the feedback signal is 1.38 mhz, then the period difference is approximately 75.36 ns (|1/1,250,000 ? 1/1,380,000| 75.36 ns). frequency clamp the AD9557 dpll features a digital tuning word clamp that ensures that the dpll output frequency stays within a defined range. this feature is very useful to eliminate undesirable behavior in cases where the reference input clocks may be unpredictable. the tuning word clamp is also useful to guarantee that the apll never loses lock, by ensuring that the apll vco frequency stays within its tuning range. frequency tuning word history the AD9557 has the ability to track the history of the tuning word samples generated by the dpll digital loop filter output. it does so by periodically computing the average tuning word value over a user-specified interval. this average tuning word is used during holdover mode to maintain the average frequency when no input references are present. loop control state machine switchover switchover occurs when the loop controller switches directly from one input reference to another. the AD9557 handles a reference switchover by briefly entering holdover mode, loading the new dpll parameters, and then immediately recovering. during the switchover event, however, the AD9557 preserves the status of the lock detectors to avoid phantom unlock indications. holdover the holdover state of the dpll is typically used when none of the input references are present, although the user can also manually engage holdover mode. in holdover mode, the output frequency remains constant. the accuracy of the AD9557 in holdover mode is dependent on the device programming and availability of tuning word history. recovery from holdover when in holdover mode and a valid reference becomes available, the device exits holdover operation. the loop state machine restores the dpll to closed-loop operation, locks to the selected reference, and sequences the recovery of all the loop parameters based on the profile settings for the active reference. note that, if the user holdover bit is set, the device does not automatically exit holdover when a valid reference is available. however, automatic recovery can occur after clearing the user holdover bit (bit 6 in register 0x0a01).
data sheet AD9557 rev. a | page 33 of 92 system clock (sysclk) system clock inputs functional description the sysclk circuit provides a low jitter, stable, high frequency clock for use by the rest of the chip. the xoa and xob pins connect to the internal sysclk multiplier. the sysclk multiplier can synthesize the system clock by connecting a crystal resonator across the xoa and xob input pins or by connecting a low frequency clock source. the optimal signal for the system clock input is either a crystal in the 50 mhz range or an ac-coupled square wave with a 1 v p-p amplitude. system clock period for the AD9557 to accurately measure the frequency of incoming reference signals, the user must enter the system clock period into the nominal system clock period registers (register 0x0103 to register 0x0105). the sysclk period is entered in units of nanoseconds (ns). system clock details there are two internal paths for the sysclk input signal: low frequency non-xtal (lf) and crystal resonator (xtal). using a tcxo for the system clock is a common use for the lf path. applications requiring dpll loop bandwidths of less than 50 hz or high stability in holdover require a tcxo. as an alternative to the 49.152 mhz crystal for these applications, the AD9557 reference design uses a 19.2 mhz tcxo, which offers excellent holdover stability and a good combination of low jitter and low spurious content. the 1.8 v differential receiver connected to the xoa and xob pins is self-biased to a dc level of ~1 v, and ac coupling is strongly recommended. when a 3.3 v cmos oscillator is in use, it is important to use a voltage divider to reduce the input high voltage to a maximum of 1.8 v. see figure 34 for details on connecting a 3.3 v cmos tcxo to the system clock input. the non-xtal input path permits the user to provide an lvpecl, lvds, 1.8 v cmos, or sinusoidal low frequency clock for multiplication by the integrated sysclk pll. the lf path handles input frequencies from 3.5 mhz up to 100 mhz. however, when using a sinusoidal input signal, it is best to use a frequency that is in excess of 20 mhz. otherwise, the resulting low slew rate can lead to substandard noise performance. note that the non-xtal path includes an optional 2 frequency multiplier to double the rate at the input to the sysclk pll and potentially reduce the pll in-band noise. however, to avoid exceeding the maximum pfd rate of 150 mhz, the 2 frequency multiplier is only for input frequencies that are below 75 mhz. t he non-xtal path also includes an input divider (m) that is programmable for divide-by-1, -2, -4, or -8. the purpose of the divider is to limit the frequency at the input to the pll to less than 150 mhz (the maximum pfd rate). t he xtal path enables the connection of a crystal resonator (typically 10 mhz to 50 mhz) across the xoa and xob pins. an internal amplifier provides the negative resistance required to induce oscillation. the internal amplifier expects an at cut, fundamental mode crystal with a maximum motional resistance of 100 . the following crystals, listed in alphabetical order, may meet these criteria. analog devices, inc., does not guarantee their operation with the AD9557 , nor does analog devices endorse one crystal supplier over another. the AD9557 reference design uses a 49.152 mhz crystal, which is high performance, low spurious content, and readily available. ? avx/kyocera cx3225sb ? ecs ecx-32 ? epson/toyocom tsx-3225 ? fox fx3225bs ? ndk nx3225sa ? siward sx-3225 ? suntsu scm10b48-49.152 mhz system clock multiplier the sysclk pll multiplier is an integer-n design with an integrated vco. it provides a means to convert a low frequency clock input to the desired system clock frequency, f sys (750 mhz to 805 mhz). the sysclk pll multiplier accepts input signals of between 3.5 mhz and 600 mhz, but frequencies that are in excess of 150 mhz require the system clock p-divider to ensure compliance with the maximum pfd rate (150 mhz). the pll contains a feedback divider (n) that is programmable for divide values between 4 and 255. pdiv sysclk ndiv sysclk ff osc sys _ _ = where: f osc is the frequency at the xoa and xob pins. sysclk_ndiv is the value stored in register 0x0100. sysclk_pdiv is the system clock p divider that is determined by the setting of register 0x0101[2:1]. if the system clock doubler is used, the value of sysclk_ndiv should be half of its original value. the system clock multiplier features a simple lock detector that compares the time difference between the reference and feedback edges. the most common cause of the sysclk multiplier not locking is a non-50% duty cycle at the sysclk input while the system clock doubler is enabled.
AD9557 data sheet rev. a | page 34 of 92 system clock stability timer because the reference monitors depend on the system clock being at a known frequency, it is important that the system clock be stable before activating the monitors. at initial power- up, the system clock status is not known, and, therefore, it is reported as being unstable. after the part has been programmed, the system clock pll (if enabled) eventually locks. when a stable operating condition is detected, a timer is run for the duration that is stored in the system clock stability period registers. if, at any time during this waiting period, the condition is violated, the timer is reset and halted until a stable condition is reestablished. after the specified period elapses, the AD9557 reports the system clock as stable.
data sheet AD9557 rev. a | page 35 of 92 output pll (apll) a diagram of the output pll (apll) is shown in figure 39 . lf cap vco2 3.35ghz to 4.05ghz pfd from dpll to clock distribution lf cp integer divider output pll divider (apll) n2 09197-138 figure 39. output pll block diagram the apll provides the frequency upconversion from the dpll output to the 3.35 ghz to 4.05 ghz range, while also providing noise filtering on the dpll output. the apll reference input is the output of the dpll. the feedback divider is an integer divider. the loop filter is partially integrated with the one external 6.8 nf capacitor. the nominal loop bandwidth for this pll is 250 khz, with 68 degrees of phase margin. the frequency wizard that is included in the evaluation software configures the apll, and the user should not need to make changes to the apll settings. however, there may be special cases where the user may wish to adjust the apll loop bandwidth to meet a specific phase noise requirement. the easiest way to change the apll loop bw is to adjust the apll charge pump current in register 0x0400. there is sufficient stability (680 of phase margin) in the apll default settings to permit a broad range of adjustment without causing the apll to be unstable. the user should contact analog devices directly if more detail is needed. calibration of the apll must be performed at startup and whenever the nominal input frequency to the apll changes by more than 100 ppm, although the apll maintains lock over voltage and temperature extremes without recalibration. calibration centers the dc operating voltage at the input to the apll vco. apll calibration at startup can be accomplished during initial register loading by following the instructions in the device register programming using a register setup file section of this datasheet. t o recalibrate the apll vco after the chip has been running, the user should first input the new settings (if any). ensure that the system clock is still locked and stable, and that the dpll is in free run mode with the free run tuning word set to the same output frequency that is used when the dpll is locked. u se the following steps to calibrate the apll vco: 1. ensure that the system clock is locked and stable. 2. ensure that the dpll is in user free run mode (register 0x0a01[5] = 1b), and the free run tuning word is set. 3. write register 0x0405 = 0x20. 4. write register 0x0005 = 0x01. 5. write register 0x0405 = 0x21. 6. write register 0x0005 = 0x01. 7. monitor the apll status using bit 2 in register 0x0d01.
AD9557 data sheet rev. a | page 36 of 92 clock distribution rf div 1 3 to 11 from dpll (3.35ghz to 4.05ghz) rf div 2 3 to 11 chip reset sync m0 m1 10-bit integer 360khz to 1250mhz 10-bit integer channel sync block max 1.25ghz max 1.25ghz channel sync (to m0 and m1) out0 out0 out1 out1 09197-139 figure 40. clock distribution block diagram a diagram of the clock distribution block appears in figure 40 . clock dividers the channel divider blocks, m0 and m1, are 10-bit integer dividers with a divide range of 1 to 1023. the channel divider block contains duty cycle correction that guarantees 50% duty cycle for both even and odd divide ratios. output power-down the output drivers can be individually powered down. output enable each of the output channels offers independent control of enable/ disable functionality via the distribution enable register. the distribution outputs use synchronization logic to control enable/disable activity to avoid the production of runt pulses and ensure that outputs with the same divide ratios become active/inactive in unison. ou tput mode th e user has independent control of the operating mode of each of the four output channels via the output clock distribution registers (address 0x0500 to address 0x0509). the operating mode control includes ? l ogic family and pin functionality ? output drive strength ? output polarity ? divide ratio ? phase of each output channel out0 provides 3.3 v cmos, in addition to 1.8 v cmos modes. out1 has 1.8 v cmos, lvds, and hstl modes. all cmos drivers feature a cmos drive strength that allows the user to choose between a strong, high performance cmos driver, or a lower power setting with less emi and crosstalk. the best setting is application dependent. for applications where lvpecl levels are required, the user should choose the hstl mode, and ac-couple the output signal. see the input/output termination recommendations section for recommended termination schemes. clock distribution synchronization divider synchronization t he dividers in the clock distribution channels can be synchronized with each other. a t power-up, the clock dividers are held static until a sync signal is initiated by the channel sync block. the following are possible sources of a sync signal, and these settings are found in register 0x0500: ? direct sync via bit 2 of register 0x0500 ? direct sync via a sync op code (0xa1) in the eeprom storage sequence during eeprom loading ? dpll phase or frequency lock ? a rising edge of the selected reference input ? the sync pin ? a multifunction pin configured for the sync signal the apll lock detect signal gates the sync signal from the channel sync block shown in figure 40 . the channel dividers receive a sync signal from the channel sync block only if the apll is calibrated and locked, unless the apll locked controlled sync bit (register 0x0405[3]) is set. a channel can be programmed to ignore the sync function by setting the mask channel 1 sync and mask channel 0 sync bits (bits[5:4]) in register 0x0500. when programmed to ignore the sync, the channel ignores both the user initiated sync signal and the zero delay initiated sync signals, and the channel divider starts toggling, provided that the apll is calibrated and locked, or if bit 3 (apll locked controlled sync bit), register 0x0405, is set. i f the output sync function is to be controlled using an m pin, use the following steps: 1. first, enable the m pins by writing register 0x0200 = 0x01. 2. issue an i/o update (register 0x0005 = 0x01). 3. set the appropriate m pin function. if this process is not followed, a sync pulse is issued automatically.
data sheet AD9557 rev. a | page 37 of 92 status and control multifunction pins (m3 to m0) the AD9557 has four digital cmos i/o pins (m3 to m0) that are configurable for a variety of uses. to use these functions, the user must enable them by writing a 0x01 to register 0x0200. the function of these pins is programmable via the register map. each pin can control or monitor an assortment of internal functions, based on the contents of register 0x0201 to register 0x0204. to monitor an internal function with a multifunction pin, write a logic 1 to the most significant bit of the register associated with the desired multifunction pin. the value of the seven least significant bits of the register defines the control function, as shown in table 124 . to control an internal function with a multifunction pin, write a logic 0 to the most significant bit of the register associated with the desired multifunction pin. the monitored function depends on the value of the seven least significant bits of the register, as shown in table 125 . if more than one multifunction pin operates on the same control signal, then internal priority logic ensures that only one multi- function pin serves as the signal source. the selected pin is the one with the lowest numeric suffix. for example, if both m0 and m3 operate on the same control signal, m0 is used as the signal source and the redundant pins are ignored. a t power-up, the multifunction pins can force the device into certain configurations, as defined in the initial pin programming section. this functionality, however, is valid only during power- up or following a reset, after which the pins can be reconfigured via the serial programming port or via the eeprom. i f the output sync function is to be controlled using an m pin, 1. first, enable the m pins by writing register 0x0200 = 0x01. 2. issue an i/o update (register 0x0005 = 0x01). 3. set the appropriate m pin function. if this process is not followed, a sync pulse is issued automatically. irq pin the AD9557 has a dedicated interrupt request (irq) pin. bits[1:0] of the irq pin output mode register (register 0x0209) control how the irq pin asserts an interrupt, based on the value of the two bits, as follows: 00the irq pin is high impedance when deasserted and active low when asserted and requires an external pull-up resistor. 01the irq pin is high impedance when deasserted and active high when asserted and requires an external pull-down resistor. 10the irq pin is logic 0 when deasserted and logic 1 when asserted. 11the irq pin is logic 1 when deasserted and logic 0 when asserted. (this is the default operating mode.) the AD9557 asserts the irq pin when any bit in the irq monitor register (address 0x0d02 to address 0x0d07) is a logic 1. each bit in this register is associated with an internal function that is capable of producing an interrupt. furthermore, each bit of the irq monitor register is the result of a logical and of the associated internal interrupt signal and the corresponding bit in the irq mask register (address 0x020a to address 0x020e). that is, the bits in the irq mask register have a one-to-one correspondence with the bits in the irq monitor register. when an internal function produces an interrupt signal and the associated irq mask bit is set, then the corresponding bit in the irq monitor register is set. the user should be aware that clearing a bit in the irq mask register removes only the mask associated with the internal interrupt signal. it does not clear the corresponding bit in the irq monitor register. the irq pin is the result of a logical or of all the irq monitor register bits. thus, the AD9557 asserts the irq pin as long as any irq monitor register bit is a logic 1. note that it is possible to have multiple bits set in the irq monitor register. therefore, when the AD9557 asserts the irq pin, it may indicate an interrupt from several different internal functions. the irq monitor register provides the user with a means to interrogate the AD9557 to determine which internal function produced the interrupt. typically, when the irq pin is asserted, the user interrogates the irq monitor register to identify the source of the interrupt request. after servicing an indicated interrupt, the user should clear the associated irq monitor register bit via the irq clearing register (address 0x0a04 to address 0x0a09). the bits in the irq clearing register have a one-to-one correspondence with the bits in the irq monitor register. note that the irq clearing register is autoclearing. the irq pin remains asserted until the user clears all of the bits in the irq monitor register that indicate an interrupt. it is also possible to collectively clear all of the irq monitor register bits by setting the clear all irqs bit in the reset function register (register 0x0a03, bit 1). note that this is an autoclearing bit. setting this bit results in deassertion of the irq pin. alternatively, the user can program any of the multifunction pins to clear all irqs. this allows the user to clear all irqs by means of a hardware pin rather than by using a serial i/o port operation.
AD9557 data sheet rev. a | page 38 of 92 watchdog timer the watchdog timer is a general-purpose programmable timer. to set the timeout period, the user writes to the 16-bit watchdog timer register (address 0x0x0210 and address 0x0211). a value of 0b in this register disables the timer. a nonzero value sets the timeout period in milliseconds (ms), giving the watchdog timer a range of 1 ms to 65.535 sec. the relative accuracy of the timer is approximately 0.1% with an uncertainty of 0.5 ms. if enabled, the timer runs continuously and generates a timeout event whenever the timeout period expires. the user has access to the watchdog timer status via the irq mechanism and the multifunction pins (m0 to m3). in the case of the multifunction pins, the timeout event of the watchdog timer is a pulse that lasts 32 system clock periods. there are two ways to reset the watchdog timer (thereby preventing it from causing a timeout event). the first is by writing a logic 1 to the autoclearing clear watchdog bit in the reset functions register (register 0x0a03, bit 0). alternatively, the user can program any of the multifunction pins to reset the watchdog timer. this allows the user to reset the timer by means of a hardware pin rather than by using a serial i/o port operation. eeprom eeprom overview the AD9557 contains an integrated 2048-byte, electrically erasable, programmable read-only memory (eeprom). the AD9557 can be configured to perform a download at power-up via the multifunction pins (m2 to m3), but uploads and downloads can also be performed on demand via the eeprom control registers (address 0x0e00 to address 0x0e03). the eeprom provides the ability to upload and download configuration settings to and from the register map. figure 41 shows a functional diagram of the eeprom. register 0x0e10 to register 0x0e3f represent a 53-byte eeprom storage sequence area (referred to as the scratch pad in this section) that enables the user to store a sequence of instructions for transferring data to the eeprom from the device settings portion of the register map. note that the default values for these registers provide a sample sequence for saving/retrieving all of the AD9557 eeprom-accessible registers. figure 41 shows the connectivity between the eeprom and the controller that manages data transfer between the eeprom and the register map. the controller oversees the process of transferring eeprom data to and from the register map. there are two modes of operation handled by the controller: saving data to the eeprom (upload mode) or retrieving data from the eeprom (download mode). in either case, the controller relies on a specific instruction set. eeprom (0x000 to 0x7ff) data data d a t a eeprom address pointer m3 m2 register map device settings (0x0a00 to 0x0a0d) scratch pad (0x0e10 to 0x0e3f) serial input/output port condition (0e01 [3:0]) scratch pad address pointer device settings address pointer eeprom controller 09197-024 figure 41. eeprom functional diagram
data sheet AD9557 rev. a | page 39 of 92 table 21. eeprom controller instruction set instruction value (hex) instruction type bytes required description 0x00 to 0x7f data 3 a data instruction tells the controller to transfer data to or from the device settings part of the register map. a data instru ction requires two additional bytes that, together, indicate a starting address in the re gister map. encoded in the data instruction is the number of bytes to transfer, which is one more than the instruction value. 0x80 i/o update 1 when the controller encounters this instruction while downloading from the eeprom, it issues a soft i/o update. 0xa0 calibrate 1 when the controller encounters this instruction while downloading from the eeprom, it initiates a system clock calibration sequence. 0xa1 distribution sync 1 when the controller encounters this instruction while downloading from the eeprom, it issues a sync pulse to the output distribution synchronization. 0xb0 to 0xcf condition 1 b1 to cf are condition instructions and correspond to condition 1 through condition 31, respectively. b0 is the nu ll condition instruction. see the eeprom conditional processing section for details. 0xfe pause 1 when the controller encounters this inst ruction in the eeprom storage sequence area while uploading to the eeprom, it ho lds both the register area address pointer and the eeprom address pointer at its last va lue. this allows storage of more than one instruction sequence in the eeprom. no te that the controller does not copy this instruction to the eeprom during upload. 0xff end 1 when the controller encounters this inst ruction in the eeprom storage sequence area while uploading to the eeprom, it rese ts both the register area address pointer and the eeprom address pointer and then enters an idle state. when the controller encounters this instruction while downloading from the eeprom, it resets the eeprom address pointer and then enters an idle state. eeprom instructions table 21 lists the eeprom controller instruction set. the controller recognizes all instruction types, whether it is in upload or download mode, except for the pause instruction, which is recognized only in upload mode. the i/o update, calibrate, distribution sync, and end instruc- tions are mostly self-explanatory. the others, however, warrant further detail, as described in the following paragraphs. data instructions are those that have a value from 0x000 to 0x7ff. a data instruction tells the controller to transfer data between the eeprom and the register map. the controller requires the following two parameters to carry out the data transfer: ? the number of bytes to transfer ? the register map target address the controller decodes the number of bytes to transfer directly from the data instruction itself by adding one to the value of the instruction. for example, the 1a data instruction has a decimal value of 26; therefore, the controller knows to transfer 27 bytes (one more than the value of the instruction). when the controller encounters a data instruction, it knows to read the next two bytes in the scratch pad because these contain the register map target address. note that, in the eeprom scratch pad, the two registers that comprise the address portion of a data instruction have the msb of the address in the d7 position of the lower register address. the bit weight increases from left to right, from the lower register address to the higher register address. furthermore, the starting address always indicates the lowest numbered register map address in the range of bytes to transfer. that is, the controller always starts at the register map target address and counts upward regardless of whether the serial i/o port is operating in i 2 c, spi lsb-first, or spi msb-first mode. as part of the data transfer process during an eeprom upload, the controller calculates a 1-byte checksum and stores it as the final byte of the data transfer. as part of the data transfer process during an eeprom download, however, the controller again calculates a 1-byte checksum value but compares the newly calculated checksum with the one that was stored during the upload process. if an upload/download checksum pair does not match, the controller sets the eeprom fault status bit. if the upload/download checksums match for all data instructions encountered during a download sequence, the controller sets the eeprom complete status bit. condition instructions are those that have a value from b0 to cf. the b1 to cf condition instructions represent condition 1 to condition 31, respectively. the b0 condition instruction is special because it represents the null condition (see the eeprom conditional processing section).
AD9557 data sheet rev. a | page 40 of 92 a pause instruction, like an end instruction, is stored at the end of a sequence of instructions in the scratch pad. when the controller encounters a pause instruction during an upload sequence, it keeps the eeprom address pointer at its last value. this way the user can store a new instruction sequence in the scratch pad and upload the new sequence to the eeprom. the new sequence is stored in the eeprom address locations immediately following the previously saved sequence. this process is repeatable until an upload sequence contains an end instruction. the pause instruction is also useful when used in conjunction with condition processing. it allows the eeprom to contain multiple occurrences of the same registers, with each occurrence linked to a set of conditions (see the eeprom conditional processing section). eeprom upload to upload data to the eeprom, the user must first ensure that the write enable bit (register 0x0e00, bit 0) is set. then, on setting the autoclearing save to eeprom bit (register 0x0e02, bit 0), the controller initiates the eeprom data storage process. uploading eeprom data requires that the user first write an instruction sequence into the scratch pad registers. during the upload process, the controller reads the scratch pad data byte- by-byte, starting at register 0x0e10 and incrementing the scratch pad address pointer, as it goes, until it reaches a pause or end instruction. as the controller reads the scratch pad data, it transfers the data from the scratch pad to the eeprom (byte-by-byte) and increments the eeprom address pointer accordingly, unless it encounters a data instruction. a data instruction tells the controller to transfer data from the device settings portion of the register map to the eeprom. the number of bytes to transfer is encoded within the data instruction, and the starting address for the transfer appears in the next two bytes in the scratch pad. when the controller encounters a data instruction, it stores the instruction in the eeprom, increments the eeprom address pointer, decodes the number of bytes to be transferred, and increments the scratch pad address pointer. then it retrieves the next two bytes from the scratch pad (the target address) and increments the scratch pad address pointer by 2. next, the controller transfers the specified number of bytes from the register map (beginning at the target address) to the eeprom. when it completes the data transfer, the controller stores an extra byte in the eeprom to serve as a checksum for the transferred block of data. to account for the checksum byte, the controller increments the eeprom address pointer by one more than the number of bytes transferred. note that, when the controller transfers data associated with an active register, it actually transfers the buffered contents of the register (see the buffered/active registers section for details on the difference between buffered and active registers). this allows for the transfer of nonzero autoclearing register contents. note that conditional processing (see the eeprom conditional processing section) does not occur during an upload sequence. eeprom download an eeprom download results in data transfer from the eeprom to the device register map. to download data, the user sets the autoclearing load from the eeprom bit (register 0x0e03, bit 1). this commands the controller to initiate the eeprom download process. during download, the controller reads the eeprom data byte-by-byte, incrementing the eeprom address pointer as it goes, until it reaches an end instruction. as the controller reads the eeprom data, it executes the stored instructions, which includes transferring stored data to the device settings portion of the register map whenever it encounters a data instruction. note that conditional processing (see the eeprom conditional processing section) is applicable only when downloading. automatic eeprom download following a power-up, an assertion of the reset pin, or a soft reset (register 0x0000, bit 5 = 1), if the pincontrol pin is low, and m3 and m2 are either high or low (see ), the instruction sequence stored in the eeprom executes automatically with one of eight conditions. if m3 and m2 are left floating and the pincontrol pin is low, the eeprom is bypassed and the factory defaults are used. in this way, a previously stored set of register values downloads automatically on power-up or with a hard or soft reset. see the section for details regarding conditional processing and the way it modifies the download process. table 22 eeprom conditional processing table 22. eeprom setup m3 m2 id eeprom download? low low 1 yes, eeprom condition 1 low open 2 yes, eeprom condition 2 low high 3 yes, eeprom condition 3 open low 4 yes, eeprom condition 4 open open 0 no open high 5 yes, eeprom condition 5 high low 6 yes, eeprom condition 6 high open 7 yes, eeprom condition 7 high high 8 yes, eeprom condition 8
data sheet AD9557 rev. a | page 41 of 92 eeprom conditional processing the condition tag board is a table maintained by the eeprom controller. when the controller encounters a condition instruct- tion, it decodes the b1 through cf instructions as condition = 1 through condition = 8, respectively, and tags that particular condition in the condition tag board. however, the b0 condition instruction decodes as the null condition, for which the controller clears the condition tag board, and subsequent download instructions execute unconditionally (until the controller encounters a new condition instruction). the condition instructions allow conditional execution of eeprom instructions during a download sequence. during an upload sequence, however, they are stored as is and have no effect on the upload process. note that, during eeprom downloads, the condition instructions themselves and the end instruction always execute unconditionally. conditional processing makes use of two elements: the condition (from condition 1 to condition 8) and the condition tag board. the relationships among the condition, the condition tag board, and the eeprom controller appear schematically in figure 42 . during download, the eeprom controller executes or skips instructions, depending on the value of the condition and the contents of the condition tag board. note, however, that condition instructions and the end instruction always execute unconditionally during download. if condition = 0, then all instructions during download execute unconditionally. if condition 0 and there are any tagged conditions in the condition tag board, then the controller executes instructions only if the condition is tagged. if the condition is not tagged, then the controller skips instructions until it encounters a condition instruction that decodes as a tagged condition. note that the condition tag board allows for multiple conditions to be tagged at any given moment. this conditional processing mechanism enables the user to have one download instruction sequence with many possible outcomes, depending on the value of the condition and the order in which the controller encounters condition instructions. th e condition is a 5-bit value with 32 possibilities. condition = 0 is the null condition. when the null condition is in effect, the eeprom controller executes all instructions unconditionally. the remaining eight possibilities (that is, condition = 1 through condition = 8) modify the eeprom controllers handling of a download sequence. the condition originates from one of two sources (see figure 42 ), as follows: ? fncinit, bits[7:3], which is the state of the m2 to m3 multifunction pins at power-up (see table 22 ) ? register 0x0e01, bits[3:0] if register 0x0e01, bits[4:0] 0, then the condition is the value that is stored in register 0x0e01, bits[4:0]; otherwise, the condition is fncinit, bits[7:3]. note that a nonzero condition that is present in register 0x0e01, bits[4:0] takes precedence over fncinit, bits[7:3]. example condition 3 and condition 13 are tagged eeprom eeprom controller upload procedure condition handler download procedure condition tag board 1 6 5 4 3 2 11 10 9 8 7 30 31 24 23 22 21 20 19 18 17 16 15 14 13 12 25 26 27 28 29 if b1 ? instruction ? cf, then tag decoded condition execute/skip instruction(s) scratch pad cond ition condition = 0e01, bits[3:0] else condition = fncinit, bits[7:3] endif m3 m2 if instruction = b0, then clear all tags fncinit, bits[7:3] register 0x0e01, bits[3:0] store condition instructions as they are read from the scratch pad. watch for occurrence of condition instructions during download. if {no tags} or {condition = 0} execute instructions else if {condition is tagged} execute instructions else skip instructions endif endif 5 5 5 if {0e01, bits[3:0] ? 0} 09197-025 figure 42. eeprom conditional processing
AD9557 data sheet rev. a | page 42 of 92 table 23 lists a sample eeprom download instruction sequence. it illustrates the use of condition instructions and how they alter the download sequence. the table begins with the assumption that no conditions are in effect. that is, the most recently executed condition instruction is either b0 or no conditional instructions have been processed. table 23. eeprom conditional processing example instruction action 0x08 0x01 0x00 transfer the system clock register contents, regardless of the current condition. 0xb1 tag condition 1. 0x19 0x04 0x00 transfer the clock distribution register contents only if tag condition = 1. 0xb2 tag condition 2. 0xb3 tag condition 3. 0x07 0x05 0x00 transfer the reference input register contents only if tag condition = 1, 2, or 3. 0x0a calibrate the system clock only if tag condition = 1, 2, or 3. 0xb0 clear the tag condition board. 0x80 execute an i/o update, regardless of the value of the tag condition. 0x0a calibrate the system clock, regardless of the value of the tag condition. st oring multiple device setups in eeprom co nditional processing makes it possible to create a number of different device setups, store them in eeprom, and download a specific setup on demand. to do so, first program the device control registers for a specific setup. then, store an upload sequence in the eeprom scratch pad with the following general form: 1. condition instruction (b1 to cf) to identify the setup with a specific condition (1 to 31) 2. data instructions (to save the register contents), along with any required calibrate and/or i/o update instructions 3. pause instruction (fe) w ith the upload sequence written to the scratch pad, perform an eeprom upload (register 0x0e02, bit 0). r eprogram the device control registers for the next desired setup. then store a new upload sequence in the eeprom scratch pad with the following general form: 1. condition instruction (b0) 2. the next desired condition instruction (b1 to cf, but different from the one used during the previous upload to identify a new setup) 3. data instructions (to save the register contents) along with any required calibrate and/or i/o update instructions 4. pause instruction (fe) with the upload sequence written to the scratch pad, perform an eeprom upload (register 0x0e02, bit 0). repeat the process of programming the device control registers for a new setup, storing a new upload sequence in the eeprom scratch pad (step 1 through step 4), and executing an eeprom upload (register 0x0e02, bit 0) until all of the desired setups have been uploaded to the eeprom. note that, on the final upload sequence stored in the scratch pad, the pause instruction (fe) must be replaced with an end instruction (ff). to download a specific setup on demand, first store the condition associated with the desired setup in register 0x0e01, bits[4:0]. then perform an eeprom download (register 0x0e03, bit 1). alternatively, to download a specific setup at power-up, apply the required logic levels necessary to encode the desired condition on the m2 to m3 multifunction pins. then power up the device; an automatic eeprom download occurs. the condition (as established by the m2 to m3 multifunction pins) guides the download sequence and results in a specific setup. keep in mind that the number of setups that can be stored in the eeprom is limited. the eeprom can hold a total of 2048 bytes. each nondata instruction requires one byte of storage. each data instruction, however, requires n + 4 bytes of storage, where n is the number of transferred register bytes and the other four bytes include the data instruction itself (one byte), the target address (two bytes), and the checksum calculated by the eeprom controller during the upload sequence (one byte). programming the eeprom to configure an m pin to control synchronization of the clock distribution a s pecial eeprom loading sequence is required to use the eeprom to load the registers and to use an m pin to enable/disable outputs. to control the output sync function by using an m pin, perform the following steps: 1. enable the m pins by writing register 0x0200 = 0x01. 2. issue an i/o update (register 0x0005 = 0x01). 3. set the appropriate m pin function (see the clock distribution synchronization section for details). if this sequence is not performed, a sync pulse is issued automatically.
data sheet AD9557 rev. a | page 43 of 92 th e following changes write register 0x0200 first and then issue an i/o update before writing the remaining m pin configuration registers in register 0x0201 to register 0x0208. th e default eeprom loading sequence from register 0x0e10 to register 0x0e16 is unchanged. the following steps must be inserted into the eeprom storage sequence: 1. r0x0e17 = 0x00 # write one byte 2. r0x0e18 = 0x02 # at register 0x0200 3. r0x0e19 = 0x00 # 4. r0x0e1a = 0x80 # op code for i/o update r0x0e1b = 0x10 # transfer 17 instead of 18 bytes 5. r0x0e1c = 0x02 # transfer starts at register address 6. r0x0e1d = 0x01 # 0x0201 instead of 0x0200 th e rest of the eeprom loading sequence is the same as the default eeprom loading sequence, except that the register address of the eeprom storage sequence is shifted down four bytes from the default. for example, ? r 0x0e1e = default value of register 0x0e1a = 0x2e ? r 0x0e1f = default value of register 0x0e1b = 0x03 ? r 0x0e20 = default value of register 0x0e1c = 0x00 ? ? r0x0e40 = default value of register 0x0e1c = 0x3c = 0xff ? (end of data)
AD9557 data sheet rev. a | page 44 of 92 serial control port the AD9557 serial control port is a flexible, synchronous serial communications port that provides a convenient interface to many industry-standard microcontrollers and microprocessors. the serial control port is compatible with most synchronous transfer formats, including i2c, motorola spi, and intel ssr protocols. the serial control port allows read/write access to the AD9557 register map. in spi mode, single or multiple byte transfers are supported. the spi port configuration is programmable via register 0x0000. this register is integrated into the spi control logic rather than in the register map and is distinct from the i 2 c register 0x0000. it is also inaccessible to the eeprom controller. although the AD9557 supports both the spi and i 2 c serial port protocols, only one or the other is active following power-up (as determined by the m0 and m1 multifunction pins during the startup sequence). that is, the only way to change the serial port protocol is to reset the device (or cycle the device power supply). spi/i2c port selection because the AD9557 supports both spi and i2c protocols, the active serial port protocol depends on the logic state of the pincontrol, m1, and m0 pins. the pincontrol pin must be low, and the state of the m0 and m1 pins determines the i 2 c address, or if spi mode is enabled. see table 24 for the i 2 c address assignments. table 24. spi/i2c serial port setup m1 m0 spi/i2c low low spi low open i2c, 1101000 low high i2c, 1101001 open low i2c, 1101010 open open i2c, 1101011 open high i2c, 1101100 high low i2c, 1101101 high open i2c, 1101110 high high i2c, 1101111 spi serial port operation pin descriptions the sclk (serial clock) pin serves as the serial shift clock. this pin is an input. sclk synchronizes serial control port read and write operations. the rising edge sclk registers write data bits, and the falling edge registers read data bits. the sclk pin supports a maximum clock rate of 40 mhz. the sdio (serial data input/output) pin is a dual-purpose pin and acts as either an input only (unidirectional mode) or as both an input and an output (bidirectional mode). the AD9557 default spi mode is bidirectional. the sdo (serial data output) pin is useful only in unidirectional i/o mode. it serves as the data output pin for read operations. the cs (chip select) pin is an active low control that gates read and write operations. this pin is internally connected to a 30 k pull-up resistor. when cs is high, the sdo and sdio pins go into a high impedance state. spi mode operation the spi port supports both 3-wire (bidirectional) and 4-wire (unidirectional) hardware configurations and both msb-first and lsb-first data formats. both the hardware configuration and data format features are programmable. by default, the AD9557 uses the bidirectional msb-first mode. the reason that bidirectional is the default mode is so that the user can still write to the device, if it is wired for unidirectional operation, to switch to unidirectional mode. assertion (active low) of the cs pin initiates a write or read operation to the spi port. for data transfers of three bytes or fewer (excluding the instruction word), the device supports the AD9557 cs stalled high mode (see ). in this mode, the table 25 cs pin can be temporarily deasserted on any byte boundary, allowing time for the system controller to process the next byte. cs can be deasserted only on byte boundaries, however. this applies to both the instruction and data portions of the transfer. during stall high periods, the serial control port state machine enters a wait state until all data is sent. if the system controller decides to abort a transfer midstream, the state machine must be reset either by completing the transfer or by asserting the cs pin for at least one complete sclk cycle (but less than eight sclk cycles). deasserting the cs pin on a nonbyte boundary terminates the serial transfer and flushes the buffer. in streaming mode (see table 25 ), any number of data bytes can be transferred in a continuous stream. the register address is automatically incremented or decremented. cs must be deasserted at the end of the last byte that is transferred, thereby ending the stream mode. table 25. byte transfer count w1 w0 bytes to transfer 0 0 1 0 1 2 1 0 3 1 1 streaming mode
data sheet AD9557 rev. a | page 45 of 92 communication cycleinstruction plus data the spi protocol consists of a two-part communication cycle. the first part is a 16-bit instruction word that is coincident with the first 16 sclk rising edges and a payload. the instruction word provides the AD9557 serial control port with information regarding the payload. the instruction word includes the r/ w bit that indicates the direction of the payload transfer (that is, a read or write operation). the instruction word also indicates the number of bytes in the payload and the starting register address of the first payload byte. write if the instruction word indicates a write operation, the payload is written into the serial control port buffer of the AD9557 . data bits are registered on the rising edge of sclk. the length of the transfer (1, 2, or 3 bytes or streaming mode) depends on the w0 and w1 bits (see table 25 ) in the instruction byte. when not streaming, cs can be deasserted after each sequence of eight bits to stall the bus (except after the last byte, where it ends the cycle). when the bus is stalled, the serial transfer resumes when cs is asserted. deasserting the cs pin on a nonbyte boundary resets the serial control port. reserved or blank registers are not skipped over automatically during a write sequence. therefore, the user must know what bit pattern to write to the reserved registers to preserve proper operation of the part. generally, it does not matter what data is written to blank registers, but it is customary to write 0s. most of the serial port registers are buffered (refer to the buffered/active registers section for details on the difference between buffered and active registers). therefore, data written into buffered registers does not take effect immediately. an additional operation is required to transfer buffered serial control port contents to the registers that actually control the device. this is accomplished with an i/o update operation, which is performed in one of two ways. one is by writing a logic 1 to register 0x0005, bit 0 (this bit is autoclearing). the other is to use an external signal via an appropriately programmed multifunction pin. the user can change as many register bits as desired before executing an i/o update. the i/o update operation transfers the buffer register contents to their active register counterparts. read the AD9557 supports the long instruction mode only. if the instruction word indicates a read operation, the next n 8 sclk cycles clock out the data from the address specified in the instruction word. n is the number of data bytes read and depends on the w0 and w1 bits of the instruction word. the readback data is valid on the falling edge of sclk. blank registers are not skipped over during readback. a readback operation takes data from either the serial control port buffer registers or the active registers, as determined by register 0x0004, bit 0. spi instruction word (16 bits) the msb of the 16-bit instruction word is r/ w , which indicates whether the instruction is a read or a write. the next two bits, w1 and w0, indicate the number of bytes in the transfer (see ). the final 13 bits are the register address (a12 to a0), which indicates the starting register address of the read/write operation (see ). table 25 table 27 spi msb-/lsb-first transfers the AD9557 instruction word and payload can be msb first or lsb first. the default for the AD9557 is msb first. the lsb- first mode can be set by writing a 1 to register 0x0000, bit 6. immediately after the lsb-first bit is set, subsequent serial control port operations are lsb first. when msb-first mode is active, the instruction and data bytes must be written from msb to lsb. multibyte data transfers in msb-first format start with an instruction byte that includes the register address of the most significant payload byte. subsequent data bytes must follow, in order, from high address to low address. in msb-first mode, the serial control port internal address generator decrements for each data byte of the multibyte transfer cycle. when register 0x0000, bit 6 = 1 (lsb first), the instruction and data bytes must be written from lsb to msb. multibyte data transfers in lsb-first format start with an instruction byte that includes the register address of the least significant payload byte, followed by multiple data bytes. the serial control port internal byte address generator increments for each byte of the multibyte transfer cycle. for multibyte msb-first (default) i/o operations, the serial control port register address decrements from the specified starting address toward address 0x0000. for multibyte lsb-first i/o operations, the serial control port register address increments from the starting address toward address 0x1fff. reserved addresses are not skipped during multibyte i/o operations; therefore, the user should write the default value to a reserved register and 0s to unmapped registers. note that it is more efficient to issue a new write command than to write the default value to more than two consecutive reserved (or unmapped) registers. table 26. streaming mode (no addresses are skipped) write mode address direction stop sequence lsb first increment 0x0000 ... 0x1fff msb first decrement 0x1fff ... 0x0000 table 27. serial control port, 16-bit instruction word, msb first msb lsb i15 i14 i13 i12 i11 i10 i9 i8 i7 i6 i5 i4 i3 i2 i1 i0 r/ w w1 w0 a12 a11 a10 a9 a8 a7 a6 a5 a4 a3 a2 a1 a0
AD9557 data sheet rev. a | page 46 of 92 cs sclk don't care sdio a12 w0 w1 r/w a11 a10 a9 a8 a7 a6 a5 a4 a3 a2 a1 a0 d7 d6 d5 d4 d3 d2 d1 d0 d7 d6 d5 d4 d3 d2 d1 d0 don't care don't care don't care 16-bit instruction header register (n) data register (n ? 1) data 09197-029 figure 43. serial control port writemsb firs t, 16-bit instruction, two bytes of data cs sclk sdio sdo register (n) data 16-bit instruction header register (n ? 1) data register (n ? 2) data register (n ? 3) data a12 w0w1 r/w a11 a10 a9 a8 a7 a6 a5 a4 a3 a2 a1 a0 don't care don't care don't care don't care d7 d6 d5 d4 d3 d2 d1 d0 d7 d6 d5 d4 d3 d2 d1 d0 d7 d6 d5 d4 d3 d2 d1 d0 d7 d6 d5 d4 d3 d2 d1 d0 09197-030 figure 44. serial control port readmsb firs t, 16-bit instruction, four bytes of data t s don't care don't care w1w0a12a11a10a9a8a7a6a5d4d3d2d1d0 don't care don't care r/w t ds t dh t high t low t clk t c cs s clk sdio 09197-031 figure 45. serial control port writemsb firs t, 16-bit instruction, timing measurements data bit n ? 1 data bit n cs sclk sdio sdo t dv 09197-032 figure 46. timing diagram for serial control port register read cs sclk don't care don't care 16-bit instruction header register (n) data register (n + 1) data sdio don't care don't care a0 a1 a2 a3 a4 a5 a6 a7 a8 a9 a10 a11 a12 d1d0r/w w1w0 d2 d3 d4 d5 d6 d7 d0 d1 d2 d3 d4 d5 d6 d7 09197-033 figure 47. serial control port writelsb firs t, 16-bit instruction, two bytes of data
data sheet AD9557 rev. a | page 47 of 92 cs sclk sdio t high t low t clk t s t ds t dh t c bit n bit n + 1 09197-034 figure 48. serial control port timingwrite table 28. serial control port timing parameter description t ds setup time between data and the rising edge of sclk t dh hold time between data and the rising edge of sclk t clk period of the clock setup time between the cs falling edge and the sclk rising edge (start of the communication cycle) t s setup time between the sclk rising edge and cs rising edge (end of the communication cycle) t c t high minimum period that sclk should be in a logic high state t low minimum period that sclk should be in a logic low state t dv sclk to valid sdio and sdo (see figure 46 )
AD9557 data sheet rev. a | page 48 of 92 i 2 c serial port operation the i 2 c interface has the advantage of requiring only two control pins and is a de facto standard throughout the i2c industry. however, its disadvantage is programming speed, which is 400 kbps maximum. the AD9557 i2c port design is based on the i2c fast mode standard; therefore, it supports both the 100 khz standard mode and 400 khz fast mode. fast mode imposes a glitch tolerance requirement on the control signals. that is, the input receivers ignore pulses of less than 50 ns duration. the AD9557 i2c port consists of a serial data line (sda) and a serial clock line (scl). in an i2c bus system, the AD9557 is connected to the serial bus (data bus sda and clock bus scl) as a slave device; that is, no clock is generated by the AD9557. the AD9557 uses direct 16-bit memory addressing instead of traditional 8-bit memory addressing. the AD9557 allows up to seven unique slave devices to occupy the i 2 c bus. these are accessed via a 7-bit slave address that is transmitted as part of an i 2 c packet. only the device that has a matching slave address responds to subsequent i 2 c commands. table 24 lists the supported device slave addresses. i 2 c bus characteristics a summary of the various i 2 c protocols appears in table 29 . table 29. i 2 c bus abbreviation definitions abbreviation definition s start sr repeated start p stop a acknowledge a nonacknowledge w write r read the transfer of data is shown in figure 49 . one clock pulse is generated for each data bit transferred. the data on the sda line must be stable during the high period of the clock. the high or low state of the data line can change only when the clock signal on the scl line is low. data line stable; data valid change of data allowed sda scl 09197-035 figure 49. valid bit transfer start/stop functionality is shown in figure 50 . the start condition is characterized by a high-to-low transition on the sda line while scl is high. the start condition is always generated by the master to initialize a data transfer. the stop condition is characterized by a low-to-high transition on the sda line while scl is high. the stop condition is always generated by the master to terminate a data transfer. every byte on the sda line must be eight bits long. each byte must be followed by an acknowledge bit; bytes are sent msb first. the acknowledge bit (a) is the ninth bit attached to any 8-bit data byte. an acknowledge bit is always generated by the receiving device (receiver) to inform the transmitter that the byte has been received. it is done by pulling the sda line low during the ninth clock pulse after each 8-bit data byte. the nonacknowledge bit ( a ) is the ninth bit attached to any 8- bit data byte. a nonacknowledge bit is always generated by the receiving device (receiver) to inform the transmitter that the byte has not been received. it is done by leaving the sda line high during the ninth clock pulse after each 8-bit data byte. sda start condition stop condition scl s p 09197-036 figure 50. start and stop conditions 12 89 12 3to 7 3to 7 89 10 sda scl s msb ack from slave receiver ack from slave receiver p 09197-037 figure 51. acknowledge bit
data sheet AD9557 rev. a | page 49 of 92 data transfer process the master initiates data transfer by asserting a start condition. this indicates that a data stream follows. all i2c slave devices connected to the serial bus respond to the start condition. the master then sends an 8-bit address byte over the sda line, consisting of a 7-bit slave address (msb first) plus an r/ w bit. this bit determines the direction of the data transfer, that is, whether data is written to or read from the slave device (0 = write, 1 = read). the peripheral whose address corresponds to the transmitted address responds by sending an acknowledge bit. all other devices on the bus remain idle while the selected device waits for data to be read from or written to it. if the r/ w bit is 0, the master (transmitter) writes to the slave device (receiver). if the r/ w bit is 1, the master (receiver) reads from the slave device (transmitter). the format for these commands is described in the data transfer format section. data is then sent over the serial bus in the format of nine clock pulses: one data byte (eight bits) from either master (write mode) or slave (read mode) followed by an acknowledge bit from the receiving device. the number of bytes that can be transmitted per transfer is unrestricted. in write mode, the first two data bytes immediately after the slave address byte are the internal memory (control registers) address bytes, with the high address byte first. this addressing scheme gives a memory address of up to 2 16 ? 1 = 65,535. the data bytes after these two memory address bytes are register data written to or read from the control registers. in read mode, the data bytes after the slave address byte are register data written to or read from the control registers. when all data bytes are read or written, stop conditions are established. in write mode, the master (transmitter) asserts a stop condition to end data transfer during the 10 th clock pulse following the acknowledge bit for the last data byte from the slave device (receiver). in read mode, the master device (receiver) receives the last data byte from the slave device (transmitter) but does not pull sda low during the ninth clock pulse. this is known as a nonacknowledge bit. by receiving the nonacknowledge bit, the slave device knows that the data transfer is finished and enters idle mode. the master then takes the data line low during the low period before the 10 th clock pulse, and high during the 10 th clock pulse to assert a stop condition. a start condition can be used in place of a stop condition. furthermore, a start or stop condition can occur at any time, and partially transferred bytes are discarded. 12 89 12 3to 7 3to 7 891 0 ack from slave receiver ack from slave receiver sda scl s msb p 09197-038 figure 52. data transfer process (master write mode, 2-byte transfer) 12 89 12 3to 7 3to 7 8910 ack from master receiver non-ack from master receiver sda scl s p 09197-039 figure 53. data transfer process (m aster read mode, 2-byte transfer)
AD9557 data sheet rev. a | page 50 of 92 data transfer format write byte formatthe write byte protocol is used to write a register address to the ram, starting from the specified ram addre ss. s slave address w a ram address high byte a ram address low byte a ram data 0 a ram data 1 a ram data 2 a p send byte formatthe send byte protocol is used to set up the register address for subsequent reads. s slave address w a ram address high byte a ram address low byte a p receive byte formatthe receive byte protocol is used to read the data byte(s) from ram, starting from the current address. s slave address r a ram data 0 a ram data 1 a ram data 2 a p read byte formatthe combined format of the send byte and the receive byte. s slave address w a ram address high byte a ram address low byte a sr slave address r a ram data 0 a ram data 1 a ram data 2 a p i2c serial port timing ss r p s sda scl t sp t hd; sta t su; sta t su; dat t hd; dat t hd; sta t su; sto t buf t r t f t r t f 09197-040 t high t low figure 54. i2c serial port timing table 30. ic timing definitions parameter description f scl serial clock t buf bus free time between stop and start conditions t hd; sta repeated hold time start condition t su; sta repeated start condition setup time t su; sto stop condition setup time t hd; dat data hold time t su; dat date setup time t low scl clock low period t high scl clock high period t r minimum/maximum receive scl and sda rise time t f minimum/maximum receive scl and sda fall time t sp pulse width of voltage spikes that must be suppressed by the input filter
data sheet AD9557 rev. a | page 51 of 92 programming the i/o registers the register map spans an address range from 0x0000 through 0x0e3c. each address provides access to 1 byte (eight bits) of data. each individual register is identified by its four-digit hexadecimal address (for exampl e, register 0x0a10). in some cases, a group of addresses collectively defines a register. in general, when a group of registers defines a control parameter, the lsb of the value resides in the d0 position of the register with the lowest address. the bit weight increases right to left, from the lowest register address to the highest register address. note that the eeprom storage sequence registers (address 0x0e10 to address 0x0e3c) are an exception to the above convention (see the eeprom instructions section). buffered/active registers there are two copies of most registers: buffered and active. the value in the active registers is the one that is in use. the buffered registers are the ones that take effect the next time the user writes 0x01 to the i/o update register (register 0x0005). buffering the registers allows the user to update a group of registers (like the digital loop filter coefficients) at the same time, which avoids the potential of unpredictable behavior in the part. registers with an l in the option column are live, meaning that they take effect the moment the serial port transfers that data byte. autoclear registers an a in the option column of the register map identifies an autoclear register. typically, the active value for an autoclear register takes effect following an i/o update. the bit is cleared by the internal device logic upon completion of the prescribed action. register access restrictions read and write access to the register map may be restricted depending on the register in question, the source and direction of access, and the current state of the device. each register can be classified into one or more access types. when more than one type applies, the most restrictive condition is the one that applies. whenever access is denied to a register, all attempts to read the register return a 0 byte, and all attempts to write to the register are ignored. access to nonexistent registers is handled in the same way as for a denied register. regular access registers with regular access do not fall into any other category. both read and write access to registers of this type can be from either the serial ports or the eeprom controller. however, only one of these sources can have access to a register at any given time (access is mutually exclusive). when the eeprom controller is active, in either load or store mode, it has exclusive access to these registers. read-only access an r in the option column of the register map identifies read- only registers. access is available at all times, including when the eeprom controller is active. note that read-only registers (r) are inaccessible to the eeprom, as well. exclusion from eeprom access an e in the option column of the register map identifies a register with contents that are inaccessible to the eeprom. that is, the contents of this type of register cannot be transferred directly to the eeprom or vice versa. note that read-only registers (r) are inaccessible to the eeprom, as well.
AD9557 data sheet rev. a | page 52 of 92 thermal performance table 31. thermal parameters for the 40-lead lfcsp package symbol thermal characteristic using a jedec51-7 plus jedec51-5 2s2p test board 1 value 2 unit ja junction-to-ambient thermal resistance, 0.0 m/sec airflow per jedec jesd51-2 (still air) 30.2 c/w jma junction-to-ambient thermal resistance, 1.0 m/sec airflow per jedec jesd51-6 (moving air) 26.4 c/w jma junction-to-ambient thermal resistance, 2.5 m/sec airflow per jedec jesd51-6 (moving air) 23.6 c/w jb junction-to-board thermal resistance, 0.0 m/sec airflow per jedec jesd51-8 (still air) 16.3 c/w jc junction-to-case thermal resistance (die-to-h eat sink) per mil-std 883, method 1012.1 2.2 c/w jt junction-to-top-of-package characterization parameter, 0 m/sec airflow per jedec jesd51-2 (still air) 0.2 c/w 1 the exposed pad on the bottom of the package must be soldered to ground to achieve the specified thermal performance. 2 results are from simulations. the pcb is a jedec multilayer ty pe. thermal performance for actual applications requires careful inspection of the conditions in the application to determine if they are similar to those assumed in these calculations. the AD9557 is specified for a case temperature (t case ). to ensure that t case is not exceeded, an airflow source can be used. use the following equation to determine the junction temperature on the application pcb: t j = t case + ( jt pd ) where: t j is the junction temperature (c). t case is the case temperature (c) measured by the customer at the top center of the package. jt is the value as indicated in table 31 . pd is the power dissipation (see the table 3 ). va lu e s of ja are provided for package comparison and pcb design considerations. ja can be used for a first order approxi- mation of t j by the following equation: t j = t a + ( ja pd ) where t a is the ambient temperature (c). va lu e s of jc are provided for package comparison and pcb design considerations when an external heat sink is required. va lu e s of jb are provided for package comparison and pcb design considerations.
data sheet AD9557 rev. a | page 53 of 92 power supply partitions the AD9557 power supplies are divided into four groups: dvdd3, dvdd, avdd3, and avdd. all power and ground pins should be connected, even if certain blocks of the chip are powered down. recommended configuration for 3.3 v switching supply a popular power supply arrangement is to power the AD9557 with the output of a 3.3 v switching power supply. when the AD9557 is powered using 3.3 v switching power supplies, all of the 3.3 v supplies can be connected to the 3.3 v switcher output, and a 0.1 f bypass capacitor should be placed adjacent to each 3.3 v power supply pin. configuration for 1.8 v supply when 1.8 v supplies are preferred, it is recommended that an ldo regulator, such as the adp222 , be used to generate the 1.8 v supply from the 3.3 v supply. the adp222 offers excellent power supply rejection in a small (2 mm 2 mm) package. it has two 1.8 v outputs. one output can be used for the dvdd pins (pin 6, pin 34, and pin 35), and the other output can drive the avdd pins. the adp7104 is another good choice for converting 3.3 v to 1.8 v. the close-in noise of the adp7104 is lower than that of the adp222 ; therefore, it may be better suited for applications where close-in phase noise is critical and the AD9557 dpll loop bandwidth is <50 hz. in such cases, all 1.8 v supplies can be connected to one adp7104 . use of ferrite beads on 1.8 v supplies to ensure the very best output-to-output isolation, one ferrite bead should be used instead of a bypass capacitor for each of the following avdd pins: pin 11, pin 17, and pin 18. the ferrite beads should be placed in between the 1.8 v ldo output and each pin listed above. ferrite beads that have low (<0.7 ) dc resistance and approximately 600 impedance at 100 mhz are suitable for this application. see table 2 for the current consumed by each group. refer to figure 20 , figure 21 , and figure 22 for information on the power consumption vs. output frequency.
AD9557 data sheet rev. a | page 54 of 92 pin program function description the AD9557 supports both hard pin and soft pin program function, with the on-chip rom containing the predefined configurations. when a pin program function is enabled and initiated, the selected, predefined configuration is transferred from the rom to the corresponding registers to configure the part into the desired state. overview of on-chip rom features in put/output frequency translation configuration the AD9557 has one on-chip rom that contains a total of 256 different input-output frequency translation configurations for independent selection of 16 input frequencies and 16 output frequencies. each input/output frequency translation configuration assumes that all input frequencies are the same and all the output frequencies are the same. each configuration reprograms the following registers/parameters: ? reference input period register ? reference divider r register ? digital pll feedback divider register (fractional part frac1, modulus part mod1 and integer part n1) free run ? tuning word register ? output pll feedback divider n2 register ? rf divider register ? clock distribution channel divider register a ll configurations are set to support one single system clock frequency as 786.432 mhz (16 the default 49.152 mhz system clock reference frequency). f our different system clock pll configurations ? ref = 49.152 mhz xo (2 on, n = 8) ? ref = 49.152 mhz xtal (2 on, n = 8) ? ref = 24.756 mhz xtal (2 on, n = 16) ? ref = 98.304 mhz xo (2 off, n = 8) four different dpll loop bandwidths ? 1 hz, 10 hz, 50 hz, 100 hz dpll phase margin ? normal phase margin (70) ? high phase margin (88.5) the rom also contains an apll vco calibration bit. this bit is used to program register 0x0405[0] (from 0) to 1 to generate a low-high transition to automatically initiate apll vco cal. table 32. preset input frequencies for hard pin and soft pin programming hard pin program pincontrol = high soft pin program pincontrol = low, register 0x0c01[3:0] freq id frequency (mhz) frequency description m0 pin b3 b2 b1 b0 0 0.008 8 khz 0 0 0 0 0 1 19.44 19.44 mhz ? 0 0 0 1 2 25 25 mhz 1 0 0 1 0 table 33. preset output frequencies for hard pin and soft pin programming hard pin program pincontrol = high soft pin program pincontrol = low, register 0x0c01[7:4] freq id frequency (mhz) frequency description m3 pin m2 pin m1 pin b7 b6 b5 b4 0 19.44 19.44 mhz 0 0 0 0 0 0 0 1 25 25 mhz 0 0 ? 0 0 0 1 2 125 125 mhz 0 0 1 0 0 1 0 3 156.7071 156.25 mhz 1027/1024 0 ? 0 0 0 1 1 4 622.08 622.08 mhz 0 ? ? 0 1 0 0 5 625 625 mhz 0 ? 1 0 1 0 1 6 644.53125 625 mhz 33/32 0 1 0 0 1 1 0 7 657.421875 657.421875 mhz 0 1 ? 1 1 1 8 660.184152 657.421875 mhz 239/238 0 1 1 1 0 0 0 9 666.5143 622.08 mhz 255/238 ? 0 0 1 0 0 1 10 669.3266 622.08 mhz 255/237 ? 0 ? 1 0 1 0 11 672.1627 622.08 mhz 255/236 ? 0 1 1 0 1 1 12 690.5692 644.53125 mhz 255/238 ? ? 0 1 1 0 0
data sheet AD9557 rev. a | page 55 of 92 hard pin program pincontrol = high register 0x0c01[7:4] soft pin program pincontrol = low, freq id frequency (mhz) frequency description m3 pin m2 pin m1 pin b7 b6 b5 b4 13 693.4830 644.53125 mhz 255/237 ? ? ? 1 1 0 1 14 698.8124 622.08 mhz 255/237 ? ? 1 1 1 1 0 15 704.380580 657.421875 mhz 255/238 ? 1 0 1 1 1 1 table 34. system clock configuration in hard pin and soft pin programming modes hard pin program pincontrol = high, irq pin soft pin program pincontrol = low, register 0x0c02[1:0] equivalent system clock pll register settings freq id frequency (mhz) system clock configuration irq pin bit 1 bit 0 0 49.152 xtal mode, doubler on, n = 8 0 0 0 0001, 0000, 1000 1 49.152 xtal mode off, doubler on, n = 8 ? 0 1 2 24.576 xtal mode, doubler on, n = 16 1 1 0 3 98.304 xtal mode off, doubler off, n = 8 n/a 1 1 hard pin programming mode the state of the pincontrol pin at power-up controls whether or not the chip is in hard pin programming mode. setting the pincontrol pin high disables the i 2 c protocol, although the register map can be accessed via the spi protocol. the m0 pin selects one of three input frequencies, and the m3 to m1 pins select one of 16 possible output frequencies. see table 32 and table 33 for details. the system clock configuration is controlled by the state of the irq pin at startup (see table 34 ). the digital pll loop bandwidth, reference input frequency accuracy tolerance ranges, and dpll phase margin selection are not available in hard pin programming mode unless the user uses the serial port to change their default values. when in hard pin programming mode, the user must set register 0x0200[0] = 1 to activate the irq, ref status, and pll lock status signals at the multifunction pins. s oft pin programming mode overview t he soft pin programming function is controlled by a dedicated register section (address 0x0c00 to address 0x0c08). the purpose of soft pin programming is to use the register bits to mimic the hard pins for the configuration section. when in soft pin programming mode, both the spi and i 2 c ports are available. ? address 0x0c00[0] enables accessibility to address 0x0c01 and address 0x0c02 (soft pin section 1). this bit must be set in soft pin mode. ? address 0x0c03[0] enables accessibility to address 0x0c04 to address 0x0c06 (soft pin section 2). this bit must be set in soft pin mode. ? address 0x0c01[3:0] select one of 16 input frequencies. ? address 0x0c01[7:4] select one of 16 output frequencies. ? address 0x0c02[1:0] select the system clock configuration. ? address 0x0c06[1:0] select one of four input frequency tolerance ranges. ? address 0x0c06[3:2] select one of four dpll loop bandwidths. ? address 0x0c06[4] selects the dpll phase margin. ? address 0x0c04[3:0] scale the refa and refb input frequency down by divide-by-1, -4, -8, or -16 independently. for example, when address 0x0c01[3:0] = 0101 to select 622.08 mhz input frequency for both refa and refb, setting address 0x0c04[1:0] = 0x01 scales down the refa input frequency to 155.52 mhz (= 622.08 mhz/4). this is done by internally scaling the r divider for refa up by 4 and the refa period up by 4. ? address 0x0c05[3:0] scale the channel 0 and channel 1 output frequency down by divide-by-1, divide-by-4, divide-by-8, or divide-by-16.
AD9557 data sheet rev. a | page 56 of 92 register map register addresses that are not listed in table 35 are not used, and writing to those registers has no effect. the user should write the default value to sections of registers marked reserved. r = read only. a = autoclear. e = excluded from eeprom loading. l = live (i/o u pdate not required for register to take effect or for a read-only register to be updated). table 35. register map reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def serial control port configuration and part identification 0x0000 l, e spi control sdo enable lsb first/ increment address soft reset reserved 00 0x0000 l i2c control reserved soft reset reserved 00 0x0004 readback control reserved read buffer register 00 0x0005 a, l i/o update reserved i/o update 00 0x0006 l user scratch pad user scratch pad[7:0] 00 0x0007 l user scratch pad[15:8] 00 0x000a r, l silicon rev silicon revision[7:0] 21 0x000b r, l reserved reserved 0d 0x000c r, l part id clock part family id[7:0] 01 0x000d r, l clock part family id[15:8] 00 system clock 0x0100 sysclk config pll feedback divider system clock n divider[7:0] 08 0x0101 reserved load from rom (reserved) sysclk xtal enable sysclk p divider[1:0] sysclk doubler enable 09 or 19 0x0102 reserved reserved 00 0x0103 sysclk period nominal system clock period (fs)[7:0] (1 ns at 1 ppm accuracy) 0e 0x0104 nominal system clock period (fs)[15:8] (1 ns at 1 ppm accuracy) 67 0x0105 reserved nominal system clock period[20:16] 13 0x0106 sysclk stability system clock stability period (ms)[7:0] 32 0x0107 system clock stability period (ms)[15:8] 00 0x0108 a reserved reset sysclk stab timer (autoclear) system clock stability period (ms)[19:16] (not autoclearing) 00 general configuration 0x0200 en_mpin reserved enable m pins and irq pin function 00 0x0201 m0func m0 output/ a input e function[6:0] b0 0x0202 m1func m1output/ a input e function[6:0] b1 0x0203 m2func m2 output/ a input e function[6:0] c0 0x0204 m3func m3 output/ a input e function[6:0] c1 0x0205 reserved b2 0x0206 reserved b3 0x0207 reserved c2 0x0208 reserved c3
data sheet AD9557 rev. a | page 57 of 92 reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def 0x0209 irq pin output mode reserved status signal at irq pin[1:0] use irq pin for status signal irq pin driver type[1:0] 1f 0x020a irq mask reserved sysclk unlocked sysclk locked apll unlocked apll locked apll cal complete apll cal started 00 0x020b reserved pin program end sync distribution watchdog timer eeprom fault eeprom complete 00 0x020c switching closed freerun holdover frequency unlocked frequency locked phase unlocked phase locked 00 0x020d reserved history updated frequency unclamped frequency clamped phase slew unlimited phase slew limited 00 0x020e reserved refb validated refb fault cleared refb fault reserved refa validated refa fault cleared refa fault 00 0x020f reserved reserved 00 0x0210 watchdog timer 1 watchdog timer (ms)[7:0] 00 0x0211 watchdog timer (ms)[15:8] 00 0x0300 free run frequency tw 30-bit free run frequency tuning word[7:0] 11 0x0301 30-bit free run frequency tuning word[15:8] 15 0x0302 30-bit free run frequency tuning word[23:16] 64 0x0303 reserved 30-bit free run frequency tuning word[29:24] 1b 0x0304 digital oscillator control reserved dco 4-level output reserved (must be 1b) reserved 10 0x0305 reserved reserved 00 0x0306 dpll frequency clamp lower limit of pull-in range[7:0] 51 0x0307 lower limit of pull-in range[15:8] b8 0x0308 reserved lower limit of pull-in range[19:16] 02 0x0309 upper limit of pull-in range[7:0] 3e 0x030a upper limit of pull-in range[15:8] 0a 0x030b reserved upper limit of pull-in range[19:16] 0b 0x030c closed-loop phase lock offset (0.5 ms) fixed phase lock offset (signed; ps)[7:0] 00 0x030d fixed phase lock offset (signed; ps)[15:8] 00 0x030e fixed phase lock offset (signed; ps)[23:16] 00 0x030f reserved fixed phase lock offset (signed; ps)[29:24] 00 0x0310 incremental phase lock offset step size (ps/step)[7:0] (up to 65.5 ns/step) 00 0x0311 incremental phase lock offset step size (ps/step)[15:8] (up to 65.5 ns/step) 00 0x0312 phase slew rate limit phase slew rate limit (s/sec)[7:0] (315 s/sec up to 65.536 ms/sec) 00 0x0313 phase slew rate limit (s/sec)[15:8] (315 s/sec up to 65.536 ms/sec) 00 0x0314 holdover history history accumulation timer (ms)[7:0] (up to 65 seconds) 0a 0x0315 history accumulation timer (ms)[15:8] (up to 65 seconds) 00 0x0316 history mode reserved single sample fallback persistent history incremental average 00 0x0317 l base loop filter a coefficient set (high phase margin) hpm alpha-0[7:0] 8c 0x0318 l hpm alpha-0[15:8] ad 0x0319 l reserved hpm alpha-1[6:0] 4c 0x031a l hpm beta-0[7:0] f5 0x031b l hpm beta-0[15:8] cb 0x031c l reserved hpm beta-1[6:0] 73 0x031d l hpm gamma-0[7:0] 24 0x031e l hpm gamma-0[15:8] d8 0x031f l reserved hpm gamma-1[6:0] 59 0x0320 l hpm delta-0[7:0] d2 0x0321 l hpm delta-0[15:8] 8d 0x0322 l reserved hpm delta-1[6:0] 5a
AD9557 data sheet rev. a | page 58 of 92 reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def 0x0323 l base loop filter a coefficient set (normal phase margin of 70o) npm alpha-0[7:0] 24 0x0324 l npm alpha-0[15:8] 8c 0x0325 l reserved npm alpha-1[6:0] 49 0x0326 l npm beta-0[7:0] 55 0x0327 l npm beta-0[15:8] c9 0x0328 l reserved npm beta-1[6:0] 7b 0x0329 l npm gamma-0[7:0] 9c 0x032a l npm gamma-0[15:8] fa 0x032b l reserved npm gamma-1[6:0] 55 0x032c l npm delta-0[7:0] ea 0x032d l npm delta-0[15:8] e2 0x032e l reserved npm delta-1[6:0] 57 output pll (apll) 0x0400 apll charge pump output pll (apll) charge pump[7:0] 81 0x0401 apll n divider output pll (apll) feedback n divider[7:0] 14 0x0402 reserved reserved 00 0x0403 apll loop filter control apll loop filter control[7:0] 07 0x0404 reserved bypass internal rzero 00 0x0405 apll vco control reserved (default: 0x2) apll locked controlled sync disable reserved manual apll vco cal (not auto- clearing) 20 0x0406 reserved reserved 00 0x0407 rf divider rf divider 2[3:0] rf divider 1[3:0] 44 0x0408 reserved rf divider start-up mode reserved pd rf divider 2 pd rf divider 1 02 output clock distribution 0x0500 distribution output sync reserved mask channel 1 sync mask channel 0 sync reserved sync source selection auto sync mode 02 0x0501 channel 0 enable 3.3 v cmos driver out0 format[2:0] out0 polarity[1:0] out0 drive strength enable out0 10 0x0502 channel 0 divider[7:0] 00 0x0503 reserved channel 0 pd select rf divider 2 channel 0 divider[9:8] 00 0x0504 reserved channel 0 divider phase[5:0] 00 0x0505 channel 1 reserved out1 format[2:0] out1 polarity[1:0] out1 drive strength enable out1 10 0x0506 reserved 10 0x0507 channel 1 divider[7:0] 03 0x0508 reserved channel 1 pd select rf divider 2 channel 1 divider[9:8] 00 0x0509 reserved channel 1 divider phase[5:0] 00 0x050a reserved 10 0x050b reserved 10 0x050c reserved 00 0x050d reserved 00 0x050e reserved 00 0x050f reserved 10 0x0510 reserved 03 0x0511 reserved 00 0x0512 reserved 00 0x0513 reserved 00 0x0514 reserved 00 0x0515 reserved 00
data sheet AD9557 rev. a | page 59 of 92 reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def reference inputs 0x0600 reference power- down reserved reserved refb power- down refa power- down 00 0x0601 reference logic type reserved refb logic type[1:0] refa logic type[1:0] 00 0x0602 reference priority reserved refb priority[1:0] refa priority[1:0] 00 0x0603 reserved reserved 00 profile a (for refa) 0x0700 l reference period (up to 1.1 ms) nominal reference period (fs), bits[7:0] (default: 51.44 ns =1/(19.44 mhz) for default system clock setting) c9 0x0701 l nominal period (fs), bits[15:8] ea 0x0702 l nominal period (f s), bits[23:16] 10 0x0703 l nominal period (f s), bits[31:24] 03 0x0704 l nominal period (f s), bits[39:32] 00 0x0705 l frequency tolerance inner tolerance (1 ppm), bits[7:0] (for reference invalid to valid; 50% down to 1 ppm) (default: 5%) 14 0x0706 l inner tolerance (1 ppm), bits[15:8] (for reference invalid to valid; 50% down to 1 ppm) 00 0x0707 l reserved inner tolerance, bits[19:16] 00 0x0708 l outer tolerance (1 ppm), bits[7:0] (for reference valid to invalid; 50% down to 1 ppm) (default: 10%) 0a 0x0709 l outer tolerance (1 ppm), bits[15:8] (for reference valid to invalid; 50% down to 1 ppm) 00 0x070a l reserved outer tolerance, bits[19:16] 00 0x070b l validation validation timer (ms), bits[7:0] (up to 65.5 seconds) 0a 0x070c l validation timer (ms), bits[15:8] (up to 65.5 seconds] 00 0x070d l reserved reserved 00 0x070e l select base loop filter reserved sel high pm base loop filter 00 0x070f l dpll loop bw digital pll loop bw scaling factor[7:0] (default: 0x01f4 = 50 hz) f4 0x0710 l digital pll loop bw scaling factor[15:8] 01 0x0711 l reserved bw scaling factor[16] 00 0x0712 l dpll r divider (20 bits) r divider[7:0] c5 0x0713 l r divider[15:8] 00 0x0714 l reserved enable refa divide-by-2 r divider[19:16] 00 0x0715 dpll n divider (17 bits) digital pll feedback dividerinteger part n1[7:0] 6b 0x0716 digital pll feedback divide rinteger part n1[15:8] 07 0x0717 reserved digital pll feedback divider integer part n1[16] 00 0x0718 dpll fractional feedback divider (24 bits) digital pll fractional feedback dividerfrac1[7:0] 04 0x0719 digital pll fractional feedback dividerfrac1[15:8] 00 0x071a digital pll fractional feedback dividerfrac1[23:16] 00 0x071b dpll fractional feedback divider modulus (24 bits) digital pll feedback divider modulusmod1[7:0] 05 0x071c digital pll feedback divider modulusmod1[15:8] 00 0x071d digital pll feedback divider modulusmod1[23:16] 00 0x071e l lock detectors phase lock threshold[7:0] (ps) bc 0x071f l phase lock threshold[15:8] (ps) 02 0x0720 l phase lock fill rate[7:0] 0a 0x0721 l phase lock drain rate[7:0] 0a 0x0722 l frequency lock threshold[7:0] bc 0x0723 l frequency lock threshold[15:8] 02 0x0724 l frequency lock threshold[23:16] 00 0x0725 l frequency lock fill rate[7:0] 0a 0x0726 l frequency lock drain rate[7:0] 0a
AD9557 data sheet rev. a | page 60 of 92 reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def profile b (for refb) 0x0740 l reference period (up to 1.1 ms) nominal period (fs), bits[7:0] (default: 125 s = 1/(8 khz) fo r default system clock setting) 00 0x0741 l nominal period (fs), bits[15:8] a2 0x0742 l nominal period (f s), bits[23:16] 94 0x0743 l nominal period (f s), bits[31:24] 1a 0x0744 l nominal period (f s), bits[39:32] 1d 0x0745 l frequency tolerance inner tolerance (1 ppm), bits[7:0] (for reference invalid to valid; 50% down to 1 ppm) (default: 5%) 14 0x0746 l inner tolerance (1 ppm), bits[15:8] (for reference invalid to valid; 50% down to 1 ppm) 00 0x0747 l reserved inner tolerance, bits[19:16] 00 0x0748 l outer tolerance (1 ppm), bits[7:0] (for reference valid to invalid; 50% down to 1 ppm] (default: 10%) 0a 0x0749 l outer tolerance (1 ppm), bits[15:8] (for reference valid to invalid; 50% down to 1 ppm) 00 0x074a l reserved outer tolerance, bits[19:16] 00 0x074b l validation validation timer (ms), bits[7:0] (up to 65.5 seconds) 0a 0x074c l validation timer (ms), bits[15:8] (up to 65.5 seconds) 00 0x074d l reserved 00 0x074e l select base loop filter reserved sel high pm base loop filt 00 0x074f l dpll loop bw digital pll loop bandwidth scaling factor[7:0] (default: 0x01f4 = 50 hz) f4 0x0750 l digital pll loop bandwidth scaling factor[15:8] 01 0x0751 l reserved bw scaling factor[16] 00 0x0752 l dpll r divider (20 bits) r divider[7:0] 00 0x0753 l r divider[15:8] 00 0x0754 l reserved enable refb divide-by-2 r divider[19:16] 00 0x0755 dpll n divider (17 bits) digital pll feedback dividerinteger part n1[7:0] 1f 0x0756 digital pll feedback divide rinteger part n1[15:8] 5b 0x0757 reserved digital pll feedback divider integer part n1[16] 00 0x0758 dpll fractional feedback divider (24 bits) digital pll fractional feedback dividerfrac1[7:0] 00 0x0759 digital pll fractional feedback dividerfrac1[15:8] 00 0x075a digital pll fractional feedback dividerfrac1[23:16] 00 0x075b dpll fractional feedback divider modulus (24 bits) digital pll feedback divider modulusmod1[7:0] 01 0x075c digital pll feedback divider modulusmod1[15:8] 00 0x075d digital pll feedback divider modulusmod1[23:16] 00 0x075e l lock detectors phase lock threshold[7:0] (ps) bc 0x075f l phase lock threshold[15:8] (ps) 02 0x0760 l phase lock fill rate[7:0] 0a 0x0761 l phase lock drain rate[7:0] 0a 0x0762 l frequency lock threshold[7:0] bc 0x0763 l frequency lock threshold[15:8] 02 0x0764 l frequency lock threshold[23:16] 00 0x0765 l frequency lock fill rate[7:0] 0a 0x0766 l frequency lock drain rate[7:0] 0a 0x0780 reserved c9 0x0781 reserved ea 0x0782 reserved 10 0x0783 reserved 03 0x0784 reserved 00 0x0785 reserved 14 0x0786 reserved 00 0x0787 reserved 00 0x0788 reserved 0a
data sheet AD9557 rev. a | page 61 of 92 reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def 0x0789 reserved 00 0x078a reserved 00 0x078b reserved 0a 0x078c reserved 00 0x078d reserved 00 0x078e reserved 00 0x078f reserved f4 0x0790 reserved 01 0x0791 reserved 00 0x0792 reserved c5 0x0793 reserved 00 0x0794 reserved 00 0x0795 reserved 6b 0x0796 reserved 07 0x0797 reserved 00 0x0798 reserved 04 0x0799 reserved 00 0x079a reserved 00 0x079b reserved 05 0x079c reserved 00 0x079d reserved 00 0x079e reserved bc 0x079f reserved 02 0x07a0 reserved 0a 0x07a1 reserved 0a 0x07a2 reserved bc 0x07a3 reserved 02 0x07a4 reserved 00 0x07a5 reserved 0a 0x07a6 reserved 0a 0x07c0 reserved 00 0x07c1 reserved a2 0x07c2 reserved 94 0x07c3 reserved 1a 0x07c4 reserved 1d 0x07c5 reserved 14 0x07c6 reserved 00 0x07c7 reserved 00 0x07c8 reserved 0a 0x07c9 reserved 00 0x07ca reserved 00 0x07cb reserved 0a 0x07cc reserved 00 0x07cd reserved 00 0x07ce reserved 00 0x07cf reserved f4 0x07d0 reserved 01 0x07d1 reserved 00 0x07d2 reserved 00 0x07d3 reserved 00 0x07d4 reserved 00 0x07d5 reserved 1f 0x07d6 reserved 5b 0x07d7 reserved 00 0x07d8 reserved 00 0x07d9 reserved 00 0x07da reserved 00 0x07db reserved 01
AD9557 data sheet rev. a | page 62 of 92 reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def 0x07dc reserved 00 0x07dd reserved 00 0x07de reserved bc 0x07df reserved 02 0x07e0 reserved 0a 0x07e1 reserved 0a 0x07e2 reserved bc 0x07e3 reserved 02 0x07e4 reserved 00 0x07e5 reserved 0a 0x07e6 reserved 0a operational controls 0x0a00 power-down soft reset exclude regmap dco pd sysclk pd ref input pd tdc pd apll pd clock dist pd full pd 00 0x0a01 loop mode reserved user holdover user freerun ref switchover mode[2:0] reserved user ref in manual switchover mode 00 0x0a02 cal/sync reserved soft sync clock dist reserved 00 0x0a03 a clear/reset functions reserved clear lf clear cci reserved clear auto sync clear tw history clear all irqs clear watchdog 00 0x0a04 a irq clearing reserved sysclk unlocked sysclk locked apll unlocked apll locked apll cal ended apll cal started 00 0x0a05 a reserved pin program end sync clock dist watchdog timer eeprom fault eeprom complete 00 0x0a06 a switching closed freerun holdover frequency unlocked frequency locked phase unlocked phase locked 00 0x0a07 a reserved history updated frequency unclamped frequency clamped phase slew unlimited phase slew limited 00 0x0a08 a reserved refb validated refb fault cleared refb fault reserved refa validated refa fault cleared refa fault 00 0x0a09 a reserved reserved 00 0x0a0a a increment phase offset reserved reset phase offset decrement phase offset increment phase offset 00 0x0a0b a manual reference validation reserved force timeout b force timeout a 00 0x0a0c manual reference invalidation reserved ref mon override b ref mon override a 00 0x0a0d static reference validation reserved ref mon bypass b ref mon bypass a 00 quick in-out frequency soft pin configuration 0x0c00 l, e enable soft pin section 1 reserved en soft pin section 1 00 0x0c01 l, e soft pin section 1 output frequency selection[3:0] input frequency selection[3:0] 00 0x0c02 l, e reserved sysclk pll ref sel[1:0] 00 0x0c03 l, e enable soft pin section 2 reserved en soft pin section 2 00 0x0c04 l, e soft pin section 2 reserved refb frequency scale[1:0] refa frequency scale[1:0] 00 0x0c05 l, e reserved channel 1 output frequency scale[1:0] channel 0 output frequency scale[1:0] 00 0x0c06 l, e reserved sel high pm base loop filter dpll loop bw[1:0] ref input frequency tolerance[1:0] 00 0x0c07 l, a, e soft pin transfer reserved soft pin start transfer 00 0x0c08 l, e soft pin reset reserved soft pin reset 00
data sheet AD9557 rev. a | page 63 of 92 reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def read-only status (accessible during eeprom transactions) 0x0d00 r, l eeprom reserved pin program rom load process fault detected load in progress save in progress n/a 0x0d01 r, l sysclk and pll status reserved dpll_apll_ lock all plls locked apll vco status apll cal in process apll lock sysclk stable sysclk lock detect n/a 0x0d02 r, l irq monitor events reserved sysclk unlocked sysclk locked apll unlocked apll lock detected apll cal ended apll cal started n/a 0x0d03 r, l reserved pin program end output dist sync watchdog timer eeprom fault eeprom complete n/a 0x0d04 r, l switching closed freerun holdover frequency unlocked frequency locked phase unlocked phase locked n/a 0x0d05 r, l reserved history updated frequency unclamped frequency clamped phase slew unlimited phase slew limited n/a 0x0d06 r, l reserved refb validated refb fault cleared refb fault reserved refa validated refa fault cleared refa fault n/a 0x0d07 r, l reserved n/a 0x0d08 r dpll reserved offset slew limiting frequency lock phase lock loop switching holdover active freerun n/a 0x0d09 r reserved frequency clamped history available active reference priority reserved current active reference n/a 0x0d0a r reserved n/a 0x0d0b r refa/refb b valid b fault b fast b slow a valid a fault a fast a slow n/a 0x0d0c r reserved n/a 0x0d0d r holdover history tuning word readback[31:0] n/a 0x0d0e r n/a 0x0d0f r n/a 0x0d10 r n/a 0x0d11 r lock detector phase tub phase tub [7:0] n/a 0x0d12 r reserved phase tub[11:8] n/a 0x0d13 r lock detector frequency tub frequency tub[7:0] n/a 0x0d14 r reserved frequency tub[11:8] n/a nonvolatile memory (eeprom) control 0x0e00 e write protect reserved write enable 00 0x0e01 e condition reserved conditional value[3:0] 00 0x0e02 a, e save reserved save to eeprom 00 0x0e03 a, e load reserved load from eeprom reserved 00
AD9557 data sheet rev. a | page 64 of 92 reg addr (hex) opt name d7 d6 d5 d4 d3 d2 d1 d0 def eeprom storage sequence 0x0e10 e eeprom id data: two bytes 01 0x0e11 e address 0x0006 00 0x0e12 e 06 0x0e13 e system clock data: nine bytes 08 0x0e14 e address 0x0100 01 0x0e15 e 00 0x0e16 e i/o update action: i/o update 80 0x0e17 e general data: 18 bytes 11 0x0e18 e address 0x0200 02 0x0e19 e 00 0x0e1a e dpll data: 47 bytes 2e 0x0e1b e address 0x0300 03 0x0e1c e 00 0x0e1d e apll data: nine bytes 08 0x0e1e e address 0x0400 04 0x0e1f e 00 0x0e20 e clock dist data: 22 bytes 15 0x0e21 e address 0x0500 05 0x0e22 e 00 0x0e23 e i/o update action: i/o update 80 0x0e24 e reference inputs data: four bytes 03 0x0e25 e address:0x0600 06 0x0e26 e 00 0x0e27 e reserved 01 0x0e28 e reserved 06 0x0e29 e 40 0x0e2a e profile refa data: 39 bytes 26 0x0e2b e address 0x0700 07 0x0e2c e 00 0x0e2d e profile refb data: 39 bytes 26 0x0e2e e address 0x0740 07 0x0e2f e 40 0x0e30 e reserved 26 0x0e31 e reserved 07 0x0e32 e 80 0x0e33 e reserved 26 0x0e34 e reserved 07 0x0e35 e reserved c0 0x0e36 e i/o update action: i/o update 80 0x0e37 e operational controls data: 14 bytes 0d 0x0e38 e address 0x0a00 0a 0x0e39 e 00 0x0e3a e calibrate apll action: calibrate output pll a0 0x0e3b e i/o update action: i/o update 80 0x0e3c e end of data action: end of data ff 0x0e3d to 0xe45 e unused unused (available for additional eeprom instructions) 00
data sheet AD9557 rev. a | page 65 of 92 register map bit descriptions serial port configuration (reg ister 0x0000 to register 0x0005) table 36. serial configuration (note that the conten ts of register 0x0000 are not stored to the eeprom.) address bits bit name description 0x0000 7 sdo enable enables spi port sdo pin. 1 = 4-wire (sdo pin enabled). 0 (default) = 3-wire. 6 lsb first/increment address bit order for spi port. 1 = least significant bit and byte first. register addresses are automatically incremented in multibyte transfers. 0 (default) = most significant bit and byte first. register addresses are automatically de ccremented in multibyte transfers. 5 soft reset device reset (invokes an eeprom download or pin program rom download if eeprom or pin program is enabled. see the eeprom section and pin configuration and function descriptions for details. [4:0] reserved reserved. table 37. readback control address bits bit name description 0x0004 [7:1] reserved reserved. 0 read buffer register for buffered registers, serial port read-back reads from actual (active) registers instead of the buffer. 1 = reads buffered values that take ef fect on next assertion of i/o update. 0 (default) = reads values currently applied to the devices internal logic. table 38. soft i/o update address bits bit name description [7:1] reserved reserved. 0x0005 0 i/o update writing a 1 to this bit transfers the data in th e serial i/o buffer registers to the devices internal control registers. unless a register is marked as live (as indicated by an l in the opt column of the register map), the user must write to this bit before any register settings can take effect and before a read-o nly register can be up dated with the most current value. this is an autoclearing bit. table 39. user scratch pad address bits bit name description 0x0006 [7:0] user scratch pad[7:0] 0x0007 [7:0] user scratch pad[15:8] user programmable eeprom id registers. thes e registers enable users to write a unique code of their choosing to keep track of revi sions to the eeprom register loading. it has no effect on part operation. 0 = default. silicon revision (register 0x000a) table 40. silicon revision address bits bit name description 0x000a [7:0] silicon revision this read-only register identifies the revision level of the AD9557 . clock part serial id (register 0x000c to register 0x000d) table 41. clock part family id address bits bit name description 0x000c [7:0] clock part family id[7:0] this read-only register (along with re gister 0x000d) uniquely identifies an AD9557 or ad9558 . no other part in the adi ad95xx family has a value of 0x0001 in these two registers. default: 0x01 for the AD9557 and ad9558. 0x000d [7:0] clock part family id[15:8] this re gister is a continuation of register 0x000c. default: 0x00 for the AD9557 and ad9558.
AD9557 data sheet rev. a | page 66 of 92 system clock (register 0x0100 to register 0x0108) table 42. system clock pll feedback divider (n3 divider) address bits bit name description 0x0100 [7:0] sysclk n3 divider system clock pll feed back divider value: 4 n3 255 (default: 0x08). table 43. sysclk configuration address bits bit name description [7:5] reserved reserved. 4 load from rom (reserved) this reserved bit has no function. 0 (default) = power-on default and rom not loaded. 1 = rom values are loaded into the register space. 3 sysclk xtal enable enables the crystal main taining amplifier for the system clock input. 1 (default) = crystal mode (crystal maintaining amplifier enabled). 0 = external xo or other system clock source. [2:1] sysclk p divider system clock input divider. 00 (default) = 1. 01 = 2. 10 = 4. 11 = 8. 0x0101 0 sysclk doubler enable enable clock double r on system clock input to reduce noise. 0 = disable. 1 (default) = enable. table 44. nominal system clock period address bits bit name description 0x0103 [7:0] system clock period, bits[7:0]. default: 0x0e. 0x0104 [7:0] nominal system clock period (fs) system clock period, bits[15:8]. default: 0x67. [7:5] reserved reserved. 0x0105 [4:0] nominal system clock period (fs) system clock period, bits[20:16]. default: 0x13. table 45. system clock stability period address bits bit name description 0x0106 [7:0] system clock period, bits[7:0]. default: 0x32 (0x000032 = 50 ms). 0x0107 [7:0] system clock stability period (ms) system clock period, bits[15:8]. default: 0x00. [7:5] reserved reserved. 4 reset sysclk stability timer this autoclearing bit resets the system clock stability timer. 0x0108 [3:0] system clock stability period system clock period, bits[19:16]. default: 0x00.
data sheet AD9557 rev. a | page 67 of 92 general configuration (register 0x0200 to register 0x0214) multifunction pin control (m3 to m0) and irq pi n control (register 0x0200 to register 0x0209) note that the default setting for the m3 to m0 multifunction pins and the irq pin is that of a 3-level logic input at startup. setting bit 1 in register 0x0200 to 1 enables normal m3 to m0 pin functionality. table 46. multifunction pins (m0 to m3) control address bits bit name description 0x0200 [7:1] reserved 0 enable m pins and irq pin function 0 (default) = disables the function of the m pins and the irq pin control register (address 0x0201 to address 0x0209); the m pins and irq pin are in 3-level logic input state. 1 = the m pins and irq pin are out of 3-level logic input state and enable the binary function of the m pins and the irq pin control registers (address 0x0201 to address 0x0209). 0x0201 7 m0 output/input in/out control for m0 pin. 0 = input (2-level logic control pin). 1 (default) = output (2-level logic status pin). [6:0] function see table 124 and table 125. default: 0xb0 = refa valid. 0x0202 7 m1 output/input in/out control for m1 pin (same as m0). [6:0] function see table 124 and table 125. default: 0xb1 = refb valid. 0x0203 7 m2 output/input in/out control for m2 pin (same as m0). [6:0] function see table 124 and table 125. default: 0xc0 = refa active. 0x0204 7 m3 output/input in/out control for m3 pin (same as m0). [6:0] function see table 124 and table 125. default: 0xc1 = refb active. 0x0205 [7:0] reserved reserved. 0x0206 [7:0] reserved reserved. 0x0207 [7:0] reserved reserved. 0x0208 [7:0] reserved reserved. table 47. irq pin output mode address bits bit name description 0x0209 [7:5] reserved reserved [4:3] status signal at irq pin[1:0] this sele ction is valid only when address 0x0209[2] = 1 00 = dpll phase locked 01 = dpll frequency locked 10 = system clock pll locked 11 (default) = (dpll phase locked) and (system clock pll locked) and (apll locked) 2 use irq pin for status signal 0 = uses irq pin to monitor irq event 1 (default) = uses irq pin to monitor internal status signals [1:0] irq pin driver type select the output mode of the irq pin 00 = nmos, open drain (requires an external pull-up resistor) 01 = pmos, open drain (requires an external pull-down resistor) 10 = cmos, active high 11 (default) = cmos, active low
AD9557 data sheet rev. a | page 68 of 92 irq mask (register 0x020a to register 0x020f) the irq mask register bits form a one-to-one correspondence with the bits of the irq monitor register (0x0d02 to 0x0d09). when set to logic 1, the irq mask bits enable the corresponding irq monitor bits to indicate an irq event. the default for all irq mask bit s is logic 0, which prevents the irq monitor from detecting any internal interrupts. table 48. irq mask for sysclk address bits bit name description [7:6] reserved reserved 5 sysclk unlocked enables irq for indicating a sysc lk pll state transition from locked to unlocked 4 sysclk locked enables irq for indicating a sysc lk pll state transition from unlocked to locked 3 apll unlocked enables irq for indicating a apll state transition from locked to unlocked 2 apll locked enables irq for indicating a apll st ate transition from unlocked to locked 1 apll cal complete enables irq for indicating that apll (lcvco) calibration has completed 0x020a 0 apll cal started enables irq for indicating that apll (lcvco) calibration has begun table 49. irq mask for distribution sync, watchdog timer, and eeprom address bits bit name description [7:5] reserved reserved 4 pin program end enables irq for indicating succe ssful completion of an pin program rom load 3 sync distribution enables irq for indicating a distribution sync event 2 watchdog timer enables irq for indicati ng expiration of the watchdog timer 1 eeprom fault enables irq for indicating a faul t during an eeprom load or save operation 0x020b 0 eeprom complete enables irq for indicating successful completion of an eeprom load or save operation table 50. irq mask for the digital pll address bits bit name description 7 switching enables irq for indicating that the dpll is switching to a new reference 6 closed enables irq for indicating that the dpll has entered closed-loop operation 5 freerun enables irq for indicating that the dpll has entered free run mode 4 holdover enables irq for indicating that the dpll has entered holdover mode 3 frequency unlocked enables irq for indica ting that the dpll lost frequency lock 2 frequency locked enables irq for indicating that the dpll has acquired frequency lock 1 phase unlocked enables irq for indica ting that the dpll lost phase lock 0x020c 0 phase locked enables irq for indicating that the dpll has acquired phase lock table 51. irq mask for history update, frequency limit and phase slew limit address bits bit name description [7:5] reserved reserved 4 history updated enables irq for indicating the occurrence of a tuning word history update 3 frequency unclamped enables irq for indicating a freque ncy limit state transition from clamped to unclamped 2 frequency clamped enables irq for indicating a state transition of the frequency limiter from unclamped to clamped 1 phase slew unlimited enables irq for indicating a state transition of the phase slew limiter from slew limiting to not slew limiting 0x020d 0 phase slew limited enables irq for indicating a state transition of the phase slew limiter from not slew limiting to slew limiting
data sheet AD9557 rev. a | page 69 of 92 table 52. irq mask for reference inputs address bits bit name description 7 reserved reserved 6 refb validated enables irq for indicating that refb has been validated 5 refb fault cleared enables irq for indicating that refb has been cleared of a previous fault 4 refb fault enables irq for indica ting that refb has been faulted 3 reserved reserved 2 refa validated enables irq for indicating that refa has been validated 1 refa fault cleared enables irq for indicating that refa has been cleared of a previous fault 0x020e 0 refa fault enables irq for indica ting that refa has been faulted 0x020f [7:0] reserved reserved table 53. watchdog timer 1 1 address bits bit name description 0x0210 [7:0] watchdog timer bits[7:0] default: 0x00 0x0211 [7:0] watchdog timer (ms) watchdog timer bits[15:8] default: 0x00 1 note that the watchdog timer is expressed in units of milliseconds (ms). the default value is 0 (disabled). dpll configuration (register 0x0300 to register 0x032e) table 54. free run frequency tuning word 1 address bits bit name description 0x0300 [7:0] free run frequency tuning word bits[7:0]; default: 0x11 0x0301 [7:0] 30-bit free run frequency tuning word free run frequency tuning word bits[15:8]; default: 0x15 0x0302 [7:0] free run frequency tunin g word bits[23:9]; default: 0x64 0x0303 [7:6] reserved reserved [5:0] 30-bit free run frequency word free run frequency tuning word bits[29:24]: default: 0x1b 1 note that the default free run tuning word is 0x1b641511, which is used fo r 8 khz/19.44 mhz = 622. 08 mhz translation. table 55. digital oscillator control address bits bit name description [7:6] reserved default: 00b 5 dco 4-level output 0 (default) = dco 3-level output mode 1 = enables dco 4-level output mode 4 reserved reserved (must be set to 1b) 0x0304 [3:0] reserved reserved (default: 0x0)
AD9557 data sheet rev. a | page 70 of 92 table 56. dpll frequency clamp address bits bit name description 0x0306 [7:0] lower limit pull-in range bits[7:0] default: 0x51 0x0307 [7:0] lower limit of pull-in range (expressed as a 20-bit frequency tuning word) lower limit pull-in range bits[15:8] default: 0xb8 0x0308 [7:4] reserved default: 0x0 [3:0] lower limit of pull-in range lower limit pull-in range bits[19:16] default: 0x2 0x0309 [7:0] upper limit pull-in range bits[7:0] default: 0x3e 0x030a [7:0] upper limit of pull-in range (expressed as a 20-bit frequency tuning word) upper limit pull-in range bits[15:8] default: 0x0a 0x030b [7:4] reserved default: 0x0 [3:0] upper limit of pull-in range upper limit pull-in range bits[19:16] default: 0xb table 57. fixed closed-loop phase lock offset address bits bit name description 0x030c [7:0] fixed phase lock offset bits[7:0] default: 0x00 0x030d [7:0] fixed phase lock offset bits[15:8] default 0x00 0x030e [7:0] fixed phase lock offset (signed; ps) fixed phase lock offset bits[23:16] default: 0x00 0x030f [7:6] reserved reserved; default: 0x0 [5:0] fixed phase lock offset (signed; ps) fixed phase lock offset bits[29:24] default: 0x00 table 58. incremental closed-loop phase lock offset step size 1 address bits bit name description 0x0310 [7:0] incremental phase lock offset step size bits[7:0]. default: 0x00. this controls the static phase offset of the dpll while it is locked. 0x0311 [7:0] incremental phase lock offset step size (ps) incremental phase lock offset step size bits[15:8] default: 0x00. this controls the static phase offset of the dpll while it is locked. 1 note that the default incremental closed-loop phase lock offset step size va lue is 0x0000 = 0 (0 ns). table 59. phase slew rate limit address bits bit name description 0x0312 [7:0] phase slew rate limit bits[7:0]. default: 0x00. this register controls the maximum allowable phase slewing during transients and reference switching. the default phase slew rate limit is 0, or disabled. minimum useful value is 310 s/sec. 0x0313 [7:0] phase slew rate limit (s/sec) phase slew rate limit bits[15:8] . default: 0x00.
data sheet AD9557 rev. a | page 71 of 92 table 60. history accumulation timer address bits bit name description 0x0314 [7:0] history accumulation timer bits[7:0]. default: 0x0a. for register 0x0314 and register 0x0315, 0x000a = 10 ms. maximum is 65 sec. this register controls the amount of tuning word averaging used to determine the tuning word used in holdover. never program a timer value of zero. the default value is 0x000a = 10 decimal, which equates to 10 ms. 0x0315 [7:0] history accumulation timer (ms) history accumulation timer bits[15:8]. default: 0x00. table 61. history mode address bits bit name description [7:5] reserved reserved. 4 single sample fallback controls holdover history. if tuning word history is not available for the reference that was active just prior to holdover, then: 0 (default) = uses the free run frequency tuning word register value. 1 = uses the last tuning word from the dpll. 3 persistent history controls holdover history in itialization. when switching to a new reference: 0 (default) = clear the tuning word history. 1 = retain the previous tuning word history. 0x0316 [2:0] incremental average history mode value from 0 to 7 (default: 0). when set to non-zero, causes the first hist ory accumulation to update prior to the first complete averaging period. after the first full interval, updates occur only at the full period. 0 (default) = update only after the full interval has elapsed. 1 = update at 1/2 the full interval. 2 = update at 1/4 and 1/2 of the full interval. 3 = update at 1/8, 1/4, and 1/2 of the full interval. ... 7 = update at 1/256, 1/128, 1/64, 1/32, 1/16, 1/8, 1/4, and 1/2 of the full interval. table 62. base digital loop filter with high phase margin (pm = 88.5, bw = 0.1 hz, third pole frequency = 10 hz, n1 = 1) 1 address bits bit name description 0x0317 [7:0] alpha-0 coefficient linear bits[7:0]. default: 0x8c 0x0318 [7:0] hpm alpha-0 linear alpha-0 coefficient linear bits[15:8] 7 reserved reserved 0x0319 [6:0] hpm alpha-1 exponent alpha-1 coefficient exponent bits[6:0] 0x031a [7:0] hpm beta-0 linear beta -0 coefficient linear bits[7:0] 0x031b [7:0] beta-0 coeffi cient linear bits[15:8] 0x031c 7 reserved reserved [6:0] hpm beta-1 exponent beta-1 coefficient exponent bits[6:0] 0x031d [7:0] hpm gamma-0 linear gamma-0 coefficient linear bits[7:0] 0x031e [7:0] gamma-0 coefficient linear bits[15:8] 0x031f 7 reserved reserved [6:0] hpm gamma-1 exponent gamma-1 coefficient exponent bits[6:0] 0x0320 [7:0] delta-0 coefficient linear bits[7:0] 0x0321 [7:0] hpm delta-0 linear delta-0 coefficient linear bits[15:8] 0x0322 7 reserved reserved [6:0] hpm delta-1 exponent delta-1 coefficient exponent bits[6:0] 1 note that the base digital loop filter coefficients (, , , and ) have the following general form: x(2y), where x is the lin ear component and y is the exponential component of the coefficient. the value of the linear component (x) constitutes a fraction, where 0 x 1. the exponential co mponent (y) is a signed integer.
AD9557 data sheet rev. a | page 72 of 92 table 63. base digital loop filter with normal phase margin (pm = 70, bw = 0.1 hz, pole frequency = 2 hz, n1 = 1) 1 address bits bit name description 0x0323 [7:0] alpha-0 coefficient linear bits [7:0] 0x0324 [7:0] npm alpha-0 linear alpha-0 coefficient linear bits [15:8] 7 reserved reserved 0x0325 [6:0] npm alpha-1 exponent alpha-1 coefficient exponent bits [6:0] 0x0326 [7:0] npm beta-0 linear beta-0 coefficient linear bits [7:0] 0x0327 [7:0] beta-0 coefficient linear bits [15:8] 0x0328 7 reserved reserved [6:0] npm beta-1 exponent beta-1 coefficient exponent bits [6:0] 0x0329 [7:0] npm gamma-0 linear gamma-0 coefficient linear bits [7:0] 0x032a [7:0] gamma-0 coefficient linear bits [15:8] 0x032b 7 reserved reserved [6:0] npm gamma-1 exponent gamma-1 coefficient exponent bits [6:0] 0x032c [7:0] npm delta-0 linear delta-0 coefficient linear bits [7:0] 0x032d [7:0] delta-0 coefficient linear bits [15:8] 0x032e 7 reserved reserved [6:0] npm delta-1 exponent delta-1 coefficient exponent bits [6:0] 1 note that the digital loop filter base coefficients (, , , and ) have the general form: x(2 y ), where x is the linear component and y the exponential component of the coefficient. the value of the linear comp onent (x) constitutes a fraction, where 0 x 1. the exponential component (y) is a signed integer. output pll configuration (regis ter 0x0400 to register 0x0408) table 64. output pll setting 1 address bits bit name description 0x0400 [7:0] output pll (apll) charge pump current lsb = 3.5 a 00000001b = 1 lsb; 00000010b = 2 lsb 11111111b = 255 lsb default: 0x81 = 451 a cp current 0x0401 [7:0] apll n divider division = 14 to 255 default: 0x14 = divide-by-20 0x0402 [7:0] reserved reserved 0x0403 [7:6] apll loop filter control pole 2 resistor, rp2; default: 0x07 rp2 () bit 7 bit 6 500 (default) 0 0 333 0 1 250 1 0 200 1 1 [5:3] zero resistor, rzero rzero () bit 5 bit 4 bit 3 1500 (default) 0 0 0 1250 0 0 1 1000 0 1 0 930 0 1 1 1250 1 0 0 1000 1 0 1 750 1 1 0 680 1 1 1
data sheet AD9557 rev. a | page 73 of 92 address bits bit name description pole 1 cp1 cp1 (pf) bit 2 bit 1 bit 0 [2:0] 0 20 80 100 20 40 100 120 (default) 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 [7:1] reserved default: 0x00 0x0404 0 bypass internal rzero 0 (default) = uses the internal rzero resistor. 1 = bypasses the internal rzero resistor (makes rzero = 0 and requires the use of a series external zero resistor). [7:4] reserved default: 0x2 3 apll locked controlled sync disable 0 (default) = the clock distribution sync function is not enabled until the output pll (apll) is calibrated and locked. after apll calibration and lock, the output cloc k distribution sync is armed, and the sync function for the clock outputs is under the control of register 0x0500. 1 = overrides the lock detector state of the output pll; allows register 0x0500 to control the output sync function, regardless of the apll lock status. [2:1] reserved default: 00b 0x0405 0 manual apll vco calibration 1 = initiates vco calibration. (calibration occurs on low-to-high transition). 0 (default) = does nothing. this is not an autoclearing bit. 1 note that the default ap ll loop bw is 180 khz. table 65. reserved address bits bit name description 0x0406 [7:0] reserved default: 0x00 table 66. rf divider setting address bits bit name description [7:4] rf divider 2 division 0000/0001 = 3 0010 = 4 0011 = 5 0100 = 6 (default) 0101 = 7 0110 = 8 0111 = 9 1000 = 10 1001 = 11 0x0407 [3:0] rf divider 1 division 0000/0001 = 3 0010 = 4 0011 = 5 0100 = 6 (default) 0101 = 7 0110 = 8 0111 = 9 1000 = 10 1001 = 11 [7:5] reserved reserved. 4 rf divider start-up mode 0 (default) = rf dividers are held in power-down until the apll feedback divider is detected. this ensures proper rf divider operation, exiting full power-down. 1 = rf dividers are not held in power-down until the apll feedback divider is detected. [3:2] reserved reserved. 1 pd rf divider 2 0 = enables rf divider 2. 1 (default) = powers down rf divider 2. 0x0408 0 pd rf divider 1 0 (default) = enables rf divider 1. 1 = powers down rf divider 1.
AD9557 data sheet rev. a | page 74 of 92 output clock distribution (reg ister 0x0500 to register 0x0515) table 67. distribution output synchronization settings address bits bit name description [7:6] reserved reserved. 5 mask channel 1 sync masks the synchronous reset to the channel 1 divider. 0 (default) = unmasked. the output drivers do not toggle until a sync pulse occurs. 1 = masked. setting this bit asynchronously rele ases channel 1 from the static sync state, thus allowing the channel 1 divider to toggle. channel 1 ignores all sync events while this bit is set. setting this bit do es not enable the output driver s connected to this channel. in addition, the output distrib ution sync also depends on the setting of register 0x0405[3]. 4 mask channel 0 sync masks the synchronous reset to the channel 0 divider. 0 (default) = unmasked. the output drivers do not toggle until a sync pulse occurs. 1 = masked. setting this bit asynchronously rele ases channel 0 from the static sync state, thus allowing the channel 0 divider to toggle. channel 0 ignores all sync events while this bit is set. setting this bit do es not enable the output driver s connected to this channel. in addition, the output distrib ution sync also depends on the setting of register 0x0405[3]. 3 reserved reserved. 2 sync source selection selects the sync sour ce for the clock distribution output channels. 0 (default) = direct. the sync pulse occurs on the next i/o update. 1 = active reference. note that the output distribution sync also depends on the apll being calibrated and locked, unless register 0x0405[3] = 1b. 0x0500 [1:0] automatic sync mode autosync mode. 00 = disabled. a sync command must be issued manually or by using the sync mask bits in this register (bits[5:4]). 01 = sync on dpll frequency lock. 10 (default) = sync on dpll phase lock. 11 = reserved. table 68. distribution out0 setting address bits bit name description 0x0501 7 enable 3.3 v cmos driver 0 (default) = disables 3.3 v cmos driver, and out0 logic is controlled by register 0x0501[6:4 ] 1 = enables 3.3 v cmos driver as operating mode of out0. this bit should be set to 1b only if bits[6:4] are in cmos mode. [6:4] out0 format these bits set the out0 driver mode. 000 = pd, tristate. 001 (default) = hstl. 010 = lvds. 011 = reserved. 100 = cmos, both outputs active. 101 = cmos, p output active, n output power-down. 110 = cmos, n output active, p output power-down. 111 = reserved. [3:2] out0 polarity controls the out0 polarity. 00 (default) = positive, negative. 01 = positive, positive. 10 = negative, positive. 11 = negative, nevative. 1 out0 drive strength controls the output drive capability of out0. 0 (default) = cmos: low drive strength; lvds: 3.5 ma nominal. 1 = cmos: normal drive strength; lvds: 4.5 ma nominal (lvds boost mode). note that this is only in 3.3 v cmos mode for cmos strength. 1.8 v cmos has only the low drive strength. 0 enable out0 enables/disables (1b/0b) out0 1.8 v driver (default is disabled). this bit does not enable/disable out0 if bit 7 of this register is set to 1.
data sheet AD9557 rev. a | page 75 of 92 table 69. distribution channel 0 divider setting address bits bit name description 0x0502 [7:0] channel 0 divider 10-bit ch annel 0 divider, bits[7:0] (lsb). division equals channel 0 divider, bits[9:0] + 1. ([9:0] = 0 is divide-by-1, [9:0] = 1 is divide-by-2[9:0] = 1023 is divide-by-1024) 0x0503 [7:4] reserved reserved 3 channel 0 pd 0 (default) = normal operation. 1 = powers down channel 0. 2 select rf divider for channel 2 1 = selects rf divider 2 as prescaler for channel 0 divider. 0 (default) =selects rf divider 1 as prescaler for channel 0 divider. [1:0] channel 0 divider 10-bit ch annel divider, bits[9:8] (msb). [7:6] reserved reserved. 0x0504 [5:0] channel 0 divider phase divider initial phase after sync relative to the divider input clock (from the rf divider output). lsb is ? of a period of the divider input clock. phase = 0 is no phase offset. phase = 1 is ? a period offset. table 70. distribution out1 setting address bits bit name description 7 reserved reserved. [6:4] out1 format these bits set the out1 driver mode. 000 = pd, tristate. 001 (default) = hstl. 010 = lvds. 011 = reserved. 100 = cmos, both outputs active. 101 = cmos, p output active, n output pd. 110 = cmos, n output active, p output pd. 111 = reserved. [3:2] out1 polarity these bits configure the out1 polar ity in cmos mode and are active only in cmos mode. 00 (default) = positive, negative. 01 = positive, positive. 10 = negative, positive. 11 = negative, negative. 1 out1 drive strength controls the output drive capability of out1. 0 (default) = lvds: 3.5 ma nominal. 1 = lvds: 4.5 ma nominal (lvds boost mode). no cmos control because out1 is 1.8 v cmos only. 0x0505 0 enable out1 setting this bit enables the out1 driver (default is disabled). 0x0506 [7:0] reserved reserved. table 71. distribution channel 1 divider setting address bits bit name description 0x0507 [7:0] channel 1 divider the same control for channe l 1 divider as in register 0x0502 for channel 0 divider 0x0508 [7:0] channel 1 divider the same control for channel 1 divider as in register 0x0503 for channel 0 divider 0x0509 [7:0] channel 1 divider the same control for channe l 1 divider as in register 0x0504 for channel 0 divider
AD9557 data sheet rev. a | page 76 of 92 reference inputs (register 0x0600 to register 0x0602) table 72. reference power-down 1 address bits bit name description [7:2] reserved reserved. 1 refb power-down powers down refb input receiver. 0 (default) = not powered down. 1 = powered down. 0x0600 0 refa power-down powers down refa input receiver. 0 (default) = not powered down. 1 = powered down. 1 when all bits are set, the reference receiver section enters a deep sleep mode. table 73. reference logic family address bits bit name description [7:4] reserved reserved. [3:2] refb logic type selects logic family for refb in put receiver; only refb_p is used in cmos mode. 00 (default) = differential. 01 = 1.2 v to 1.5 v cmos. 10 = 1.8 v to 2.5 v cmos. 11 = 3.0 v to 3.3 v cmos. 0x0601 [1:0] refa logic type the refa logic type settings are the same as register 0x0601[3:2] for refb. table 74. reference priority setting address bits bit name description [7:4] reserved reserved. [3:2] refb priority user assigned priority level (0 to 3) of the reference associated with refb, which ranks that reference relative to the others. 00 (default) = 0. 01 = 1. 10 = 2. 11 = 3. 0x0602 [1:0] refa priority the refa priority settings are the same as in register 0x0602[3:2] for refb.
data sheet AD9557 rev. a | page 77 of 92 dpll profile registers (regis ter 0x0700 to register 0x0766) note that the default value of the refa profile is as follows: input frequency = 19.44 mhz, output frequency = 622.08 mhz/155.5 2 mhz, loop bandwidth = 400 hz, normal phase margin, inner tolerance = 5%, and outer tolerance = 10%. the default value of refb profile is as follows: input frequenc y = 8 khz, output frequency = 622.08 mhz/155.52 mhz, loop bandwi dth = 100 hz, normal phase margin, inner tolerance = 5%, and outer tolerance = 10%. refa profile (register 0x0700 to register 0x0726) table 75. reference periodrefa profile address bits bit name description 0x0700 [7:0] nominal reference peri od bits[7:0] (default: 0xc9) 0x0701 [7:0] nominal reference peri od bits[15:8] (default: 0xea) 0x0702 [7:0] nominal reference period bits[23:16] (default: 0x10) 0x0703 [7:0] nominal reference period bits[31:24] (default: 0x03) 0x0704 [7:0] nominal reference period (fs) nominal reference period bits[39:32] (default: 0x00) default for register 0x0700 to register 0x 0704 = 0x000310eac9 = 51.44 ns (1/19.44 mhz) table 76. reference period tolerancerefa profile address bits bit name description 0x0705 [7:0] input reference frequency monitor inner tolerance bits [7:0] (default: 0x14). 0x0706 [7:0] inner tolerance input reference frequency monitor inner tolerance bit [15:8] (default: 0x00). [7:4] reserved reserved. 0x0707 [3:0] inner tolerance input reference frequency monitor inner tolerance bits[19:16]. default for register 0x0705 to regist er 0x0707 = 0x000014 = 20 (5% or 50,000 ppm). the stratum 3 clock requires inner tolerance of 9.2 ppm and outer tolerance of 12 ppm; an smc clock requires an outer tolerance of 48 ppm. the allowable range for the inner tolera nce is 0x0000a (10%) to 0xfffff (1 ppm). the tolerance of the input frequency monitor is only as accurate as the system clock frequency. 0x0708 [7:0] input reference frequency monitor ou ter tolerance bits [7:0] (default: 0x0a). 0x0709 [7:0] outer tolerance input reference frequency monitor outer tolerance bits[15:8] (default: 0x00). [7:4] reserved reserved. 0x070a [3:0] outer tolerance input reference frequency monito r outer tolerance bits[19:16] . default for register 0x0708 to register 0x070a = 0x00000a = 10 (10% or 100,000 ppm). the stratum 3 clock requires an inner tolera nce of 9.2 ppm and outer tolerance of 12 ppm; an smc clock requires an outer tolerance of 48 ppm. the outer tolerance register setting should al ways be smaller than the inner tolerance. table 77. reference validati on timerrefa profile address bits bit name description 0x070b [7:0] validation timer bits[7:0] (default: 0x0a). this is the amount of time a reference input must be valid before it is declared valid by the reference input monitor (default: 10 ms). 0x070c [7:0] validation timer (ms) validation timer bits[15:8] (default: 0x00). table 78. reserved register address bits bit name description 0x070d [7:0] reserved default: 0x00 table 79. dpll base loop filter selectionrefa profile address bits bit name description [7:1] reserved default: 0x00 0x070e 0 sel high pm base loop filter 0 = base loop fi lter with normal (70) phase margin (default) 1 = base loop filter with high (88.5) phase margin (0.1 db peaking in the closed-loop transfer function for loop bandwidths 2 khz; setting this bit is also recommended for loop bandwidths > 2khz)
AD9557 data sheet rev. a | page 78 of 92 table 80. dpll loop bw scaling factorrefa profile 1 address bits bit name description 0x070f [7:0] digital pll loop bandwidth scal ing factor, bits[7:0] (default: 0xf4). 0x0710 [7:0] dpll loop bw scaling factor (unit of 0.1 hz) digital pll loop bandwidth scaling fa ctor, bits[15:8] (default: 0x01). the default for register 0x070f to register 0x0710 = 0x01f4 = 500 (50 hz loop bandwidth. the loop bandwidth should always be less than the dpll phase detector frequency divided by 20. [7:1] reserved default: 0x00. 0x0711 0 bw scaling factor digital pll loop bandwi dth scaling factor, bit 16 (default: 0b). 1 note that the default dpll loop bandwidth is 50.4 hz. table 81. r dividerrefa profile address bits bit name description 0x0712 [7:0] dpll integer reference divider (minus 1), bits[7:0 ] (default: 0xc5) 0x0713 [7:0] r divider dpll integer reference divider, bits[15:8] (default: 0x00) [7:5] reserved default: 0x0 4 enable refa div2 enables the reference input divide-by-2 for refa 0 = bypass the divide-by-2 (default) 1 = enable the divide-by-2 0x0714 [3:0] r divider dpll integer reference divider, bits[19:16] (default: 0x0) the default for register 0x0712 to register 0x0714 = 0x000c5 = 197 (which equals r = 198) table 82. integer part of fractional feedback divider n1refa profile address bits bit name description 0x0715 [7:0] dpll integer feedback divider (minus 1), bits[7:0 ] (default: 0x6b) 0x0716 [7:0] integer part n1 dpll integer feedback divider, bits[15:8] (default: 0x07) [7:1] reserved default: 0x00 0x0717 0 integer part n1 dpll integer feedback divider, bit 16 (default: 0b) the default for register 0x0715 to register 0x717 = 0x0076b = (which equals n1 = 1900) table 83. fractional part of fractional feedback divider frac1refa profile address bits bit name description 0x0718 [7:0] the numerator of the fractional-n feedback divider, bits[7:0] (default: 0x04) 0x0719 [7:0] the numerator of the fractional-n f eedback divider, bits[15:8] (default: 0x00) 0x071a [7:0] digital pll fractional feedback dividerfrac1 the numerator of the fractional-n feedba ck divider, bits[23:16] (default: 0x00) table 84. modulus of fractional feedback divider mod1refa profile address bits bit name description 0x071b [7:0] the denominator of the fractional-n feedback divider, bi ts[7:0] (default: 0x05) 0x071c [7:0] the denominator of the fractional-n feedback divider, bits [15:8] (default: 0x00) 0x071d [7:0] digital pll feedback divider modulusmod1 the denominator of the fractional-n feedba ck divider, bits[23: 16] (default: 0x00) table 85. phase and frequency lock detector controlsrefa profile address bits bit name description 0x071e [7:0] phase lock thresh old, bits[7:0] (default: 0xbc ); default of 0x02bc = 700 ps 0x071f [7:0] phase lock threshold phase lock threshold, bits[15:8] (default: 0x02) 0x0720 [7:0] phase lock fill rate ph ase lock fill rate, bits[7:0] (default: 0x0a = 10 code/pfd cycle) 0x0721 [7:0] phase lock drain rate ph ase lock drain rate, bits[7:0] (def ault: 0x0a = 10 code/pfd cycle) 0x0722 [7:0] frequency lock threshold, bits[7:0 ] (default: 0xbc); defa ult of 0x02bc = 700 ps 0x0723 [7:0] frequency lock threshol d, bits[15:8] (default: 0x02) 0x0724 [7:0] frequency lock threshold frequency lock threshold, bits[23:16] (default: 0x00) 0x0725 [7:0] frequency lock fill rate frequency lock fill rate, bits[7:0] (default: 0x0a = 10 code/pfd cycle) 0x0726 [7:0] frequency lock drain rate frequency lock drai n rate, bits[7:0] (default: 0x0a = 10 code/pfd cycle)
data sheet AD9557 rev. a | page 79 of 92 refb profile (register 0x0740 to register 0x0766) the refb profile registers, register 0x0740 to register 0x0766, are identical to the refa profile registers, register 0x0700 to register 0x0726. operational controls (registe r 0x0a00 to register 0x0a0d) table 86. general power-down address bits bit name description 7 soft reset exclude regmap resets device but retain programmed register values (default is not reset) 6 dco power-down places dco in deep sleep mode (default is not powered down) 5 sysclk power-down places sysclk input and pll in deep sleep mode (defaul t is not powered down) 4 reference input power-down places reference clock inp uts in deep sleep mode (de fault is not powered down) 3 tdc power-down places the time-to-digital converte r in deep sleep mode (default is not powered down) 2 apll power-down places the output pll in deep sleep mode (default is not powered down) 1 clock dist power-down places the clock distribution ou tputs in deep sleep mode (de fault is not powered down) 0x0a00 0 full power-down places the entire device in deep sleep mode (default is not powered down) table 87. loop mode address bits bit name description 7 reserved reserved. 6 user holdover forces the device into holdover mode (default is not forced holdover mode). if a tuning word history is available, then the history tuning word specifies the dco output frequency. otherwise, the free run frequency tuning word register specifies the dco output frequency. the phase and frequency lock detectors are forced into the unlocked state. 5 user freerun forces the device into user free ru n mode (default is not forced user free run mode). the free run frequency tuning word register specifies the dco output frequency. when the user freerun bit is set, it overrides the user holdover bit (address 0x0a01, bit 6). [4:2] ref switchover mode selects the operating mode of the reference switching state machine. reference switchover mode, bits[2:0]; register 0x0a01[4:2] reference selection mode 000 (default) automatic revertive mode 001 automatic non-revertive mode 010 manual reference select (with automatic fallback mode) 011 manual reference select mode (with auto-holdover) 100 full manual mode (no auto-holdover) 101 not used 110 not used 111 not used 0x0a01 1 reserved reserved. 0 user reference in manual switchover mode input reference when reference switchover mode (register 0x0a01, bits[4:2]) = 100. 0 (default) = input reference a. 1 = input reference b. table 88. cal/sync address bits bit name description [7:2] reserved default: 0x00 1 soft sync clock distribution setting this bit initiates sy nchronization of the clock dist ribution output (default: 0b). nonmasked outputs stall when value is 1b, rest art is initialized on 1b to 0b transition. 0x0a02 0 reserved default: 0b.
AD9557 data sheet rev. a | page 80 of 92 reset functions (register 0x0a03) table 89. reset functions address bits bit name description 7 reserved default: 0b. 6 clear lf setting this bit clears the digi tal loop filter (intended as a debug tool). 5 clear cci setting this bit clears the cci filter (intended as a debug tool). 4 reserved default: 0b. 3 clear auto sync setting this bit resets the au tomatic synchronization logic (see register 0x0500). 2 clear tw history setting this bit resets the tuning word history logic (part of holdover functionality). 1 clear all irqs setting this bit clears the entire irq monitor register (register 0x0d02 to register 0x0d07). it is the equivalent of setting all the bits of the irq clearing register (register 0x0a04 to 0x0a0d). 0x0a03 (autoclear) 0 clear watchdog timer setting this bit resets the watchdog timer (see register 0x0210 and re gister 0x0211). if the timer times out, it simply starts a new timing cycl e. if the timer has not yet timed out, it restarts at time zero without causing a timeout event. continuously resetting the watchdog timer at intervals of less than its timeout period prev ents the watchdog timer from generating a timeout event. irq clearing (register 0x0a04 to register 0x0a09) the irq clearing registers are identical in format to the irq monitor registers (register 0x0d02 to register 0x0d09). when set to logic 1, an irq clearing bit resets the corresponding irq monitor bit, thereby canceling the interrupt request for the indicated event. the irq clearing register is an autoclearing register. table 90. irq clearing for sysclk address bits bit name description [7:6] reserved reserved 5 sysclk unlocked clears sysclk unlocked irq 4 sysclk locked clears sysclk locked irq 3 apll unlocked clears output pll unlocked irq 2 apll locked clears output pll locked irq 1 apll cal ended clears apll calibration complete irq 0x0a04 0 apll cal started clears apll calibration started irq table 91. irq clearing for distributi on sync, watchdog timer and eeprom address bits bit name description [7:5] reserved reserved 4 pin program end clears pin program end irq 3 sync clock distribution clears distribution sync irq 2 watchdog timer clears watchdog timer irq 1 eeprom fault clears eeprom fault irq 0x0a05 0 eeprom complete clears eeprom complete irq table 92. irq clearing for the digital pll address bits bit name description 7 switching clears switching irq 6 closed clears closed irq 5 freerun clears free run irq 4 holdover clears holdover irq 3 frequency unlocked clears frequency unlocked irq 2 frequency locked clears frequency locked irq 1 phase unlocked clears phase unlocked irq 0x0a06 0 phase locked clears phase locked irq
data sheet AD9557 rev. a | page 81 of 92 table 93. irq clearing for history update, frequency limit, and phase slew limit address bits bit name description [7:5] reserved reserved 4 history updated clears history updated irq 3 frequency unclamped clears frequency unclamped irq 2 frequency clamped clears frequency clamped irq 1 phase slew unlimited clea rs phase slew unlimited irq 0x0a07 0 phase slew limited clea rs phase slew limited irq table 94. irq clearing for reference inputs address bits bit name description 7 reserved reserved 6 refb validated clears refb validated irq 5 refb fault cleared clears refb fault cleared irq 4 refb fault clears refb fault irq 3 reserved reserved 2 refa validated clears refa validated irq 1 refa fault cleared clears refa fault cleared irq 0x0a08 0 refa fault clears refa fault irq 0x0a09 [7:0] reserved reserved incremental phase offset control and manual referenc e validation (register 0x0a0a to register 0x0a0d) table 95. incremental phase offset control address bits bit name description [7:3] reserved reserved 2 reset phase offset resets the incremental phase offset to zero. this is an autoclearing bit. 1 decrement phase offset decrements the incremental phase offset by the amount specified in the incremental phase lock offset step size register (register 0x0312 to register 0x0313). this is an autoclearing bit. 0x0a0a 0 increment phase offset increments the incremental phase offset by the amount specified in the incremental phase lock offset step size register (register 0x0312 to register 0x0313). this is an autoclearing bit. table 96. manual reference validation address bits bit name description [7:2] reserved reserved. 1 force timeout b setting this autoclearing bit em ulates timeout of the validation timer for reference b and allows the user to make refb valid immediately. 0x0a0b 0 force timeout a setting this autoclearing bit emul ates timeout of the validation timer for reference a and allows the user to make refa valid immediately. [7:2] reserved reserved. 1 ref mon override b overrides the reference monitor ref fault signal for reference b. setting this bit forces refb to be invalid and is a useful way to force a refe rence switch away from refb (default: 0b). 0x0a0c 0 ref mon override a overrides the reference monitor ref fault signal for reference a. setting this bit forces refa to be invalid and is a useful way to force a reference switch away from refa (default: 0). [7:2] reserved reserved. 1 ref mon bypass b setting this bit bypasses the reference monitor for reference b and starts the refb validation timer. by first setting this bit, and then setting the force timeout b bit, refb is valid for use by the dpll. however, the user should not set this bit at exactly the same time as the force timeout bit (default: 0). 0x0a0d 0 ref mon bypass a setting this bit bypasses the reference monitor for reference a and starts the refa validation timer. by first setting this bit, and then setting the force timeout b bit, refa is valid for use by the dpll. however, the user should not set this bit at exactly the same time as the force timeout bit (default: 0).
AD9557 data sheet rev. a | page 82 of 92 quick in/out frequency soft pin configurat ion (register 0x0c00 to register 0x0c08) table 97. soft pin program setting address bits bit name description [7:1] reserved reserved 0x0c00 0 enable soft pin section 1 0 (default) = disables the function of soft pin re gisters in soft pin section 1 (register 0x0c01 and register 0x0c02). 1 = enables the function of soft pin registers in soft pin section 1 (register 0x0c01 and register 0x0c02) when the pincontrol pin is low at startup and/or reset. the register in soft pin section 1 configures the part into one of 256 preconfigured input-to- output frequency translations stored in the on-chip rom. the registers in soft pin section 1 (register 0x 0c00 to register 0x0c02) are ignored when the pincontrol pin is high at power-up and/or rese t (which means the hard pin program is enabled). [7:4] output frequency selection selects one of 16 predefined output frequencie s as ouptut frequency of the desired frequency translation and reprogram the free run tw, n2, rf div, and m0 to m3 divider with the value stored in the rom. 0x0c01 [3:0] input frequency selection selects one of 16 predefined input frequencies as the input frequency of the desired frequency translation and reprogram the reference period, r divider, n1, frac1, and mod1 in four ref profiles with the value stored in the rom. 0x0c02 [7:2] reserved reserved. [1:0] system clock pll ref selection selects one of the four predefined system pll references for the desired frequency translation and reprogram the system pll configuration with the value stored in the rom. to load values from rom, user must writ e register 0x0c07[0] = 1 after writing this value. register 0x0c02[1:0] equivalent system clock pll settings, register 0x0100 to register 0x101[3:0] system pll ref bit 1 bit 0 12 bits 1 0 0 24.576 mhz xtal, 2 on, n = 16 2 0 1 49.152 mhz xtal, 2 on, n = 8 3 1 0 24.576 mhz xo, 2 off, n = 32 4 1 1 49.152 mhz xo, 2 off, n = 16 0x0c03 [7:1] reserved reserved. 0 enable soft pin section 2 0 (default) = disables the function of soft pin re gisters in soft pin section 2 (register 0x0c04 to register 0x0c06). 1 = enables the function of soft pin registers in soft pin section 2 (register 0x0c04 to register 0x0c06) when pincontrol pin is low. [7:4] reserved reserved. [3:2] refb frequency scale scales selected input frequency (define d by register 0x0c01[3:0]) for refb. 00 (default) = divide-by-1. 01 = divide-by-4. 10 = divide-by-8. 11 = divide-by-16. for example, if the selected input frequenc y is 622.08 mhz and register 0x0c04[3:2] = 11b, the new input frequency should be 622.08 mhz/16 = 38.8 mhz 0x0c04 [1:0] refa frequency scale scales selected input frequency (define d by register 0x0c01[3:0]) for refa. 00 (default) = divide-by-1. 01 = divide-by-4. 10 = divide-by-8. 11 = divide-by-16. 0x0c05 [7:4] reserved reserved. [3:2] channel 1 output frequency scale scales selected output fr equency (defined by register 0x0c01 [7:4]) for channel divider 1 output. 00 (default) = divide-by-1. 01 = divide-by-4. 10 = divide-by-8. 11 = divide-by-16. [1:0] channel 0 output frequency scale scales selected output fr equency (defined by register 0x0c01 [7:4]) for channel divider 0 output. 00 (default) = divide-by-1. 01 = divide-by-4. 10 = divide-by-8. 11 = divide-by-16.
data sheet AD9557 rev. a | page 83 of 92 address bits bit name description 0x0c06 [7:5] reserved reserved 4 sel high pm base loop filter 0 = base loop filter with normal (70) phase margin (default). 1 = base loop filter with high (88.5) phase margin. (<0.1 db peaking in closed-loop transfer function). [3:2] dpll loop bw scales the dpll loop bw while in soft pin mode. 00 (default) = 50 hz. 01 = 1 hz. 10 = 10 hz. 11 = 100 hz. [1:0] reference input frequency tolerance scales the input frequency tolerance while in soft pin mode. 00 (default) = outer tolerance: 10%; inner tolerance: 8% (for general conditions). 01 = outer tolerance: 12 ppm; inner tolerance: 9.6 ppm (for stratum 3). 10 = outer tolerance: 48 ppm; inner tolerance: 38 ppm (for smc clock standard). 11 = outer tolerance: 200 ppm; inner tolerance: 160 ppm (for xtal system clock). [7:1] reserved reserved. 0x0c07 0 soft pin start transfer autoclearing register. 1 = initiates rom downlo ad without resetting the part/register map. after rom download is complete, this register is reset. [7:1] reserved reserved. 0x0c08 0 soft pin reset autoclearing register; resets the part like soft reset (register 0x0000[5]), except that this reset function initiates a soft pin rom download with out resetting the part/register map. after rom download is complete, this register is pulled back to zero. status readback (register 0x0d00 to register 0x0d14) all bits in register 0x0d00 to register 0x0d14 are read only. to show the latest status, these registers require an i/o update (register 0x0005 = 0x01) immediately before being read. table 98. eeprom status address bits bit name description [7:4] reserved reserved. 3 pin program rom load process the control logic sets this bit when data is being read from the rom. 2 fault detected an error occurred while savi ng data to or loading data from the eeprom. 1 load in progress the control logic sets this bit while data is being read from the eeprom. 0x0d00 0 save in progress the control logic sets this bit while data is bein g written to the eeprom. table 99. sysclk status address bits bit name description 7 reserved reserved. 6 dpll_apll_lock indicates the status of the dpll and apll. 0 = either the dpll or the apll is unlocked. 1 = both the dpll and apll are locked. 5 all plls locked indicates the status of the system clock pll, apll, and dpll. 0 = system clock pll or apll or dpll is unlocked. 1 = all three plls (system clock pll, apll, and dpll) are locked. 4 apll vco status 1 = ok. 0 = off/clocks are missing. 3 apll cal in process the control logic holds this bit set while the amplitude calibration of the apll vco is in progress. 2 apll lock indicates the status of the apll. 0 = unlocked. 1 = locked. 1 system clock stable the control logic sets this bit when the device considers the system clock to be stable (see the system clock stability timer section). 0 = not stable (the system clock stability timer has not expired yet). 1 = stable (the system clock stability timer has expired). 0x0d01 0 sysclk lock detect indicates the status of the system clock pll. 0 = unlocked. 1 = locked.
AD9557 data sheet rev. a | page 84 of 92 irq monitor (register 0x0d02 to register 0x0d07 if not masked via the irq mask registers (register 0x0209 and register 0x020a), the appropriate irq monitor bit is set to logic 1 when the indicated event occurs. these bits are cleared only via the irq clearing registers (register 0x0a04 to register 0a0b), the rese t all irqs bit (register 0x0a03[1]), or a device reset. table 100. irq monitor for sysclk address bits bit name description [7:6] reserved reserved. 5 sysclk unlocked indicates a sysclk pll st ate transition from locked to unlocked 4 sysclk locked indicates a sysclk pll stat e transition from unlocked to locked 3 apll unlocked indicates an output pll state transition from locked to unlocked 2 apll locked indicates an output pll state transition from unlocked to locked 1 apll cal ended indicates that apll calibration is complete 0x0d02 0 apll cal started indicates that apll in apll calibration has begun table 101. irq monitor for distribution sync, watchdog timer and eeprom address bits bit name description [7:5] reserved reserved 4 pin program end indicates successful completion of a rom load operation 3 output distribution sync indicates a distribution sync event 2 watchdog timer indicates expiration of the watchdog timer 1 eeprom fault indicates a fault during an eeprom load or save operation 0x0d03 0 eeprom complete indicates successful completi on of an eeprom load or save operation table 102. irq monitor for the digital pll address bits bit name description 7 switching indicates that the dpll is switching to a new reference 6 closed indicates that the dpll has entered closed-loop operation 5 freerun indicates that the dpll has entered free run mode 4 holdover indicates that the dpll has entered holdover mode 3 frequency unlocked indicates that the dpll has lost frequency lock 2 frequency locked indicates that the dpll has acquired frequency lock 1 phase unlocked indicates that the dpll has lost phase lock 0x0d04 0 phase locked indicates that the dpll has acquired phase lock table 103. irq monitor for history update , frequency limit and phase slew limit address bits bit name description [7:5] reserved reserved 4 history updated indicates the occurrence of a tuning word history update 3 frequency unclamped indicates a frequency limiter state transition from clamped to unclamped 2 frequency clamped indicates a frequency limiter state transition from unclamped to clamped 1 phase slew unlimited indicates a phase slew limiter st ate transition from slew limiting to not slew limiting 0x0d05 0 phase slew limited indicates a phase slew limiter state transition from not slew limiting to slew limiting table 104. irq monitor for reference inputs address bits bit name description 7 reserved reserved 6 refb validated indicates that refb has been validated 5 refb fault cleared indicates that refb has been cleared of a previous fault 4 refb fault indicates that refb has been faulted 3 reserved reserved 2 refa validated indicates that refa has been validated 1 refa fault cleared indicates that refa has been cleared of a previous fault 0x0d06 0 refa fault indicates that refa has been faulted 0x0d07 [7:0] reserved reserved
data sheet AD9557 rev. a | page 85 of 92 dpll status, input reference status, holdover history, and dpll lock detect tub levels (register 0x0d08 to register 0x0d14) table 105. dpll status address bits bit name description 7 reserved reserved 6 offset slew limiting the current clos ed-loop phase offset is rate limited 5 frequency lock the dpll has achieved frequency lock 4 phase lock the dpll has achieved phase lock 3 loop switching the dpll is in the process of a reference switchover 2 holdover the dpll is in holdover mode 1 active the dpll is active (that is, operating in a closed-loop condition) 0x0d08 0 freerun the dpll is free run (that is, operating in an open-loop condition) [7:6] reserved default: 0b 5 frequency clamped the upper or lower frequency tuning word clamp is in effect 4 history available there is sufficient tuning word history available for holdover operation [3:2] active reference priority priority value of the currently active reference 00 = highest priority 11 = lowest priority 0x0d09 1 reserved default: 0b 0 current active reference index of the currently active reference 0 = reference a 1 = reference b table 106. reserved register address bits bit name description 0x0d0a [7:0] reserved reserved table 107. input reference status address bits bit name description 7 b valid refb is valid for use (it is unfaulted, and its validation timer has expired). 6 b fault refb is not valid for use. 5 b fast this bit indicates that the frequency of re fb is higher than allowed by its profile settings. 4 b slow this bit indicates that the frequency of re fb is lower than allowed by its profile settings. 3 a valid refa is valid for use (it is unfa ulted and its validation timer has expired). 2 a fault refa is not valid for use. 1 a fast this bit indicates that the frequency of re fa is higher than allowed by its profile settings. 0x0d0b 0 a slow this bit indicates that the frequency of re fa is lower than allowed by its profile settings. 0x0d0c [7:0] reserved reserved. table 108. holdover history 1 address bits bit name description 0x0d0d [7:0] tuning word readback bits[7:0] 0x0d0e [7:0] tuning word readback bits[15:8] 0x0d0f [7:0] tuning word readback bits[23:9] 0x0d10 [7:0] tuning word readback tuning word readback bits[31:24] 1 note that these registers contain the curre nt 30-bit dco frequency tuning word that is generated by the tuning word history lo gic.
AD9557 data sheet rev. a | page 86 of 92 table 109. digital pll lock detect tub levels address bits bit name description 0x0d11 [7:0] read-only digital pll lock detect bathtub level[7:0] (see the dpll frequency lock detector section). 0x0d12 [7:4] reserved. [3:0] phase tub read-only digital pll lock detect bathtub level[11:8] (see the dpll frequency lock detector section). 0x0d13 [7:0] frequency tub read-only digital pll lock detect bathtub level[7:0] (see the dpll phase lock detector section). 0x0d14 [7:4] reserved reserved. [3:0] frequency tub read-only digital pll lock detect bathtub level[11:8] (see the dpll phase lock detector section). eeprom control (register 0x0e00 to register 0x0e3c) table 110. eeprom control address bits bit name description [7:1] reserved reserved. 0x0e00 0 write enable eeprom write enable/protect. 0 (default) = eeprom write protected 1 = eeprom write enabled. [7:4] reserved reserved. 0x0e01 [3:0] conditional value when set to a non-zero value, es tablishes the condition for eepr om downloads. default: 0. [7:1] reserved reserved. 0x0e02 0 save to eeprom uploads data to the eeprom (see the eeprom storage sequence (register 0x0e10 to register 0x0e3c) section). [7:2] reserved reserved. 1 load from eprom downloads data from the eeprom. 0x0e03 0 reserved reserved. eeprom storage sequence (register 0x0e10 to register 0x0e3c) the default settings of register 0x0e10 to register 0x0e3c contain the default eeprom instruction sequence. the tables in this section provide descriptions of the register defaults, assuming that the controller has been instructed to carry out an eeprom storage sequence in which all of the registers are stored and loaded by the eeprom. table 111. eeprom storage sequence for system clock settings address bits bit name description 0x0e10 [7:0] eeprom id the default value of this register is 0x01, which th e controller interprets as a data instruction. its decimal value is 1, so this tells the controller to transfer two bytes of data (1 + 1), beginning at the address specified by the next two bytes. the controller stores 0x01 in the eeprom and increments the eeprom address pointer. 0x0e11 [7:0] 0x0e12 [7:0] the default value of these two registers is 0x0006. note that register 0x0e11 and register 0x0e12 are the most significant and least significant bytes of the target address, respectively. because the previous register contains a data instruction, thes e two registers define a starting address (in this case, 0x0006). the controller stores 0x0006 in the eeprom and increments the eeprom pointer by 2. it then transfers two bytes from the register map (beginning at address 0x0006) to the eeprom and increments the eeprom address pointer by 3 (two data bytes and one checksum byte). the two bytes transferred correspond to the system clock parameters in the register map. 0x0e13 [7:0] system clock the default value of this register is 0x08, which th e controller interprets as a data instruction. its decimal value is 8, so this tells the controller to transfer nine bytes of data (8 + 1), beginning at the address specified by the next two bytes. the controller stores 0x08 in the eeprom and increments the eeprom address pointer. 0x0e14 [7:0] 0x0e15 [7:0] the default value of these two registers is 0x0100. note that register 0x0e14 and register 0x0e15 are the most significant and least significant bytes of the target address, respectively. because the previous register contains a data instruction, thes e two registers define a starting address (in this case, 0x0100). the controller stores 0x0100 in the eeprom and increments the eeprom pointer by 2. it then transfers nine bytes from the register map (beginning at address 0x0100) to the eeprom and increments the eeprom address pointer by 10 (nine data bytes and one checksum byte). the nine bytes transferred correspond to the syst em clock parameters in the register map. 0x0e16 [7:0] i/o update the default value of this register is 0x80, which the controller interprets as an i/o update instruction. the controller stores 0x80 in the eeprom and increments the eeprom address pointer.
data sheet AD9557 rev. a | page 87 of 92 table 112. eeprom storage sequence for general configuration settings address bits bit name description 0x0e17 [7:0] general the default value of this register is 0x11, which the controller interprets as a data instruction. its decimal value is 17, so this tells the controller to transfer 18 bytes of data (17 + 1), beginning at the address specified by the next two bytes. the controller stores 0x11 in the eeprom and increments the eeprom address pointer. 0x0e18 [7:0] 0x0e19 [7:0] the default value of these two registers is 0x0200. note that register 0x0e18 and register 0x0e19 are the most significant and le ast significant bytes of the target address, respectively. because the previous register contains a data instruction, these two registers define a starting address (in this case, 0x0200). the controller stores 0x0200 in the eeprom and increments the eeprom pointer by 2. it then transfers 18 bytes from the register map (beginning at address 0x0200) to the eeprom and increments the eeprom address pointer by 19 (18 data bytes and one checksum byte). the 18 bytes transferred correspond to the general configuration parameters in the register map. table 113. eeprom storage sequence for dpll settings address bits bit name description 0x0e1a [7:0] dpll the default value of this register is 0x2e, which the controller interprets as a data instruction. its decimal value is 46, so this tells the controller to transfer 47 bytes of data (46 + 1), beginning at the address specified by the next two bytes. the controller stores 0x2e in the eeprom and increments the eeprom address pointer. 0x0e1b [7:0] 0x0e1c [7:0] the default value of these two registers is 0x03. note that register 0x0e1b and register 0x0e1c are the most significant and least significant bytes of the target address, respectively. because the previous register contains a data instruction, these two registers define a starting address (in this case, 0x0300). the controller stores 0x0300 in the eeprom and increments the eeprom pointer by 2. it then transfers 47 bytes from the register map (beginning at address 0x0300) to the eeprom and increments the eeprom address pointer by 48 (47 data bytes and one checksum byte). the 47 bytes transferred correspond to the dpll parameters in the register map. table 114. eeprom storage sequence for apll settings address bits bit name description 0x0e1d [7:0] apll the default value of this register is 0x08, which the controller interprets as a data instruction. its decimal value is 8, so this tells the controller to transfer nine bytes of data (8 + 1), beginning at the address specified by the next two bytes. the controller stores 0x08 in the eeprom and increments the eeprom address pointer. 0x0e1e [7:0] 0x0e1f [7:0] the default value of these two registers is 0x0400. note that register 0x0e1e and register 0x0e1f are the most significant and le ast significant bytes of the target address, respectively. because the previous register contains a data instruction, these two registers define a starting address (in this case, 0x040 0). the controller stor es 0x0400 in the eeprom and increments the eeprom pointer by 2. it then transfers nine bytes from the register map (beginning at a ddress 0x0400) to the eeprom and increments the eeprom address pointer by 10 (nine data bytes and one checksum byte). the nine bytes transferred correspond to apll parameters in the register map. table 115. eeprom storage sequence for clock distribution settings address bits bit name description 0x0e20 [7:0] clock distribution the default value of this register is 0x15, which the controller interprets as a data instruction. its decimal value is 21, so this tells the controller to transfer 22 bytes of data (21+1), beginning at the address specified by the next two bytes. the controller stores 0x15 in the eeprom and increments the eeprom address pointer. 0x0e21 [7:0] 0x0e22 [7:0] the default value of these two registers is 0x0500. note that register 0x0e21 and register 0x0e22 are the most significant and le ast significant bytes of the target address, respectively. because the previous register contains a data instruction, these two registers define a starting address (in this case, 0x050 0). the controller stor es 0x0500 in the eeprom and increments the eeprom pointer by 2. it then transfers 22 bytes from the register map (beginning at address 0x0500) to the eeprom and increments the eeprom address pointer by 23 (22 data bytes and one checksum byte). the 22 bytes transferred correspond to the clock distribution parameters in the register map. 0x0e23 [7:0] i/o update the default value of this register is 0x80, whic h the controller interprets as an i/o update instruction. the controller stores 0x80 in the eeprom and incremen ts the eeprom address pointer.
AD9557 data sheet rev. a | page 88 of 92 table 116. eeprom storage sequence for reference input settings address bits bit name description 0x0e24 [7:0] reference inputs the default value of this register is 0x03, which th e controller interprets as a data instruction. its decimal value is 3, so this tells the controller to transfer four bytes of data (3 + 1), beginning at the address specified by the next two bytes. the controller stores 0x03 in the eeprom and increments the eeprom address pointer. 0x0e25 [7:0] 0x0e26 [7:0] the default value of these two registers is 0x0600. note that register 0x0e25 and register 0x0e26 are the most significant and least significant bytes of the target address, respectively. because the previous register contains a data instruction, thes e two registers define a starting address (in this case, 0x0600). the controller stores 0x0600 in the eeprom and increments the eeprom pointer by 2. it then transfers four bytes from the register map (beginning at address 0x0600) to the eeprom and increments the eeprom address pointer by 5 (four data bytes and one checksum byte). the four bytes transferred correspond to the reference inputs parameters in the register map. table 117. reserved address bits bit name description 0x0e27 [7:0] reserved reserved. 0x0e28 [7:0] 0x0e29 [7:0] reserved reserved. table 118. eeprom storage sequence for refa profile settings address bits bit name description 0x0e2a [7:0] refa profile the default value of this register is 0x26, which th e controller interprets as a data instruction. its decimal value is 38, so this tells the controller to transfer 39 bytes of data (38 + 1), beginning at the address specified by the next two bytes. the controller stores 0x26 in the eeprom and increments the eeprom address pointer. 0x0e2b [7:0] 0x0e2c [7:0] the default value of these two registers is 0x0700. note that register 0x0e2b and register 0x0e2c are the most significant and least significant bytes of the target address, respectively. because the previous register contains a data instruction, thes e two registers define a starting address (in this case, 0x0700). the controller stores 0x0700 in the eeprom and increments the eeprom pointer by 2. it then transfers 39 bytes from the register map (beginning at address 0x0700) to the eeprom and increments the eeprom address pointer by 40 (39 data bytes and one checksum byte). the 39 bytes transferred correspond to the refa profile parameters in the register map. table 119. eeprom storage sequence for refb profile settings address bits bit name description 0x0e2d [7:0] refb profile the default value of this register is 0x26, which th e controller interprets as a data instruction. its decimal value is 38, so this tells the controller to transfer 39 bytes of data (38 + 1), beginning at the address specified by the next two bytes. the controller stores 0x26 in the eeprom and increments the eeprom address pointer. 0x0e2e [7:0] 0x0e2f [7:0] the default value of these two registers is 0x0740. note that register 0x0e2e and register 0x0e2f are the most significant and least significant bytes of the target address, respectively. because the previous register contains a data instruction, thes e two registers define a starting address (in this case, 0x0740). the controller stores 0x0740 in the eeprom and increments the eeprom pointer by 2. it then transfers 39 bytes from the register map (beginning at address 0x0740) to the eeprom and increments the eeprom address pointer by 40 (39 data bytes and one checksum byte). the 39 bytes transferred correspond to the refb profile parameters in the register map. 0x0e30 to 0x0e35 [7:0] reserved reserved. 0x0e36 [7:0] i/o update the default value of this register is 0x80, which the controller interprets as an i/o update instruction. the controller stores 0x80 in the eeprom and increments the eeprom address pointer.
data sheet AD9557 rev. a | page 89 of 92 table 120. eeprom storage sequence for operational control settings address bits bit name description 0x0e37 [7:0] operational controls the default value of this register is 0x0d, which th e controller interprets as a data instruction. its decimal value is 13, so this tells the controller to transfer 14 bytes of data (13 + 1), beginning at the address specified by the next two bytes. the controller stores 0x0d in the eeprom and increments the eeprom address pointer. 0x0e38 [7:0] 0x0e39 [7:0] the default value of these two registers is 0x0a00. note that register 0x0e38 and register 0x0e39 are the most significant and least significant bytes of the target address, respectively. because the previous register contains a data instruction, these two registers define a starting address (in this case, 0x0a00). the controller stores 0x0a00 in the eeprom and increments the eeprom pointer by 2. it then transfers 14 bytes from the register map (beginning at address 0x0a00) to the eeprom and increments the eeprom address po inter by 15 (14 data bytes and one checksum byte). the 14 bytes transferred correspond to the op erational controls parameters in the register map. table 121. eeprom storage sequence for apll calibration address bits bit name description 0x0e3a [7:0] calibrate apll the default value of this register is 0xa0, which the controller interprets as a calibrate instruction. the controller stores 0xa0 in the eeprom and increments the eeprom address pointer. 0x0e3b [7:0] i/o update the default value of this register is 0x80, which the controller interprets as an i/o update instruction. the controller stores 0x80 in the eeprom and increments the eeprom address pointer. table 122. eeprom storage sequence for end of data address bits bit name description 0x0e3c [7:0] end of data the default value of this register is 0xff, which the controller interprets as an end instruction. the controller stores this instruction in the eeprom, resets the eeprom address pointer, and enters an idle state. note that if this is a pause ra ther than an end instruction, the controller actions are the same except that the controller increments the eepr om address pointer rather than resetting it. table 123. available for additional eeprom instructions address bits bit name description 0x0e3d to 0xe45 [7:0] unused this area is availabl e for additional eeprom instructions.
AD9557 data sheet rev. a | page 90 of 92 table 124. multifunction pin output functions (d7 = 1) register value output function equivalent status register 0x80 static logic 0 none 0x81 static logic 1 none 0x82 system clock divided by 32 none 0x83 watchdog timer output none 0x84 eeprom upload in progress register 0x0d00, bit 0 0x85 eeprom download in progress register 0x0d00, bit 1 0x86 eeprom fault detected register 0x0d00, bit 2 0x87 sysclk pll lock detected register 0x0d01, bit 0 0x88 sysclk pll stable register 0x0d01, bit 1 0x89 output pll locked register 0x0d01, bit 2 0x8a apll calibration in process register 0x0d01, bit 3 0x8b apll input reference present register 0x0d01, bit 4 0x8c all plls locked register 0x0d01, bit 5 (dpll phase lock) and (apll lock) and (sys pll lock) 0x8d (dpll phase lock) and (apll lo ck) register 0x0d01, bit 6 0x8e reserved 0x8f reserved 0x90 dpll free run register 0x0d08, bit 0 0x91 dpll active register 0x0d08, bit 1 0x92 dpll in holdover register 0x0d08, bit 2 0x93 dpll in reference switchover register 0x0d08, bit 3 0x94 dpll phase locked register 0x0d08, bit 4 0x95 dpll frequency locked register 0x0d08, bit 5 0x96 dpll phase slew limited register 0x0d08, bit 6 0x97 dpll frequency clamped register 0x0d09, bit 5 0x98 tuning word history available register 0x0d09, bit 4 0x99 tuning word history updated register 0x0d05, bit 4 0x9a to 0x9f reserved 0xa0 reference a fault register 0x0d0b, bit 2 0xa1 reference b fault register 0x0d0b, bit 6 0xa2 reserved 0xa3 reserved 0xa4 to ax2f reserved 0xb0 reference a valid register 0x0d0b, bit 3 0xb1 reference b valid register 0x0d0b, bit 7 0xb2 reserved 0xb3 reserved 0xb4 to 0xbf reserved 0xc0 reference a active register 0x0d09, bit 0 0xc1 reference b active register 0x0d09, bit 0 0xc2 reserved 0xc3 reserved 0xc4 to 0xcf reserved 0xd0 clock distribution sync pulse register 0x0d03, bit 3 0xd1 soft pin configuration in process register 0x0d03, bit 4 0xd2 to 0xff reserved
data sheet AD9557 rev. a | page 91 of 92 table 125. multifunction pin input functions (d7 = 0) register value input function equivalent control register 0x00 reserved, high-z input 0x01 i/o update register 0x0005, bit 0 0x02 full power-down register 0x0a00, bit 0 0x03 clear watchdog register 0x0a03, bit 0 0x04 clear all irqs register 0x0a03, bit 1 0x05 tuning word history reset register 0x0a03, bit 2 0x06 to 0x0e reserved 0x10 user holdover register 0x0a01, bit 6 0x11 user free run register 0x0a01, bit 5 0x12 reset incremental phase offset register 0x0a0a, bit 2 0x13 increment incremental phase offset register 0x0a0a, bit 0 0x14 decrement incremental phase offset register 0x0a0a, bit 1 0x15 to 0x1f reserved 0x20 override reference monitor a register 0x0a0c, bit 0 0x21 override reference monitor b register 0x0a0c, bit 1 0x22 to 0x2f reserved 0x30 force validation timeout a register 0x0a0b, bit 0 0x31 force validation timeout b register 0x0a0b, bit 1 0x32 to 0x3f reserved 0x40 enable out0 register 0x0501, bit 0 0x41 enable out1 register 0x0505, bit 0 0x42 to 0x45 reserved 0x46 enable out0 and out1 register 0x0501 and register 0x0505, bit 0 0x47 sync clock distribution o utputs register 0x0a02, bit 1 0x48 to 0xff reserved
AD9557 data sheet rev. a | page 92 of 92 outline dimensions compliant to jedec standards mo-220-vjjd-2 1 40 10 29 28 20 19 4.65 4.50 sq 4.35 top view coplanarity 0.08 4.50 ref 0.50 0.40 0.30 0.50 bsc pin 1 indicator 0.60 max 0.60 max 0.25 min exposed pad bottom view pin 1 indicator 0.30 0.23 0.18 0.20 ref 12 max 0.80 max 0.65 typ 1.00 0.85 0.80 0.05 max 0.02 nom seating plane for proper connection of the exposed pad, refer to the pin configuration and function descriptions section of this data sheet. 05-19-2010-a 6.10 6.00 sq 5.90 5.85 5.75 sq 5.65 figure 55. 40-lead lead frame chip scale package [lfcsp_vq] 6 mm 6 mm body, very thin quad (cp-40-13) dimensions shown in millimeters ordering guide model 1 temperature range package description package option AD9557bcpz ?40c to +85c 40-lead lead frame chip scale package [lfcsp_vq] cp-40-13 AD9557bcpz-reel7 ?40c to +85c 40-lead lead frame chip scale package [lfcsp_vq] cp-40-13 AD9557/pcbz ?40c to +85c evaluation board cp-40-13 1 z = rohs compliant part. i 2 c refers to a communications protocol originally developed by philips semiconductors (now nxp semiconductors). ?2011C2012 analog devices, inc. all rights reserved. trademarks and registered trademarks are the property of their respective owners. d09197-0-3/12(a)

▲Up To Search▲

Price & Availability of AD9557

	To Download AD9557 Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .