View WM8912GEFLV_4743169.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

w wm8912 ultra low power dac with headphone driver for portable audio applications wolfson microelectronics plc to receive regular email updates, sign up at http://www.wolfsonmicro.com/enews production data, september 2010, rev 4.0 copyright ? 2010 wolfson microelectronics plc description the wm8912 is a high performance ultra-low power stereo dac optimised for portable audio applications. the device features stereo ground-referenced headphone amplifiers using the wolfson ?class-w? amplifier techniques - incorporating an innovative dual-mode charge pump architecture - to optimise efficiency and power consumption during playback. the ground-referenced outputs eliminate headphone coupling capacitors. the outputs include common mode feedback paths to reject ground noise. control sequences for audio path setup can be pre-loaded and executed by an integrated control write sequencer to reduce software driver development and minimise pops and clicks via wolfson?s silentswitch? technology. a dynamic range controller provides compression and level control to support a wide range of portable recording applications. anti-clip and quick release features offer good performance in the presence of loud impulsive noises. retune tm mobile 5-band parametric equaliser with fully programmable coefficients is integrated for optimization of speaker characteristics. common audio sampling frequencies are supported from a wide range of external clocks, either directly or generated using the integrated fll. the wm8912 can operate directly from a single 1.8v switched supply. for optimal power consumption, the digital core can be operated from a 1.0v supply. features ? 3.8mw quiescent power consumption for dac to headphone playback ? dac snr 96db typical, thd -86db typical ? class w ground-referenced headphone driver - 28mw per channel into 30 at <1% thd - 32mw per channel into 15 at <1% thd ? dynamic range controller ? retune? mobile parametric equalizer ? integrated control write sequencer for pop minimised start- up and shutdown ? single register write for default start-up and shutdown sequences ? on-chip fll provides all necessary clocks ? dac supports standard sample rates from 8khz to 96khz ? 32-pin qfn package (4x4mm, 0.4mm pitch) applications ? portable multimedia players ? multimedia handsets ? handheld gaming
wm8912 production data w pd, rev 4.0, september 2010 2 block diagram
production data wm8912 w pd, rev 4.0, september 2010 3 table of contents description ....................................................................................................... 1 features............................................................................................................. 1 applications ..................................................................................................... 1 block diagram ................................................................................................. 2 table of contents ......................................................................................... 3 pin configuration ........................................................................................... 6 ordering information .................................................................................. 6 pin description ................................................................................................ 7 absolute maximum ratings ......................................................................... 8 recommended operating conditions ..................................................... 8 electrical characteristics ...................................................................... 9 terminology ............................................................................................................ 9 common test conditions .................................................................................... 9 output signal path ............................................................................................. 10 charge pump .......................................................................................................... 11 fll ........................................................................................................................... ... 11 other parameters .............................................................................................. 11 power consumption .................................................................................... 12 common test conditions .................................................................................. 12 power consumption measurements ............................................................ 13 signal timing requirements ..................................................................... 14 common test conditions .................................................................................. 14 master clock ........................................................................................................ 14 audio interface timing ...................................................................................... 15 master mode ................................................................................................................... ..................................... 15 slave mode .................................................................................................................... ....................................... 16 control interface timing ................................................................................ 17 digital filter characteristics ............................................................... 18 dac filter responses......................................................................................... 19 de-emphasis filter responses ........................................................................ 20 device description ....................................................................................... 21 introduction ......................................................................................................... 21 dynamic range control (drc) .................. ....................................................... 22 compression/limiting capabilities ............................................................................................. .................. 22 gain limits ................................................................................................................... .......................................... 24 dynamic characteristics ....................................................................................................... ......................... 24 anti-clip control ............................................................................................................. .................................. 25 quick release control ......................................................................................................... ........................... 26 gain smoothing ................................................................................................................ ................................... 26 initialisation ................................................................................................................ ........................................ 27 retune tm mobile parametric equalizer (eq) .............................................. 28 default mode (5-band parametric eq) ........................................................................................... ............. 28 retune tm mobile mode .................................................................................................................. ..................... 29 eq filter characteristics ..................................................................................................... ......................... 29 digital mixing ......................................................................................................... 31 dac interface routing and control ............................................................................................. ............. 31 dac interface volume boost .................................................................................................... ..................... 31
wm8912 production data w pd, rev 4.0, september 2010 4 digital-to-analogue converter (dac) ......................................................... 32 dac digital volume control .................................................................................................... ...................... 32 dac soft mute and soft un-mute .................................................................................................................. 34 dac mono mix .................................................................................................................. ...................................... 35 dac de-emphasis ............................................................................................................... ................................... 35 dac sloping stopband filter ................................................................................................... ...................... 36 dac oversampling ratio (osr) .................................................................................................. ..................... 36 output signal path ............................................................................................. 37 output signal paths enable .................................................................................................... ...................... 38 headphone / line output signal paths enable ................................................................................... ..... 38 output volume control ......................................................................................................... ......................... 41 analogue outputs............................................................................................... 44 headphone outputs ? hpoutl and hpoutr ......................................................................................... ....... 44 line outputs ? lineoutl and lineoutr .......................................................................................... .............. 44 external components for ground referenced outputs.................................................................. 45 reference voltages and master bias ......................................................... 46 charge pump .......................................................................................................... 46 dc servo .................................................................................................................. 48 dc servo enable and start-up .................................................................................................. .................... 48 dc servo active modes ......................................................................................................... ........................... 51 dc servo readback.............................................................................................................. .............................. 53 digital audio interface ..................................................................................... 53 master and slave mode operation................................................................................................ .............. 53 operation with tdm ............................................................................................................ ............................... 54 bclk frequency ................................................................................................................ .................................. 55 audio data formats (normal mode) .............................................................................................. ............... 55 audio data formats (tdm mode) ................................................................................................. ................... 57 digital audio interface control ................................................................... 59 audio interface output tri-state .............................................................................................. ................. 59 bclk and lrclk control ........................................................................................................ .......................... 60 companding .................................................................................................................... ...................................... 61 digital pull-up and pull-down ................................................................................................. ..................... 63 clocking and sample rates ............................................................................. 64 sysclk control ................................................................................................................ .................................. 65 control interface clocking .................................................................................................... ..................... 66 clocking configuration ........................................................................................................ ......................... 66 dac clock control ............................................................................................................. ............................... 67 opclk control ................................................................................................................. ................................... 67 toclk control ................................................................................................................. ................................... 68 dac operation at 88.2k / 96k .................................................................................................. .......................... 69 frequency locked loop (fll) .......................................................................... 70 free-running fll clock ........................................................................................................ ........................... 73 example fll calculation ....................................................................................................... .......................... 74 gpio outputs from fll ......................................................................................................... ............................ 75 example fll settings .......................................................................................................... .............................. 75 general purpose input/output (gpio).......................................................... 76 irq/gpio1 ..................................................................................................................... ........................................... 76 bclk/gpio4 .................................................................................................................... ......................................... 77 interrupts ............................................................................................................. 78 control interface .............................................................................................. 80 control write sequencer ............................................................................... 83 initiating a sequence ......................................................................................................... ............................... 83 programming a sequence ........................................................................................................ ....................... 84 default sequences ............................................................................................................. .............................. 86
production data wm8912 w pd, rev 4.0, september 2010 5 start-up sequence ............................................................................................................. ............................... 87 shutdown sequence ............................................................................................................. ............................ 88 power-on reset ................................................................................................... 90 quick start-up and shutdown ........................................................................ 91 quick start-up (default sequence) ............................................................................................. ............... 91 fast start-up from standby .................................................................................................... ...................... 92 quick shutdown (default sequence) ............................................................................................. ............ 93 software reset and chip id ............................................................................. 93 register map ................................................................................................... 94 register bits by address ... .............................................................................. 97 applications information ....................................................................... 126 recommended external components ........................................................ 126 package dimensions .................................................................................. 127 important notice ........................................................................................ 128 address .......................................................................................................... 128
wm8912 production data w pd, rev 4.0, september 2010 6 pin configuration the wm8912 is supplied in a 32-pin qfn package. ordering information device temperature range package moisture sensitivity level peak soldering temperature wm8912gefl/v -40c to +85c 32-lead qfn (4x4x0.4mm, lead-free) msl3 260c wm8912gefl/rv -40c to +85c 32-lead qfn (4x4x0.4mm, lead-free, tape and reel) msl3 260c note: reel quantity = 3,500
production data wm8912 w pd, rev 4.0, september 2010 7 pin description name qfn-32 type description irq/gpio1 1 digital input / output gpio1 / interrupt sclk 2 digital input control interface clock input sda 3 digital input / output control interface data input / output dbvdd 4 supply digital buffer supply (powers audio interface and control interface) dgnd 5 supply digital ground (return path for dcvdd and dbvdd) dcvdd 6 supply digital core supply cpvdd 7 supply charge pump power supply cpca 8 analogue input charge pump flyback capacitor pin cpgnd 9 supply charge pump ground cpcb 10 analogue input charge pump flyback capacitor pin cpvoutp 11 analogue output charge pump positive supply decoupling (powers hpoutl/r, lineoutl/r) cpvoutn 12 analogue output charge pump negative supply decoupling (powers hpoutl/r, lineoutl/r) hpoutl 13 analogue output left headphone output (line or headphone output) hpoutfb 14 analogue output headphone output ground loop noise rejection feedback hpoutr 15 analogue output right headphone output (line or headphone output) lineoutl 16 analogue output left line output 1 (line output) lineoutfb 17 analogue output line output ground loop noise rejection feedback lineoutr 18 analogue output right line output 1 (line output) dnc 19 n/a do not connect dnc 20 n/a do not connect vmidc 21 analogue output midrail voltage decoupling capacitor agnd 22 supply analogue power return avdd 23 supply analogue power supply dnc 24 n/a do not connect dnc 25 n/a do not connect dnc 26 n/a do not connect dnc 27 n/a do not connect mclk 28 digital input master clock for dac bclk/gpio4 29 digital input / output audio interface bit clock / gpio4 lrclk 30 digital input / output audio interface left / right clock dnc 31 n/a do not connect dacdat 32 digital input dac digital audio data gnd_paddle 33 die paddle note: it is recommended that the qfn ground paddle is connected to analogue ground on the application pcb.
wm8912 production data w pd, rev 4.0, september 2010 8 absolute maximum ratings absolute maximum ratings are stress ratings only. permanent damage to the device may be caused by continuously operating at or beyond these limits. device functional operating limits and guaranteed performance specifications are given under electrical characteristics at the test conditions specified. esd sensitive device. this device is manufactured on a cmos process. it is therefore generically susceptible to damage from excessive static voltages. proper esd precautions must be taken during handling and storage of this device. wolfson tests its package types according to ipc/jedec j-std-020b for moisture sensitivity to determine acceptable storage conditions prior to surface mount assembly. these levels are: msl1 = unlimited floor life at <30 c / 85% relative humidity. not normally stored in moisture barrier bag. msl2 = out of bag storage for 1 year at <30 c / 60% relative humidity. supplied in moisture barrier bag. msl3 = out of bag storage for 168 hours at <30 c / 60% relative humidity. supplied in moisture barrier bag. the moisture sensitivity level for each package type is specified in ordering information. condition min max avdd, dcvdd -0.3v +2.5v dbvdd, -0.3v +4.5v cpvdd -0.3v +2.2v hpoutl, hpoutr, lineoutl, lineoutr (cpvdd + 0.3v) * -1 cpvdd + 0.3v voltage range digital inputs dgnd -0.3v dbvdd +0.3v voltage range analogue inputs agnd -0.3v avdd +0.3v temperature range, t a -40 c +85 c storage temperature after soldering -65 c +150 c notes: 1. analogue and digital grounds must always be within 0.3v of each other. 2. all digital and analogue supplies are completely independent from each other; there is no restriction on power supply sequencing. 3. hpoutl, hpoutr, lineoutl, lineoutr are outputs, and should not normally become connected to dc levels. however, if the limits above are exceeded, then damage to the wm8912 may occur. recommended operating conditions parameter symbol min typ max unit digital supply range (core) dcvdd 0.95 1.0 1.98 v digital supply range (buffer) dbvdd 1.42 1.8 3.6 v analogue supplies range avdd 1.71 1.8 2.0 v charge pump supply range cpvdd 1.71 1.8 2.0 v ground dgnd, agnd, cpgnd 0 v operating temperature (ambient) t a -40 +25 +85 c
production data wm8912 w pd, rev 4.0, september 2010 9 electrical characteristics terminology 1. signal-to-noise ratio (db) ? snr is the difference in level between a full scale output signal and the device output noise with no signal applied, measured over a bandwidth of 20hz to 20khz. this ratio is also called idle channel noise. (no auto-zero or automute function is employed). 2. total harmonic distortion (db) ? thd is the difference in level between a 1khz full scale sinewave output signal and the first seven harmonics of the output signal. the amplitude of the fundamental frequency of the output signal is compared to the rms value of the next seven harmonics and expressed as a ratio. 3. total harmonic distortion + noise (db) ? thd+n is the difference in level between a 1khz full scale sine wave output signal and all noise and distortion products in the audio band. the amplitude of the fundamental reference frequency of the output signal is compared to the rms value of all other noise and distortion products and expressed as a ratio. 4. channel separation (db) ? is a measure of the coupling between left and right channels. a full scale signal is applied to the left channel only, the right channel amplitude is measured. then a full scale signal is applied to the right channel only and the left channel amplitude is measured. the worst case channel separation is quoted as a ratio. 5. channel level matching (db) ? measures the difference in gain between the left and the right channels. 6. power supply rejection ratio (db) ? psrr is a measure of ripple attenuation between the power supply pin and an output path. with the signal path idle, a small signal sine wave is summed onto the power supply rail, the amplitude of the sine wave is measured at the output port and expressed as a ratio. 7. all performance measurements carried out with 20khz aes17 low pass filter for distortion measurements, and an a-weighted filter for noise measurement. failure to use such a filter will result in higher thd and lower snr and dynamic range readings than are found in the electrical characteristics. the low pass filter removes out of band noise; although it is not audible it may affect dynamic specification values. common test conditions unless otherwise stated, the following test conditions apply throughout the following sections: ? dcvdd = 1.0v ? dbvdd = 1.8v ? avdd = cpvdd =1.8v ? ambient temperature = +25c ? audio signal: 1khz sine wave, sampled at 48khz with 24-bit data resolution ? sysclk_src = 0 (system clock comes direct from mclk, not from fll). additional, specific test conditions are given within the relevant sections below.
wm8912 production data w pd, rev 4.0, september 2010 10 output signal path stereo playb ack to headphones - dac input to hpoutl+hpoutr pins with 15 load test conditions: hpoutl_vol = hpoutr_vol = 111001b (0db) parameter symbol test conditions min typ max unit output power (per channel) p o 1% thd r load = 30 28 0.92 -0.76 mw vrms dbv 1% thd r load = 15 32 0.69 -3.19 mw vrms dbv dc offset dc servo enabled, calibration complete. -1.5 +1.5 mv signal to noise ratio snr a-weighted 90 96 db total harmonic distortion + noise thd+n r l =30 ; p o =2mw -91 db r l =30 ; p o =20mw -84 r l =15 ; p o =2mw -87 -80 r l =15 ; p o =20mw -85 channel separation 1khz signal, 0dbfs 100 db 10khz signal, 0dbfs 90 channel level matching 1khz signal, 0dbfs +/-1 db power supply rejection ratio psrr 217hz, 100mvpk-pk 75 db 1khz, 100mv pk-pk 70 stereo playback to line-out - dac input to lineoutl+lineoutr pins with 10k / 50pf load test conditions: lineoutl_vol = lineoutr_vol = 111001b (0db) parameter symbol test conditions min typ max unit full scale output signal level dac 0dbfs output at 0db volume 1.0 0 2.83 vrms dbv vpk-pk dc offset dc servo enabled. calibration complete. -1.5 +1.5 mv signal to noise ratio snr a-weighted 90 96 db total harmonic distortion + noise thd+n 10k load -85 -70 db channel separation 1khz signal, 0dbfs 100 db 10khz signal, 0dbfs 90 channel level matching 1khz signal, 0dbfs +/-1 db power supply rejection ratio psrr 217hz, 100mvpk-pk 62 db 1khz, 100mv pk-pk 62 output pgas (hp, line) parameter test conditions min typ max unit minimum pga gain setting -57 db maximum pga gain setting 6 db pga gain step size 1 db pga gain accuracy +6db to -40db -1.5 +1.5 db pga gain accuracy -40db to -57db -1 +1 db mute attenuation hpoutl/r 85 db lineoutl/r 85 db
production data wm8912 w pd, rev 4.0, september 2010 11 charge pump parameter test conditions min typ max unit start-up time 260 s cpca normal mode cpvdd v low power mode cpvdd/2 v cpcb normal mode -cpvdd v low power mode -cpvdd/2 v external component requirements to achieve specified headphone output power and performance flyback capacitor (between cpca and cpcb) at 2v 1 2.2 f cpvoutn capacitor at 2v 2 2.2 f cpvoutp capacitor at 2v 2 2.2 f fll parameter symbol test conditions min typ max unit input frequency f ref fll_clk_ref_div = 00 0.032 13.5 mhz fll_clk_ref_div = 01 0.064 27 mhz lock time 2 ms free-running mode start-up time vmid enabled 100 s free-running mode frequency accuracy reference supplied initially +/-10 % no reference provided +/-30 % other parameters vmid reference parameter test conditions min typ max unit midrail reference voltage (vmid pin) -3% avdd/2 +3% v charge up time (from fully discharged to +5% or -10% of vmid) external capacitor 4.7 f 890 s digital inputs / outputs parameter symbol test conditions min typ max unit input high level v ih 0.7 dbvdd v input low level v il 0.3 dbvdd v output high level v oh i oh = +1ma 0.9 dbvdd v output low level v ol i ol = -1ma 0.1 dbvdd v
wm8912 production data w pd, rev 4.0, september 2010 12 power consumption the wm8912 power consumption is dependent on many parameters. most significantly, it depends on supply voltages, sample rates, mode of operation, and output loading. the power consumption on each supply rail varies approximately with the square of the voltage. power consumption is greater at fast sample rates than at slower ones. when the digital audio interface is operating in master mode, the dbvdd current is significantly greater than in slave mode. (note also that power savings can be made by using mclk as the bclk source in slave mode.) the output load conditions (impedance, capacitance and inductance) can also impact significantly on the device power consumption. common test conditions unless otherwise stated, the following test conditions apply throughout the following sections: ? ambient temperature = +25c ? audio signal = quiescent (zero amplitude) ? sample rate = 48khz ? mclk = 12.288mhz ? audio interface mode = slave (lrclk_dir=0, bclk_dir=0) ? sysclk_src = 0 (system clock comes direct from mclk, not from fll) additional, variant test conditions are quoted within the relevant sections below. where applicable, power dissipated in the headphone or line loads is included.
production data wm8912 w pd, rev 4.0, september 2010 13 power consumption measurements stereo playb ack to h eadphones - dac input to hpoutl+hpoutr pins with 30 ? load. test conditions: vmid_res = 01 (for normal operation) cp_dyn_pwr = 1 (class-w, charge pump controlled by real-time audio level) variant test conditions avdd dcvdd dbvdd cpvdd total v ma v ma v ma v ma mw 48khz sample rate 1.80 1.69 1.00 0.76 1.80 0.00 1.80 0.31 4.38 8khz sample rate 1.80 1.69 1.00 0.18 1.80 0.00 1.80 0.31 3.80 48khz, po = 0.1mw/channel 1khz sine wave 0dbfs hpout_vol= -25db dac_vol= 0db 1.80 1.71 1.00 0.77 1.80 0.00 1.80 1.99 7.45 48khz, po = 1mw/channel 1khz sine wave 0dbfs hpout_vol= -15db dac_vol= 0db 1.80 1.73 1.00 0.77 1.80 0.00 1.80 5.61 13.99 48khz sample rate, master mode, fll enabled, mclk input frequency = 13mhz 1.80 1.82 1.00 1.05 1.80 0.73 1.80 0.30 6.18 48khz sample rate, master mode, fll enabled, mclk input frequency = 32.768khz 1.80 1.83 1.00 0.94 1.80 0.76 1.80 0.29 6.14 stereo playback to line-out - dac input to lineoutl+lineoutr or hpoutl+hpoutr pins with 10k ? / 50pf load test conditions : vmid_res = 01 (for normal operation) cp_dyn_pwr = 1 (class-w, charge pump controlled by real-time audio level) variant test conditions avdd dcvdd dbvdd cpvdd total v ma v ma v ma v ma mw 48khz sample rate 1.8 1.67 1 0.76 1.8 0.00 1.8 0.36 4.43 8khz sample rate 1.8 1.67 1 0.18 1.8 0.00 1.8 0.36 3.86 48khz, po = 0dbfs 1khz sine wave 1.8 1.78 1 0.77 1.8 0.00 1.8 2.27 8.09 off note: dc servo calibration is retained in this state as long as dcvdd is supplied. this allows fast, pop suppressed start-up from the off state. variant test conditions avdd dcvdd dbvdd cpvdd total v ma v ma v ma v ma mw off (default settings) no clocks applied 1.8 0.01 1 0.00 1.8 0.00 1.8 0.01 0.04 off (default settings) dacdat, mclk, bclk, and lrclk applied 1.8 0.01 1 0.02 1.8 0.00 1.8 0.01 0.06
wm8912 production data w pd, rev 4.0, september 2010 14 signal timing requirements common test conditions unless otherwise stated, the following test conditions apply throughout the following sections: ? ambient temperature = +25c ? dcvdd = 1.0v ? dbvdd = avdd = cpvdd = 1.8v ? dgnd = agnd = cpgnd = 0v additional, specific test conditions are given within the relevant sections below. master clock mclk t mclkl t mclkh t mclky figure 1 master clock timing master clock timing parameter symbol test conditions min typ max unit mclk cycle time t mclky mclk_div=1 40 ns mclk_div=0 80 ns mclk duty cycle t mclkds 60:40 40:60
production data wm8912 w pd, rev 4.0, september 2010 15 audio interface timing master mode bclk (output) lrclk (output) dacdat (input) t dht t dst t dl figure 2 audio interface timing ? master mode test conditions dcvdd = 1.0v, avdd = dbvdd = cpvdd = 1.8v, dgnd=agnd=cpgnd =0v, t a = +25 o c, master mode, fs=48khz, mclk=256fs, 24-bit data, unless otherwise stated. parameter symbol min typ max unit audio interface timing - master mode lrclk propagation delay from bclk falling edge t dl 20 ns dacdat setup time to bclk rising edge t dst 20 ns dacdat hold time from bclk rising edge t dht 10 ns
wm8912 production data w pd, rev 4.0, september 2010 16 slave mode bclk (input) lrclk (input) dacdat (input) t ds t dh t lrh t lrsu t bch t bcl t bcy figure 3 audio interface timing ? slave mode test conditions dcvdd = 1.0v, avdd = dbvdd = cpvdd = 1.8v, dgnd=agnd=cpgnd =0v, t a = +25 o c, slave mode, fs=48khz, mclk=256fs, 24-bit data, unless otherwise stated. parameter symbol min typ max unit audio interface timing - slave mode bclk cycle time t bcy 50 ns bclk pulse width high t bch 20 ns bclk pulse width low t bcl 20 ns lrclk set-up time to bclk rising edge t lrsu 20 ns lrclk hold time from bclk rising edge t lrh 10 ns dacdat hold time from bclk rising edge t dh 10 ns dacdat set-up time to bclk rising edge t ds 20 ns note: bclk period must always be greater than or equal to mclk period.
production data wm8912 w pd, rev 4.0, september 2010 17 control interface timing figure 4 control interface timing test conditions dcvdd = 1.0v, avdd = dbvdd = cpvdd = 1.8v, dgnd=agnd=cpgnd =0v, t a =+25 o c, slave mode, fs=48khz, mclk = 256fs, 24-bit data, unless otherwise stated. parameter symbol min typ max unit sclk frequency 400 khz sclk low pulse-width t 1 1300 ns sclk high pulse-width t 2 600 ns hold time (start condition) t 3 600 ns setup time (start condition) t 4 600 ns data setup time t 5 100 ns sda, sclk rise time t 6 300 ns sda, sclk fall time t 7 300 ns setup time (stop condition) t 8 600 ns data hold time t 9 900 ns pulse width of spikes that will be suppressed t ps 0 5 ns
wm8912 production data w pd, rev 4.0, september 2010 18 digital filter characteristics parameter test conditions min typ max unit dac normal filter passband +/- 0.05db 0 0.454 fs -6db 0.5 fs passband ripple 0.454 fs +/- 0.03 db stopband 0.546 fs stopband attenuation f > 0.546 fs -50 db dac sloping stopband filter passband +/- 0.03db 0 0.25 fs +/- 1db 0.25 fs 0.454 fs -6db 0.5 fs passband ripple 0.25 fs +/- 0.03 db stopband 1 0.546 fs 0.7 fs stopband 1 attenuation f > 0.546 fs -60 db stopband 2 0.7 fs 1.4 fs stopband 2 attenuation f > 0.7 fs -85 db stopband 3 1.4 fs stopband 3 attenuation f > 1.4 fs -55 db dac filters mode group delay normal 16.5 / fs sloping stopband 18 / fs terminology 1. stop band attenuation (db) ? the degree to which the frequency spectrum is attenuated (outside audio band) 2. pass-band ripple ? any variation of the frequency response in the pass-band region
production data wm8912 w pd, rev 4.0, september 2010 19 dac filter responses magnitude(db) -0.005 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 frequency (fs) figure 5 dac digital filter frequency response; (normal mode); sample rate > 24khz figure 6 dac digital filter ripple (normal mode) magnitude(db) -0.5 -0.45 -0.4 -0.35 -0.3 -0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 frequency (fs) figure 7 dac digital filter frequency response; (sloping stopband mode); sample rate <= 24khz figure 8 dac digital filter ripple (sloping stopband mode)
wm8912 production data w pd, rev 4.0, september 2010 20 de-emphasis filter responses magnitude(db) -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 0 5000 10000 15000 20000 frequency (hz) magnitude(db) -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 frequency (hz) figure 9 de-emphasis digital filter response (32khz) figure 10 de-emphasis error (32khz) magnitude(db) -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 0 5000 10000 15000 20000 25000 frequency (hz) magnitude(db) -0.1 -0.05 0 0.05 0.1 0.15 0.2 0 5000 10000 15000 20000 25000 frequency (hz) figure 11 de-emphasis digital filter response (4 4.1khz) figure 12 de-emphasis error (44.1khz) magnitude(db) -12 -10 -8 -6 -4 -2 0 0 5000 10000 15000 20000 25000 30000 frequency (hz) magnitude(db) -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0 5000 10000 15000 20000 25000 30000 frequency (hz) figure 13 de-emphasis digital filter response (48khz) figure 14 de-emphasis error (48khz)
production data wm8912 w pd, rev 4.0, september 2010 21 device description introduction the wm8912 is a high performance ultra-low power stereo dac optimised for portable audio applications. powerful digital signal processing (dsp) makes it ideal for small portable devices. two stereo pairs of ground-referenced class-w outputs are provided, suitable for driving a stereo headphone and stereo line load simultaneously. the ground-referenced outputs are powered from an integrated charge pump, enabling high quality, power efficient outputs without requirement for dc blocking capacitors. a dc servo circuit is available for dc offset correction, thereby suppressing pops and further reducing power consumption. ground loop feedback is provided on the headphone outputs and on the line outputs, providing rejection of noise on the ground connections. all outputs use wolfson silentswitch? technology for pop and click suppression. the stereo dacs are of hi-fi quality, using a 24-bit low-order oversampling architecture to deliver optimum performance. a flexible clocking arrangement supports all commonly used dac sample rates, either directly from an external mclk or with the use of the integrated frequency locked loop (fll) for additional flexibility. dac soft mute and un-mute is available for pop-free music playback. the integrated dynamic range controller (drc) and retune tm mobile 5-band parametric equaliser (eq) provide further processing capability of the digital audio paths. the drc provides compression and signal level control to improve the handling of unpredictable signal levels. ?anti-clip? and ?quick release? algorithms improve intelligibility in the presence of transients and impulsive noises. the eq provides the capability to tailor the audio path according to the frequency characteristics of an earpiece or loudspeaker, and/or according to user preferences. the wm8912 has a highly flexible digital audio interface, supporting a number of protocols, including i 2 s, dsp, msb-first left/right justified, and can operate in master or slave modes. pcm operation is supported in the dsp mode. a-law and -law companding are also supported. time division multiplexing (tdm) is available to allow multiple devices to stream data simultaneously on the same bus, saving space and power. the system clock (sysclk) provides clocking for the dacs, dsp core, digital audio interface and other circuits. sysclk can be derived directly from the mclk pin or via the integrated fll, providing flexibility to support a wide range of clocking schemes. typical portable system mclk frequencies and commonly used sample rates from 8khz to 48khz are all supported. the clocking circuits are configured automatically from the sample rate and from the mclk / sysclk ratio. the integrated fll can be used to generate sysclk from a wide variety of different reference sources and frequencies. the fll can accept a wide range of reference frequencies, which may be high frequency (e.g. 13mhz) or low frequency (eg. 32.768khz). the fll is tolerant of jitter and may be used to generate a stable sysclk from a less stable input signal. the integrated fll can be used as a free-running oscillator, enabling autonomous clocking of the headphone charge pump and dc servo if required. the wm8912 uses a standard 2-wire control interface, providing full software control of all features, together with device register readback. an integrated control write sequencer enables automatic scheduling of control sequences; commonly-used signal configurations may be selected using ready- programmed sequences, including time-optimised control of the wm8912 pop suppression features. it is an ideal partner for a wide range of industry standard microprocessors, controllers and dsps. unused circuitry can be disabled under software control, in order to save power; low leakage currents enable extended standby/off time in portable battery-powered applications. two gpio pins may be configured for miscellaneous input/output functions such as button/accessory detect inputs, or for clock, system status, or programmable logic level output for control of additional external circuitry. interrupt logic, status readback and de-bouncing options are supported within this functionality.
wm8912 production data w pd, rev 4.0, september 2010 22 dynamic range control (drc) the dynamic range controller (drc) is a circuit which can be enabled in the digital dac playback path. the function of the drc is to adjust the signal gain in conditions where the input amplitude is unknown or varies over a wide range, e.g. when recording from microphones built into a handheld system. the drc can apply compression and automatic level control to the signal path. it incorporates ?anti-clip? and ?quick release? features for handling transients in order to improve intelligibility in the presence of loud impulsive noises. the drc is enabled in the digital dac playback path by setting drc_ena and drc_dac_path, as shown in table 1. both bits must be set for drc operation. register address bit label default description r40 (28h) drc control 0 15 drc_ena 0 drc enable 0 = disabled 1 = enabled 14 drc_dac_pat h 0 drc path select 0 = reserved 1 = dac path table 1 drc enable compression/limiting capabilities the drc supports two different compression regions, separated by a ?knee? at input amplitude t. for signals above the knee, the compression slope drc_hi_comp applies; for signals below the knee, the compression slope drc_lo_comp applies. the overall drc compression characteristic in ?steady state? (i.e. where the input amplitude is near- constant) is illustrated in figure 15. drc_knee_ip (y0) 0db d r c _ h i _ c o m p d r c _ l o _ c o m p drc input amplitude (db) drc output amplitude (db) drc_knee_op ?knee? figure 15 drc compression characteristic
production data wm8912 w pd, rev 4.0, september 2010 23 the slope of the drc response is determined by register fields drc_hi_comp and drc_lo_comp respectively. a slope of 1 indicates constant gain in this region. a slope less than 1 represents compression (i.e. a change in input amplitude produces only a smaller change in output amplitude). a slope of 0 indicates that the target output amplitude is the same across a range of input amplitudes; this is infinite compression. the ?knee? in figure 15 is determined by register fields drc_knee_ip and drc_knee_op. parameter y0, the output level for a 0db input, is not specified directly, but can be calculated from the other parameters, using the equation. the drc compression parameters are defined in table 2. register address bit label default description r43 (2bh) drc control 3 10:5 drc_knee_ip [5:0] 00_0000 input signal at the compressor 'knee'. 000000 = 0db 000001 = -0.75db 000010 = -1.5db ? (-0.75db steps) 111100 = -45db 111101 to 111111 = reserved 4:0 drc_knee_op [4:0] 0_0000 output signal at the compressor 'knee'. 00000 = 0db 00001 = -0.75db 00010 = -1.5db ? (-0.75db steps) 11110 = -22.5db 11111 = reserved r42 (2ah) drc control 2 5:3 drc_hi_comp [2:0] 000 compressor slope (upper region) 000 = 1 (no compression) 001 = 1/2 010 = 1/4 011 = 1/8 100 = 1/16 101 = 0 110 to 111 = reserved 2:0 drc_lo_comp [2:0] 000 compressor slope (lower region) 000 = 1 (no compression) 001 = 1/2 010 = 1/4 011 = 1/8 100 = 0 101 to 111 = reserved table 2 drc compression control
wm8912 production data w pd, rev 4.0, september 2010 24 gain limits the minimum and maximum gain applied by the drc is set by register fields drc_mingain and drc_maxgain. these limits can be used to alter the drc response from that illustrated in figure 15. if the range between maximum and minimum gain is reduced, then the extent of the dynamic range control is reduced. the maximum gain prevents quiet signals (or silence) from being excessively amplified. register address bit label default description r41 (29h) drc control 1 3:2 drc_mingain [1:0] 10 minimum gain the drc can use to attenuate audio signals 00 = 0db (default) 01 = -6db 10 = -12db 11 = -18db 1:0 drc_maxgain [1:0] 00 maximum gain the drc can use to boost audio signals 00 = 12db 01 = 18db (default) 10 = 24db 11 = 36db table 3 drc gain limits dynamic characteristics the dynamic behaviour determines how quickly the drc responds to changing signal levels. note that the drc responds to the average (rms) signal amplitude over a period of time. drc_atk determines how quickly the drc gain decreases when the signal amplitude is high. drc_dcy determines how quickly the drc gain increases when the signal amplitude is low. these register fields are described in table 4. note that the register defaults are suitable for general purpose microphone use. register address bit label default description r41 (29h) drc control 1 15:12 drc_atk [3:0] 0011 gain attack rate (seconds/6db) 0000 = reserved 0001 = 182 s 0010 = 363 s 0011 = 726 s (default) 0100 = 1.45ms 0101 = 2.9ms 0110 = 5.8ms 0111 = 11.6ms 1000 = 23.2ms 1001 = 46.4ms 1010 = 92.8ms 1011-1111 = reserved
production data wm8912 w pd, rev 4.0, september 2010 25 register address bit label default description 11:8 drc_dcy [3:0] 0010 gain decay rate (seconds/6db) 0000 = 186ms 0001 = 372ms 0010 = 743ms (default) 0011 = 1.49s 0100 = 2.97s 0101 = 5.94s 0110 = 11.89s 0111 = 23.78s 1000 = 47.56s 1001-1111 = reserved table 4 drc attack and decay rates note: for detailed information about drc attack and decay rates, please see wolfson application note wan0247. anti-clip control the drc includes an anti-clip feature to avoid signal clipping when the input amplitude rises very quickly. this feature uses a feed-forward technique for early detection of a rising signal level. signal clipping is avoided by dynamically increasing the gain attack rate when required. the anti-clip feature is enabled using the drc_anticlip bit. note that the feed-forward processing increases the latency in the input signal path. for low-latency applications (e.g. telephony), it may be desirable to reduce the delay, although this will also reduce the effectiveness of the anti-clip feature. the latency is determined by the drc_ff_delay bit. if necessary, the latency can be minimised by disabling the anti-clip feature altogether. the drc anti-clip control bits are described in table 5. register address bit label default description r40 (28h) drc control 0 5 drc_ff_delay 1 feed-forward delay for anti-clip feature 0 = 5 samples 1 = 9 samples time delay can be calculated as 5/f s or 9/ f s , where f s is the sample rate. 1 drc_anticlip 1 anti-clip enable 0 = disabled 1 = enabled table 5 drc anti-clip control note that the anti-clip feature operates entirely in the digital domain. it cannot be used to prevent signal clipping in the analogue domain nor in the source signal. analogue clipping can only be prevented by reducing the analogue signal gain or by adjusting the source signal.
wm8912 production data w pd, rev 4.0, september 2010 26 quick release control the drc includes a quick-release feature to handle short transient peaks that are not related to the intended source signal. for example, in handheld microphone recording, transient signal peaks sometimes occur due to user handling, key presses or accidental tapping against the microphone. the quick release feature ensures that these transients do not cause the intended signal to be masked by the longer time constants of drc_dcy. the quick-release feature is enabled by setting the drc_qr bit. when this bit is enabled, the drc measures the crest factor (peak to rms ratio) of the input signal. a high crest factor is indicative of a transient peak that may not be related to the intended source signal. if the crest factor exceeds the level set by drc_qr_thr, then the normal decay rate (drc_dcy) is ignored and a faster decay rate (drc _qr_dcy) is used instead. the drc quick-release control bits are described in table 6. register address bit label default description r40 (28h) drc control 0 2 drc_qr 1 quick release enable 0 = disabled 1 = enabled r41 (29h) drc control 1 7:6 drc_qr_thr [1:0] 01 quick release crest factor threshold 00 = 12db 01 = 18db (default) 10 = 24db 11 = 30db 5:4 drc_qr_dcy [1:0] 00 quick release decay rate (seconds/6db) 00 = 0.725ms (default) 01 = 1.45ms 10 = 5.8ms 11 = reserved table 6 drc quick-release control gain smoothing the drc includes a gain smoothing filter in order to prevent gain ripples. a programmable level of hysteresis is also used to control the drc gain. this improves the handling of very low frequency input signals whose period is close to the drc attack/decay time. drc gain smoothing is enabled by default and it is recommended to use the default register settings. the extent of the gain smoothing filter may be adjusted or disabled using the control fields described in table 7. register address bit label default description r40 (28h) drc control 0 12:11 drc_gs_hyst _lvl [1:0] 00 gain smoothing hysteresis threshold 00 = low 01 = medium (recommended) 10 = high 11 = reserved 3 drc_gs_ena 1 gain smoothing enable 0 = disabled 1 = enabled 0 drc_gs_hyst 1 gain smoothing hysteresis enable 0 = disabled 1 = enabled table 7 drc gain smoothing
production data wm8912 w pd, rev 4.0, september 2010 27 initialisation when the drc is initialised, the gain is set to the level determined by the drc_startup_gain register field. the default setting is 0db, but values from -3db to +6db are available, as described in table 8. register address bit label default description r40 (28h) drc control 0 10:6 drc_startup_ gain [4:0] 00110 initial gain at drc start-up 00000 = -3db 00001 = -2.5db 00010 = -2db 00011 = -1.5db 00100 = -1db 00101 = -0.5db 00110 = 0db (default) 00111 = 0.5db 01000 = 1db 01001 = 1.5db 01010 = 2db 01011 = 2.5db 01100 = 3db 01101 = 3.5db 01110 = 4db 01111 = 4.5db 10000 = 5db 10001 = 5.5db 10010 = 6db 10011 to 11111 = reserved table 8 drc initialisation
wm8912 production data w pd, rev 4.0, september 2010 28 retune tm mobile parametric equalizer (eq) the retune tm mobile parametric equaliser is a circuit that can be enabled in the dac path. the function of the eq is to adjust the frequency characteristic of the output to compensate for unwanted frequency characteristics in the loudspeaker (or other output transducer). it can also be used to tailor the response according to user preferences, for example to accentuate or attenuate specific frequency bands to emulate different sound profiles or environments such as concert hall, rock etc. the eq is enabled using the eq_ena bit as shown in table 9. register address bit label default description r134 (86h) eq1 0 eq_ena 0 eq enable 0 = eq disabled 1 = eq enabled table 9 retune tm mobile parametric eq enable the eq can be configured to operate in two modes - ?default? mode or ?retune tm mobile? mode. default mode (5-band parametric eq) in default mode, the cut-off / centre frequencies are fixed as per table 10. the filter bandwidths are also fixed in default mode. the gain of the individual bands (-12db to +12db) can be controlled as described in table 11. note that the cut-off / centre frequencies noted in table 10 are applicable to a dac sample rate of 48khz. when using other sample rates, these frequencies will be scaled in proportion to the selected sample rate. eq band cut-off/centre frequency 1 100 hz 2 300 hz 3 875 hz 4 2400 hz 5 6900 hz table 10 eq band cut-off / centre frequencies register address bit label default description r135 (87h) eq2 4:0 eq_b1_gain [4:0] 01100b (0db) eq band 1 gain (see table 12 for gain range) r136 (88h) eq3 4:0 eq_b2_gain [4:0] 01100b (0db) eq band 2 gain (see table 12 for gain range) r137 (89h) eq4 4:0 eq_b3_gain [4:0] 01100b (0db) eq band 3 gain (see table 12 for gain range) r138 (8ah) eq5 4:0 eq_b4_gain [4:0] 01100b (0db) eq band 4 gain (see table 12 for gain range) r139 (8bh) eq6 4:0 eq_b5_gain [4:0] 01100b (0db) eq band 5 gain (see table 12 for gain range) table 11 eq band gain control
production data wm8912 w pd, rev 4.0, september 2010 29 eq gain setting gain (db) 00000 -12 00001 -11 00010 -10 00011 -9 00100 -8 00101 -7 00110 -6 00111 -5 01000 -4 01001 -3 01010 -2 01011 -1 01100 0 01101 +1 01110 +2 01111 +3 10000 +4 10001 +5 10010 +6 10011 +7 10100 +8 10101 +9 10110 +10 10111 +11 11000 +12 11001 to 11111 reserved table 12 eq gain control retune tm mobile mode retune tm mobile mode provides a comprehensive facility for the user to define the cut-off/centre frequencies and filter bandwidth for each eq band, in addition to the gain controls already described. this enables the eq to be accurately customised for a specific transducer characteristic or desired sound profile. the eq enable and eq gain controls are the same as defined for the default mode. the additional coefficients used in retune tm mobile mode are held in registers r140 to r157. these coefficients are derived using tools provided in wolfson?s wisce? evaluation board control software. please contact your local wolfson representative for more details. eq filter characteristics the filter characteristics for each frequency band are shown in figure 16 to figure 20. these figures show the frequency response for all available gain settings, using default cut-off/centre frequencies and bandwidth.
wm8912 production data w pd, rev 4.0, september 2010 30 -15 -10 -5 0 5 10 15 1 10 100 1000 10000 100000 frequency (hz) gain (db) -15 -10 -5 0 5 10 15 1 10 100 1000 10000 100000 frequency (hz) gain (db) figure 16 eq band 1 ? low freq shelf filter response figure 17 eq band 2 ? peak filter response -15 -10 -5 0 5 10 15 1 10 100 1000 10000 100000 frequency (hz) gain (db) -15 -10 -5 0 5 10 15 1 10 100 1000 10000 100000 frequency (hz) gain (db) figure 18 eq band 3 ? peak filter response figure 19 eq band 4 ? peak filter response -15 -10 -5 0 5 10 15 1 10 100 1000 10000 100000 frequency (hz) gain (db) figure 20 eq band 5 ? high freq shelf filter response
production data wm8912 w pd, rev 4.0, september 2010 31 digital mixing the digital dac data can be controlled in various ways to support a range of different usage modes. data from either of the digital audio interface channels can be routed to either the left or the right dac. the dacs can be configured as a mono mix of the two audio channels. see "digital audio interface control" for more information on the audio interface. the wm8912 provides a dynamic range control (drc) feature, which can apply compression and gain adjustment in the digital domain to the dac signal path. this is effective in controlling signal levels under conditions where input amplitude is unknown or varies over a wide range. see ?dynamic range control (drc)? for further details. the w8912 also incorporates the retune tm mobile 5-band parametric equaliser with fully programmable coefficients for optimization of speaker characteristics or for tailoring the response according to user preferences. see ?retune tm mobile parametric equalizer (eq)? for further details. dac interface routing and control the input data source for each dac can be changed under software control using register bits aifdacl_src and aifdacr_src. the polarity of each dac input may also be modified using register bits dacl_datinv and dacr_datinv. these register bits are described in table 13. register address bit label default description r24 (18h) audio interface 0 12 dacl_datinv 0 left dac invert 0 = left dac output not inverted 1 = left dac output inverted 11 dacr_datinv 0 right dac invert 0 = right dac output not inverted 1 = right dac output inverted 5 aifdacl_src 0 left dac data source select 0 = left dac outputs left interface data 1 = left dac outputs right interface data 4 aifdacr_src 1 right dac data source select 0 = right dac outputs left interface data 1 = right dac outputs right interface data table 13 dac routing and control dac interface volume boost a digital gain function is available at the audio interface to boost the dac volume when a small signal is received on dacdat. this is controlled using register bits dac_boost [1:0]. to prevent clipping at the dac input, this function should not be used when the boosted dac data is expected to be greater than 0dbfs. the digital interface volume is controlled as shown in table 14. register address bit label default description r24 (18h) audio interface 0 10:9 dac_boost [1:0] 00 dac input volume boost 00 = 0db 01 = +6db (input data must not exceed -6dbfs) 10 = +12db (input data must not exceed -12dbfs) 11 = +18db (input data must not exceed -18dbfs) table 14 dac interface volume boost
wm8912 production data w pd, rev 4.0, september 2010 32 digital-to-analogue converter (dac) the wm8912 dacs receive digital input data from the dacdat pin. the digital audio data is converted to oversampled bit streams in the on-chip, true 24-bit digital interpolation filters. the bitstream data enters two multi-bit, sigma-delta dacs, which convert them to high quality analogue audio signals. the wolfson smartdac? architecture offers reduced power consumption, whilst also delivering a reduction in high frequency noise and sensitivity to clock jitter. it also uses a dynamic element matching technique for high linearity and low distortion. the analogue outputs from the dacs are sent directly to the output pgas (see ?output signal path?). the dacs are enabled by the dacl_ena and dacr_ena register bits. register address bit label default description r18 (12h) power management 6 3 dacl_ena 0 left dac enable 0 = dac disabled 1 = dac enabled 2 dacr_ena 0 right dac enable 0 = dac disabled 1 = dac enabled table 15 dac enable control dac digital volume control the output level of each dac can be controlled digitally over a range from -71.625db to 0db in 0.375db steps. the level of attenuation for an eight-bit code is detailed in table 17. the dac_vu bit controls the loading of digital volume control data. when dac_vu is set to 0, the dacl_vol or dacr_vol control data is loaded into the respective control register, but does not actually change the digital gain setting. both left and right gain settings are updated when a 1 is written to dac_vu. this makes it possible to update the gain of both channels simultaneously. register address bit label default description r30 (1eh) dac digital volume left 8 dac_vu n/a dac volume update writing a 1 to this bit causes left and right dac volume to be updated simultaneously 7:0 dacl_vol [7:0] 1100_0000 (0db) left dac digital volume 00h = mute 01h = -71.625db 02h = -71.250db ? (0.375db steps) c0h to ffh = 0db (see table 17 for volume range) r31 (1fh) dac digital volume right 8 dac_vu n/a dac volume update writing a 1 to this bit causes left and right dac volume to be updated simultaneously 7:0 dacr_vol [7:0] 1100_0000 (0db) right dac digital volume 00h = mute 01h = -71.625db 02h = -71.250db ? (0.375db steps) c0h to ffh = 0db (see table 17 for volume range) table 16 dac digital volume control
production data wm8912 w pd, rev 4.0, september 2010 33 dacl_vol or dacr_vol volume (db) dacl_vol or dacr_vol volume (db) dacl_vol or dacr_vol volume (db) dacl_vol or dacr_vol volume (db) 0h mute 40h -48.000 80h -24.000 c0h 0.000 1h -71.625 41h -47.625 81h -23.625 c1h 0.000 2h -71.250 42h -47.250 82h -23.250 c2h 0.000 3h -70.875 43h -46.875 83h -22.875 c3h 0.000 4h -70.500 44h -46.500 84h -22.500 c4h 0.000 5h -70.125 45h -46.125 85h -22.125 c5h 0.000 6h -69.750 46h -45.750 86h -21.750 c6h 0.000 7h -69.375 47h -45.375 87h -21.375 c7h 0.000 8h -69.000 48h -45.000 88h -21.000 c8h 0.000 9h -68.625 49h -44.625 89h -20.625 c9h 0.000 ah -68.250 4ah -44.250 8ah -20.250 cah 0.000 bh -67.875 4bh -43.875 8bh -19.875 cbh 0.000 ch -67.500 4ch -43.500 8ch -19.500 cch 0.000 dh -67.125 4dh -43.125 8dh -19.125 cdh 0.000 eh -66.750 4eh -42.750 8eh -18.750 ceh 0.000 fh -66.375 4fh -42.375 8fh -18.375 cfh 0.000 10h -66.000 50h -42.000 90h -18.000 d0h 0.000 11h -65.625 51h -41.625 91h -17.625 d1h 0.000 12h -65.250 52h -41.250 92h -17.250 d2h 0.000 13h -64.875 53h -40.875 93h -16.875 d3h 0.000 14h -64.500 54h -40.500 94h -16.500 d4h 0.000 15h -64.125 55h -40.125 95h -16.125 d5h 0.000 16h -63.750 56h -39.750 96h -15.750 d6h 0.000 17h -63.375 57h -39.375 97h -15.375 d7h 0.000 18h -63.000 58h -39.000 98h -15.000 d8h 0.000 19h -62.625 59h -38.625 99h -14.625 d9h 0.000 1ah -62.250 5ah -38.250 9ah -14.250 dah 0.000 1bh -61.875 5bh -37.875 9bh -13.875 dbh 0.000 1ch -61.500 5ch -37.500 9ch -13.500 dch 0.000 1dh -61.125 5dh -37.125 9dh -13.125 ddh 0.000 1eh -60.750 5eh -36.750 9eh -12.750 deh 0.000 1fh -60.375 5fh -36.375 9fh -12.375 dfh 0.000 20h -60.000 60h -36.000 a0h -12.000 e0h 0.000 21h -59.625 61h -35.625 a1h -11.625 e1h 0.000 22h -59.250 62h -35.250 a2h -11.250 e2h 0.000 23h -58.875 63h -34.875 a3h -10.875 e3h 0.000 24h -58.500 64h -34.500 a4h -10.500 e4h 0.000 25h -58.125 65h -34.125 a5h -10.125 e5h 0.000 26h -57.750 66h -33.750 a6h -9.750 e6h 0.000 27h -57.375 67h -33.375 a7h -9.375 e7h 0.000 28h -57.000 68h -33.000 a8h -9.000 e8h 0.000 29h -56.625 69h -32.625 a9h -8.625 e9h 0.000 2ah -56.250 6ah -32.250 aah -8.250 eah 0.000 2bh -55.875 6bh -31.875 abh -7.875 ebh 0.000 2ch -55.500 6ch -31.500 ach -7.500 ech 0.000 2dh -55.125 6dh -31.125 adh -7.125 edh 0.000 2eh -54.750 6eh -30.750 aeh -6.750 eeh 0.000 2fh -54.375 6fh -30.375 afh -6.375 efh 0.000 30h -54.000 70h -30.000 b0h -6.000 f0h 0.000 31h -53.625 71h -29.625 b1h -5.625 f1h 0.000 32h -53.250 72h -29.250 b2h -5.250 f2h 0.000 33h -52.875 73h -28.875 b3h -4.875 f3h 0.000 34h -52.500 74h -28.500 b4h -4.500 f4h 0.000 35h -52.125 75h -28.125 b5h -4.125 f5h 0.000 36h -51.750 76h -27.750 b6h -3.750 f6h 0.000 37h -51.375 77h -27.375 b7h -3.375 f7h 0.000 38h -51.000 78h -27.000 b8h -3.000 f8h 0.000 39h -50.625 79h -26.625 b9h -2.625 f9h 0.000 3ah -50.250 7ah -26.250 bah -2.250 fah 0.000 3bh -49.875 7bh -25.875 bbh -1.875 fbh 0.000 3ch -49.500 7ch -25.500 bch -1.500 fch 0.000 3dh -49.125 7dh -25.125 bdh -1.125 fdh 0.000 3eh -48.750 7eh -24.750 beh -0.750 feh 0.000 3fh -48.375 7fh -24.375 bfh -0.375 ffh 0.000 table 17 dac digital volume range
wm8912 production data w pd, rev 4.0, september 2010 34 dac soft mute and soft un-mute the wm8912 has a soft mute function. when enabled, this gradually attenuates the volume of the dac output. when soft mute is disabled, the gain will either gradually ramp back up to the digital gain setting, or return instantly to the digital gain setting, depending on the dac_unmute_ramp register bit. to mute the dac, this function must be enabled by setting dac_mute to 1. soft mute mode would typically be enabled (dac_unmute_ramp = 1) when using dac_mute during playback of audio data so that when dac_mute is subsequently disabled, the sudden volume increase will not create pop noise by jumping immediately to the previous volume level (e.g. resuming playback after pausing during a track). soft mute mode would typically be disabled (dac_unmute_ramp = 0) when un-muting at the start of a music file, in order that the first part of the track is not attenuated (e.g. when starting playback of a new track, or resuming playback after pausing between tracks). dac muting and un-muting using volume control bits dacl_vol and dacr_vol. dac muting and un-muting using the dac_mute bit. if soft mute mode is not enabled (dac_unmute_ramp = 0): setting the dac_mute bit causes the volume to ramp down at a rate controlled by dac_muterate. clearing the dac_mute bit causes the volume to return to the un-muted level immediately. dac muting and un-muting using the dac_mute bit. if soft mute mode is enabled (dac_unmute_ramp = 1): setting the dac_mute bit causes the volume to ramp down. clearing the dac_mute bit causes the volume to ramp up to the un-muted level at a rate controlled by dac_muterate. figure 21 dac mute control the volume ramp rate during soft mute and un-mute is controlled by the dac_muterate bit. ramp rates of fs/32 and fs/2 can be selected, as shown in table 18. the ramp rate determines the rate at which the volume is increased or decreased. the actual ramp time depends on the extent of the difference between the muted and un-muted volume settings.
production data wm8912 w pd, rev 4.0, september 2010 35 register address bit label default description r33 (21h) dac digital 1 10 dac_mutera te 0 dac soft mute ramp rate 0 = fast ramp (fs/2, maximum ramp time is 10.7ms at fs=48k) 1 = slow ramp (fs/32, maximum ramp time is 171ms at fs=48k) 9 dac_unmute _ramp 0 dac soft mute mode 0 = disabling soft-mute (dac_mute=0) will cause the dac volume to change immediately to dacl_vol and dacr_vol settings 1 = disabling soft-mute (dac_mute=0) will cause the dac volume to ramp up gradually to the dacl_vol and dacr_vol settings 3 dac_mute 1 dac soft mute control 0 = dac un-mute 1 = dac mute table 18 dac soft-mute control dac mono mix a dac digital mono-mix mode can be enabled using the dac_mono register bit. this mono mix will be output on whichever dac is enabled. to prevent clipping, a -6db attenuation is automatically applied to the mono mix. the mono mix is only supported when one or other dac is disabled. when the mono mix is selected, then the mono mix is output on the enabled dac only; there is no output from the disabled dac. if dacl_ena and dacr_ena are both set, then stereo operation applies. register address bit label default description r33 (21h) dac digital 1 12 dac_mono 0 dac mono mix 0 = stereo 1 = mono (mono mix output on enabled dac) table 19 dac mono mix dac de-emphasis digital de-emphasis can be applied to the dac playback data (e.g. when the data comes from a cd with pre-emphasis used in the recording). de-emphasis filtering is available for sample rates of 48khz, 44.1khz and 32khz. see ?digital filter characteristics? for details of de-emphasis filter characteristics. register address bit label default description r33 (21h) dac digital 1 2:1 deemph [1:0] 00 dac de-emphasis control 00 = no de-emphasis 01 = 32khz sample rate 10 = 44.1khz sample rate 11 = 48khz sample rate table 20 dac de-emphasis control
wm8912 production data w pd, rev 4.0, september 2010 36 dac sloping stopband filter two dac filter types are available, selected by the register bit dac_sb_filt. when operating at lower sample rates (e.g. during voice communication) it is recommended that the sloping stopband filter type is selected (dac_sb_filt=1) to reduce out-of-band noise which can be audible at low dac sample rates. see ?digital filter characteristics? for details of dac filter characteristics. register address bit label default description r33 (21h) dac digital 1 11 dac_sb_filt 0 selects dac filter characteristics 0 = normal mode 1 = sloping stopband mode table 21 dac sloping stopband filter dac oversampling ratio (osr) the dac oversampling rate is programmable to allow power consumption versus audio performance trade-offs. the default oversampling rate is low for reduced power consumption; using the higher osr setting improves the dac signal-to-noise performance. register address bit label default description r33 (21h) dac digital 1 6 dac_osr128 0 dac oversample rate select 0 = low power (normal osr) 1 = high performance (double osr) table 22 dac oversampling control
production data wm8912 w pd, rev 4.0, september 2010 37 output signal path the outputs hpoutl and lineoutl are derived from the left dac output, whilst the outputs hpoutr and lineoutr are derived from the right dac output, as illustrated in figure 22. a feedback path for common mode noise rejection is provided at hpoutfb and lineoutfb for the headphone and line outputs respectively. this pin must be connected to ground for normal operation. each analogue output can be separately enabled; independent volume control is also provided for each output. the signal paths and associated control registers are illustrated in figure 22. figure 22 output signal path and control registers
wm8912 production data w pd, rev 4.0, september 2010 38 output signal paths enable the output pgas for each analogue output pin can be enabled and disabled using the register bits described in table 23. note that the headphone outputs and line outputs are also controlled by fields located within register r90 and r94, which provide suppression of pops & clicks when enabling and disabling these signal paths. these registers are described in the following ?headphone / line output signal paths enable? section. under recommended usage conditions, all the control bits associated with enabling the headphone outputs and the line outputs will be configured by scheduling the default start-up and shutdown sequences as described in the ?control write sequencer? section. in these cases, the user does not need to set the register fields in r14, r15, r90 and r94 directly. register address bit label default description r14 (0eh) power management 2 1 hpl_pga_ena 0 left headphone output enable 0 = disabled 1 = enabled 0 hpr_pga_ena 0 right headphone output enable 0 = disabled 1 = enabled r15 (0fh) power management 3 1 lineoutl_pga_ ena 0 left line output enable 0 = disabled 1 = enabled 0 lineoutr_pga _ena 0 right line output enable 0 = disabled 1 = enabled table 23 output signal paths enable note that, to enable the output pgas and multiplexers, the reference voltage vmid and the bias current must also be enabled. see ?reference voltages and master bias? for details of the associated controls vmid_res and bias_ena. headphone / line output signal paths enable the output paths can be actively discharged to agnd through internal resistors if desired. this is desirable at start-up in order to achieve a known output stage condition prior to enabling the vmid reference voltage. this is also desirable in shutdown to prevent the external connections from being affected by the internal circuits. the ground-referenced headphone outputs and line outputs are shorted to agnd by default; the short circuit is removed on each of these paths by setting the applicable fields hpl_rmv_short, hpr_rmv_short, lineoutl_rmv_short or lineoutr_rmv_short. the ground-referenced headphone output and line output drivers are designed to suppress pops and clicks when enabled or disabled. however, it is necessary to control the drivers in accordance with a defined sequence in start-up and shutdown to achieve the pop suppression. it is also necessary to schedule the dc servo offset correction at the appropriate point in the sequence (see ?dc servo?). table 24 and table 25 describe the recommended sequences for enabling and disabling these output drivers.
production data wm8912 w pd, rev 4.0, september 2010 39 sequence headphone enable lineout enable step 1 hpl_ena = 1 hpr_ena = 1 lineoutl_ena = 1 lineoutr_ena = 1 step 2 hpl_ena_dly = 1 hpr_ena_dly = 1 lineoutl_ena_dly = 1 lineoutr_ena_dly = 1 step 3 dc offset correction dc offset correction step 4 hpl_ena_outp = 1 hpr_ena_outp = 1 lineoutl_ena_outp = 1 lineoutr_ena_outp = 1 step 5 hpl_rmv_short = 1 hpr_rmv_short = 1 lineoutl_rmv_short = 1 lineoutr_rmv_short = 1 table 24 headphone / line output enable sequence sequence headphone disable lineout disable step 1 hpl_rmv_short = 0 hpr_rmv_short = 0 lineoutl_rmv_short = 0 lineoutr_rmv_short = 0 step 2 hpl_ena = 0 hpl_ena_dly = 0 hpl_ena_outp = 0 hpr_ena = 0 hpr_ena_dly = 0 hpr_ena_outp = 0 lineoutl_ena = 0 lineoutl_ena_dly = 0 lineoutl_ena_outp = 0 lineoutr_ena = 0 lineoutr_ena_dly = 0 lineoutr_ena_outp = 0 table 25 headphone / line output disable sequence the register bits relating to pop suppression control are defined in table 26 below. register address bit label default description r90 (5ah) analogue hp 0 7 hpl_rmv_shor t 0 removes hpl short 0 = hpl short enabled 1 = hpl short removed for normal operation, this bit should be set as the final step of the hpl enable sequence. 6 hpl_ena_outp 0 enables hpl output stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the dc offset cancellation has been scheduled. 5 hpl_ena_dly 0 enables hpl intermediate stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the output signal path has been configured, and before dc offset cancellation is scheduled. this bit should be set with at least 20us delay after hpl_ena. 4 hpl_ena 0 enables hpl input stage 0 = disabled 1 = enabled for normal operation, this bit should be set as the first step of the hpl enable sequence.
wm8912 production data w pd, rev 4.0, september 2010 40 register address bit label default description 3 hpr_rmv_sho rt 0 removes hpr short 0 = hpr short enabled 1 = hpr short removed for normal operation, this bit should be set as the final step of the hpr enable sequence. 2 hpr_ena_outp 0 enables hpr output stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the dc offset cancellation has been scheduled. 1 hpr_ena_dly 0 enables hpr intermediate stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the output signal path has been configured, and before dc offset cancellation is scheduled. this bit should be set with at least 20us delay after hpr_ena. 0 hpr_ena 0 enables hpr input stage 0 = disabled 1 = enabled for normal operation, this bit should be set as the first step of the hpr enable sequence. r94 (5eh) analogue lineout 0 7 lineoutl_rmv _short 0 removes lineoutl short 0 = lineoutl short enabled 1 = lineoutl short removed for normal operation, this bit should be set as the final step of the lineoutl enable sequence. 6 lineoutl_ena_ outp 0 enables lineoutl output stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the dc offset cancellation has been scheduled. 5 lineoutl_ena_ dly 0 enables lineoutl intermediate stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the output signal path has been configured, and before dc offset cancellation is scheduled. this bit should be set with at least 20us delay after lineoutl_ena. 4 lineoutl_ena 0 enables lineoutl input stage 0 = disabled 1 = enabled for normal operation, this bit should be set as the first step of the lineoutl enable sequence.
production data wm8912 w pd, rev 4.0, september 2010 41 register address bit label default description 3 lineoutr_rmv _short 0 removes lineoutr short 0 = lineoutr short enabled 1 = lineoutr short removed for normal operation, this bit should be set as the final step of the lineoutr enable sequence. 2 lineoutr_ena _outp 0 enables lineoutr output stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the dc offset cancellation has been scheduled. 1 lineoutr_ena _dly 0 enables lineoutr intermediate stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the output signal path has been configured, and before dc offset cancellation is scheduled. this bit should be set with at least 20us delay after lineoutr_ena. 0 lineoutr_ena 0 enables lineoutr input stage 0 = disabled 1 = enabled for normal operation, this bit should be set as the first step of the lineoutr enable sequence. table 26 headphone / line output pop suppression control output volume control each analogue output can be independently controlled using the registers described in table 27 (for headphone outputs) and table 28 (for line outputs). see also the ?analogue outputs? section for details of these output pins, including recommended external components. the volume and mute status of each analogue output can be controlled individually using the register bits described in table 27 and table 28. to prevent ?zipper noise? when a volume adjustment is made, a zero-cross function is provided on all output paths. when this function is enabled, volume updates will not take place until a zero-crossing is detected. in the event of a long period without zero-crossings, a timeout will apply. the timeout must be enabled by setting the toclk_ena bit, as defined in ?clocking and sample rates?. the volume update bits control the loading of the output driver volume data. for example, when hpout_vu is set to 0, the headphone volume data can be loaded into the respective control register, but will not actually change the gain setting. the left and right headphone volume settings are updated when a 1 is written to hpout_vu. this makes it possible to update the gain of a left/right pair of output paths simultaneously.
wm8912 production data w pd, rev 4.0, september 2010 42 register address bit label default description r57 (39h) analogue out1 left 8 hpoutl_mute 0 left headphone output mute 0 = un-mute 1 = mute 7 hpout_vu 0 headphone output volume update writing a 1 to this bit will update hpoutl and hpoutr volumes simultaneously. 6 hpoutlzc 0 left headphone output zero cross enable 0 = disabled 1 = enabled 5:0 hpoutl_vol [5:0] 10_1101 left headphone output volume 000000 = -57db 000001 = -56db (? 1db steps) 111001 = 0db (? 1db steps) 111110 = +5db 111111 = +6db r58 (3ah) analogue out1 right 8 hpoutr_mute 0 right headphone output mute 0 = un-mute 1 = mute 7 hpout_vu 0 headphone output volume update writing a 1 to this bit will update hpoutl and hpoutr volumes simultaneously. 6 hpoutrzc 0 right headphone output zero cross enable 0 = disabled 1 = enabled 5:0 hpoutr_vol [5:0] 10_1101 right headphone output volume 000000 = -57db 000001 = -56db (? 1db steps) 111001 = 0db (? 1db steps) 111110 = +5db 111111 = +6db table 27 volume control for hpoutl and hpoutr
production data wm8912 w pd, rev 4.0, september 2010 43 register address bit label default description r59 (3bh) analogue out2 left 8 lineoutl_mute 0 left line output mute 0 = un-mute 1 = mute 7 lineout_vu 0 line output volume update writing a 1 to this bit will update lineoutl and lineoutr volumes simultaneously. 6 lineoutlzc 0 left line output zero cross enable 0 = disabled 1 = enabled 5:0 lineoutl_vol [5:0] 11_1001 left line output volume 000000 = -57db 000001 = -56db (? 1db steps) 111001 = 0db (? 1db steps) 111110 = +5db 111111 = +6db r60 (3ch) analogue out2 right 8 lineoutr_mut e 0 right line output mute 0 = un-mute 1 = mute 7 lineout_vu 0 line output volume update writing a 1 to this bit will update lineoutl and lineoutr volumes simultaneously. 6 lineoutrzc 0 right line output zero cross enable 0 = disabled 1 = enabled 5:0 lineoutr_vol [5:0] 11_1001 right line output volume 000000 = -57db 000001 = -56db (? 1db steps) 111001 = 0db (? 1db steps) 111110 = +5db 111111 = +6db table 28 volume control for lineoutl and lineoutr
wm8912 production data w pd, rev 4.0, september 2010 44 analogue outputs the wm8912 has four analogue output pins: ? headphone outputs, hpoutl and hpoutr, with feedback hpoutfb ? line outputs, lineoutl and lineoutr, with feedback lineoutfb the output signal paths and associated control registers are illustrated in figure 22. headphone outputs ? hpoutl and hpoutr the headphone outputs are designed to drive 16 ? or 32 ? headphones. these outputs are ground- referenced, i.e. no series capacitor is required between the pins and the headphone load. they are powered by an on-chip charge pump (see ?charge pump? section). signal volume at the headphone outputs is controlled as shown in table 27. the ground-referenced outputs incorporates a common mode, or ground loop, feedback path which provides rejection of system-related ground noise. the return path for the hpoutl and hpoutr outputs is via hpoutfb. this pin must be connected to ground for normal operation of the headphone output. no register configuration is required. line outputs ? lineoutl and lineoutr the line outputs are identical to the headphone outputs in design. they are ground-referenced and powered by the on-chip charge pump. signal volume at the line outputs is controlled as shown in table 28. note that these outputs are intended for driving line loads, as the charge pump powering both the headphone and line outputs can only provide sufficient power to drive one set of headphones at any given time. the ground-referenced outputs incorporates a common mode, or ground loop, feedback path which provides rejection of system-related ground noise. the return path for the lineoutl and lineoutr outputs is via lineoutfb. this pin must be connected to ground for normal operation of the line output. no register configuration is required.
production data wm8912 w pd, rev 4.0, september 2010 45 external components for ground referenced outputs it is recommended to connect a zobel network to the ground-referenced outputs hpoutl, hpoutr, lineoutl and lineoutr in order to ensure best audio performance in all applications. the components of the zobel network have the effect of dampening high frequency oscillations or instabilities that can arise outside the audio band under certain conditions. possible sources of these instabilities include the inductive load of a headphone coil or an active load in the form of an external line amplifier. the capacitance of lengthy cables or pcb tra cks can also l ead to amplifier instability. the zobel network should comprise of a 20 resistor and 100nf capacitor in series with each other, as illustrated in figure 23. note that the zobel network is recommended for best audio quality and amplifier stability in all cases. figure 23 zobel network components for hpoutl, hpoutr, lineoutl and lineoutr
wm8912 production data w pd, rev 4.0, september 2010 46 reference voltages and master bias this section describes the analogue reference voltage and bias current controls. note that, under the recommended usage conditions of the wm8912, these features will be configured by scheduling the default start-up and shutdown sequences as described in the ?control write sequencer? section. in these cases, the user does not need to set these register fields directly. the analogue circuits in the wm8912 require a mid-rail analogue reference voltage, vmid. this reference is generated from avdd via a programmable resistor chain. vmid is enabled by setting the vmid_ena register bit. the programmable resistor chain is configured by vmid_res [1:0], and can be used to optimise the reference for normal operation, low power standby or for fast start-up as described in table 29. for normal operation, the vmid_res field should be set to 01. the analogue circuits in the wm8912 require a bias current. the normal bias current is enabled by setting bias_ena. note that the normal bias current source requires vmid to be enabled also. register address bit label default description r5 (05h) vmid control (0) 2:1 vmid_res [1:0] 00 vmid divider enable and select 00 = vmid disabled (for off mode) 01 = 2 x 50k divider (for normal operation) 10 = 2 x 250k divider (for low power standby) 11 = 2 x 5k divider (for fast start-up) 0 vmid_ena 0 enable vmid master bias current source 0 = disabled 1 = enabled r4 (04h) bias control (0) 0 bias_ena 0 enables the normal bias current generator (for all analogue functions) 0 = disabled 1 = enabled table 29 reference voltages and master bias enable charge pump the wm8912 incorporates a dual-mode charge pump which generates the supply rails for the headphone and line output drivers, hpoutl, hpoutr, and lineoutl and lineoutr. the charge pump has a single supply input, cpvdd, and generates split rails cpvoutp and cpvoutn according to the selected mode of operation. the charge pump connections are illustrated in figure 24 (see the ?electrical characteristics? section for external component values). an input decoupling capacitor may also be required at cpvdd, depending upon the system configuration. figure 24 charge pump external connections
production data wm8912 w pd, rev 4.0, september 2010 47 the charge pump is enabled by setting the cp_ena bit. when enabled, the charge pump adjusts the output voltages (cpvoutp and cpvoutn) as well as the switching frequency in order to optimise the power consumption according to the operating conditions. this can take two forms, which are selected using the cp_dyn_pwr register bit. ? register control (cp_dyn_pwr = 0) ? dynamic control (cp_dyn_pwr = 1) under register control, the hpoutl_vol, hpoutr_vol, lineoutl_vol and lineoutr_vol register settings are used to control the charge pump mode of operation. under dynamic control, the audio signal level in the dac is used to control the charge pump mode of operation. this is the wolfson ?class w? mode, which allows the power consumption to be optimised in real time. under the recommended usage conditions of the wm8912, the charge pump will be enabled by running the default headphone start-up sequence as described in the ?control write sequencer? section. (similarly, it will be disabled by running the shutdown sequence.) in these cases, the user does not need to write to the cp_ena bit. the charge pump operating mode defaults to register control; dynamic control may be selected by setting the cp_dyn_pwr register bit, if appropriate. note that the charge pump clock is derived from internal clock sysclk; this may derived from mclk directly or else using the fll output, as determined by the sysclk_src bit. u nder normal circumstances an external clock signal must be present for the charge pump to function. however, the fll has a free-running mode that does not require an external clock but will generate an internal clock suitable for running the charge pump. the clock division from sysclk is handled transparently by the wm8912 without user intervention, as long as sysclk and sample rates are set correctly. refer to the ?clocking and sample rates? section for more detail on the fll and clocking configuration. the charge pump control fields are described in table 30. register address bit label default description r98 (62h) charge pump 0 0 cp_ena 0 enable charge-pump digits 0 = disable 1 = enable r104 (68h) class w (0) 0 cp_dyn_pwr 0 enable dynamic charge pump power control 0 = charge pump controlled by volume register settings (class g) 1 = charge pump controlled by real- time audio level (class w) class w is recommended for lowest power consumption table 30 charge pump control
wm8912 production data w pd, rev 4.0, september 2010 48 dc servo the wm8912 provides four dc servo circuits, two on the headphone outputs hpoutl and hpoutr and two on the line outputs lineoutl and lineoutr, to remove dc offset from these ground- referenced outputs. when enabled, the dc servos ensure that the dc level of these outputs remains within 1mv of ground. removal of the dc offset is important because any deviation from gnd at the output pin will cause current to flow through the load under quiescent conditions, resulting in increased power consumption. additionally, the presence of dc offsets can result in audible pops and clicks at power up and power down. the recommended usage of the dc servo is initialised by scheduling the default start-up sequence as described in the ?control write sequencer? section. the default start-up sequence executes a series of dc offset corrections, after which the measured offset correction is maintained on the headphone output channels. if a different usage is required, e.g. if a periodic dc offset correction is required, then the default start-up sequence may be modified according to specific requirements. the relevant control fields are described in the following paragraphs and are defined in table 31. dc servo enable and start-up the dc servo circuits are enabled on hpoutl and hpoutr by setting dcs_ena_chan_0 and dcs_ena_chan_1 respectively. similarly, the dc servo circuits are enabled on lineoutl and lineoutr by setting dcs_ena_chan_2 and dcs_ena_chan_3 respectively when the dc servo is enabled, the dc offset correction can be commanded in a number of different ways, including single-shot and periodically recurring events. writing a logic 1 to dcs_trig_startup_ n initiates a series of dc offset measurements and applies the necessary correction to the associated output; (?n? = 3 for lineoutr channel, 2 for lineoutl channel, 1 for hpoutr channel, 0 for hpoutl channel). on completion, the output will be within 1mv of agnd. this is the dc servo mode selected by the default start-up sequence. completion of the dc offset correction triggered in this way is indicated by the dcs_startup_complete field, as described in table 31. typically, this operation takes 86ms per channel. writing a logic 1 to dcs_trig_dac_wr_ n causes the dc offset correction to be set to the value contained in the dcs_dac_wr_val_ n fields in registers r73 to r76. this mode is useful if the required offset correction has already been determined and stored; it is faster than the dcs_trig_startup_ n mode, but relies on the accuracy of the stored settings. completion of the dc offset correction triggered in this way is indicated by the dcs_dac_wr_complete field, as described in table 31. typically, this operation takes 2ms per channel. when using either of the dc servo options above, the status of the dc offset correction process is indicated by the dcs_cal_complete field; this is the logical or of the dcs_startup_complete and dcs_dac_wr_complete fields. the dc servo control fields associated with start-up operation are described in table 31. it is important to note that, to minimise audible pops/clicks, the start-up and dac write modes of dc servo operation should be commanded as part of a control sequence which includes muting and shorting of the headphone outputs; a suitable sequence is defined in the default start-up sequence. register address bit label default description r68 (44h) dc servo 1 7 dcs_trig_star tup_3 0 writing 1 to this bit selects start-up dc servo mode for lineoutr. in readback, a value of 1 indicates that the dc servo start-up correction is in progress. 6 dcs_trig_star tup_2 0 writing 1 to this bit selects start-up dc servo mode for lineoutl. in readback, a value of 1 indicates that the dc servo start-up correction is in progress.
production data wm8912 w pd, rev 4.0, september 2010 49 register address bit label default description 5 dcs_trig_star tup_1 0 writing 1 to this bit selects start-up dc servo mode for hpoutr. in readback, a value of 1 indicates that the dc servo start-up correction is in progress. 4 dcs_trig_star tup_0 0 writing 1 to this bit selects start-up dc servo mode for hpoutl. in readback, a value of 1 indicates that the dc servo start-up correction is in progress. 3 dcs_trig_dac_ wr_3 0 writing 1 to this bit selects dac write dc servo mode for lineoutr. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 2 dcs_trig_dac_ wr_2 0 writing 1 to this bit selects dac write dc servo mode for lineoutl. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 1 dcs_trig_dac_ wr_1 0 writing 1 to this bit selects dac write dc servo mode for hpoutr. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 0 dcs_trig_dac_ wr_0 0 writing 1 to this bit selects dac write dc servo mode for hpoutl. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. r67 (43h) dc servo 0 3 dcs_ena_chan _3 0 dc servo enable for lineoutr 0 = disabled 1 = enabled 2 dcs_ena_chan _2 0 dc servo enable for lineoutl 0 = disabled 1 = enabled 1 dcs_ena_chan _1 0 dc servo enable for hpoutr 0 = disabled 1 = enabled 0 dcs_ena_chan _0 0 dc servo enable for hpoutl 0 = disabled 1 = enabled r73 (49h) dc servo 6 7:0 dcs_dac_wr_v al_3 [7:0] 0000 0000 dc offset value for lineoutr in dac write dc servo mode in two's complement format. in readback, the current dc offset value is returned in two's complement format. two?s complement format: lsb is 0.25mv. range is +/-32mv
wm8912 production data w pd, rev 4.0, september 2010 50 register address bit label default description r74 (4ah) dc servo 7 7:0 dcs_dac_wr_v al_2 [7:0] 0000 0000 dc offset value for lineoutl in dac write dc servo mode in two's complement format. in readback, the current dc offset value is returned in two's complement format. two?s complement format: lsb is 0.25mv. range is +/-32mv r75 (4bh) dc servo 8 7:0 dcs_dac_wr_v al1 [7:0] 0000 0000 dc offset value for hpoutr in dac write dc servo mode in two's complement format. in readback, the current dc offset value is returned in two's complement format. two?s complement format: lsb is 0.25mv. range is +/-32mv r76 (4ch) dc servo 9 7:0 dcs_dac_wr_v al0 [7:0] 0000 0000 dc offset value for hpoutl in dac write dc servo mode in two's complement format. in readback, the current dc offset value is returned in two's complement format. two?s complement format: lsb is 0.25mv. range is +/-32mv r77 (4dh) dc servo readback 0 11:8 dcs_cal_comp lete [3:0] 0000 dc servo complete status [3] - lineoutr [2] - lineoutl [1] - hpoutr [0] - hpoutl 0 = dac write or start-up dc servo mode not completed. 1 = dac write or start-up dc servo mode complete. 7:4 dcs_dac_wr_ complete [3:0] 0000 dc servo dac write status [3] - lineoutr [2] - lineoutl [1] - hpoutr [0] - hpoutl 0 = dac write dc servo mode not completed. 1 = dac write dc servo mode complete.
production data wm8912 w pd, rev 4.0, september 2010 51 register address bit label default description 3:0 dcs_startup_ complete [3:0] 0000 dc servo start-up status [3] - lineoutr [2] - lineoutl [1] - hpoutr [0] - hpoutl 0 = start-up dc servo mode not completed.. 1 = start-up dc servo mode complete. table 31 dc servo enable and start-up modes dc servo active modes the dc servo modes described above are suitable for initialising the dc offset correction circuit on the line and headphone outputs as part of a controlled start-up sequence which is executed before the signal path is fully enabled. additional modes are available for use whilst the signal path is active; these modes may be of benefit following a large change in signal gain, which can lead to a change in dc offset level. periodic updates may also be desirable to remove slow drifts in dc offset caused by changes in parameters such as device temperature. the dc servo circuit is enabled on hpoutr and hpoutl by setting dcs_ena_chan_1 and dcs_ena_chan_0 respectively, as described earlier in table 31. similarly, the dc servo circuit is enabled on lineoutr and lineoutl by setting dcs_ena_chan_3 and dcs_ena_chan_2 respectively. writing a logic 1 to dcs_trig_single_ n initiates a single dc offset measurement and adjustment to the associated output; (?n? = 3 for lineoutr channel, 2 for lineoutl channel, 1 for hpoutr channel, 0 for hpoutl channel). this will adjust the dc offset correction on the selected channel by no more than 1lsb (0.25mv). setting dcs_timer_period_01 or dcs_timer_period_23 to a non-zero value will cause a single dc offset measurement and adjustment to be scheduled on a periodic basis. periodic rates ranging from every 0.52s to in excess of 2 hours can be selected. writing a logic 1 to dcs_trig_series_ n initiates a series of dc offset measurements and applies the necessary correction to the associated output. the number of dc servo operations performed is determined by dcs_series_no_01 or dcs_series_no_23. a maximum of 128 operations may be selected, though a much lower value will be sufficient in most applications. the dc servo control fields associated with active modes (suitable for use on a signal path that is in active use) are described in table 32. register address bit label default description r68 (44h) dc servo 1 15 dcs_trig_sing le_3 0 writing 1 to this bit selects a single dc offset correction for lineoutr. in readback, a value of 1 indicates that the dc servo single correction is in progress. 14 dcs_trig_sing le_2 0 writing 1 to this bit selects a single dc offset correction for lineoutl. in readback, a value of 1 indicates that the dc servo single correction is in progress. 13 dcs_trig_sing le_1 0 writing 1 to this bit selects a single dc offset correction for hpoutr. in readback, a value of 1 indicates that the dc servo single correction is in progress.
wm8912 production data w pd, rev 4.0, september 2010 52 register address bit label default description 12 dcs_trig_sing le_0 0 writing 1 to this bit selects a single dc offset correction for hpoutl. in readback, a value of 1 indicates that the dc servo single correction is in progress. 11 dcs_trig_seri es_3 0 writing 1 to this bit selects a series of dc offset corrections for lineoutr. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 10 dcs_trig_seri es_2 0 writing 1 to this bit selects a series of dc offset corrections for lineoutl. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 9 dcs_trig_seri es_1 0 writing 1 to this bit selects a series of dc offset corrections for hpoutr. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 8 dcs_trig_seri es_0 0 writing 1 to this bit selects a series of dc offset corrections for hpoutl. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. r71 (47h) dc servo 4 6:0 dcs_series_n o_23 [6:0] 010_1010 number of dc servo updates to perform in a series event for lineoutl/lineoutr. 0 = 1 updates 1 = 2 updates ... 127 = 128 updates r72 (48h) dc servo 5 6:0 dcs_series_n o_01 [6:0] 010 1010 number of dc servo updates to perform in a series event for hpoutl/hpoutr. 0 = 1 updates 1 = 2 updates ... 127 = 128 updates r69 (45h) dc servo 2 11:8 dcs_timer_pe riod_23 [3:0] 1010 time between periodic updates for lineoutl/lineoutr. time is calculated as 0.256s x (2^period) 0000 = off 0001 = 0.52s 1010 = 266s (4min 26s) 1111 = 8519s (2hr 22s) 3:0 dcs_timer_pe riod_01 [3:0] 1010 time between periodic updates for hpoutl/hpoutr. time is calculated as 0.256s x (2^period) 0000 = off 0001 = 0.52s 1010 = 266s (4min 26s) 1111 = 8519s (2hr 22s) table 32 dc servo active modes
production data wm8912 w pd, rev 4.0, september 2010 53 dc servo readback the current dc offset value for each line and headphone output channel can be read in two?s complement format from the dcs_dac_wr_val_n [7:0] bit fields in registers r73, r74, r75 and r76. note that these values may form the basis of settings that are subsequently used by the dc servo in dac write mode. digital audio interface the digital audio interface is used for inputting dac data to the wm8912. the digital audio interface uses three pins: ? dacdat: dac data input ? lrclk: left/right data alignment clock ? bclk: bit clock, for synchronisation the clock signals bclk and lrclk can be outputs when the wm8912 operates as a master, or inputs when it is a slave (see ?master and slave mode operation?, below). four different audio data formats are supported: ? left justified ? right justified ? i2s ? dsp mode all four of these modes are msb first. they are described in ?audio data formats (normal mode)?, below. refer to the ?signal timing requirements? section for timing information. time division multiplexing (tdm) is available in all four data format modes. the wm8912 can be programmed to receive data in one of two time slots. pcm operation is supported using the dsp mode. master and slave mode operation the wm8912 digital audio interface can operate in master or slave mode, as shown in figure 25 and figure 26. figure 25 master mode figure 26 slave mode in master mode, bclk is derived from sysclk via a programmable division set by bclk_div. in master mode, lrclk is derived from bclk via a programmable division set by lrclk_rate. the bclk input to this divider may be internal or external, allowing mixed master and slave modes.
wm8912 production data w pd, rev 4.0, september 2010 54 the direction of these signals and the clock frequencies are controlled as described in the ?digital audio interface control? section. bclk and lrclk can be enabled as outputs in slave mode, allowing mixed master/slave operation - see ?digital audio interface control?. operation with tdm time division multiplexing (tdm) allows multiple devices to transfer data simultaneously on the same bus. the wm8912 supports tdm in master and slave modes for all data formats and word lengths. tdm is enabled and configured using register bits defined in the ?digital audio interface control? section. figure 27 tdm with wm8912 as master figure 28 tdm with other dac as master figure 29 tdm with processor as master note: the wm8912 is a 24-bit device. if the user operates the wm8912 in 32-bit mode then the 8 lsbs will be ignored on the dac input. it is recommended to add a pull-down resistor to the dacdat line in tdm mode in this case.
production data wm8912 w pd, rev 4.0, september 2010 55 bclk frequency the bclk frequency is controlled relative to sysclk by the bclk_div divider. internal clock divide and phase control mechanisms ensure that the bclk and lrclk edges will occur in a predictable and repeatable position relative to each other and relative to the data for a given combination of dac sample rate and bclk_div settings. bclk_div is defined in the ?digital audio interface control? section. see also ?clocking and sample rates? section for more information. audio data formats (normal mode) in right justified mode, the lsb is available on the last rising edge of bclk before a lrclk transition. all other bits are transmitted before (msb first). depending on word length, bclk frequency and sample rate, there may be unused bclk cycles after each lrclk transition. figure 30 right justified audio interface (assuming n-bit word length) in left justified mode, the msb is available on the first rising edge of bclk following a lrclk transition. the other bits up to the lsb are then transmitted in order. depending on word length, bclk frequency and sample rate, there may be unused bclk cycles before each lrclk transition. figure 31 left justified audio interface (assuming n-bit word length) in i 2 s mode, the msb is available on the second rising edge of bclk following a lrclk transition. the other bits up to the lsb are then transmitted in order. depending on word length, bclk frequency and sample rate, there may be unused bclk cycles between the lsb of one sample and the msb of the next.
wm8912 production data w pd, rev 4.0, september 2010 56 n n-1 n-2 3 2 1 n n-1 n-2 3 2 1 left channel right channel msb lsb input word length (wl) 1/fs lrclk bclk dacdat 1 bclk 1 bclk figure 32 i2s justified audio interface (assuming n-bit word length) in dsp mode, the left channel msb is available on either the 1 st (mode b) or 2 nd (mode a) rising edge of bclk (selectable by aif_lrclk_inv) following a rising edge of lrclk. right channel data immediately follows left channel data. depending on word length, bclk frequency and sample rate, there may be unused bclk cycles between the lsb of the right channel data and the next sample. in device master mode, the lrc output will resemble the frame pulse shown in figure 33 and figure 34. in device slave mode, figure 35 and figure 36, it is possible to use any length of frame pulse less than 1/fs, providing the falling edge of the frame pulse occurs greater than one bclk period before the rising edge of the next frame pulse. figure 33 dsp mode audio interface (mode a, aif_lrclk_inv=0, master) figure 34 dsp mode audio interface (mode b, aif_lrclk_inv=1, master)
production data wm8912 w pd, rev 4.0, september 2010 57 figure 35 dsp mode audio interface (mode a, aif_lrclk_inv=0, slave) figure 36 dsp mode audio interface (mode b, aif_lrclk_inv=1, slave) pcm operation is supported in dsp interface mode. mono pcm data received by the wm8912 will be treated as left channel data. this data may be routed to the left/right dacs as described in the ?digital mixing? section. audio data formats (tdm mode) tdm is supported in master and slave mode and is enabled by the aifdac_tdm register bit. all audio interface data formats support time division multiplexing (tdm) for dac data. two time slots are available (slot 0 and slot 1), selected by the aifdac_tdm_chan register bit. when tdm is enabled, bclk frequency must be high enough to allow data from both time slots to be transferred. the relative timing of slot 0 and slot 1 depends upon the selected data format as shown in figure 37 to figure 41. figure 37 tdm in right-justified mode
wm8912 production data w pd, rev 4.0, september 2010 58 figure 38 tdm in left-justified mode figure 39 tdm in i 2 s mode 1/fs lrclk bclk dacdat 1 bclk slot 0 left slot 0 right slot 1 left slot 1 right falling edge can occur anywhere in this area 1 bclk figure 40 tdm in dsp mode a figure 41 tdm in dsp mode b
production data wm8912 w pd, rev 4.0, september 2010 59 digital audio interface control the register bits controlling audio data format, word length, left/right channel data source and tdm are summarised in table 33. register address bit label default description r24 (18h) audio interface 0 5 aifdacl_sr c 0 left dac data source select 0 = left dac outputs left channel data 1 = left dac outputs right channel data 4 aifdacr_sr c 1 right dac data source select 0 = right dac outputs left channel data 1 = right dac outputs right channel data r25 (19h) audio interface 1 13 aifdac_tdm 0 dac tdm enable 0 = normal dacdat operation 1 = tdm enabled on dacdat 12 aifdac_tdm _chan 0 dacdat tdm channel select 0 = dacdat data input on slot 0 1 = dacdat data input on slot 1 7 aif_bclk_in v 0 bclk invert 0 = bclk not inverted 1 = bclk inverted 4 aif_lrclk_i nv 0 lrc polarity / dsp mode a-b select. right, left and i2s modes ? lrc polarity 0 = not inverted 1 = inverted dsp mode ? mode a-b select 0 = msb is available on 2nd bclk rising edge after lrc rising edge (mode a) 1 = msb is available on 1st bclk rising edge after lrc rising edge (mode b) 3:2 aif_wl [1:0] 10 digital audio interface word length 00 = 16 bits 01 = 20 bits 10 = 24 bits 11 = 32 bits 1:0 aif_fmt [1:0] 10 digital audio interface format 00 = right justified 01 = left justified 10 = i2s 11 = dsp table 33 digital audio interface data control audio interface output tri-state register bit aif_tris can be used to tri-state the audio interface pins as described in table 34. all digital audio interface pins will be tri-stated by this function, regardless of the state of other registers which control these pin configurations. register address bit label default description r25 (19h) audio interface 1 8 aif_tris 0 audio interface tristate 0 = audio interface pins operate normally 1 = tristate all audio interface pins table 34 digital audio interface tri-state control
wm8912 production data w pd, rev 4.0, september 2010 60 bclk and lrclk control the audio interface can be programmed to operate in master mode or slave mode using the bclk_dir and lrclk_dir register bits. in master mode, the bclk and lrclk signals are generated by the wm8912 when either of the dacs is enabled. in slave mode, the bclk and lrclk clock outputs are disabled by default to allow another digital audio interface to drive these pins. it is also possible to force the bclk or lrclk signals to be output using bclk_dir and lrclk_dir, allowing mixed master and slave modes. the bclk_dir and lrclk_dir fields are defined in table 35. register address bit label default description r25 (19h) audio interface 1 6 bclk_dir 0 audio interface bclk direction 0 = bclk is input 1 = bclk is output r26 (1ah) audio interface 2 4:0 bclk_div [4:0] 0_0100 bclk frequency (master mode) 00000 = sysclk 00001 = sysclk / 1.5 00010 = sysclk / 2 00011 = sysclk / 3 00100 = sysclk / 4 00101 = sysclk / 5 00110 = sysclk / 5.5 00111 = sysclk / 6 01000 = sysclk / 8 (default) 01001 = sysclk / 10 01010 = sysclk / 11 01011 = sysclk / 12 01100 = sysclk / 16 01101 = sysclk / 20 01110 = sysclk / 22 01111 = sysclk / 24 10000 = sysclk / 25 10001 = sysclk / 30 10010 = sysclk / 32 10011 = sysclk / 44 10100 = sysclk / 48 r27 (1bh) audio interface 3 11 lrclk_dir 0 audio interface lrc direction 0 = lrc is input 1 = lrc is output 10:0 lrclk_rate [10:0] 000_0100 _0000 lrc rate (master mode) lrc clock output = bclk / lrclk_rate integer (lsb = 1) valid range: 8 to 2047 table 35 digital audio interface clock control
production data wm8912 w pd, rev 4.0, september 2010 61 companding the wm8912 supports a-law and -law companding on the digital receive (dac) path as shown in table 36. register address bit label default description r24 (18h) audio interface 0 1 dac_comp 0 dac companding enable 0 = disabled 1 = enabled 0 dac_compmode 0 dac companding type 0 = -law 1 = a-law table 36 companding control companding involves using a piecewise linear approximation of the following equations (as set out by itu-t g.711 standard) for data compression: -law (where =255 for the u.s. and japan): f(x) = ln( 1 + |x|) / ln( 1 + ) -1 x 1 a-law (where a=87.6 for europe): f(x) = a|x| / ( 1 + lna) x 1/a f(x) = ( 1 + lna|x|) / (1 + lna) 1/a x 1 the companded data is also inverted as recommended by the g.711 standard (all 8 bits are inverted for -law, all even data bits are inverted for a-law). the data will be transmitted as the first 8 msbs of data. companding converts 13 bits ( -law) or 12 bits (a-law) to 8 bits using non-linear quantization. this provides greater precision for low amplitude signals than for high amplitude signals, resulting in a greater usable dynamic range than 8 bit linear quantization. the companded signal is an 8-bit word comprising sign (1 bit), exponent (3 bits) and mantissa (4 bits). 8-bit mode is selected whenever dac_comp=1. the use of 8-bit data allows samples to be passed using as few as 8 bclk cycles per lrclk frame. when using dsp mode b, 8-bit data words may be transferred consecutively every 8 bclk cycles. 8-bit mode (without companding) may be enabled by setting dac_compmode=1 when dac_comp=0. bit7 bit [6:4] bit [3:0] sign exponent mantissa table 37 8-bit companded word composition
wm8912 production data w pd, rev 4.0, september 2010 62 u-law companding 0 20 40 60 80 100 120 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 normalised input companded output 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 normalised output figure 42 -law companding a-law companding 0 20 40 60 80 100 120 0 0.2 0.4 0.6 0.8 1 normalised input companded output 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 normalised output figure 43 a-law companding
production data wm8912 w pd, rev 4.0, september 2010 63 digital pull-up and pull-down the wm8912 provides integrated pull-up and pull-down resistors on each of the mclk, dacdat, lrclk and bclk pins. this provides a flexible capability for interfacing with other devices. each of the pull-up and pull-down resistors can be configured independently using the register bits described in table 38. register address bit label default description r126 (7eh) digital pulls 7 mclk_pu 0 mclk pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 6 mclk_pd 0 mclk pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled 5 dacdat_pu 0 dacdat pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 4 dacdat_pd 0 dacdat pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled 3 lrclk_pu 0 lrclk pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 2 lrclk_pd 0 lrclk pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled 1 bclk_pu 0 bclk pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 0 bclk_pd 0 bclk pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled table 38 digital audio interface pull-up and pull-down control
wm8912 production data w pd, rev 4.0, september 2010 64 clocking and sample rates the internal clocks for the wm8912 are all derived from a common internal clock source, sysclk. this clock is the reference for the dacs, dsp core functions, digital audio interface, dc servo control and other internal functions. sysclk can either be derived directly from mclk, or may be generated from a frequency locked loop (fll) using mclk, bclk or lrclk as a reference. many commonly-used audio sample rates can be derived directly from typical mclk frequencies; the fll provides additional flexibility for a wide range of mclk frequencies. to avoid audible glitches, all clock configurations must be set up before enabling playback. the fll can be used to generate a free-running clock in the absence of an external reference source; see ?frequency locked loop? for further details. the wm8912 supports automatic clocking configuration. the programmable dividers associated with the dacs, dsp core functions and dc servo are configured automatically, with values determined from the clk_sys_rate and sample_rate fields. the user must also configure the opclk (if required), the toclk (if required) and the digital audio interface. oversample rates of 64fs or 128fs are supported (based on a 48khz sample rate). a 256khz clock, supporting a number of internal functions, is derived from sysclk. the dc servo control is clocked from sysclk. a gpio clock, opclk, can be derived from sysclk and output on a gpio pin to provide clocking to other devices. this clock is enabled by opclk_ena and controlled by opclk_div. a slow clock, toclk, is used to de-bounce the button/accessory detect inputs, and to set the timeout period for volume updates when zero-cross detect is used. this clock is enabled by toclk_ena and controlled by toclk_rate, toclk_rate_x4 and toclk_rate_div16. in master mode, bclk is derived from sysclk via a programmable divider set by bclk_div. in master mode, the lrclk is derived from bclk via a programmable divider lrclk_rate. the lrclk can be derived from an internal or external bclk source, allowing mixed master/slave operation. the control registers associated with clocking and sample rates are shown in table 39 to table 43. the overall clocking scheme for the wm8912 is illustrated in figure 44.
production data wm8912 w pd, rev 4.0, september 2010 65 figure 44 clocking scheme sysclk control the sysclk_src bit is used to select the source for sysclk. the source may be either the selected mclk source or the fll output. the mclk source can be inverted or non-inverted, as selected by the mclk_inv bit. the selected source may also be adjusted by the mclk_div divider to generate sysclk. these register fields are descri bed in table 39. see ?frequency locked loop (fll)? for more details of the frequency locked loop clock generator. the sysclk signal is enabled by register bit clk_sys_ena. this bit should be set to 0 when reconfiguring clock sources. it is not recommended to change sysclk_src while the clk_sys_ena bit is set. the following operating frequency limits must be observed when configuring sysclk. failure to observe these limits will result in degraded noise performance and/or incorrect dac functionality. ? sysclk 3mhz ? if dac_osr128 = 1 then sysclk 6mhz ? if dac_mono = 1, then sysclk 64 x fs ? if dac_mono = 0, then sysclk 128 x fs note that dac mono mode (dac_mono = 1) is only valid when one or other dac is disabled. if both dacs are enabled, then the minimum sysclk for clocking the dacs is 128 x fs.
wm8912 production data w pd, rev 4.0, september 2010 66 the sysclk control register fields are defi ned in table 39. register address bit label default description r22 (16h) clock rates 2 15 mclk_inv 0 mclk invert 0 = mclk not inverted 1 = mclk inverted 14 sysclk_src 0 sysclk source select 0 = mclk 1 = fll output 2 clk_sys_ena 0 system clock enable 0 = disabled 1 = enabled r20 (14h) clock rates 0 0 mclk_div 0 enables divide by 2 on mclk 0 = sysclk = mclk 1 = sysclk = mclk / 2 table 39 mclk and sysclk control control interface clocking register map access is possible with or without a master clock (mclk). however, if clk_ sys_ena has been set to 1, then a master clock must be present for control register read/write operations. if clk_sys_ena = 1 and mclk is not present, then register access will be unsuccessful. (note that read/write access to register r22, containing clk_sys_ena, is always possible.) if it cannot be assured that mclk is present when accessing the register map, then it is required to set clk_sys_ena = 0 to ensure correct operation. note that mclk is always required when using hpoutl, hpoutr, lineoutl or lineoutr. clocking configuration the wm8912 supports a wide range of standard audio sample rates from 8khz to 96khz. the automatic clocking configuration simplifies the configuration of the clock dividers in the wm8912 by deriving most of the required parameters from a minimum number of user registers. the sample_rate field selects the sample rate, fs, of the dac. the clk_sys_rate fields must be set according to the ratio of sysclk to fs. when these fields are set correctly, the sample rate decoder circuit automatically determines the clocking configuration for all other circuits within the wm8912. a high performance mode of dac operation can be selected by setting the dac_osr128 bit; in 48khz sample mode, the dac_osr128 feature results in 128x oversampling. audio performance is improved, but power consumption is also increased.
production data wm8912 w pd, rev 4.0, september 2010 67 register address bit label default description r33 (21h) dac digital 1 6 dac_osr128 0 dac oversample rate select 0 = low power (normal osr) 1 = high performance (double osr) r21 (15h) clock rates 1 13:10 clk_sys_rat e [3:0] 0011 selects the sysclk / fs ratio 0000 = 64 0001 = 128 0010 = 192 0011 = 256 0100 = 384 0101 = 512 0110 = 768 0111 = 1024 1000 = 1408 1001 = 1536 2:0 sample_rate [2:0] 101 selects the sample rate (fs) 000 = 8khz 001 = 11.025khz, 12khz 010 = 16khz 011 = 22.05khz, 24khz 100 = 32khz 101 = 44.1khz, 48khz 110 to 111 = reserved table 40 automatic clocking configuration control dac clock control the clocking of the dac circuits is derived from clk_dsp, which is enabled by clk_dsp_ena. clk_dsp is generated from sysclk which is s eparately enabled, using the register bit clk_sys_ena. note that higher performance dac operation can be achieved by increasing the dac oversample rate - see table 40. register address bit label default description r22 (16h) clock rates 2 1 clk_dsp_ena 0 dsp clock enable 0 = disabled 1 = enabled table 41 adc / dac clock control opclk control a clock output (opclk) derived from sysclk may be output on a gpio pin. this clock is enabled by register bit opclk_ena, and its frequency is controlled by opclk_div. this output of this clock is also dependent upon the gpio register settings described under ?general purpose input/output (gpio)?.
wm8912 production data w pd, rev 4.0, september 2010 68 register address bit label default description r22 (16h) clock rates 2 3 opclk_ena 0 gpio clock output enable 0 = disabled 1 = enabled r26 (1ah) audio interface 2 11:8 opclk_div [3:0] 0000 gpio output clock divider 0000 = sysclk 0001 = sysclk / 2 0010 = sysclk / 3 0011 = sysclk / 4 0100 = sysclk / 5.5 0101 = sysclk / 6 0110 = sysclk / 8 0111 = sysclk / 12 1000 = sysclk / 16 1001 to 1111 = reserved table 42 opclk control toclk control a slow clock (toclk) is derived from the internally generated 256khz clock to enable input de- bouncing and volume update timeout functions. this clock is enabled by register bit toclk_ena, and its frequency is controlled by toclk_rate and toclk_rate_x4, as described in table 43. register address bit label default description r22 (16h) clock rates 2 12 toclk_rate 0 toclk rate divider (/2) 0 = f / 2 1 = f / 1 0 toclk_ena 0 zero cross timeout enable 0 = disabled 1 = enabled r20 (14h) clock rates 0 14 toclk_rate_ div16 0 toclk rate divider (/16) 0 = f / 1 1 = f / 16 13 toclk_rate_ x4 0 toclk rate multiplier 0 = f x 1 1 = f x 4 table 43 toclk control a list of possible toclk rates is provided in table 44. toclk_rate toclk_rate_x4 toclk_rate_div16 toclk freq (hz) period (ms) 1 1 0 1000 1 0 1 0 500 2 1 0 0 250 4 0 0 0 125 8 1 1 1 62.5 16 0 1 1 31.25 32 1 0 1 15.625 64 0 0 1 7.8125 128 table 44 toclk rates
production data wm8912 w pd, rev 4.0, september 2010 69 dac operation at 88.2k / 96k the wm8912 supports dac operation at 88.2khz and 96khz sample rates. this section details specific conditions applicable to these operating modes. for dac operation at 88.2khz or 96khz sample rates, the available clocking configurations are detailed in table 45. dac operation at these sample rates is achieved by setting the sample_rate field to half the required sample rate (eg. select 48khz for 96khz mode). for dac operation at 88.2khz or 96khz sample rates, the dac_osr128 register must be set to 0. retune tm mobile can not be used during 88.2khz or 96khz operation, so eq_ena must be set to 0. the sysclk fr equency is derived from mclk. the maximum mclk frequency is defined in the ?signal timing requirements? section. sample rate register configuration clocking ratio 88.2khz sample_rate = 101 clk_sys_rate = 0001 (sysclk / fs = 128) bclk_div = 00010 lrclk_rate = 040h sysclk = 128 x fs 96khz sample_rate = 101 clk_sys_rate = 0001 (sysclk / fs = 128) bclk_div = 00010 lrclk_rate = 040h sysclk = 128 x fs table 45 dac operation at 88.2khz and 96khz sample rates
wm8912 production data w pd, rev 4.0, september 2010 70 frequency locked loop (fll) the integrated fll can be used to generate sysclk from a wide variety of different reference sources and frequencies. the fll can use either mclk, bclk or lrclk as its reference, which may be a high frequency (eg. 12.288mhz) or low frequency (eg. 32,768khz) reference. the fll is tolerant of jitter and may be used to generate a stable sysclk from a less stable input signal. the fll characteristics are summarised in ?electrical characteristics?. note that the fll can be used to generate a free-running clock in the absence of an external reference source. this is described in the ?free-running fll clock? section below. the fll is enabled using the fll_ena register bit. note that, when changing fll settings, it is recommended that the digital circuit be disabled via fll_ena and then re-enabled after the other register settings have been updated. when changing the input reference frequency f ref , it is recommended the fll be reset by setting fll_ena to 0. the fll_clk_ref_src field allows mclk, bclk or lrclk to be selected as the input reference clock. the field fll_clk_ref_div provides the option to divide the input reference (mclk, bclk or lrclk) by 1, 2, 4 or 8. this field should be set to bring the reference down to 13.5mhz or below. for best performance, it is recommended that the highest possible frequency - within the 13.5mhz limit - should be selected. the field fll_ctrl_rate controls internal functions within the fll; it is recommended that only the default setting be used for this parameter. fll_gain controls the internal loop gain and should be set to the recommended value quoted in table 48. the fll output frequency is directly determined from fll_fratio, fll_outdiv and the real number represented by fll_n and fll_k. the field fll_n is an integer (lsb = 1); fll_k is the fractional portion of the number (msb = 0.5). the fractional portion is only valid when enabled by the field fll_fracn_ena. it is recommended that fll_fracn_ena is enabled at all times. power consumption in the fll is reduced in integer mode; however, the performance may also be reduced, with increased noise or jitter on the output. if low power consumption is required, then fll settings must be chosen when n.k is an integer (ie. fll_k = 0). in this case, the fractional mode can be disabled by setting fll_fracn_ena = 0. for best fll performance, a non-integer value of n.k is required. in this case, the fractional mode must be enabled by setting fll_fracn_ena = 1. the fll settings must be adjusted, if necessary, to produce a non-integer value of n.k. the fll output frequency is generated according to the following equation: f out = (f vco / fll_outdiv) the fll operating frequency, f vco is set according to the following equation: f vco = (f ref x n.k x fll_fratio) see table 48 for the coding of the fll_outdiv and fll_fratio fields. f ref is the input frequency, as determined by fll_clk_ref_div. f vco must be in the range 90-100 mhz. frequencies outside this range cannot be supported. note that the output frequencies that do not lie within the ranges quoted above cannot be guaranteed across the full range of device operating temperatures.
production data wm8912 w pd, rev 4.0, september 2010 71 in order to follow the above requirements for f vco , the value of fll_outdiv should be selected according to the desired output f out . the divider, fll_outdiv, must be set so that f vco is in the range 90-100mhz. the available divisions are integers from 4 to 64. some typical settings of fll_outdiv are noted in table 46. output frequency f out fll_outdiv 2.8125 mhz - 3.125 mhz 011111 (divide by 32) 3.75 mhz - 4.1667 mhz 011000 (divide by 24) 5.625 mhz - 6.25 mhz 001111 (divide by 16) 11.25 mhz - 12.5 mhz 000111 (divide by 8) 18 mhz - 20 mhz 000100 (divide by 5) 22.5 mhz - 25 mhz 000011 (divide by 4) table 46 selection of fll_outdiv the value of fll_fratio should be selected as described in table 47. reference frequency f ref fll_fratio 1mhz - 13.5mhz 0h (divide by 1) 256khz - 1mhz 1h (divide by 2) 128khz - 256khz 2h (divide by 4) 64khz - 128khz 3h (divide by 8) less than 64khz 4h (divide by 16) table 47 selection of fll_fratio in order to determine the remaining fll parameters, the fll operating frequency, f vco , must be calculated, as given by the following equation: f vco = (f out x fll_outdiv) the value of fll_n and fll_k can then be determined as follows: n.k = f vco / (fll_fratio x f ref ) see table 48 for the coding of the fll_outdiv and fll_fratio fields. note that f ref is the input frequency, after division by fll_clk_ref_div, where applicable. in fll fractional mode, the fractional portion of the n.k multiplier is held in the fll_k register field. this field is coded as a fixed point quantity, where the msb has a weighting of 0.5. note that, if desired, the value of this field may be calculated by multiplying k by 2 16 and treating fll_k as an integer value, as illustrated in the following example: if n.k = 8.192, then k = 0.192 multiplying k by 2 16 gives 0.192 x 65536 = 12582.912 (decimal) apply rounding to the nearest integer = 12583 (decimal) = 3127 (hex) for best performance, fll fractional mode should always be used. therefore, if the calculations yield an integer value of n.k, then it is recommended to adjust fll_outdiv in order to obtain a non-integer value of n.k. care must always be taken to ensure that the fll operating frequency, f vco , is within its recommended limits of 90-100 mhz. the register fields that control the fll are described in table 48. example settings for a variety of reference frequencies and output frequencies are shown in table 50.
wm8912 production data w pd, rev 4.0, september 2010 72 register address bit label default description r116 (74h) fll control 1 2 fll_fracn_e na 0 fll fractional enable 0 = integer mode 1 = fractional mode fractional mode (fll_fracn_ena=1) is recommended in all cases 1 fll_osc_ena 0 fll oscillator enable 0 = disabled 1 = enabled fll_osc_ena must be enabled before enabling fll_ena. note that this field is required for free- running fll modes only. 0 fll_ena 0 fll enable 0 = disabled 1 = enabled fll_osc_ena must be enabled before enabling fll_ena. r117 (75h) fll control 2 13:8 fll_outdiv [5:0] 00_0000 fll fout clock divider 00_0000 = reserved 00_0001 = reserved 00_0010 = reserved 00_0011 = 4 00_0100 = 5 00_0101 = 6 ? 11_1110 = 63 11_1111 = 64 (fout = fvco / fll_outdiv) 6:4 fll_ctrl_rat e [2:0] 000 frequency of the fll control block 000 = fvco / 1 (recommended value) 001 = fvco / 2 010 = fvco / 3 011 = fvco / 4 100 = fvco / 5 101 = fvco / 6 110 = fvco / 7 111 = fvco / 8 recommended that these are not changed from default.
production data wm8912 w pd, rev 4.0, september 2010 73 register address bit label default description 2:0 fll_fratio [2:0] 111 fvco clock divider 000 = 1 001 = 2 010 = 4 011 = 8 1xx = 16 000 recommended for high fref 011 recommended for low fref r118 (76h) fll control 3 15:0 fll_k [15:0] 0000h fractional multiply for f ref (msb = 0.5) r119 (77h) fll control 4 14:5 fll_n [9:0] 177h integer multiply for f ref (lsb = 1) 3:0 fll_gain [3:0] 0h gain applied to error 0000 = x 1 (recommended value) 0001 = x 2 0010 = x 4 0011 = x 8 0100 = x 16 0101 = x 32 0110 = x 64 0111 = x 128 1000 = x 256 recommended that these are not changed from default. r120 (78h) fll control 5 4:3 fll_clk_ref_ div [1:0] 00 fll clock reference divider 00 = mclk / 1 01 = mclk / 2 10 = mclk / 4 11 = mclk / 8 mclk (or other input reference) must be divided down to <=13.5mhz. for lower power operation, the reference clock can be divided down further if desired. 1:0 fll_clk_ref_ src [1:0] 00 fll clock source 00 = mclk 01 = bclk 10 = lrclk 11 = reserved table 48 fll register map free-running fll clock the fll can generate a clock signal even when no external reference is available. however, it should be noted that the accuracy of this clock is reduced, and a reference source should always be used where possible. note that, in free-running mode, the fll is not sufficiently accurate for hi-fi dac applications. however, the free-running mode is suitable for clocking most other functions, including the write sequencer, charge pump, dc servo and class w output driver. if an accurate reference clock is available at fll start-up, then the fll should be configured as described above. the fll will continue to generate a stable output clock after the reference input is stopped or disconnected.
wm8912 production data w pd, rev 4.0, september 2010 74 if no reference clock is available at the time of starting up the fll, then an internal clock frequency of approximately 12mhz can be generated by enabling the fll analogue oscillator using the fll_osc_ena register bit, and setting f out clock divider to divide by 8 (fll_outdiv = 07h), as defined in table 48. under recommended operating conditions, the fll output may be forced to approximately 12mhz by then enabling the fll_frc_nco bit and setting fll_frc_nco_val to 19h (see table 49). the resultant sysclk delivers the r equired clock frequencies for the class w output driver, dc servo, charge pump and other functions. note that the value of fll_frc_nco_val may be adjusted to control f out , but care should be taken to maintain the correct relationship between sysclk and the aforementioned functional blocks. register address bit label default description r248 (f8h) fll nco test 1 5:0 fll_frc_nco _val [5:0] 01_1001 fll forced oscillator value valid range is 000000 to 111111 0x19h (011001) = 12mhz approx (note that this field is required for free-running fll modes only) r247 (f7h) fll nco test 0 0 fll_frc_nco 0 fll forced control select 0 = normal 1 = fll oscillator controlled by fll_frc_nco_val (note that this field is required for free-running fll modes only) table 49 fll free-running mode in both cases described above, the fll must be selected as the sysclk source by setting sysclk_src (see table 39). note that, in the absence of any reference clock, the fll output is subject to a very wide tolerance. see ?electrical characteristics? for details of the fll accuracy. example fll calculation to generate 12.288 mhz output (f out ) from a 12.000 mhz reference clock (f ref ): ? set fll_clk_ref_div in order to generate f ref <=13.5mhz: fll_clk_ref_div = 00 (divide by 1) ? set fll_ctrl_rate to the recommended setting: fll_ctrl_rate = 000 (divide by 1) ? sett fll_gain to the recommended setting: fll_gain = 0000 (multiply by 1) ? set fll_outdiv for the required output frequency as shown in table 46:- f out = 12.288 mhz, therefore fll_outdiv = 07h (divide by 8) ? set fll_fratio for the given reference frequency as shown in table 47: f ref = 12mhz, therefore fll_fratio = 0h (divide by 1) ? calculate f vco as given by f vco = f out x fll_outdiv:- f vco = 12.288 x 8 = 98.304mhz ? calculate n.k as given by n.k = f vco / (fll_fratio x f ref ): n.k = 98.304 / (1 x 12) = 8.192 ? determine fll_n and fll_k from the integer and fractional portions of n.k:- fll_n is 8. fll_k is 0.192 ? confirm that n.k is a fractional quantity and set fll_fracn_ena: n.k is fractional. set fll_fracn_ena = 1. note that, if n.k is an integer, then an alternative value of fll_fratio should be selected in order to produce a fractional value of n.k.
production data wm8912 w pd, rev 4.0, september 2010 75 gpio outputs from fll the wm8912 has an internal signal which indicates whether the fll lock has been achieved. the fll lock status is an input to the interrupt control circuit and can be used to trigger an interrupt event - see ?interrupts?. the fll lock signal can be output directly on a gpio pin as an external indication of fll lock. see ?general purpose input/output (gpio)? for details of how to configure a gpio pin to output the fll lock signal. the fll clock can be output directly on a gpio pin as a clock signal for other circuits. note that the fll clock may be output even if the fll is not selected as the wm8912 sysclk source. the clocking configuration is illustrated in figure 44. see ?general purpose input/output (gpio)? for details of how to configure a gpio pin to output the fll clock. example fll settings table 50 provides example fll settings for generating common sysclk fr equencies from a variety of low and high frequency reference inputs. f ref f out fll_clk_ ref_div f vco fll_n fll_k fll_ fratio fll_ outdiv fll_ fracn _ena 32.768 khz 12.288 mhz divide by 1 (0h) 98.304 mhz 187 (0bbh) 0.5 (8000h) 16 (4h) 8 (7h) 1 32.768 khz 11.288576 mhz divide by 1 (0h) 90.308608 mhz 344 (158h) 0.5 (8000h) 8 (3h) 8 (7h) 1 32.768 khz 11.2896 mhz divide by 1 (0h) 90.3168 mhz 344 (158h) 0.53125 (8800h) 8 (3h) 8 (7h) 1 48 khz 12.288 mhz divide by 1 (0h) 98.304 mhz 256 (100h) 0 (0000h) 8 (3h) 8 (7h) 0 12.000 mhz 12.288 mhz divide by 1 (0h) 98.3040 mhz 8 (008h) 0.192 (3127h) 1 (0h) 8 (7h) 1 12.000 mhz 11.289597 mhz divide by 1 (0h) 90.3168 mhz 7 (007h) 0.526398 (86c2h) 1 (0h) 8 (7h) 1 12.288 mhz 12.288 mhz divide by 1 (0h) 98.304 mhz 8 (008h) 0 (0000h) 1 (0h) 8 (7h) 0 12.288 mhz 11.2896 mhz divide by 1 (0h) 90.3168 mhz 7 (007h) 0.35 (599ah) 1 (0h) 8 (7h) 1 13.000 mhz 12.287990 mhz divide by 1 (0h) 98.3040 mhz 7 (007h) 0.56184 (8fd5h) 1 (0h) 8 (7h) 1 13.000 mhz 11.289606 mhz divide by 1 (0h) 90.3168 mhz 6 (006h) 0.94745 (f28ch) 1 (0h) 8 (7h) 1 19.200 mhz 12.287988 mhz divide by 2 (1h) 98.3039 mhz 5 (005h) 0.119995 (1eb8h) 1 (0h) 8 (7h) 1 19.200 mhz 11.289588 mhz divide by 2 (1h) 90.3168 mhz 4 (004h) 0.703995 (b439h) 1 (0h) 8 (7h) 1 table 50 example fll settings
wm8912 production data w pd, rev 4.0, september 2010 76 general purpose input/output (gpio) the wm8912 provides two multi-function pins which can be configured to provide a number of different functions. these are digital input/output pins on the dbvdd power domain. the gpio pins are: ? irq/gpio1 ? bclk/gpio4 each general purpose i/o pin can be configured to be a gpio input or configured as one of a number of output functions. signal de-bouncing can be selected on gpio input pins for use with jack/button detect applications. table 51 lists the functions that are available on each of the gpio pins. gpio pin function gpio pins irq / gpio1 bclk / gpio4 gpio input (including jack/button detect) yes yes gpio output yes yes bclk no yes interrupt (irq) yes yes fll lock output yes yes fll clock output yes yes table 51 gpio functions irq/gpio1 the irq/gpio1 pin is configured using the register bits described in table 52. by default, this pin is irq output with pull-down resistor enabled. register address bit label default description r121 (79h) gpio control 1 5 gpio1_pu 0 gpio1 pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 4 gpio1_pd 1 gpio1 pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled 3:0 gpio1_sel [3:0] 0100 gpio1 function select 0000 = gpio input 0001 = clock output (f=sysclk/opclkdiv) 0010 = logic '0' output 0011 = logic '1' output 0100 = irq output (default) 0101 = fll lock output 0110 = reserved 0111 = reserved 1000 = reserved 1001 = fll clock output 1010 to 1111 = reserved table 52 irq/gpio1 control
production data wm8912 w pd, rev 4.0, september 2010 77 bclk/gpio4 the bclk/gpio4 pin is configured using the register bits described in table 53. by default, this pin provides the bclk function associated with the digital audio interface. the bclk function can operate in slave mode (bclk input) or in master mode (bclk output), depending on the bclk_dir register bit as described in the ?digital audio interface? section. it is possible to configure the bclk/gpio4 pin to provide various gpio functions; in this case, the bclk function is provided using the mclk pin. note that the bclk function is always in slave mode (bclk input) in this mode. to select the gpio4 functions, it is required to set bclk_dir = 0 (see table 35) and to set gpio_bclk_mode_ena = 1 (see table 53 below). in this configuration, the mclk input is used as the bit-clock (bclk) for the digital audio interface. when the bclk/gpio4 pin is configured as gpio4, then the pin function is determined by the gpio_bclk_sel register field. register address bit label default description r124 (7ch) gpio control 4 7 gpio_bclk_mode_ ena 0 selects bclk/gpio4 pin function 0 = bclk/gpio4 is used as bclk 1 = bclk/gpio4 is used as gpio. mclk provides the bclk in the aif in this mode. 3:0 gpio_bclk_sel [3:0] 0000 bclk/gpio4 function select: 0000 = gpio input (default) 0001 = clock output (f=sysclk/opclkdiv) 0010 = logic '0' output 0011 = logic '1' output 0100 = irq output 0101 = fll lock output 0110 = reserved 0111 = reserved 1000 = reserved 1001 = fll clock output 1010 to 1111 = reserved table 53 bclk/gpio4 control
wm8912 production data w pd, rev 4.0, september 2010 78 interrupts the interrupt controller has multiple inputs; these include the gpio input pins and other internal signals. any combination of these inputs can be used to trigger an interrupt output (irq) event. wm8912 interrupt events may be triggered in response to external gpio inputs, fll lock status or write sequencer status. note that the gpio inputs are only supported as interrupt events when the respective pin is configured as a gpio input. there is an interrupt status field associated with each of the irq inputs. these are contained in the interrupt status register (r127), as described in table 54. the status of the irq inputs can be read from this register at any time, or in response to the interrupt output being signalled via a gpio pin. individual mask bits can select or deselect different functions from the interrupt controller. these are listed within the interrupt status mask register (r128), as described in table 54. note that the interrupt status fields remain valid, even when masked, but the masked bits will not cause the interrupt (irq) output to be asserted. the interrupt (irq) output represents the logical ?or? of all unmasked irq inputs. the bits within the interrupt status register (r127) are latching fields and, once set, are not reset until a ?1? is written to the respective register bit in the interrupt status register. the interrupt (irq) output is not reset until each of the unmasked irq inputs has been reset. each of the irq inputs can be individually inverted in the interrupt function, enabling either active high or active low behaviour on each irq input. the polarity inversion is controlled using the bits contained in the interrupt polarity register (r129). each of the irq inputs can be debounced to ensure that spikes and transient glitches do not assert the interrupt output. this is selected using the bits contained in the interrupt debounce register (r130). the wm8912 interrupt controller circuit is illustrated in figure 45. the associated control fields are described in table 54. figure 45 interrupt controller
production data wm8912 w pd, rev 4.0, september 2010 79 register address bit label default description r127 (7fh) interrupt status 10 irq 0 logical or of all other interrupt flags 9 gpio_bclk_eint 0 gpio4 interrupt 0 = interrupt not set 1 = interrupt is set cleared when a ?1? is written 8 wseq_eint 0 write sequence interrupt 0 = interrupt not set 1 = interrupt is set cleared when a ?1? is written. note that the read value of wseq_eint is not valid whilst the write sequencer is busy 5 gpio1_eint 0 gpio1 interrupt 0 = interrupt not set 1 = interrupt is set cleared when a ?1? is written 2 fll_lock_eint 0 fll lock interrupt 0 = interrupt not set 1 = interrupt is set cleared when a ?1? is written r128 (80h) interrupt status mask 9 im_gpio_bclk_ei nt 1 gpio4 interrupt mask 0 = do not mask interrupt 1 = mask interrupt 8 im_wseq_eint 1 write sequencer interrupt mask 0 = do not mask interrupt 1 = mask interrupt 5 im_gpio1_eint 1 gpio1 interrupt mask 0 = do not mask interrupt 1 = mask interrupt 2 im_fll_lock_ein t 1 fll lock interrupt mask 0 = do not mask interrupt 1 = mask interrupt r129 (81h) interrupt polarity 9 gpio_bclk_eint_ pol 0 gpio4 interrupt polarity 0 = active high 1 = active low 8 wseq_eint_pol 0 write sequencer interrupt polarity 0 = active high (interrupt is triggered when wseq is busy) 1 = active low (interrupt is triggered when wseq is idle) 5 gpio1_eint_pol 0 gpio1 interrupt polarity 0 = active high 1 = active low 2 fll_lock_eint_p ol 0 fll lock interrupt polarity 0 = active high (interrupt is triggered when fll lock is reached) 1 = active low (interrupt is triggered when fll is not locked) r130 (82h) interrupt debounce 9 gpio_bclk_eint_ db 0 gpio4 interrupt debounce 0 = disabled 1 = enabled
wm8912 production data w pd, rev 4.0, september 2010 80 register address bit label default description 8 wseq_eint_db 0 write sequencer interrupt debounce enable 0 = disabled 1 = enabled 5 gpio1_eint_db 0 gpio1 input debounce 0 = disabled 1 = enabled 2 fll_lock_eint_d b 0 fll lock debounce 0 = disabled 1 = enabled table 54 interrupt control control interface the wm8912 is controlled by writing to registers through a 2-wire serial control interface. readback is available for all registers, including chip id, power management status and gpio status. note that, if it cannot be assured that mclk is present when accessing the register map, then it is required to set clk_sys_ena = 0 to ensure correct operation. see ?clocking and sample rates? for details of clk_sys_ena. the wm8912 is a slave device on the control interface; sclk is a clock input, while sda is a bi- directional data pin. to allow arbitration of multiple slaves (and/or multiple masters) on the same interface, the wm8912 transmits logic 1 by tri-stating the sda pin, rather than pulling it high. an external pull-up resistor is required to pull the sda line high so that the logic 1 can be recognised by the master. in order to allow many devices to share a single 2-wire control bus, every device on the bus has a unique 8-bit device id (this is not the same as the 8-bit address of each register in the wm8912). the wm8912 device id is 0011 0100 (34h). the lsb of the device id is the read/write bit; this bit is set to logic 1 for ?read? and logic 0 for ?write?. the wm8912 operates as a slave device only. the controller indicates the start of data transfer with a high to low transition on sda while sclk remains high. this indicates that a device id, register address and data will follow. the wm8912 responds to the start condition and shifts in the next eight bits on sda (8-bit device id including read/write bit, msb first). if the device id received matches the device id of the wm8912, then the wm8912 responds by pulling sda low on the next clock pulse (ack). if the device id is not recognised or the r/w bit is ?1? when operating in write only mode, the wm8912 returns to the idle condition and waits for a new start condition and valid address. if the device id matches the device id of the wm8912, the data transfer continues as described below. the controller indicates the end of data transfer with a low to high transition on sda while sclk remains high. after receiving a complete address and data sequence the wm8912 returns to the idle state and waits for another start condition. if a start or stop condition is detected out of sequence at any point during data transfer (i.e. sda changes while sclk is high), the device returns to the idle condition. the wm8912 supports the following read and write operations: ? single write ? single read ? multiple write using auto-increment ? multiple read using auto-increment
production data wm8912 w pd, rev 4.0, september 2010 81 the sequence of signals associated with a single register write operation is illustrated in figure 46. figure 46 control interface register write the sequence of signals associated with a single register read operation is illustrated in figure 47. a1 d7 d1 a7 sda sclk device id device id a0 d6 b9 b8 note: the sda pin is driven by both the master and slave devices in turn to transfer device address, register address, data and ack responses r/w d0 b15 register address data bits b15 ? b8 start (write) ack ack ack r/w (read) b1 b0 b7 data bits b15 ? b8 ack stop ack rpt start figure 47 control interface register read the control interface also supports other register operations, as listed above. the interface protocol for these operations is summarised below. the terminology used in the following figures is detailed in table 55. note that multiple write and multiple read operations are supported using the auto-increment mode. this feature enables the host processor to access sequential blocks of the data in the wm8912 register map faster than is possible with single register operations. terminology description s start condition sr repeated start a acknowledge (sda low) a not acknowledge (sda high) p stop condition r/w readnotwrite 0 = write 1 = read [white field] data flow from bus master to wm8912 [grey field] data flow from wm8912 to bus master table 55 control interface terminology figure 48 single register write to specified address
wm8912 production data w pd, rev 4.0, september 2010 82 figure 49 single register read from specified address figure 50 multiple register write to specified address using auto-increment figure 51 multiple register read from specified address using auto-increment figure 52 multiple register read from last address using auto-increment
production data wm8912 w pd, rev 4.0, september 2010 83 control write sequencer the control write sequencer is a programmable unit that forms part of the wm8912 control interface logic. it provides the ability to perform a sequence of register write operations with the minimum of demands on the host processor - the sequence may be initiated by a single operation from the host processor and then left to execute independently. default sequences for start-up and shutdown are provided (see ?default sequences? section). it is recommended that these default sequences are used unless changes become necessary. when a sequence is initiated, the sequencer performs a series of pre-defined register writes. the host processor informs the sequencer of the start index of the required sequence within the sequencer?s memory. at each step of the sequence, the contents of the selected register fields are read from the sequencer?s memory and copied into the wm8912 control registers. this continues sequentially through the sequencer?s memory until an ?end of sequence? bit is encountered; at this point, the sequencer stops and an interrupt status flag is asserted. for cases where the timing of the write sequence is important, a programmable delay can be set for specific steps within the sequence. note that the control write sequencer?s internal clock is derived from the internal clock sysclk. an external mclk signal must be present when using the control write sequencer, and sysclk must be enabled by setting clk_ sys_ena (see ?clocking and sample rates?). the clock division from mclk is handled transparently by the wm8912 without user intervention, as long as mclk and sample rates are set correctly. initiating a sequence the register fields associated with running the control write sequencer are described in table 56. the write sequencer clock is enabled by setting the wseq_ena bit. note that the operation of the control write sequencer also requires the internal clock sysclk to be enabled via the clk_sys_ena (see ?clocking and sample rates?). the start index of the required sequence must be written to the wseq_start_index field. setting the wseq_start bit initiates the sequencer at the given start index. the write sequencer can be interrupted by writing a logic 1 to the wseq_abort bit. the current status of the write sequencer can be read using two further register fields - when the wseq_busy bit is asserted, this indicates that the write sequencer is busy. note that, whilst the control write sequencer is running a sequence (indicated by the wseq_busy bit), normal read/write operations to the control registers cannot be supported. (the write sequencer registers and the software reset register can still be accessed when the sequencer is busy.) the index of the current step in the write sequencer can be read from the wseq_current_index field; this is an indicator of the sequencer?s progress. on completion of a sequence, this field holds the index of the last step within the last commanded sequence. when the write sequencer reaches the end of a sequence, it asserts the wseq_eint flag in register r127 (see table 54 within the ?interrupts? section). this flag can be used to generate an interrupt event on completion of the sequence. note that the wseq_eint flag is asserted to indicate that the write sequencer is not busy. register address bit label default description r108 (6ch) write sequencer 0 8 wseq_ena 0 write sequencer enable. 0 = disabled 1 = enabled r111 (6fh) write sequencer 3 9 wseq_abort 0 writing a 1 to this bit aborts the current sequence and returns control of the device back to the serial control interface.
wm8912 production data w pd, rev 4.0, september 2010 84 register address bit label default description 8 wseq_start 0 writing a 1 to this bit starts the write sequencer at the memory location indicated by the wseq_start_index field. the sequence continues until it reaches an ?end of sequence? flag. at the end of the sequence, this bit will be reset by the write sequencer. 5:0 wseq_start_ index [5:0] 00_0000 sequence start index. this is the memory location of the first command in the selected sequence. 0 to 31 = ram addresses 32 to 48 = rom addresses 49 to 63 = reserved r112 (70h) write sequencer 4 9:4 wseq_curre nt_index [5:0] 00_0000 sequence current index (read only): this is the location of the most recently accessed command in the write sequencer memory. 0 wseq_busy 0 sequencer busy flag (read only): 0 = sequencer idle 1 = sequencer busy note: it is not possible to write to control registers via the control interface while the sequencer is busy. table 56 write sequencer control - initiating a sequence programming a sequence a sequence consists of write operations to data bits (or groups of bits) within the control registers. the register fields associated with programming the control write sequencer are described in table 57. for each step of the sequence being programmed, the sequencer index must be written to the wseq_write_index field. the values 0 to 31 correspond to all the available ram addresses within the write sequencer memory. (note that memory addresses 32 to 48 also exist, but these are rom addresses, which are not programmable.) having set the index as described above, register r109 must be written to (containing the control register address, the start bit position and the field width applicable to this step of the sequence). also, register r110 must be written to (containing the register data, the end of sequence flag and the delay time required after this step is executed). after writing to these two registers, the next step in the sequence may be programmed by updating wseq_write_index and repeating the procedure. wseq_addr is an 8-bit field containing the control register address in which the data should be written. wseq_data_start is a 4-bit field which identifies the lsb position within the selected control register to which the data should be written. setting wseq_data_start = 0100 will cause 1-bit data to be written to bit 4. with this setting, 4-bit data would be written to bits 7:4 and so on. wseq_data_width is a 3-bit field which identifies the width of the data block to be written. this enables selected portions of a control register to be updated without any concern for other bits within the same register, eliminating the need for read-modify-write procedures. values of 0 to 7 correspond to data widths of 1 to 8 respectively. for example, setting wseq_data_width = 010 will cause a 3-bit data block to be written. note that the maximum value of this field corresponds to an 8-bit data block; writing to register fields greater than 8 bits wide must be performed using two separate operations of the control write sequencer.
production data wm8912 w pd, rev 4.0, september 2010 85 wseq_data is an 8-bit field which contains the data to be written to the selected control register. the wseq_data_width field determines how many of these bits are written to the selected register; the most significant bits (above the number indicated by wseq_data_width) are ignored. wseq_delay is a 4-bit field which controls the waiting time between the current step and the next step in the sequence. the total delay time per step (including execution) is given by: t = k (2 wseq_delay + 8) where k = 62.5 s (under recommended operating conditions) this gives a useful range of execution/delay times from 562 s up to 2.048s per step. wseq_eos is a 1-bit field which indicates the end of sequence. if this bit is set, then the control write sequencer will automatically stop after this step has been executed. register address bit label default description r108 (6ch) write sequencer 0 4:0 wseq_write_ index [4:0] 0_0000 sequence write index. this is the memory location to which any updates to r109 and r110 will be copied. 0 to 31 = ram addresses r109 (6dh) write sequencer 1 14:12 wseq_data_ width [2:0] 000 width of the data block written in this sequence step. 000 = 1 bit 001 = 2 bits 010 = 3 bits 011 = 4 bits 100 = 5 bits 101 = 6 bits 110 = 7 bits 111 = 8 bits 11:8 wseq_data_ start [3:0] 0000 bit position of the lsb of the data block written in this sequence step. 0000 = bit 0 ? 1111 = bit 15 7:0 wseq_addr [7:0] 0000_000 0 control register address to be written to in this sequence step. r110 (6eh) write sequencer 2 14 wseq_eos 0 end of sequence flag. this bit indicates whether the control write sequencer should stop after executing this step. 0 = not end of sequence 1 = end of sequence (stop the sequencer after this step). 11:8 wseq_delay [3:0] 0000 time delay after executing this step. total delay time per step (including execution)= 62.5 s (2^wseq_delay + 8) 7:0 wseq_data [7:0] 0000_000 0 data to be written in this sequence step. when the data width is less than 8 bits, then one or more of the msbs of wseq_data are ignored. it is recommended that unused bits be set to 0. table 57 write sequencer control - programming a sequence
wm8912 production data w pd, rev 4.0, september 2010 86 note that a ?dummy? write can be inserted into a control sequence by commanding the sequencer to write a value of 0 to bit 0 of register r255 (ffh). this is effectively a write to a non-existent register location. this can be used in order to create placeholders ready for easy adaptation of the sequence. for example, a sequence could be defined to power-up a mono signal path from dacl to headphone, with a ?dummy? write included to leave space for easy modification to a stereo signal path configuration. dummy writes can also be used in order to implement additional time delays between register writes. dummy writes are included in the default start-up sequence ? see table 59. in summary, the control register to be written is set by the wseq_addr field. the data bits that are written are determined by a combination of wseq_data_start, wseq_data_width and wseq_data. this is illustrated below for an example case of writing to the vmid_res field within register r5 (05h). figure 53 control write sequencer example in this example, the start position is bit 01 (wseq_data_start = 0001b) and the data width is 2 bits (wseq_data_width = 0001b). with these settings, the control write sequencer would updated the control register r5 [2:1] with the contents of wseq_data [1:0]. default sequences when the wm8912 is powered up, two control write sequences are available through default settings in both ram and rom memory locations. the purpose of these sequences, and the register write required to initiate them, is summarised in table 58. a single register write will initiate the sequence in both cases. wseq start index wseq finish index purpose to initiate 0 (00h) 22 (16h) start-up sequence write 0100h to register r111 (6fh) 25 (19h) 39 (27h) shutdown sequence write 0119h to register r111 (6fh) table 58 write sequencer default sequences note on shutdown sequence: the instruction at index address 25 (19h) shorts the outputs lineoutl and lineoutr. if the line outputs are not in use at the time the sequence is run, then the sequence could, instead, be started at index address 26.
production data wm8912 w pd, rev 4.0, september 2010 87 index addresses 0 to 24 may be programmed to users? own settings at any time, as described in ?programming a sequence?. users? own settings remain in memory and are not affected by software resets (i.e. writing to register r0). however, any non-default sequences are lost when the device is powered down. start-up sequence the start-up sequence is initiated by writing 0100h to register r111 (6fh). this single operation starts the control write sequencer at index address 0 (00h) and executes the sequence defined in table 59. for typical clocking configurations with mclk=12.288mhz, this sequence takes approximately 300ms to run. note that, for fast startup, step 18 may be overwritten with dummy data in order to achieve startup within 50ms (see ?quick start-up and shutdown?). wseq index register address width start data delay eos description 0 (00h) r4 (04h) 5 bits bit 0 1ah 0h 0b bias_ena = 0 (delay = 0.5625ms) 1 (01h) r5 (05h) 8 bits bit 0 47h 6h 0b vmid_res [1:0] = 11b vmid_ena = 1 (delay = 4.5ms) 2 (02h) r5 (05h) 2 bits bit 1 01h 0h 0b vmid_res [1:0] = 01b (delay = 0.5625ms) 3 (03h) r4 (04h) 1 bit bit 0 01h 0h 0b bias_ena = 1 (delay = 0.5625ms) 4 (04h) r14 (0eh) 2 bits bit 0 03h 0h 0b hpl_pga_ena = 1 hpr_pga_ena = 1 (delay = 0.5625ms) 5 (05h) r15 (0fh) 2 bits bit 0 03h 0h 0b lineoutl_pga_ena = 1 lineoutr_pga_ena = 1 (delay = 0.5625ms) 6 (06h) r22 (16h) 1 bit bit 1 01h 0h 0b clk_dsp_ena = 1 (delay = 0.5625ms) 7 (07h) r18 (12h) 2 bits bit 2 03h 5h 0b dacl_ena = 1 dacr_ena = 1 (delay = 2.5ms) 8 (08h) r255 (ffh) 1 bit bit 0 00h 0h 0b dummy write for expansion (delay = 0.5625ms) 9 (09h) r4 (04h) 1 bit bit 4 00h 0h 0b (delay = 0.5625ms) 10 (0ah) r98 (62h) 1 bit bit 0 01h 6h 0b cp_ena = 1 (delay = 4.5ms) 11 (0bh) r255 (ffh) 1 bit bit 0 00h 0h 0b dummy write for expansion (delay = 0.5625ms) 12 (0ch) r90 (5ah) 8 bits bit 0 11h 0h 0b hpl_ena = 1 hpr_ena = 1 (delay = 0.5625ms) 13 (0dh) r94 (5eh) 8 bits bit 0 11h 0h 0b lineoutl_ena = 1 lineoutr_ena = 1 (delay = 0.5625ms) 14 (0eh) r90 (5ah) 8 bits bit 0 33h 0h 0b hpl_ena_dly = 1 hpr_ena_dly = 1 (delay = 0.5625ms) 15 (0fh) r94 (5eh) 8 bits bit 0 33h 0h 0b lineoutl_ena_dly = 1 lineoutr_ena_dly = 1
wm8912 production data w pd, rev 4.0, september 2010 88 wseq index register address width start data delay eos description (delay = 0.5625ms) 16 (10h) r67 (43h) 4 bits bit 0 0fh ch 0b dcs_ena_chan_0 = 1 dcs_ena_chan_1 = 1 dcs_ena_chan_2 = 1 dcs_ena_chan_3 = 1 (delay = 0.5625ms) 17 (11h) r68 (44h) 8 bits bit 0 f0h 0h 0b dcs_trig_startup_0 = 1 dcs_trig_startup_1 = 1 dcs_trig_startup_2 = 1 dcs_trig_startup_3 = 1 (delay = 256.5ms) 18 (12h) r255 (ffh) 1 bit bit 0 00h 0h 0b dummy write for expansion (delay = 0.5625ms) 19 (13h) r90 (5ah) 8 bits bit 0 77h 0h 0b hpl_ena_outp = 1 hpr_ena_outp = 1 (delay = 0.5625ms) 20 (14h) r94 (5eh) 8 bits bit 0 77h 0h 0b lineoutl_ena_outp = 1 lineoutr_ena_outp = 1 (delay = 0.5625ms) 21 (15h) r90 (5ah) 8 bits bit 0 ffh 0h 0b hpl_rmv_short = 1 hpr_rmv_short = 1 (delay = 0.5625ms) 22 (16h) r94 (5eh) 8 bits bit 0 ffh 0h 1b lineoutl_rmv_short = 1 lineoutr_rmv_short = 1 end of sequence 23 (17h) r255 (ffh) 1 bit bit 0 00h 0h 0b spare 24 (18h) r255 (ffh) 1 bit bit 0 00h 0h 0b spare table 59 start-up sequence shutdown sequence the shutdown sequence is initiated by writing 0119h to register r111 (6fh). this single operation starts the control write sequencer at index address 25 (19h) and executes the sequence defined in table 60. for typical clocking configurations with mclk=12.288mhz, this sequence takes approximately 350ms to run. wseq index register address width start data delay eos description 25 (19h) r94 (5eh) 8 bits bit 0 77h 0h 0b lineoutl_rmv_short = 0 lineoutr_rmv_short = 0 (delay = 0.5625ms) 26 (1ah) r90 (5ah) 8 bits bit 0 77h 0h 0b hpl_rmv_short = 0 hpr_rmv_short = 0 (delay = 0.5625ms) 27 (1bh) r90 (5ah) 8 bits bit 0 00h 0h 0b hpl_ena_outp = 0 hpl_ena_dly = 0 hpl_ena = 0 hpr_ena_outp = 0 hpr_ena_dly = 0 hpr_ena = 0 (delay = 0.5625ms)
production data wm8912 w pd, rev 4.0, september 2010 89 wseq index register address width start data delay eos description 28 (1ch) r94 (5eh) 8 bits bit 0 00h 0h 0b lineoutl_ena_outp = 0 lineoutl_ena_dly = 0 lineoutl_ena = 0 lineoutr_ena_outp = 0 lineoutr_ena_dly = 0 lineoutr_ena = 0 (delay = 0.5625ms) 29 (1dh) r67 (43h) 4 bits bit 0 00h 0h 0b dcs_ena_chan_0 = 0 dcs_ena_chan_1 = 0 dcs_ena_chan_2 = 0 dcs_ena_chan_3 = 0 (delay = 0.5625ms) 30 (1eh) r98 (62h) 1 bit bit 0 00h 0h 0b cp_ena = 0 (delay = 0.5625ms) 31 (1fh) r18 (12h) 2 bits bit 2 00h 0h 0b dacl_ena = 0 dacr_ena = 0 (delay = 0.5625ms) 32 (20h) r22 (16h) 1 bit bit 1 00h 0h 0b clk_dsp_ena = 0 (delay = 0.5625ms) 33 (21h) r14 (0eh) 2 bits bit 0 00h 0h 0b hpl_pga_ena = 0 hpr_pga_ena = 0 (delay = 0.5625ms) 34 (22h) r15 (0fh) 2 bits bit 0 00h 0h 0b lineoutl_pga_ena = 0 lineoutr_pga_ena = 0 (delay = 0.5625ms) 35 (23h) r4 (04h) 1 bit bit 0 00h 0h 0b bias_ena = 0 (delay = 0.5625ms) 36 (24h) r5 (05h) 1 bit bit 0 00h ch 0b vmid_ena = 0 (delay = 256.5ms) 37 (25h) r5 (05h) 1 bit bit 0 00h 9h 0b vmid_ena = 0 (delay = 32.5ms) 38 (26h) r5 (05h) 8 bits bit 0 00h 0h 0b vmid_res [1:0] = 00 vmid_ena = 0 (delay = 0.5625ms) 39 (27h) r4 (04h) 2 bits bit 0 00h 0h 1b bias_ena = 0 end of sequence table 60 shutdown sequence
wm8912 production data w pd, rev 4.0, september 2010 90 power-on reset the wm8912 includes an internal power-on-reset (por) circuit, which is used to reset the digital logic into a default state after power up. the por circuit is powered from avdd and monitors dcvdd. the internal por si gnal is asserted low when avdd and dcvdd are below minimum thresholds. the specific behaviour of the circuit will vary, depending on the relative timing of the supply voltages. typical scenarios are illustrated in figure 54 and figure 55. dcvdd v pord_on 0v avdd 0v v pora v pora_off lo hi internal por device ready por active por active por undefined figure 54 power on reset timing - avdd enabled first figure 55 power on reset timing - dcvdd enabled first
production data wm8912 w pd, rev 4.0, september 2010 91 the por si gnal is undefined until avdd has exceeded the minimum threshold, v pora once this threshold has been exceeded, por is asserted low and the chip is held in reset. in this condition, all writes to the control interface are ignored. once avdd and dcvdd have reached their respective power on thresholds, por is released high, all registers are in their default state, and writes to the control interface may take place. note that a minimum power-on reset period, t por , applies even if avdd and dcvdd have zero rise time. (this specification is guaranteed by design rather than test.) on power down, por is asserted low w hen any of avdd or dcvdd falls below their respective power-down thresholds. typical power-on reset parameters for the wm8912 are defined in table 61. symbol description typ unit v pora avdd threshold below which por is undefined 0.25 v v pora_on power-on threshold (avdd) 1.15 v v pora_off power-off threshold (avdd) 1.12 v v pord_on power-on threshold (dcvdd) 0.57 v v pord_off power-off threshold (dcvdd) 0.55 v t por minimum power-on reset period 9.5 s table 61 typical power-on reset parameters notes: 1. if avdd and dcvdd suffer a brown-out (i.e. drop below the minimum recommended operating level but do not go below v pora_off or v pord_off ) then the chip does not reset and resumes normal operation when the voltage is back to the recommended level again. 2. the chip enters reset at power down when avdd or dcvdd falls below v pora_off or v pord_off . this may be important if the supply is turned on and off frequently by a power management system. 3. the minimum t por period is maintained even if dcvdd and avdd have zero rise time. this specification is guaranteed by design rather than test. quick start-up and shutdown the wm8912 has the capability to perform a quick start-up and shutdown with a minimum number of register operations. this is achieved using the control write sequencer, which is configured with default start-up settings that set up the device for dac playback via headphone and line output. assuming a 12.288mhz external clock, the start-up sequence configures the device for 48khz playback mode. the default start-up sequence requires three register write operations. the default shutdown sequence requires just a single register write. the minimum procedure for executing the quick start- up and shutdown sequences is described below. see ?control write sequencer? for more details. after the default start-up sequence has been performed, the dc offset correction values will be held in memory, provided that power is maintained and a software reset is not performed. fast start-up using the stored values of dc offset correction is also possible, as described below. quick start-up (default sequence) an external clock must be applied to mclk. assuming 12.288mhz input clock, the start-up sequence will take approximately 300ms to complete. the following register operations will initiate the quick start-up sequence.
wm8912 production data w pd, rev 4.0, september 2010 92 register address value description r108 (6ch) write sequencer 0 0100h wseq_ena = 1 wseq_write_index = 0_0000 r111 (6fh) write sequencer 3 0100h wseq_abort = 0 wseq_start = 1 wseq_start_index = 0_0000 r33 (21h) dac digital 1 0000h dac_mono = 0 dac_sb_filt = 0 dac_muterate = 0 dac_unmute_ramp = 0 dac_osr128 = 0 dac_mute = 0 deemph = 00 table 62 quick start-up control the wseq_busy bit (in register r112, see table 56) will be set to 1 while the sequence runs. when this bit returns to 0, the device has been set up and is ready for dac playback operation. fast start-up from standby the default start-up sequence runs the dc servo to remove dc offsets from the outputs. the offset for this path selection is then stored in memory. provided that power is maintained to the chip, and a software reset is not performed, then the dc offset correction will be held in memory on the wm8912. this allows the dc servo calibrations to be omitted from the start-up sequence if the offset correction has already been performed. by omitting this part of the start-up sequence, a fast start-up time of less than 50ms can be achieved. the register write sequence described in table 63 replaces the default dc servo operation with dummy operations, allowing a fast start-up to be achieved, assuming the device is initially in a standby condition with dc offset correction previously performed. note that, if power is removed from the wm8912 or if a software reset is performed, then the default sequence will be restored, and the dc offset correction will be necessary on the output paths once more. register address value description r108 (6ch) write sequencer 0 0111h wseq_ena = 1 wseq_write_index = 1_0001 r109 (6dh) write sequencer 1 00ffh wseq_data_width = 000 wseq_data_start = 0000 wseq_addr = 1111_1111 r110 (6eh) write sequencer 2 0000h wseq_eos = 0 wseq_delay = 0000 wseq_data = 0000_0000 r111 (6fh) write sequencer 3 0100h wseq_abort = 0 wseq_start = 1 wseq_start_index = 00_0000 r33 (21h) dac digital 1 0000h dac_mono = 0 dac_sb_filt = 0 dac_muterate = 0 dac_unmute_ramp = 0 dac_osr128 = 0 dac_mute = 0 deemph = 00 table 63 fast start-up from standby control
production data wm8912 w pd, rev 4.0, september 2010 93 the wseq_busy bit (in register r112, see table 56) will be set to 1 while the sequence runs. when this bit returns to 0, the device has been set up and is ready for dac playback operation. quick shutdown (default sequence) the default shutdown sequences assumes the initial device conditions are as configured by the default start-up sequence. assuming 12.288mhz input clock, the shutdown sequence will take approximately 350ms to complete. the following register operation will initiate the default shutdown sequence. register address value description r111 (6fh) write sequencer 3 0119h wseq_abort = 0 wseq_start = 1 wseq_start_index = 19h table 64 quick shutdown control the wseq_busy bit (in register r112, see table 56) will be set to 1 while the sequence runs. when this bit returns to 0, the system clock can be disabled (clk_sys_ena=0) and mclk can be stopped. software reset and chip id a software reset can be commanded by writing to register r0. this is a read-only register field and the contents will not be affected by writing to this register. the chip id can be read back from register r0. register address bit label default description r0 (00h) sw reset and id 15:0 sw_rst_de v_id1 [15:0] 8904h writing to this register resets all registers to their default state. reading from this register will indicate device id 8904h. table 65 software reset and chip id
wm8912 production data w pd, rev 4.0, september 2010 94 register map dec addr hex addr name 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 bin default 000sw reset and id 1000_1001_0000_0100 4 04 bias control 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 bias_ena 0000_0000_0001_1000 5 05 vmid control 0 0 0 0 0 0 0 0 0 0 0 0 0 0 vmid_ena 0000_0000_0000_0000 14 0e power management 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 hpl_pga_ena hpr_pga_ena 0000_0000_0000_0000 150fpower management 300000000000000 lineoutl_pga _ ena lineoutr_pga _ena 0000_0000_0000_0000 18 12 power management 6 0 0 0 0 0 0 0 0 0 0 0 0 dacl_ena dacr_ena 0 0 0000_0000_0000_0000 20 14 clock rates 0 1 toclk_rate_di v16 toclk_rate_x 4 0 1 1 0 0 0 1 0 1 1 1 1 mclk_div 1000_1100_0101_1110 21 15 clock rates 1 0 0 0000000 0000_1100_0000_0101 22 16 clock rates 2 mclk_inv sysclk_src 0 toclk_rate 0 0 0 0 0 0 0 0 opclk_ena clk_sys_ena clk_dsp_ena toclk_ena 0000_0000_0000_0000 24 18 audio interface 0 0 0 0 dacl_datinv dacr_datinv 0 0 1 aifdacl_src aifdacr_src 0 0 dac_comp dac_compmod e 0000_0000_0101_0000 25 19 audio interface 1 0 0 aifdac_tdm aifdac_tdm_chan 000 aif_tris aif_bclk_inv bclk_dir 0 aif_lrclk_inv 0000_0000_0000_1010 26 1a audio interface 2 0 0 0 0 111 0000_0000_1110_0100 27 1b audio interface 3 0 0 0 0 lrclk_dir 0000_0000_0100_0000 30 1e dac digital volume left 0 0 0 0 0 0 0 dac_vu 0000_000p_1100_0000 31 1f dac digital volume right 0 0 0 0 0 0 0 dac_vu 0000_000p_1100_0000 33 21 dac digital 1 0 0 0 dac_mono dac_sb_filt dac_muterat e dac_unmute_ ramp 0 0 dac_osr128 0 0 dac_mute 0 0000_0000_0000_1000 40 28 drc 0 drc_ena drc_dac_path 0 drc_ff_delay 0 drc_gs_ena drc_qr drc_anticlip drc_gs_hyst 0000_0001_1010_1111 41 29 drc 1 0011_0010_0100_1000 42 2a drc 2 0 0 0 0 0 0 0 0 0 0 0000_0000_0000_0000 43 2b drc 3 0 0 0 0 0 0000_0000_0000_0000 57 39 analogue out1 left 0 0 0 0 0 0 0 hpoutl_mute hpout_vu hpoutlzc 0000_0000_p010_1101 58 3a analogue out1 right 0 0 0 0 0 0 0 hpoutr_mute hpout_vu hpoutrzc 0000_0000_p010_1101 59 3b analogue out2 left 0 0 0 0 0 0 0 lineoutl_mute lineout_vu lineoutlzc 0000_0000_p011_1001 60 3c analogue out2 right 0 0 0 0 0 0 0 lineoutr_mute lineout_vu lineoutrzc 0000_0000_p011_1001 6743dc servo 0 000000000000 dcs_ena_chan _3 dcs_ena_chan _2 dcs_ena_chan _1 dcs_ena_chan _0 0000_0000_0000_0000 68 44 dc servo 1 dcs_trig_sing le_3 dcs_trig_sing le_2 dcs_trig_sing le_1 dcs_trig_sing le_0 dcs_trig_seri es_3 dcs_trig_seri es_2 dcs_trig_seri es_1 dcs_trig_seri es_0 dcs_trig_sta rtup_3 dcs_trig_sta rtup_2 dcs_trig_sta rtup_1 dcs_trig_sta rtup_0 dcs_trig_dac _wr_3 dcs_trig_dac _wr_2 dcs_trig_dac _wr_1 dcs_trig_dac _wr_0 pppp_pppp_pppp_pppp 69 45 dc servo 2 0 0 0 0 0000 1010_1010_1010_1010 7147dc servo 4 000000000 1010_1010_1010_1010 7248dc servo 5 000000000 1010_1010_1010_1010 lrclk_rate[10:0] lineoutr_vol[5:0] lineoutl_vol[5:0] dcs_series_no_01[6:0] hpoutl_vol[5:0] drc_qr_dcy[1:0] hpoutr_vol[5:0] dcs_timer_period_23[3:0] drc_knee_op[4:0] dcs_timer_period_01[3:0] drc_atk[3:0] drc_dcy[3:0] drc_maxgain[1:0] drc_knee_ip[5:0] drc_mingain[1:0] drc_qr_thr[1:0] drc_lo_comp[2:0] dcs_series_no_23[6:0] deemph[1:0] aif_wl[1:0] dacr_vol[7:0] dacl_vol[7:0] drc_hi_comp[2:0] opclk_div[3:0] bclk_div[4:0] sw_rst_dev_id1[15:0] vmid_res[1:0] clk_sys_rate[3:0] aif_fmt[1:0] dac_boost[1:0] sample_rate[2:0] drc_startup_gain[4:0] drc_gs_hyst_lvl[1:0]
production data wm8912 w pd, rev 4.0, september 2010 95 dec addr hex addr name 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 bin default 7349dc servo 6 00000000 0000_0000_0000_0000 744adc servo 7 00000000 0000_0000_0000_0000 754bdc servo 8 00000000 0000_0000_0000_0000 764cdc servo 9 00000000 0000_0000_0000_0000 77 4d dc servo readback 0 0 0 0 0 0000_0000_0000_0000 905aanalogue hp 000000000 hpl_rmv_sho rt hpl_ena_outp hpl_ena_dly hpl_ena hpr_rmv_sho rt hpr_ena_outp hpr_ena_dly hpr_ena 0000_0000_0000_0000 945eanalogue lineout 000000000 lineoutl_rmv _short lineoutl_ena_ outp lineoutl_ena_ dly lineoutl_ena lineoutr_rmv _short lineoutr_ena _outp lineoutr_ena _dly lineoutr_ena 0000_0000_0000_0000 9862charge pump 0000000000000000cp_ena0000_0000_0000_0000 104 68 class w 0 0000000000000 1 0 cp_dyn_pwr 0000_0000_0000_0100 1086cwrite sequencer 00000000wseq_ena000 0000_0000_0000_0000 109 6d write sequencer 1 0 0000_0000_0000_0000 110 6e write sequencer 2 0 wseq_eos 0 0 0000_0000_0000_0000 1116fwrite sequencer 3000000wseq_abortwseq_start00 0000_0000_0000_0000 11270write sequencer 4000000 0 0 0 wseq_busy 0000_0000_0000_0000 11674fll control 10000000000000 fll_fracn_ena fll_osc_ena fll_ena 0000_0000_0000_0000 117 75 fll control 2 0 0 0 0 0000_0000_0000_0111 118 76 fll control 3 0000_0000_0000_0000 119 77 fll control 4 0 0 0010_1110_1110_0000 12078fll control 500000000000 1 0000_0000_0000_0100 121 79 gpio control 1 0000 000000 gpio1_pu gpio1_pd 0000_0000_0001_0100 124 7c gpio control 4 0000 0000 gpio_bclk_mo de_ena 000 0000_0000_0000_0000 1267edigital pulls 00000000mclk_pumclk_pddacdat_pudacdat_pdlrclk_pulrclk_pdbclk_pubclk_pd0000_0000_0000_0000 1277finterrupt status00000irq gpio_bclk_ein t wseq_eint 0 0 gpio1_eint 0 0 fll_lock_eint 0 0 xxxx_xppp_pppp_pppp 12880interrupt status mask000000 im_gpio_bclk_ eint im_wseq_eint 1 1 im_gpio1_eint 1 1 im_fll_lock_e int 1 1 1111_1111_1111_1111 12981interrupt polarity000000 gpio_bclk_ein t_pol wseq_eint_po l 00 gpio1_eint_po l 00 fll_lock_eint _pol 0 0 0000_0000_0000_0000 13082interrupt debounce000000 gpio_bclk_ein t_db wseq_eint_db 0 0 gpio1_eint_db 0 0 fll_lock_eint _db 0 0 0000_0000_0000_0000 13486 eq1 000000000000000eq_ena0000_0000_0000_0000 dcs_cal_complete[3:0] dcs_startup_complete[3:0] dcs_dac_wr_val_0[7:0] dcs_dac_wr_complete[3:0] wseq_current_index[5:0] wseq_write_index[4:0] wseq_addr[7:0] wseq_delay[3:0] fll_n[9:0] gpio1_sel[3:0] dcs_dac_wr_val_3[7:0] dcs_dac_wr_val_1[7:0] wseq_start_index[5:0] fll_gain[3:0] dcs_dac_wr_val_2[7:0] fll_ctrl_rate[2:0] fll_fratio[2:0] fll_k[15:0] fll_outdiv[5:0] wseq_data_width[2:0] wseq_data[7:0] wseq_data_start[3:0] fll_clk_ref_div[1:0] fll_clk_ref_src[1:0] gpio_bclk_sel[3:0]
wm8912 production data w pd, rev 4.0, september 2010 96 dec addr hex addr name 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 bin default 13587 eq2 00000000000 0000_0000_0000_1100 13688 eq3 00000000000 0000_0000_0000_1100 13789 eq4 00000000000 0000_0000_0000_1100 1388a eq5 00000000000 0000_0000_0000_1100 1398b eq6 00000000000 0000_0000_0000_1100 140 8c eq7 0000_1111_1100_1010 141 8d eq8 0000_0100_0000_0000 142 8e eq9 0000_0000_1101_1000 143 8f eq10 0001_1110_1011_0101 144 90 eq11 1111_0001_0100_0101 145 91 eq12 0000_1011_0111_0101 146 92 eq13 0000_0001_1100_0101 147 93 eq14 0001_1100_0101_1000 148 94 eq15 1111_0011_0111_0011 149 95 eq16 0000_1010_0101_0100 150 96 eq17 0000_0101_0101_1000 151 97 eq18 0001_0110_1000_1110 152 98 eq19 1111_1000_0010_1001 153 99 eq20 0000_0111_1010_1101 154 9a eq21 0001_0001_0000_0011 155 9b eq22 0000_0101_0110_0100 156 9c eq23 0000_0101_0101_1001 157 9d eq24 0100_0000_0000_0000 247f7fll nco test 0000000000000000 fll_frc_nco 0000_0000_0000_0000 248f8fll nco test 10000000000 0000_0000_0001_1001 fll_frc_nco_val[5:0] eq_b3_pg[15:0] eq_b5_a[15:0] eq_b4_a[15:0] eq_b4_b[15:0] eq_b3_c[15:0] eq_b3_b[15:0] eq_b2_a[15:0] eq_b2_pg[15:0] eq_b1_gain[4:0] eq_b2_gain[4:0] eq_b1_b[15:0] eq_b3_a[15:0] eq_b2_b[15:0] eq_b2_c[15:0] eq_b3_gain[4:0] eq_b4_gain[4:0] eq_b1_pg[15:0] eq_b5_gain[4:0] eq_b1_a[15:0] eq_b4_c[15:0] eq_b4_pg[15:0] eq_b5_b[15:0] eq_b5_pg[15:0]
production data wm8912 w pd, rev 4.0, september 2010 97 register bits by address register address bit label default description refer to r0 (00h) sw reset and id 15:0 sw_rst_dev _id1 [15:0] 1000_1001 _0000_010 0 writing to this register resets all registers to their default state. reading from this register will indicate device id 8904h. register 00h sw reset and id register address bit label default description refer to r4 (04h) bias control 0 0 bias_ena 0 enables the normal bias current generator (for all analogue functions) 0 = disabled 1 = enabled register 04h bias control 0 register address bit label default description refer to r5 (05h) vmid control 0 2:1 vmid_res [1:0] 00 vmid divider enable and select 00 = vmid disabled (for off mode) 01 = 2 x 50k divider (for normal operation) 10 = 2 x 250k divider (for low power standby) 11 = 2 x 5k divider (for fast start-up) 0 vmid_ena 0 enable vmid master bias current source 0 = disabled 1 = enabled register 05h vmid control 0 register address bit label default description refer to r14 (0eh) power managemen t 2 1 hpl_pga_en a 0 left headphone output enable 0 = disabled 1 = enabled 0 hpr_pga_en a 0 right headphone output enable 0 = disabled 1 = enabled register 0eh power management 2 register address bit label default description refer to r15 (0fh) power managemen t 3 1 lineoutl_pg a_ena 0 left line output enable 0 = disabled 1 = enabled 0 lineoutr_p ga_ena 0 right line output enable 0 = disabled 1 = enabled register 0fh power management 3
wm8912 production data w pd, rev 4.0, september 2010 98 register address bit label default description refer to r18 (12h) power managemen t 6 3 dacl_ena 0 left dac enable 0 = dac disabled 1 = dac enabled 2 dacr_ena 0 right dac enable 0 = dac disabled 1 = dac enabled register 12h power management 6 register address bit label default description refer to r20 (14h) clock rates 0 14 toclk_rate _div16 0 toclk rate divider (/16) 0 = f / 1 1 = f / 16 13 toclk_rate _x4 0 toclk rate multiplier 0 = f x 1 1 = f x 4 0 mclk_div 0 enables divide by 2 on mclk 0 = sysclk = mclk 1 = sysclk = mclk / 2 register 14h clock rates 0 register address bit label default description refer to r21 (15h) clock rates 1 13:10 clk_sys_ra te [3:0] 0011 selects the sysclk / fs ratio 0000 = 64 0001 = 128 0010 = 192 0011 = 256 0100 = 384 0101 = 512 0110 = 768 0111 = 1024 1000 = 1408 1001 = 1536 2:0 sample_rat e [2:0] 101 selects the sample rate (fs) 000 = 8khz 001 = 11.025khz, 12khz 010 = 16khz 011 = 22.05khz, 24khz 100 = 32khz 101 = 44.1khz, 48khz 110 to 111 = reserved register 15h clock rates 1
production data wm8912 w pd, rev 4.0, september 2010 99 register address bit label default description refer to r22 (16h) clock rates 2 15 mclk_inv 0 mclk invert 0 = mclk not inverted 1 = mclk inverted 14 sysclk_src 0 sysclk source select 0 = mclk 1 = fll output 12 toclk_rate 0 toclk rate divider (/2) 0 = f / 2 1 = f / 1 3 opclk_ena 0 gpio clock output enable 0 = disabled 1 = enabled 2 clk_sys_en a 0 system clock enable 0 = disabled 1 = enabled 1 clk_dsp_en a 0 dsp clock enable 0 = disabled 1 = enabled 0 toclk_ena 0 zero cross timeout enable 0 = disabled 1 = enabled register 16h clock rates 2 register address bit label default description refer to r24 (18h) audio interface 0 12 dacl_datinv 0 left dac invert 0 = left dac output not inverted 1 = left dac output inverted 11 dacr_datin v 0 right dac invert 0 = right dac output not inverted 1 = right dac output inverted 10:9 dac_boost [1:0] 00 dac digital input volume boost 00 = 0db 01 = +6db (input data must not exceed -6dbfs) 10 = +12db (input data must not exceed -12dbfs) 11 = +18db (input data must not exceed -18dbfs) 5 aifdacl_src 0 left dac data source select 0 = left dac outputs left channel data 1 = left dac outputs right channel data 4 aifdacr_sr c 1 right dac data source select 0 = right dac outputs left channel data 1 = right dac outputs right channel data 1 dac_comp 0 dac companding enable 0 = disabled 1 = enabled 0 dac_compm ode 0 dac companding type 0 = -law 1 = a-law register 18h audio interface 0
wm8912 production data w pd, rev 4.0, september 2010 100 register address bit label default description refer to r25 (19h) audio interface 1 13 aifdac_tdm 0 dac tdm enable 0 = normal dacdat operation 1 = tdm enabled on dacdat 12 aifdac_tdm_ chan 0 dacdat tdm channel select 0 = dacdat data input on slot 0 1 = dacdat data input on slot 1 8 aif_tris 0 audio interface tristate 0 = audio interface pins operate normally 1 = tristate all audio interface pins 7 aif_bclk_inv 0 bclk invert 0 = bclk not inverted 1 = bclk inverted 6 bclk_dir 0 audio interface bclk direction 0 = bclk is input 1 = bclk is output 4 aif_lrclk_in v 0 lrc polarity / dsp mode a-b select. right, left and i2s modes ? lrc polarity 0 = not inverted 1 = inverted dsp mode ? mode a-b select 0 = msb is available on 2nd bclk rising edge after lrc rising edge (mode a) 1 = msb is available on 1st bclk rising edge after lrc rising edge (mode b) 3:2 aif_wl [1:0] 10 digital audio interface word length 00 = 16 bits 01 = 20 bits 10 = 24 bits 11 = 32 bits 1:0 aif_fmt [1:0] 10 digital audio interface format 00 = right justified 01 = left justified 10 = i2s 11 = dsp register 19h audio interface 1 register address bit label default description refer to r26 (1ah) audio interface 2 11:8 opclk_div [3:0] 0000 gpio output clock divider 0000 = sysclk 0001 = sysclk / 2 0010 = sysclk / 3 0011 = sysclk / 4 0100 = sysclk / 5.5 0101 = sysclk / 6 0110 = sysclk / 8 0111 = sysclk / 12 1000 = sysclk / 16 1001 to 1111 = reserved
production data wm8912 w pd, rev 4.0, september 2010 101 register address bit label default description refer to 4:0 bclk_div [4:0] 0_0100 bclk frequency (master mode) 00000 = sysclk 00001 = sysclk / 1.5 00010 = sysclk / 2 00011 = sysclk / 3 00100 = sysclk / 4 00101 = sysclk / 5 00110 = sysclk / 5.5 00111 = sysclk / 6 01000 = sysclk / 8 (default) 01001 = sysclk / 10 01010 = sysclk / 11 01011 = sysclk / 12 01100 = sysclk / 16 01101 = sysclk / 20 01110 = sysclk / 22 01111 = sysclk / 24 10000 = sysclk / 25 10001 = sysclk / 30 10010 = sysclk / 32 10011 = sysclk / 44 10100 = sysclk / 48 register 1ah audio interface 2 register address bit label default description refer to r27 (1bh) audio interface 3 11 lrclk_dir 0 audio interface lrc direction 0 = lrc is input 1 = lrc is output 10:0 lrclk_rate [10:0] 000_0100_ 0000 lrc rate (master mode) lrc clock output = bclk / lrclk_rate integer (lsb = 1) valid range: 8 to 2047 register 1bh audio interface 3 register address bit label default description refer to r30 (1eh) dac digital volume left 8 dac_vu 0 dac volume update writing a 1 to this bit causes left and right dac volume to be updated simultaneously 7:0 dacl_vol [7:0] 1100_0000 left dac digital volume 00h = mute 01h = -71.625db 02h = -71.250db ? (0.375db steps) c0h to ffh = 0db register 1eh dac digital volume left
wm8912 production data w pd, rev 4.0, september 2010 102 register address bit label default description refer to r31 (1fh) dac digital volume right 8 dac_vu 0 dac volume update writing a 1 to this bit causes left and right dac volume to be updated simultaneously 7:0 dacr_vol [7:0] 1100_0000 right dac digital volume 00h = mute 01h = -71.625db 02h = -71.250db ? (0.375db steps) c0h to ffh = 0db register 1fh dac digital volume right register address bit label default description refer to r33 (21h) dac digital 1 12 dac_mono 0 dac mono mix 0 = stereo 1 = mono (mono mix output on enabled dac) 11 dac_sb_filt 0 selects dac filter characteristics 0 = normal mode 1 = sloping stopband mode 10 dac_mutera te 0 dac soft mute ramp rate 0 = fast ramp (fs/2, maximum ramp time is 10.7ms at fs=48k) 1 = slow ramp (fs/32, maximum ramp time is 171ms at fs=48k) 9 dac_unmute _ramp 0 dac soft mute mode 0 = disabling soft-mute (dac_mute=0) will cause the dac volume to change immediately to dacl_vol and dacr_vol settings 1 = disabling soft-mute (dac_mute=0) will cause the dac volume to ramp up gradually to the dacl_vol and dacr_vol settings 6 dac_osr128 0 dac oversample rate select 0 = low power (normal osr) 1 = high performance (double osr) 3 dac_mute 1 dac soft mute control 0 = dac un-mute 1 = dac mute 2:1 deemph [1:0] 00 dac de-emphasis control 00 = no de-emphasis 01 = 32khz sample rate 10 = 44.1khz sample rate 11 = 48khz sample rate register 21h dac digital 1
production data wm8912 w pd, rev 4.0, september 2010 103 register address bit label default description refer to r40 (28h) drc 0 15 drc_ena 0 drc enable 1 = enabled 0 = disabled 14 drc_dac_pa th 0 drc path select 0 = reserved 1 = dac path 12:11 drc_gs_hys t_lvl [1:0] 00 gain smoothing hysteresis threshold 00 = low 01 = medium (recommended) 10 = high 11 = reserved 10:6 drc_startu p_gain [4:0] 0_0110 initial gain at drc startup 00000 = -3db 00001 = -2.5db 00010 = -2db 00011 = -1.5db 00100 = -1db 00101 = -0.5db 00110 = 0db (default) 00111 = 0.5db 01000 = 1db 01001 = 1.5db 01010 = 2db 01011 = 2.5db 01100 = 3db 01101 = 3.5db 01110 = 4db 01111 = 4.5db 10000 = 5db 10001 = 5.5db 10010 = 6db 10011 to 11111 = reserved 5 drc_ff_del ay 1 feed-forward delay for anti-clip feature 0 = 5 samples 1 = 9 samples time delay can be calculated as 5/fs or 9/ fs, where fs is the sample rate. 3 drc_gs_ena 1 gain smoothing enable 0 = disabled 1 = enabled 2 drc_qr 1 quick release enable 0 = disabled 1 = enabled 1 drc_anticli p 1 anti-clip enable 0 = disabled 1 = enabled 0 drc_gs_hys t 1 gain smoothing hysteresis enable 0 = disabled 1 = enabled register 28h drc 0
wm8912 production data w pd, rev 4.0, september 2010 104 register address bit label default description refer to r41 (29h) drc 1 15:12 drc_atk [3:0] 0011 gain attack rate (seconds/6db) 0000 = instantaneous 0001 = 363us 0010 = 726us 0011 = 1.45ms (default) 0100 = 2.9ms 0101 = 5.8ms 0110 = 11.6ms 0111 = 23.2ms 1000 = 46.4ms 1001 = 92.8ms 1010 = 185.6ms 1011-1111 = reserved 11:8 drc_dcy [3:0] 0010 gain decay rate (seconds/6db) 0000 = 186ms 0001 = 372ms 0010 = 743ms (default) 0011 = 1.49s 0100 = 2.97s 0101 = 5.94s 0110 = 11.89s 0111 = 23.78s 1000 = 47.56s 1001-1111 = reserved 7:6 drc_qr_thr [1:0] 01 quick release crest factor threshold 00 = 12db 01 = 18db (default) 10 = 24db 11 = 30db 5:4 drc_qr_dcy [1:0] 00 quick release decay rate (seconds/6db) 00 = 0.725ms (default) 01 = 1.45ms 10 = 5.8ms 11 = reserved 3:2 drc_mingain [1:0] 10 minimum gain the drc can use to attenuate audio signals 00 = 0db (default) 01 = -6db 10 = -12db 11 = -18db 1:0 drc_maxgai n [1:0] 00 maximum gain the drc can use to boost audio signals 00 = 12db 01 = 18db (default) 10 = 24db 11 = 36db register 29h drc 1
production data wm8912 w pd, rev 4.0, september 2010 105 register address bit label default description refer to r42 (2ah) drc 2 5:3 drc_hi_com p [2:0] 000 compressor slope r0 000 = 1 (no compression) 001 = 1/2 010 = 1/4 011 = 1/8 100 = 1/16 101 = 0 110 = reserved 111 = reserved 2:0 drc_lo_com p [2:0] 000 compressor slope r1 000 = 1 (no compression) 001 = 1/2 010 = 1/4 011 = 1/8 100 = 0 101 = reserved 11x = reserved register 2ah drc 2 register address bit label default description refer to r43 (2bh) drc 3 10:5 drc_knee_ip [5:0] 00_0000 compressor threshold t (db) 000000 = 0db 000001 = -0.75db 000010 = -1.5db ? (-0.75db steps) 111100 = -45db 111101 = reserved 11111x = reserved 4:0 drc_knee_o p [4:0] 0_0000 compressor amplitude at threshold yt (db) 00000 = 0db 00001 = -0.75db 00010 = -1.5db ? (-0.75db steps) 11110 = -22.5db 11111 = reserved register 2bh drc 3
wm8912 production data w pd, rev 4.0, september 2010 106 register address bit label default description refer to r57 (39h) analogue out1 left 8 hpoutl_mut e 0 left headphone output mute 0 = un-mute 1 = mute 7 hpout_vu 0 headphone output volume update writing a 1 to this bit will update hpoutl and hpoutr volumes simultaneously. 6 hpoutlzc 0 left headphone output zero cross enable 0 = disabled 1 = enabled 5:0 hpoutl_vol [5:0] 10_1101 left headphone output volume 000000 = -57db 000001 = -56db (? 1db steps) 111001 = 0db (? 1db steps) 111110 = +5db 111111 = +6db register 39h analogue out1 left register address bit label default description refer to r58 (3ah) analogue out1 right 8 hpoutr_mut e 0 right headphone output mute 0 = un-mute 1 = mute 7 hpout_vu 0 headphone output volume update writing a 1 to this bit will update hpoutl and hpoutr volumes simultaneously. 6 hpoutrzc 0 right headphone output zero cross enable 0 = disabled 1 = enabled 5:0 hpoutr_vol [5:0] 10_1101 right headphone output volume 000000 = -57db 000001 = -56db (? 1db steps) 111001 = 0db (? 1db steps) 111110 = +5db 111111 = +6db register 3ah analogue out1 right
production data wm8912 w pd, rev 4.0, september 2010 107 register address bit label default description refer to r59 (3bh) analogue out2 left 8 lineoutl_mu te 0 left line output mute 0 = un-mute 1 = mute 7 lineout_vu 0 line output volume update writing a 1 to this bit will update lineoutl and lineoutr volumes simultaneously. 6 lineoutlzc 0 left line output zero cross enable 0 = disabled 1 = enabled 5:0 lineoutl_vo l [5:0] 11_1001 left line output volume 000000 = -57db 000001 = -56db (? 1db steps) 111001 = 0db (? 1db steps) 111110 = +5db 111111 = +6db register 3bh analogue out2 left register address bit label default description refer to r60 (3ch) analogue out2 right 8 lineoutr_m ute 0 right line output mute 0 = un-mute 1 = mute 7 lineout_vu 0 line output volume update writing a 1 to this bit will update lineoutl and lineoutr volumes simultaneously. 6 lineoutrzc 0 right line output zero cross enable 0 = disabled 1 = enabled 5:0 lineoutr_v ol [5:0] 11_1001 right line output volume 000000 = -57db 000001 = -56db (? 1db steps) 111001 = 0db (? 1db steps) 111110 = +5db 111111 = +6db register 3ch analogue out2 right
wm8912 production data w pd, rev 4.0, september 2010 108 register address bit label default description refer to r67 (43h) dc servo 0 3 dcs_ena_ch an_3 0 dc servo enable for lineoutr 0 = disabled 1 = enabled 2 dcs_ena_ch an_2 0 dc servo enable for lineoutl 0 = disabled 1 = enabled 1 dcs_ena_ch an_1 0 dc servo enable for hpoutr 0 = disabled 1 = enabled 0 dcs_ena_ch an_0 0 dc servo enable for hpoutl 0 = disabled 1 = enabled register 43h dc servo 0 register address bit label default description refer to r68 (44h) dc servo 1 15 dcs_trig_si ngle_3 0 writing 1 to this bit selects a single dc offset correction for lineoutr. in readback, a value of 1 indicates that the dc servo single correction is in progress. 14 dcs_trig_si ngle_2 0 writing 1 to this bit selects a single dc offset correction for lineoutl. in readback, a value of 1 indicates that the dc servo single correction is in progress. 13 dcs_trig_si ngle_1 0 writing 1 to this bit selects a single dc offset correction for hpoutr. in readback, a value of 1 indicates that the dc servo single correction is in progress. 12 dcs_trig_si ngle_0 0 writing 1 to this bit selects a single dc offset correction for hpoutl. in readback, a value of 1 indicates that the dc servo single correction is in progress. 11 dcs_trig_se ries_3 0 writing 1 to this bit selects a series of dc offset corrections for lineoutr. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 10 dcs_trig_se ries_2 0 writing 1 to this bit selects a series of dc offset corrections for lineoutl. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 9 dcs_trig_se ries_1 0 writing 1 to this bit selects a series of dc offset corrections for hpoutr. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 8 dcs_trig_se ries_0 0 writing 1 to this bit selects a series of dc offset corrections for hpoutl. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 7 dcs_trig_st artup_3 0 writing 1 to this bit selects start-up dc servo mode for lineoutr. in readback, a value of 1 indicates that the dc servo start-up correction is in progress. 6 dcs_trig_st artup_2 0 writing 1 to this bit selects start-up dc servo mode for lineoutl. in readback, a value of 1 indicates that the dc servo
production data wm8912 w pd, rev 4.0, september 2010 109 register address bit label default description refer to start-up correction is in progress. 5 dcs_trig_st artup_1 0 writing 1 to this bit selects start-up dc servo mode for hpoutr. in readback, a value of 1 indicates that the dc servo start-up correction is in progress. 4 dcs_trig_st artup_0 0 writing 1 to this bit selects start-up dc servo mode for hpoutl. in readback, a value of 1 indicates that the dc servo start-up correction is in progress. 3 dcs_trig_da c_wr_3 0 writing 1 to this bit selects dac write dc servo mode for lineoutr. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 2 dcs_trig_da c_wr_2 0 writing 1 to this bit selects dac write dc servo mode for lineoutl. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 1 dcs_trig_da c_wr_1 0 writing 1 to this bit selects dac write dc servo mode for hpoutr. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. 0 dcs_trig_da c_wr_0 0 writing 1 to this bit selects dac write dc servo mode for hpoutl. in readback, a value of 1 indicates that the dc servo dac write correction is in progress. register 44h dc servo 1 register address bit label default description refer to r69 (45h) dc servo 2 11:8 dcs_timer_p eriod_23 [3:0] 1010 time between periodic updates for lineoutl/lineoutr. time is calculated as 0.256s x (2^period) 0000 = off 0001 = 0.52s 1010 = 266s (4min 26s) 1111 = 8519s (2hr 22s) 3:0 dcs_timer_p eriod_01 [3:0] 1010 time between periodic updates for hpoutl/hpoutr. time is calculated as 0.256s x (2^period) 0000 = off 0001 = 0.52s 1010 = 266s (4min 26s) 1111 = 8519s (2hr 22s) register 45h dc servo 2 register address bit label default description refer to r71 (47h) dc servo 4 6:0 dcs_series_ no_23 [6:0] 010_1010 number of dc servo updates to perform in a series event for lineoutl/lineoutr. 0 = 1 updates 1 = 2 updates ... 127 = 128 updates register 47h dc servo 4
wm8912 production data w pd, rev 4.0, september 2010 110 register address bit label default description refer to r72 (48h) dc servo 5 6:0 dcs_series_ no_01 [6:0] 010_1010 number of dc servo updates to perform in a series event for hpoutl/hpoutr. 0 = 1 updates 1 = 2 updates ... 127 = 128 updates register 48h dc servo 5 register address bit label default description refer to r73 (49h) dc servo 6 7:0 dcs_dac_w r_val_3 [7:0] 0000_0000 dc offset value for lineoutr in dac write dc servo mode in two's complement format. in readback, the current dc offset value is returned in two's complement format. two?s complement format: lsb is 0.25mv. range is +/-32mv register 49h dc servo 6 register address bit label default description refer to r74 (4ah) dc servo 7 7:0 dcs_dac_w r_val_2 [7:0] 0000_0000 dc offset value for lineoutl in dac write dc servo mode in two's complement format. in readback, the current dc offset value is returned in two's complement format. two?s complement format: lsb is 0.25mv. range is +/-32mv register 4ah dc servo 7 register address bit label default description refer to r75 (4bh) dc servo 8 7:0 dcs_dac_w r_val_1 [7:0] 0000_0000 dc offset value for hpoutr in dac write dc servo mode in two's complement format. in readback, the current dc offset value is returned in two's complement format. two?s complement format: lsb is 0.25mv. range is +/-32mv register 4bh dc servo 8
production data wm8912 w pd, rev 4.0, september 2010 111 register address bit label default description refer to r76 (4ch) dc servo 9 7:0 dcs_dac_w r_val_0 [7:0] 0000_0000 dc offset value for hpoutl in dac write dc servo mode in two's complement format. in readback, the current dc offset value is returned in two's complement format. two?s complement format: lsb is 0.25mv. range is +/-32mv register 4ch dc servo 9 register address bit label default description refer to r77 (4dh) dc servo readback 0 11:8 dcs_cal_co mplete [3:0] 0000 dc servo complete status [3] - lineoutr [2] - lineoutl [1] - hpoutr [0] - hpoutl 0 = dac write or start-up dc servo mode not completed. 1 = dac write or start-up dc servo mode complete. 7:4 dcs_dac_w r_complete [3:0] 0000 dc servo dac write status [3] - lineoutr [2] - lineoutl [1] - hpoutr [0] - hpoutl 0 = dac write dc servo mode not completed. 1 = dac write dc servo mode complete. 3:0 dcs_startu p_complete [3:0] 0000 dc servo start-up status [3] - lineoutr [2] - lineoutl [1] - hpoutr [0] - hpoutl 0 = start-up dc servo mode not completed.. 1 = start-up dc servo mode complete. register 4dh dc servo readback 0
wm8912 production data w pd, rev 4.0, september 2010 112 register address bit label default description refer to r90 (5ah) analogue hp 0 7 hpl_rmv_sh ort 0 removes hpoutl short 0 = hpoutl short enabled 1 = hpoutl short removed for normal operation, this bit should be set as the final step of the hpl enable sequence. 6 hpl_ena_ou tp 0 enables hpoutl output stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the dc offset cancellation has been scheduled. 5 hpl_ena_dl y 0 enables hpoutl intermediate stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the output signal path has been configured, and before dc offset cancellation is scheduled. this bit should be set with at least 20us delay after hpl_ena. 4 hpl_ena 0 enables hpoutl input stage 0 = disabled 1 = enabled for normal operation, this bit should be set as the first step of the hpl enable sequence. 3 hpr_rmv_sh ort 0 removes hpoutr short 0 = hpoutr short enabled 1 = hpoutr short removed for normal operation, this bit should be set as the final step of the hpr enable sequence. 2 hpr_ena_ou tp 0 enables hpoutr output stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the dc offset cancellation has been scheduled. 1 hpr_ena_dl y 0 enables hpoutr intermediate stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the output signal path has been configured, and before dc offset cancellation is scheduled. this bit should be set with at least 20us delay after hpr_ena. 0 hpr_ena 0 enables hpoutr input stage 0 = disabled 1 = enabled for normal operation, this bit should be set as the first step of the hpr enable sequence. register 5ah analogue hp 0
production data wm8912 w pd, rev 4.0, september 2010 113 register address bit label default description refer to r94 (5eh) analogue lineout 0 7 lineoutl_rm v_short 0 removes lineoutl short 0 = lineoutl short enabled 1 = lineoutl short removed for normal operation, this bit should be set as the final step of the lineoutl enable sequence. 6 lineoutl_en a_outp 0 enables lineoutl output stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the dc offset cancellation has been scheduled. 5 lineoutl_en a_dly 0 enables lineoutl intermediate stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the output signal path has been configured, and before dc offset cancellation is scheduled. this bit should be set with at least 20us delay after lineoutl_ena. 4 lineoutl_en a 0 enables lineoutl input stage 0 = disabled 1 = enabled for normal operation, this bit should be set as the first step of the lineoutl enable sequence. 3 lineoutr_r mv_short 0 removes lineoutr short 0 = lineoutr short enabled 1 = lineoutr short removed for normal operation, this bit should be set as the final step of the lineoutr enable sequence. 2 lineoutr_en a_outp 0 enables lineoutr output stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the dc offset cancellation has been scheduled. 1 lineoutr_en a_dly 0 enables lineoutr intermediate stage 0 = disabled 1 = enabled for normal operation, this bit should be set to 1 after the output signal path has been configured, and before dc offset cancellation is scheduled. this bit should be set with at least 20us delay after lineoutr_ena. 0 lineoutr_en a 0 enables lineoutr input stage 0 = disabled 1 = enabled for normal operation, this bit should be set as the first step of the lineoutr enable sequence. register 5eh analogue lineout 0 register address bit label default description refer to r98 (62h) charge pump 0 0 cp_ena 0 enable charge-pump digits 0 = disable 1 = enable register 62h charge pump 0
wm8912 production data w pd, rev 4.0, september 2010 114 register address bit label default description refer to r104 (68h) class w 0 0 cp_dyn_pwr 0 enable dynamic charge pump power control 0 = charge pump controlled by volume register settings (class g) 1 = charge pump controlled by real-time audio level (class w) class w is recommended for lowest power consumption. register 68h class w 0 register address bit label default description refer to r108 (6ch) write sequencer 0 8 wseq_ena 0 write sequencer enable. 0 = disabled 1 = enabled 4:0 wseq_write _index [4:0] 0_0000 sequence write index. this is the memory location to which any updates to r109 and r110 will be copied. 0 to 31 = ram addresses register 6ch write sequencer 0 register address bit label default description refer to r109 (6dh) write sequencer 1 14:12 wseq_data_ width [2:0] 000 width of the data block written in this sequence step. 000 = 1 bit 001 = 2 bits 010 = 3 bits 011 = 4 bits 100 = 5 bits 101 = 6 bits 110 = 7 bits 111 = 8 bits 11:8 wseq_data_ start [3:0] 0000 bit position of the lsb of the data block written in this sequence step. 0000 = bit 0 ? 1111 = bit 15 7:0 wseq_addr [7:0] 0000_0000 control register address to be written to in this sequence step. register 6dh write sequencer 1
production data wm8912 w pd, rev 4.0, september 2010 115 register address bit label default description refer to r110 (6eh) write sequencer 2 14 wseq_eos 0 end of sequence flag. this bit indicates whether the control write sequencer should stop after executing this step. 0 = not end of sequence 1 = end of sequence (stop the sequencer after this step). 11:8 wseq_delay [3:0] 0000 time delay after executing this step. total delay time per step (including execution)= 62.5 s (2^wseq_delay + 8) 7:0 wseq_data [7:0] 0000_0000 data to be written in this sequence step. when the data width is less than 8 bits, then one or more of the msbs of wseq_data are ignored. it is recommended that unused bits be set to 0. register 6eh write sequencer 2 register address bit label default description refer to r111 (6fh) write sequencer 3 9 wseq_abor t 0 writing a 1 to this bit aborts the current sequence and returns control of the device back to the serial control interface. 8 wseq_start 0 writing a 1 to this bit starts the write sequencer at the memory location indicated by the wseq_start_index field. the sequence continues until it reaches an ?end of sequence? flag. at the end of the sequence, this bit will be reset by the write sequencer. 5:0 wseq_start _index [5:0] 00_0000 sequence start index. this is the memory location of the first command in the selected sequence. 0 to 31 = ram addresses 32 to 48 = rom addresses 49 to 63 = reserved register 6fh write sequencer 3 register address bit label default description refer to r112 (70h) write sequencer 4 9:4 wseq_curr ent_index [5:0] 00_0000 sequence current index (read only): this is the location of the most recently accessed command in the write sequencer memory. 0 wseq_busy 0 sequencer busy flag (read only): 0 = sequencer idle 1 = sequencer busy note: it is not possible to write to control registers via the control interface while the sequencer is busy. register 70h write sequencer 4
wm8912 production data w pd, rev 4.0, september 2010 116 register address bit label default description refer to r116 (74h) fll control 1 2 fll_fracn_e na 0 fll fractional enable 0 = integer mode 1 = fractional mode fractional mode (fll_fracn_ena=1) is recommended in all cases 1 fll_osc_en a 0 fll oscillator enable 0 = disabled 1 = enabled fll_osc_ena must be enabled before enabling fll_ena. note that this field is required for free-running fll modes only. 0 fll_ena 0 fll enable 0 = disabled 1 = enabled fll_osc_ena must be enabled before enabling fll_ena. register 74h fll control 1 register address bit label default description refer to r117 (75h) fll control 2 13:8 fll_outdiv [5:0] 00_0000 fll fout clock divider 00_0000 = reserved 00_0001 = reserved 00_0010 = reserved 00_0011 = 4 00_0100 = 5 00_0101 = 6 ? 11_1110 = 63 11_1111 = 64 (fout = fvco / fll_outdiv) 6:4 fll_ctrl_ra te [2:0] 000 frequency of the fll control block 000 = fvco / 1 (recommended value) 001 = fvco / 2 010 = fvco / 3 011 = fvco / 4 100 = fvco / 5 101 = fvco / 6 110 = fvco / 7 111 = fvco / 8 recommended that these are not changed from default. 2:0 fll_fratio [2:0] 111 fvco clock divider 000 = 1 001 = 2 010 = 4 011 = 8
production data wm8912 w pd, rev 4.0, september 2010 117 register address bit label default description refer to 1xx = 16 000 recommended for high fref 011 recommended for low fref register 75h fll control 2 register address bit label default description refer to r118 (76h) fll control 3 15:0 fll_k [15:0] 0000_0000 _0000_000 0 fractional multiply for fref (msb = 0.5) register 76h fll control 3 register address bit label default description refer to r119 (77h) fll control 4 14:5 fll_n [9:0] 01_0111_0 111 integer multiply for fref (lsb = 1) 3:0 fll_gain [3:0] 0000 fll gain applied to error 0000 = x 1 (recommended value) 0001 = x 2 0010 = x 4 0011 = x 8 0100 = x 16 0101 = x 32 0110 = x 64 0111 = x 128 1000 = x 256 recommended that these are not changed from default. register 77h fll control 4 register address bit label default description refer to r120 (78h) fll control 5 4:3 fll_clk_ref _div [1:0] 00 fll clock reference divider 00 = mclk / 1 01 = mclk / 2 10 = mclk / 4 11 = mclk / 8 mclk (or other input reference) must be divided down to <=13.5mhz. for lower power operation, the reference clock can be divided down further if desired. 1:0 fll_clk_ref _src [1:0] 00 fll clock source 00 = mclk 01 = bclk 10 = lrclk 11 = reserved register 78h fll control 5
wm8912 production data w pd, rev 4.0, september 2010 118 register address bit label default description refer to r121 (79h) gpio control 1 5 gpio1_pu 0 gpio1 pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 4 gpio1_pd 1 gpio1 pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled 3:0 gpio1_sel [3:0] 0100 gpio1 function select 0000 = gpio input 0001 = clock output (f= sysclk/opclkdiv) 0010 = logic '0' output 0011 = logic '1' output 0100 = irq output (default) 0101 = fll lock output 0110 = reserved 0111 = reserved 1000 = reserved 1001 = fll clock output 1010 to 1111 = reserved register 79h gpio control 1 register address bit label default description refer to r124 (7ch) gpio control 4 7 gpio_bclk_ mode_ena 0 selects bclk/gpio4 pin function 0 = bclk/gpio4 is used as bclk 1 = bclk/gpio4 is used as gpio. mclk provides the bclk in the aif in this mode. 3:0 gpio_bclk_s el [3:0] 0000 bclk/gpio4 function select: 0000 = gpio input (default) 0001 = clock output (f= sysclk/opclkdiv) 0010 = logic '0' output 0011 = logic '1' output 0100 = irq output 0101 = fll lock output 0110 = reserved 0111 = reserved 1000 = reserved 1001 = fll clock output 1010 to 1111 = reserved register 7ch gpio control 4
production data wm8912 w pd, rev 4.0, september 2010 119 register address bit label default description refer to r126 (7eh) digital pulls 7 mclk_pu 0 mclk pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 6 mclk_pd 0 mclk pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled 5 dacdat_pu 0 dacdat pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 4 dacdat_pd 0 dacdat pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled 3 lrclk_pu 0 lrclk pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 2 lrclk_pd 0 lrclk pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled 1 bclk_pu 0 bclk pull-up resistor enable 0 = pull-up disabled 1 = pull-up enabled 0 bclk_pd 0 bclk pull-down resistor enable 0 = pull-down disabled 1 = pull-down enabled register 7eh digital pulls register address bit label default description refer to r127 (7fh) interrupt status 10 irq 0 logical or of all other interrupt flags 9 gpio_bclk_e int 0 gpio4 interrupt 0 = interrupt not set 1 = interrupt is set cleared when a ?1? is written 8 wseq_eint 0 write sequence interrupt 0 = interrupt not set 1 = interrupt is set cleared when a ?1? is written. note that the read value of wseq_eint is not valid whilst the write sequencer is busy 5 gpio1_eint 0 gpio1 interrupt 0 = interrupt not set 1 = interrupt is set cleared when a ?1? is written 2 fll_lock_ei nt 0 fll lock interrupt 0 = interrupt not set 1 = interrupt is set cleared when a ?1? is written register 7fh interrupt status
wm8912 production data w pd, rev 4.0, september 2010 120 register address bit label default description refer to r128 (80h) interrupt status mask 9 im_gpio_bcl k_eint 1 gpio4 interrupt mask 0 = do not mask interrupt 1 = mask interrupt 8 im_wseq_ein t 1 write sequencer interrupt mask 0 = do not mask interrupt 1 = mask interrupt 5 im_gpio1_ein t 1 gpio1 interrupt mask 0 = do not mask interrupt 1 = mask interrupt 2 im_fll_lock _eint 1 fll lock interrupt mask 0 = do not mask interrupt 1 = mask interrupt register 80h interrupt status mask register address bit label default description refer to r129 (81h) interrupt polarity 9 gpio_bclk_e int_pol 0 gpio4 interrupt polarity 0 = active high 1 = active low 8 wseq_eint_ pol 0 write sequencer interrupt polarity 0 = active high (interrupt is triggered when wseq is busy) 1 = active low (interrupt is triggered when wseq is idle) 5 gpio1_eint_ pol 0 gpio1 interrupt polarity 0 = active high 1 = active low 2 fll_lock_ei nt_pol 0 fll lock interrupt polarity 0 = active high (interrupt is triggered when fll lock is reached) 1 = active low (interrupt is triggered when fll is not locked) register 81h interrupt polarity register address bit label default description refer to r130 (82h) interrupt debounce 9 gpio_bclk_e int_db 0 gpio4 interrupt debounce 0 = disabled 1 = enabled 8 wseq_eint_ db 0 write sequencer interrupt debounce enable 0 = disabled 1 = enabled 5 gpio1_eint_ db 0 gpio1 input debounce 0 = disabled 1 = enabled 2 fll_lock_ei nt_db 0 fll lock debounce 0 = disabled 1 = enabled register 82h interrupt debounce
production data wm8912 w pd, rev 4.0, september 2010 121 register address bit label default description refer to r134 (86h) eq1 0 eq_ena 0 eq enable 0 = eq disabled 1 = eq enabled register 86h eq1 register address bit label default description refer to r135 (87h) eq2 4:0 eq_b1_gain [4:0] 0_1100 gain for eq band 1 00000 = -12db 00001 = -11db (? 1db steps) 01100 = 0db (? 1db steps) 11000 = +12db 11001 to 11111 = reserved register 87h eq2 register address bit label default description refer to r136 (88h) eq3 4:0 eq_b2_gain [4:0] 0_1100 gain for eq band 2 00000 = -12db 00001 = -11db (? 1db steps) 01100 = 0db (? 1db steps) 11000 = +12db 11001 to 11111 = reserved register 88h eq3 register address bit label default description refer to r137 (89h) eq4 4:0 eq_b3_gain [4:0] 0_1100 gain for eq band 3 00000 = -12db 00001 = -11db (? 1db steps) 01100 = 0db (? 1db steps) 11000 = +12db 11001 to 11111 = reserved register 89h eq4
wm8912 production data w pd, rev 4.0, september 2010 122 register address bit label default description refer to r138 (8ah) eq5 4:0 eq_b4_gain [4:0] 0_1100 gain for eq band 4 00000 = -12db 00001 = -11db (? 1db steps) 01100 = 0db (? 1db steps) 11000 = +12db 11001 to 11111 = reserved register 8ah eq5 register address bit label default description refer to r139 (8bh) eq6 4:0 eq_b5_gain [4:0] 0_1100 gain for eq band5 00000 = -12db 00001 = -11db (? 1db steps) 01100 = 0db (? 1db steps) 11000 = +12db 11001 to 11111 = reserved register 8bh eq6 register address bit label default description refer to r140 (8ch) eq7 15:0 eq_b1_a [15:0] 0000_1111 _1100_101 0 eq band 1 coefficient a register 8ch eq7 register address bit label default description refer to r141 (8dh) eq8 15:0 eq_b1_b [15:0] 0000_0100 _0000_000 0 eq band 1 coefficient b register 8dh eq8 register address bit label default description refer to r142 (8eh) eq9 15:0 eq_b1_pg [15:0] 0000_0000 _1101_100 0 eq band 1 coefficient pg register 8eh eq9
production data wm8912 w pd, rev 4.0, september 2010 123 register address bit label default description refer to r143 (8fh) eq10 15:0 eq_b2_a [15:0] 0001_1110 _1011_010 1 eq band 2 coefficient a register 8fh eq10 register address bit label default description refer to r144 (90h) eq11 15:0 eq_b2_b [15:0] 1111_0001 _0100_010 1 eq band 2 coefficient b register 90h eq11 register address bit label default description refer to r145 (91h) eq12 15:0 eq_b2_c [15:0] 0000_1011 _0111_010 1 eq band 2 coefficient c register 91h eq12 register address bit label default description refer to r146 (92h) eq13 15:0 eq_b2_pg [15:0] 0000_0001 _1100_010 1 eq band 2 coefficient pg register 92h eq13 register address bit label default description refer to r147 (93h) eq14 15:0 eq_b3_a [15:0] 0001_1100 _0101_100 0 eq band 3 coefficient a register 93h eq14 register address bit label default description refer to r148 (94h) eq15 15:0 eq_b3_b [15:0] 1111_0011 _0111_001 1 eq band 3 coefficient b register 94h eq15 register address bit label default description refer to r149 (95h) eq16 15:0 eq_b3_c [15:0] 0000_1010 _0101_010 0 eq band 3 coefficient c register 95h eq16
wm8912 production data w pd, rev 4.0, september 2010 124 register address bit label default description refer to r150 (96h) eq17 15:0 eq_b3_pg [15:0] 0000_0101 _0101_100 0 eq band 3 coefficient pg register 96h eq17 register address bit label default description refer to r151 (97h) eq18 15:0 eq_b4_a [15:0] 0001_0110 _1000_111 0 eq band 4 coefficient a register 97h eq18 register address bit label default description refer to r152 (98h) eq19 15:0 eq_b4_b [15:0] 1111_1000 _0010_100 1 eq band 4 coefficient b register 98h eq19 register address bit label default description refer to r153 (99h) eq20 15:0 eq_b4_c [15:0] 0000_0111 _1010_110 1 eq band 4 coefficient c register 99h eq20 register address bit label default description refer to r154 (9ah) eq21 15:0 eq_b4_pg [15:0] 0001_0001 _0000_001 1 eq band 4 coefficient pg register 9ah eq21 register address bit label default description refer to r155 (9bh) eq22 15:0 eq_b5_a [15:0] 0000_0101 _0110_010 0 eq band 5 coefficient a register 9bh eq22 register address bit label default description refer to r156 (9ch) eq23 15:0 eq_b5_b [15:0] 0000_0101 _0101_100 1 eq band 1 coefficient b register 9ch eq23
production data wm8912 w pd, rev 4.0, september 2010 125 register address bit label default description refer to r157 (9dh) eq24 15:0 eq_b5_pg [15:0] 0100_0000 _0000_000 0 eq band 5 coefficient pg register 9dh eq24 register address bit label default description refer to r247 (f7h) fll nco test 0 0 fll_frc_nc o 0 fll forced control select 0 = normal 1 = fll oscillator controlled by fll_frc_nco_val (note that this field is required for free-running fll modes only) register f7h fll nco test 0 register address bit label default description refer to r248 (f8h) fll nco test 1 5:0 fll_frc_nc o_val [5:0] 01_1001 fll forced oscillator value valid range is 000000 to 111111 0x19h (011001) = 12mhz approx (note that this field is required for free-running fll modes only) register f8h fll nco test 1
wm8912 production data w pd, rev 4.0, september 2010 126 applications information recommended external components figure 56 recommended external components notes: 1. decoupling capacitors x5r ceramic capacitor is recommended for capacitors c7, c8, c13, c14, c15, c17 and c18. the positioning of c7 is very important - this should be as close to the wm8912 as possible. capacitors c17 and c18 should also be positioned as close to the wm8912 as possible. 2. charge pump capacitors specific recommendations for c12, c17 and c18 are provided in table 66. note that two different recommendations are provided for c17 and c18; either of these components is suitable, depending upon size requirements and availability. the positioning of c12 is very important - this should be as close to the wm8912 as possible. it is important to select a suitable capacitor type for the charge pump. note that the capacitance may vary with dc voltage; care is required to ensure that required capacitance is achieved at the applicable operating voltage, as specified in table 66. the capacitor datasheet should be consulted for this information. component required capacitance value part number voltage type size c12 (cpca-cpcb) 1 f at 2vdc 2.2 f kemet c0402c225m9pac 6.3v x5r 0402 c17 (cpvoutp) c18 (cpvoutn) 2 f at 2vdc 2.2 f murata grm188r61a225ke34d 10v x5r 0603 4.7 f murata grm155r60j475m_eia 6.3v x5r 0402 table 66 charge pump capacitors 3. zobel networks the zobel network shown in figure 56 is required on hpoutl, hpoutr, lineoutl and lineoutr whenever that output is enabled. stability of these ground-referenced outputs across all process corners cannot be guaranteed without the zobel network components. (note that, if any ground-referenced output pin is not required, the zobel network components can be omitted from the output pin, and the pin can be left floating.) the zobel network requirement is detailed further in the applications note wan_0212 ?class w headphone impedance compensation?. zobel networks (c9, c10, c11, c16, r3, r4, r5, r6) should be positioned reasonably close to the wm8912.
production data wm8912 w pd, rev 4.0, september 2010 127 package dimensions fl: 32 pin qfn plastic package 4 x 4 x 0.75 mm body, 0.40 mm lead pitch e2 b b 15 a 8 9 e c 0.08 c ccc a a1 c a3 seating plane 1 l index area (d/2 x e/2) top view d c aaa 2 x c aaa 2 x e 1 17 24 25 32 d2 b c bbb m a 5 4 notes: 1. dimension b applies to metallized terminal and is measured between 0.15 mm and 0.25 mm from terminal tip. 2. all dimensions are in millimetres. 3. the terminal #1 identifier and terminal numbering convention shall conform to jedec 95-1 spp-002. 4. coplanarity applies to the exposed heat sink slug as well as the terminals. 5. refer to application note wan_0118 for further information regarding pcb footprints and qfn package soldering. 6. this drawing is subject to change without notice. a3 g b exposed lead detail 1 detail 1 exposed ground paddle 6 bottom view side view 16 dm067.a dimensions (mm) symbols min nom max note a a1 a3 0.70 0.75 0.8 0.05 0.035 0 0.203 ref b d d2 e e2 e l 0.25 0.15 4.00 bsc 2.75 2.7 2.65 0.40 bsc 0.35 0.40 0.45 1 2 2 4.00 bsc 2.75 2.7 2.65 0.10 aaa bbb ccc ref: 0.05 0.10 tolerances of form and position 0.2 0.5 g
wm8912 production data w pd, rev 4.0, september 2010 128 important notice wolfson microelectronics plc (?wolfson?) products and services are sold subject to wolfson?s terms and conditions of sale, delivery and payment supplied at the time of order acknowledgement. wolfson warrants performance of its products to the specifications in effect at the date of shipment. wolfson reserves the right to make changes to its products and specifications or to discontinue any product or service without notice. customers should therefore obtain the latest version of relevant information from wolfson to verify that the information is current. testing and other quality control techniques are utilised to the extent wolfson deems necessary to support its warranty. specific testing of all parameters of each device is not necessarily performed unless required by law or regulation. in order to minimise risks associated with customer applications, the customer must use adequate design and operating safeguards to minimise inherent or procedural hazards. wolfson is not liable for applications assistance or customer product design. the customer is solely responsible for its selection and use of wolfson products. wolfson is not liable for such selection or use nor for use of any circuitry other than circuitry entirely embodied in a wolfson product. wolfson?s products are not intended for use in life support systems, appliances, nuclear systems or systems where malfunction can reasonably be expected to result in personal injury, death or severe property or environmental damage. any use of products by the customer for such purposes is at the customer?s own risk. wolfson does not grant any licence (express or implied) under any patent right, copyright, mask work right or other intellectual property right of wolfson covering or relating to any combination, machine, or process in which its products or services might be or are used. any provision or publication of any third party?s products or services does not constitute wolfson?s approval, licence, warranty or endorsement thereof. any third party trade marks contained in this document belong to the respective third party owner. reproduction of information from wolfson datasheets is permissible only if reproduction is without alteration and is accompanied by all associated copyright, proprietary and other notices (including this notice) and conditions. wolfson is not liable for any unauthorised alteration of such information or for any reliance placed thereon. any representations made, warranties given, and/or liabilities accepted by any person which differ from those contained in this datasheet or in wolfson?s standard terms and conditions of sale, delivery and payment are made, given and/or accepted at that person?s own risk. wolfson is not liable for any such representations, warranties or liabilities or for any reliance placed thereon by any person. address wolfson microelectronics plc westfield house 26 westfield road edinburgh eh11 2qb tel :: +44 (0)131 272 7000 fax :: +44 (0)131 272 7001 email :: sales@wolfsonmicro.com

▲Up To Search▲

Price & Availability of WM8912GEFLV

	To Download WM8912GEFLV Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .