View WM8976GEFLRV_5188340.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

w wm8976 low power codec with speaker driver wolfson microelectronics plc www.wolfsonmicro.com preliminary technical data, february 2005, rev 2.0 copyright ? 2005 wolfson microelectronics plc . description the wm8976 is a low power, high quality codec designed for portable applications such as digital still camera or digital camcorder. the device integrates a preamp for a differential mic, and includes drivers for speakers, headphone and differential or stereo line output. external component requirements are reduced as no separate microphone or headphone amplifiers are required. advanced on-chip digital signal processing includes a 5-band equaliser, a mixed signal automatic level control for the microphone or line input through the adc as well as a purely digital limiter function for record or playback. additional digital filtering options are available in the adc path, to cater for application filtering such as wind noise reduction. the wm8976 codec can operate as a master or a slave. an internal pll can generate all required audio clocks for the codec from common reference clock frequencies, such as 12mhz and 13mhz. the wm8976 operates at analogue supply voltages from 2.5v to 3.6v, although the digital core can operate at voltages down to 1.8v to save power. the speaker outputs and out3/4 line outputs can run from a 5v supply if increased output power is required. individual sections of the chip can also be powered down under software control. features ? stereo codec: ? dac snr 98db, thd -84db (a weighted @ 48khz) ? adc snr 90db, thd -80db (a weighted @ 48khz) ? on-chip headphone driver with capless option - 40mw output power into 16 ? / 3.3v spkvdd ? 0.9w output power into 8 ? btl speaker / 5v spkvdd - capable of driving piezo speakers - stereo speaker drive configuration ? mic preamps: ? differential or single-ended microphone interface - programmable preamp gain - psuedo differential input with common mode rejection - programmable alc / noise gate in adc path ? low-noise bias supplied for electret microphones ? other features: ? enhanced 3-d function for improved stereo separation ? digital playback limiter ? 5-band equaliser (record or playback) ? programmable adc high pass filter (wind noise reduction) ? programmable adc notch filter ? aux inputs for stereo analog input signals or beep ? on-chip pll supporting 12, 13, 19.2mhz and other clocks ? low power, low voltage - 2.5v to 3.6v (digital core: 1.8v to 3.6v) - power consumption <30mw all-on with 2.5v supplies ? 5x5mm 32-pin qfn package applications ? stereo camcorder or dsc
wm8976 preliminary technical data w ptd rev 2.0 february 2005 2 table of contents description .......................................................................................................1 features.............................................................................................................1 applications .....................................................................................................1 pin configuration...........................................................................................3 ordering information ..................................................................................3 pin description ................................................................................................4 absolute maximum ratings.........................................................................5 recommended operating conditions .....................................................5 electrical characteristics ......................................................................6 terminology............................................................................................................ 8 speaker output thd versus power ......................................................10 power consumption ....................................................................................11 signal timing requirements .....................................................................12 system clock timing ........................................................................................... 12 audio interface timing C master mode ........................................................ 12 audio interface timing C slave mode............................................................ 13 control interface timing C 3-wire mode .................................................... 14 control interface timing C 2-wire mode .................................................... 15 internal power on reset circuit ..........................................................16 recommended power up/down sequence .................................................. 18 device description .......................................................................................22 introduction ......................................................................................................... 22 input signal path ................................................................................................. 24 analogue to digital converter (adc).......................................................... 29 input limiter / automatic level control (alc) .......................................... 32 output signal path ............................................................................................. 36 3d stereo enhancement .................................................................................... 42 analogue outputs............................................................................................... 43 digital audio interfaces................................................................................... 56 audio sample rates ............................................................................................. 63 master clock and phase locked loop (pll) ............................................... 63 general purpose input/output...................................................................... 65 output switching (jack detect)..................................................................... 66 control interface.............................................................................................. 68 resetting the chip .............................................................................................. 69 power supplies .................................................................................................... 69 power management ............................................................................................ 70 register map...................................................................................................72 digital filter characteristics ...............................................................74 terminology.......................................................................................................... 74 dac filter responses......................................................................................... 75 adc filter responses......................................................................................... 75 highpass filter..................................................................................................... 76 5-band equaliser .................................................................................................. 77 applications information .........................................................................81 recommended external components .......................................................... 81 important notice ..........................................................................................83 address: .................................................................................................................. 83
preliminary technical data wm8976 w ptd rev 2.0 february 2005 3 pin configuration ordering information order code temperature range package moisture sensitivity level peak soldering temperature wm8976gefl/v -25 c to +85 c 32-pin qfn (5 x 5 mm) (lead free) msl3 260 o c wm8976gefl/rv -25 c to +85 c 32-pin qfn (5 x 5 mm) (lead free, tape and reel) msl3 260 o c note: reel quantity = 3,500
wm8976 preliminary technical data w ptd rev 2.0 february 2005 4 pin description pin name type description 1 lip analogue input left mic pre-amp positive input 2 lin analogue input left mic pre-amp negative input 3 l2/gpio2 analogue input left channel line input/secondary mic pre-amp positive input/gpio pin 4 dnc do not connect leave this pin floating 5 dnc do not connect leave this pin floating 6 dnc do not connect leave this pin floating 7 lrc digital input / output dac and adc sample rate clock 8 bclk digital input / output digital audio port clock 9 adcdat digital output adc digital audio data output 10 dacdat digital input dac digital audio data input 11 mclk digital input master clock input 12 dgnd supply digital ground 13 dcvdd supply digital core logic supply 14 dbvdd supply digital buffer (i/o) supply 15 csb/gpio1 digital input / output 3-wire mpu chip select / general purpose input/output 1 16 sclk digital input 3-wire mpu clock input / 2-wire mpu clock input 17 sdin digital input / output 3-wire mpu data input / 2-wire mpu data input/acknowledge 18 mode digital input control interface selection 19 auxl analogue input left auxillary input 20 auxr analogue input right auxillary input 21 out4 analogue output buffered midrail headphone pseudo-ground, or right line output or mon o mix output 22 out3 analogue output buffered midrail headphone pseudo-ground, or left line output 23 rout2 analogue output second right output, or btl speaker driver positive output 24 spkgnd supply speaker ground (feeds speaker amp and out3/out4) 25 lout2 analogue output second left output, or btl speaker driver negative output 26 spkvdd supply speaker supply (feed speaker amp only) 27 vmid reference decoupling for adc and dac reference voltage 28 agnd supply analogue ground (feeds adc and dac) 29 rout1 analogue output headphone output right 30 lout1 analogue output headphone output left 31 avdd supply analogue supply (feeds adc and dac) 32 micbias analogue output microphone bias
preliminary technical data wm8976 w ptd rev 2.0 february 2005 5 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 dbvdd, dcvdd, avdd supply voltages -0.3v +3.63v spkvdd supply voltage -0.3v +7v voltage range digital inputs dgnd -0.3v dvdd +0.3v voltage range analogue inputs agnd -0.3v avdd +0.3v operating temperature range, t a -25 c +85 c storage temperature prior to soldering 30 c max / 85% rh max 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. recommended operating conditions parameter symbol test conditions min typ max unit digital supply range (core) dcvdd 1.62 3.6 v digital supply range (buffer) dbvdd 1.8 3.6 v analogue core supply range avdd 2.5 3.6 v analogue output supply range spkvdd 2.5 5.5 v ground dgnd,agnd, spkgnd 0 v
wm8976 preliminary technical data w ptd rev 2.0 february 2005 6 electrical characteristics test conditions dcvdd=1.8v, avdd=dbvdd=spkvdd = 3.3v, t a = +25 o c, 1khz signal, fs = 48khz, 24-bit audio data unless otherwise stated. parameter symbol test conditions min typ max unit microphone preamp inputs (lip, lin, rip, rin, l2, r2) full-scale input signal level C note this changes with avdd (note 1) v infs pgaboost = 0db inppgavol = 0db 1.0 0 vrms dbv mic pga equivalent input noise at 35.25db gain tbd uv r micin gain set to 35.25db 1.6 k ? r micin gain set to 0db 47 k ? r micin gain set to -12db 75 k ? r micip micp2inppga = 1 94 k ? input resistance r micip micp2inppga = 0 tbd k ? input capacitance c micin 10 pf recommended coupling cap c coup 220 pf mic programmable gain amplifier (pga) programmable gain -12 35.25 db programmable gain step size guaranteed monotonic 0.75 db mute attenuation tbd db selectable input gain boost (0/+20db) boost disabled 0 db gain boost on pga input boost enabled 20 db gain range from auxl or l2 input to boost/mixer -12 +6 db gain step size to boost/mixer 3 db auxilliary analogue inputs (auxl, auxr) full-scale input signal level (0db) C note this changes with avdd v infs avdd/3.3 0 vrms dbv r auxinlmin left input boost and mixer enabled, at max gain 4.3 k ? r auxinltyp left input boost and mixer enabled, at 0db gain 8.6 k ? r auxinlmax left input boost and mixer enabled, at min gain 39.1 k ? r auxinrmin r ight input boost, mixer and beep enabled, at max gain 3 k ? r auxinrtyp r ight input boost, mixer and beep enabled, at 0db gain 6 k ? input resistance (note 2) r auxinrmax r ight input boost, mixer and beep enabled, at min gain 29 k ? input capacitance c micin 10 pf
preliminary technical data wm8976 w ptd rev 2.0 february 2005 7 test conditions dcvdd=1.8v, avdd=dbvdd=spkvdd = 3.3v, t a = +25 o c, 1khz signal, fs = 48khz, 24-bit audio data unless otherwise stated. parameter symbol test conditions min typ max unit automatic level control (alc) target record level -28.5 -6 db programmable gain -12 35.25 gain hold time (note 3,5) t hold mclk = 12.288mhz (note 3) 0, 2.67, 5.33, 10.67, , 43691 (time doubles with each step) ms alcmode=0 (alc), mclk=12.288mhz (note 3) 3.3, 6.6, 13.1, , 3360 (time doubles with each step) gain ramp-up (decay) time (note 4,5) t dcy alcmode=1 (limiter), mclk=12.288mhz (note 3) 0.73, 1.45, 2.91, , 744 (time doubles with each step) ms alcmode=0 (alc), mclk=12.288mhz (note 3) 0.83, 1.66, 3.33, , 852 (time doubles with each step) gain ramp-down (attack) time (note 4,5) t atk alcmode=1 (limiter), mclk=12.288mhz (note 3) 0.18, 0.36, 0.73, , 186 (time doubles with each step) ms mute attenuation tbd db analogue to digital converter (adc) signal to noise ratio (note 6,7) a-weighted, 0db gain 90 db total harmonic distortion (note 8) full-scale, 0db gain -80 db digital to analogue converter (dac) to line-out (lout1, rout1 with 10k ? ? ? ? / 50pf load) pga gains set to 0db, out34boost=0 avdd/3.3 full-scale output pga gains set to 0db, out34boost=1 1.5x (avdd/3.3) vrms signal to noise ratio (note 6,7) snr a-weighted tbd 98 db total harmonic distortion (note 8) thd r l = 10k ? full-scale signal -84 db channel separation (note 9) 1khz signal 80 100 db output mixers (lmx1, rmx1) pga gain range into mixer -15 0 +6 db pga gain step into mixer 3 db analogue outputs (lout1, rout1, lout2, rout2) programmable gain range -57 0 +6 db programmable gain step size monotonic 1 db mute attenuation 1khz, full scale signal 85 db headphone output (lout1, rout1 with 32 ? ? ? ? load) 0db full scale output voltage (note 10) avdd/3.3 vrms signal to noise ratio snr a-weighted tbd 93 db r l = 16 ? , po=20mw avdd=3.3v 0.008 -81 % db total harmonic distortion thd r l = 32 ? , po=20mw avdd=3.3v 0.007 - 83 % db
wm8976 preliminary technical data w ptd rev 2.0 february 2005 8 test conditions dcvdd=1.8v, avdd=dbvdd=spkvdd = 3.3v, t a = +25 o c, 1khz signal, fs = 48khz, 24-bit audio data unless otherwise stated. parameter symbol test conditions min typ max unit speaker output (lout2, rout2 with 8 ? ? ? ? bridge tied load, invrout2=1) spkboost=0 spkvdd/3.3 vrms full scale output voltage, 0db gain. (note 10) spkboost=1 (spkvdd/3.3)*1.5 output power p o output power is very closely correlated with thd; see below p o =200mw, r l = 8 ? , spkvdd=3.3v 0.04 -68 % db p o =320mw, r l = 8 ? , spkvdd=3.3v 1.0 -40 % db p o =500mw, r l = 8 ? , spkvdd=5v 0.02 -74 % db total harmonic distortion thd p o =860mw, r l = 8 ? , spkvdd=5v 1.0 -40 % db spkvdd=3.3v, r l = 8 ? 90 db signal to noise ratio snr spkvdd=5v, r l = 8 ? 90 db power supply rejection ratio 50 db out3/out4 outputs (with 10k ? ? ? ? / 50pf load) out3boost=0/ out4boost=0 spkvdd/3.3 vrms full-scale output voltage, 0db gain (note 10) out3boost=1 out4boost=1 (spkvdd/3.3)*1.5 vrms signal to noise ratio (note 6,7) snr a-weighted tbd 98 db total harmonic distortion (note 8) thd r l = 10 k ? full-scale signal -84 db channel separation (note 9) 1khz signal 80 100 db microphone bias mbvsel=0 0.9*avdd v bias voltage v micbias mbvsel=1 0.75*avdd v bias current source i micbias 3 ma output noise voltage vn 1k to 20khz 15 nv/ hz digital input / output input high level v ih 0.7 dbvdd v input low level v il 0.3 dbvdd v output high level v oh i ol =1ma 0.9 dbvdd v output low level v ol i oh -1ma 0.1xdbvdd v input capacitance tbd pf input leakage tbd pa terminology 1. input level to lip in pseudo-differential configurations is limited to a maximum of -3db or thd+n performance will be reduced. 2. note when beep path is not enabled then auxl and auxr have the same input impedances. 3. hold time is the length of time between a signal detected being too quiet and beginning to ramp up the gain. it does not apply to ramping down the gain when the signal is too loud, which happens without a delay. 4. ramp-up and ramp-down times are defined as the time it takes for the pga to sweep across 90% of its gain range. 5. all hold, ramp-up and ramp-down times scale proportionally with mclk 6. signal-to-noise ratio (db) C snr is a measure of the difference in level between the full scale output and the output with no signal applied. (no auto-zero or automute function is employed in achieving these results).
preliminary technical data wm8976 w ptd rev 2.0 february 2005 9 7. dynamic range (db) C dr is a measure of the difference between the highest and lowest portions of a signal. normally a thd+n measurement at 60db below full scale. the measured signal is then corrected by adding the 60db to it. (e.g. thd+n @ -60db= -32db, dr= 92db). 8. thd+n (db) C thd+n is a ratio, of the rms values, of (noise + distortion)/signal. 9. channel separation (db) C also known as cross-talk. this is a measure of the amount one channel is isolated from the other. normally measured by sending a full scale signal down one channel and measuring the other. 10. the maximum output voltage can be limited by the speaker power supply. if out3boost, out4boost or spkboost is set then s pkvdd s hould be 1.5xavdd to prevent clipping taking place in the output stage (when pga gains are set to 0db).
wm8976 preliminary technical data w ptd rev 2.0 february 2005 10 speaker output thd versus power speaker power vs thd+n (8ohm btl load) -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 0.00 50.00 100.00 150.00 200.00 250.00 300.00 350.00 400.00 450.00 500.00 output power (mw) thd+n (db) avdd=spkvdd=dbvdd=3.3, dcvdd=1.8 speaker power vs thd+n (8ohm btl load) -100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 0.00 100.00 200.00 300.00 400.00 500.00 600.00 700.00 800.00 900.00 1000.00 output power (mw) thd+n (db) avdd=dbvdd=3.3v, spkvdd=5v, dcvdd=1.8v
preliminary technical data wm8976 w ptd rev 2.0 february 2005 11 power consumption estimated power consumption for typical scenarios are shown below. for more information on power consumption of individual blocks, see "estimated supply currents" section. unless otherwise specified, all supply voltages are 3.3v. mode avdd dcvdd units off 0 0 ma sleep (vref maintained) 0.1 0 ma record (8khz, pll enabled) 5.3 0.8 ma stereo hp playback (44.1khz, pll enabled) 6.6 4.3 ma table 1 power consumption notes: 1. dbvdd supply current is not shown, as this is determined by loading on the digital i/o pins.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 12 signal timing requirements system clock timing mclk t mclkl t mclkh t mclky figure 1 system clock timing requirements test conditions dcvdd=1.8v, dbvdd=avdd=spkvdd=3.3v, dgnd=agnd=spkgnd=0v, t a = +25 o c, slave mode fs = 48khz, mclk = 256fs, 24-bit data, unless otherwise stated. parameter symbol min typ max unit system clock timing information mclk system clock cycle time t mclky tbd ns mclk duty cycle t mclkds 60:40 40:60 audio interface timing C master mode figure 2 digital audio data timing C master mode (see control interface)
preliminary technical data wm8976 w ptd rev 2.0 february 2005 13 test conditions dcvdd=1.8v, dbvdd=avdd=spkvdd=3.3v, dgnd=agnd=spkgnd=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 data input timing information frame propagation delay from bclk falling edge t dl 10 ns adcdat propagation delay from bclk falling edge t dda 10 ns dacdat setup time to bclk rising edge t dst 10 ns dacdat hold time from bclk rising edge t dht 10 ns audio interface timing C slave mode figure 3 digital audio data timing C slave mode test conditions dcvdd=1.8v, dbvdd=avdd=spkvdd=3.3v, dgnd=agnd=spkgnd=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 data input timing information bclk cycle time t bcy 50 ns bclk pulse width high t bch 20 ns bclk pulse width low t bcl 20 ns frame set-up time to bclk rising edge t lrsu 10 ns frame hold time from bclk rising edge t lrh 10 ns dacdat hold time from bclk rising edge t dh 10 ns adcdat propagation delay from bclk falling edge t dd 10 ns note: bclk period should always be greater than or equal to mclk period.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 14 control interface timing C 3-wire mode figure 4 control interface timing C 3-wire serial control mode test conditions dcvdd = 1.8v, dbvdd = avdd = spkvdd = 3.3v, dgnd = agnd = spkgnd = 0v, t a =+25 o c, slave mode, fs=48khz, mclk = 256fs, 24-bit data, unless otherwise stated. parameter symbol min typ max unit program register input information sclk rising edge to csb rising edge t scs 80 ns sclk pulse cycle time t scy 200 ns sclk pulse width low t scl 80 ns sclk pulse width high t sch 80 ns sdin to sclk set-up time t dsu 40 ns sclk to sdin hold time t dho 40 ns csb pulse width low t csl 40 ns csb pulse width high t csh 40 ns csb rising to sclk rising t css 40 ns pulse width of spikes that will be suppressed t ps 0 5 ns
preliminary technical data wm8976 w ptd rev 2.0 february 2005 15 control interface timing C 2-wire mode sdin sclk t 3 t 1 t 6 t 2 t 7 t 5 t 4 t 3 t 8 t 9 figure 5 control interface timing C 2-wire serial control mode test conditions dcvdd=1.8v, dbvdd=avdd=spkvdd=3.3v, dgnd=agnd=spkgnd=0v, t a =+25 o c, slave mode, fs=48khz, mclk = 256fs, 24-bit data, unless otherwise stated. parameter symbol min typ max unit program register input information sclk frequency 0 400 khz sclk low pulse-width t 1 1.3 us 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 sdin, sclk rise time t 6 300 ns sdin, 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
wm8976 preliminary technical data w ptd rev 2.0 february 2005 16 internal power on reset circuit figure 6 internal power on reset circuit schematic the wm8976 includes an internal power-on-reset circuit, as shown in figure 6, which is used reset the digital logic into a default state after power up. the por circuit is powered from avdd and monitors dvdd. it asserts porb low if avdd or dvdd is below a minimum threshold. figure 7 typical power up sequence where avdd is powered before dvdd figure 7 shows a typical power-up sequence where avdd comes up first. when avdd goes above the minimum threshold, v pora , there is enough voltage for the circuit to guarantee porb is asserted low and the chip is held in reset. in this condition, all writes to the control interface are ignored. now avdd is at full s upply level. next dvdd rises to v pord_on and porb is released high and all registers are in their default state and writes to the control interface may take place. on power down, where avdd falls first, porb is asserted low whenever avdd drops below the minimum threshold v pora_off .
preliminary technical data wm8976 w ptd rev 2.0 february 2005 17 figure 8 typical power up sequence where dvdd is powered before avdd figure 8 shows a typical power-up sequence where dvdd comes up first. first it is assumed that dvdd is already up to specified operating voltage. when avdd goes above the minimum threshold, v pora , there is enough voltage for the circuit to guarantee porb is asserted low and the chip is held in reset. in this condition, all writes to the control interface are ignored. when avdd rises to v pora_on , porb is released high and all registers are in their default state and writes to the control interface may take place. on power down, where dvdd falls first, porb is asserted low whenever dvdd drops below the minimum threshold v pord_off . symbol min typ max unit v pora 0.4 0.6 0.8 v v pora_on 0.9 1.2 1.6 v v pora_off 0.4 0.6 0.8 v v pord_on 0.5 0.7 0.9 v v pord_off 0.4 0.6 0.8 v table 2 typical por operation (typical values, not tested) notes: 1. if avdd and dvdd 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 will not reset and will resume normal operation when the voltage is back to the recommended level again. 2. the chip will enter reset at power down when avdd or dvdd 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 dvdd and avdd have zero rise time. this specification is guaranteed by design rather than test.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 18 recommended power up/down sequence in order to minimise output pop and click noise, it is recommended that the wm8976 device is powered up and down using one of the following sequences: power up when not using the output 1.5x boost stage: 1. turn on external power supplies. wait for supply voltage to settle. 2. set biasen = 1, bufioen = 1 and also the vmidsel[1:0] bits in the power management 1 register. r1 = 0x00d. *refer notes 1 and 2. 3. wait for the vmid supply to settle. *refer note 2. 4. set vroi = 1. r49 = 0x003. 5. set l/rout1en = 1. r1 = 0x180. 6. set l/rmixen and dacen = 1. r3 = 0x00f. 7. enable other mixers as required. 8. enable other outputs as required. power up when using the output 1.5x boost stage: 1. turn on external power supplies. wait for supply voltage to settle. 2. set biasen = 1, bufioen = 1, bufdcopen = 1 and also the vmidsel[1:0] bits in the power management 1 register. r1 = 0x10d. *refer notes 1 and 2. 3. wait for the vmid supply to settle. *refer note 2. 4. set vroi = 1 and spkboost = 1. r49 = 0x 007. 5. set l/rout2en = 1, l/rmixen = 1 and dacen = 1. r3 = 0x06f. *note 3. 6. enable other mixers as required. 7. enable other outputs as required. power down (all cases): 1. disable power management register 1 by setting r1[8:0]=0x00. 2. disable all other output stages. 3. remove external power supplies. notes: 1. this step enables the internal device bias buffer and the vmid buffer for unassigned inputs/outputs. this will provide a startup reference voltage for all inputs and outputs. this will cause the inputs and outputs to ramp towards vmid (not using output 1.5x boost) or 1.5 x (avdd/2) (using output 1.5x boost) in a way that is controlled and predictable (see note 2). 2. choose the value of the vmidsel bits based on the startup time (vmidsel=10 for slowest startup, vmidsel=11 for fastest startup). startup time is defined by the value of the vmidsel bits (the reference impedance) and the external decoupling capacitor on vmid. 3. setting dacen to off while operating in x1.5 boost mode will cause the vmid voltage to drop to avdd/2 midrail level and cause an output pop. to avoid this de-select dac from mixer inputs before disabling the dac. in addition to the power on sequence, it is recommended that the zero cross functions are used when changing the volume in the pgas to avoid any audible pops or clicks.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 19 v pora dgnd internal por active device ready no power v por_off power supply por i 2 s clocks adc internal state t midrail_on analogue inputs adcdat pin gd adcen bit power down init normal operation normal operation init pd power down adc enabled adc enabled adc off t adcint dnc inppgaen bit t adcint inppga enabled dnc gd gd gd por por undefined vmid enabled vmidsel/ biasen bits avdd/2 t midrail_off (note 1) (note 2) (note 3) (note 4) v por_on figure 9 adc power up and down sequence (not to scale) symbol min typical max unit t midrail_on 300 ms t midrail_off 6 s t adcint 2/fs n /fs adc group delay 29/fs n /fs table 3 typical adc power up and down sequence notes: 1. the analogue input pin charge time, t midrail_on, is determined by the vmid pin charge time. this time is dependent upon the value of vmid decoupling capacitor and vmid pin input resistance and avdd power supply rise time. 2. the analogue input pin discharge time, t midrail_off, is determined by the vmid pin discharge time. this time is dependent upon the value of vmid decoupling capacitor and vmid pin input resistance and avdd power supply decay time. the time, t midrail_off , is measured using a 1 f capacitor and will vary dependent upon the value of input coupling capacitor. 3. the timings t midrail_on and t midrail_off are measured using a 1 f capacitor and will vary dependent upon the value of input coupling capacitor. the measurements were taken at 10% and 90% points on the charge time curve. 4. while the adc is enabled there will be lsb data bit activity on the adcdat pin due to system noise but no significant digital output will be present. 5. the vmidsel and bi asen bits must be set to enable analogue input midrail voltage and for normal adc operation. 6. adcdat data output delay from power up - with power supplies starting from 0v - is determined primarily by the vmid charge time. adc initialisation and power management bits may be set immediately after por is released; vmid charge time will be significantly longer and will dictate when the device is stabilised for analogue input. 7. adcdat data output delay at power up from device standby (power supplies already applied) is determined by adc initialisation time, 2/fs.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 20 figure 10 dac power up and down sequence (not to scale) symbol min typical max unit t line_midrail_on 60 ms t line_midrail_off 6 s t hp_midrail_on 150 ms t hp__midrail_off 9 s t dacint 2/fs n /fs dac group delay 29/fs n /fs table 4 typical dac power up and down sequence notes: 1. the lineout charge time, t line_midrail_on, is mainly determined by the vmid pin charge time. this time is dependent upon the value of vmid decoupling capacitor and vmid pin input resistance and avdd power supply rise time. the values above were measured using a 1 f capacitor and were taken at 10% and 90% points on the charge time curve. 2. it is not advisable to allow dacdat data input during initialisation of the dac. if the dac data value is not zero at point of initialisation, then this is likely to cause a pop noise on the analogue outputs. the same is also true if the dacdat is removed at a none zero value, and no mute function has been applied to the signal beforehand. 3. the lineout discharge time, t line_midrail_off, is dependent upon the value of the lineout coupling capacitor and the leakage resistance path to ground. the values above were measured using a 1 f capacitor and were taken at 10% and 90% points on the discharge time curve.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 21 4. the headphone charge time, t hp_midrail_on, is mainly determined by the headphone output coupling capacitor charge time. this time is dependent upon the value of the headphone coupling capacitor. the values above were measured using a 100 f capacitor and were taken at 10% and 90% points on the charge time curve. 5. the headphone discharge time, t hp_midrail_off, is dependent upon the value of the headphone coupling capacitor and the leakage resistance path to ground. the values above were measured using a 100 f capacitor and were taken at 10% and 90% points on the discharge time curve. 6. the vmidsel and biasen bits must be set to enable analogue output midrail voltage and for normal dac operation.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 22 device description introduction the wm8976 is a low power audio codec combining a high quality stereo audio dac and mono adc, with flexible line and microphone input and output processing. applications for this device include digital camcorders, and digital still cameras with mono record and mono or stereo playback capability. features the chip offers great flexibility in use, and so can support many different modes of operation as follows: microphone input a microphone pre-amp is provided, allowing for a microphone to be pseudo-differentially connected, with user-defined gain using internal resistors. the provision of the common mode input pin allows for rejection of common mode noise on the microphone input (level depends on gain setting chosen). a microphone bias is output from the chip which can be used to bias the microphone. the signal routing can be configured to allow manual adjustment of mic level, or indeed to allow the alc loop to control the level of mic signal that is transmitted. total gain through the microphone path of up to +55.25db can be selected. pga and alc operation a programmable gain amplifier is provided in the input path to the adc. this may be used manually or in conjunction with a mixed analogue/digital automatic level control (alc) which keeps the recording volume constant. line inputs (auxl, auxr) the inputs, auxl and auxr, can be used as a stereo line input or as an input for warning tones (or beeps) etc. the auxl input can be summed into the record path, along with the microphone preamp output. adc the adc uses a multi-bit high-order oversampling architecture to deliver optimum performance with low power consumption. hi-fi dac the stereo hi-fi dac provides high quality audio playback suitable for all portable audio hi-fi type applications, including mp3 players and portable disc players of all types. output mixers flexible mixing is provided on the outputs of the device. a stereo mixer is provided for the stereo headphone or line outputs, lout1/rout1, and additional summers on the out3/out4 outputs allow for an optional differential or stereo line output on these pins. gain adjustment pgas are provided for the lout1/rout1 and lout2/rout2 outputs, and signal switching is provided to allow for all possible signal combinations. the output buffers can be configured in several ways, allowing support of up to three sets of external transducers; ie stereo headphone, btl speaker, and btl earpiece may be connected simultaneously. thermal implications should be considered before simultaneous full power operation of all outputs is attempted. alternatively, if a speaker output is not required, the lout2 and rout2 pins might be used as a stereo headphone driver, (disable output invert buffer on rout2). in that case two sets of headphones might be driven, or the lout2 and rout2 pins used as a line output driver.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 23 out3 and out4 can be configured to provide an additional stereo lineout from the output of the dacs, the mixers or the input microphone boost stages. alternatively out4 can be configured as a mono mix of left and right dacs or mixers, or simply a buffered version of the chip midrail reference voltage. out3 can also be configured as a buffered vmid output. this voltage may then be used as a headphone pseudo ground allowing removal of the large ac coupling capacitors often used in the output path. audio interfaces the wm8976 has a standard audio interface, to support the transmission of mono or stereo data to and from the chip. this interface is a 3 wire standard audio interface which supports a number of audio data formats including i2s, dsp/pcm mode (a burst mode in which lrc sync plus 2 data packed words are transmitted), msb-first, left justified and msb-first, right justified, and can operate in master or slave modes. control interfaces to allow full software control over all its features, the wm8976 offers a choice of 2 or 3 wire mpu control interface. it is fully compatible and an ideal partner for a wide range of industry standard microprocessors, controllers and dsps. selection between the modes is via the mode pin. in 2 wire mode the address of the device is fixed as 0011010. clocking schemes wm8976 offers the normal audio dac clocking scheme operation, where 256fs mclk is provided to the dac and adc. a pll is included which may be used to generate these clocks in the event that they are not available from the system controller. this pll uses an input clock, typically the 12mhz usb or ilink clock, to generate high quality audio clocks. if this pll is not required for generation of these clocks, it can be reconfigured to generate alternative clocks which may then be output on the gpio pin and used elsewhere in the system. power control the design of the wm8976 has given much attention to power consumption without compromising performance. it operates at very low voltages, and includes the ability to power off any unused parts of the circuitry under software control, and includes standby and power off modes. operation scenarios flexibility in the design of the wm8976 allows for a wide range of operational scenarios, some of which are proposed below: camcorder; the provision of a microphone preamplifier allows support for both internal and external microphones. all drivers for speaker, headphone and line output connections are integrated. the selectable application filters after the adc provide for features such as wind noise reduction, or mechanical noise reducing filters. digital still camera recording; support for digital recording is similar to the camcorder case. but additionally if the dsc supports mp3 playback, and perhaps recording, the ability of the adc to support full 48ks/s high quality recording increases device flexibility. analog fm tuner support an analog stereo fm tuner might be connected to the aux inputs of wm8976, and the stereo signal listened to via headphones, if required.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 24 input signal path the wm8976 has a number of flexible analogue inputs. the input pga stage is followed by a boost/mix stage which drives into the adc. the input path has three input pins which can be configured in a variety of ways to accommodate single-ended, differential or dual differential microphones. there are two auxiliary input pins, and the auxl input can be fed into to the input boost/mix stage as well as driving into the output path. a bypass path exists from the output of the boost/mix stage into the output left mixer. microphone inputs the wm8976 can accommodate a variety of microphone configurations including single ended and differential inputs. the input to the differential input pga are lin, lip and l2. in single-ended microphone input configuration the microphone signal should be input to lin or rin and the internal nor gate configured to clamp the non-inverting input of the input pga to vmid. figure 11 microphone input pga circuit
preliminary technical data wm8976 w ptd rev 2.0 february 2005 25 the input pga is enabled by the ippgaenl register bit. register address bit label default description r2 power management 2 2 inppgaenl 0 left channel input pga enable 0 = disabled 1 = enabled table 5 input pga enable register settings register address bit label default description 0 lip2inppga 1 connect lip pin to left channel input pga amplifier positive terminal. 0 = lip not connected to input pga 1 = input pga amplifier positive terminal connected to lip (constant input impedance) 1 lin2inppga 1 connect lin pin to left channel input pga negative terminal. 0=lin not connected to input pga 1=lin connected to input pga amplifier negative terminal. r44 input control 2 l2_2inppga 0 connect l2 pin to left channel input pga positive terminal. 0=l2 not connected to input pga 1=l2 connected to input pga amplifier positive terminal (constant input impedance). table 6 input pga control input pga volume controls the input microphone pga has a gain range from -12db to +35.25db in 0.75db steps. the gain from the lin input to the pga output and from the l2 amplifier to the pga output are always common and controlled by the register bits inppgavoll[5:0]. these register bits also affect the lip pin when lip2inppga=1, the l2 pin when l2_2inppga=1 and the l2 pin when l2_2inppga=1. when the automatic level control (alc) is enabled the input pga gain is controlled automatically and the inppgavoll bits should not be used.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 26 register address bit label default description 5:0 inppgavoll 010000 input pga volume 000000 = -12db 000001 = -11.25db . 010000 = 0db . 111111 = 35.25db 6 inppgamutel 0 mute control for input pga: 0=input pga not muted, normal operation 1=input pga muted (and disconnected from the following input boost stage). 7 inppgazcl 0 input pga zero cross enable: 0=update gain when gain register changes 1=update gain on 1 st zero cross after gain register write. r45 input pga volume control 8 inpgaupdate not latched inpga volume does not update until a 1 is written to inpgaupdate table 7 input pga volume control
preliminary technical data wm8976 w ptd rev 2.0 february 2005 27 auxilliary inputs there are two auxilliary inputs, auxl and auxr which can be used for a variety of purposes such as stereo line inputs or as a beep input signal to be mixed with the outputs. the auxl input can be used as a line input to the input boost stage which has gain adjust of - 12db to +6db in 3db steps (plus off). see the input boost section for further details. the auxl/r inputs can also be mixed into the output channel mixers, with a gain of -15db to +6db plus off. in addition the auxr input can be summed into the right speaker output path (rout2) with a gain adjust of -15 to +6db. this allows a beep input to be output on the speaker outputs only without affecting the headphone or lineout signals. input boost the stereo input pga stage is followed by an input boost circuit. the input boost circuit has 3 selectable inputs: the input microphone pga output, the aux amplifier output and the l2 input pin (can be used as a line input, bypassing the input pga). these three inputs can be mixed together and have individual gain boost/adjust as shown in figure 12. figure 12 input boost stage the input pga path can have a +20db boost (pgaboostl=1), a 0db pass through (pgaboostl=0) or be completely isolated from the input boost circuit (inppgamutel=1). register address bit label default description r47 input boost control 8 pgaboost 1 boost enable for input pga: 0 = pga output has +0db gain through input boost stage. 1 = pga output has +20db gain through input boost stage. table 8 input boost stage control the auxilliary amplifier path to the boost stages is controlled by the auxl2boostvol[2:0] register bits. when auxl2boostvol=000 this path is completely disconnected from the boost stage. settings 001 through to 111 control the gain in 3db steps from -12db to +6db. the l2 path to the boost stage is controlled by the lip2boostvol[2:0] register bits. when l2_2boostvol=000 the l2 input pin is completely disconnected from the boost stage. settings 001 through to 111 control the gain in 3db steps from -12db to +6db.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 28 register address bit label default description 2:0 auxl2boostvol 000 controls the auxilliary amplifer to the input boost stage: 000=path disabled (disconnected) 001=-12db gain through boost stage 010=-9db gain through boost stage 111=+6db gain through boost stage r47 input boost control 6:4 l2_2boostvol 000 controls the l2 pin to the input boost stage: 000=path disabled (disconnected) 001=-12db gain through boost stage 010=-9db gain through boost stage 111=+6db gain through boost stage table 9 input boost stage control the boost stage is enabled under control of the boostenl register bit. register address bit label default description r2 power management 2 4 boostenl 0 left channel input boost enable 0 = boost stage off 1 = boost stage on table 10 input boost enable control microphone biasing circuit the micbias output provides a low noise reference voltage suitable for biasing electret type microphones and the associated external resistor biasing network. refer to the applications information section for recommended external components. the micbias voltage can be altered via the mbvsel register bit. w hen m bvsel=0, micbias=0.9*avdd and when m bvsel=1, micbias=0.6*avdd. the output can be enabled or disabled using the micben control bit. register address bit label default description r1 power management 1 4 micben 0 microphone bias enable 0 = off (high impedance output) 1 = on table 11 microphone bias enable control register address bit label default description r44 input control 8 mbvsel 0 microphone bias voltage control 0 = 0.9 * avdd 1 = 0.6 * avdd table 12 microphone bias voltage control the internal micbias circuitry is shown in figure 13. note that the maximum source current capability for micbias is 3ma. the external biasing resistor therefore must be large enough to limit the micbias current to 3ma.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 29 figure 13 microphone bias schematic analogue to digital converter (adc) the wm8976 uses a multi-bit, oversampled sigma-delta adc. the use of multi-bit feedback and high oversampling rates reduces the effects of jitter and high frequency noise. the adc full scale input level is proportional to avdd. with a 3.3v supply voltage, the full scale level is 1.0v rms . any voltage greater than full scale may overload the adc and cause distortion. adc digital filters the adc filters perform true 24 bit signal processing to convert the raw multi-bit oversampled data from the adc to the correct sampling frequency to be output on the digital audio interface. the digital filter path for each adc channel is illustrated in figure 14. figure 14 adc digital filter path the adc is enabled by the adcenl register bit. register address bit label default description r2 power management 2 0 adcenl 0 enable adc: 0 = adc disabled 1 = adc enabled table 13 adc enable control agnd mbvsel=0 micbias = 1.8 x vmid = 0.9 x avdd vmid internal resistor internal resistor mb mbvsel=1 micbias = 1.2 x vmid = 0.6 x avdd
wm8976 preliminary technical data w ptd rev 2.0 february 2005 30 the polarity of the output signal can also be changed under software control using the adclpol register bit. the oversampling rate of the adc can be adjusted using the adcosr register bit. with adcosr=0 the oversample rate is 64x which gives lowest power operation and when adcosr=1 the oversample rate is 128x which gives best performance. register address bit label default description 0 adclpol 0 adc channel polarity adjust: 0=normal 1=inverted r14 adc control 3 adcosr128 0 adc oversample rate select: 0=64x (lower power) 1=128x (best performance) table 14 adc control selectable high pass filter a selectable high pass filter is provided. to disable this filter set hpfen=0. the filter has two modes controlled by hpfapp. in audio mode (hpfapp=0) the filter is first order, with a cut-off frequency of 3.7hz. in application mode (hpfapp=1) the filter is second order, with a cut-off frequency selectable via the hpfcut register. the cut-off frequencies when hpf app=1 are shown in table 16. register address bit label default description 8 hpfen 1 high pass filter enable 0=disabled 1=enabled 7 hpfapp 0 select audio mode or application mode 0=audio mode (1 st order, fc = ~3.7hz) 1=application mode (2 nd order, fc = hpfcut) r14 adc control 6:4 hpfcut 000 application mode cut-off frequency see table 16 for details. table 15 adc enable control fs (khz) sr=101/100 sr=011/010 sr=001/000 hpfcut 8 11.025 12 16 22.05 24 32 44.1 48 000 82 113 122 82 113 122 82 113 122 001 102 141 153 102 141 153 102 141 153 010 131 180 156 131 180 156 131 180 156 011 163 225 245 163 225 245 163 225 245 100 204 281 306 204 281 306 204 281 306 101 261 360 392 261 360 392 261 360 392 110 327 450 490 327 450 490 327 450 490 111 408 563 612 408 563 612 408 563 612 table 16 high pass filter cut-off frequencies (hpfapp=1) note that the high pass filter values (when hpfapp=1) work on the basis that the sr register bits are set correctly for the actual sample rate as shown in table 16.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 31 programmable notch filter a programmable notch filter is provided. this filter has a variable centre frequency and bandwidth, programmable via two coefficients, a0 and a1. a0 and a1 are represented by the register bits nfa0[13:0] and nfa1[13:0]. because these coefficient values require four register writes to setup there is an nfu (notch filter update) flag which should be set only when all four registers are setup. register address bit label default description 6:0 nfa0[13:7] 0 notch filter a0 coefficient, bits [13:7] 7 nfen 0 notch filter enable: 0=disabled 1=enabled r27 notch filter 1 8 nfu 0 notch filter update. the notch filter values used internally only update when one of the nfu bits is set high. 6:0 nfa0[6:0] 0 notch filter a0 coefficient, bits [6:0] r28 notch filter 2 8 nfu 0 notch filter update. the notch filter values used internally only update when one of the nfu bits is set high. 6:0 nfa1[13:7] 0 notch filter a1 coefficient, bits [13:7] r29 notch filter 3 8 nfu 0 notch filter update. the notch filter values used internally only update when one of the nfu bits is set high. 6:0 nfa1[6:0] 0 notch filter a1 coefficient, bits [6:0] r30 notch filter 4 8 nfu 0 notch filter update. the notch filter values used internally only update when one of the nfu bits is set high. table 17 notch filter function the coefficients are calculated as follows: ) 2 / tan( 1 ) 2 / tan( 1 0 b b w w a + ? = ) cos( ) 1 ( 0 0 1 w a a + ? = where: s c f f w / 2 0 = s b b f f w / 2 = f c = centre frequency in hz, f b = -3db bandwidth in hz, f s = sample frequency in hz the actual register values can be determined from the coefficients as follows: nfa0 = -a0 x 2 13 nfa1 = -a1 x 2 12 digital adc volume control the output of the adc can be digitally attenuated over a range from C127db to 0db in 0.5db steps. the gain for a given eight-bit code x is given by: 0.5 (g-255) db for 1 g 255; mute for g = 0
wm8976 preliminary technical data w ptd rev 2.0 february 2005 32 register address bit label default description 7:0 adcvol [7:0] 11111111 ( 0db ) adc digital volume control 0000 0000 = digital mute 0000 0001 = -127db 0000 0010 = -126.5db ... 0.5db steps up to 1111 1111 = 0db r15 adc digital volume 8 adcvu not latched adc volume does not update until a 1 is written to adcvu table 18 adc digital volume control input limiter / automatic level control (alc) the wm8976 has an automatic pga gain control circuit, which can function as an input peak limiter or as an automatic level control (alc). in input peak limiter mode (alcmode bit = 1), a digital peak detector detects when the input signal goes above a predefined level and will ramp the pga gain down to prevent the signal becoming too large for the input range of the adc. when the signal returns to a level below the threshold, the pga gain is slowly returned to its starting level. the peak limiter cannot increase the pga gain above its static level. figure 15 input peak limiter operation in alc mode (alcmode bit = 0) the circuit aims to keep a constant recording volume irrespective of the input signal level. this is achieved by continuously adjusting the pga gain so that the signal level at the adc input remains constant. a digital peak detector monitors the adc output and changes the pga gain if necessary.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 33 figure 16 alc operation the alc/limiter function is enabled by setting the register bit alcsel. when enabled, the recording volume can be programmed between C6db and C28.5db (relative to adc full scale) using the alclvl register bits. an upper limit for the pga gain can be imposed by setting the alcmax control bits and a lower limit for the pga gain can be imposed by setting the alcmin control bits. alchld, alcdcy and alcatk control the hold, decay and attack times, respectively: hold time is the time delay between the peak level detected being below target and the pga gain beginning to ramp up. it can be programmed in power-of-two (2 n ) steps, e.g. 2.67ms, 5.33ms, 10.67ms etc. up to 43.7s. alternatively, the hold time can also be set to zero. the hold time is not active in limiter mode (alcmode = 1). the hold time only applies to gain ramp-up, there is no delay before ramping the gain down when the signal level is above target. decay (gain ramp-up) time is the time that it takes for the pga gain to ramp up and is given as a time per gain step, time per 6db change and time to ramp up over 90% of its range. the decay time can be programmed in power-of-two (2 n ) steps, from 3.3ms/6db, 6.6ms/6db, 13.1ms/6db, etc. to 3.36s/6db. attack (gain ramp-down) time is the time that it takes for the pga gain to ramp down and is given as a time per gain step, time per 6db change and time to ramp down over 90% of its range. the attack time can be programmed in power-of-two (2 n ) steps, from 832us/6db, 1.66ms/6db, 3.328us/6db, etc. to 852ms/6db. nb, in peak limiter mode the gain control circuit runs approximately 4x faster to allow reduction of fast peaks. attack and decay times for peak limiter mode are given below. the hold, decay and attack times given in table 19 are constant across sample rates so long as the sr bits are set correctly. e.g. when sampling at 48khz the sample rates stated in table 19 will only be correct if the sr bits are set to 000 (48khz). i f the actual sample rate was only 44.1khz then the hold, decay and attack times would be scaled down by 44.1/48.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 34 register address bit label default description 8 alcsel 0 alc function select 0=alc disabled 1=alc enabled 5:3 alcmaxgain [2:0] 111 (+35.25db) set maximum gain of pga 111=+35.25db 110=+29.25db 101=+23.25db 100=+17.25db 011=+11.25db 010=+5.25db 001=-0.75db 000=-6.75db r32 alc control 1 2:0 alcmingain [2:0] 000 (-12db) set minimum gain of pga 000=-12db 001=-6db 010=0db 011=+6db 100=+12db 101=+18db 110=+24db 111=+30db 7:4 alchld [3:0] 0000 (0ms) alc hold time before gain is increased. 0000 = 0ms 0001 = 2.67ms 0010 = 5.33ms (time doubles with every step) 1111 = 43.691s 3:0 alclvl [3:0] 1011 (-12db) alc target C sets signal level at adc input 0000 = -28.5db fs 0001 = -27.0db fs (1.5db steps) 1110 = -7.5db fs 1111 = -6db fs r33 alc control 2 8 alczc 0 (zero cross off) alc uses zero cross detection circuit.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 35 8 alcmode 0 determines the alc mode of operation: 0=alc mode 1=limiter mode. decay (gain ramp-up) time (alcmode ==0) per step per 6db 90% of range 0000 410us 3.3ms 24ms 0001 820us 6.6ms 48ms 0010 1.64ms 13.1ms 192ms (time doubles with every step) 0011 (13ms/6db) 1010 or high er 420ms 3.36s 24.576s decay (gain ramp-up) time (alcmode ==1) per step per 6db 90% of range 0000 90.8us 726.4us 5.26ms 0001 181.6us 1.453ms 10.53ms 0010 363.2us 2.905ms 21.06ms (time doubles with every step) 7:4 alcdcy [3:0] 0011 (2.9ms/6db) 1010 93ms 744ms 5.39s alc attack (gain ramp-down) time (alcmode == 0) per step per 6db 90% of range 0000 104us 832us 6ms 0001 208us 1. 664ms 12ms 0010 416us 3. 328ms 24.1ms (time doubles with every step) 0010 (832us/6db) 1010 or high er 106ms 852ms 6.18s alc attack (gain ramp-down) time (alcmode == 1) per step per 6db 90% of range 0000 22.7us 182.4us 1.31ms 0001 45.4us 363.2us 2.62ms 0010 90.8us 726.4us 5.26ms (time doubles with every step) r34 alc control 3 3:0 alcatk [3:0] 0010 (182us/6db) 1010 23.2ms 186ms 1.348s table 19 alc control registers
wm8976 preliminary technical data w ptd rev 2.0 february 2005 36 alc clip protection to prevent clipping when a large signal occurs just after a period of quiet, the alc circuit includes a clip protection function. if the adc input signal exceeds 87.5% of full scale (C1.16db), the pga gain is ramped down at the maximum attack rate (as when alcatk = 0000), until the signal level falls below 87.5% of full scale. this function is automatically enabled whenever the alc is enabled. note: if atk = 0000, then the clip protection circuit makes no difference to the operation of the alc. it is designed to prevent clipping when long attack times are used. noise gate when the signal is very quiet and consists mainly of noise, the alc function may cause noise pumping, i.e. loud hissing noise during silence periods. the wm8976 has a noise gate function that prevents noise pumping by comparing the signal level at the input pins against a noise gate threshold, ngth. the noise gate cuts in when: signal level at adc [db] < ngth [db] + pga gain [db] + mic boost gain [db] this is equivalent to: signal level at input pin [db] < ngth [db] the pga gain is then held constant (preventing it from ramping up as it normally would when the signal is quiet). the table below summarises the noise gate control register. the ngth control bits set the noise gate threshold with respect to the adc full-scale range. the threshold is adjusted in 6db steps. levels at the extremes of the range may cause inappropriate operation, so care should be taken with setCup of the function. note that the noise gate only works in conjunction with the alc function. register address bit label default description 2:0 ngth 000 noise gate threshold: 000=-39db 001=-45db 010=-51db (6db steps) 111=-81db r35 alc noise gate control 3 ngaten 0 noise gate function enable 1 = enable 0 = disable table 20 alc noise gate control output signal path the wm8976 output signal paths consist of digital application filters, up-sampling filters, stereo hi-fi dacs, analogue mixers, speaker, stereo headphone and stereo line/mono/midrail output drivers. the digital filters and dac are enabled by register bits dacenl and dacenr. the mixers and output drivers can be separately enabled by individual control bits (see analogue outputs). thus it is possible to utilise the analogue mixing and amplification provided by the wm8976, irrespective of whether the dacs are running or not. the wm8976 dacs receive digital input data on the dacdat pin. the digital filter block processes the data to provide the following functions: digital volume control graphic equaliser a digital peak limiter. sigma-delta modulation high performance sigma-delta audio dac converts the digital data into an analogue signal.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 37 figure 17 dac digital filter path the analogue outputs from the dacs can then be mixed with the aux analogue inputs and the adc analogue input. the mix is fed to the output drivers for headphone (lout1/rout1), speaker (lout2/rout2) or line (out3/out4). out3 and out4 have additional mixers which allow them to output different signals to the headphone and speaker outputs. digital playback (dac) path digital data is passed to the wm8976 via the flexible audio interface and is then passed through a variety of advanced digital filters as shown in figure 17 to the hi-fi dacs. the dacs are enabled by the dacenl/r register bits. register address bit label default description 0 dacenl 0 left channel dac enable 0 = dac disabled 1 = dac enabled r3 power management 3 1 dacenr 0 right channel dac enable 0 = dac disabled 1 = dac enabled table 21 dac enable control the wm8976 also has a soft mute function, which gradually attenuates the volume of the digital signal to zero. when removed, the gain will ramp back up to the digital gain setting. register address bit label default description 0 daclpol 0 left dac output polarity: 0 = non-inverted 1 = inverted (180 degrees phase shift) 1 dacrpol 0 right dac output polarity: 0 = non-inverted 1 = inverted (180 degrees phase shift) 2 amute 0 automute enable 0 = amute disabled 1 = amute enabled 3 dacosr128 0 dac oversampling rate: 0=64x (lowest power) 1=128x (best performance) r10 dac control 6 softmute 0 softmute enable: 0=disabled 1=enabled table 22 dac control register the digital audio data is converted to oversampled bit streams in the on-chip, true 24-bit digital interpolation filters. the bitstream data enters the multi-bit, sigma-delta dacs, which convert it to a high quality analogue audio signal. the multi-bit dac architecture reduces high frequency noise and sensitivity to clock jitter. it also uses a dynamic element matching technique for high linearity and low distortion. the dac output defaults to non-inverted. setting daclpol will invert the dac output phase on the left channel and dacrpol inverts the phase on the right channel.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 38 automute the dac has an automute function which applies an analogue mute when 1024 consecutive zeros are detected. the mute is released as soon as a non-zero sample is detected. automute can be disabled using the amute control bit. digital hi-fi dac volume (gain) control the signal volume from each hi-fi dac can be controlled digitally. the gain and attenuation range is C127db to 0db in 0.5db steps. the level of attenuation for an eight-bit code x is given by: 0.5 (x-255) db for 1 x 255; mute for x = 0 register address bit label default description 7:0 daclvol [7:0] 11111111 ( 0db ) left dac digital volume control 0000 0000 = digital mute 0000 0001 = -127db 0000 0010 = -126.5db ... 0.5db steps up to 1111 1111 = 0db r11 left dac digital volume 8 dacvu not latched dac left and dac right volume do not update until a 1 is written to dacvu (in reg 11 or 12) 7:0 dacrvol [7:0] 11111111 ( 0db ) right dac digital volume control 0000 0000 = digital mute 0000 0001 = -127db 0000 0010 = -126.5db ... 0.5db steps up to 1111 1111 = 0db r12 right dac digital volume 8 dacvu not latched dac left and dac right volume do not update until a 1 is written to dacvu (in reg 11 or 12) table 23 dac digital volume control note: an additional gain of up to +12db can be added using the gain block embedded in the digital peak limiter circuit (see dac output limiter section). dac 5-band equaliser a 5-band graphic equaliser function which can be used to change the output frequency levels to suit the environment. this can be applied to the adc or dac path and is described in the 5-band equaliser section for further details on this feature. dac 3-d enhancement the wm8976 has an advanced digital 3-d enhancement feature which can be used to vary the perceived stereo separation o f the left and right channels. like the 5-band equaliser this feature can be applied to either the record path or the plaback path but not both simultaneously. see the 3-d stereo enhancement section for further details on this feature. dac digital output limiter the wm8976 has a digital output limiter function. the operation of this is shown in figure 18. in this diagram the upper graph shows the envelope of the input/output signals and the lower graph shows the gain characteristic.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 39 figure 18 dac digital limiter operation the limiter has a programmable upper threshold which is close to 0db. referring to figure 18, in normal operation (limboost=000 => limit only) signals below this threshold are unaffected by the limiter. signals above the upper threshold are attenuated at a specific attack rate (set by the limatk register bits) until the signal falls below the threshold. the limiter also has a lower threshold 1db below the upper threshold. when the signal falls below the lower threshold the signal is amplified at a specific decay rate (controlled by limdcy register bits) until a gain of 0db is reached. both threshold levels are controlled by the limlvl register bits. the upper threshold is 0.5db above the value programmed by limlvl and the lower threshold is 0.5db below the limlvl value. volume boost the limiter has programmable upper gain which boosts signals below the threshold to compress the dynamic range of the signal and increase its perceived loudness. this operates as an alc function with limited boost capability. the volume boost is from 0db to +12db in 1db steps, controlled by the limboost register bits. the output limiter volume boost can also be used as a stand alone digital gain boost when the limiter is disabled.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 40 register address bit label default description 3:0 limatk 0010 limiter attack time (per 6db gain change) for 44.1khz sampling. note that these will scale with sample rate. 0000=94us 0001=188s 0010=375us 0011=750us 0100=1.5ms 0101=3ms 0110=6ms 0111=12ms 1000=24ms 1001=48ms 1010=96ms 1011 to 1111=192ms 7:4 limdcy 0011 limiter decay time (per 6db gain change) for 44.1khz sampling. note that these will scale with sample rate: 0000=750us 0001=1.5ms 0010=3ms 0011=6ms 0100=12ms 0101=24ms 0110=48ms 0111=96ms 1000=192ms 1001=384ms 1010=768ms 1011 to 1111=1.536s r24 dac digital limiter control 1 8 limen 0 enable the dac digital limiter: 0=disabled 1=enabled 3:0 limboost 0000 limiter volume boost (can be used as a stand alone volume boost when limen=0): 0000=0db 0001=+1db 0010=+2db (1db steps) 1011=+11db 1100=+12db 1101 to 1111=reserved r25 dac digital limiter control 2 6:4 limlvl 000 programmable signal threshold level (determines level at which the limiter starts to operate) 000=-1db 001=-2db 010=-3db 011=-4db 100=-5db 101 to 111=-6db table 24 dac digital limiter control
preliminary technical data wm8976 w ptd rev 2.0 february 2005 41 5-band graphic equaliser a 5-band graphic eq is provided, which can be applied to the adc or dac path, together with 3d enhancement, under control of the eq3dmode register bit. register address bit label default description r18 eq control 1 8 eq3dmode 1 0 = equaliser and 3d enhancement applied to adc path 1 = equaliser and 3d enhancement applied to dac path table 25 eq and 3d enhancement dac or adc path select the equaliser consists of low and high frequency shelving filters (band 1 and 5) and three peak filters for the centre bands. each has adjustable cut-off or centre frequency, and selectable boost (+/- 12db in 1db steps). the peak filters have selectable bandwidth. register address bit label default description 4:0 eq1g 01100 (0db) band 1 gain control. see table 31 for details. r18 eq band 1 control 6:5 eq1c 01 band 1 cut-off frequency: 00=80hz 01=105hz 10=135hz 11=175hz table 26 eq band 1 control register address bit label default description 4:0 eq2g 01100 (0db) band 2 gain control. see table 31 for details. 6:5 eq2c 01 band 2 centre frequency: 00=230hz 01=300hz 10=385hz 11=500hz r19 eq band 2 control 8 eq2bw 0 band 2 bandwidth control 0=narrow bandwidth 1=wide bandwidth table 27 eq band 2 control register address bit label default description 4:0 eq3g 01100 (0db) band 3 gain control. see table 31 for details. 6:5 eq3c 01 band 3 centre frequency: 00=650hz 01=850hz 10=1.1khz 11=1.4khz r20 eq band 3 control 8 eq3bw 0 band 3 bandwidth control 0=narrow bandwidth 1=wide bandwidth table 28 eq band 3 control
wm8976 preliminary technical data w ptd rev 2.0 february 2005 42 register address bit label default description 4:0 eq4g 01100 (0db) band 4 gain control. see table 31 for details 6:5 eq4c 01 band 4 centre frequency: 00=1.8khz 01=2.4khz 10=3.2khz 11=4.1khz r21 eq band 4 control 8 eq4bw 0 band 4 bandwidth control 0=narrow bandwidth 1=wide bandwidth table 29 eq band 4 control register address bit label default description 4:0 eq5g 01100 (0db) band 5 gain control. see table 31 for details. r22 eq band 5 gain control 6:5 eq5c 01 band 5 cut-off frequency: 00=5.3khz 01=6.9khz 10=9khz 11=11.7khz table 30 eq band 5 control gain register gain 00000 +12db 00001 +11db 00010 +10db . (1db steps) 01100 0db 01101 -1db 11000 to 11111 -12db table 31 gain register table 3d stereo enhancement the wm8976 has a digital 3d enhancement option to artificially increase the separation between the left and right channels. selection of 3d for playback is controlled by register bit eq3dmode. switching the eq/3d bit from record to playback or from playback to record may only be done when adc and dac are disabled. the wm8976 control interface will only allow eq3dmode to be changed when adc and dac are disabled (ie adcenl = 0, dacenl = 0 and dacenr = 0). the depth3d setting controls the degree of stereo expansion. when 3d enhancement is used, it may be necessary to attenuate the signal by 6db to avoid limiting.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 43 register address bit label default description r41 (29h) 3d 3:0 depth3d[3:0] 0000 stereo depth 0000: 0% (minimum 3d effect) 0001: 6.67% .... 1110: 93.3% 1111: 100% (maximum 3d effect) table 32 3d stereo enhancement function analogue outputs the wm8976 has three sets of stereo analogue outputs. these are: ? lout1 and rout1 which are normally used to drive a headphone load. ? lout2 and rout2 C normally used to drive an 8 ? btl speaker. ? out3 and out4 C can be configured as a stereo line out (out3 is left output and out4 is right output). out4 can also be used to provide a mono mix of left and right channels. lout2, rout2, out3 and out4 are supplied from spkvdd and are capable of driving up to 1.5vrms signals as shown in figure 19. lout1 and rout1 are supplied from avdd and can only drive out a 1v rms signal (avdd/3.3). lout1, rout1, lout2 and rout2 have individual analogue volume pgas with -57db to +6db ranges. there are four output mixers in the output signal path, the left and right channel mixers which control the signals to speaker, headphone (and optionally the line outputs) and also dedicated out3 and out4 mixers. left and right output channel mixers the left and right output channel mixers are shown in figure 19. these mixers allow the aux inputs, the adc bypass and the dac left and right channels to be combined as desired. this allows a mono mix of the dac channels to be done as well as mixing in external line-in from the aux or speech from the input bypass path. the aux and bypass inputs have individual volume control from -15db to +6db and the dac volume can be adjusted in the digital domain if required. the output of these mixers goes to both the headphone (lout1 and rout1) and speaker (lout2 and rout2) and can optionally go to the out3 and out4 mixers.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 44 figure 19 left/right output channel mixers
preliminary technical data wm8976 w ptd rev 2.0 february 2005 45 register address bit label default description 5 dacr2lmix 0 right dac output to left output mixer 0 = not selected 1 = selected r49 output mixer control 6 dacl2rmix 0 left dac output to right output mixer 0 = not selected 1 = selected 0 dacl2lmix 0 left dac output to left output mixer 0 = not selected 1 = selected 1 bypl2lmix 0 left bypass path (from the left channel input boost output) to left output mixer 0 = not selected 1 = selected 4:2 byplmixvol 000 left bypass volume contol to output channel mixer: 000 = -15db 001 = -12db 101 = 0db 110 = +3db 111 = +6db 5 auxl2lmix 0 left auxilliary input to left channel output mixer: 0 = not selected 1 = selected r50 left channel output mixer control 8:6 auxlmixvol 000 aux left channel input to left mixer volume control: 000 = -15db 001 = -12db 101 = 0db 110 = +3db 111 = +6db 0 dacr2rmix 0 right dac output to right output mixer 0 = not selected 1 = selected 5 auxr2rmix 0 right auxiliary input to right channel output mixer: 0 = not selected 1 = selected r51 right channel output mixer control 8:6 auxrmixvol 000 aux right channel input to right mixer volume control: 000 = -15db 001 = -12db 101 = 0db 110 = +3db 111 = +6db
wm8976 preliminary technical data w ptd rev 2.0 february 2005 46 2 lmixen 0 left output channel mixer enable: 0 = disabled 1= enabled r3 power management 3 3 rmixen 0 right output channel mixer enable: 0 = disabled 1 = enabled table 33 left and right output mixer control headphone outputs (lout1 and rout1) the headphone outputs, lout1 and rout1 can drive a 16 ? or 32 ? headphone load, either through dc blocking capacitors, or dc coupled without any capacitor. each headphone output has an analogue volume control pga with a gain range of -57db to +6db as shown in figure 22. figure 20 headphone outputs lout1 and rout1 headphone output using dc blocking capacitors: dc coupled headphone output: figure 21 recommended headphone output configurations when dc blocking capacitors are used, then their capacitance and the load resistance together determine the lower cut-off frequency, f c . increasing the capacitance lowers f c , improving the bass response. smaller capacitance values will diminish the bass response. assuming a 16 ? load and c1, c2 = 220 f: f c = 1 / 2 r l c 1 = 1 / (2 x 16 ? x 220 f) = 45 hz in the dc coupled configuration, the headphone ground is connected to the vmid pin. the vmid pin can be configured as a dc output driver by setting the out3mute register bit. the dc voltage on vmid in this configuration is equal to the dc offset on the lout1 and rout1 pins therefore no dc blocking capacitors are required. this saves space and material cost in portable applications.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 47 note that out3 and out4 have an optional output boost of 1.5x. when these are configured in this output boost mode (out3boost/out4boost=1) then the vmid value of these outputs will be equal to 1.5xavdd/2 and will not match the vmid of the headphone drivers. do not use the dc coupled output mode in this configuration. it is recommended to connect the dc coupled outputs only to headphones, and not to the line input of another device. although the built-in short circuit protection will prevent any damage to the headphone outputs, such a connection may be noisy, and may not function properly if the other device is grounded. speaker outputs (lout2 and rout2) the outputs lout2 and rout2 are designed to drive an 8 ? btl speaker but can optionally drive two headphone loads of 16 ? /32 ? or a line output (see headphone output and line output sections, respectively). each output has an individual volume control pga, an output boost/level shift bit, a mute and an enable as shown in figure 22. lout2 and rout2 output the left and right channel mixer outputs respectively. the rout2 signal path also has an optional invert. the amplifier used for this invert can be used to mix in the auxr signal with an adjustable gain range of -15db -> +6db. this allows a beep signal to be applied only to the speaker output without affecting the hp or line outputs. figure 22 speaker outputs lout2 and rout2
wm8976 preliminary technical data w ptd rev 2.0 february 2005 48 the signal to be output on lout2/rout2 comes from the left/right mixer circuits and can be any combination of the dac output, the bypass path (output of the input boost stage) and the aux input. the lout2/rout2 volume is controlled by the lout2vol/ rout2vol register bits. note that gains over 0db may cause clipping if the signal is large. the lout2mute/ rout2mute register bits cause the speaker outputs to be muted (the output dc level is driven out). the output pins remains at the same dc level (dcop), so that no click noise is produced when muting or un-muting the speaker output stages also have a selectable gain boost of 1.5x (3.52db). when this boost is enabled the output dc level is also level shifted (from avdd/2 to 1.5xavdd/2) to prevent the signal from clipping. a dedicated amplifier bufdcop, as shown in figure 23, is used to perform the dc level shift operation. this buffer must be enabled using the bufdcopen register bit for this operating mode. it should also be noted that if spkvdd is not equal to or greater than 1.5xavdd this boost mode may result in signals clipping. table 35 summarises the effect of the spkboost control bits. register address bit label default description r49 output control 2 spkboost 0 0 = spk gain = -1; dc = avdd / 2 1 = spk gain = +1.5; dc = 1.5 x avdd / 2 r1 power management 1 8 bufdcopen 0 dedicated buffer for dc level shifting output stages when in 1.5x gain boost configuration. 0=buffer disabled 1=buffer enabled (required for 1.5x gain boost) table 34 speaker boost stage control spkboost output stage gain output dc level output stage configuration 0 1x (0db) avdd/2 inverting 1 1.5x (3.52db) 1.5xavdd/2 non-inverting table 35 output boost stage details register address bit label default description 0 beepen 0 0 = mute auxr beep input 1 = enable auxr beep input 3:1 beepvol 000 auxr input to rout2 inverter gain 000 = -15db ... 111 = +6db 4 invrout2 0 inverts rout2 output 0 = not inverted 1 = inverted r43 beep control 5 muterpga2inv 0 mute input to invrout2 mixer table 36 auxr C rout2 beep mixer function
preliminary technical data wm8976 w ptd rev 2.0 february 2005 49 register address bit label default description 7 lout2zc 0 speaker volume zero cross enable: 1 = change gain on zero cross only 0 = change gain immediately 6 lout2mute 0 left speaker output mute: 0 = normal operation 1 = mute 5:0 lout2vol 111001 left speaker output volume: 000000 = -57db ... 111001 = 0db ... 111111 = +6db r54 lout2 (spk) volume control 8 spkvu not latched lout2 and rout2 volumes do not update until a 1 is written to spkvu (in reg 54 or 55) 7 rout2zc 0 speaker volume zero cross enable: 1 = change gain on zero cross only 0 = change gain immediately 6 rout2mute 0 right speaker output mute: 0 = normal operation 1 = mute 5:0 rout2vol 111001 right speaker output volume: 000000 = -57db ... 111001 = 0db ... 111111 = +6db r55 rout2 (spk) volume control 8 spkvu not latched lout2 and rout2 volumes do not update until a 1 is written to spkvu (in reg 54 or 55) table 37 speaker volume control
wm8976 preliminary technical data w ptd rev 2.0 february 2005 50 register address bit label default description 7 lout1zc 0 headphone volume zero cross enable: 1 = change gain on zero cross only 0 = change gain immediately 6 lout1mute 0 left headphone output mute: 0 = normal operation 1 = mute 5:0 lout1vol 111001 left headphone output volume: 000000 = -57db ... 111001 = 0db ... 111111 = +6db r52 lout1 volume control 8 hpvu not latched lout1 and rout1 volumes do not update until a 1 is written to hpvu (in reg 52 or 53) 7 rout1zc 0 headphone volume zero cross enable: 1 = change gain on zero cross only 0 = change gain immediately 6 rout1mute 0 right headphone output mute: 0 = normal operation 1 = mute 5:0 rout1vol 111001 right headphone output volume: 000000 = -57db ... 111001 = 0db ... 111111 = +6db r53 rout1 volume control 8 hpvu not latched lout1 and rout1 volumes do not update until a 1 is written to hpvu (in reg 52 or 53) table 38 out1 volume control zero cross timeout a zero-cross timeout function is also provided so that if zero cross is enabled on the input or output pgas the gain will automatically update after a timeout period if a zero cross has not occurred. this is enabled by setting slowclken. the timeout period is dependent on the clock input to the digital and is equal to 2 21 * input clock period. register address bit label default description r7 additional control 0 slowclken 0 slow clock enable. used for both the jack insert detect debounce circuit and the zero cross timeout. 0 = slow clock disabled 1 = slow clock enabled table 39 timeout clock enable control
preliminary technical data wm8976 w ptd rev 2.0 february 2005 51 out3/out4 mixers and output stages the out3/out4 pins can provide an additional stereo line output, a mono output, or a pseudo ground connection for headphones. there is a dedicated analogue mixer for out3 and one for out4 as shown in figure 24. the out3 and out4 output stages are powered from spkvdd and spkgnd. the indivi dually controllable outputs also incorporate an optional 1.5x boost and level shirting stage. figure 24 out3 and out4 mixers out3 can provide a buffered midrail headphone pseudo-ground, or a left line output. out4 can provide a buffered midrail headphone pseudo-ground, a right line output, or a mono mix output.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 52 register address bit label default description 6 out3mute 0 0 = output stage outputs out3 mixer 1 = output stage muted C drives out vmid. can be used as vmid buffer in this mode. 3 out4_2out3 0 out4 mixer output to out3 0 = disabled 1= enabled 2 bypl2out3 0 adc input to out3 0 = disabled 1= enabled 1 lmix2out3 0 left dac mixer to out3 0 = disabled 1= enabled r56 out3 mixer control 0 ldac2out3 1 left dac output to out3 0 = disabled 1= enabled 6 out4mute 0 0 = output stage outputs out4 mixer 1 = output stage muted C drives out vmid. can be used as vmid buffer in this mode. 5 halfsig 0 0 = out4 normal output 1 = out4 attenuated by 6db 4 lmix2out4 0 left dac mixer to out4 0 = disabled 1= enabled 3 ldac2out4 0 left dac to out4 0 = disabled 1= enabled 1 rmix2out4 0 right dac mixer to out4 0 = disabled 1= enabled r57 out4 mixer control 0 rdac2out4 1 right dac output to out4 0 = disabled 1= enabled table 40 out3/out4 mixer registers the out3 and out4 output stages each have a selectable gain boost of 1.5x (3.52db). when this boost is enabled the output dc level is also level shifted (from avdd/2 to 1.5xavdd/2) to prevent the signal from clipping. a dedicated amplifier bufdcop, as shown in figure 25, is used to perform the dc level shift operation. this buffer must be enabled using the bufdcopen register bit for this operating mode. it should also be noted that if spkvdd is not equal to or greater than 1.5xavdd this boost mode may result in signals clipping. table 35 summarises the effect o f the out3boost and out4boost control bits.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 53 figure 26 outputs out3 and out4 register address bit label default description 3 out3boost 0 0 = out3 output gain = -1; dc = avdd / 2 1 = out3 output gain = +1.5 dc = 1.5 x avdd / 2 r49 output control 4 out4boost 0 0 = out4 output gain = -1; dc = avdd / 2 1 = out4 output gain = +1.5 dc = 1.5 x avdd / 2 r1 power management 1 8 bufdcopen 0 dedicated buffer for dc level shifting output stages when in 1.5x gain boost configuration. 0=buffer disabled 1=buffer enabled (required for 1.5x gain boost) table 41 out3 and out4 boost stages control out3boost/ out4boost output stage gain output dc level output stage configuration 0 1x avdd/2 inverting 1 1.5x 1.5xavdd/2 non-inverting table 42 out3/out4 output boost stage details
wm8976 preliminary technical data w ptd rev 2.0 february 2005 54 enabling the outputs each analogue output of the wm8976 can be separately enabled or disabled. the analogue mixer associated with each output has a separate enable. all outputs are disabled by default. to save power, unused parts of the wm8976 should remain disabled. the sleep bit should only be set on to reduce residual device currents once all the other power management bits have been set to off. outputs can be enabled at any time, but it is not recommended to do so when bufio is disabled (bufioen=0) or when bufdcop is disabled (bufdcopen=0) when configured in output boost mode, as this may cause pop noise (see power management and applications information sections). register address bit label default description 2 bufioen 0 unused input/output tie off buffer enable 6 out3mixen 0 out3 mixer enable 7 out4mixen 0 out4 mixer enable r1 power management 1 8 bufdcopen 0 output stage 1.5xavdd/2 driver enable 6 sleep 0 0 = normal operation 1 = removes residual currents when device is in standby mode. 7 lout1en 0 lout1 output enable r2 power management 2 8 rout1en 0 rout1 output enable 2 lmixen 0 left mixer enable 3 rmixen 0 right mixer enable 5 spkpen 0 rout2 output enable 6 spknen 0 lout2 output enable 7 out3en 0 out3 enable r3 power management 3 8 out4en 0 out4 enable note: all enable bits are 1 = on, 0 = off table 43 output stages power management control thermal shutdown the speaker outputs can drive very large currents. to protect the wm8974 from overheating a thermal shutdown circuit is included . if the device temperature reaches approximately 125 0 c and the thermal shutdown circuit is enabled (tsden=1) then the speaker amplifiers will be disabled if tsden is set. the thermal shutdown may also be configured to generate an interrupt. see the gpio and interrupt controller section for details. register address bit label default description r49 output control 1 tsden 1 thermal shutdown enable 0 : thermal shutdown disabled 1 : thermal shutdown enabled table 44 thermal shutdown unused analogue inputs/outputs whenever an analogue input/output is disabled, it remains connected to a voltage source (either avdd/2 or 1.5xavdd/2 as appropriate) through a resistor. this helps to prevent pop noise when the output is re-enabled. the resistance between the voltage buffer and the output pins can be controlled using the vroi contol bit. the default impedance is low, so that any capacitors on the outputs can charge up quickly at start-up. if a high impedance is desired for disabled outputs, vroi can then be set to 1, increasing the resistance to about 30k ? . register address bit label default description r49 0 vroi 0 vref (avdd/2 or 1.5xavdd/2) to analogue output resistance 0: approx 1k ? 1: approx 30 k ? table 45 disabled outputs to vref resistance
preliminary technical data wm8976 w ptd rev 2.0 february 2005 55 a dedicated buffer is available for tying off unused analogue i/o pins as shown in figure 27. this buffer can be enabled using the bufioen register bit. if the spkboost, out3boost or out4boost bits are set then the relevant outputs will be tied to the output of the dc level shift buffer at 1.5xavdd/2 when disabled. figure 27 summarises the tie-off options for the speaker and mono output pins. figure 27 unused input/output pin tie-off buffers l/rout2en/ out3/4en out3boost/ out4boost/ spkboost vroi output configuration 0 0 0 1k ? tie-off to avdd/2 0 0 1 30k ? tie-off to avdd/2 0 1 0 1k ? tie-off to 1.5xavdd/2 0 1 1 30k ? tie-off to 1.5xavdd/2 1 0 x output enabled (dc level=avdd/2) 1 1 x output enabled (dc level=1.5xavdd/2) table 46 unused output pin tie-off options
wm8976 preliminary technical data w ptd rev 2.0 february 2005 56 digital audio interfaces the audio interface has four pins: ? adcdat: adc data output ? dacdat: dac data input ? lrc: data left/right alignment clock ? bclk: bit clock, for synchronisation the clock signals bclk, and lrc can be outputs when the wm8976 operates as a master, or inputs when it is a slave (see master and slave mode operation, below). five different audio data formats are supported: ? left justified ? right justified ? i 2 s ? dsp mode a ? dsp mode b all of these modes are msb first. they are described in audio data formats, below. refer to the electrical characteristic section for timing information. master and slave mode operation the wm8976 audio interface may be configured as either master or slave. as a master interface device the wm8976 generates bclk and lrc and thus controls sequencing of the data transfer on adcdat and dacdat. to set the device to master mode register bit ms should be set high. in slave mode (ms=0), the wm8976 responds with data to clocks it receives over the digital audio interfaces. audio data formats in left justified mode, the msb is available on the first rising edge of bclk following an lrc 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 lrc transition. figure 28 left justified audio interface (assuming n-bit word length) in right justified mode, the lsb is available on the last rising edge of bclk before a lrc 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 lrc transition.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 57 figure 29 right 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 lrc 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. figure 30 i 2 s audio interface (assuming n-bit word length) in dsp/pcm 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 lrp) following a rising edge of lrc. 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 31 and figure 32. in device slave mode, figure 33 and figure 34, 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 31 dsp/pcm mode audio interface (mode a, lrp=0, master)
wm8976 preliminary technical data w ptd rev 2.0 february 2005 58 figure 32 dsp/pcm mode audio interface (mode b, lrp=1, master) figure 33 dsp/pcm mode audio interface (mode a, lrp=0, slave) figure 34 dsp/pcm mode audio interface (mode b, lrp=0, slave)
preliminary technical data wm8976 w ptd rev 2.0 february 2005 59 register address bit label default description 1 adclrswap 0 controls whether adc data appears in right or left phases of lrc clock: 0=adc data appear in left phase of lrc 1=adc data appears in right phase of lrc 2 daclrswap 0 controls whether dac data appears in right or left phases of lrc clock: 0=left dac data appears in left phase of lrc and vice versa 1=left dac data appears in right phase of lrc and vice versa 4:3 fmt 10 audio interface data format select: 00=right justified 01=left justified 10=i 2 s format 11= dsp/pcm mode 6:5 wl 10 word length 00=16 bits 01=20 bits 10=24 bits 11=32 bits (see note) 7 lrcp lrc clock polarity 0=normal 1=inverted r4 audio interface control 8 bcp bclk polarity 0=normal 1=inverted table 47 audio interface control note: right justified mode will only operate with a maximum of 24 bits. audio interface control the register bits controlling audio format, word length and master / slave mode are summarised below. each audio interface can be controlled individually. register bit ms selects audio interface operation in master or slave mode. in master mode bclk, and lrc are outputs. the frequency of bclk and lrc in master mode are controlled with bclkdiv. these are divided down versions of master clock. this may result in short bclk pulses at the end of a lrc if there is a non-integer ratio of bclks to lrc clocks.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 60 register address bit label default description 0 ms 0 sets the chip to be master over lrc and bclk 0=bclk and lrc clock are inputs 1=bclk and lrc clock are outputs generated by the wm8976 (master) 4:2 bclkdiv 000 configures the bclk and lrc output frequency, for use when the chip is master over bclk. 000=divide by 1 (bclk=mclk) 001=divide by 2 (bclk=mclk/2) 010=divide by 4 011=divide by 8 100=divide by 16 101=divide by 32 110=reserved 111=reserved 7:5 mclkdiv 010 sets the scaling for either the mclk or pll clock output (under control of clksel) 000=divide by 1 001=divide by 1.5 010=divide by 2 011=divide by 3 100=divide by 4 101=divide by 6 110=divide by 8 111=divide by 12 r6 clock generation control 8 clksel 1 controls the source of the clock for all internal operation: 0=mclk 1=pll output table 48 clock control loopback setting the loopback register bit enables digital loopback. when this bit is set the output data from the adc audio interface is fed directly into the dac data input. companding the wm8976 supports a-law and -law companding on both transmit (adc) and receive (dac) sides. companding can be enabled on the dac or adc audio interfaces by writing the appropriate value to the dac_comp or adc_comp register bits respectively.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 61 register address bit label default description 0 loopback 0 digital loopback function 0=no loopback 1=loopback enabled, adc data output is fed directly into dac data input. 2:1 adc_comp 0 adc companding 00=off 01=reserved 10=-law 11=a-law 4:3 dac_comp 0 dac companding 00=off 01=reserved 10=-law 11=a-law r5 companding control 5 wl8 0 0 = off 1 = device operates in 8-bit mode table 49 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) } for x 1/a f(x) = ( 1 + lna|x|) / (1 + lna) } for 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. the input data range is separated into 8 levels, allowing low amplitude signals better precision than that of high amplitude signals. this is to exploit the operation of the human auditory system, where louder sounds do not require as much resolution as quieter sounds. the companded signal is an 8- bit word containing sign (1-bit), exponent (3-bits) and mantissa (4-bits). setting the wl8 register bit allows the device to operate with 8-bit data . in this m ode it is possible to use 8 bclks per lrc frame. when using dsp mode b, this allows 8-bit data words to be output consecutively every 8 bclks and can be used with 8-bit data words using the a-law and u-law companding functions. bit8 bit[7:4] bit[3:0] sign exponent mantissa table 50 8-bit companded word composition
wm8976 preliminary technical data w ptd rev 2.0 february 2005 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 35 u-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 36 a-law companding
preliminary technical data wm8976 w ptd rev 2.0 february 2005 63 audio sample rates the wm8976 sample rates for the adc and the dacs are set using the sr register bits. the cutoffs for the digital filters and the alc attack/decay times stated are determined using these values and assume a 256fs master clock rate. if a sample rate that is not explicitly supported by the sr register settings is required then the closest sr value to that sample rate should be chosen, the filter characteristics and the alc attack, decay and hold times will scale appropriately. register address bit label default description r7 additional control 3:1 sr 000 approximate sample rate (configures the coefficients for the internal digital filters): 000=48khz 001=32khz 010=24khz 011=16khz 100=12khz 101=8khz 110-111=reserved table 51 sample rate control master clock and phase locked loop (pll) the wm8976 has an on-chip phase-locked loop (pll) circuit that can be used to: generate master clocks for the wm8976 audio functions from another external clock, e.g. in telecoms applications. generate and output (on pin csb/gpio1) a clock for another part of the system that is derived from an existing audio master clock. figure 37 shows the pll and internal clocking arrangment on the wm8976. the pll can be enabled or disabled by the pllen register bit. register address bit label default description r1 power management 1 5 pllen 0 pll enable 0=pll off 1=pll on table 52 pllen control bit
wm8976 preliminary technical data w ptd rev 2.0 february 2005 64 figure 37 pll and clock select circuit the pll frequency ratio r = f 2 /f 1 (see figure 37) can be set using the register bits pllk and plln: plln = int r pllk = int (2 24 (r-plln)) example: mclk=12mhz, required clock = 12.288mhz. r should be chosen to ensure 5 < plln < 13. there is a fixed divide by 4 in the pll and a selectable divide by n after the pll which should be set to divide by 2 to meet this requirement. enabling the divide by 2 sets the required f 2 = 4 x 2 x 12.288mhz = 98.304mhz. r = 98.304 / 12 = 8.192 plln = int r = 8 k = int ( 2 24 x (8.192 C 8)) = 3221225 = 3126e9h register address bit label default description 4 pllprescale 0 divide mclk by 2 before input to pll r36 pll n value 3:0 plln 1000 integer (n) part of pll input/output frequency ratio. use values greater than 5 and less than 13. r37 pll k value 1 5:0 pllk [23:18] 0ch r38 pll k value 2 8:0 pllk [17:9] 093h r39 pll k value 3 8:0 pllk [8:0] 0e9h fractional (k) part of pll1 input/output frequency ratio (treat as one 24-digit binary number). table 53 pll frequency ratio control the pll performs best when f 2 is around 90mhz. its stability peaks at n=8. some example settings are shown in table 54.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 65 mclk (mhz) (f1) desired output (mhz) f2 (mhz) prescale divide postscale divide r n (hex) k (hex) 12 11.29 90.3168 1 2 7.5264 7 86c227 12 12.288 98.304 1 2 8.192 8 3126e9 13 11.29 90.3168 1 2 6.947446 6 f28bd5 13 12.288 98.304 1 2 7.561846 7 8fd526 14.4 11.29 90.3168 1 2 6.272 6 45a1cb 14.4 12.288 98.304 1 2 6.826667 6 d3a06d 19.2 11.29 90.3168 2 2 9.408 9 6872b0 19.2 12.288 98.304 2 2 10.24 a 3d70a4 19.68 11.29 90.3168 2 2 9.178537 9 2db493 19.68 12.288 98.304 2 2 9.990243 9 fd80a0 19.8 11.29 90.3168 2 2 9.122909 9 1f76f8 19.8 12.288 98.304 2 2 9.929697 9 ee009f 24 11.29 90.3168 2 2 7.5264 7 86c227 24 12.288 98.304 2 2 8.192 8 3126e9 26 11.29 90.3168 2 2 6.947446 6 f28bd5 26 12.288 98.304 2 2 7.561846 7 8fd526 27 11.29 90.3168 2 2 6.690133 6 boac94 27 12.288 98.304 2 2 7.281778 7 482297 table 54 pll frequency examples general purpose input/output the wm8976 has two dual purpose input/output pins. ? csb/gpio1: csb / gpio pin ? l2/gpio2: left channel line input / headphone detection input the gpio2 function is provided for use as a jack detection input. the gpio1 and gpio4 functions are provided for use as jack detection inputs or general purpose outputs. the default configuration for the csb/gpio1 pin is to be inputs. when setup as an input, the csb/gpio1 pin can either be used as csb or for jack detection, depending on how the mode pin is set. table 48 illustrates the functionality of the gpio1 pin when used as general purpose outputs.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 66 register address bit label default description 2:0 gpio1sel 000 csb/gpio1 pin function select: 000= input (csb/jack detection: depending on mode setting) 001= reserved 010=temp ok 011=amute active 100=pll clk o/p 101=pll lock 110=logic 1 111=logic 0 3 gpio1pol 0 gpio1 polarity invert 0=non inverted 1=inverted r8 gpio control 5:4 opclkdiv 00 pll output clock division ratio 00=divide by 1 01=divide by 2 10=divide by 3 11=divide by 4 table 55 csb/gpio control note: if mode is set to 3 wire mode, csb/gpio1 shall be used as csb input irrespective of the gpio1sel[2:] bits. note that slowclken must be enabled when using the jack detect function. for further details of the jack detect operation see the output switching section. output switching (jack detect) when the device is configured with a 2-wire interface the csb/gpio1 pin can be used as a switch control input to automatically disable one set of outputs and enable another. the l2/gpio2 pin can also be used for this purpose. for example, when a headphone is plugged into a jack socket then it may be desirable to disable the speaker (e.g. when one of the gpio pins is connected to a mechanical switch in the headphone socket to detect plug-in). the gpio pins have an internal de-bounce circuit when in this mode in order to prevent the output enables from toggling multiple times due to input glitches. this de-bounce circuit is clocked from a slow clock with period 2 21 x mclk. note that slowclken must be enabled when using the jack detect function. note that the gpiopol bits are not relevant for jack detection, it is the signal detected at the pin which is used. the switching on/off of the outputs is fully configurable by the user. each output, out1, out2, out3 and out4 has 2 enables associated with it. out1_en_0, out2_en_0, out3_en_0 and out4_en_0 are the output enable signal which is used if the selected jack detection pin is at logic 0 (after de-bounce). out1_en_1, out2_en_1, out3_en_1 and out4_en_1 are the output enable signal which is used if the selected jack detection pin is at logic 1 (after de-bounce). similar to the output enables, vmid, which can be driven out of out3 can be configured to be on/off depending on the jack detection input polarity using the vmid_en_0 and vmid_en_1 bits. the jack detection enables work as follows: all out_en signals have an and function performed with their normal enable signals (in table 37). thus for an normal output enable which is on (1), if the selected jack detection enable (controlled by selected jack detection pin polarity) is a 0 it shall turn the output off. if the normal enable signal is already off (0), the jack detection signal shall have no effect due to the and function. during jack detection if the user desires an output to be un-changed whether the jack is in or not, both the jd_en settings i.e. out1_en0 and out1_en1, should be set to 1.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 67 the vmid_en signal has an or function performed with the normal vmid driver enable. if the vmid_en signal is to have no effect to normal functi onality when jack detection is enabled, it should set to 0 for both enables. if jack detection is not enabled, the output enables default to all 1s, allowing the outputs to be controlled as normal via the normal output enables found in table 37, similarly the vmid_en signal defaults to 0 allowing the vmid driver to be controlled via the normal enable bit. register address bit label default description 5:4 jd_sel 00 pin selected as jack detection input 00 = gpio1 01 = gpio2 10 = reserved 11 = reserved 6 jd_en 0 jack detection enable 0=disabled 1=enabled r9 gpio control 8:7 jd_vmid 00 [7] vmid_en_0 [8] vmid_en_1 3:0 jd_en0 0 output enables when selected jack detection input is logic 0. [0]= out1_en_0 [1]= out2_en_0 [2]= out3_en_0 [3]= out4_en_0 r13 7:4 jd_en1 0 output enables when selected jack detection input is logic 1 [4]= out1_en_1 [5]= out2_en_1 [6]= out3_en_1 [7]= out4_en_1 table 56 jack detect register control bits
wm8976 preliminary technical data w ptd rev 2.0 february 2005 68 control interface selection of control mode and 2-wire mode address the control interface can operate as either a 3-wire or 2-wire mpu interface. the mode pin determines the 2 or 3 wire mode as shown in table 57. the wm8976 is controlled by writing to registers through a serial control interface. a control word consists of 16 bits. the first 7 bits (b15 to b9) are address bits that select which control register is accessed. the remaining 9 bits (b8 to b0) are register bits, corresponding to the 9 bits in each control register. mode interface format low 2 wire high 3 wire table 57 control interface mode selection 3-wire serial control mode in 3-wire mode, every rising edge of sclk clocks in one data bit from the sdin pin. a rising edge on csb/gpio latches in a complete control word consisting of the last 16 bits. figure 38 3-wire serial control interface 2-wire serial control mode the wm8976 supports software control via a 2-wire serial bus. many devices can be controlled by the same bus, and each device has a unique 7-bit device address (this is not the same as the 7-bit address of each register in the wm8976). the wm8976 operates as a 2-wire slave device only. the controller indicates the start of data transfer with a high to low transition on sdin while sclk remains high. this indicates that a device address and data will follow. all devices on the 2-wire bus respond to the start condition and shift in the next eight bits on sdin (7-bit address + read/write bit, msb first). if the device address received matches the address of the wm8976, then the wm8976 responds by pulling sdin low on the next clock pulse (ack). if the address is not recognised or the r/w bit is 1 when operating in write only mode, the wm8976 returns to the idle condition and wait for a new start condition and valid address. during a write, once the wm8976 has acknowledged a correct address , the controller sends the first byte of control data (b15 to b8, i.e. the wm8976 register address plus the first bit of register data). the wm8976 then acknowledges the first data byte by pulling sdin low for one clock pulse. the controller then sends the second byte of control data (b7 to b0, i.e. the remaining 8 bits of register data), and the wm8976 acknowledges again by pulling sdin low. transfers are complete when there is a low to high transition on sdin while sclk is high. after a complete sequence the wm8976 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. sdin changes while sclk is high), the device jumps to the idle condition.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 69 sdin sclk register address and 1st register data bit device address (7 bits) rd / wr bit ack (low) control byte 1 (bits 15 to 8) control byte 1 (bits 7 to 0) remaining 8 bits of register data stop start ack (low) ack (low) figure 39 2-wire serial control interface in 2-wire mode the wm8976 has a fixed device address, 0011010. resetting the chip the wm8976 can be reset by performing a write of any value to the software reset register (address 0 hex). this will cause all register values to be reset to their default values. in addition to this there is a power-on reset (por) circuit which ensures that the registers are set to default when the device is powered up. power supplies the wm8976 can use up to four separate power supplies: avdd and agnd: analogue supply, powers all analogue functions except the speaker output and mono output drivers. avdd can range from 2.5v to 3.6v and has the most significant impact on overall power consumption (except for power consumed in the headphone). a large avdd slightly improves audio quality. spkvdd and spkgnd: headphone and speaker supplies, power the speaker and mono output drivers. spkvdd can range from 2.5v to 5v. spkvdd can be tied to avdd, but it r equires separate layout and decoupling capacitors to curb harmonic distortion. with a larger spkvdd, louder headphone and speaker outputs can be achieved with lower distortion. if spkvdd is lower than avdd, the output signal may be clipped. dcvdd: digital core supply, powers all digital functions except the audio and control interfaces. dcvdd can range from 1.8v to 3.6v, and has no effect on audio quality. the return path for dcvdd is dgnd, which is shared with dbvdd. dbvdd can range from 1.8v to 3.6v. dbvdd return path is through dgnd. it is possible to use the same supply voltage for all four supplies. however, digital and analogue supplies should be routed and decoupled separately on the pcb to keep digital switching noise out of the analogue signal paths. recommended power up/down seqence in order to minimise output pop and click noise it is recommended that the device is powered up in a controlled sequence. in addition to this it is recommended that the zero cross functions are used when changing the volume in the pgas.
wm8976 preliminary technical data w ptd rev 2.0 february 2005 70 power management saving power by reducing oversampling rate the default mode of operation of the adc and dac digital filters is in 64x oversampling mode. under the control of adcosr and dacosr the oversampling rate may be doubled. 64x oversampling results in a slight decrease in noise performance compared to 128x but lowers the power consumption of the device. register address bit label default description r10 dac control 3 dacosr128 0 dac oversample rate select 0 = 64x (lowest power) 1 = 128x (best snr) r14 adc control 3 adcosr128 0 adc oversample rate select 0 = 64x (lowest power) 1 = 128x (best snr) table 58 adc and dac oversampling rate selection vmid the analogue will not work unless vmid is enabled (vmidsel 00). the impedance of the vmid resistor string, together with the decoupling capacitor on the vmid pin will determine the startup time of the vmid circuit. register address bit label default description r1 power management 1 1:0 vmidsel 00 reference string impedance to vmid pin (detemines startup time): 00=off (open circuit) 01=75k ? 10=300k ? 11=5k ? (for fastest startup) table 59 vmid impedance control biasen the analogue amplifiers will not operate unless bi asen is enabled. register address bit label default description r1 power management 1 3 biasen 0 analogue amplifier bias control table 60 analogue bias control estimated supply currents when either the dacs or adc are enabled it is estimated that approximately 4ma will be drawn from dcvdd when fs=48khz (this will be lower at lower sample rates). when the pll is enabled an additional 700 microamps will be drawn from dcvdd.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 71 table 61 shows the estimated 3.3v avdd current drawn by various circuits, by register bit. register bit avdd current (milliamps) bufdcopen 0.1 out4mixen 0.2 out3mixen 0.2 pllen 1.4 (with clocks applied) micben 0.5 biasen 0.3 bufioen 0.1 vmidsel 10k=>0.3, less than 0.1 for 100k/500k rout1en 0.4 lout1en 0.4 boostenl 0.2 inpgaenl 0.2 adcenl x64 (adcosr=0)=>2.6, x128 (adcosr=1)=>4.9 out4en 0.2 out3en 0.2 spknen 1ma from spkvdd + 0.2ma from avdd in 5v mode spkpen 1ma from spkvdd + 0.2ma from avdd in 5v mode rmixen 0.2 lmixen 0.2 dacenr x64 (dacosr=0)=>1.8, x128(dacosr=1)=>1.9 dacenl x64 (dacosr=0)=>1.8, x128(dacosr=1)=>1.9 table 61 avdd supply current
wm8976 preliminary technical data w ptd rev 2.0 february 2005 72 register map addr b[15:9] deft val dec hex register name b8 b7 b6 b5 b4 b3 b2 b1 b0 (hex) 0 00 software reset software reset 1 01 power manage?t 1 bufdcop en out4mix en out3mix en pllen micben biasen bufioen vmidsel 000 2 02 power manage?t 2 rout1en lout1en sleep 0 boostenl 0 nppgaenl 0 adcenl 000 3 03 power manage?t 3 out4en out3en spknen spkpen 0 rmixen lmixen dacenr dacenl 000 4 04 audio interface bcp lrcp wl fmt dlrswap alrswap mono 050 5 05 companding ctrl 0 0 0 wl8 dac_comp adc_comp loopback 000 6 06 clock gen ctrl clksel mclkdiv bclkdiv 0 ms 140 7 07 additional ctrl 0 0 0 0 0 sr lowclken 000 8 08 gpio stuff 0 0 0 opclkdiv gpio1pol gpio1sel[2:0] 000 9 09 jack detect control jd_vmid1 jd_vmid0 jd_en jd_sel 0 000 000 10 0a dac control 0 0 soft mute 0 0 dacosr 128 amute dacrpol daclpol 000 11 0b left dac digital vol dacvu dacvoll 0ff 12 0c right dac dig?l vol dacvu dacvolr 0ff 13 0d jack detect control jd_en1 jd_en0 000 14 0e adc control hpfen hpfapp hpfcut adcosr 128 0 0 adclpol 100 15 0f adc digital vol adcvu adcvol 0ff 18 12 eq1 ? low shelf eq3dmode 0 eq1c eq1g 12c 19 13 eq2 ? peak 1 eq2bw 0 eq2c eq2g 02c 20 14 eq3 ? peak 2 eq3bw 0 eq3c eq3g 02c 21 15 eq4 ? peak 3 eq4bw 0 eq4c eq4g 02c 22 16 eq5 ? high shelf 0 0 eq5c eq5g 02c 24 18 dac limiter 1 limen limdcy limatk 032 25 19 dac limiter 2 0 0 limlvl limboost 000 27 1b notch filter 1 nfu nfen nfa0[13:7] 000 28 1c notch filter 2 nfu 0 nfa0[6:0] 000 29 1d notch filter 3 nfu 0 nfa1[13:7] 000 30 1e notch filter 4 nfu 0 nfa1[6:0] 000 32 20 alc control 1 alcsel 0 alcmax alcmin 038 33 21 alc control 2 alczc alchld alclvl 00b 34 22 alc control 3 alcmode alcdcy alcatk 032 35 23 noise gate 0 0 0 0 0 ngen ngth 000 36 24 pll n 0 0 0 0 pll_pre scale plln[3:0] 008 37 25 pll k 1 0 0 0 pllk[23:18] 00c 38 26 pll k 2 pllk[17:9] 093 39 27 pll k 3 pllk[8:0] 0e9 41 29 3d control depth3d 000 43 2b beep control invrout2 beepvol beepen 000 44 2c input ctrl micbvsel 0 0 0 l2_2 inppga lin2 inppga lip2 inppga 003 45 2d inp pga gain ctrl inpga update inppgazc inppga mute inppgavol 010 47 2f adc boost ctrl pgaboost 0 l2_2boostvol 0 auxl2boostvol 100 49 31 output ctrl dacl2 rmix dacr2 lmix out4 boost out3 boost spkr boost tsden vroi 002 50 32 left mixer ctrl auxlmixvol a uxl2lmix byplmixvol bypl2lmix acl2lmix 001
preliminary technical data wm8976 w ptd rev 2.0 february 2005 73 51 33 right mixer ctrl auxrmixvol uxr2rmix 0 0 acr2rmix 001 52 34 lout1 (hp) volume ctrl update lout1zc lout1 mute lout1vol 039 53 35 rout1 (hp) volume ctrl update rout1zc rout1 mute rout1vol 039 54 36 lout2 (spk) volume ctrl update lout2zc lout2 mute lout2vol 039 55 37 rout2 (spk) volume ctrl update rout2zc rout2 mute rout2vol 039 56 38 out3 mixer ctrl 0 0 out3 mute 0 0 out4_ 2out3 bypl2 out3 lmix2 out3 ldac2 out3 001 57 39 out4 (mono) mixer ctrl 0 0 out4 mute halfsig lmix2 out4 ldac2 out4 0 rmix2 out4 rdac2 out4 001 table 62 wm8976 register map
wm8976 preliminary technical data w ptd rev 2.0 february 2005 74 digital filter characteristics parameter test conditions min typ max unit adc filter +/- 0.025db 0 0.454fs passband -6db 0.5fs passband ripple +/- 0.025 db stopband 0.546fs stopband attenuation f > 0.546fs -60 db group delay 21/fs adc high pass filter -3db 3.7 -0.5db 10.4 high pass filter corner frequency -0.1db 21.6 hz dac filter +/- 0.035db 0 0.454fs passband -6db 0.5fs passband ripple +/-0.035 db stopband 0.546fs stopband attenuation f > 0.546fs -80 db group delay 29/fs table 63 digital filter characteristics terminology 1. stop band attenuation (db) C the degree to which the frequency spectrum is attenuated (outside audio band) 2. pass-band ripple C any variation of the frequency response in the pass-band region
preliminary technical data wm8976 w ptd rev 2.0 february 2005 75 dac filter responses -120 -100 -80 -60 -40 -20 0 00.511.522.53 frequency (fs) response (db) -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0 0.1 0.2 0.3 0.4 0.5 frequency (fs) response (db) figure 40 dac digital filter frequency response figure 41 dac digital filter ripple adc filter responses -120 -100 -80 -60 -40 -20 0 0 0.5 1 1.5 2 2.5 3 frequency (fs) response (db) -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0 0.1 0.2 0.3 0.4 0.5 frequency (fs) response (db) figure 42 adc digital filter frequency response figure 43 adc digital filter ripple
wm8976 preliminary technical data w ptd rev 2.0 february 2005 76 highpass filter the wm8976 has a selectable digital highpass filter in the adc filter path. this filter has two modes, audio and applications. in audio mode the filter is a 1 st order iir with a cut-off of around 3.7hz. in applications mode the filter is a 2 nd order high pass filter with a selectable cut-off frequency. -40 -35 -30 -25 -20 -15 -10 -5 0 5 0 5 10 15 20 25 30 35 40 45 frequency (hz) response (db) figure 44 adc highpass filter response, hpfapp=0 -60 -50 -40 -30 -20 -10 0 10 0 200 400 600 800 1000 1200 frequency (hz) response (db) -80 -70 -60 -50 -40 -30 -20 -10 0 10 0 200 400 600 800 1000 1200 frequency (hz) response (db) figure 45 adc highpass filter responses (48khz), hpfapp=1, all cut-off settings shown. figure 46 adc highpass filter responses (24khz), hpfapp=1, all cut-off settings shown. -90 -80 -70 -60 -50 -40 -30 -20 -10 0 10 0 200 400 600 800 1000 1200 frequency (hz) response (db) figure 47 adc highpass filter responses (12khz), hpfapp=1, all cut-off settings shown.
preliminary technical data wm8976 w ptd rev 2.0 february 2005 77 5-band equaliser the wm8976 has a 5-band equaliser which can be applied to either the adc path or the dac path. the plots from figure 48 to figure 61 show the frequency responses of each filter with a sampling frequency of 48khz, firstly showing the different cut-off/centre frequencies with a gain of 12db, and secondly a sweep of the gain from -12db to +12db for the lowest cut-off/centre frequency of each filter. 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) figure 48 eq band 1 low frequency shelf filter cut-offs figure 49 eq band 1 gains for lowest cut-off frequency 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) figure 50 eq band 2 C peak filter centre frequencies, eq2bw=0 figure 51 eq band 2 C peak filter gains for lowest cut-off frequency, eq2bw=0 10 -2 10 -1 10 0 10 1 10 2 10 3 10 4 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) figure 52 eq band 2 C eq2bw=0, eq2bw=1
wm8976 preliminary technical data w ptd rev 2.0 february 2005 78 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) figure 53 eq band 3 C peak filter centre frequencies, eq3 b figure 54 eq band 3 C peak filter gains for lowest cut-off frequency, eq3bw=0 10 -2 10 -1 10 0 10 1 10 2 10 3 10 4 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) figure 55 eq band 3 C eq3bw=0, eq3bw=1
preliminary technical data wm8976 w ptd rev 2.0 february 2005 79 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) figure 56 eq band 4 C peak filter centre frequencies, eq3 b figure 57 eq band 4 C peak filter gains for lowest cut-off frequency, eq4bw=0 10 -2 10 -1 10 0 10 1 10 2 10 3 10 4 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) figure 58 eq band 4 C eq3bw=0, eq3bw=1 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 frequency (hz) magnitude (db) figure 59 eq band 5 high frequency shelf filter cut-offs figure 60 eq band 5 gains for lowest cut-off frequency
wm8976 preliminary technical data w ptd rev 2.0 february 2005 80 figure 61 shows the result of having the gain set on more than one channel simultaneously. the blue traces show each band (lowest cut-off/centre frequency) with 12db gain. the red traces show the cumulative effect of all bands with +12db gain and all bands -12db gain, with eqxbw=0 for the peak filters. 10 -1 10 0 10 1 10 2 10 3 10 4 10 5 -15 -10 -5 0 5 10 15 20 frequency (hz) magnitude (db) figure 61 cumulative frequency boost/cut
preliminary technical data wm8976 w ptd rev 2.0 february 2005 81 applications information recommended external components figure 62 external component diagram
wm8976 preliminary technical data w ptd rev 2.0 february 2005 82 package diagram dm030.d fl: 32 pin qfn plastic package 5 x 5 x 0.9 mm body, 0.50 mm lead pitch notes: 1. dimension b applied to metallized terminal and is measured between 0.25 mm and 0.30 mm from terminal tip. dimension l1 repr esents terminal pull back from package side wall. maximum of 0.1mm is acceptable. where terminal pull back exists, only upper half of lead is visible on package side wall due to half etching of leadframe. 2. falls within jedec, mo-220 with the exception of d2, e2: d2,e2: larger pad size chosen which is just outside jedec specification 3. all dimensions are in millimetres 4. this drawing is subject to change without notice. 5. shape and size of corner tie bar may vary with package terminal count. corner tie bar is connected to exposed pad internall y see detail b e2 e2/2 b b 16 15 a 8 9 e 5 corner tie bar c 0.08 c ccc a a1 c (a3) seating plane 1 b d2 l d2/2 see detail a index area (d/2 x e/2) top view d c aaa 2 x c aaa 2 x e detail b terminal tip r datum e e/2 l1 1 detail a b c bbb m a 32x b l 32x k l1 r 1 1 0 . 5 6 6 m m 0 . 4 3 m m 5 corner tie bar symbols dimensions (mm) min nom max note a a1 a3 b d d2 e e2 e l l1 r 0.85 0.90 1.00 0.05 0.02 0 0.2 ref 0.30 0.23 0.18 5.00 3.4 3.3 3.2 0.5 bsc 0.35 0.4 0.45 0.1 b(min)/2 1 2 2 1 k 0.20 aaa bbb ccc ref: 0.15 0.10 0.10 jedec, mo-220, variation vhhd-2 tolerances of form and position 4.90 5.10 5.00 4.90 5.10 3.4 3.3 3.2 exposed centre pad 1 17 24 25 32
preliminary technical data wm8976 w ptd rev 2.0 february 2005 83 important notice wolfson microelectronics plc (wm) reserve the right to make changes to their products or to discontinue any product or service without notice, and advise customers to obtain the latest version of relevant information to verify, before placing orders, that information being relied on is current. all products are sold subject to the wm terms and conditions of sale supplied at the time of order acknowledgement, including those pertaining to warranty, patent infringement, and limitation of liability. wm warrants performance of its products to the specifications applicable at the time of sale in accordance with wms standard warranty. testing and other quality control techniques are utilised to the extent wm deems necessary to support this warranty. specific testing of all parameters of each device is not necessarily performed, except those mandated by government requirements. in order to minimise risks associated with customer applications, adequate design and operating safeguards must be used by the customer to minimise inherent or procedural hazards. wolfson products are not authorised for use as critical components in life support devices or systems without the express written approval of an officer of the company. life support devices or systems are devices or systems that are intended for surgical implant into the body, or support or sustain life, and whose failure to perform when properly used in accordance with instructions for use provided, can be reasonably expected to result in a significant injury to the user. a critical component is any component of a life support device or system whose failure to perform can be reasonably expected to cause the failure of the life support device or system, or to affect its safety or effectiveness. wm assumes no liability for applications assistance or customer product design. wm does not warrant or represent that any license, either express or implied, is granted under any patent right, copyright, mask work right, or other intellectual property right of wm covering or relating to any combination, machine, or process in which such products or services might be or are used. wms publication of information regarding any third partys products or services does not constitute wms approval, license, warranty or endorsement thereof. reproduction of information from the wm web site or datasheets is permissible only if reproduction is without alteration and is accompanied by all associated warranties, conditions, limitations and notices. representation or reproduction of this information with alteration voids all warranties provided for an associated wm product or service, is an unfair and deceptive business practice, and wm is not responsible nor liable for any such use. resale of wms products or services with statements different from or beyond the parameters stated by wm for that product or service voids all express and any implied warranties for the associated wm product or service, is an unfair and deceptive business practice, and wm is not responsible nor liable for any such use. address: wolfson microelectronics plc westfield house 26 westfield road edinburgh eh11 2qb united kingdom tel :: +44 (0)131 272 7000 fax :: +44 (0)131 272 7001 email :: sales@wolfsonmicro.com

▲Up To Search▲

Price & Availability of WM8976GEFLRV

	To Download WM8976GEFLRV Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .