• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 461 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AIC3204 Pop on Head Phone Amp Power Up

Bill Finger
Genius 3735 points

I am experiencing a loud pop when the AIC3204 codec on our electronics board powers up.  This happens despite use of the slow head phone power up feature.  What seems to happen is that the codec will slowly bring the DC blocking cap up to voltage, but will lag the target output voltage by some millivolts.  Finally, when the slow charging feature times out, the output recovers those last few millivolts suddenly, producing a loud pop.  I am using Sennheiser HDA 200 headphones, so the sound is overwhelming.

 Is there any workaround for this?

I can take a photo if anyone is interested in seeing the effect. 

  • 0
    TI__Mastermind 46195 points

    Hi Bill,

    I am assume you are using the HP outputs.

    P1_R20_D1-0 controls the resistor (Rp) used in the charge phase of the HP outputs. At the instant the HP outs are powered, you will have a resistor divider formed by Rp and the headphone load (RL) driven by VCM. In your case, your headphones are 40-ohms. The larger Rp is, the less this initial pop will be. After this initial pop, the series capacitor (Cs) will charge to VCM based on the time constant (Rp*Cs). The bigger Rp, the longer the required charge time.

    P1_R20_D5-2 controls a timer that switches from charging mode to playback mode. This timer uses the Rp value as an input and assumes its time constant based on 47uF. So if you program Rp = 25k, a time constant of 5, the mode will change from charging to playback in (25K*47u) seconds. 5 time constants is a good choice, because the voltage at the cap is close its target VCM value. However, if you change your physical cap value from 47uF to say, 200uF, then the time will expire much sooner because it will take much longer to the cap to charge and the timer assumes 47uF. In this case you would need a setting 5*200u/47u = 21 time constants to compensate. In such case, you cannot use 25k, you need Rp smaller (6k or 2k).

    To see this on a scope, play a sinewave (50mVrms or so) and load your configuration. You will see the sinewave playing when the timer expires. If it expires too early, then increase the timer value (P1_R20_D5-2).

    Regards,

    J-

  • 0 in reply to J-
    Genius 3735 points

    J,

    Thank you for your rapid response. Please see the attached photo. Trace 1 is the AIC3204 side of the headphone decoupling cap. As you can see, it asymptotically approaches the final value, minus an offset. Then, when charging mode is complete, the voltage sharply increases to the target voltage. Increasing the time constant does not significantly alter the offset voltage.

    Regards,

    Bill

     

  • 0 in reply to Bill Finger
    TI__Mastermind 46195 points

    Hi Bill,

    Here is a plot I took a while ago. The wider portion of a trace is a sinetone.

    Can you try 0.9V common mode? From your plot it looks as if you are using 1.65V or so. Also, are you using the AIC3254EVM-K or your system?

    Regards,

    J-

  • 0 in reply to J-
    Genius 3735 points

    I agree that your plot is nicer than my plot. 

     I have tried the 0.9V common mode, and it did not affect the result.

    We are using a custom board with the AIC3204, not the AIC3254.