• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 29 subscribers
  • Views 1600 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.

DRV632: Pops with VDD removal

radar
Expert 3550 points
Part Number: DRV632
Other Parts Discussed in Thread: PCM1681, , PCM1608

 

I have a customer situation that has popped up (bad pun) and could use an assist.  The system uses a PCM1608 and the DRV632; and we are having issues with a ‘pop’ when power is removed unceremoniously; not common but still possible in our system.  Basically the PCM1681 DAC generates a pop because the output is biased to mid-rail; this is common to most DACs I think. The DRV632 copes with this if it is muted but does pass a pop through as the rail collapses.

So we do have a voltage monitor on the supply line for the DRV32; it has a #FLAG open drain output that senses the battery voltage collapse and triggers the #MUTE on the DRV632; but this is ineffective if the DRV632 rail also collapses.  Does the UVP function operate differently than the MUTE function as it relates to a low voltage/shutdown operation on the DRV632?  The datasheet isn’t super clear as to whether this UVP function is there for eliminating pops, limiting current in a low VDD situation, or anything.

Any ideas welcome!

  • 0
    TI__Guru** 109395 points
    Hi Radar,

    Actually UVP feature may be useful here. There is some information in data sheet at the bottom of page 10.
    The UVP will disable the driver if the reference voltage is below the threshold, which you will set with a resistor divider. It's main purpose is to monitor supply rail and disable the device before a pop comes in from previous stage.

    Best regards,
    -Ivan Salazar
    Audio Applications Engineer - Low Power Audio & Actuators
  • 0 in reply to Ivan Salazar
    TI__Expert 3550 points

    Ivan,

    Thanks- gets a little more confusion as I dig in, but bear with me; we'll get there.  I had asked for a few plots just to get a better handle on this; and this is what it's looking like thus far, WITHOUT UVP used (this was the unfortunate part; when they took the plot they used a fresh board without the UVP mod... THe customer REPORTED that when they made a UVP mod it "didn't fix the issue"; but in the absence of telling me any resistor values or a schematic I don't truly know what that meant.  SO I have asked for this to be done and the plot recaptured so I can see the before/after to see if UVP truly does anything; but the plot without UVP implemented raises a little suspicion all by itself, so in the interim while i get the second plot, let me bounce this off of you.

    So channel 1 is 3.3V to the DRV632, its VDD supply; CH2 is /MUTE, CH3 is the input to the DRV632, and CH4 the DRV632 output.

    So what we see is that /MUTE DOES assert before the VDD drops; so we were sort of thinking that would prevent the DRVoutput from collapsing.  Now maybe this is all moot because VDD is going away- if you look closely the purple trace; the input; does start to fade and the DRV does NOT drop until VDD drops; suggesting MUTE is in fact doing something; but maybe just can't protect against VDD loss.  That might all be fair to your points on trying to use the UVP function, so when we get that plot it maybe will tell us something; but it's at least been communicated to/through me that it didn't fix the issue.  I just wanted to plot in front of you in case it sparked some other thoughts.

  • 0 in reply to radar
    TI__Expert 3550 points

    Okay, still having issues.  A divider on UVP isn't necessarily helping, insomuch as it drops pretty quickly when VDD goes away, so as a test I pulled UVP low (aka tweezer test) before pulling power; under the premise that if we could get it to work we could figure out a way with some caps and a FET or something to create a circuit to get the order right; but UVP doesn't seem to do much for us.  Here's the same plot- this time showing UVP instead of MUTE (MUTE is already low); and you can still see the output doing this 'pop' as VDD is removed.

    So any ideas- how can we prevent an audio pop from occurring when we pull power with the DRV632?  This is supposed to be a feature of this device (and it's not that we're pulling power- this is just what normal shutdown looks like- VDD goes to 0V- looks pretty clean actually).

  • 0 in reply to radar
    TI__Guru** 109395 points
    Radar,

    The Mute/UVP feature of this device is actually a 80 dB attenuation.
    As seen in the plots, the pop at the input is relatively high like close to 2 V. This is being attenuated but 80 dB seems to be not enough for such a big transient.
    Another thing to take into account is that DRV632 have feedback paths that may pass the pop present at the input to the output.
    You can try DRV632 performance by disconnecting its input from PCM1608 and removing VDD from DRV632. There should be no pop.

    Best regards,
    -Ivan Salazar
    Audio Applications Engineer - Low Power Audio & Actuators