• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 30 subscribers
  • Views 516 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.

TAS5825M: DSP is producing a pop noise randomly during playback in 1% of playback sessions.

Federico Beckhoff
Prodigy 40 points
Part Number: TAS5825M


We have a BT speaker product in development. It uses an Actions ATS2835P as BT SoC and system processor, which configures the TAS with I2C and streams I2S from BT source. We have a 2.0 BTL configuration.
We are using the Process Flow 1, using PEQ, DRC, Clipper, etc.

I2C timing requirements are met. 

Here is a scope shot of the output and decoded I2S stream during a pop noise event.

Here is another capture.



Is this a known issue? The rate of occurrence seems t be 1% but often the pops will then continue until power cycle (i.e. configuration happens again) is done.
I have the I2C stream from the configuration from that session if needed.

  • 0
    TI__Mastermind 20541 points

    hi Federico

    this hasn't happen before.

    do you have any method about how to reproduce this issue?

    also, can you share me your ppc3 file, then i can check whether having any configuration problem.

    tks

    jesse

  • 0 in reply to Jesse Ji
    Prodigy 40 points

    Hi Jesse,
    Thanks for the quick reply. I have sent you the configuration file to the email address listed in your contact here in the forum. 
    The only way we are aware on how to reproduce it is just power cycling and starting playback, I2S stream until it appears. The current occurrence rate is 1%.

    Thank you and look forward to your feedback.

    Best,

    Federico

  • 0 in reply to Federico Beckhoff
    TI__Mastermind 20541 points

    hi Federico

    i got the configuration file.

    about this issue, please let me confirm some question.

    1. how the i2s stream work? does the lrclk change with different music or some setting?

    2. one more thing is it happen on how many chips? 

    3. have you tried on our EVM board? can it be reproduced.

    4. have you tried disable the clipper model in your ppc3 file and put the -0.1dB in AGL to see whether it can be reproduced?

    thanks.

    jesse

  • 0 in reply to Jesse Ji
    TI__Mastermind 20541 points

    hi Federico

    one more thing want to confirm.

    is this below the pop? it seems high related to the i2s data.

    also can you help enlarge the capture of i2s clock including bclk and LRclk ?

    i want to confirm its spec.

  • 0 in reply to Jesse Ji
    Prodigy 40 points

    Hi Jesse,

    1. how the i2s stream work? does the lrclk change with different music or some setting? 
    No, it seems to be always the same.

    2. At the moment we are not 100% sure if it is really being produced from the TAS alone but it is happening on all chips it seem, it is not an outlier.

    3. We have not tried to reproduce it with the EVK yet we have used the EVK for prototyping and did not experience the pop, if I am correct. Although playback time was very limited.

    4. have you tried disable the clipper model in your ppc3 file and put the -0.1dB in AGL to see whether it can be reproduced? We will try this.

    Here is some of the zoomed in I2S stream:





    BCLK Period: ~ 340ns
    BCLK Freq: 2.941 MHz
    Period L: 160ns
    Period H: 160ns

    LRCLK Period: 20.84us
    LRCLK Frequency: 47.985Khz

    TSL: 160ns
    TLS: 160ns

    That pop you showed in the picture (analog output waveform) is the pop we are talking about, yes. 

  • 0 in reply to Federico Beckhoff
    TI__Mastermind 20541 points

    hi, Fed

    That pop you showed in the picture (analog output waveform) is the pop we are talking about, yes. 

    -> this seems strange, you can see when the pop happen, your i2s data is also having strange points.

    that is purposely removed data or it is not on purpose?

    tks

    jesse

  • 0 in reply to Jesse Ji
    Prodigy 40 points

    Hi Jesse,
    The real root cause is not yet 100% clear but it seems it is indeed coming from the BT SoC. Thank you for the support.