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.

PCM1754-Q1: output noise even when I2S is not configurated

Part Number: PCM1754-Q1
Other Parts Discussed in Thread: TPA6211A1, PCM1754

Hi Expert,

Our customer use PCM1754-Q1 + TPA6211 in their project. The schematic is as below. When power on and unmute pcm1754-q1(mute pin is low) and they don't even initialize I2S, they can hear noise from speaker. Is that normal? What kind of reason can cause the noise? I ask them to test pin8&9 using oscilloscope but didn't find noise. Could you please help and give some hints for debugging? Thansk.

BR,

Elec Cheng

  • Hi Elec,

    Can you share a capture of the output noise? Is this present all time like white noise or just a brief transient like pop noise?
    You should try to AC couple to GND the inputs of the TPA6211A1 amplifier and check if there is still noise. Basically let's isolate each device on the signal path.

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • Hi Ivan,

    Thanks for your reply. The detailed test result is attached.

    Is this present all time like white noise or just a brief transient like pop noise?

    I guess it's like white noise. 

    You should try to AC couple to GND the inputs of the TPA6211A1 amplifier and check if there is still noise.

    Do you mean set oscilloscope as AC couple to test? I suppose the customer use DC couple for the but maybe you can still find out the noise. Thanks.

    Just adding one more question. For I2S format as below, how to decide the number of  "N" the LSB? Is it fixed? 

    rt1176-PCM1754 audio noise problem.docx

    BR,

    Elec Cheng

  • Elec,

    Do you mean set oscilloscope as AC couple to test? I suppose the customer use DC couple for the but maybe you can still find out the noise. Thanks.

    No, I mean short the input side of C228 and C229 to GND in order to separate the amplifier input from the DAC output.

    Can you clarify if there is noise on the following cases?:

    • TPA6211A1 unpowered / PCM1754 unpowered
    • TPA6211A1 powered/ PCM1754 unpowered
    • TPA6211A1 unpowered / PCM1754 powered and muted
    • TPA6211A1 unpowered / PCM1754 powered and unmuted
    • TPA6211A1 powered/ PCM1754 powered
    • TPA6211A1 powered/ PCM1754 powered and unmuted
    Just adding one more question. For I2S format as below, how to decide the number of  "N" the LSB? Is it fixed? 

    The number of offset bits would depend on the host configuration, eg. if host is slot length is 32-bit, but there is 24-bit word length, then you'll have 8-bit offset (N = 8).
    Notice that PCM1754 supports 16-bit and 24-bit word length for I2S format.

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • Hi Ivan,

    Thanks for your reply. Below is some feedback from the customer. Please help. We got a quite tight schedule since they will DV soon. Thanks.

    • TPA6211A1 unpowered / PCM1754 unpowered---no noise 
    • TPA6211A1 powered/ PCM1754 unpowered---no noise 
    • TPA6211A1 unpowered / PCM1754 powered and muted---no noise 
    • TPA6211A1 unpowered / PCM1754 powered and unmuted---no noise 
    • TPA6211A1 powered/ PCM1754 powered and muted---have no noise 
    • TPA6211A1 powered/ PCM1754 powered and unmuted---have noise 
    They also power and unmute PCM1754, power and mute TPA6211, there is no noise. The noise comes only when both PCM1754 and TPA6211 are powered and unmuted, even before sending I2S. From my understanding, I guess the noise is added by PCM1754-q1.
    Besides this, the customer also have other teat items. They tie the test point to GND separately, which means they only tie one point at one time. The orking condition is TPA6211A1 powered/ PCM1754 powered and muted without I2S sending. The result is very interesting.
    • Tie TP74 to GND---noise disappear
    • Tie TP73 to GND---noise is smaller half
    • Tie TP202 to GND---noise is smaller half
    • Tie TP203 to GND---no change.
    I don't really understand why they can have this kind of test result. Please help ASAP since they are pushing. The detailed schematic is attached earlier. Let me know if you need anything else. Thanks.
    BR,
    Elec Cheng
  • Elec,

    Please share captures of their digital input signals, most likely the format or timing is not correct.
    There is not much you can configure on PCM1754, however if the input data is not correct in terms of edge detection, offset bits and length. The input data must follow the description on the capture you shared before:

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • Hi Ivan,

    There are two abnormal situation:

    •  the noise can be heard even before sending I2S. 
    • after sending I2S, the noise is louder

    For the second case, it's probably related with I2S. Surely I will work with the customer to get the I2S waveform. But for the first question, could you please also give some hints since the customer and I are working to solve this problem first? Thanks.

    BR,

    Elec Cheng

  • Hi Elec,

    I would not recommend to unmute the device while there is no I2S, what would be the use case to unmute while there is no audio being played?
    Anyways, I'm checking what else could be related.

    For the second case, we need to make sure the format is correct, please try different edge and offset configurations from the host side.

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • Elec,

    Please also share captures of their input clocks SCK, BCK and LRCK.
    These clocks must be within the specified ranges and SCK to LRCK ratio is important as well:

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • Hi Ivan,

    Yes I agree with you that no I2S and unmute is an unusual case. But the noise does exist, which is very strange. So we highlight this for finding some potential problems. This problem seems irrelated with I2S so we simply treat them as two question and follow at the same. This problem has been pending for one week but we still don't have much progress. The customer is pushing since it's near DV. I really need your help for this. Thanks for your understanding.

    I will test the waveform as you suggest. 

    BR.

    Elec Cheng

  • Hi Elec,

    I'll follow up with some direct messages.

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • Hi Elec,

    Any update on this issue.

    The TPA6211A1 can be checked stand alone with PCM1754 Powered off and the TPA6211 inputs being fed from a standard signal generator.

    A 1khz sine wave can be given to the device and sound from speaker can be checked for Noise. If there is no noise the device is fine.

     

    The PCM1754 should also be tested stand alone with digital input and given. Analog output should be seen on an oscilloscope.

    If this also is fine then the issue may be a grounding issue .