• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 30 subscribers
  • Views 2063 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

How to improve a noise of PCM1802

ttd
Mastermind 7670 points
Other Parts Discussed in Thread: PCM1802

 

Dear Support Team,

 

I have questions about PCM1802.

Our customer faces noise issue of it, and audio speaker sounds noise during no sound data(like a mute).

We want to improve it.

 

1.)

When VinL and VinR connect ground(like a mute situation), LSB 10bit data[9:0] of I2S is toggling(I2S 24bit mode).

We expect I2S data is zero(or more few toggling of I2S data such as only LSB 1bit), but I think this LSB data toggling is a little large for only PCM1802 noise.

Do you have any idea to improve it except Vcc, Vref?

If it is the spec of PCM1802, please let me know about it.

2.)

I2S data is 24bit and Vref is 2.5V(0.5 vcc).

In this condition, is following understanding correct?

I can't find the format of audio digital data in datasheet.

 

■PCM1802 Settings

SCKI    ：12.288MHz

LRCK(Sampling Freq)：48kHz

BCK(Bit clock)  : 64x48kHz = 3.072MHz

Interface mode  : Master mode (MODE1=1, MODE0=1)

Format          : I2S 24bit (FMT1=0, FMT0=1)

OSR             : Oversampling ratio select=1, x128fs

BYPASS          : Low (Normal mode)

Best Regards,

y.i

  • 0
    TI__Mastermind 27850 points
    Hi y,

    1. When you say grounded does this mean that the input at the pin of the PCM1802 is a 0 V or is it at Vref? If you are inputting 0V at the input and have a DC blocking cap on the input, then the pin should be at Vref. If the pin is actually at 0 V then you are near full scale. Have you measured this with any audio test equipment?

    2. Yes this is the correct understanding of I2S data.

    3. These settings look to be valid.

    Justin
  • 0 in reply to Justin Bohr
    Mastermind 7670 points
    Hi Justin,

    Thank you for your answer.
    I insert my commnet to your answer.
    Could you please check it?

    >1. When you say grounded does this mean that the input at the pin of the PCM1802 is a 0 V or is it at Vref?
    →Customer tested both input.
    I think that the difference of I2S data will be available between them, but currently both results of ADC look almost the same.
    The both results show that 10 bit data from LSB(24bit mode) are toggling.

    System is following and speaker sounds unexpected noise coutinuasliy during no sound data.
    If sound data input from Sound IC to ADC, it is correct behavior and sound is expected data.

    Sound IC → ADC(PCM1802)→Serializer(FPD Link)→Desirializer(FPD Link)→DAC→Audio AMP→Speaker

    They are investigating about VREF1、VREF2、VCC、VDD、DGND、AGND, because I think that they are affected by noise for ADC.

    >If you are inputting 0V at the input and have a DC blocking cap on the input, then the pin should be at Vref.
    →I think that this input is converted 0x0000 by ADC.
    Is it correct for normal operation?

    >If the pin is actually at 0 V then you are near full scale. Have you measured this with any audio test equipment?
    →I think that this input is converted 0x8000(Under of full scale) by ADC.
    Is it correct for normal operation?

    Best Regards,
    y.i
  • 0 in reply to ttd
    TI__Mastermind 27850 points
    Hi y.i,

    Yes general noise or offsets on the Analog supply and reference pins will cause noise on the ADC. This is why we recommend that there be decoupling capacitors as near to the part as possible for these pins to try and eliminate these factors. Is it possible to get a scope shot of the input?

    Yes those converted values are correct operation for I2S.

    Justin
  • 0 in reply to Justin Bohr
    Mastermind 7670 points
    Hi Justin,

    Thank you for your reply. I understand general operation of I2S.
    I think that customer's ADC looks strange behavior. Now I have been requesting scope shot of the input.
    If I get it , I will update to you.

    Best Regards,
    y.i
  • 0 in reply to ttd
    Mastermind 7670 points

     

    Hi Justin,

    I got following two snapshots including FFT analyze.

    These are Vcc and GND for AGDN and DGND.

    AGND and DGND are single point ground.

    It seems that Vcc noise causes around 124kHz frequently noise on GND.

    I think that these look a little big GND noise.

    Customer has been trying to improve it.

    If you have any advice , please let me know.

    ■VCC

    ■GND

     

     P.S

    Do you have PSRR for PCM1802?

    I can't find it on datasheet.

    Best Regards,

    y.i

  • 0 in reply to ttd
    Mastermind 7670 points
    Hi Justin,

    We are investigating this issue.
    If you have any idea, please let me know.

    Best Regards,
    y.i
  • 0 in reply to ttd
    TI__Mastermind 27850 points
    Hi y.i,

    We do not have PSRR numbers for the PCM1802, as we assume there will be a clean voltage available in an audio system. Is it possible for you to post a schematic of the system and or the layout?

    Justin