• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 30 subscribers
  • Views 542 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.

PCM9211: strange behavior with DIR

Ralf Koschnicke
Prodigy 80 points
Part Number: PCM9211
Other Parts Discussed in Thread: PCM1794

Hello,

I noticed the following strange behavior:

The signal comes in via RXIN6 and is sent to TXOUT and RECOUT1 in parallel:

Register 34: B11000110

Register 36: B00000110

 

From the MPO0 output the signal goes to a DAC.

If TXOUT is routed to MPO0 (register 78: B00001101), an impulse response of the DAC measured via ADC shows significantly more ringing than if the same signal is routed via RECOUT1 (register 78: B00001111)

At the same time, the measurement of frequency response and THD is absolutely identical. I also recorded a piece of audio as a digital signal at the output of the PCM9211 and did a bit comparison. In both cases, the audio is exactly bit-identical. Thus: No problem is seen in static measurements and with real audio signal. Differences only become apparent with a short pulse (see pictures).

Does anyone have any idea what's happening here?

Furthermore: If I send the signal not as AES3, but as I2S or left justified or right justified signal from the MAIN OUT to the DAC, the result is the same picture as TXOUT: a lot of ringing.

 If I remember correctly, the DIR is bypassed via RECOUT. So it looks like something strange is happening in DIR.

By the way, all measurements at 192kHz.

Best wishes

Ralf

  • 0
    Prodigy 80 points

    Another addendum: It is unlikely that the matter has anything to do with clock jitter. The DAC (PCM1794) can draw its master clock from the AES3 signal (via CS8416) or it gets another clock separately. This in turn can be the DIR of the PCM9211 or it has been improved again by a precision PLL. The picture is the same every time.
    Even though the clock always comes from the PCM9211 via a separate line, the same difference can be seen between RECOUT and TXOUT, regardless of whether with or without the precision PLL.
    Therefore I would say: definitely something to do with the signal processing in the PCM9211 and not a clock issue.
    And I can also give the signal to the DAC completely bypassing the PCM9211. Then it looks like the RECOUT. So the one with less ringing is the right thing. As soon as the DIR is in play, there is ringing...

  • 0 in reply to Ralf Koschnicke
    TI__Mastermind 31541 points

    Hello Ralf,

    From what you are describing, I don't think it is clk issue and it has to do with some internal circuitry of DIR. I don't have the internal  schematics to give a suggestions of what might be the cause but let me take a look at this some more and see if I can find anything. I will get back to you by Thursday.

    Regards,

    Arash

  • 0 in reply to Arash Loloee
    Prodigy 80 points

    Hello Arash,
    Thank you for your reply.

    I have now made extensive listening comparisons and am sure that there will be no sonic disadvantages in real use. That was actually to be expected, given that the digital output signal is bit correct and the clock cannot be the cause. But the strange impulse response made me a bit nervous that something subtle could be going wrong in the time domain...

    But I noticed something else. If I set the DAC to left justified instead of I2S (PCM9211 still I2S output), the Dirac comes through without amplified ringing. (Frequency response is shifted upwards by 6dB).
    Could it be that the PCM9211 swallows a bit in case of such a short pulse - and only with such a short pulse - and then the bit assignment in the output signal is no longer correct?

    Best wishes

    Ralf

  • 0 in reply to Ralf Koschnicke
    TI__Mastermind 31541 points

    Hello Ralf, thanks for the update. I could not find any schematic or internal document that I could explain what you see but since we have never had any report of any issue for this part, i believe it should not be any concern to use the part for your application.

    If the PCM expects  the data to be I2S, then if data is send in another format, still it threats it as if it was I2S and really doesn't know the data was RJ or LJ. So as far as it is concern it does what it thinks it is the correct data, but in reality from outside, we know the hand shake was not correct and bit was shifted and thus the outcome would be shifted accordingly. Having said that, you might be right and it might be swallowing a bit  on that short pulse!  Hard to tell what happens inside the block  for this specific case. 

    Regards,

    Arash