• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 807 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.

Original question:

TLV320ADC3101 audio offset

TLV320ADC3101: DC offset questions about the reason, impact and suggestion

Nick Hsiao19
Intellectual 645 points

Part Number: TLV320ADC3101

Hi Teams,

I searched that ADC3101 has DC offset issue, and I had confirmed different input selection config with different DC offset on my side. So here I have some questions need your help.

  1. Does the DC offset come from input selection mixer? Because the offset is changing while different input selection.
  2. What is the reason to make DC offset?
  3. DC offset means the lower dynamic range, I want to know what severity it is.

In below picture, the R channel DC offset is around -46dBFS, so the max dynamin range is 46dB? 

But actually, the ideal full scale is 0.7Vrms(ADC3101 spec), when I reduce the input down to 0.67Vrms, the clipping is gone. (in the picture, the 2k, 3k harmonic comes from signal generator itself). So I can say, the dynamic range impact is 20*log(0.67/0.7)=~0.5dB?

Which one is correct?

  1. I found the PDM input still have DC offset, but turned the input mixer off(NC), the offset still exist. Do you have any suggestion without using any HPF or offset compensation dither voltage?

  • 0
    TI__Mastermind 23050 points

    Hi Nick,

    Assuming you have AC coupled your inputs, the DC offset comes from the front end amplifier stage and the analog mixing can have some impact since this changes the signal path and has an impact on the input impedance. The offset will be gained up by the PGA, which is why it changes with different PGA settings. One thing to mention is that audio products in general are almost never trimmed for DC precision because it is such common practice to AC couple and use a HPF to filter out any offset that might be introduced from the ADC front end. I'm not sure why you are opposed to using the HPF and offset compensation options? These are the two mechanisms through which you can account for the offset in your signal.

    The impact the offset has to dynamic range is minimal, 0.5dBFS is a very common full scale error since again these devices do not get trimmed for DC precision. Some of our newer device in the TLV320ADCx140 family have much better trimming, but there is still more gain/offset error than you would find in a precision ADC because this is typically not critical in audio. For applications like beamforming where the matching is important, we provide gain and phase calibration features that can be used to match channels to a very high degree of accuracy.

    Best,

    Zak

  • 0 in reply to Zak Kaye
    Intellectual 645 points

    Hi Zak,

    Thanks, it is much clear now.

    To figure out the offset comes from, is important to us to judge how bad the dynamic range we lose. As I mentioned at point #3, I feel confused that the DC offset is about -46dBFS, so the dynamic range should be 46dB, right? At this condition, even I can set HPF in DSP, but I got the clipping in analog front-end, and it is irreversible.

  • 0 in reply to Nick Hsiao19
    TI__Mastermind 23050 points

    Hey Nick,

    This can be a little confusing since we tend to look at audio ADC outputs in terms of dBFS instead of raw codes or voltage. A -46dBFS offset for this converter would be approximately 5mV. You can still swing above and below this though so it does not throw away the bottom half of your range. It does impact your dynamic range slightly because now your input signal will hit max code 5mV earlier than you would have expected in this case, so instead of 1Vp being equal to full scale, it's actually 0.995Vp. So your second calculation of the effect on dynamic range is the more accurate one. This will change a bit with your PGA gain of course as you amplify the offset more

    Best,

    Zak

  • 0 in reply to Zak Kaye
    Intellectual 645 points

    Hi Zak,

    I totally understand the point, and it meets my experiment result. thanks for your help!! I would apply DC block filter in mini-DSP to fix the offset and provide the enough headroom for further processing.