• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 810 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:

AFE5818: HPF or DC compensation distortion?

AFE5818: PGA HPF distortion

Vassili Ivanov40
Prodigy 220 points
Part Number: AFE5818

I confirm that PGA HFP in AFE5818 cause high distortion in the signal, for higher input signal amplitude. LNA HPF does not cause any distortion for any signal amplitude. The only remedy against this distortion is to disable PGA HFP. The drawback of disabling PGA HPF is that PGA HFP is much more efficient in removal of DC boas vs LNA HPF.

The same issue is described in https://e2e.ti.com/support/data-converters/f/73/t/783963  (thread was closed but not actually resolved), but in that topic the problem was not localized to PGA HPF.

  • 0
    TI__Mastermind 49906 points
    Hi Vassili,

    Can you provide your input signal conditions?
    I will be able to provide a response by mid-next week.
  • 0 in reply to Praveen Aroul
    Prodigy 220 points

    Hi Praveen,

    To reproduce high distortion try LNA gain = 12 dB, PGA gain = 24 dB, Attenuator = 0 dB or disabled. -6 dB clipping =  disabled. LPA filter  = any. LNA HPF is not important.

    Apply 100 kHz signal from generator and increase the signal amplitude to about 80% of full range. Test PGA HPA = ON and OFF. 

    Note that distortion happens only for signals of large amplitude at least 50% of full ADC range p-p. For 25% of full range or less there is no significant distortion.

    Best Regards, Vassili

  • 0 in reply to Vassili Ivanov40
    TI__Expert 6040 points
    Vassili,

    Can you please share your captured distortion waveforms? we would like to take a look. Then Praveen can compare to his measuremnt.

    your description shows the signal of PGA HPA or offset correction circuit is saturated and can't correct the low frequency signal. it might be expected when PGA output is big. if your singal is 100KHz, we do not suggest turning on the HPF filter . that PGA HPF is designed for RF signal, like >1MHz. not for low frequency signals.

    we do see distortion when PGA HPF is on and signal freq is low, like 20KHz. So it might not be a suprise when you see some distoriton at 100KHz signals.

    Thanks!

    Xiaochen
  • 0 in reply to Xiaochen (Sean)Xu
    Prodigy 220 points

    Hi Sean,

    I created 2 screenshots with PGA LPF enabled and disabled (in my software). ADC clock was 80 MHz, 12-bit mode, the input signal was 100 kHz. Full scale on screen is - 32,768‬ to + 32,752 (12-bit scaled to 16-bit by multiplying by factor 16 - 4 low bits are zeros).  I just do not see any button on this forum to attach a file.

  • 0 in reply to Vassili Ivanov40
    TI__Expert 6040 points

    Vassili,

    your obersvation is correct. we got the same resutls as well. the reason is that PGA HPF filter is set at 80Khz, it is so close to signal frequency. thus it will start to distort the signal. when you care about the 100khz HD2, you can disable the HPF in PGA. our intention is to apply the filter to Mhz signal only.  

    two plots  one is diabled HPF, the other is enable the HPF. 

    Thanks!

  • 0 in reply to Xiaochen (Sean)Xu
    Prodigy 220 points

    Hi Sean,

    Signal is distorted by PGA HFP (HD2) only if the signal amplitude is high enough (but within the range for rest of the signal path chain). Small amplitude signal is not distorted. It looks like the signal within PGA HPF filter goes rail-to-rail or close. LNA HPF does not have such effects. Errata note should be created for this issue.

  • 0 in reply to Vassili Ivanov40
    TI__Expert 6040 points

    Thanks Vassili for the details! we noted these.