About Ad9613 error

I wonder why this happens to my ad9613.

The problem is that ADC can get first channel data correctly. The problem is that I cannot get the data from channel 2, even if I get the data without doing anything(means nothing goes into ADC),

the data noise is over 200mv.

(This data is channel 2 data, using Labview with hsc-adc-evalcz driver,  nothing going in to channel 2).

So, my question is , what may cause this problem to ad9613? Are there any patterns that cause this kind of problem?

I connected ad9613, referenced with ug-293. the set of ad9613 is all default.

  • 0
    •  Analog Employees 
    on Mar 13, 2019 2:51 PM

    Hi,

    The two ADC cores within the AD9613 share a single interleaved DDR LVDS output port. The fact you are able to capture data correctly from ADC Channel A through this single port, but are seeing bit errors when decoding ADC Channel B through this same port suggests the issue may be related to noise at the frontend of ADC Channel B. I would expect FPGA connectivity or significant capture code timing errors on this single data output port to effect capture on both ADC channels, not just channel B.

    Are you using our ADI AD9613 Eval Brd, or your own system AD9613 brd design?

    One Digital debug approach would be to enable the internal digital test patterns via Register 0x0D. These test patterns bypass the analog input frontend and inject digital patterns into the backend output data stream. It would be useful to see if these test patterns can be decoded properly from the atypical Channel B output.

    One additional Analog debug approach would be to physically tie the two VinB+/- Inputs (pins 61&62) together (don't ground them). This will force 0V differential on the Channel B Ain which should result in a static midscale output coding for Channel B (+/- any inherent DC offset on Channel B). If you still see the erratic code output on Channel B it could suggest a damaged ADC or FPGA receiver timing/decoding error on Channel B. If the Channel B output is quiet with the analog inputs tied together it suggests you have external noise coupling into the Channel B inputs. Floating the undriven Analog Inputs is not the same as tying them together to force 0V differential.

    Best Regards,

    TonyM