I am currently trying to verify if this ADC chip is right for my design. I am running it at a sampling rate of 3GSPS with an LMFS of 82820. I am trying to record a LFM signal that starts at 450MHz and ends at 1050MHz. Below is the frequency response of my captured data.
Unfortunately, the response is rather "wavey." You can see a sharp increase in power in the band of my LFM however, so the energy is getting through. If I take a FFT of just the noise floor it is equally as wavey. So equal in fact, that I can pretty much compensate for the wavey-ness of my first capture using the noise floor. This leads me to believe that the input signal into the ADC isn't the issue here. Below are the frequency of the noise floor and the compensated LFM signal:
Have you seen anything like this before? Could it be some register setting in the ADC? When I put in a CW tone it seems to map it correctly, so I am not too skeptical of how I am mapping the JESD lanes, but I haven't ruled them out entirely.
Any help would be appreciated, thank you.