I've observed severe distortion in I and Q sinewave waveforms. Please see the attached Word document for a detailed description.
Do you see the same when using the on board crystal and default device tree?
I don't think we have seen that.
Can you please observe signal at reference clock input and 1.3V supply rail using an oscilloscope?
Are you What algorithm was used to set the Manual Gain Control for RX1? Have you tried an Automatic Gain Control setting? It looks to me like some kind of saturation effect.
When measuring the reference clock input and 1.3V supply rail using an oscilloscope; be sure this is with the IC connected to these, not with them open.
When I originally posted this problem on April 22, 2015 I included detailed conditions (including a Dropbox video) so that the problem could be aacurately replicated. I have seen very similar symptoms now on three different "platforms" (1. Analog Devices FMCOMMS3-EBZ / Avnet ZedBoard, 2. Ettus Research B200, and 3. my company's design). I have used multiple external generators for these tests to eliminate them as possible culprits. In view of all this to my mind the symptoms strongly suggest a flaw in the AD9361. I'm surprised that Analog Devices has not undertaken to replicate this problem. I really do think they need to take this issue more seriously.
Retrieving data ...