AD9364 Signal Distortion at Rx in MGC mode

imBilal said:

but when I change the mode to MGC and give a small gain (i.e. 0-30, that does not cause overflow) then the signal at the Rx has some noise in it.

Can you check the FFT plot in this case? There might be saturation in the RX path. 

imBilal said:

The second issue is that the here adc and dac valid is high once every 1 cycle (1 valid in 2 cycles). But with AD9361, I have observed that the adc and dac valid is high once every 3 cycle (1 valid in 4 cycles).

Are you referring to Aux ADC valid signal? How are you probing the DAC valid signal?

I have not connected any FFT IP and I am using the No-OS implementation so there is no spectrum available. If there was some saturation, then the values on the adc would also be very high but this is not the case here. Since when my MGC is at lower value, how can there be saturation with lower values. My external signal is also -30dBm. 

I am probing the dac valid signal in the ILA. I am unaware of AUX adc signal.

If the gain set for the RX path is too low a compared to the input signal, then the signal amplitude will be very less and get buried in noise. You can try increasing the input signal amplitude with the same gain  and you will see clean spectrum at the output.

I just observed that some times when I program the device, I don't get any noise but some times with the same settings the Rx data is very noisy. I am unable to understand this behavior. 

What is the possible issue with that?

Is the interface tuning passing successfully for all the power-ups?

 If not, can you try running the digital interface tuning? refer to the below link:

https://wiki.analog.com/resources/eval/user-guides/ad-fmcomms2-ebz/interface_timing_validation 

Yes interface tuning is passing successfully each time.

Are you still facing the issue? If yes, please re open a new post with full details and the issue. Also, try using internal BIST and check if the same issue persists with multiple boots.