I am debugging some spurs I am seeing in my receive path in a custom AD9364 board. I am running the no-OS code. My board uses a 38.4 MHz external reference. The configuration I am using for the following plots is 61.44 MHz sample rate and 40 MHz analog filer bandwidth.
I have attached the following spectral plots from Matlab of the ADC output. This is with the RF input capped.
In the following plots, the dB axis is not calibrated to dBm but I have measured the levels. In spur.jpg, the strongest spur is about -75 dBm with the LO set using the API function to 1.919700000 GHz. When I retune the LO by 300 kHz to 1.920000000 GHz, the spur reduces by about 30 dB.
It leads me to believe something with the dividers and counters at certain desired LO frequencies causes these distortions.
Any guidance is appreciated.