We incorporated AD9361 in one of our 2.3-2.4 GHz range Radio card designs. ad9361 is configured to work in FDD, single data rate, dual port mode. As a part of basic testing of the board we used sin wave samples to push from the TX port and captured in the spectrum analyzer. We observed that the spectrum coming out of the board is clean. But when we supplied the same kind of sin wave from signal generator in to the RX port (with 1 MHz offset from centre frequency) and plotted the power spectrum for the received samples, we observed a lot of spurs distributed over the spectrum with ~145 KHz spacing along with the actual sin wave frequency. The difference in the power levels corresponding to the desired and undesired frequencies is about 10dBm.
1. We probed all the power sources coming to AD9361 to check if there are any spurs introduced by supply itself (which is ADP5050), but except the switching frequency (600 KHz) there are no considerable spikes in the spectrum.
2. We bypassed the entire path from RX antenna to LNA, and fed the signal just before the BPF. The new path only contains BPF-->BALUN-->AD9361. Still the received samples exhibit same kind of spurs in the spectrum.
3. When no signal is fed at the RX Port the same spurs are observed in spectrum.
NOTE: The register set for ad9361 is generated using the software for AD9361 Evaluation module by choosing fc=2.35 GHz, SDR,CMOS Mode, SPI control to ENSM.
Can you please help to find what could be the source of the spurs in the spectrum, is there anything to do with calibrations in order to suppress these spurs.
Thanks and Regards