Dear AD9361 folks,
I'm working on a TDD system with the pin-control ENSM. My problem is a signifficant noise affecting samples received from ADC. What you can see in the figure below are the samples received from ADC through util_ad9361_adc_fifo on both Rx channels. The last line is a valid in line, so you can see that the head of my Rx frame reaches my IP core near 26,080 offset. As the tranmitter does transmit anything, only the noise should be visible in all I,Q components. However one of them is contaminated by some extra noise.
The noise would also appear if the receiver gets some useful data from the transmitter. Below you can see the moment when the head of the signal preamble starts, but the I-component of Rx1 (the first line) is degraded by the noise, again.
My question is what can cause this kind of noise? It can appear in I, Q, or both signal components, and can affect either Rx1 or Rx2 channel.
It is visible if the signal is transmitted over the air or by a cable (in the second case one Tx channel is splitted to both Rx channels)
Some details about my system:
Are you keeping the RX and the TX LO at the same frequency? If yes, can you try by keeping the LO of both the paths at different frequencies?
In which RX path(RX1 or RX2) are you seeing the degradation when there is no signal at the input? Are you seeing the increase in noise sometimes only in I component and not in Q component?
thank you for your response.
In my setup, both Rx and Tx LO work at the same frequency (2.5 GHz). I tried to catch the noise after changing the Tx frequency to 2.3 GHz, as you suggested, but today there was almost none noise event. I mean, it appeared only 2 times in about 50 trials (only with the same Tx and Rx frequency).
When switching Tx LO frequency, sometimes I received a message: 'ad9361 spi1.0: Calibration TIMEOUT (0x287, 0x2)'.
Previously, the noise could appear in Rx1 path, Rx2 path or in both paths. Yes, sometimes it occured only in I or only in Q component.
I'll be back having some more observations.
There is VCO pulling issue in AD9361 because of which if the TX and RX Lo are set at same or close frequencies, spurs get added in the spectrum which may be causing this increase in noise.
I tried to collect some more data for analysis, but I failed. The phenomenon has not appeared anymore. Thank you for your help.