We are working with FMCOMMS3 board. We are doing the following two tests:
1.- With a termination in Rx1 we observe the level of the received signal (noise) for different Rx gain levels.
2.- Connecting Tx1 to Rx1 we transmit 0 and again we observe the level of the received signal (noise) for different Rx gain levels
In both tests Tx attenuation is 0dB.
The results are showed in the figures (the shape is due to the use of the RxFIR).
In this first figure (test 1) we observe that:
- The noise level does not increase linearily with the value RxGain until values above 50 dB.
- With a RxGain = 30 dB the noise level is practicaly equal to the RxGain = 0 case. We believe that this is due to the jump in the LNA gain level (from 4dB to 13dB) which reduces the noise before LPF.
In this second figure we notice that:
- With RxGain = 70dB the receiver is saturated (not printed in the figure) even when we transmit 0 from Tx1 to Rx1.
- The noise level increases linearily with the Rx gain from RxGain = 20dB.
We believe that the reason of this behaviour is a noise from the transmitter. Is that correct? What could the source of this noise be?
Could the reason be an incorrect configuration of the ad9361?