I'm using picozed sdr1x1 with libiio IIO system object for data streaming. Without using any profiles, and calibration mode are set in auto, I'm sending 960000 I and Q samples to transceiver with 50000 samples per frame. When I saw the time domain response, the system is taking atleast 4 million samples to start tracking the waveform. If I reduce the no. of samples per frame, the settling time is less but I'm seeing lots of disturbances in between the frames leading to the trade-off. I'm also witnessing DC disturbance in between frames, which is unavoidable I understand, but is there any other way to reduce this disturbance or make it to correct the offset quickly? Also is there any reliable way to measure the settling time of TRx (with and without profiles)?
The above waveform is the response of ad9364 for FM waveform, the shape of the waveform is almost preserved but the DC level is disturbed in the beginning of each frame(with the phase inversion of 180).