I know you are not there to teach me signal processing, even though you have done quite some times . I wan to know if you have any ideas the why of this behavior (see image). I am not sure it is the DC offset, or some sort of aliasing the thing is that it also happens when I use the BIST on the TX, but it is less perceptible, see detail in figure 2 (Pay attention that the graph is with 10000 samples, just to make the phenomena clearer). The image below, figure1, is the packet I generated, and you can see the effect is strange. As the DC offset is set to auto I hoped it would have an envelop more or less constant. Could be a problem of sampling or of a "loose" harmonic, but I set the RF bandwidth to the minimum 1MHz and I had no perceptible change. The TX and RX sampling rates are also the same.. Any ideas?
The main parameters:
Calibration mode: auto
RX/TX RF Bandwidth: 1MHz
RX/TX Sampling rate: 8MSPS
RX/TX LO Frequency: 2.4GHz
RX gain: fast_attack/Manual gain 5
Cable between TX board 1 and RX board 2
Any ideas why the I/Q vary this way?