We are analyzing Tx impairments of the FMCOMMS3 board. We have connected an external LNA (20dB) to overcome the noise floor of the spectrum analyzer. Under these conditions we transmit 0 and we get the following spectrum signal around our carrier frequency (2.8GHz). What we observe in the spectrum analyzer is a carrier at 2.8GHz and an unknown noise shaped.
We thought this noise could be phase noise (as we discussed in About ad9361 Tx noise ) but when we changed the TX bandwidth we observed that the noise were filtered by the BB LPF placed after the DAC (as shown in the next figure). Therefore the source of this noise must be before this filter (dither?).
In order to discard the phase noise we performed another test. We transmited a high DC component but this noise did not increase so it is not a multiplicative noise as the phase noise.
what could be the source of this noise? Is it possible to reduce its power?