Yes, I see, the spurious are higher than the output frequency and the power meter reads the level sum,
so will use a Ytf (Yig tuned filter) preselector to avoid this.
Anyway, using a calibrated spectrum analyzer with direct connection sma-sma there are? level errors.
Reference from 3 to 4 GHz -30dBm no errors(0dB),? from 4.5 GHz? to 5.4 GHz? are -2dB.
Tinysa is a very good test instrument ,? I ask if may be improved this frequency flatness by software.
