Hugen
Can I suggest you use the same values of R25 and R24 in the attenuator circuit for the CH1 input as those used in for the reference SA612?
And then set R22 to come to 50 ohm? This will reduce the amount of calibration deduced compensation needed for S21 measurements and thus noise.
Interesting you found the unbalance less noisy. In my experiments I found the balanced less noisy. Not sure why????. Possibly because I was doing experiments on an unshielded PCB and we have rather strong DAB and mobile phone signals here and in a balanced configuration these could be eliminated because of the SA612 differential input????
--
NanoVNA Wiki:
/g/nanovna-users/wiki/homeNanoVNA Files:
/g/nanovna-users/filesErik, PD0EK
