I work with an N5222A 10MHz to 26.5GHz VNA. I use segmented sweeps to optimize SNR and speed. Since the VNA is a coupler based architecture, low frequency performance is not great in terms of uncertainty for this particular VNA.
I work in the signal integrity field and when it comes to step responses, the low frequency components are important. Unfortunately, due to the uncertainty of these low frequency measurements (especially reflections), my end-game time domain plots after IDFT are very not repeatable. One way to fix this would be to have a hybrid architecture where a VSWR bridge architecture covers the "lower" frequencies. But that's not an option for me.
I would like to find a way to Average say my first 1 or two segments for example 10 times using point averaging type and then do no averaging for the "higher" frequency segments.
This seems like this would be something be available to me. I can't find out how to do it though. I understand this will impact speed but I don't care about that.
I'm using Application Code A.09.90.17 on XP operating system for the record.