You're going to tell me I'm doing it wrong, but I've been shorting the fixture, punching 10kHz into the span and ref/chn ref, loading the crystal, then centering the peak. Span to 500 Hz and centering again. Then off to advanced for the fun stuff. Pressing zoom a dozen times wasn't working for me.
That method should work fine, maybe even a little better than the method I describe in the manual. It comes down to how flat the Nano's response is in the range we are working over a small span. Again, we are normalizing the data (subtracting off the error). I show using a 1MHz span which is much different than a 500Hz span. Zooming into 500Hz, then taking the baseline with the short, then reinstalling the DUT gains us very little (from my testing) as the Nano is fairly flat. Still, 100K is going to be much better than 1MHz. I will change the manual to reflect this.
Using the term "ported" VERY loosely, because of how this software has evolved, starting with the old HP8754A, ported to Nano, ported to my other VNAs, then ported to the Plus, restructured to a common code base, then ported back to the Nano....
.... it's not a well thought out, polished bit of software by any means. To make things easier to maintain and keep more code common, some features work better on different products. The old Nano is very slow compared to the V2+4. Treating the old system like the new, then doing something simple like Zoom took a lot of time. I am starting to diverge from the common code for some functions to improve their performance. Really, that new V2+4 would have been the best solution for all of this had they considered the measurements you are trying to make. Seeing spikes in the lastest Hugen firmware tells me that the old Nano will continue to be a problem.
The button that starts the Zoom process is released when it completes. You could push it 1000 times and if you don't wait for it to finish, you won't get there.
BTW, I was playing with the bandwidth in the glitchy firmware and discovered that it defaults to 4000Hz. Tried it at the more standard 1000Hz and far less glitchy in the xtal data gathering. going down to 330Hz pretty much eradicated it.
Either way, the glitches didn't seem to show up in the 3D graphs.
I don't know what you mean by defaults to 4KHz. The span?. No doubt we saw in your first data that with a wide span, you may not see these glitches. They may appear at random. If you like collecting random data, I would run the Hugen's. If you want something more stable and slow use eddy 0.8.0. If you want something faster, glitch free and old, I would hunt down that last version RadioListener posted (keeping in mind that my software is now crippled to support the slower eddy firmware).
Our host sells a hand held meter that while not open source, does have user updateible firmware. Like the Nano, the firmware was in constant limbo. Like the Nano, I gave up trying to track it. Some people like to play with new releases of OSs, firmware.... I just want the thing to work well enough that I can use it (I've given up on modern software/firmware not having problems).