Time domain vector network analysis code

Am 12.04.2018 um 15:51 schrieb Phil Hobbs:

The founder of PSPL has still all the app notes on his personal web site. (and they are buried on one of my disks).

Cheers, Gerhard

I've made my own fft and ifft programs using the GSL library for similar purposes. The source for each is just under 100 lines of C.

The difference between a VNA and a TDR is that the latter has no directional coupler, so the data is the superposition of the incident and the reflected signal. Oh, and that the VNA measures a frequency response, so an IFFT begets an impulse response, whereas a TDR uses steps. OK, no big deal, I know.

Jeroen Belleman

What I'd like is a program that takes the TDR trace and constructs an R-L-C equivalent model.

I sometimes fiddle a Spice model to match the waveform. Some insider knowledge, or DC measurements, add a little guidance there.

Sweet. FWIW there's yet another open source Octave microwave toolbox available. [1] Among other things, it can import and plot Touchstone files on a Smith chart. (It's probably easy enough to migrate the toolbox to Matlab.)

Note.

1. formatting link

Thank you,

Half useful, I suppose. However, I prefer old school. I have specific instruments for bespoke purposes, so my VNA is a HP one from the 1980s. Only goes up to 1.3Ghz but that's more than enough for my purposes. I'm never completely confident that anything 'interpolated' or whatever the jargon is, is going to tell the truth. I have a morbid fear of artifacts (if anyone can tell me what the fancy Greek term is for this phobia they'll win a 5 shilling postal order.) Analogue rools!