I have the nanoVNA V2 plus, the SVA1032X and the LibreVNA.
I'd recommend the LibreVNA (full 2 port S params), but the nanoVNA V3 (aka VNA6000) is apparently just around the corner, so I would wait and see how that thing performs if you only need to measure S11 and S21 without flipping the DUT manually.
I'll be reviewing the LibreVNA in a couple of weeks, but the TLDR is:
- Software is not very polished (features are a bit hidden, may freeze up after a few hours of continuous use for some users), but has some advanced features found only in VNAs many times more expensive (TRL cal, time gating, 2x thru de-embedding etc.)
- Hardware can get hot and readings may drift, but not a problem if you stick a heatsink (optionally a fan) on it.
- 6GHz or even higher if using harmonics, giving good time domain resolution for the price.
- Full 2-port (measures S11, S21, S22, S12)
- I think it is the best value for money in the low cost VNA market right now, open source everything is a big plus.
The SVA1032X IMO is primarily a SA with a VNA tacked on it. The features and usability have improved over time, but is not nearly as powerful as the LibreVNA in terms of being a VNA (T/R vs full two port, advanced de-embedding etc. etc.) . I rarely use the SVA1032X as a VNA after I got my LibreVNA.
Thanks. Can you comment on the lowest frequency the LibreVNA is useful at? The purpose I have in mind right now is looking at PDN (power distribution network) and measuring combinations of decoupling capacitors. Measurements go from sub 1MHz to several hundred MHz (close to 1GHz). I have two VNAs; one goes from 10Hz to 300MHz and the other from 5MHz to 4GHz. There is overlap but for some measurements it sucks needing two instruments.
The LibreVNA is designed for 10MHz and up IIRC, but is usable down to 100kHz in my experience (and I do use it down that low), though the S21 dynamic range is only around 40-50 dB from 100k to 1M, not the 90dB dynamic range higher up in frequency.