On 10/6/21 8:20 AM, Victor Reijs wrote:
What I don't understand: If 'weird' measurements can be handled/calibrated
with code in the NanoVNA (which is a smallish device), why would it not be
possible to have proper calibration, without crashes (even of 'weird'
measurements) on a PC (aka NanoVNA Saver)?
It could be an artifact of how scikit-rf works or how NanoVNA-Saver is coded.
For example, the uncalibrated nanovna returns gamma for ch0 as reflected measurement/source measurement
It's easy to imagine raw reflected > raw source, so |gamma| is greater than 1.
It might be that the code doesn't allow for that.
Saver, in general, doesn't have a lot of error recovery built in - it's a nice program, but there's plenty of places where if something is wrong, it just dies, and leaves it up to you to go look at the source code to figure out how to fix it.
The plotting routines are probably more brittle than the scikit-rf routines - that's sort of typical for Python - PyQT5 works fairly well, but there are times when some assumption isn't met internally (e.g. in the data you pass it) and it gets confused.
Op di 5 okt. 2021 om 23:14 schreef Roger Need via groups.io <sailtamarack=
[email protected]>:
As I mentioned earlier the calibration routines in NanoVNA Saver depend on
having calibrated input over a wide range from the NanoVNA according to
Rune. This may be because the algorithm used in Saver does not do much
bounds checking and would not work well with raw data. I have had
instances where Saver just aborts at the conclusion of a cal routine and I
have to start all over gain.
