Except to tell a stupidity, I prefer, to make the calibration on all the range of frequency which I will use, with many points (therefore many segments). Like: 20 segments for 1-30 MHz with smoothing 25-6 (2020 points with my -F). It takes time but only once. Then I use this calibration file for the current sub-bands.
Why :
You will see that the calibration file contains errors which will show up as straight lines in the graphs (like "parasites"). As these errors are in the calibration file they are found, in fact, by the calculations in nanovna-saver, on all the measurements (at the same frequency) and it has a very bad effect.
So, I correct in this single configuration file the faults by interpolation (by hand).
--
Fran?ois
toggle quoted message
Show quoted text
-----Message d'origine-----
De la part de Lou W7HV via
15 octobre 2022 14:45
