Indeed, it is optimized for 50 ohms, but I don’t know that I’d say the precision loses much of its meaning as you get away from that.
Can you quantify your assertion?
Ultimately, the accuracy is determined by the uncertainty of the raw I/Q measurements, which is around 1 part in a few thousand, and then whatever that does when calculating reflection coefficients. At big mismatches the signal into the reflected port of the bridge is large, so the SNR is high (70-80 dB?) which is 1 part in 10,000.
As they say in school, show your work - give an example of a measurement uncertainty with an unknown near 50 ohms, and then one of one near, say, 500 ohms, and 5k.
The Copper Mountain folks have an article on various ways to measure Z, and there’s differences in how the errors stack up whether you’re doing series or shunt techniques. That might be a place to start.
toggle quoted message
Show quoted text
On Feb 22, 2025, at 06:47, Team-SIM SIM-Mode via groups.io <sim31_team@...> wrote:
?Hi
We have the impression that the NanoVNA is as an impedance meter, a Q-meter or a frequency analyzer, while it is only a transmission line network analyzer, that is to say we have to do with impedances that turn around 50 Ohm with a SWR that can go to 10:1 max beyond which the precision loses much of its meaning,
in fact the S11 measurements are made at the base of a physical 50 Ohm resistive divider bridge, it is sufficiently precise around 50 Ohm but it loses its precision as we move away from this value, saying very correct measurements is between 5 and 500 Ohm otherwise the interpretation of the measurements becomes very compromised,
thus the Z-renormalization should not used to the extrem much from 50 Ohm and it will be tedious as we are moving away from 50 Ohm, so renormalizing the Z to 440 Ohm is really at the limit of validity of these ingenious calculation methods which facilitate the measurements with the greatest simplicity and least chance of implementation errors.
so with this good results, we are already Happy .
73's Nizar
