I would think the port 2 (labeled Ch1) on nanoVNA and nanoVNA derived would be
pretty close to 50 ohms - it’s a ~17 dB pad. And Port 1 (labeled Ch0) is a
resistive bridge - so its impedance is slightly affected by the Z of the
stimulus source and the receiver on the other port of the bridge. At HF
frequencies, I’d be surprised of the Z is far off. (But I can measure it… I
have two NanoVNAs - I’ll do that later today)
In fact, I would think that if the cal process comes back with something
significantly different from 50 ohms, something’s broken.
The versions of the NanoVNA (NanoVNA2) that have electronic switches in front
of the ports - yeah, those can be pretty wonky and the switches are sensitive
to ESD damage.
On Fri, 23 Jan 2026 18:18:44 -0800, Brian Beezley <k6sti@att.net> wrote:
AC0C said:
"Brian, would you consider the N2PK VNA in the "fancy" category rearding
your calibration comment?"
Jeff, I don't think the N2PK measures all four S-Parameters nor does it
do a 12-term calibration. I was thinking of the HP 8753C and 8753D,
which are sophisticated VNAs that a surprising number of hams have.
I confused some things in my last message. The NanoVNA compensates for
the port 1 impedance when doing a single-port calibration, but I don't
believe the stock firmware compensates for it when doing a two-port
calibration. Also it is the port 1 impedance that was so far off for the
particular NanoVNA-H4 N6LF tested, not port 2.
I'd be interested in seeing an .s2p file for the commercial test
fixture. If the S21 method is good enough without going to the trouble
of making a Y21 measurement, that would be good to know.
Brian
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|