Well, I just spent two hours re-calibrating everything; results using
Zplots were marginally better, but still not acceptable.
Re-reading Dan's notes on Zplots' "Measure Transmission Line
Parameters" feature, he describes two methods: the first uses only S11
measurements taken with open and shorted lines; the second method uses
both S11 and S21 data in Touchstone s2p format files - presumably taken
with the line terminated in 50 Ohms - on which he comments: "This method
may be used for lines of any characteristic impedance; it is not
restricted to just "50 ohm" type lines."
That comment seems to apply *only* to the second method, which means my
attempts to use the first method on 450 Ohm line were doomed to failure!
In retrospect, I'm not too surprised that you can't extract the TL
coefficients for high impedance line easily using S11 S/C and O/C
measurements; you'd need to accurately measure impedances as high as
30,000 Ohms!
The method I have been implementing manually looks similar to Dan's
"second method"; but unfortunately the VNA2180 can't produce Touchstone
s2p files, so I can't use Zplots to automate the process.
Looks like I'll have to revert to the manual number crunching :)
Steve G3TXQ
On 10/08/2013 09:36, Steve Hunt wrote:
Jim,
I'm comparing like for like: TLD's loss predictions for the line with
a 50 Ohm load, vs my measured results with a 50 Ohm load.
I'm pretty sure my line is equivalent to Wireman 551, and therefore
would expect Zplots should deliver similar K0,K1 and K2 coefficients
to those Dan has for 551.
I have a couple of hours free this morning so will re-run the open and
short circuit tests.
73,
Steve G3TXQ
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|