[time-nuts] VNA design

Dr. David Kirkby drkirkby at gmail.com
Mon Jun 2 19:44:00 UTC 2014


On 2 Jun 2014 19:07, "Alexander Pummer" <alexpcs at ieee.org> wrote:
>
> I do not wanted to discourage any body, but building the hardware of a
network analyzer is not a simple task, and requires substantial
instrumentation, software could solve hardware problems to certain limit
only
> 73
> KJ6UHN

VERY TRUE

The larger the hardware errors,  the larger the errors that need to removed
so the accuracy of measurements suffer with the slightest change in
temperature.

Adapters with a poor return loss BEFORE the calibration plane can still
cause problems with stability of measurements.

Despite error correction, it is not uncommon to do things like improve the
match at test ports with attenuators.

I think it would be unwise to embark on designing a VNA unless one has at
least used one first.  I don't think designing a VNA is the best way to get
one. Perhaps after buying one (8510 or 8753 is probably best) then
designing one is likely to be more productive.

Dave.



More information about the Time-nuts_lists.febo.com mailing list