Waveguide offset pieces (spacers) for calibration

[logbook]

I thought I had understood waveguide offsets (an empty waveguide spacer section) to estimate mismatch errors, but it seems I was over-confident. I bought a quarter wave offset waveguide section to reverse the mismatch error in my scalar measurement system and therefore give me an uncertainty limit. But if the mismatch phasor is at right angles to the incident phasor, 180 degrees on the mismatch gives the same scalar answer!

I guess that is why calibration kits have lamda/8, lamda/4 and 3 lamda/8 offsets. It looks like any single offset can always be phased up to give no change at all between the normal and offset readings. Hence I would need to use at least two offsets.

Is there some simple rule concerning the use of offsets and the resulting scalar changes in the measured quantity?

 

[Higgler]

Normal scalars have open short cal averaging, and so your 1/4 wave purchase is correct for use in full waveguide band measurements. Yes, you may have exactly the wrong phasing at one frequency, but the assumption in scalars is that you go over a wide frequency range and can see the ripple due to the equipments imperfection.

If you slightly better calibration, you could add others such as the 1/8 wavelength. You can add dielectric to your short to lengthen it too and see the ripple positioning change.

The next big step is to buy or make a sliding short if you need much better. Eventually you'll need a vector analyzer to get really low VSWR accuracy.

kch

 

[logbook]

Interesting.

My scalar measurement is essentially at a spot frequency using a power meter. I was going to put in a spacer to estimate the measurement uncertainty due to the mismatch factor.

I don’t think a sliding short would be of use in this situation.