Best way of impedance measurement ?

[yechi]

What is the best way to measure the impedance(S11 with network analyzer), (for example the output matched impedance of a PA) at high frequencies (above 5 GHz) with the minimum errors due to low impedance, parasitic capacitance, and soldered connector influence ?

 

[Skogsgurra]

You do need a network analyzer! Homebrew boxes do not work at those frequencies. Unless you are very, very good in that part of the spectrum - and then you probably needn't ask...

Seen some network analyzers on eBay lately.

Gunnar Englund

http://www.gke.org

 

[yechi]

More specific, is there away to minimize measurement errors due to the measurement itself (deembedding) ?

 

[Skogsgurra]

1/4 wavelength is 15 mm. It is not easy. I would not even think of it. Someone else?

Gunnar Englund

http://www.gke.org

 

[logbook]

Measuring output impedance with a VNA is not necessarily a good idea anyway. The technique using a VNA back-drives an output. How representative is that going to be?

The thing you are trying to measure is source match or source reflection coefficient. One way is to use a long cable and sweep the input frequency. You then need a power or voltage detector on the output. The cable changes the reflected signal phase continuously as the frequency is swept. This method is good for wideband sources.

If you have a narrow band source then you need a directional coupler, a power meter and a sliding short circuit. This kit is more than 10x cheaper than a VNA and gives a better answer.

 

[Biochemist]

Sorry to barge in on this thread. I am also interested in measuring impedance. Is it possible to buy/put together from commercial parts, a network analyzer that measures in the visible range of the electromagnetic spectrum?

I have a liquid nano structure that I would like to analyze it's permeability and permittivity in order to perhaps explain it optical/light scattering/sound altering properties. HP only sells them up to the GHz range.

I am a biochemist. I do not have a clue.

 

[logbook]

Anritsu/Agilent/OML/ABmm do network analysers into the mmwave region (say 1000GHz). You are then talking far infra-red.

However people also make optical spectrum analysers (OSA)

http://www.promax.es/ing/productos/familias.asp?subfamilia=37

http://www.npl.co.uk/photonics/training/osa.html

Once you have an optical source (laser) and an OSA you can start to make insertion loss measurements.