[?] Measuring antennas with a Vector Impedance Analyser

Jul 18, 2010 4 Replies

I've been trying to optimise a quarter-wave antenna on a 433MHz transmitter using a T100 Vector Impedance Analyser.



Attaching just the antenna directly to the T100 shows that it is already a good impedance match to 50 ohms, so I thought I'd check it out whilst it is actually attached to the hand-held case into which the transmitter is fitted.



To do this I cut a length of RG174 co-ax cable (terminated in an SMA free plug) to be an exact half-wavelength long at 433MHz (remembering the velocity factor). I connected the cable to the T100's SMA connector and the other end of it to the SMA socket (other feed cable removed) mounted on the transmitter case. I then connected the antenna being tested to the SMA socket on the case. The transmitter case does offer a 'ground-plane' of sorts, although it is nothing like a quarter-wave in radius/length..



I had thought that the antenna impedance shown at an exact half-wavelength along a 50 ohm co-ax cable would have been fairly similar to that of the first test (but just at 433MHz), but the second measurement bears no sensible relationship to the first, and I'm puzzling to understand why..



No doubt it'll be immediately apparent to the experts in this NG that I'm not very familiar with the subtleties of antenna measurements. If someone who is knowledgeable about them reads this posting and would care to let me know what I'm doing wrong, I'd be most grateful to hear their comments and recommendations as to the correct way to make these measurements..



TIA - Dave


David C.Chapman - (dcchapman@minda.co.uk) --------------------------------------------------------------------------

Need a picture or five.

What sort of ground plane was the antenna operating against in the first case? What sort of ground plane (if any) was the antenna operating against in the second?

If you're just using the handheld transmitter case as the counterpoise, then your antenna performance will be more or less sensitive to damn near anything touching or close to the transmitter. Specifically, having a bunch of conductors (like a 1/2 wavelength transmission line) sticking out of the case will totally screw up your measurement.

So: pictures.

Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" was written for you. See details at http://www.wescottdesign.com/actfes/actfes.html

David Chapman Inscribed thus:

I would expect a resonant quarter wave to exhibit around 35 ohms assuming that it had an effective ground plane. Which your T100 is not !

Thats only true if the load is purely resistive at the frequency where the cable is a half wavelength.

As far as the quarter wave is concerned, against an infinite ground plane, ie one with a radius => 1/4 wave. Realistically is doesn't matter too much if the transmitter can feed the required power into it via the matching network without distress.

Best Regards: Baron.

One "gotcha" is the "exact half wavelength long" connecting cable. Was it _constructed_ to 1/2 wave, or _measured_ (using your T100) as 1/2 wave after construction? *If* what I have heard is true, a short piece of coax cut from a large reel can depart quite a bit from nominal specs. "Quite a bit" = ?? but I was told "as much as 20%". It's hearsay, and I find it hard to believe, but the part about measuring the thing is not. Even assuming a perfect 66 vf and 50 (+/- 2) ohm z construction error can move it a bit from 1/2 wave

Ed

Many thanks to those who took the trouble to reply (directly and via this NG) to my posting.

All comments and suggestions have been duly noted - I now know quite a bit more about the tricky world of antenna measurements.

ATB - Dave

David C.Chapman - (dcchapman@minda.co.uk) --------------------------------------------------------------------------

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required