On 15 Jun 2006 11:51:24 -0700, "
wrote:

I'm wondering about calculating antenna efficiency from the Q as
measured at the feedpoint.

I know the definition of antenna Q has been hashed out a bit on this
group, so let me be specific and say that I'm measuring Q by looking at
the half power bandwidth around resonance as I would with an LCR
circuit, or maybe I'm looking at the generalization for a different
voltage down from the maximum, so I can use the 2:1 SWR bandwidth as
per the description he

http://lists.contesting.com/_amps/2006-01/msg00179.html

Once you know the Q, you need to know the antenna's radiation
resistance using some method, so that you can figure out the loss
resistance.

Is seems reasonable to model the antenna in EZNEC with lossless
elements and use the feedpoint resistance at resonance as the radiation
resistance and then subtract that off from the R calculated from Q=X/R.

Then all the loss elements are empirically determined, ground losses,
bolted connections, traps, etc etc.

Any comments on this? It seems a good idea to me but I'd welcome some
holes poked in it. I know that accurately modeling things like
multiband verticals and so forth even in the lossless case might be
tricky.


SWR bandwidth is pretty much meaningless in this effort. If you model
the antenna and then use the reactance slope with respect to frequency
around the resonant frequency then you can calculate an equivalent L &
C for the radiator. With this and the feedpoint resistance you can
calculate Q, for whatever it's worth.

Close to resonance the reactance slope is very near linear and the
real part is nearly constant. I use Excel's solver to find L & C and
go from there. With coupled elements (Yagis, etc) I'm not sure this
method is particularly useful, but I'm not sure whether knowing the Q
is very useful in the first instance.