At 10:53 AM 5/13/2007, Peter Voelpel wrote:
>From: Jim Lux [mailto:email@example.com]
>At 02:51 PM 5/11/2007, Peter Voelpel wrote:
> >>Those figures are EZNEC simulated
> >Which figures?
>The figures given by Optibeam on their homepage
> >In general, decent NEC models of different antennas should be
>intercomparable. "decent" being the operative word.. It
> >takes a fair amount of skill and work to generate good models, especially
>if there are tapered elements, traps, or lumped >components. The other
>thing that will definitely be an issue for a model is anything where the
>feedline or other stuff
> >is part of the antenna, whether deliberately or not.
> >But, for comparisons of run of the mill Yagi-Uda arrays, with attention to
>segmentation and such, one should be able to
> >get comparative data within 1/2 dB or so.
>Try to convince Steve.
>Seems he is not a believing in antenna simulation.
I'm a believer in modeling (A good part of my job
depends on modeling), but not necessarily in model builders.
>But I don´t believe in gain figures which are not measured in the main lobe
And there's a key problem. Most of us don't
operate antennas (particularly for HF) in
anything resembling "free space", so you have the
antenna pattern overlaid on the "ground
reflection" pattern. Add to that the fact that
many antennas, at typical heights, have their
main lobe (with ground effects) substantially
above the desired take-off-angle for the signal
of interest, and you're set up for all sorts of inconsistencies.
You have the gain at the main lobe (generally too
high elevation).. the gain at the signal's angle
of arrival (not known with any degree of
accuracy, and you're in a part of the pattern
where 1 degree makes a big gain difference).. and
the gain measured in a test scenario, essentially at zero elevation angle.
I think the empirical measurements in this case
serve as an extremely valuable "reality check"...
you can model the antenna, and if the number you
get from empirical measurements is 10dB
different, something is going on (unaccounted for loss in the antenna, etc.)
Modeling serves an extremely valuable service in
assessing the "sensitivity" of the antenna to
changes in configuration. If changing the
spacing of an element by a few cm or it's angle
by a couple degrees causes radical changes in
modeled performance, that does not bode well for
an amateur installation, where the elements are
waving around in the wind, birds sit on the antenna, etc.
TowerTalk mailing list