View Single Post
  #11   Report Post  
Old November 4th 03, 06:45 PM
Roy Lewallen
 
Posts: n/a
Default

I hope the readers will forgive me if I considered the radiation of the
two cases to be equal, not worrying about a couple of dB difference in
the range of -30 dBi. Actually, I know of no way to ascertain the total
radiation from the stub or wire alone, since it occurs at all azimuths
and elevations, producing both horizontal and vertical components, and
adds to and modify's the vertical's pattern. What I meant to say was
that the radiation characteristics are certainly very similar, and both
have the same general effect on the vertical's current distribution.
And, both for exactly the same reason. Cecil's earlier statement that
the wire radiates while the stub does not is certainly and demonstrably
not true, and the 2 dB difference in field strength isn't at all
evidence that one radiates more total energy than the other.

As for the statement that "EZNEC doesn't account for is the phase delay
through a bugcatcher coil", that's entirely true. As I've said several
times now, an EZNEC coil "load" is a lumped element model, which has
equal currents at its two terminals. A coil with significant physical
length doesn't behave like a lumped inductor, and therefore not like the
EZNEC model. I believe, but have no proof, that approximating a lengthy
coil with a combination of wire and load models will produce reasonable
results, but that's the best you can do with NEC based programs like
EZNEC. (Or with MININEC-based programs for that matter.)

Anyone who attempts to model a lengthy coil as a lumped "load" component
won't get results that closely model reality, for the same reason that
anyone who attempts to model a long wire as a short wire will be
disappointed. Neither should be a surprise.

Roy Lewallen, W7EL

Cecil Moore wrote:
Roy Lewallen wrote:

The stub produces just as much horizontally polarized radiation as the
wire.



Not true. The wire produces 2 dB more radiation than the stub. Given
that the stub is located in a high current region compared to the wire,
it is significant how much the stub doesn't radiate. If you replace
the stub with an equal length of single wire, it radiates 4 dB more
than the stub.

Run your stub vertical model with an elevation plot, and azimuth angle
of 90 degrees. Click FF Tab. Note the magnitude of the horizontal
component -- roughly -30 dBi. Then repeat with the experimental model
with the single horizontal wire.



Thanks, Roy, that's an angle I had not looked at. Results are above.

As I mentioned in my lengthy posting, the radiation from the stub
isn't a large part of the overall field, and this certainly shows it.
But it's certainly enough to disturb the vertical's current. Exactly
the same thing holds for the straight wire. Common mode current is
common mode current. No magic, no mysterious phenomena "not accounted
for" by EZNEC.



What EZNEC doesn't account for is the phase delay through a bugcatcher
coil which is an appreciable percentage of a wavelength. EZNEC is incapable
of modeling a bugcatcher coil. The only coil that EZNEC is capable of
modeling
is one that does not and cannot exist in reality.

Therefo One cannot use EZNEC to try to prove the current is the
same at both ends of a bugcatcher coil which is what kicked off
this entire discussion.