Re: NEC-LIST:NEC2 and Beverage antennas

>At 10:51 PM -0800 1/1/02, Charles Hutton wrote:
>>When I compare the results of my NEC2 models of Beverages with my previous
>>calculations, I am seeing some significant differences. Also, my NEC2
>>Beverage models aren't matching published data in papers such as Litva and
>>Rook's report and also Knight....
>
>Others are better qualified to comment, but I'll stick my neck out.
>My impression is that NEC-2 is inaccurate for antennas very near
>ground and/or extending into the ground, as a Beverage and its
>termination(s) are/do. Better to use NEC-4.
>
>-Chuck
>--
>The NEC-List mailing list <nec-list_at_gweep.ca>
>http://www.gweep.ca/mailman/listinfo.cgi/nec-list

NEC-2 should be accurate for wires very near to the ground, down to
about 1.e-7 wavelengths height as long as the wire radius is about
1/4 or less of the height for the thin wire approximation. The
problem with NEC-2 is to model the termination, since NEC-2 cannot
model wires in the ground. You can terminate the wire on the ground
with the "GE1" option. That does not correspond to any case in the
real world, but provides a termination. Then you can play with the
terminating impedance to minimize the standing wave on the wire. The
front-to-back ratio is very sensitive to the reflection at the
termination.

A simpler antenna to model is just a straight horizontal wire with
source and termination about 1/4 wavelength from each end. I use a
post processing program that takes the current on the wire with a
source at one end, and flips it over and adds it to the original
current with a complex multiplying factor. That way you can try any
terminating load in post processing, since the load is the same as a
voltage source with V=Z*I. The program averages the second
derivative of the magnitude of current and tries to adjust the
terminating impedance to find a minimum. It usually reduces the
standing wave to negligible levels. Then it can fit a linear
regression to the phase and log of magnitude of the current to get
the propagation constant. The result is usually in close agreement
with analytic solutions for the propagation constant on an infinite
horizontal wire over ground. I cannot send the post processing
program, since it uses a modified version of NEC that writes out the
current expanded as A + B*sin(ks) + C*cos(ks).

Jerry Burke
LLNL

-- 
The NEC-List mailing list <nec-list_at_gweep.ca>
http://www.gweep.ca/mailman/listinfo.cgi/nec-list