Topband: radial layout and pattern skew

[Tom Rauch]

> In a comparison test I did a year or so ago, the results
of EZNEC3 and
> EZNEC4 (which uses the NEC-4 engine and can model
in-ground radials) were
> compared.  In EZNEC3, the radials were modeled slightly
above ground and
> in the NEC-4 program, they were modeled on (in contact
with) the ground.
> I was surprised at how closely the two programs agreed on
patterns and
> gain of the same antenna.
>
> In EZNEC3, the high accuracy Sommerfield-Norton ground
model was used and
> in the program with the NEC-4 engine (modeling was done by
the editor of
> the ARRL Antenna Book, Dean Straw, N6BV), the high
accuracy ground model
> was used.
>
> Therefore, if EZNEC3 (which uses the NEC-2 engine) was
inaccurate, then
> so was the program that used the NEC-4 engine.

Actually that only proves the particular model you used
agreed on all three. It doesn't mean anything else. We
cannot use an unverified model to verify another unverified
model, even if we compare it to a dozen other unverified
models! Someplace you actually have to measure the real
antenna.

If we look at Belroses NEC-2 model of a low dipole compared
to an actual FS plot made by Haagn, NEC-2 was off by 5dB at
low heights for a dipole!!! It under-estimated loss.

If read the warnings attached to Eznec, it specifically
warns about things like wires near earth.

73 Tom