"Roy Lewallen" wrote in message
...
Reg Edwards wrote:

The only program I am reasonably familiar with is the several years old
free
EZNEC. I don't know whether it has been updated or not and I make very
little use of it. Come to think of it, I don't make much use of my own
programs either.

Regarding shallow buried radials in conjunction with a vertical, have
you
tried my recent program RADIALS2 ?

It is intended to demonstrate performance of the radials themselves in a
given ground rather than antenna performance. Which I suspect is the
reverse
of NEC-4.

As you probably know, the effects of above-ground radials change very
rapidly as they get within a few inches of the ground surface. But once
in
the ground they tend to remain static.

RADIALS2 uses an entirely different, unconventional form of performance
analysis. If other programs don't take soil permittivity into account at
HF,
predictions must lose accuracy. Are the inputs and outputs of NEC-4 in
a
form suitable for a direct comparison with my simple program?

Yes. I made a few comparisons long ago, shortly after you introduced
your program, and found major disagreement. NEC-4 approximately agrees
with the measurements made long ago by Brown, Lewis, and Epstein (whom I
know you've never heard of), once you make reasonable assumptions of
ground conductivity and dielectric constant. Your program gives very
different answers. At the time, I concluded that there's considerable
coupling between radials, which your program doesn't seem to account
for. Interested readers should look in the google archives for postings
in this group on the thread "Ground Radials" in July 1998 and
"Evaluation of G4FGQ Freeware Antenna Software" in September 1998.


But in view of the large uncertainties involving ground conditions,
accuracy
is not worth making much of a song and dance about.

True, but in the past, you've used the results from your program to
reach conclusions about radial systems that I didn't, and don't, believe
to be valid. (See the threads mentioned above.) I don't think it's wise
to draw conclusions from a program that gives results which are
demonstrably very different from the only measurements regarded to be
reasonably well made.

Roy Lewallen, W7EL
==============================

Roy, it's gratifying to see, once again, you take notice of my sayings.

Such as, I repeat -

"Fact 4. Computer programs do not tell gospel truths. They are at least as
unreliable as their human programmers."
----
Reg.