On Tue, 11 Oct 2005 19:20:35 GMT, "Frank"
wrote:
I did not notice that somebody had attempted what you suggest. I
translated
the coordinates into regular NEC2, and it does run. It appears to violate
some NEC2 criteria, but not sure how critical the model is.
Which criteria?
This is from L. B. Cebik. "Basic Antenna Modeling: A Hands-On Tutorial",
page 2-9 published by Nittany Scientific; "Among the most important
conventions to adopt is to begin at one end of each antenna element and to
proceed from that end to the other without changing directions in
mid-stream". I have seen array pattern reversals when this is not followed.
There will also be current discontinuities. To be honest, with such a high
segmentation this will probably not effect the pattern of such a stucture,
or for that matter the input impedance. Just for curiosity I will give it a
try, and see if it makes much difference. Also, although it should not
cause any problems, I would not have used a single segment for the source,
just the end segment from the upper element.
With the exception of the sleeve wires, the model follows this
convention and a trial with different end-to-end connections for the
sleeve gives identical results.
If someone doubts that the sleeve is effective or the model of same is
invalid, as I said before, remove it and place a parallel resonant
trap at the top of the "coax" running from ground to the bottom of the
antenna. The results will be very (but not exactly) similar.
With a one amp source, there will be a current standing wave on the
"coax" with a peak amplitude of approximately 1/2 amp. Changing the
height above ground changes this dramatically and the angle of maximum
radiation above ground changes dramatically as well.
Those wanting to spend more time with it can try adding wires to each
end of the sleeve, tying the wires together; changing the length of
the sleeve and re-resonating the rod, and so forth.
Because the top of the sleeve is a multiwire junction I prefer to use
a separate wire to hold the source.
Obviously I
cannot use the "Minninec" ground, so have substituted an average S/M
ground
with the coax end about an inch above the ground. I had assumed the
antenna
was for HF, so it is probably impractically high, causing multiple lobing.
The multiple lobes are what you should see and are exactly my point.
I had assumed the antenna was to be used on HF, still comparing models with
similar parameters will provide meaningful information.
A free space model might provide more meaningful results.
If you can operate your antennas in free space then they would be
meaningful.
True, but the pattern is much simplified. Even so, in free space, I do see
evidence of minor patter ripple. I have also read (Cebik again) where the
Mininec ground can produce eroneous results.
If you can't/won't buy Roy's fine EZNEC program, then may I suggest
4nec2 at zero cost or MultiNEC at nominal cost as alternatives to bare
NEC.
I use Nittany Scientific's (www.nittany-scientific.com) NEC-Win Pro, which
seems to be a fairly good implementation of NEC2. The program does contain
simplified (spread sheet) data entry, but I prefer to enter in basic code,
which is apparently not available in EZNEC.
Since I'm a long time client of Roy's and a beta tester for MultiNEC,
I use EZNEC with MultiNEC as a shell. I get the best of both worlds
and MultiNEC will also invoke Arie's fine program, which I use for the
neat full-color 3-D plotting. EZNEC keeps me honest with all of the
segment length checking, antenna viewing and other fine features.
MultiNEC offers full spreadsheet entry, and other features too
numerous to mention. It writes EZNEC input files just dandy. It will
do the same with your Nec-Win.