Danny Richardson wrote in
:

....
Works okay for me. Nice paper ( I stole the NEC deck).

Hi Danny,

You will see that the FR card has a symbolic variable for the frequency,
you will need to feed it a float value. (I used a PERL script to run a
series of models by tailoring the basic NEC deck, executing it, and
summarising the results.)

Differently to many of the commercial Cobras, my model used 2mm dia
copper and wider spacing. Most seem to use smaller, even very small
spacings where proximity effect becomes significant, smaller conductors,
and possibly steel cored copper clad conductors.

IMHO, the interesting aspect of the article is the loss mechanism of
linear loading, it makes sense once you see the current magnitudes and
phase and realise this thing at some frequencies carries large conductor
currents for low net current moment.

I have been exploring a 13m unloaded vertical with 2 radials for each of
80, 40, 30, & 20 metre bands, and there is a different but similar (!!!)
effect where at some frequencies the radials carry large individual
currents (larger in magnitude than the current in the vertical) with a
peak in loss at these frequencies, see
http://www.vk1od.net/multibandunload...al/13mEV-1.gif .

Owen

On Thu, 26 Apr 2007 03:34:29 GMT, Owen Duffy wrote:

Danny Richardson wrote in
:

...
Works okay for me. Nice paper ( I stole the NEC deck).

Hi Danny,

You will see that the FR card has a symbolic variable for the frequency,
you will need to feed it a float value. (I used a PERL script to run a
series of models by tailoring the basic NEC deck, executing it, and
summarising the results.)

[snip]

Owen

No problem Owen, I'll use MultiNec which very easily allows me to run
"test cases" using variables with no programming required on my part.
G.

Danny

Danny Richardson wrote in
:
....
No problem Owen, I'll use MultiNec which very easily allows me to run
"test cases" using variables with no programming required on my part.

Ok. There are other tools that allow symbolic variables, and of course NEC
allows specification of a range of frequencies on the FR card.

The main reason for using PERL to build, run and sumamrise is the summarise
stage. At that point I merge the NEC output with real transmission line
models, L-tuner models etc... and without wading through 5MB of NEC output
by hand. The system view is a bigger picture than the NEC output.

Owen