[Top] [All Lists]

Re: [TowerTalk] Taper changes using EZNEC+

Subject: Re: [TowerTalk] Taper changes using EZNEC+
From: Jim Lux <>
Date: Fri, 11 May 2012 08:08:50 -0700
List-post: <">>
On 5/11/12 7:29 AM, wrote:
> I would like to change tubing diameters on an antenna that I have
> previously designed.  Is there a way do this and maintain the same
> element resonate frequency using EZNEC+?  I tried "Show Stepped Dia
> Correction" thinking it would be easy see how much longer or shorter the
> new element is, however there seems to be no way to make the new and old
> correction diameters identical.  It would be nice if I could set a
> standard taper correction size. Any suggestions?

Isn't that taper correction an empirical formula? (from Dave Leeson W6NL 
nee W6QHS)  It converts a specific taper schedule into an equivalent 
constant diameter and length, with the calibration done by comparison 
against a real antenna.

So, if you have Dave's book (or the algorithms in there), you can 
probably figure it out.

It's possible that someone has done a numerical model to generate other 
cases (without needing actual range tests to confirm), but I haven't 
seen one.

Joe Reiser (W1JR) has a write up on Yagi Antenna Design in 
Communications Quarterly (google for it) which describes all the various 
correction schemes that have been used. Some were empirical, some are 
based on theory (W6NL's), etc.  Others have their own schemes for taper 
correction, and they all typically work better for some more restricted 
  problem space (e.g. K1FO UHF designs).

You could also model the tube, not as a single wire, but as a circular 
array of wires, which overall are tapered.  The usual rule of thumb from 
Jerry Burke on modeling a "big" thing with "small" things is that the 
surface area of the wires should match the surface area of the thing 
being modeled.

So if you have a 2 cm diameter tube (surface area = 6.28 sq cm/cm 
length), and you want to model it with 8 smaller wires, the smaller 
wires should be 1/4 cm in diameter, spaced on a 2cm diameter circle.

Lots of segments in the resulting model, of course.
Some sort of program or spreadsheet is handy to generate the gazillion 
segments, or make strategic use of the GM card to replicate and rotate 
the copies.

If you have access to NEC4, it does a better job with discontinuities in 
the model, particularly if you have very short segments joining others 
(NEC4 uses a different set of basis functions than NEC2 for each of the 
segments, so the numerical precision problem is less)


TowerTalk mailing list

<Prev in Thread] Current Thread [Next in Thread>