[TowerTalk] Antenna Modeling Services

jimlux jimlux at earthlink.net
Sat Apr 7 21:11:31 EDT 2018

On 4/7/18 5:30 PM, Grant Saviers wrote:
> I agree with Wes, you can make a lot of progress with the tool you have 
> plus AutoEZ.  The iterating on designs to understand the issues and 
> optimize for what you want is very time consuming. Dozens of hours, 
> maybe 100s.  More than that for the arrays I modeled and discarded.  At 
> skilled engineer consultant rates that's going to be expensive vs an 
> commercial antenna or proven handbook design.  If I were such a person 
> in a business I'd rather take a fee plus a piece of the action, one 
> off's don't sound so attractive IMO.

Relatively few people are looking to engineer arrays as a product, so 
I'm not sure a piece of the action is worth looking at.  If I had to 
guess, I'd say that *all* HF arrays are "one-off".

Well, maybe not the sort of modular thing being used at Issoudun - 
that's a sort of product line.  But anyone building a 500kW broadcast 
station is probably going to do one-off enegineering of all kinds<grin>

> It will also require an EZNEC/NEC commercial license as a paid 
> consultant is a business.  That rules out most if not all folks on 
> towertalk (unless a commercial licensee pops up).

If it's verticals, and you're not looking to model ground currents or 
buried radials (which would require NEC4), then NEC2 would work just 
fine - and NEC2 is essentially "license-free" (EZNEC is not, of course).

I'd say that trying to model buried radials on an array of verticals is 
probably not worth doing - NEC4 doesn't support variations in soil EM 
properties under your array, and unless you're like N6RK or the VOA 
Delano site with a flat piece of farmland out in the middle of other 
flat farmland where the topsoil is thousands of feet deep, the 
variations from soil are bigger than your modeling error.

What I would do is model it with several different soil conditions 
ranging from perfect to poor and see what the difference is, i.e. do a 
sensitivity analysis.  As you've pointed out that's the real value of 
modeling - you can see if changing something by 6" changes things 

I like R.W. Hamming's comment: "The purpose of computing is insight, not 

More information about the TowerTalk mailing list