[TowerTalk] Re: Accuracy of models

[Jim Lux]

----- Original Message -----
From: "Tom Rauch" <w8ji at contesting.com>
To: "Jim Lux" <jimlux at earthlink.net>; "Jim Brown"
<jim at audiosystemsgroup.com>; <towertalk at contesting.com>; "Pete Smith"
<n4zr at contesting.com>
Sent: Tuesday, April 13, 2004 6:08 AM
Subject: Accuracy of models


> Jim,
>
> This is a point that gnaws at me:
>
> > By the way, this whole RF impedance of the soil thing is why a lot of
> those
> > ground radial analyses leave something to be desired.  They tend to rely
> on
> > DC analysis, and ignore things like capacitive coupling from wire to
> ground,
> > etc.
>
> There was a recent article about optimizing a radial system based on the
> soil. The article described for a given amount of wire how it should be
> distributed for optimum performance.

I have some analytical quibbles with the QST articles.  For one thing,
optimizing the length is not necessarily telling you useful things,
especially when you constrain the problem to arrays of equal length,
equiangularly spaced radials. What you really might want to do is optimized
based on mass of the copper (is few short fat wires better than lots of long
skinny ones?) or on the time to install (probably length is a good variable,
then, independent of wire diameter). You'd also want to consider branching
arrangements or ones with varying lengths. Admittedly, you've got to start
somewhere, though.

 The recent article was based on analysis (in an earlier set of articles
from QEX or QST, as I recall) originally deriving from Brown, for 1 MHz,
which is a fairly different animal from, say, 14-30 MHz, where the
dielectric properties become more important than the conductivity.


>
> What is your opinion of following those or similar guidelines?
>
> Also, I'm still looking for a coil Q program for HF loading inductors.
I'll have some links later this afternoon.