[TowerTalk] modeling monster stacks

[Jim Lux]

For those contemplating modeling multiple SteppIR things, etc., there's 
some useful resources out there:

http://www.ece.rutgers.edu/~orfanidi/ewa

is an online electromagnetics textbook which is quite useful.  Even better, 
Sophocles Orfanidis, the author, has included a pile of matlab programs to 
do useful stuff like calculate mutual interactions among elements in a 
Yagi, calculate patterns, etc.

However, given that modern computers run NEC pretty fast, you might be 
better off just modeling in NEC rather than using an analytical 
approximation.  There's a nifty program in perl that does parameter 
substitution into a NEC deck: 
<http://www.si-list.org/NEC_Archives/nec_param_util.Z>nec_param_util<http://www.si-list.org/NEC_Archives/nec_param_util.Z>.Z 
available from the NEC archives ( use the new URL: 
http://wwww.si-list.org/swindex2.html)

There's also a couple of NEC toolkits in Tcl/tk and python on the archive 
from W5GFE, but I haven't used them yet.

What would be wonderful is if someone could develop some generic pattern 
evaluation routines that would take a NEC output pattern and extract some 
sort of figure of merit.  This would, of course, require some way to 
specify what "good" is, and I think that's the rub.

With fully adjustable antennas, the pattern's not fixed.  You can change it 
to suit the conditions right now.  That's a different sort of design 
paradigm than the fixed pattern against a statistical summary (which is 
done quite well by HFTA).  The HFTA model (If I can be so bold as to 
presume on what Dean was thinking) is that you want to pick the height(s) 
for your one antenna configuration that is the "best overall".    But, with 
a fully adjustable antenna, you can change the elevation pattern (and the 
az pattern too).. using HFTA (or more accurately, the statistical arrival 
angle data) for planning might be like trying to pick the one best 
direction for your fixed-non-rotatable antenna.

I think one could come up with some suitable statistical analysis of the 
propagation data that would lend itself to an analysis of variable 
elevation patterns.  Perhaps, rather than working on matching the 
percentiles against the pattern, you could analyze arrivals to identify 
some "nominal cases" for various angles, evaluate that angle against a 
pattern optimized for that angle., etc.

Even more useful, you might want to do some interference suppression 
analysis. It might be more handy to use all that adjustability to knock out 
thunderstorm interference from a distance, or that strong SWL broadcaster, 
etc.

Jim, W6RMK