[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

Re: some more helix analysis (if you still care...).. & I Do!

Thanks, Scott.
Your simulation is very close to mine, at 36 and 52 degrees on the side lobes (for 18 turns).  The rtf files were great, but I don't think the BB passes them on.  I don't understand the 10.log[Eaf(i,1)^2] function ("array field"?).  The other two functions are self-explanatory.

I now have a bridge borrowed from W4MVB and will attempt to improve the match between the 4 helices before I do anything else, but my "un-matched" observations of the 52 degree lobe(s) is for a single 28-turn helix it was down 6-10 dB (as predicted), but for the quad arrangement at 1.5 WL spacing it was at parity with the main lobe and the quad had no more than 2-3 (perhaps less) dB gain over the single--certainly not the 5 dB predicted by your NEC 2 analysis.   

Onced matched, it will be easier for me to reduce the axial length (number of turns) of the helices than adjust the spacing:  my matching setup is based on using a fixed 1 wl matching section.  I'll report what I find.  What might make a whole lot more sense for collecting empirical data would be to produce 4 separate helix w/cup models and use coax matching sections and then simply rearrange the spacing.  Wish I had thought of that first :-(
Jerry, K5OE

Scott Townley <nx7u@mindspring.com> writes:

<< Hi Jerry (and anyone else who's interested)-
Don't know if you're still chasing helical arrays but I did a bit of
analysis that you may be interested in.
I ran a NEC2 analysis on a 18T helix w/o cup and got a pattern that
looks pretty much like you described (and showed in your NEC4WIN plot). 
It has pretty hefty sidelobes around the 33 and 51 degree marks, around
So, with array spacing of 1.474wl, the array null (there is only one) is
at 19.8 degrees (well inside the element sidelobes) and the array
grating lobe (again, only one) is at 42.7 degrees.  So unfortunately the
array does nothing to suppress the element sidelobes.  That explains
your observation of quite high sidelobes at those two angles.  The
"version 1" .rtf file has some calculations to illustrate.  BTW you do
get 19.2dBic out of the array but as we've discussed those sidelobes
won't help the G/T any.
Let's try to cut those sidelobes down, by *decreasing* the array spacing
so that the array null will cut down the element sidelobe (kind of
working the problem in the opposite direction, since usually in arrays
of highly directive elements you use the element to cut down the array
grating lobes).  At spacing 0.9wl (the "version 2" .rtf file) you get
good sidelobe suppression; all sidelobes look to be <-14dB.  A much
better situation; and you still get 19.2dBi out of the array.
Obviously the limiting factor in this implementation is the element
pattern.  Realizing a lower sidelobe helix will go a long ways towards
improving the end result.  And that's next to go out, if you're
I apologize for the somewhat choppy nature of the .rtf files but it's
the most efficient way I know of to get stuff out of Mathcad, and I
figure many more have MS word or other .rtf compatible software then
have Mathcad.

Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org