[TowerTalk] Re: 80 Meter Hybrid Triangle Loop

Tom Rauch w8ji at contesting.com
Sun Nov 7 08:10:31 EST 2004


> Thanks for all the input so far on the loop. I
> neglected to post the basis for my loop project. I
> have read a lot of information from Cebik and others
> on loops. The main article that intrigued me for this
> project is as follows:
>
> http://www.cebik.com/fdim/fdim5.html
>
> "Do the VOMBA! Vertically Oriented Multi-Band
> Antennas"

First three general comments not specifically about the
loop:

While the models are interesting, there are a few
shortcomings. First, pattern alone isn't very useful unless
you are planning a receiving-only antenna.
People often focus far too much on TO angle and where the
main lobe is.

What we really should compare is the gain at the absolute
angle and  direction where we want to work, not where the
peak is. I've watched friends actually give up gain at low
angles because they don't want a high TO angle. My advice is
to ignore any information considering pattern only, unless
you convert or normalize it indicate the relative gain of
other antennas you are considering.

The second shortfall Guy already partically addressed. We
should ALWAYS consider the feedline in arrays that have
"unusual" feed impedances. Even ladder line is not always
negligable loss when feeding astronomical SWR loads. Also
impedance zooms all over the place with frequency and
feedline length when SWR is high. Say for example I
terminate 400 ohm line (typical for 450 ohm ladder line)
with a  30:1 SWR because of load reactance. The input
impedance of the line would move between 40 times 400 and
400/40 in impedance.  That's a range that falls between 10
ohms and 16,000 ohms, depending on line length. This makes a
high-Q frequency sensitve antenna that is a bitch to match
on multiple frequencies, and it can also mean SWR is high.

If I ONLY model and antenna and don't consider the feedline
or power factor correction losses, even a 50 ft long dipole
looks OK on 160 meters!!!! It doesn't look bad at all
compared to a regular dipole. Yet everyone knows if
installed you'd be dog-gnat signal grade.

I'd view any arm-waving about any antenna with skepticism if
the input impedance has high reactance or a very odd
resistance. The feedline really needs to be included. ARRL's
TLA  should always be run on antennas with high reactance or
high SWR on the feedline. Always.

See: http://www.w8ji.com/short_dipoles_and_problems.htm

The third suggestion is to put the feedline in the pattern
models. We can model antenna all day and model wires and
make antennas that work but are impossible to obtain the
model performance because of common mode voltage or current
issues. It's easy to add the feedline (it can almost always
be modeled as a single additional wire with a load or source
to represent a balun).  The end-fed half waves like the PAR
antenna and IMAX 2000 are good examples of systems that are
feedline length critical for pattern. Even a 1/4 wl
groundplane pattern is affected by feedline length!!!

See:
http://www.w8ji.com/end-fed_vertical_j-pole_and_horizontal_zepp.htm

http://www.w8ji.com/verticals_and_baluns.htm

It is a waste of time to focus on pattern and frequency
agility when we leave out key players like gain and
feedlines. If we ignore feedlines and gain (loss) we can
design hundreds of antennas that work well on paper but
won't work in life.

 The section that I am using for my project is titled;
> "The Hybrid Triangle"

OK, but what in particular catches your fancy? Is it the
fact the physical shape that makes it easy to install?  Does
the pattern fit the needs of directions you want to work? Do
you want to use it on multiple bands?

I'm not trying to talk you out of it, but to understand what
you look for in an antenna. This is like a dating service.

73 Tom




More information about the TowerTalk mailing list