Interesting comments re 4-squares. I currently have up a "wire compromise in
the woods" 4 square for 80 that seems to work quite well. Here's what I have
In the center of the 4-square is a 69 foot R25 tower with non-conducting guys.
On the tower are two 15m yagis and a 6m yagi. My 4-square verticals are wires.
They go vertical for 48 feet and then head towards the top of the tower- like
inverted L's, except the horizontal part of the L slopes upwards. There are
ropes tied out to trees to keep them up. Radials are on the ground. I use a
Comtek hybrid coupler.
I detuned the tower by dropping a wire from about 20 feet up on a sidearm. This
loop is tuned to resonance with a capacitor. I see about a 1 ohm difference in
the resistive part of each vertical's Z with the detuning in vs. out.
As far as performance, the 4-square is always better than a dipole (nearly
flat) at 90 feet for dx stations. For the USA west coast the 4 square is also
quite a bit better than the dipole (the dipole favors NE/SW). Even for working
W1 from my qth (MS) I often use the 4-square.
The reason I chose inverted-L-like elements was that they are easy to make
exactly identical, and I get the maximum height possible. With this geometry,
the pull-out ropes going to trees don't have to all be at the same height.
In my opinion, if you are height limited, you want to stay away from elevated
radials. If you elevate your radials, you are giving up valuable vertical
distance which will further lower the feedpoint Z and increase ground losses.
I don't think the advice to model the central tower is very helpful. Modeling
the typical tower is nearly impossible. Just getting the effective diameter of
a triangular tower is nontrivial. Plus add stacked yagis,
insulated/noninsulated yagi elements on top, etc. Just plan on detuning anyway.
I thought that the hybrid-type controllers (at least the Comtek I have) were
really meant to work at 50 ohms. Then for a fullsize vertical, 36 ohms + some
ground loss is not too far from 50. Obviously they seem to work ok over a
TowerTalk mailing list