A cheap 40M DX machine: the twin half-square array

Written by admin On August - 19 - 2010 0 Comment

The array's geometry depends on some well-placed supports such as trees.

Overview — the basic half-square

A simple half-square antenna for 40M is cheap and easy to build. For the CW end of the band (7.050 Mhz), start with one piece of wire about 36′ long. That’s the first leg. Then cut a piece 70′ long (the horizontal top wire) plus 36′ long for the far leg.

Find two trees, and hang the wires so they look like a big staple in the air. Feed it with 50-ohm coax — the shield connects to the top of the first leg, and the center conductor to the top horizontal wire. Check the SWR, make the legs longer or shorter to tune the antenna, and you’re ready to go.

The bottoms of each leg can be any distance off the ground — I’ve had mine just a few inches, all the way up to 10 feet off the ground. The whole thing can fit under branches just 37′ in the air, if that’s what you have to work with.

And if you can’t go that high, the bottoms of each leg can be bent inward in an L-shape without any significant change in the antenna gain or pattern. Rudy Severns wrote a great article about the half-square for the1996 ARRL Antenna Compendium, explaining the effects of bending the leg elements.

I arranged my single half-square so it fired broadside in two directions: the US and JA. It worked very, very well.

Twin half-squares — doubling up

Following advice from the late LB Cebik W4RNL, I added a second half-square 20.4 feet away from my original US/JA half-square, and used coax stubs to each antenna to make the whole array reversible.

See LB Cebik’s article, “Two to One” (free registration required)

The design employs 11’5″ stubs of 50-ohm coax connected to each antenna (top corner of each). The stubs are switched with a pair of DPDT relays, so when one half-square is fed via one of the stubs, the other stub is shorted at the relay box, turning that antenna into a reflector.

LB didn’t specify how to wire the DPDT relays, but on the left you’ll see how I wired mine. It’s not complicated, but it is easy to get confused with all the wires soldered to your relay terminals. Take your time and check everything twice three times.

The theoretical gain of the array is better than 6 dBi at 17 degrees takeoff angle. That gain might not seem like a lot, but compare it with a single half-square, which is by itself a solid DX antenna and offers just 3.3 dBi.

The two-element array offers a vertically polarized signal with nice low angles — and below about 20 degrees, the gain is actually better than a two-element yagi (though not a super-high yagi). I was able to switch this array from the US-Caribbean to JA and Asia.

Pattern switched to the east.

Pattern to the west.

August 2010 — a new location for the half-square array

In the spring of 2010, I took down the US/JA half-square array as it was too close to my SteppIR 40M dipole and 80M 2-element vertical array. This was the part of a domino effect that began in Jan. 2009, when I found that a nearby 40M EU raised-vertical array severely skewed the half-square array’s pattern.

In an attempt to improve my chances of working Europe on 40M in the fall contests, in August 2010 I returned to the great twin half-square concept. This time, the array is down the back yard in the pine woods. I was fortunate to have trees in the right places to arrange the elements so they fire at Europe and New Zealand/W6.

The location down in the woods is far enough away to avoid interaction with the SteppIR dipole, though it is actually now between the vertical elements of my 80M array. Time will tell if that’s a problem, but so far no SWR changes in the big antenna, and the half-squares are playing well.

November 2010 — interaction with 80M vertical array

After several weekends of trekking down to the half-square array in the woods, I have determined that the array doesn’t work well in my location. The reason has nothing to do with the half-squares. It has everything to do with the 80M wire vertical array beside it.

I believe the interaction with the 80M array made the half-square array, and the twin vertical array before it, poor performers.

Using the Autek VA-1 RF analyzer, I was able to measure the feedpoint impedance of the half-square array when pointed either north or south.

The north half-square is bang-on 50 ohms at 7.050 Mhz, with nearly zero reactance. Pretty darn good piece of wire. If I removed the south element, I’d have a pretty great but low-gain single half-square to play with.

Just 20′ away to the south, the south half-square, however, has its horizontal wire just a foot away from one of my 67′ tall 80M wire verticals. It has at least TWO resonant frequencies, about 200 khz either side of the 40M CW band. Changing the length of the half-square’s tails does shift the resonant frequency, but not by the amounts you’d expect (i.e. adding 5″ to each leg should move the resonant frequency down by 100 khz. It doesn’t, in my case.) It is simply too well-coupled to the adjacent 80M wire.

So, I will have to try yet another location for the half-square array. But not this winter.

Instead, I am moving to the next experiment to get 40M action over the pole: a 2-element 40M quad, hanging from the same rope that’s currently supporting the half-square array. The quad loops will be quite a bit further from the 80M verticals, so here’s hoping there is far less interaction for either of the antenna systems.

Additional thoughts on the twin half-square array

If you can put the twin half-square array in the air far enough from clutter and other vertical metal, do it! This may be the cheapest, lowest, no-radials-required, no fancy networks, way to get great gain on 40M. If you have some rope, trees, junkbox wire, a couple of DPDT relays, and some coax laying around, you have all it takes to build this antenna.

It worked extremely well when I had it aimed at the US/JA, and I would do it again in a heartbeat if that part of the yard was still clear of antennas.

Antennas, DXing, Uncategorized

Enjoy this Post? Share or cast your vote!

E-mail RSS Del.icio.us Digg Reddit Twitter

You must be logged in to post a comment.