Wednesday 23 May 2012

The 'Snowflake' That Melted

If you've read any of my stuff, or had a QSO with me, then you'll know I'm a very big fan of loop antennas. So, I was very interested indeed to come across David Shortess's (W7PTL) design for a convoluted full wave loop antenna, a PVC pipe variant of the Pfeiffer quad antenna.

My very beautiful - but very deaf - 2-element Pfeiffer quad.  Needs more work...


David's design is squarely aimed at my view of life - antennas that are cheap, practical and made of commonly-available materials.

David's instructions are very clear, even for a pleb like me. In fact, David was kind enough to lend some insight and help when I was putting my loop together.

I had all the PVC pipe for next to nothing money. The junction pieces are significantly more expensive, but the glue is cheap enough. The wire can be just about any old junk.

A side view.  Note the old cement mixer stand - the most useful thing you'll find for amateur radio!

I worked quite hard over a couple of months in the depths of winter to get what my kids now called the 'snowflake' antenna (I think the name should stick!)

 One fine day, I made a simple aluminium boom with shelf brackets as mounting points (a leftover from an abandoned quad loop - have you seen how big these things are, even for 10m?) I strung the loops quite easily onto the boom with cable ties (ty-wraps in American English!) and connected up.

I couldn't quite believe my ears - nothing! Not a dicky bird. A two element Pfeiffer quad was as deaf as a proverbial dormouse. Quite why is a total mystery, and I have yet to test just one loop on its own, in case some bizarre interaction was going on.

I'll have to work a bit more on working this loop out, because it's very compact and potentially very useful. It's even very wind resistant, which is a must for this QTH.

 Even so, it's still a loop that needs mounting up high, and that is always much, much harder than books and magazine articles ever admit is the case. Getting a wire antenna up to 30 feet is hard enough; just try it with a substantial structure like a two-element beam!

Getting Bigger...

Having discovered the almost mythical powers of a delta loop antenna - a simple triangle of wire - I decided I'd venture out into the ocean of activity that is the 20m band.

The first variant of a 20m delta.  The base was only 0.7m off the ground, but it worked global DX very well.  The tension put into the pole with this sloping arrangement is actually very good for high wind resistance, but it restricted the garden space too much, so I opted for a 1m-taller pole and a non-sloping delta, which has to be retracted much more often in windy conditions.

First, I had to get some wire.  Most folk like to use old bits of wire and old domestic wiring to make their antennas, which is very much in the low-cost spirit of this blog.

But here, wind is a big problem, so something tough is required.  The wire also needs to be pretty light, because fishing poles, though very useful and cheap, are useless at supporting anything more than a handful of grams' weight.

That's why, with my wires, I do splash out on kevlar-cored wire.  This is exceptionally robust.  It starts with a tough plastic outer, which is in itself extremely strong; I've tried to pull it apart with two pliers but never managed it.  Then there's a braided, very lightweight wire beyond this, and then a substantial core of kevlar fibre that you can pull a car with (no, don't try it on a main road!)

Kevlar-cored wire.  Light, very flexible, super-strong and durable.  A bit pricey, though!


Each metre of kevlar wire weighs-in at about  11.8g.  Compare this with plastic-coated Flex Weave, also a very robust wire - which weighs very nearly three times as much at 31g per metre.  Enamelled solid copper at about 16 SWG is about 17g per metre, but very much stiffer and so much more difficult to work with for large, retractable loops.

So kevlar, whilst expensive at between 85p - £1 per metre, is extremely durable, wonderfully flexible and extremely light.  So I feel it's well worth the money.

The final version of the 20m delta has no slope, and is about 1.2m off the ground at the base.  This experiment with a 3m base proved too unstable in high winds due to the bendy nature of fishing pole top sections.  A stiffer pole should be fine.
I didn't find a 20m delta easy to control in terms of SWR.  Modelling suggests a resonant 20m loop should be about 21.3m or so in length, with a native SWR in excess of 3:1.  The exact length always depends on the environment, but the modelling proved pretty accurate.

But my fairly simple, discrete-settings inductor ATU doesn't do such a good job of matching the loop; at peak power, it's getting an SWR of about 1.4:1.  With a coax feed and 4:1 balun, the losses are higher than I would like; I expect that only about 85W is being radiated with a transmitter output of 100W.  You could use an automatic ATU at the feed point, and it would undoubtedly match the antenna, but at quite what price in terms of losses, I can't say.  Auto-ATUs are also pretty expensive.

The QSL card proves the delta works on long path!  Image courtesy of my friend Paul Simmonds.


From my portable operations experience, I've found a 20m delta fed with 300-Ohm 'ladder' line is a very much better option than coax-and-balun.  Even at high SWRs, twin wire losses are extremely low - essentially nothing in comparison to coax.  My ATU with a twin wire still gives about 1.4:1 - 1.5:1 SWR at 100W with the delta, but the transceiver is perfectly happy with that.

Not bad for a pocket money antenna and 100W.  Thanks to Jim for the lovely coin!

And, best of all, a twin-fed delta can be matched to any band above its design wavelength (though this can be achieved quite readily and successfully with a 4:1 balun as well - I've worked very good, 'real' DX with such an arrangement.)

Erwin's not only far away on short path, but almost always has the biggest pile-ups you can imagine.  The 20m delta has made it to Erwin on a couple of occasions by now.  Image courtesy YB0DJ (and I have the real QSL card to prove it!)


Those of a more yagi bent will say that wire loops are all well and good, but lack gain and are noisy.  Well, you have to be careful how to compare your fruits.  A yagi for 20m is not a structure for the fainthearted.  You'll need a pretty robust (read: expensive) support structure, a means of rotating and some good guying, especially for a windy location like mine.  You can build a 20m yagi very easily, but it will be big - a loaded version will have elements at least 5m long, and a non-loaded version much longer.


And, as this lovely video shows, even with eight burly men at the controls, getting even a modest, loaded yagi into the air is no easy task - by any means:



True, a yagi has very good gain, made better by rejecting the rear and sides.  But that's not always a good thing.  In fact, having got used to hearing signals from every direction (deltas are, though, very good at rejecting high-angle, local signals), I think I might find using a beam a bit of a frustration; I do wonder how many unexpected contacts I might lose with a yagi compared to how many I gain with an omnidirectional delta.

Here's a nice comparison of a G5RV antenna against a delta loop, which very much reflects my own experiences:



And here's another comparison - a commercial vertical (cost: in excess of £450) against a delta loop (cost: pocket money):



Next, we're off to the beach for some portable delta loop working...