The Large and Powerful Spider Beam

Written by admin On January - 23 - 2016 Comments Off on The Large and Powerful Spider Beam

The Large, Powerful Spider Beam

Originally written in May 2003

Just in time for WPX CW 2003, I finally got my new homebrew spider beam up in the air.

This antenna, invented by Con DF4SA, offers full-size yagis (with directors and reflectors in V configuration) interlaced on a 32-foot boom. You get 3 elements on 20m, 3 elements on 15m, and 4 elements on 10m. It’s a powerful package.

My construction method for this antenna was identical to the way I built my hex-style beam last summer — a PVC pipe serves as the center post, which runs up through the middle of a circle of plywood. The spreaders are U-bolted to the plywood.

Got everything I needed, except the fishing poles, from Home Depot.

Parts List

4 — 16.5-foot-long Classic Crappie Poles from Cabelas.com
$9.99 each
1 — round 1-foot diameter piece of 1″ plywood
1 — can of black acrylic “plastic’ spray paint
1 — roll of 3-conductor “Romex” house wire
insulation stripped off with a utility knife
1 — roll of 30lb-test nylon monofilament fishing line
1 — roll of bright white nylon cord (mason line)
8 — stainless steel U-bolts (1.75-inch)
4 — steel “L”-shaped shelving brackets
1 — 4-foot long piece of 1.5-inch PVC tubing
4 — stainess steel screw-in eyebolts
2 — brass machine screws, plus a nut and wing nut for each

Assorted miscellaneous bits — some crazy glue, doweling, old broom handles, screws, etc.

A Shorter Spider
The Cabelas fishing poles are great — five pieces that nest down into the largest 3.5-foot section. I found the smallest two sections too whippy to hold a straight line, so I took them out. To make up for one of those removed sections, I built 3-foot add-ons of 1″ PVC pipe reinforced with old broom handles (they fit perfectly).

These extensions were inserted into the butt-ends of the fishing poles, and also into the hub. They readily handle the compression of the hub’s U-bolts.

So, the end result was that I had spreaders of just 13.5 feet, not the 16.5 feet called for in Con’s very wonderful design.

The great news is that all the elements still fit into this shorter design, except for the 20M driven (dipole). I decided that, for me, an acceptable trade-off of size vs. performance would be to linear load the 20M dipole. This was done by winding about 1 meter (OK, 3 feet or so) of each side into a coil wrapped around 3-inch-long pieces of 1″ PVC.

I used a bead of 5-minute epoxy to keep the wire nicely shaped on the forms. This made the 20M driven element fit my spreader “wingspan” perfectly. Again, every other element was cut exactly to Con’s measurements.

I experimented with element spacings to accommodate the shorter spreader lengths, looking for best SWR and best perceived receive strength. Not scientific at all, but I am happy with what I ended up with.

(See the updates below for information about how I extended the spreaders and removed the loading coils in August 2003).

Anecdotal Performance Info
I don’t yet have a tower (one day…) and can only get my antennas up about 27-30 feet. Too low to really get much from a yagi (or a dipole for that matter), but I have had a 5-band hex-style beam on that same pole from October 2002 to May 2003 and was able to compare it to the spider on the back lawn (3 feet high in the test stand).

The hex and spider were very close, but now that the spider is at 27 feet, I am already hearing stations I’ve never heard before from this QTH.

Con’s design had me hooked from the second I saw it on his Web site. I waited anxiously for the day that he would release his design specs, and when they arrived I started gathering all the parts to build one. For the past four weeks I puttered around, building the thing and adjusting, etc.

Tonight, after work, it took my wife and I over an hour to get the antenna up off the lawn and onto the steel pole I use, bolted to a lower eve of the house.

The moment it was up, I ran to the shack to compare signals with my G5RV. Heard a KH7 station at 20 over S9… and down at S5 with the doublet. Other signals confirmed the differential between antennas. No comparison, and nor should there be — this cheap, simple wire beam is doing its thing. Can’t wait for WPX to start at 0000z tomorrow to see what I can hear, and what I can work.

Update: Here’s how I fared

CQ WW WPX CW Contest results
Year         Score    Qs  Mults  Antenna

2002      9,000    63     45  G5RV at 30′
2003    268,348   412    233  Spider beam <30′

See my writeup about the 2003 contest

Con, thank you for dreaming up this design — and sharing it with us! Although I had to modify my version to suit the parts I had available, I think many, many hams will enjoy building a spider beam yagi as I did.

Update: August 2003

With the A-index at 65 or so, and a severe solar event under way, I thought it would be a good time to update my spider beam story.

I built the all-homebrew antenna and put it up in late May. Last week, I took it down and made some modifications.

Initially, I used 13.5-foot spreaders (made from fiberglass “Crappie rod” arms from Cabelas.com) instead of the full-length 16.5-foot spreaders in Con’s design. That meant shortening the 20M driven
element by several feet, and using closer-than-recommended element spacing.

To make the 20M dipole short enough to fit the spreaders, I installed coils on PVC forms just out from the feedpoints. I thought it was a good trade-off between size and performance. I was wrong. The 20M performance was not very good, even compared to my G5RV at the same height of just under 30 feet (10M).

With that original configuration, both 15M and 10M were very good, with great forward gain and good side and front-to-back rejection. On 20M, despite a 1.5:1 SWR from 14.000 to 14.100, signals were clearly attenuated — and the loading coils were my prime suspects.

With the help of my wife and sons (7 and 9 yrs.) [2016 update: one is now in his third year of university, and the other is at law school… dads, don’t blink or you’ll miss the best years with your kids], I eventually managed to take down the antenna and get it into a WorkMate on the back lawn to work on it.

The experience of putting it up, taking it down and putting it up again proved to me just how resilient this design really is — it snagged on maple brances, snagged on the cedar roof shakes, snagged on the eavestroughs, balanced almost vertically resting on just one spreader… and nothing broke! I think it’s the German engineering, hi.

Once it was down, my first task was to extend the lateral (driven element) spreaders. I am using telescoping fiberglass rods, which initially came in five nesting sections. The very thinnest 5th section (.4 inch diameter) was removed at the start, and the 4th section was simply pushed into the other sections.

On the lateral spreaders only, I fished out that 4th section and used it to extend the spreaders to the full 16.5-foot length. That allowed me to uncoil the 20M driven element wire so it returned to a full- sized dipole.

The result of this extension? A big, very noticeable improvement — the antenna really shows up the G5RV in many instances.

Since using the spider beam, my contest scores have gone up slightly over last year, despite declining conditions. My station is still a peanut whistle (100W max, and no tower yet), and I am limited by a huge RF filter (mountain) blocking quite a bit of eastern horizon from our back yard, but every dB of gain helps to overcome the QTH deficiencies.

So, I now have 13.5-foot boom spreaders, and 16.5-foot lateral spreaders for the driven elements. The element spacings are still shorter than Con’s design, but not too short to be effective.

Next time the antenna is down, I will make the 15M element a bit longer as it has 3:1 SWR in the CW portion of the band (although the tuner easily tunes it down to 1.2:1 or so). The 15M resonant frequency appears to be about 400 khz too high.

Update: 2004

On an unseasonably balmy day in December 2003, I completely re-did the element spacing to conform almost exactly to Con’s original design dimensions. Even used the proper guy string layout.

The fiberglass spreaders were a bit too whippy for the guying to work properly, but the antenna was solid. Flexing allowed the 20m reflector to tangle a bit in the end of the 15m reflector’s fishing line, but it’s nothing major.

In March 2004 I took down the spider beam in preparation for moving to a new QTH.

At the new house I put up the MA40 tower with a Mosley CL33 triband yagi on top. In 2009, the Mosley was replaced with a 3-element Steppir yagi with 40M dipole element.

 

Antennas, Contesting, DXing, General

Enjoy this Post? Share or cast your vote!

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

Comments are closed.