Taming a 3,5 MHz Yagi

Last updated 2012-04-02

Notice: © 2001 to 2012, Chris R. Burger. This document may be reproduced as required for personal use, and may be freely referenced from other Web sites. However, publication elsewhere requires express written permission from the author.

This article is based on something I wrote on the Towertalk reflector in 2001. Towertalk was moderated by Steve Morris K7LXC. I've changed a few phrases to take account of subsequent developments and a few changes in preferred terminology.

Some History

When I started contesting in the early Eighties, even 7 MHz beams were not that commonplace. The few big guns who used them were loud --PY5EG, I2VRN and a few others.

As they proliferated, my awe-struck teenage eyes shifted to 3,5 MHz beams. There were but a handful again--I5NPH, OH1RY and W2HCW spring to mind. I promised myself at age 20 that I would have a rotatable beam for 3,5 MHz before my 40th birthday.

Fortunately for me, N6BT and others came up with some design philosophies that helped to bring it within reach when I reached my mid-thirties. Thinner diameter elements, using thick walls and high-grade materials, had reduced the wind load of practical 3,5 MHz elements to the point that a much flimsier tower could hold one aloft.

I remember the legend that a 3,5 MHz beam is a life-long maintenance commitment. However, I'm a firm believer that if things are done right, they will survive. Doing it within an amateur budget is another matter, though.

With support from Tom Schiller N6BT of Force 12 antennas, I finally got my own 3,5 MHz beam flying around my 34th birthday. I was a Force 12 Magnum 2/3, with two elements on 3,5 MHz and three elements on 7 MHz, collocated on the same boom. Although it's loaded on both bands, it's still a monster by most standards. The boom is something like 15 m long; boom diameter is something over 150 mm, and the whole thing weighs over 100 kg (50', 6" and 220 lbs if you are metric-impaired). However, because of its loaded elements and relatively light weight, it doesn't demand a megabuck tower. In my case, my tower is 42 m (138') tall, with this monster at 36 m (118'). The tower has a face width of over 1 m (3'), and weighs in at over two tonnes. It is guyed at three levels and in three directions. It is not a toy by any standards, and really taxed my engineering design, construction and erection skills to their limits. The tower is still standing sixteen years later, with no mechanical maintenance.

Tuning this beast was an overwhelming undertaking. Each 3,5 MHz element went up and down over a dozen times, to get the four band segments tuned exactly right. The entire beam went up and down three times before the feedpoint was properly matched. The whole process took two guys (me and ZS6AWK) around 20 hours. This doesn't sound like much, unless you try and squeeze it into your work schedule! I couldn't get the 7 MHz portion working at all; resonance was easy to achieve, but at high SWR. Low SWR was easy to achieve, but not in the band. Because of other demands on my time and because the availability of another 7 MHz beam made it a low priority, I never did solve the problem.


Using the thing was a lot of fun. Signals that could not even be imagined on the wire antennas and the loaded tower were audible. Given our high QRN level, almost anything that was audible was also workable. My biggest surprise came when I was calling several audible mults after sunrise in CQWW CW. I suddenly realised that it was broad daylight. The beam could still hear signals at least half an hour after the other antennas had lost the last ones.

The first problem came with the rotator. The OR2800P simply broke under the sideways load imposed by the cable drive. The cable drive was based on a W3LPL idea. The manufacturer blamed me for exerting an excessive side load. However, the calculated overturning moment caused by the tension in the cable at maximum torque specified for the rotator (well in excess of the actual torque exerted by the antenna) was much less than the overturning moment caused by the rated wind load at the top of the mast clamp. In practice, any beam would be mounted higher, causing greater overturning moments. In other words: The force exerted was much less than they claimed that the rotor could comfortably handle. The brittle fracture appears to be due to a casting fault.

Modifying the tower for a chain drive took considerable creative engineering, as a split sprocket had to be machined to fit around the main mast without removing the mast. Fortunately my friend Tjerk Lammers ZS6P had a good machine shop and was willing to help. The mast above the rotator was given additional support to absorb the overturning moment, in the form of a thrust bearing placed above the drive sprocket and the rotator.

After fixing the rotator, I had more than two years of service from the beam. It was a wonderful performer, both on Phone and CW. I estimate that I had less than 50 hours of use all told, given how inactive I was.

The beam suddenly died one day, and then resurrected itself intermittently for some time. Fixing the problem involved taking the beam down and replacing the balun. The original Bencher BY1 had unsoldered itself, despite relatively modest power levels. In case you're wondering, I've never used tubes with handles! I've learned a lesson here: Don't try and save on baluns. Having to haul a low-band beam up and down is worth more than just a fistful of dollars. After fixing the balun and hauling the antenna back up, the resonant point was about 10 kHz low. Clearly, more work would be required.

I never got to fix the resonance problem, as other demands required my attention. Eventually, the control cables to the relay boxes started dropping off as cable ties hardened in the sun. Ultra-violet radiation at our altitude (1500 m or 5000 feet) has its drawbacks. Copper thieves also came and collected some of the control and feed lines at ground level.

As of 2012, the beam is still up. It seems mechanically sound, with some play in azimuth. The beam rotates by a few degrees to and fro in heavy winds.

Would I do it again?

Was it worth it? Yes and no.

On the one hand, I just had to try it, and there wasn't another one around to try. Like it or not, I had to build it myself. I also had to keep my promise to myself. It really was an awesome performer. For someone who lives in Thunderstorm Alley, actually hearing modest stations was quite an experience.

On the other hand, given the magnitude of the project, it's hardly a sane decision. Paying that price for something like on S unit over a vertical is not a reasonable decision. If you really need the gain, a vertical array will give you the same thing with instant direction switching, and far less cost. The vertical array's main disadvantage is the space required, but even that disadvantage can be ameliorated to some extent by using close-spaced triangular arrays and creative radial arrangements. Perhaps the only significant advantage of the Yagi is its horizontal polarisation, which helps a little with the noise level.

We all know that these decisions are not taken fully on reason, though. You can't knock a low band beam unless you've tried it!

To summarise: I had two failures in about three years. The beam is still up after 13 years, so presumably with a bit of TLC it could have remained in service.

The rotator failure was probably due to a faulty shaft, but could have been avoided by using a three-bearing arrangement such as is in place now. The other failure was due to an unreliable balun, something that could have been avoided if I'd had the luxury of walking into a dealer's premises and touchy-feeling a few different types. I had to order blindly from paper, possibly attaching too much significance to price. I guess if I'd had the benefit of experience with either of those subjects before the installation, the beam might well still be operating. Whether this kind of reliability could be achieved in less benign climates, though, is anyone's guess.

Return to ZS6EZ's Radio Page