Stealth Antenna – Moxon Element Separators, 15m (top) & 17m (bottom)

Moxons in the Attic (Part 1)

A few months ago, I wrote an article on building a stealth antenna farm. Since I live in the land of CC&Rs, antennas must be “dual use” such as a vertical hidden inside a PVC flagpole, or low dipoles and inverted Vees hidden in trees. I spent many years as an avid contester and DX-chaser to appreciate the logic of stacked beams on towers to enhance the thrill of the hunt. Thanks to Layne, AE1N, I checked out the website of Jeff, AC0C (www.acoc.com) for some ideas of how to build a multi-band station in one’s attic without the condo association vigilantes running him out of town. Jeff spent countless hours crawling around his attic to construct multi-element antennas for 160 through 6 meters. Spurred on by Jeff’s success, I decided to explore the attic of my garage to see what I could do. As I described in the MAY Nashua ARC bulletin, I settled on building Moxon antennas for 15 and 17 meters.

An old adage about antenna building states that an antenna must be built-in lousy weather in order to work right. Thirty years in New Hampshire lent credence to this axiom as I spent many a cold, windy day on a tower doing antenna work. In Florida, a similar law applies: build an antenna in the summer months while sweating profusely rather than during the comfortable winter weather. Again this makes sense: DX and contesting fill up the winter months to have time to mess with antennas. It is also important to remember that, during such endeavors, you will become enamored with you antenna as you take breaks to warm your body (in NH) or drink a gallon of water (in FL), all the while cursing this law of antenna building.

But I’m getting ahead of myself. My garage attic is roughly 20 x 20 feet with an apex of about five feet that runs north-south. I had selected the Moxon design because a conventional 2-element beam would not fit in the space available. I elected to build a Moxon for 15 and 17 meters that would fit in the space available without having to encounter obstacles like the ventilation duct work. I was also fortunate in that my home is one of the older models that do not have foil-backed insulation inside the roof that creates a radiation-proof box. The joists junctions are reinforced with metal plates as part of hurricane building codes. My plan was to attach the wires to the roof trusses and stay away from these plates as much as possible to avoid interaction.

The 17-meter Moxon is a little over seven feet between elements while the 15-meter Moxon is about six feet between elements. I used a piece of half-inch PVC pipe as a template to mark the joists for the 17-meter antenna. The antenna is about 2.5 feet above the attic floor for a total height of eleven feet above the ground. I had modeled it at 13 feet so I figured it would be close enough. The 15-meter Moxon is about 15 inches above the 17-meter one. When viewed from the top, the antennas look like concentric rectangular loops.

Over the course of several weeks, I grunted, groaned and sweated my way back and forth measuring and installing the wires. I worked during the morning hours before I was soaked before 10:00 AM. I found myself wishing I could have my five-year-old grandson help me. He can stand upright and is plenty flexible to maneuver around the joists. While I did not have to worry about the obvious safety issue of working on a tower, I did at times feel I was a candidate for the NFL concussion protocol from bumping my head. I tried using my cycling helmet but it interfered with my headlamp. Another similarity to tower work is that I had to make N trips back and forth in the attic for stuff I forgot. This is, however, much more bearable that climbing up and down a tower to get what I forgot.

Figure 1 shows a view toward the south end of the attic. The two pieces of PVC form the element separators for the 15-meter Moxon (top wire) and the 17-meter Moxon (bottom wire).

Figure 1 – Moxon Element Separators, 15m (top) & 17m (bottom)
Figure 1 – Moxon Element Separators, 15m (top) & 17m (bottom)

Figure 2 shows the reflector elements for each antenna as secured to the joists, looking north through the attic. The white standoff fasteners are coax cable tie-downs that I found at the hardware store. Standard house wiring fasteners would have worked but they leave little room for pulling wires if I needed to make adjustments. (My first attempt was to use duct tape to hold the elements up. However, the heat soon made them droop.)

Figure 2 – Moxon Reflector Elements, 15 m (top) & 17m (bottom)
Figure 2 – Moxon Reflector Elements, 15 m (top) & 17m (bottom)

Figure 3 shows one corner of the director of each antenna looking east. The duct work to the left is part of the ventilation system while the open duct vents directly from the garage below. The yellow fence standoff on the upper antenna is the bend point for one end of the 15-meter director. Not visible to the left is a similar bend point for the 17-meter director.

Stealth Antenna - Looking East-from the Reflectors
Figure 3 – Looking East-from the Reflectors

Figure 4 shows the temporary feed points for each antenna.

Stealth Antenna – Feedpoint for 15m (top) & 17m (bottom)
Figure 4 – Feedpoint for 15m (top) & 17m (bottom)

The figures above show the project to date. I installed the 17-meter antenna first and measured its SWR performance with my analyzer. I found that it resonated beautifully at 16.7 MHz with a 1.1:1 SWR while bulging to 3.9:1 at 18.1 MHz. I shortened each element by a foot and ran measurements again, this time the resonant point moved up to 17.3 MHz (1.3:1 SWR) and the SWR at 18.1 MHz dropped to 2.7:1. I folded the elements back another four inches on each end and measured the response. I observed the SWR bottoming out at 1.5:1 at 18.1 MHz where I wanted to be. As a point of interest, I modeled a Moxon designed for 16.7 MHz and noticed the elements were about two feet longer than a Moxon designed for 18.1 MHz, close to the twenty inches I had to shorten the elements. Apparently, there is some interaction with the wiring that runs along the attic floor near the edges.

Armed with the satisfaction I was on the right track, I installed the 15-meter Moxon above the 17-meter antenna. I hooked up the analyzer and fired it up only to find to that its “resonant” point was a dismal 3.0:1 SWR at 22.9 MHz, rising to 3.9:1 at 21.1 MHz. This meant my antenna was too short. I went back to EZNEC, opened the standard dipole model and plugged in 22.9 MHz and found that its length was very close to the overall driven element for a Moxon designed for 21.1 MHz. I lengthened each element by five inches as a starting point to see what would happen. The result was no change in SWR at 22.9 MHz while dropping slightly to 3.6:1at 21.1 MHz. Hmm, looks like I need to get a little smarter about this.

Stay tuned for Part 2 to find out. (Don’t you hate that?)

Ed, K2TE

If you have found a spelling error, please, notify us by selecting that text and pressing Ctrl+Enter.

Follow or Contact Us:

One thought on “Moxons in the Attic (Part 1)”

  1. Put those lawn chairs about 4 feet apart and feed them in parallel through your tuner and you have a 2-element broadside array on 10- and 15-meters!

Comments are closed.