40 Meter Dilemma: Interactions & Size

For almost as long as this blog has existed 40 meters has been a recurring theme. In the beginning when my station was small my focus was on how to achieve good DX performance from a small antenna. Peruse the 2013 archives and see all of my failed experiments! Later that grew into how to achieve good DX and contest performance with less than full size yagis. I have not yet answered the question to my satisfaction.

It seems I am not alone. Among readers of the blog it is the articles on 40 meter wire yagis that are by far the most popular. The most popular of all is for the design of a 3-element inverted vee reversible yagi. It has many good features, including: low cost, good gain, F/B and SWR bandwidth, electrically reversible and it can be erected with a couple of moderately high supports.

Much to the surprise of some readers I've corresponded with I've never built that 3-element wire yagi. However I did build the reversible 2-element version a long time ago and it worked very well. It is only more recently that I've returned to the idea of building a wire yagi for 40 meters fixed on Europe, primarily for contests. I arranged the tower that way.

Of the larger, rotatable small yagi designs I've played with over the years, and ultimately rejected, are:

• Coil loaded elements: too much loss for reasonable component selection and tuning difficulty.
• W6NL style Moxon: modest gain and narrow gain bandwidth but excellent F/B and SWR.
• NW3Z "V" yagi: low gain for its size and tuning difficulty but excellent F/B and SWR.

Currently my only 40 meter antenna is a Cushcraft XM240 up 22 meters because I ran out of time in 2019 to build 40 meter yagis for the 150' tower. I want a mid-height, high gain fixed yagi on Europe and a rotatable full size yagi at the top. This is a very productive contest band that demands the best antennas I can manage.

The dilemma

The south facing panorama below provides context. The 150' tower on the left will have the rotatable 40 meter yagi on top with a 10 meter yagi above. The TH7 will be relocated to ~40' and probably stacked with the TH6 at 75' for shorter paths in the south through west quadrant. A fixed 40 meter yagi of some type will be placed ~100' if I can resolve interaction issues with the top yagi. One or two 10 meter yagis on the upper 50' will be used for stacking.

The 140' tower further away is for the 15 and 20 meter stacks. The lower yagis of the stacks can be seen through the trees. The 40/10 tower is roughly in the direction of Europe from the 15/20 tower, and they are 60 meters apart. The 22 meter tower with the XM240 and A50-6 is on the right. It is 65 meters from the 15/20 tower and 50 meters from the 40/10 tower. The 80 meter vertical yagi is in the foreground. Low band receive antennas are in the bush well off to the left (east).

It was the limitations of candidate designs that gave me the impetus to work out the logistics of a full size 3-element yagi. The challenge is so great that I reconsidered. The difficulties of a full size 40 meter yagi include the following:

• Interactions with 15: Although 60 meters (4λ) away the 15 meter stack will point at the 40/10 tower when working Europe. As we'll see this can be a problem. I already have minor guy wire interactions on the first 15 meter yagi and I don't want to add to the difficulty. The TH6 and TH7 on the 40/10 tower will suffer from interactions on 15 meters.
• Size and weight is a construction and maintenance hazard: No robust 3-element full size 40 meter yagi weighs less than 250 lb (75 kg) and it can weigh up to 400 lb (120 kg). There are ways to build, lift and service it without a crane but these involve significant risk. I can do it but I'd rather not.
• Cost: In addition to the metal and fasteners there is the lifetime cost of equipment and services to raise and maintain the antenna. As I and my helpers get older I'll have to hire riggers.

Modelling of interactions with the full size 40 meter yagi shows that they are significant. I modelled a wire yagi with sloping elements to minimize interactions with a top yagi and the 15 meter stack however its performance is impacted by ground absorption of the large vertically polarized energy -- unless you are pointing it over salt water and I'm not. Horizontal HF antennas are usually best except on the lowest bands where horizontal antennas suffer from ground loss and high elevation angles.

To keep the 40 meter antennas horizontal and reduce interaction with the 15 meter antennas requires moving the 3λ/2 resonance well outside the 15 meter band. As it turns out there are ways to accomplish this. These designs also offer a path to reducing cost and size.

Loaded elements, done with coils, capacity hats or both, shift the resonance of the 3^rd harmonic of a 40 meter dipole. Yagis behave differently but have a similar shift. For example, the XM240 with its coils and capacity hats resonates well below the 15 meter band. For the time being I use it on 17 meters since it performs moderately well as a long dipole.

Unfortunately it is not quite so simple. There are many ways to load elements and many ways to combine them into yagis. Candidates designs must be tested for interactions on 15 meters and other bands. Sometimes pushing a resonance out of one band puts it onto another. It's time to do some computer modelling!

Interaction with resonant 40 meter antennas

Let's start with a dipole. The wire dipole shown here is resonant at 7.1 MHz and its 3λ/2 resonance is 21.65 MHz. It is typical that the harmonic resonance is a little higher than 3× the fundamental.

My 5-element 15 meter yagi will be used to test for interaction. First I will place it above the 40 meter dipole at several heights: 10, 5 and 2.5 meters. These separation distances are typical of antennas sharing a mast or tower. The elements are parallel (as depicted) to represent the worst case.

Pattern distortion is modest, with the shape of the main lobe altered by a fraction of a decibel. Minor (rear) lobes are more distorted and vertically asymmetric with magnitudes up to 10 db worse. Gain is identical while rejection of signals off the back suffers but is not catastrophic. The SWR change (not shown) is negligible.

The more interesting case is when the yagi is pointed directly at the 40 meter yagi. For this I put them at the same height and vary the distance between them -- 60, 30 and 15 meters -- as measured from the front of the yagi.

This analysis is in free space so by "height" I mean that the antennas are coplanar. I have removed the ground so that we can deal with one thing at a time. When ground is added later the pattern will show differences but won't change the key findings.

At 60 meters distance the pattern distortion is notable with the appearance of many minor lobes. As the distance decreases there are fewer but stronger minor lobes. Rejection of unwanted directions is impaired though not to an extreme. As the antennas get closer the main lobe widens, the minor lobes grow and the gain declines slightly. The SWR is barely changed, with variations of only 0.1 to 0.2 at the closest distance of 15 meters.

Of course this is dipole and not a yagi. A 40 meter yagi is more pertinent for my station so we'll do that test next. The differences that may affect the interaction are several: each yagi element has a different resonant frequency; array impedance is different from that of a single element; and a 40 meter yagi covers a large area.

I'll use a full-size 3-element 40 meter yagi developed for an earlier article. Because of the sizes involved and the use of towers the experimental setup is modified. For vertical stacking the distances are doubled and the yagi centres vertically aligned. When pointing the 15 meter yagi at the 40 meter yagi the distances are centre to centre.

It is no surprise that the interaction is worse than with the 15 meter yagi above the 40 meter dipole even at the greater distances being: 20, 10 and 5 meters. Gain and directionality suffer quite a lot at the closest vertical spacing of 5 meters. It is a bad idea to put two antennas of this type on the same tower!

As before the SWR hardly changes at all. The lesson is not to rely on the SWR (feed impedance) to judge whether interactions are present.

The patterns are far worse when the 15 meter yagi is behind (pointing at) the 40 meter yagi. I skipped the 15 meter centre-to-centre distance since it is unrealistic and the interaction extreme. Even when the heights of the two antennas are offset you should expect the interaction to be a problem though different since the main lobe is an ellipse with a large area.

The SWR does exhibit noticable change in the 30 meter separation scenario yet remains low across the band. Perhaps surprisingly it simply isn't a problem.

Of course these configurations are worst case. When the 40 meter yagi is well off the 15 meter yagi's main lobe or the 40 meter yagi is pointed elsewhere the interaction will be far less and usually not too concerning. I will be in a worst case scenario when both the 15 meter stack and 40 meter yagi point to Europe (northeast). Actually it's worse because pointing the 40 meter yagi in the opposite direction, southwest to Central America or the south Pacific, gives the same severe interaction.

This is the crux of my dilemma and thus my pursuit of alternatives.

40 meter yagis with shortened elements

As discussed earlier, elements with loads typically do not resonate on harmonic frequencies. The reactance of loads in short elements, L or C, is frequency dependent. By carefully engineering the element it is possible to avoid resonance on higher amateur bands and to reduce the size and weight of the antenna with only a small reduction of performance compared to a full size antenna, single element or yagi.

That's a challenge I am willing to tackle. There are several considerations:

• Coils must be high Q (typically > 500) to avoid excess loss in a yagi. They are also vulnerable to weather, especially precipitation that can drastically lower Q.
• Large capacity hats are vulnerable to wind and ice, putting stress on the element. For example the large horizontal capacity hats of the W6NL Moxon often require that the elements be guyed. Large capacity hats are fragile and complicate antenna lifting and maintenance.
• Array (feed) impedance of a yagi with shortened elements obscures the behaviour of the individual elements. Don't expect to find the harmonic resonances from the feed point. The XM240 with its dipole feed is better in this respect (see SWR measurement above) but a yagi with a matching network can obscure harmonic behaviour. Modelling or a special measurement mechanism may be required.
• NEC2 does not accurately calculate the resonant frequency (primarily reactance) of elements with loads. Some error is to be expected. EZNEC version 6 includes a correction for simple loads however I don't know the reliability of the algorithm. Simple coils should be fine but probably not capacity hats.

To begin let's look at the SWR sweeps of two short 40 meter dipoles. Both use 25 mm (1") constant diameter tubes since SDC (stepped diameter correction) is incompatible with capacity hats.

The coil loaded dipole is 75% of full size and the one with capacity hats is 65% of full size. Neither has a 15 meter resonance. The coil loaded dipole harmonic resonance is at 24.6 MHz. The capacity hat dipole has its harmonic resonance within the 10 meter band and an unexpected resonance at 10.5 MHz which is close to the 30 meter band. These calculations were done in free space with dipole feeds.

Although the dipole with capacity hats must be modified to avoid resonance within an amateur band it does meet the primary criterion of avoiding the 15 meter band. The resonance of the coil loaded dipole near the 12 meter band is of little concern to me due to my contest focus. It is in fact ideal since it is midway between 15 and 10 meters.

A different capacity hat design studied but not shown here is better suited to wire yagis. Its harmonic resonance is at 23.9 MHz which is far enough to minimize 15 meter interactions.

For the interaction test I'll use the coil loaded dipole and the corresponding 3-element coil-loaded yagi design from a previous article. The precise design detail will be skipped over in this initial study since it is not needed to gain an initial insight into the interaction problem.

Now that's a big improvement. The effects of interaction are negligible even when the 40 meter dipole is just 15 meters in front of the yagi. Moving the harmonic resonance 3.5 MHz above the 15 meter band is obviously beneficial.

That's a good first step so we can proceed to do the same interaction test with a 40 meter 3-element yagi using a similar short element design.

The interaction has reappeared despite the 40 meter yagi using elements that are not resonant within the 15 meter band. It isn't as bad as with the full size 40 meter dipole and 3-element yagi but bad enough to be of concern. As before the SWR is negligibly affected. EZNEC can tell us what appears to be the problem.

There is a resonance at 25 MHz and there is evidence of mutual coupling within the 15 meter band. The resonances in a yagi are not the same as in the individual elements. We can inspect the element currents for further insight.

This is becoming interesting. Notice that it is only the driven element that is exhibiting a resonance and only between the coils. The induced current is quite large so this must be the culprit. The other elements carry almost negligible current.

The only unique property of the driven element is a beta match. This is a shorted transmission line stub inductor and capacitance from shortening the driven element. It transforms the low yagi impedance to a 50 Ω match at 7.075 MHz. Somehow the combination of the beta match and the coils is creating a broad resonance where we don't want it.

Many of the better antenna switches used by contesters place a 50 Ω load on unused ports rather than shorting them to avoid similar transmission line induced resonances in disconnected antennas. The problem we have here could be a phantom or it could be real depending on how the 40 meter yagi transmission line is terminated.

Discussion

I am not going to solve the discovered interaction problems here and now. The study has generated food for thought as I consider 40 meter yagi design. Additional work and real life testing is required. For readers with similar concerns and objectives you should pay attention to these interaction issues. They not easily solved.

One can put up multiple towers in strategic relative placement to minimize interactions. Few are the hams with the ability to do this. As we've seen the distances at which interactions occur is large and not all yagi directions can be accommodated during contests. On balance I don't get overly anxious until the gain takes a significant hit.

The degree to which interactions can be tolerated depends on many operating factors including inter-station isolation, poor directionality can be a bonus when you're called by stations off the side and back and more. Several large stations I've visited must have interaction issues yet they do exceptionally well in contests. No matter how you arrange the towers and antennas you will have problematic interactions at least some of the time.

I have twice at this station stacked tri-banders, small and large, on the same mast as an XM240, which (see above) has no 15 meter resonance. I noticed only modest interaction on 15 meters but honestly this is really difficult to assess. Mostly it was hearing (and working!) more off the back than ought to be there. SWR was unaffected, as we should expect from models studied in this article. Gain was too difficult to assess but I suspect it remained good. Maybe I'd have had a different opinion with more sunspots and a crowded 15 meter band.

There are other interaction issues that I have not addressed in this study and may be significant:

• Guy wires: Despite chopping the steel guys of my towers into segments not resonant on any contest band there will be interactions. Non-resonant mutual coupling tends to increase as frequency is increased. On the lower yagis of my 15 and 20 meter stacks the 20 meter SWR is unaffected but on 15 meters it is slightly different from when it was tuned.
• Full wave loops: These antennas have a resonance on every harmonic and can wreak havoc if not employed with due care. I have none in my current station so I won't soon look at this issue.
• Boom resonance: When a yagi is turned 90° the primary interaction with the elements is eliminated and a secondary one can appear. For example, a full size 3-element 40 meter yagi is approximately resonant on 80 meters when fed at the centre of the boom -- the boom is the element and the elements are capacity hats. Some hams exploit these serendipitous resonances with gamma and omega matches on the boom. These resonances depend a great deal on yagi construction so there is no general rule to determining their frequencies. Models can help.

With the lower yagi of my 15 meter stack installed 100' high and pointing at the 40/10 tower there may be interactions with the two tri-band yagis currently on it. There is no coax (yet) to the 15 meter antenna so that assessment will have to wait.

Perfection is never attainable and we should not go overboard trying to get there. This is a hobby so we do the best we can with our resources. Ultimately we all have to do the best we can with what we have, whether the station is small or large. However don't see this as an excuse to be sloppy. Understanding can be educational and illuminating even if remedial action is impractical or not desired.

Next steps

Although I used coil loaded 40 meter elements for this study that was merely expedience: I had the models in my files. I prefer shortening with capacity hats for improved efficiency despite the mechanical challenges. I plan to explore 40 meter yagi design using elements of this type and see whether I can achieve my performance, interaction and mechanical objectives.

Once a design is in hand I will construct a 40 meter dipole and put it on the tower. There I will adjust the tuning to compensate for NEC2 inaccuracies with this type of antenna. It will use a dipole feed (split element centre) to best measure reactance right across the HF spectrum. Matching networks make measurement difficult so I'll deal with that afterwards in whatever yagi design I settle upon.

If all goes very well this year I'll have a 40 meter yagi (or two!) on the 150' tower by next winter with, I hope, minimal disruption of the 15 meter stack.