Yagi Side Resonance

RF is very sociable: it happily interacts with everything in the universe. A few interactions are significant, some are minor and the vast majority are negligible. As hams we discover or predict interactions, then deal with the problematic ones and ignore the rest. Often, interactions are unknown or dismissed, and only addressed when they are too severe to be ignored. 

As deliberately radiating devices, antennas are especially prone to interactions, as the perpetrator or as the victim, or both. As the victim, an antenna's pattern and efficiency can be degraded by interactions. Those interactions can be with structures, utilities, trees, ground other antennas within a couple of wavelengths.

When I designed my 3-element 40 meter yagi I paid particular attention to its interaction as the perpetrator and the 15 meter stack as the victim. It is big and high enough that it is not particularly prone to be a victim. This is helped by the vertical polarization of the 80 and 160 meter antennas: orthogonal antennas interact little.

Here you see a simple 80 meter dipole excited at 3.6 MHz. It is positioned 200' (60 m) laterally from the centre of the 40 meter yagi and 50' (15 m) lower. The 40 meter yagi has a substantial mutual impedance and is happily interacting. How is this possible? 

It is no surprise that the pattern of the 80 meter dipole is distorted. For an 80 meter yagi with 2 or more elements the pattern degradation would be substantial. This is of particular interest to me since I am considering a wire yagi for 80 meters suspended between my two tall towers, including the one supporting the 40 meter yagi.

A ham in the distant past had the interesting idea of making a yagi into a dipole for the next lower band. Viewed from the side a yagi is a short dipole with large capacitance hats. A typical 3-element yagi resonates at about half that of the design frequency. That is, a 3-element 20 meter yagi is resonant within or near the 40 meter band, and the same is true of a 3-element 40 meter yagi on 80 meters.

The trick is with the feed point. You must excite the boom and not the driven element. Although they are mechanically and electrically connected the 40 meter feed is neutral with respect to the boom. As is common knowledge, the elements of a yagi are electrically independent of an orthogonal boom that crosses their centres.

Yagis with elements insulated from the boom do not have this attribute. At least the outermost parasitic elements -- reflector and last director -- must be electrically connected to the boom. 

Since the boom is a continuous conductor a similar feed system is required as for a yagi with a continuous driven element. A gamma match may work. More often an omega match must be used since the yagi cannot be adjusted to be compatible with a gamma match. In an earlier article I referenced a couple of construction articles by those who have done this.

Were the elements of the yagi insulated from the boom, its 46' (14 m) length would serve pretty well as a dipole on 30 meters.  

For those modelling these effects it will be necessary to add the boom to the model. It is ordinarily omitted from the model because, other than the element-to-boom clamp, the boom is neutral. To model the boom dipole I had to split the centre wire of each element to allow connection of the wires that comprise the boom. SDC (stepped diameter correction) and other NEC2 limitations that confound accurate yagi modelling are a relatively minor source of inaccuracy when the elements are employed as capacitance hats.

Both matching networks are difficult to retrofit to an existing 40 meter yagi due to its size. For example, the calculated gamma rod length for the impedance at 3.6 MHz is 13' (4 m) long. There is no good way to do that without a crane and basket. More capacitive reactance can reduce the length by several feet but that requires rebuilding the 40 meter yagi! Perhaps a "long enough" but short gamma rod and an L-network will get you to 50 Ω.

Notice that there are other resonances, which is not unusual for an antenna with many protuberances. There is a resonance at 14.4 MHz that, due to model inaccuracies, might actually fall within the 20 meter band. I have not done the modelling to determine the magnitude of the 20 meter interaction.

Although the antenna can have a DE feed point for 40 and a boom feed point for 80 meters, it is not advisable to operate both bands concurrently without band pass filters that are up to the challenge. However this is usually only a concern for multi-op and SO2R contesters. Everyone else can utilize two transmission lines (one terminated) or a switch at the antenna.

There are pros and cons in this situation. The major benefit is the possibility of a high dipole on a band where one might otherwise only be able to deploy a wire antenna like an inverted vee at a lower height. The major disadvantage is that a yagi can interact with a horizontal antenna for the lower band when the boom is broadside to that antenna. There is no one solution since every station is different with respect to operating objectives and favoured beam directions.

That lower resonance is there whether you feed the boom or not. Putting that lower band resonance to use as a dipole may involve significant effort. The boom resonance can be defeated in the same manner as for a tower. However I've never heard of someone taking the trouble to do so. Interactions are silent so that their impact is easily mistaken for other things and therefore may never be addressed.

Out of interest I created a boom model for my 5-element 20 meter yagis. The resonance is lower than half of 14 MHz because, with 5 elements, the boom is electrically longer. The boom resonance surprisingly falls right inside the 60 meter band. Of course it almost certainly doesn't but it should be nearby. But I have no interest in that band and the interaction isn't important since it isn't a contest band.

Notice how the elements adjacent to the feed point at the boom centre are very active (directors 1 and 2), but not so for the other 3 elements. There isn't much current flowing at the extremities since the most active elements introduce a current distribution discontinuity. That is in the nature of capacitance hats. There is a similar effect with loading coils. That is why they are effective at their job of shortening an antenna.

More can be said of unusual yagi resonances, how to positively exploit them and the downside of interactions with other antennas. But I won't. Every antenna and station is unique so it's enough to raise awareness of this little understood issue. Look around your antenna farm, whether it is big or small, and consider what might be at play. Most will shrug it off since characterizing the scope of the problem and taking mitigation measures is neither easy nor straight-forward.

If I build the 80 meter wire yagi that I have tentatively inserted into my 2022 plan there are factors that lessen my worry. First, most of the time the boom of the 40 meter yagi will not be parallel to the 80 meter elements. Second, yagis have enough directivity to reduce interaction when they are at different heights. Third, I still have the 80 meter vertical yagi should a problem arise when the 40 meter yagi is pointing in a direction that increases the interaction.

When I get serious about building an 80 meter wire yagi I will develop models to explore the matter in more depth. I have no plan to convert any of my yagis as lower band dipoles since I have enough antennas to meet my needs. My concern is interactions.