XM240 and 17 Meters

The station I have built, and continue to build, is primarily focussed on HF contesting with a secondary focus (though a very close second) on HF and 6 meter DXing. Due to time, budget and other practical considerations the so-called WARC bands -- 30, 17 and 12 meters -- have gotten limited attention. While my attention is indeed elsewhere these bands are not forgotten.

I have been happy to use my 80 meter antennas on 30 meters with the help of a tuner to bridge the small disparity between 10.1 MHz and the actual third harmonic. Having a choice between a horizontal and vertical antenna is beneficial. Building a gain antenna, one with two or more elements, is not at all in my plans and may never be due to the size, cost and effort involved.

The minimum of a solar cycle does not turn 12 meters into a hotbed of activity though there are sparks of interest, such as when I worked VP6D on both CW and SSB a few weeks ago. On 12 meters I use a tuner on one of my high yagis and hope for the best. This is usually good enough because extreme height makes up for many flaws.

Which brings us to 17 meters. This band remains very useful for DXing even during a solar minimum. Yet all I have for this band is my XM240, which performs similar to a dipole at 18.1 MHz and is resonant not far above the band. The moderately high SWR is managable with a tuner. Again, height is my friend.

Of these three bands it is 17 meters that is best positioned into inducing me to consider a better antenna. It is not urgent. What it does is make me conduct a few thought experiments on how I can achieve improved performance at little cost and effort. One of these thoughts was to turn the XM240 into a 2-element 17 meter yagi. The concept of adding a 17 meter reflector to the antenna is theoretically sound so I built a computer model to see what could be done.

Before we proceed I will be very explicit regarding the limits of this exercise. I have not built this antenna and I may never do so. I don't know how well it would really work. The greatest danger is that the XM240 cannot be properly modelled by the NEC2 engine; only NEC4 is up to the task and I don't have it. My EZNEC model of the XM240 is at best a proxy of the real antenna that allows me to explore its behaviour, but not to accurately derive physical dimensions. I'll have more to say about this later.

Model

When I purchased the XM240 it was placed in storage until it could be raised at my new QTH. At that time I developed a "naive" EZNEC model of the antenna using straight tube elements along with coils and capacity hats with the same values as the actual antenna. Because NEC2 cannot accurately model elements with loads (EZNEC 6 can with its NEC2 enhancements, but only somewhat) the element lengths are not correct. The model nevertheless permitted me to explore the antenna's behaviour.

My current exploration of 17 meters uses that same model. I added a reflector element, tuning its reactance and spacing to the driven element to optimize performance at 18.1 MHz. As with the rest of the model I used straight tubing rather than a taper schedule. That can be added later should I decide to build it. The model's value is not affected by this doing this, it's just that the element length will be different with tapered tubes.

Notice that the XM240 40 meter reflector has almost zero current flowing on it. This is one key to understanding why a 17 meter reflector can be added to the antenna so simply. Indeed it the same reason why a yagi can be correctly tuned close to the ground by pointing it upward. When the 17 meter reflector is correctly tuned there is effective cancellation of the fields from it and the driven element towards the rear. The 40 meter reflector is spaced far enough that it has low mutual impedance with the other elements at 18 MHz, and therefore has negligible effect on the 17 meter reflector and driven element.

The other key to understanding the design is that the tuning of the driven element has essentially zero effect on yagi performance, that is, gain and F/B. Whatever phase shift mistuning introduces does not affect the phase relationship to the reflector. It is only a concern for matching. However this can only be relied upon if the length of the driven element is not too short or long, which would affect the mutual impedance with the reflector. You can see above that current on the driven element is low beyond the 40 meter loading elements, which keeps its electrical length on 17 meters close to that of the reflector.

Some improvement in 17 meter performance could be gleaned by reducting element spacing. I decided against that in favour of placing the reflector close to the boom centre to minimize torque imbalance. The 3 meter spacing is 0.18λ at 18.1 MHz.

Performance

The width of the 17 meter band is only 0.5%. It is therefore good enough to tune the antenna for the centre of the band and have close to optimum performance across the full 100 kHz, despite the narrow gain and F/B bandwidth typical of 2-element yagis. I used 18.1 MHz to keep to round numbers, and because I favour CW.

It is possible to raise the gain a little higher than 6.2 dbi by narrowing the element spacing and centering the frequency of maximum gain within the band. This would come at the expense of poorer F/B, and matching challenge due to the drop of radiation resistance at peak gain. I chose not to do so in this design due to the aforementioned torque imbalance and because the F/B is already quite poor, though as expected for the parameters of this 2-element yagi design. For me these are reasons enough to sacrifice 0.4 db of additional gain.

When fed without the reflector the pattern of the XM240 on 17 meters is similar to that of a dipole, with a bidirectional gain of ~2.4 dbi. On the air I do not notice any F/B so the model may be reliable in this respect.

The question to ask is whether 4 db of gain and modest F/B is worth the trouble. It can be easily argued that a separate and better performing 17 meter mono-band yagi is preferable. In my situation a separate yagi would not be placed as high as where the XM240 is currently located: 46 meters up. I am undecided. It's food for thought until the time comes that I decide to do more on 17 meters.

I have not reported on the SWR or feed point impedance because the straight tubes model of the antenna is unreliable on 17 meters. The best I can predict, based on exploring the model, is that the R component of the impedance is in the region of 25 Ω. I will not predict the reactance other than to note that it should be comfortably low.

A switchable matching unit, such as an L-network can be employed for 17 meters if an SWR of 2 or 2.5 is problematic. The network can be mounted at the mast to avoid adding weight at the end of the boom, and the network designed to match the impedance measured at that point.

Of some concern is that the Cushcraft 1:1 common mode choke operating at 18 MHz, a kilowatt and a mismatch at the feed point. A better balun is advisable or the network and switch should be located at the driven element and the balun switched out of circuit when operating 17 meters. But this is all a lot of bother for a modest improvement of 17 meter performance.

Conclusion

As said, this is purely a thought experiment to consider my future alternatives for improved 17 meter performance. I don't know if anyone else has tried to do this with an XM240 and I didn't research the question.

Computer modelling makes these types of analyses easy to do in the comfort of our shacks. It's a lot easier than modifying the antenna, especially an antenna of this size, and measuring the results. The analysis will now be shelved. Time will tell whether I return to it in the next couple of years.