Even Simpler Moxon Tuning

One of my pet-peeves of antenna building is that the antenna should be able to be optimised on-site. Simulations reflect very ideal conditions which are most likely not met in real practice. On some antennas the ground conditions, dielectric constants and resistivity, are very important parametres. More often than not we can only make educated guesses on these figures, and lump them together as being “average ground” etc. Nearby objects of metal, or even trees, do influence the result and are hard to accurately simulate. So, as a conclusion, I wish the antenna to be adjustable in some way to achieve its best.When it comes to the moxon, this antenna is no exception. Postings on mailing lists show that very few of us have used calculated wire lengths and been happy with them directly. Some reports show that the point of low SWR is too high in frequency, some have not achieved the high F/B-ratio the simulations claim will be there. Evidently there is some room for experimentation until the SWR is in the right place and the maximum F/B is where it should.

Note that this is not a deficiency of the moxon in itself, merely a natural effect depending on different ground conditions, lengths used, materials used for the spreaders as well as actual position of the antenna.

Simulations followed by a series of controlled experiments on vertical moxons at a frequency of 29.5 MHz have resulted in a slightly new approach to the tuning of the moxon. This method of tuning is presented below and is found to work very well indeed as well as being extrememly simple to implement.

Optimal F/B-ratio

As previously known the front-/backratio of the moxon is an effect of the detuning of the reflector. This detuning is always in the direction as to lower the resonant frequency of the reflector. The standard way of doing this is to make the reflector slightly larger than the driven element. Another way is, as presented elsewehere on this site, to load the reflector inductively in the middle of the longer side with a tuning stub (hair-pin match) or quite simply with a small coil. All of these devices have the effect of lowering the reflector’s frequency.

Adjusting the reflector according to the standard setup is labourious at best and quite cumbersome – as anyone who has tried it will concur. The hair-pin match is however the preferred method if one can reach it! For every little change on the hairpin one has to move oneself away several metres from the antenna before an assessment of the resulting F/B can be made. This also goes for any other temporary metal object, e.g. ladder. The coil-loading method also works well but one has to be very careful in trimming it, as the inductance required is quite small and, consequently, and changes in its value must be even smaller.

Traditionally strong emphasis has been on keeping the symmetry of the antenna at all times. Simulations as well as practical tests show that some deviations from total symmetry is acceptable and will not deteriorate performance in any noticeable way. This observation leads to another way of loading the reflector: With a simple wire. Any such, capacitive, load should be placed at a high impedance point of the antenna. But as the moxon gets it fabulous performance from the capacitive coupling between the ends of the two elements one should avoid upsetting this geometry. A satisfactory point for installing a capacitive load is at either corner of the reflector. So clamp a wire to a corner and simply let it dangle from there! For the symmetry-freaks around a symmetrical load could be used by halving the wire and attaching it to both corners, but this is not really necessary.

A longer wire lowers the frequency more than a shorter wire. And starting off with a long wire it is relatively easy to cut off inch after inch until best F/B is achieved. A thorough approach would include a small mW-transmitter at the preferred frequency some 5–10 wavelengths away in the rear direction. As the wire-reflector-combo reaches its best F/B one should have to cut in smaller increments and notice very carefully when minimum signal strength has been passed. A convenient trick is to fold an inch of the wire up, then test, and cut only after this folding has proven to improve the F/B.

Lowest SWR

It is of course pleasant if the antenna shows a small SWR. 1:1 is the goal but one should not be obsessed by this. A SWR of less than 1.5 is small enough to be used with no tuner for even the most discerning transistor finals. With a tuner at the transmitter the only negative with higher SWR is increased loss in the coax. This loss is however small in most cases with SWR:s up to 2 or 3 and not overly long runs of coax.

It turns out that once the reflector is tuned and in place almost all of the SWR depends on the size of the driven element. A too small driven element have a capactive reactance and a too large one has a inductive reactance. At resonance there is no reactance at all and with the geometry normally associated with the moxon its impedance will then be about 50 ohms pure resistance.

The wire-detuning as used above can equally well be used for the driven element: Attach a wire to one if the driven elements corners and chop it off in increments while observing the falling SWR and finish when the SWR is lowest or acceptable as to one’s liking.

The method assumes a feed point in the middle of the driven element. In that case a 50 ohm impedance will be achieved and finding it with a simple SWR-meter is easy. Purists will find that feeding with a number of exact electrical half-wavelengths feeder might give a better result. No matter what one tries, a lowest SWR will be found and it will not be too far off, if not spot on.

Alternate feedpoint

In some instances the traditional feedpoint is tricky to reach. For a vertical moxon the feedpoint can be even trickier to reach as the vertical setup is common in portable scenarios, where ladders are not possible to bring about. Alternate feedpoints have been simulated and evaluated in real tests described in the following text.

In general the impedance changes to higher resistive values and higher capacitive (-jX) reactive values as we go from the centre towards the ends of the elements. We know from above that capacitive reactances can be tuned out by longer tuning wires. This cannot, however, be stretched infinately. The reactance can only be tuned out a rather small amount.  This leads us to look for a point along the director where the resistive component of the impedance is manageable and its capacitive part can be tuned out – or at least made sufficiently small.

Tests reveal that one such point is situated about 15 percent in taken from one of the director’s corners. This point will have only a small reactive component (after trimming the tuning wire) while having a resistive value of about 200 ohms. 200 ohms is quite a long way away from our coaxial cable’s 50 or 75 ohm but will do well with a simple 4:1 balun transformer. Feeding with 300 (or even 450) ohm ladder line is also possible, in which case the balun is not warranted – at least not out of an impedance viewpoint.

Feeding from the corner is also possible but the impedance will be greater and will also have a, untunable (with tuning wire as above anyway), unwanted reactive component. The system’s F/B will be largely unaffected by this but the 4:1 balun, or even 16:1 balun tested, will see a quite severe complex impedance to transform. A balun works best, have smallest losses, with resistive loads but if one is willing to sacrifice some energy – and probably not run huge powers which will consume the poor balun – corner feed is perfectly possible.

Practical tests on a vertical setup confirms these simulated results. There is one severe caveat however: The driven element should have an overall length giving a feed point impedance of about 50 ohm when fed in the centre – even if we do not end up feeding it there! When no care of this was taken it was easy to get good SWR via the “15 percent”-feedpoint and a 4:1-balun by shortening the element. Thus pleased the author went on with testing F/B only to find that it was severely off and after readjustment of the tuning wire a only a much worse F/B was ever achieved. After replacing the shortened piece of wire and settling for the best SWR at the design frequency, the situation was back to normal and good F/B was restored.

It appears that corner feed dramatically enhances this effect and one should be observant not be lured into chopping of wire to get a lowest SWR only adding load by means of a tuning wire. Do not change the original director size.

Initial dimensions

As we want to be able to load the antenna’s elements it is important the antenna used gives us this leeway. In practice this means to start off with a much too small antenna, i.e. dimensioned for a higher frequency. Suggested dimensions can be taken from the figure below. These are expressed in terms of wavelength which makes it easy to scale the dimensions into any practical frequency. So while we aim at an antenna to work at a design frequency we start off with something physically smaller.

Suggested lengths to be used for the tuning-wire-loading to work. Then antenna has to be smaller than normally assumed but its overall geometry is the same. As these figures are rather “rough” a few percents change either way will be of no importance. Specially there is no need to take velocity factor into account. Just cut the wires according to the suggested lengths and let the tuning-wire take care of the rest. Tuning wires can be about 1 metre to start with. In the case of 29.5 MHz the initial dimensions turned out to be 356 cm high, 58 cm ends and 20 cm spacing between the elements. Reflector tuning wire was some 30 cm and the driven element tuning wire about half of that. “15 % feed”-point was 0.15*356=53 cm from the lower corner.


Suggested lengths to be used for the tuning-wire-loading to work. Then antenna has to be smaller than normally assumed but its overall geometry is the same. As these figures are rather “rough” a few percents change either way will be of no importance. Specially there is no need to take velocity factor into account. Just cut the wires according to the suggested lengths and let the tuning-wire take care of the rest. Tuning wires can be about 1 metre to start with. In the case of 29.5 MHz the initial dimensions turned out to be 356 cm high, 58 cm ends and 20 cm spacing between the elements. Reflector tuning wire was some 30 cm and the driven element tuning wire about half of that. “15 % feed”-point was 0.15*356=53 cm from the lower corner.

Measured results

First a warning. These measurements were taken with the S-meter of my TS-930S. As S-meters go this one is neither better nor worse than the average instrument. So instead of declaring this or that amount of F/B in (untrue?) decibels I will write actual values measured. And also, these values do not really say actual F/B but tell the effect of a tuned reflector versus a severly detuned one. In practice the gain of the antenna in the forward direction should be added to these figures. My SWR-meter is best at low SWR’s so a reading of more than 2 can be actual 3 or worse.

  • Centre feed

With centre feed a SWR of 1.2 was possible and F/B-values were S9+12dB reduced to less than S1 and S9+15dB reduced to about S3. With the TS-930S’s built-in 10 dB attenuator  a reduction from S9 to S0 was recorded by simply attaching the “tail”/tuning wire of the correct length. Quite an impressive demonstration.

  • 15 percent feed

With a feeding point 15 percent from the moxon’s lower corner and a 4:1 balun wound on a FT-140-77 toroid, a best SWR of 1.3 was reached in a rather flat response. Note that the design frequency of the antenna (29.5 MHz), i.e. frequency of best F/B, the SWR was not at its lowest. Lower SWR was measured at a lower frequency (about 1.2 at 28.5 MHz), but at this frequency no good F/B was ever possible. When transmitting a constant 50 watt into the antenna for 20 seconds the balun was somewhat heated indication some loss. Best F/B was from S9+10 dB down to S5.5.

  • Corner feed

Fed at a corner it was never possible to get a lower SWR than about 2. Both 4:1 and 16:1 baluns were tried, and contrary to expected from reading similar experiments on the web, the 16:1 balun was actually much worse. A better match is perhaps possible with less “orthodox” balun ratios. This was not tested. F/B was measured from S8.5 to S2, i.e. not as good as was possible with  the other feed points. Still the simplicity of having the feed point in the corner may outweigh. The 50 watt transmit test resulted in a much warmer balun than in the previous test, indicating larger losses.

Simulated diagrams

The vertically mounted moxon is simulated at a height of 3 metres and fed at the “15% feed”-point. As can be seen (can it?) there is a slight unsymmetry in its pattern. Look at the shape of the point of maximum F/B. Not much to make a fuzz about is it?


This tuning wire trimming is equally possible with horisontally mounted moxons.  The  following picture shows what to be expected from a mounting 5 metres (half a wavelength) up:


Notes and conclusions

With centre feed it was clear that the tuning wire works as expected giving a clear minimum SWR. When using the 4:1-balun or the 16:1-balun any change in impedance is divided by four or sixteen so the effect of the tuning wire was much smaller. One may argue that using a tuning wire in the driven element loses its meaning in the latter cases, except possibly to reduce losses in the balun used. It is however not easy to trim the tuning-wire for least power loss.

If you would ask me then I would try to use the normal centre feed first, it is definately easiest to trim, 15 percent feed only after some thinking and completely avoid the corner feed according to the results above. A corner feed with some other balun might work but the losses due to the reactive load are likely to be about the same anyway. But, as always, it is better to have any signal out than no signal out so by all means do use the corner feed if that is the only practical feed point!

In all but centre feed a choke balun on the feeder close to the antenna feeding point is mandatory.

Gallery | This entry was posted in Uncategorized. Bookmark the permalink.

4 Responses to Even Simpler Moxon Tuning

  1. kuchura says:

    Hi Michael,
    I like this way to tune antenna. I am going to try vertical Moxons in portable operation this summer.
    But I see no pictures in the article. Though, it is clear from the text what do they illustrate.
    73! EU1KY

    • jabcam says:

      Hi Yury,
      and thanks for the heads up. A server change made me lose the pictures, I have to restore them by hand:(

      Specially if you are close to salt water, on the beach of even place the antenna IN salt water, then the vertical Moxon excels due to increased gain! The F/B is of course always there.

      Salt water could be tricky “в минске” though…

      Good luck with your experiments!

  2. kuchura says:

    Regarding the feeding. I’ve got an idea.

    Have you ever tried to make an element, or at least a part of it, of lightweight but low-loss teflon coax, like RG-316? The current inside of its braid is different from the current outside, they should not interact because of skin effect. The VSWR inside should remain low, no matter what RF current is induced on the braid from outside.

    So, the feed point remains in the middle of element, providing a perfect match, but the cable leaves the end of antenna (or the corner of Moxon). What if I put here a ferrite choke right on the braid, or, possibly better, make a trap with several turns of the cable with trimming capacitor (e.g. made of the same coax)? No cable cutting is needed in this case. This should cut off high RF voltage from the rest of the feeder and will not detune the antenna significantly. Possibly, several more ferrite cores must be placed on the braid near the trap, to minimize proximity effect of the cable leaving antenna.

    Thinking of this way to feed halfsquare from the bottom. This antenna seems to be the simplest one to implement and try it in the garden. 🙂

  3. jabcam says:

    That is a good idea but I have not tried it yet. I had plans to do this on the G0GSF-antenna to get rid of the “feeder parallel to the radiator”-problem. It is certainly worth trying. Must have 2″ toroid around somewhere or perhaps som clamps chokes.

    If you try it, please report back!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s