So, what is the conclusion?
If I wanted the antenna work on 27.555, what will I have to put the number to moxgen? Shud I put the higher, 28.555 or the lower, 26.555?
Thank you.
If you are using bare wire, use the frequency you want. If you are using insulated wire you need to decrease the dimensions by roughly 4% as due to the velocity factor of insulated wire, it appears 4% longer to RF. Moxgen DOES NOT allow for insulated wire. If you want 27.555 as the target and you're using insulated wire, putting in 29MHz should get you close to where you want to be. The resonant frequency will be a little lower but this will be ideal for being able to access the soon to be legalised SSB frequencies on 11m
I've done quite some investigation into this and been discussing via email with Marconi. My own Moxon when built to the Moxgen dimensions was about 1.5MHz too low because I was using insulated wire. If I look at the dimensions I finally ended up with, they're roughly 4% smaller than the Moxgen ones.
Modelling in EZNEC confirms this.
I'm interested to know about the installations when it appears to generate lengths that are too long resulting in a build with a resonant frequency too low. In particular I'm interested in the height above ground, how long the coax is between the antenna socket and the analyser and how long you take the coax away from the feedpoint at right angles to the driven element and in what direction as these all have an effect on the resonant frequency. And also by resonant frequency are we talking about where X=0 or where the lowest SWR is? The lowest SWR is not the resonant frequency, the point where there's no reactance, i.e X=0, is and the SWR dip should occur several kHz away from this. If its occurring at the same point there's an issue.
The downside as I discussed with Marconi was that you can take 100 different Moxon calculators and they'll spit out 100 different set of dimensions even though they're all apparently following Les Moxon's formula or L.E Cebiks' revised formula. Fortunately modelling each of these shows that the variations are in forward gain and front to back rejection and that the differences aren't that massive. Some favour a bit more gain over less front to back and others favour more front to back at the expense of a little forward gain.
Here's the models I did for my own investigations. All are done for a centre frequency of 28.5MHz.
This is the first one.
It is the one generated by Moxgen. The resonant frequency is what we expect, the performance is what we expect.
http://1drv.ms/1lErzMc
Now if we change that to insulated wire which is what most people will be using as they'll use speaker cable or mains cable, look what happens when we still use the Moxgen dimensions. Resonant frequency drops to 26.9MHz!
http://1drv.ms/UTHhrT
This is the model of mine which was shortened from the Moxgen model to get it to tune where I wanted it. If you do a SWR sweep up to 31MHz you'll see the dip is around 30.2MHz. The SWR at 28.5MHz is quite high and the front to back ratio is only a few dB. However in real world use it was resonant at just below 28.5MHz and had F/B of around -20dB.
http://1drv.ms/1lErd8y
This is the model of mine taking into account the insulation of the actual cable used. If you do a SWR sweep on that you'll now find that its at roughly 28.5MHz, a drop of 1.7MHz over the previous model which used bare wire. F/B ratio at 28.5MHz is around -20dB which is what we expect.
http://1drv.ms/1lEroQZ
Now you'll find that if you increase the size of my wires by 5% you get pretty damned close to the Moxgen generated lengths.
Its a good practical lesson into the real world effects of using insulated wire and how it affects RF.