The following design came from a program called 'Antenna Maker' by John Agrelius, KM6HG. The program is a DOS based calculator which comes from the 16 bit generation of PC's. As I run V64 now, I have no way of running it however I provide a download mirror for the program for any interested parties.

This design was a lot more compact than the others for 6 elements and packed a fair bit more gain also:-

As can be seen, the pattern resembles a traditional yagi shape of similar gain range. The boom is 4.57m long, which couples the elements quite a bit closer than more favoured designs. Nonetheless it is a fair design, worth building I think, for those so inclined. At 11.45dB it has the highest forward gain of any quad design I came across while searching for the mix I felt was right.

Finally there was the 4 element quad I found on G4AON's site. Originally based on a 2 element portable design by VE7CA AON took the design and made it his own, for portable use. The gain was a much more realistic 9.6dBi which was below what I really wanted for MS work but it was necessary to consider all options.

The pattern is very "quad" and it has a respectable F/B. Although the gain was down on my preferred range, at 2.57m, it placed this antenna very high on the agenda, which by now was biased far more towards prevention of future damage by the wind than motivated by a hunger for forward gain.

At this point the thought struck me that I had not been active on 2m or 70cm for a long time. One thing you can do with quads that you can't with yagis, is nest them; quads for shorter wavelengths fit neatly inside each other, the challenge was on. Design a nested quad for 6m, 4m, 2m, and 70cms, all with acceptable specifications and as much as possible with any nasty side-lobes designed out.

Cubex have been making quads for a long time and are well respected in their field. The company also offers for sale, spares and materials for self-build which was ideal, as I didn't really have the stomach to source all the relevant parts - I don't leave the house too often these days and living in rural Somerset doesn't give me local access to much that is of much use.

The principal obstacle to designing a nested quad like this is the interaction between the various band elements, most elements have other elements which are a significant proportion of a wavelength and this can make life very difficult.

Without getting into the dimensions just yet, an overview of the design is above and as you can see, there is some not insubstantial proximity going on here, especially between 6m and 4m. To build this one thing above all else was mechanically essential to keep the weight and complexity down - re-use of the cross-support arms. If each band was installed exactly to its ideal spacings, each band would need its own separate set of supports. As  6m and 4m were selected to be 4 elements and 2m and 70cm were selected to be 8 elements, if you total up this based on individual band ideal spacings, one ends up with 48 diagonals. . . That's a lot of fibreglass to put up in the air and the bigger they come, the harder they fall. Total wind surface area alone made this a prime candidate for calamity! Some further optimisations would be necessary.

By use of larger diagonals and tensioning them, it would be possible to off-set elements by several centimetres which although doesn't sound much, can make the difference between a working antenna and a failure. Logically 6m would require the widest spacings, so a 'best case' 6m quad was constructed based on a 2.4m boom. This would be the working model from which to start the process of implementing the remaining bands.

I did not want to have separate supports for 6 & 4m, so 4m had to be designed on to the 6m supports. This meant the spacing would be less than ideal. By combining elements onto shared supports, I was able to reduce the diagonals from 48 to 22 which was far more realistic. I think I must have spent two weeks or more doing nothing but tweaking the design to try and extract every last available drop of forward gain while not sacrificing too much in the way of F/B or side. Bearing in mind this is a compromise antenna by any standards, come the end I was happy with what I achieved. The following plots represent what I managed to do.

6m

Once built, 4m SWR and matching was significantly better than modelled, at 1.1:1 & 48+j 1.2 ohms

4m

Once built, 4m SWR and matching was significantly better than modelled at 1.15:1 @ 53.1 -j 2.1 ohms

2m
I confess that much less effort went into 2m and 70cm

70cm
 

The final analysis

I have been using this antenna for many weeks prior to writing up this page and I must say that I am extremely pleased with it. Each band is separately fed with RH100 which is almost identical to Westflex 103 but at 10mm diameter. The SWR is at 1.1 on each band at the design Fc and the forward gain and F/B bears out the model data. I have used the 6m and 4m components on meteor scatter with very good success, I would say that I do not really notice not having 6 elements on 1WL which I was afraid would be the final outcome. Indeed at the time of writing I have been conducting tests with ZS6WAB on 4m EME and although he has not heard me as of yet, partly due to prevailing conditions, I have now successfully received my own echo off the moon at -30dB.

One final observation. In another part of the website I have stated that I have not heard anything off the moon since 2001 due to gradually increasing noise levels at this QTH. Don't get me wrong, I have tried, oh boy have I tried, even attempting phase noise cancellation, stacking yagis for noise, etc, etc. In the end I occasionally listened in (tried to) to an EME QSO in progress and while others all around me were reporting audible signals and -15dB, I was seeing nothing, zilch; deeply frustrating it was, too.

In the last few days I have started listening again and as if by magic, suddenly I can hear off the moon again. Although I remain sceptical about the reported 'quietness' of quads, based on Cebik's findings, I can nonetheless hear weak signals which I had all-but given up on several years ago. Perhaps it's coincidence, the noise level 'appears' to be the same however I can't help but wonder if the S/N ratio has improved.8

Then again, perhaps its the IC-7600 that has made the difference, I cannot say for sure.

73

Nige.

(\__/)
(='.'=)
(")_(")