So many commercial and home brew constructors use Stauff type clamps for attaching yagi elements to the boom but I have until now avoided these because I hate the huge bolt holes that are typically at least 6.5mm in diameter which is sloppy even on a grossly overkill M6 bolt. Of course these clamps were not designed for making antennas but are extremely useful as they come in pretty much every tube diameter there is.
Usually instead I have been making my own elements mounts in some way or other but I am building some HF beams and it’s just convenient to use off the shelf Stauff clamps so I decided I needed to do something about this issue. I recently assembled a commercial beam that had a tapped boom so next to no slop bolt to boom but you could move the element side to side about ±6 inches or more at the ends of a 50MHz element. I had to use a square to mount them:
So I set about as usual pondering (usually when I should be sleeping!) over how I could stop all the hideous play. Another good thing for me about the Stauff clamps is that they provide 3D CAD models for free download so I was able to download some and do some thinking. I prefer to use much smaller bolts like M4 for my antennas which is more than strong enough in a good design. My 2m and 70cms elements are held on with 2x M3 bolts which is also more than strong enough.
The recess in each half for the bolt head according to the 3D model is 12mm diameter, so I decided that a simple solution would be 2 delrin or similar washers 12mm round with a 4mm bore fitted to the half nearest the boom/mounting plate. They are not needed in the top half because the top clamp will self centre onto the element tube:
That would do it I decided. I have a very small lathe and I checked I can easily buy rod in 12mm diameter.
Next step was to get some Stauff clamps ordered to ensure they matched the 3D models I had. On measuring the bolt head recess they were actually more like 12.4mm (at the top anyway), probably due to needing slight draught for injection moulding. Hmm. Next size rod size was 14mm. I do have the lathe, but I’d rather not turn down enough rod to make dozens of washers. I wanted to just drill and part off. I decided to check the supplier notes on tolerances. Some good news there! as most of these materials are very often turned down, the tolerance was stated as +0.2 to +0.7mm. Good stuff. Let’s get some in. I ended up ordering acetal for its low absorption quality and as it was very cheap!
Once it arrived I did a test fit and the bar fitted a couple of mm deep into the bolt head recess then started getting tight, that draught taper I imagine. Ideal. So it was off to my baby lathe and get to work, first running a 4mm drill up the centre then parting off 5mm lengths:
Next to press them in. I discovered that the 16mm and 12.7mm Stauff clamps had an ever-so-slightly larger bore bolt head recess than the 10mm Stauff did. And coupled with the slight edge burr from parting off (seen in pic above) it would have been impossible to press them in by hand. However, I already have an arbour press that I got for forming bent box section boom supports, so in a couple of seconds they were fitted easily:
Soon I had a complete set of Stauff clamp pairs with captive washers on one half only:
The proof of the pudding is in the eating or something along those lines, so I thought I best test them! The day before I had drilled out the mounting plates for the 2 beams we are making so I clamped one in the drill vice and just fitted two M4 bolts into the holes to see how it worked:
That’ll do, as the Yorkshire based advert says… I am happy now that single clamps on the boom will hold a 50/70MHz element securely and square. And once you get to 2 clamp pairs on a 28MHz element or bigger there should be no play at all.
Many might say that this is all overkill and 1000s of antennas have been made and used very successfully with none of this effort. This is of course true but I like to build ALL my antennas to exacting UHF accuracy including the HF ones. I think attention to detail is worth it and my VHF/UHF contesting results using my home brew antennas makes it worthwhile.
As an avid antenna constructor myself this is a bit of an unusual post for me. It comes around as I was planning to build myself a dual 10m and 6m beam based on the DK7ZB design on this page (link). I was already running late as the Es season was well underway and in chatting with the Hereford club members Clive G8LNR said there was a tri band version of the same thing doing nothing I could borrow. It was made by VPA Systems and sold by TelTad on this page (link). This was ideal for me as it would save quite a lot of time, so I leapt at the chance and fetched it to my house to build.
This is a lightweight budget end of the market antenna with a claimed weight of 3kg and costing 193€ but that is ideal for my purposes. I retract my mast to gutter height when not in use and I don’t want heavy antennas on the aluminium mast (I’d love the Optibeam OB6-3M but it’s just too heavy).
Unravelling the bundle gave me this set of parts: which includes a set of Stauff style element mounts, stainless fittings and a single U bolt for fixing boom to mast.
The first thing of note to me was that all the element mounting holes on the 25mm square boom were tapped holes: I must admit I wasn’t feeling very confident about that in terms of mechanical strength but they seemed to tighten down securely. Note that I did tighten using the short arm of the Allen key to reduce possible over torquing.
The 6m elements are 12mm diameter rather than the 10mm specified in the DK7ZB design. I assume this was done in order to facilitate the joining ferrules used to shorten the shipping length. The corners of the 10m moxon are formed with flattened and drilled ends. The 4m elements were 10mm single piece elements:
The antenna came with no assembly instructions (from new) and Clive couldn’t remember if it was the lower or higher power version so the first thing to do was examine the choke in the feed box. Here I found the biggest disappointment so far (and worse later on!). The DK7ZB design recommends a 6 turn ferrite common mode choke on 43 core but here I found a simple ugly balun type ‘choke’: On the up side the dipole box is a good stiff ABS box and though still only 2 fixings to the boom looks like it shouldn’t droop too much like some driven elements. Also an N-type socket used rather than an SO-239 banana plug socket. The ugly ‘choke’ was made from an RG-188 based Teflon coax so this would be the lower power rated version.
A sort through the Stauff copy clamps and fixings soon revealed which clamps and fixings were for each element and the position of them is fairly self evident so assembly was commenced. As is typical with Stauff style element mounts there is loads of play between the mount and the fixings so I found the square was essential to line them up:
10m centre parts and 4m elements fitted:
Next up were the 6m elements. On my initial inspection I assumed the 6m elements were meant to slide together on the ferrules and be retained by a centre punch as I noticed one ferrule was centre popped as supplied. But a closer rummage through the fixings bag revealed 4 short self tapping screws. There seemed no other sensible place to use these so I figured they must be to secure the element halves together. In this picture the centre pop by the pink pen was as supplied, the pop on the right is mine:
I then set about drilling pilot holes and screwing the tappers in securely: The thin marker pen lines are the edges of the Stauff mount so I didn’t drill in a stupid place!
After fitting the 6m elements it looked to me like there was a design flaw about to hit me. When designing my dipole boxes for my elements in 3D CAD I always fit a model of the feeder into the assembly to ensure the coax will pass any nearby elements. I fetched an LMR400 patch lead and proved myself correct. No way at all was my feeder screwing onto the dipole box!
Hmmm. Not great. There is room in the dipole box to mount the N-type higher to avoid this issue without any effect of performance. LMR400 is same size as common RG213, surprised this issue exists. It would be possible to reverse the dipole box but the 4m Stauff holders are the same size and even closer. Poor show.
So what to do. I’d already toyed with the idea of making a separate G3TXQ choke balun box for this as there is not enough room in the dipole box to fit an FT240 toroid but I was keen to press on so thought I wouldn’t bother. But now it would give me the opportunity to utilise a short connecting cable between them to overcome this issue. So I decided that would have to be the way to go. For that I would use RG316 as it is higher rated than RG188 and also because I couldn’t fit my RG142 one into the box I had as RG142 is too stiff! This is my G3TXQ choke on 2x FT240-52 cores:
So I then also decided I may as well give Clive a free upgrade and swap out the RG188 for RG316. Bloody good job too! On opening the box again on the bench with my glasses on I spotted this!! Centre of coax NOT soldered to the dipole half!! Very poor I am sorry to say.
RG188 ugly choke replaced with RG316 version for when I return the loaned antenna to Clive G8LNR:
With the dipole centre re-fitted I now started to fit the rest of the 10m elements and immediately found another issue not covered by the VPA Systems implementation mechanically. Even fully tightened up, the nut and bolt style jubilee clip (much better than worm drive IMO) could not compress the outer tube enough to grip the inner tube. I could still move it quite easily by hand. The DK7ZB designs do use all metric tube and some of the telescopic joints do have a lot of ‘slop’ to take up. My 6 element 50MHz DK7ZB driven is exactly the same. The dipole ends came with just 2 hacksawed slits (non deburred by the way) which isn’t enough to take up all the slop. You can just see air space above my thin blacker marker pen line and the slit fully compressed together:
My only recourse was to dig out the junior hacksaw and add some slits at 90° to those supplied. That was just enough to get a secure grip onto the smaller section though you can see the edges of the now 4 slits are still almost touching:
Another point of interest now. In the previous pictures you could see my marker pen lines for the position of the smaller 10m section. That was marked at my simulated position to slighter lower the resonance point of 10m from the DK7ZB design so I could use it with a better SWR at the digital end but still OK in SSB. The DK7ZB design page recommends adding 150mm to the extended elements. VPA systems have taken another needless shortcut here.
Bearing in mind the ‘slop’ mentioned above a very short insertion means the smaller sections will hang down needlessly due to the internal play rather than gravity induced flex.
Anyway, moving on, the ends of the 10m moxon are fitted with the supplied bolts. As an aside, during a small brain fart induced by the way I had the two 10m end assemblies laid out together I thought someone had assembled the two reflector ends together and the two driven ends together. So I tried to separate them to correct that. It was NOT happening. The insulator is really firmly fixed into the the ends of the elements. Once I realised I was being stupid we were all good but with the knowledge those ends are NOT coming apart in the wind.
With all the elements now fitted the G3TXQ common mode choke and linked it to the dipole with an RG142 cable and fitted it to my home brew mast. (Ignore the aluminium angle arm-I moved that to the same side as the boom after the picture was taken):
Time to wind it up and test it.
I forgot to save any analyser plots of the initial measurements as we were keen to get it working more than anything.
First tested was 10m. I had already simulated the DK7ZB design as mentioned earlier so I had set the 10m section to my calculated width intended to slighter move the SWR dip lower in the band. As it turned out the dip was about 28.4MHz and the antenna was perfectly usable between 28.0 to 29.0 with the worst SWR about 1:1.4.
Next I looked at 6m. It was miles out. I expanded the sweep range and found the best SWR dip at 49.83MHz. I then put the sizes of the 6m elements as supplied into my 4NEC2 model and found the exact same results: Now at least I knew I had a known reference point to fix this.
4m was even worse. This was even further out but too high in frequency meaning there was no cutting to be done. I decided that although tri band would be nice my 7610 doesn’t have 4m so I would need to set up the 7300 so for now I would just ditch 4m. We removed the 4m elements and wound the antenna back up to check things hadn’t changed much. 10m resonance changed very slightly, no noticeable effect on 6m, which is encouraging for adding 4m back on at a later date.
I used the simulation to find out how much to shorten the 6m elements by then dug out the hacksaw and cut off exactly 12mm from each end.
On 10m I looked to see how much I needed to move the ends out to get closer to my preferred lower resonance point. Really it was more than 20mm, but as there is so little tube in the joints I figured 20mm would have to do.
We wound the mast back up and the SWR curves we really good, just where I wanted them.
The proof of the pudding is of course, does it work? I had a couple of hours operating on FT8. 6m wasn’t that great though I did get a couple across the pond, my first ever on 6m. I spent more time on 10m and got some decent DX. I have only ever made about 5 QSOs before on 10m with my cobweb so it was good fun. 6m QSOs are orange.
The same weekend was the 50MHz R1 contest where I had a few hours operating. I missed the best Es but got a few, and around the UK in very poor conditions locally:
Conclusion.
Whilst I appreciate this antenna is more at the budget end of the market and is nicely lightweight which I favour personally, I must say I am glad I didn’t pay for it myself. The ‘ugly’ balun is a shortcut that might very well be adequate (IF it was actually soldered !!) and work but the miles off tune 6m and 4m elements cannot be forgiven. It took me minutes in the free 4NEC2 program to prove that the 6m would never work in band. For someone who is less ‘handy’ with making antennas (eg someone likely to buy a ready made antenna…) as it was supplied only 10m was usable and that only if spring tension in the RG188 happened to make contact with the driven element.
It’s a great shame as it’s so nearly there. More attention to detail (at no real extra cost – better slits in the 10m centre parts, move N type up so you can use decent feeder, few more inches insertion on the 10m outer elements, check soldering is soldered etc) and this would be a viable alternative to making your own. For me it was a simple task to adapt what I had to work and saved me some many hours of ordering and making but I do know for less cost I would have an almost equally lightweight version but one where it could take a pigeon sitting on the ends of the 10m moxon.
Some great mast guying tips……in my humble opinion of course!
I love guy ropes. All masts are just so much safer well guyed in my opinion.
I recently made my own 3 section aluminium winch up and luffing mast for the base station (must detail that one day!) that needed guys for my own peace of mind. So I wanted some good secure ways of keeping is safely guyed when up and retracted. So the below is what I came up with with some experience in other hobbies and some research. I’m really pleased with it so thought I would share.
Anchors.
The garden is very small and fully paved but does have a 6 feet or so brick wall on two sides and a concrete post on the 3rd that I could use to guy to. So bolting a fixing to the wall was the obvious answer. I’d already bolted some eye rawl bolts in before but they were too small for the snap links I wanted to use and they didn’t fill me with 100% confidence. What came to mind as the perfect solution was bolt on hangers used in climbing walls for clipping the top ropes into. I could then bolt through the bricks and it would be bombproof. They look like this (though I would be using them upside down as our ropes go up not down!): I got stainless ones from Needlesports here (link) for under £3 each as they will be outside 24/7.
Snap Links.
These are similar to karabiners but steel and a lot cheaper! I use ultralight karabiners for my backpacking radio gear but here weight is no issue. These are from Screwfix (link) and are cheap as chips (cheaper-chips are expensive these days!) working out at a quid each.
Although these are only zinc plated, I have been very surprised at the resistance to rust. Over a year outside and no signs yet. One thing I will point out is the engaging teeth on the opening gate are very sharp and love to try to snag the rope or your finger, and the gates themselves don’t open that wide. Climbing karabiners are designed specifically to have a good width gate opening, these don’t need to so they don’t. But in this application we only clip the rope in once anyway.
With these fitted I attached the 3 guys to the mast, wound it up to full height then tied each guy into the snap link using a figure 8 knot. This does take a while as getting the knots in the right place and the right tension on each guy to keep the mast upright is a bit trial and error but it only requires doing once and you then know the guys will forever be the right length.
Guy ropes.
While we’re at it, there are many ropes you can use of course but I have gone for braid on braid polyester rope as used a lot in yachting etc. It’s quite low stretch and has good strength and should be durable in the sun and rain. I use it on all my guys now, 6mm for my lightweight backpacking guys and 8mm here and for car portable guying. I buy mine from OutdoorXscape (link) as they seem to be good on service etc. Usually I buy white with colour fleck but for this I got solid black to be low key.
Tying off when retracted.
I wanted a way to quickly and easily tie off the guys when the mast was fully retracted as I wanted the guys to take the strain of the wall mount holding the mast up but for it also to be reliable and secure. After some considerable searching I found the perfect solution. The Nite Ize Figure 9 Carabiner Large (link) turned out to be perfect. The locking off system is very quick and you can also actually set it with a decent tension. It literally takes about 3 seconds to do. I also add a half hitch to stop wind/gravity/other factors trying to release the rope, so even then it’s like 10 seconds per guy rope. These were not cheap at a about a tenner each but they are worth the lavish expense 100%.
This is the basic tie off method which I have used especially when I plan to wind the mast up next day:
The image below is how I typically leave each guy when mast is stowed retracted. You can see the safety half hitch and the spare rope I loop over the snap link. The knot at the bottom of the photo is the excess guy rope I should really cut off as I will never use it. You can also see the now redundant rawl eye bolt. Might be handy one day…
Hope this page helps you with some ideas, I have to say again, I really love the Figure 9 karabiners!
Most of us that operate portable in any way will no doubt have had interested people come to ask us what what we are up to. My portable operations these days are almost exclusively contesting and single op so I don’t really want to break off a pileup to answer questions so I decided I would make a sign answering most of the questions before they ask them! I have found it to be very effective especially when I am tent portable on a mountain summit where there is a fair amount of walker activity. Often when I am in the tent listening around the band before the contest kicks off I can hear one person of a group reading out the card to the rest of the group. It’s been a great help. You may have even seen it in one of the Practical Wireless magazines in the PW 144MHz QRP Contest results write ups.
I printed and laminated it then peg it out with tent pegs from the pound shop between the summit and my tent: It’s survived some pretty hairy and wet weather so far. I also have another one for car portable that I put in one of the windows (doubles as a sun shade in the summer evenings!)
A few people have asked about doing one themselves so I have add below a Word DOC file and DOCX file (both ZIPped up) for you you use as a start point for your own version, you can just change the images and wording a bit to suit.
It’s time that the saying about sliced bread was replaced with 3D printers. They are right up there with the internet in ‘how did we manage without it’ in my opinion. For the radio amateur they uses are never ending!
The yagi below is completely a assembled with 3D printed parts. Not only that it’s mounted to a 30mm pole with 38mm U bolts using 3D printed reducers and the guy ring has an integral 3D printed plastic bearing.
I just built a 20m moxon made with loads of 3D printed parts:
Including a custom made for the job dipole centre with strain relief and tie wrap slots for the wires and the choke:
The things you can make are really only limited by your imagination.
Here is a time lapse video of a yagi dipole box being printed:
Ok sales pitch over, you know you NEED one, what now?
Well first choose your printer. I went for the Creality Ender 5 which is at the higher end of the budget printers. Loads to choose from though.
To print anything first you need a design. This can often be downloaded from a site such as Thingiverse or you can design your own. To design your own you need 3D CAD software. Don’t panic, it’s not as bad as it sounds. There are many free options you can try like FreeCAD or TinkerCAD etc. I use Solidworks as I am lucky enough to use that for my job. Here is the box from the video in Solidworks: Once you have your design file you export it as an STL file (downloads from Thingiverse already are STL files).
The STL file then needs to be sliced for the printer to print. That is turning the finished shape into many thin layers, or slices, to make up the final part. Again, software for this can be free. A lot of people use Cura which is a free download.
Here is the box above in Cura ready for slicing. It all sounds complicated but once you have told Cura what printer you are using and what type of filament you have you can often print with the default settings very well.
Here is the 18th layer of printing that would make up this box:
Once you slice your design you save the output to a card or send it via USB and pass it to the printer and the magic begins.
There is loads of help and advice on the internet for 3D printing but it is a really useful tool for the amateur radio home builder.
Having recently received my new (2nd hand) short contest callsign I was keen to air it and with probably the biggest contest in the world, CQ WW SSB, coming up I decided I would try a single band entry and build a moxon that was as lightweight as possible to use after how well my 50MHz moxon performed for me.
First step was the trusty Moxgen application and give it the wire size to get my start dimensions. As with the 50MHz moxon I then used 4NEC2 to recalculate the sizes for insulated wire as I was using.
Here are the calculated dimensions above applied to the 3D design described below. Note these dimensions apply to 16AWG/1.5mm2 tri rated wire. Other wire sizes with different insulation thicknesses will be different.
Once I had the dimensions I modelled up the wires in Solidworks to work out how long the spreader poles would be. This turned out to be about 4m so I ordered from eBay two types of 5m long fishing poles for evaluation. To do so I weighed a known length of the wire I was using to get an approximate weight each spreader would support and used something that weight to test how much sag there would be. The windjammer style pole had a thicker top section and was best.
Once chosen I could model up the poles themselves, weigh them and set their weight correctly in Solidworks.
To make the hub, rather than a thick metal plate and chunky U bolts and something to stop those U bolts crushing the poles I decided to use lightweight 1.5mm thick sheets of aluminium in a sandwich arrangement using 3D printed parts to grip the fishing poles without crushing them. I am a big believer in ferrite chokes at the feed point so I wanted a smaller support arm for that also. After my ultralight cobweb build which was unbalanced due to its 5th arm for the feedbox and choke I also wanted this antenna to be balanced, particularly as I intended to use this on the top section of my fibreglass push up mast which is 23mm in diameter.
Solidworks is great because it will also tell you the centre of gravity of your design, which I used to find the best places to put the boom supports on my yagi designs. So I was able to see how lop sided the design was, which was quite a bit. A ferrite toroid with coax wound around it on the end of a pole is a fair imbalance. To counter this I moved the mounting point towards the feed point so the main weight could help counter it. There was still a lot of imbalance so I decided there would have to be a support for the reflector to help counter the feed arm and it would also keep the reflector in better shape in the wind. Still not enough and I didn’t want to increase the aluminium plate sizes. Hmm. So I made a wire guide for the reflector support (exactly like the ones for the spreaders) but used a relatively fat chunky M6 bolt and washers for the mass and that pretty much balanced it according to Solidworks.
The pink two way arrow shows the expected centre of gravity:
One beauty of 3D printing is you can make exactly what you need so I could make pole clamps the exact size for the fishing poles and also arrange the poles compactly together (needed for the balancing) that you couldn’t do with something like Stauff clamps. I was also able to save on a bolt per side, every little helps. I started with the main pole clamps so I could test if my thin sheet sandwich would work as expected. As the holes were all in odd locations I decided I would mark out by printing a 1:1 scale print of the 3D model of the plate and taping it to one of the sheets and centre marking the holes ready for drilling. I then taped then G clamped the two plates together to drill the holes exactly in the same place on both plates. The holes in the plates are 4.0mm and so are the holes in the 3D printed parts. All the M4 bolts fitted perfectly.
I assembled the poles on and gave the assembly shake. All seemed OK.
I was able to press on with the design and printing. To mount the antenna to a 23mm fibreglass pole without crushing it I got some 32mm U bolts and printed some load spreading clamp halves. Also I was able to print some stiff but light supports to join the ends of dipole halves. Pretty much had a full kit of parts.
Final drilling could now take place.
I did a test fit to the pole to ensure it fitted OK:
To connect the feeder I used a panel mount N type socket. I used RG316 to go to the feed point as it’s lighter than RG58,can handle more power being PTFE based and also being PTFE based can be soldered without it melting. I drew up and printed a plate for the feed point with holes to stress receive the dipole wires, slots for cable ties and clamping on the back to attach to the support arm.
Even though it’s “only” 14MHz I still like to keep the connections tidy.
The coax part done I can start final assembly. The N type socket had to be fitted first as it would be very awkward once built.
Once mechanically assembled I took it out the front of house to have enough room to rig up the wires. I allowed an extra 20mm on each end to allow for tuning. Another good thing about a full 3D model is I was able to fit my 3D printed wire guides tot he spreaders using measurements taken off the 3D model and as seen below all was spot on.
I collapsed the poles to carry it round into the back garden to fit to the mast. Once all re-extended the first thing I wanted to do was check the balance by extending the 23mm section fully to see if it leaned over. Looks good!
Pushing it up to about 30 feet to test I trimmed 10mm off each end to arrive at the below plot. Almost perfect match with the simulation.
Time to test it on the air. Hmm. SWR with real RF way up in the red. It couldn’t be but was. I wondered if the choke wasn’t just not working but was working against me? It was meant to be 17 turns of RG58 so I hoped tightly packed RG316 would have same effect. I tried a GM3SEK choke my end in the shack, no good. Hmm. Talking with a couple of friends one suggested taking the N type off the plate.
I couldn’t think this would have any affect but it was worth a try before I started cutting things.
Well snag my blade! All was good now. The metal plate attached effectively to the braid of the coax certainly had some negative effect!
So now I had a working 20m moxon beam.
As it turned out come CQ WW SSB weekend it was too windy for me to use the moxon, but I have had great success working USA on FT8 since. The below is where I was heard for a few hours with the moxon aimed for USA and working mostly USA and Canada.
Final weight of the finished moxon was 3.2kg and it does compact down small enough to go on a roof rack, possibly even inside my estate though it could be a squeeze. I aimed for the SSB section of the band but it would have probably been to aim lower in the band to get a more usable SWR at the CW end. But performance wise I’m very happy with the new moxon.
Edit: October 2021
Since the initial build I have since soldered a little wire to the end of all 4 wire ends to move the resonance down to the digital end of the band. And since that I have made a 3 section wind up aluminium mast and can now use a rotator instead of running downstairs and up the step ladder!
But more interestingly, I re-measured the moxon at about 12.3m above ground on the new mast with LBC400 coax so very low loss at 14MHz and the whole band is below 1:1.6. (I really must set the date on the analyser!)
At my location it’s very difficult to have any sign of HF antennas (or any antennas for that matter) up in the garden. I’m already using a covertly erected dipole after dark using this quick erect fishing pole mount. But I was very interested in an ATU free multiband antenna. The hexbeam is nice but too big and needing a rotator. So I looked at the G3TPW Cobwebb antenna. For those unaware this is a 5 band dipole based antenna horizontally polarised and roughly omnidirectional working on 10, 12, 15, 17 and 20m. The folded dipole style with shorting points seemed more complicated than I fancied so I looked at the simpler G3TXQ cobweb design.
This filled more boxes for me, single wires nice and easy to tune and a tidy looking feed box. However, it would still stand out in the garden due to it being some strange looking (to non hams) spoked wire thing. What I decided to try was upsweeping the spreaders to make the antenna look like a rotary washing line as there is already one in my garden. I was counting on the neighbours not noticing it had grown an extra arm and got a bit bigger. Except when actually in use I would keep it low down like a normal washing line such as this:
I searched the internet to see what people had designed but no-one had done anything quite this style. I was sure it would work due to the success of the hexbeam, but to be sure I started a thread on qrz.com and got some advice from the man himself, Steve G3TXQ, confirming there should be no reason why it wouldn’t work. Thinking about it the wire positions on this design are quite similar to those on the successful hexbeam.
Looking at existing designs all seemed to be based on heavy aluminium plates and tubes to support the spreaders. My plan to use this on a fishing pole would preclude this sort of design. Mine would have to be ultralight to have any chance of working. The spreaders were easy, I would use the 2nd and 3rd sections from telescopic fishing poles like the ones I use for masts. These two sections are long enough (bearing mind the upswept angle means the spreaders need to be slightly longer than for a flat design) and are plenty stiff enough. At the bottom they are approx 20mm diameter tapering to 8mm at the tip. Five 4 metre poles were soon ordered from eBay at a cost of £37.70 with the delivery, making them just over £7.50 each. Here they are next to an 8m and the 10m fishing pole the finished cobweb will be used on:
Next problem was the centre piece which came to be known as the spider. I mentioned what I was looking to do with a non ham engineering friend and it turned out he was about to get a 3D printer on loan. A great answer to my problem, a custom designed 3D printed ABS spider would be perfect. I got straight onto to modelling up the antenna in a 3D design program. This would also allow me to be able to confirm the fishing poles would be long enough and see how it looked, find where the wires should be based on the wire lengths used by G3TXQ and also predict quite well the finished weight. After searching the web a lot for everything to do with the G3TXQ cobweb, many threads on qrz.com with posts from Steve himself, I designed it to have the feedbox in line with the 17m wire, making the 17m wire a square and all the others bunched into the feedbox.
Here is an animation showing how the antenna should look when finished:
And here is my original design of the spider:
I sent the CAD file over to my friend for 3D printing and the printer reported it would take 95 hours to print! So he decided to strip it down to just the 6 bores (five spokes and bore for the pole) and rebuild it from separate parts that would be glued together after printing. The glue should be stronger than the ABS he assured me.
My design was originally based on the main spider with a slightly tapered bore suitable for the 5th section down on my 10m pole. Not where I planned to use it, but hedging my bets until I got one working and mounted on the pole. I would then use one of 3 sleeves to move the antenna up a section per sleeve. During the redesign we were able to come up with a better idea using end caps and also save a bit of weight, all good.
This is the new design:
And on this page the printing process and spider assembly is detailed in more detail.
The spreaders slot into the spider bores and need to locate on a pip in the bores to stop any rotation (seen in the spider images above). This is mostly for the feed box to keep it in position but also to securely locate all the spreaders. The ends of the fishing poles are not cut square so I marked a line to indicate the longest part and marked out for the slot to be made with a Dremmel 3000 and a small milling like bit:
To Dremmel the fibreglass safely I set up the hoover as a dust extractor which was very effective:
Finished notch, which is done to all 4 spreaders and the feed box spoke:
So I knew the notch was deep enough I marked the spreader fitted before the notch is made, then added another mark 3mm up the pole (depth of the pip):
Spreader fitted after notching:
Once this stage was finished I extended each spreader and feed spoke in turn and marked the smaller section at the end of the larger section to indicate how far it fitted, then took them back apart and applied some varnish to the part of the smaller section that makes the joint and extended them again adding a little extra force to ensure they are jammed nicely and hopefully the varnish should ‘glue’ them together. I then added some varnish to the outside of the joint to seal against water running down the joint and inside the spoke pooling in the ferrule. You can see in the picture below that the spreader on the left is going to need another application or two of varnish to seal that gap:
The feed box fits on the feed spoke suspended underneath the feed spoke pole. The angle of the feed spoke bore on the spider is slightly steeper to align the wire entry of the 17m straight in, one nice advantage of 3D CAD design.
The feed box is an ABS box from Farnell, their part number 244-4686, manufacturers part number TW7-5-11B. Note that Farnell do free delivery to non company accounts now. This size was chosen to be the smallest usable and (hopefully!) fit the feed balun inside:
The buses for the wires I wanted also to be lightweight but not compromise on the amount of copper so I abandoned my plan A of some copper clad FR4 when I found some 0.32mm copper sheet. Perfect. That equates to 10 ounce copper where the FR4 copper would have only been about 1 ounce and I have no dead weight in FR4:
As copper sheet is pesky to solder to I wanted to tin one side before trying to solder even more copper to it in the form of the balun and wires. First attempt with a heavy iron and a hot plate was not very successful. The solder looks awful and you can see the copper discolouration on the bare part before we gave up:
A plan B was required! A multi zone reflow oven and some solder paste was much more successful! Not something everyone has access to but I do so I may as well take advantage of it. The scratches are from scraping off the flux:
To mount the copper flat to the base of the box to maximise space for the balun and simplify the build I needed to Dremmel off the four PCB moulded mounts. For good measure I removed the same mounts from the lid. No point carrying dead weight even if small!
As I like things neat and tidy, I have taped the two copper bus plates back to back to drill a small pair of mounting holes through both together so they match:
Here are the two bus bars fitted with M2x6mm screws:
Instead of nuts and bolts to join the wires on I wanted to minimise weight again so just soldered directly to the copper. Not so good to replace wires but I like to bridge cross problems if and when it’s needed. As I didn’t want to waste space in the box and add extra weight there are no spacers for mounting the copper bus bars. They will sit flat on the bottom of the box. But as I will be soldering with it in situ, especially for the fiddly to fit Guanella current balun, I have added some Kapton tape to the box to hopefully protect it from the heat a little.
The wire I used is 16/0.2mm stranded which measured to be pretty much the same as G3TXQ describes in his text though in his photos the coloured wire looks a little chunkier to me. For the actual lengths I started with the lengths given by G3TXQ here:
I then referred to his wire thickness adjustments for hex beam wires here: to see how much things changed for wire diameters then decided that an extra 3 inches each should hopefully be more than enough. I also took note of some predicted and measured figures for the amount the resonant frequency changes when trimming provided by Jacek SP3L in the disguised cobweb thread higher up the page: Changing the wires with my construction would be a nightmare so hoped this extra length would be plenty.
Where the wires exit the feed box I’d planned to use Hellerman H20 rubber sleeving but it was too loose a fit to the wire, so instead I used adhesive lined heatshrink which would also prevent water ingress between sleeve and wire.
To solder the wires to the copper bus plates I removed them from the box and placed them onto a piece of MDF to prevent heat sinking away. Before doing so I drew around the box and marked the wire entry points so I would be soldering the wire in the right places. The pre-cut to length wires were then threaded into the box treble checking the order before soldering:
Using a 100 watt Weller mains power soldering iron with a 6mm chisel tip I was easily able to solder the wires on with sound looking joints and without melting the previously soldered joints:
Once cool I fed the wires back into the box and fitted the plate and repeated for the other side:
I’m using an N-type socket for the feed input connector as I much prefer them over the SO-239 and PL-259 UHF connectors.
Having created the design in 3D CAD I was able to use my model to tell me how long to cut down the fishing pole for the feed box arm. Using the end of the centre spider tubes as an ideal reference point for a tape measure told me the spot to mark for cutting, in [inches] and mm:
As things worked out the position for the feed box happened to span the joint of the telescopic pole. I had originally planned to use saddle clamps to attach the feed box to the pole, using the same screws for those to hold the bus plates in but I needed two sizes neither of which I could find and to save making them I just bolted the box to the pole with 2x M3 screws and nylocs. I used nylocs so I could tighten up without crushing the fibreglass pole but not have the nuts work loose.
Once the box was fitted to the pole I could then fit the balun. I wanted to fit the feed box to pole first in case once the balun was fitted it was blocking the access to the screws (quite likely). To try and reduce the damage to the edge of the box during soldering (the iron ALWAYS finds a way of melting the box when you are working inside no matter how carefully) I cut up a Pepsi can (other canned drinks are acceptable!) and folded it around the edges, hoping it would make a good heat shield:
For the feed balun I used the exact same FT 140-61 ferrite toroids as G3TXQ and RD316 (which is just RG316 with a double screen)
4 lengths of RD316 taped out for marking exactly the same lengths. Apparently the outer jacket of RD316 is repellent to all marker pens including biro, Steadler permanent markers and sharpies. So I marked the line to strip the braid off with masking tape then slit down between each piece to leave a strip on a marker: Outer jacket removed and bits of heatshrink added to keep the 2 coaxes together. This also helps stop the coax curling up all the time:
Close up showing quite closely matching outer jacket length remaining:
Both 1:1 chokes wound:
To fit the balun into the feed box I first loosened the four screw holding the two bus plates in so I could slip some thin card (from a cereal bar packet) under the ends of the bus plate where I would solder the coax (paralleled end) then soldered the coax to each plate. Then I removed the card and tightened the screws back up on the bus plates and added a drop of varnish to each one. I then fitted the other end (the series part of the balun). This job didn’t turn out to be as much of a stressful procedure as I expected. The heat shields worked perfectly and the box survived without any damage! I used some Hellerman rubber sleeves where the wires exited the feed box
As this project has taken me so long because I started make VHF & UHF yagis for contesting and doing lots of contesting, in that period I have moved away from using the fishing pole as I now have a more heavy duty pole this can go on. I did test mount it on the fishing pole and it is still viable. But it fits even better onto my 23mm diameter 3mm wall mast top section. To do this I designed and 3D printed adapter washers and simple locking collars to fit top and bottom. In the first photo you can see that without my glasses on I have fitted the collar upside down as the ridge is supposed to mate with the groove.
The wires are fitted to the spreaders with some 3D printed guides I drew up and printed. The wires pass through holes so the spreaders should self centre and tension be equalised all along the wire.
As my design is closely following the build of Steve G3TXQ I was able to use 3D CAD to determine where the wires should be attached to my spreaders if all goes to plan and at the very least give me a good start point. This could of course change if my wire turns out to be thinner or have a different velocity factor from the wire Steve used. These are the positions using the same wire lengths as on Steve’s page:
To join the ends of the dipole I used nylon cord as used on bathroom light pulls only a bit thinner, thin, strong and light. I used a tight fisherman’s knot in the cord to grip the wire and taped the string to the wire ends to keep the wires straight at the ends whilst tuning then locked off once tuned with a cable tie as well.
Tuning I did with the extremely useful MetroPWR FX700 antenna analyser, taking notes of how much I cut off each time affected the resonant frequency. I decided my target tune frequencies would be somewhere near midway between the SSB and data portions as from home data modes is very family friendly with no shouting to annoy anyone.
SWR results are as follows, measured at about 30 feet above ground.
Finished antenna pretending to be a washing line. It sure looks like one but I don’t think I quite have the disguise cracked…
As this started off to be an ultralight cobweb it’s worth stating the weight. Weighing myself alone and with the cobweb complete on the bathroom scales it comes out as 1.2kg overall.
Final Note:
It is with sadness that this took me so long to complete, and during that time Steve G3TXQ succumbed to illness and is now SK. Steve had some thoughts on the antenna that we exchanged, both via email and in this QRZ thread. It’s a great shame for me he never got to see the finished antenna he inspired. However the QRZ thread did stimulate some good experimentation on cobwebs and is well worth a read through for that alone.
Earlier this week I built a simple wire moxon antenna for 50MHz for use at home, detailed on this page (link). The impetus for this build was to contribute towards the Hereford ARS campaign to retain the RSGB VHF Championship Trophy that we won over the year long 2019 contesting calendar.
I already have a 50MHz yagi (link) for use portable when contesting but at 7.2m long it’s just too large to assemble in the garden and if I could get it to the height required for a clear view would put the fear of (deity) into the whole estate! So the plan is to add a few club points using the little moxon.
As you can see a vast difference in antennas! However I was not completely dismayed. I have been out contesting on the hills often with the big yagi and really struggled to break pileups to DX on sporadic E openings, usually with many other club members who I usually do better than just calling over the top of me. I’m always low power at 10W, but look at that beam man! So I have come to the conclusion that the vertical pattern of the long yagi is just not suited to the angle required to get good results on Es. I may be wrong as many others use reasonable beams but my post summer 50MHz contest catchphrase is:
“I bloody hate Es season!”
But the radiation plots of the moxon design I thought were very promising. Not only a great big fat front azimuth lobe (in red on graph below) with a 3dB beam width of 78° but a huge vertical shape )in blue on the graph) to it covering all take off angles you could imagine for Es. So despite having way under half the gain of ‘the beast’ that disparity in gain once we get off the main heading the moxon would ‘catch up’ a little and should also be launching skywards when required. My theory anyway!
So, contest day. I planned to do just 6 hours and as it’s more family friendly to get that over with in one session for me so I opted for Sunday morning and the last 6 hours before the contest ended. Amazingly for me I seemed to have pretty much cherry picked the best slot of the 24 hours. I was 4 hours into the contest before I got any other G squares than my own and IO81 adjacent to my square (I sit on the join line at IO82PA). But I was busy with the Es! Great fun. I only have 100W max on 6m and whilst some of the pile-ups especially to EA were hard to break it was very enjoyable! One station ( I think Italian) even commented “wow such a huge signal” after our QSO. This is on two bits of wire and some plumbing tube!
I was completely unable to raise all the ‘local’ EU countries like F, ON, PA nor could I get any GD, GI, EI, GU, GJ. If I had their country and square mults would have upped my score nicely. However my new temporary summer 50MHz contest catchphrase is:
“I bloody love Es season!”
As we are still on lockdown and my 50MHz yagi is literally too huge to fit in the garden let alone erect on my lockdown lash-up system I decided I needed to make something smaller to use at home. I didn’t have any aluminium tubing at home long enough to make a small yagi so I decided to make a moxon antenna on the recommendation of a friend. These are very compact and easy to make so it seemed like a plan. I decided on a wire based version as although I have some 12mm tube I could cobble together I didn’t have anything I could get today for the corners. Wire it is.
I’d already looked around the web and compared the various online moxon calculators and the AC6LA Moxgen program (link) and the Moxgen program seemed to be the best fit for the suggested spreader angles. (Even though I’m not using spreaders as such.) It’s dead easy, just put your desired frequency in and the wire size and click calculate:
That’s it, job done. Almost…
I’m using normal insulated wire but the calculator doesn’t cater for the change in velocity factor from insulated wire. So I decided to run it through the free 4NEC2 simulation software (link) to make the required adjustments to the dimensions so it would work with minimal fiddling after. I love building, hate fiddling. Now, before you start backing away from the PC this is quite easy to use and just needs some really simple maths to do this. Stick with me. Look at the image above and see I have selected Format NEC on the right. You just do that and click the Generate Model button and save a file ready to open in 4NEC2. Run 4NEC2 and click Open and load the file you just made.
Click that green calculator looking icon to bring up the next screen, choose frequency sweep and check the start and end frequencies are a useful range and that the step size is not too large, then click generate:
We then get a plot showing us the expected SWR curve of the antenna:
Oddly minimum SWR at 50.1MHz rather than 50.2MHz but looking good. We can click the green calculator button again and this time plot the azimuth plot we are interested in:
Which results in this plot:
We can see the moxon should have about 6dBi gain and the amazing front to back ratio it is known for. Next to nothing off the back.
This is all very well but I’m not make it with bare wire so my antenna will not look like this without some tuning. First of all we need to add in the insulation so we can see what the effect will be. On the main screen, to the left of the green calculator is a red book, for editing the information that defines the wires making up the simulated moxon. Select the Source/Load tab, and tick Show loads. We then add two lines selecting as shown below from the offered selections. For Tag, First & Last we put 0 (zero) which will apply the setting to all parts of all wires. My tri-rated wire has insulation 3mm in diameter so I enter the radius, adding the mm to ensure correct scaling is used:
Once done we can click the green calculator on that screen, select frequency sweep again, but widen the scale. I’ve gone 5MHz either side of 50MHz. You can see the resonant frequency has moved 2.5MHz due to the effect of the insulation’s velocity factor:
So if we had built to the Moxgen dimensions using my insulated wire we’d be looking at an SWR of about 2.2:1. So we need to make some simple adjustments. The dimensions of the wires are on the Geometry tab below. Looks complicated but it’s just a few repeated co-ordinates. Some with a minus ( – ) sign to make the equal around the center of the axis to plot correctly:
What we are going to do is put those numbers into symbols, or what we would call variables in programming. When you look above there is only actually 4 different numbers used so it’s not complex. You can give them any name you like, even Harold but I have gone similar to the Moxgen image further up. Just click the Symbols tab and enter as shown below. You’ll see a 5th value called Vf. I’ve already tuned it by now but pretend I haven’t:
Now we need to flip back to the Geometry tab and put the letters (W, E, DirS, RefS) where the numbers used to be:
If you were to run the SWR plot again now, it should be exactly the same, best SWR on 47.5MHz. But finally we add in the velocity factor. To the end of each of the letter symbols add without spaces
*Vf (star V f):
Now we can run the SWR plot and it will apply a Vf correction to every dimension. I found 0.945 by trial and error. They say Vf for wire is between 0.95 and 0.98. I knew mine would be the lower end as it is quite thick. Now if we run the SWR plot, changing back to the 49 to 51MHz range we get this plot:
Back in business. A point to note here. I’m aiming for the SSB section of the band, but if I wanted to cover more of the band I would still aim at the lower frequency with this design. You can see the SWR curves rises steeply on the LF side but gently on the HF side. Anyway, if we run the far field plot we now get this:
As you can see this has changed. A bit less forward gain because we are chucking some out the back now. Now as I mostly do contesting this doesn’t bother me at all but it is interesting nonetheless.
So now we have redesigned the original dimensions to my real world application of actual wire, we should be shooting for success. We just need to apply the scaling factor of 0.945 to the original sizes with a calculator and a little rounding to sensible numbers:
So for the reflector we need a wire 2051+384.5+384.5 – 2820mm long and for the driven element 2 wires 1025.5+305.5 = 1330.5mm long.
Now we can build the antenna and have a good chance of it working!
I decided to make it from a small length of 20mm boom left over from VHF yagis and find some plastic pipes from the DIY shop to support the wires in the A dimension and just stretch them between the pipes for the E direction:
I bought a terminal block strip with the intention of using the brass inserts with nylon rope to join the ends at the C direction but didn’t after so total cost of parts not lying around £3.87. I used my red yagi elemt plates to mount the pipe clips and snapped the pipes into them after cutting to size. They didn’t really grip the pipes so I just taped them on to stop them sliding sideways.
I fitted a short tail of RG223 to an N-type plug, split the other end into braid and core and soldered the two halves of the driven element on and threaded those into a hole drilled in the centre of one pipe. After testing I sealed this with liquid insulation tape:
I had a better idea for the ends of the elements at dimension C. I quickly modeled up a small plastic part like so and sent two to the 3D printer: The holes are a snug fit, tight enough to hole the wire until final testing. I marked lines on with a Sharpie 60mm apart and fitted the wires. Once tested I locked off with cable ties:
So, the acid test. What does it measure like?
SWR plot is very close but shifted down in frequency. This is because of the plastic pipes which I can’t (or don’t know how to) model in 4NEC2. Not a problem as I expected this and knew that now the wires would be ‘too long’. So I cut 10mm off the end of each element half (4 ends) and ended up with this: (just noticed my analyser clock is way out LOL)
Pretty much an exact match to the simulation with 30 seconds of trimming. Just how I like it!
The finished antenna looks like so. It actually looks better than this because this is before cutting the 10mm off each wire end.:
