that’s right, if you want it to range across the higher bands, it needs the capacitor. Doesn’t change anything on 40M. the very clever winding approach by KN5L alleviates the need for the capacitor used in either the Parfits or the AA5TB approaches. I am trying to wind one now!
and by the way… if you deploy an EFHW, there is no need for any coax. There is no radiation from the end of the wire, so, I just attach the UNUN and plug that directly into the radio. Of course this means I have to locate myself near the end of the wire. I run the wire as an inverted-L in part because some of the wire is vertical, and the higher current area is at the top of it, and in part to reduce the length of the wire that the pole has to support. I tape the wire to the pole four feet from the end of the wire, and that’s my feedline. I use a very flimsy pole, leaned against a branch or a bush. or the end stuck in a pile of rocks if no shrubbery. Also, using two traps, the wire length for 40M is shorted by the trap inductance to about 43 feet, of which 15 feet goes up the pole, so the horizontal segment is only 28 feet long. - fred kt5x
You are confusing the broadband and resonant narrowband designs.
The AA5TB designs have the cap across the secondary and are narrowband. The design Andy currently has is broadband and is similar to the HB9BCB and HB9EAJ designs.
You should read the article which includes detailed build instructions by Stephan, HB9EAJ at:
https://hb9sota.ch/wp-content/uploads%2F2021%2F08%2FPortable-7-Band-EFHW_HB9EAJ-V1.2.pdf
Thanks for the link Colin, saves me searching for it again.
I am not confused. I know that. Both work in their own way. - fred kt5x
I had also noticed once that the 1:49 balun got very hot.
I then got the antenna wire out of the wet leaves of the trees and hung more freely… and all was well. The bad SWR was probably the reason at that time.
73 Armin
Andy, is a trapped vertical antenna out of the question for you?
Sorry if you have already answered this question.
BTW, I just realized that exactly 10 years ago today I did my 1st SOTA activation with an Elecraft K1 and a trapped vertical. That would actually be a reason to celebrate, but only when those crazy storm clouds have cleared again, hi.
Heinz, I am simply playing. I’ve used 1/4wave GP verticals and coax centre fed dipoles for years. I have this trapped end-fed for convenience and it was only the discovery of the hot matching unit that prompted me to try to improve it or make something better.
As I have a few coils to wind I needed some wire. A scout about on eBay and there was someone selling assorted sizes of enamelled copper wire for a decent price. My 24SWG and 30SWG wire turned up nice and quick, excellent I thought, let’s get winding.
I just came to solder the wire and found I must have bought high-temperature wire. I have never tried to use any ECW before that the enamelling didn’t melt and evaporate when exposed to some solder on the end of a Weller #7 temp tip. Not this time, the enamelling just laughed at me. I tried the flame from the gas cooker and some wire wool and that worked. It took 45mins searching the newly reorganise shack to find my Weller #8 temperature bit and that does melt the enamelling.
#7 is 370C and #8 is 425C. I’ve been soldering through the enamelling of wire since university (1980) and this is the first time I have come across wire that needs an extra hot iron which I find quite amazing. I’m not sure if this a change to ECW that has past me by or if I have led a sheltered life. But beware… not all ECW is created equal.
Now I can solder the stuff I can build a new match unit…
Hi Andy,
Glad you sorted the soldering out. I am a scraper. Scrape the enamel off then solder. Use an ink eraser on the leads of tinned components to remove the oxide. Try to get some mechanical connection, then solder.
I agree with others that 2 turns in the primary on that core results in inadequate inductance and core saturation although I would have expected it to be OK on 20m. However Owen Duffy puts compelling arguments.
I connect my balun 2 m in from one end. Winding 3t of CAT5 wire over an enamelled 18 t secondary type 43 core.
Yes I have a coax connection, enough coax to not be under the antenna and not affecting it
The coax and the rig and me do not form part of the antenna. Much less agricultural.
No traps, covers 40/20/15 nicely.
73
Ron
VK3AFW
Hi Andy, I have always scraped off the enamel on such wire, tin it and solder it to what ever I am making.
Geoff vk3sq
Right, let’s bring this up to date.
I know the ready made unit works well enough for QSOs but is it a bad / fair / good antenna? The consensus is the match unit is sub par for 40m at least and looking back, 40m was the last band used and the coil was hot. Making a new broad band match will require some work on the lengths of the existing antenna. How much, measure and see or simulate it.
Rather than change a complete antenna that works to some degree I decided to make an AA5TB style match. I’ve made one before (and left it on a summit 
) so I know how to make them and how they work.
I know that AA5TB’s match will also find the lengths not perfect out of the existing antenna but it’s adjustable. The downside is it will put more current into the counterpoise but having a small counterpoise is not a problem. I made a unit in a small plastic box that has a 196pf Polyvaricon (a NOS Phillips unit from the 60s or 70s) and coil wound on a T68-6 core. That was the same as the one I lost but I went a bit further. I wasn’t sure just which impedance ratio to pick so with a pair of switches I can have 3 or 4 turns on the primary and 24 or 28 turns on the secondary. That gives 1:6, 1:7, 1:8 and 1:9.3. Trying with a range of resistor to simulate the high impedance of an end-fed I was able to always get better than 1.3:1 match with everything from 1k8, 2k2, 2k7, 3k3 and 5k5 resistors across the secondary on 7,10,14,18,21 & 24MHz. I’ve modded the original antenna so I can either use the original broadband match unit or this AA5TB design. The proof will be in the eating as I should be out with Paul W6PNG/M0SNA later this week where I’ll try it on 40/30/20 (and maybe some other bands along with 13cms).
I have a big box of toroids, many T50/T68/T80 sizes in type 2 and type 6 and a huge number of unknown things. Now some came from dead switchmode PSUs and were used for noise supression. They will be entirely the wrong material for this use and others maybe perfect. One was marked FX1588 which is an old (70s) Mullard Ferroxcube number. I found an old data book online which suggested it was a type 2 material. I did wind a 3-4/24-28 turn coil but it was non-functional above 3MHz.
Next is to make some of the other designs such as Stephan’s from scratch. And to buy a pocket VNA and do some experiments in categorising all the toroids I have to see if any are usable for this purpose. No point having loads of them if you don’t know what you can use them for and it will give me something to play with in my soon to be greater free time 
As usual comments favourable or otherwise welcome.
Andy,
There are already answers to your original question above, as well as some hints and recommendations.
When I read your reflections on the next steps envisaged, I’m not quite sure whether you wouldn’t mix up the 2 different concepts: (auto-)transformer (-> ferrite cores) and tunable parallel resonant circuit (-> iron powder cores).
For fine tuning on the bands 40/30/20m with any antenna setup in the field, no matter how unsuitable, a tunable parallel resonant circuit is recommended, e.g. like the people from QRPGuys use with their trapped EFHW here:
https://qrpguys.com/wp-content/uploads/2022/04/mbef_antenna_assy_021422.pdf
Once you know your needs, you will surely find the right solution for them.
Good luck with that then.
Do you already know the kit provider HB9ZHK?
The offered tiny impedance transformer (optionally for 10 or 15 watts) and the traps could be interesting for the self-construction of EFHW antennas for everyone as well as for not particularly practiced in tinkering and/or without a large component store.
I don’t own any of these products myself and don’t know the details. From the description it can be seen that the impedance transformer can be built for the transformation ratio desired by the user (e.g. 1:49 … 1:81).
Note: Just hope users will use 3 primary turns for good efficiency…
Also note that the efficiency of the impedance transformer is, theoretically, about 3% better with the FT-50A-43 ferrite core compared to the FT-50-43.
Of course, the transformer efficiency could be further improved by stacking 2 toroidal cores FT-50-43, the calculated efficiency would then be approximately:
Such an impedance transformer would certainly be usable up to and including the 15m band, then the transmission loss (s21) will probably increase too much.
Thanks for the links Heinz.
The AA5TB style tuned parallel circuit worked perfectly on its first outing. Why wouldn’t it? It’s not wise to draw too many conclusions from a single outing but the whole system seemed to work better than I can ever recall than when using the broadband match. It could have been ionospheric conditions or the physical arrangement on the mountain. (Barely enough soil to insert a tent peg then a huge rocky lump standing 100s of metres above the local ground.)
If we had had more time I’d have liked to compare the two match units and also against coax fed dipoles. But we were being eaten by midges and we had a long walk back. And I’d dumped all the other gear as I had kgs of uWave equipment.
I don’t have any type-43 material cores and the original has the windings glued in place so I cannot easily rework that. I’m just about to order some FT50-43 and FT82-43 cores to play with.
Thanks for the suggestions and guidance. This is something new to experiment with on the hills adding a bit of spice so stopping activations becoming hum-drum.
Andy, if you haven’t already placed your order, add some Fair-Rite 2643625002 (https://www.mouser.co.uk/ProductDetail/Fair-Rite/2643625002?qs=MLldULe7zY2ZDvtexqlGbA%3D%3D) There’s growing evidence from Owen Duffy, our Stephan HB9EAJ and others that this a good performing core for transformers to handle 10-20w. Efficiency helped by its increased surface area (see Owen Duffy’s recent blog posts). [Its worked well for me, but I haven’t done any comparisons so I’m not going to make any personal claims for it]. Its not an expensive core either.
Jonathan
Colin Summers MM0OPX has done exhaustive research on the subject. I’ve been following his tests on twitter. Eventually he produced this summary video:
Thanks Fraser, thats interesting, Colin also ended up using the 2643625002 for his QRP transformer.
Yes! That’s why I posted Jonathan. Great minds and all that.
Colin was testing various combos for weeks and posting regular updates on his twitter feed. I think he has been extremely thorough. Surprising results too.