Very nice work, was pleasantly surprised at you aproach. Very nice to watch.

Excellent research, great job. I love it when people go the whole 9 yards to answer questions like this

Wow Manuel! What an eye opener! Thank you for staying the course and putting in so much work to discover and document this. Real food for thought when running QRP and impacting the effective radiated power so much by introducing a matching device like this. 73 - KF6IF

I appreciate all the work and effort you put into your videos. Great help and resource for the hobby.

Thank you very much for making this study of EFHW un-uns

Definitely interesting video. Seeing how hot those small cores got was mind blowing.

That's a lot of building and testing. Thanks for sharing this with us, very helpful.👍🤓

Thank you for all the work you did winding the cores and building the spread sheet. This is very useful information. I knew wideband transformers were not lossless but didn't think they would be as lossy as they are. Thanks again. Dave K2ZU

Outstanding job, thank you very much for your presentation, the information was very useful.

This is really great stuff. Thank you. You are helping me learn a lot about practical antenna implementation.

Good engineering, Manuel. I will ejoy learning from your experiments. Especially valuable for me was the insertion loss and LOSSES POWER is applied, where core heating becomes a problem. I have always wondered about this.

Thank you for the level of effort put into doing, documenting, and sharing this work! I have some cores on their way to my house as I type. Some are toroids and some are called “multi-aperture” cores. Have you ever thought about using something like those? I am not sure they will give higher peak efficiencies but I am hoping for wider range at higher efficiencies. I will try and reproduce your results, thank you for giving enough info to do the test. I will try to do similar tests with the multi aperture cores.

Thank you! You made a great job. Your reseach asked for my question about efficience of the wide band transformers. Even at QRP power the losses are big and make a heating.

Nice work and experimenting. I have learned much from your video. Keep up the hard work!

Excellent approach and thorough findings of results! Best overview of the UnUn operation and efficiency…Have not seen such complete testing of its parameters! Certainly good to know for QRP RF and DC supply power issues. 73 WA4ITD

Extremely interesting results. Thank you

Thank you for taking the time to wind these cores and documenting the results. It looks like the differences in the physical winding between the step-up and step-down transformers may have affected the SWR numbers.

I used my endfed for the last SOTA and found I sit in one spot all the time and get a sore body. After my hikes up its nice to jump up and swap links on my dipole and that lets me stretch out a bit during the activation. By fluke I had a FT 82 43 to make my EFHW and it seems to run quite cool. Thanks for all the good work setting that all up and sharing with us the results. vk5cz ..

Very good video and very informative.

Interesting and useful measurements for EFHW transformers. It should be noted that the smaller transformers will have a lower magnetizing inductance so will have higher current in the winding. The higher current will result in a higher magnetic field strength H in the toroid core. H=N.I/(2.π.R). The larger cores will have a lower magnetic field strength H for the same current I. Also worth noting that the back to back transformer loss is due to core loss and mismatch loss. Both these effects make a direct comparison of results between transformers difficult to interpret.

Thanks Manuel! Very interesting. And yes oh boy does it take time to do these experiments. 😅

Thanks brother, very good job

Very, very interesting and informative.. Thank you for taking the time to wind all those transformers, and sharing your results.. John.. G4EIJ

Very well done Manuel 👏. I'm now working on a workable solution for the larger core I have been testing for insertion loss. It turns out that this what my myantennas use for their 250w (100w Continuous data) version. I need to get one of those digital power meters and I can do the "under load" test.

Thanks for looking into this. One thing we learn here is to avoid making assumptions without data. In the same vein we should not assume tuner loss values, they vary dramatically depending on the actual situation, configuration and settings and can be low or high. Tuners are easier to model than transformers, so we can mathematically get some data on tuners and we need to include that, as well as feedline losses in any comparisons of antenna systems. Also concentrating on a smaller frequency range we can optimize for lower losses. Trying to handle "all bands" isn't necessarily the right goal. We have better options for higher frequency bands with small antennas, there's little need to make the 80-30 meter antenna work on 20 and up. 73 de w6akb

Quite interesting!

That was excellent! Thanks for all of the work. 72 - AA4K

Thanks!

Sehr interessant

Thanks for all the testing, together with videos from MM0OPX your are both helping many see that there is a lot more to understating the EFHW UnUns. 73, M0KCB

Wow... that was one heck of a test. I did noticed that small toroids are not too great for the impedance transformer. Ohh by the way... I think the dipole will still resonate on the third harmonic.. I think thats how it is. So for 7MHz dipole will also resonate on 21MHz. At least that is what I know haha. 73

Thanks for sharing your investigations and creating a very informative video. The measurements at 5w with a power meter and a thermal camera were a great idea. While 20% losses sound terrible, it isn't necessarily a reason to abandon the EFHW for portable use, as some would have us believe. For comparison, I regularly use a resonant linked dipole as an inverted V, with the apex around 5m. On the 40m band this has a feed point impedance of around 20 ohm, due to the low height, leading to a 1.6dB loss in my 10m of RG174, and around 30% signal loss. It works just fine but probably isn't as efficient as an EFHW! The EFHW presents a feed point near to my operating positions so I can use around 2m of RG58 coax.

I have spend months building your trusdx...I got a very early kit...winding the toroids, soldering the relays, all the pins, jacks, switches, making the small tweak to the back of the display...I'm slow, evidently, and I had to do lots twice when i found my mistake from the first effort. hahah. And then. And then. I had to do the software. There was nothing on my board. I had to go back to the beginning of time and the first bit of code and start from loading the bootloader, but to do that I had to figure out how to operate a strange little device called a usbasp, and load the software for that, before I could try to load the bootloader, and what about drivers, and then I had to get the firmware working, but before I could do that I had to...well...you get the idea. So you have been on my mind a lot lately. hahah. After suffering extensively, I finally got the serial number to show. Oh jump for joy! The screen has a number! then to your site and load my call/serial #, and get that to work, then back to the little orange box of misery. And more trial and error, with more error than trial. hahah. What does this have to do with EFHW antenna core types? Well, after the orange box of now JOY came to life, I have to build a portable qrp EFHW antenna. So I have just spend the last two hours watching you like the high priest of EFHW build-ology...every word...every point...every consideration has been poured into my old age brain. I am soooo much the richer for paying attention to your work. If I had learned how to pay attention like that when I was in school, I might have gotten much better grades. hahah. I have really just one goal..get your orange box of joy, working with a laptop, a portable EFHW antenna, all using power from my ebike, in a park and making one, just one, park activation (POTA) in VarAC, or PSK or THOR (my current digital crush). And then i will call the entire adventure a success, and be happy. So, sir, thank you for the guidance. In a few more months I might be done. maybe. Don, KM4UDX

Very interesting. It would be really interesting to see the same test performed with 2x of the 37 cores stacked in top of each other. Many have suggested that the smaller diameter core can achieve greater efficiency (if not in saturation).

Excellent video, did you try nesting a 52 inside the 82 with the 3:21 winding?

Thanks a lot for your work! I wonder if part of the losses are caused by saturation/ clipping which could cause distortion and nasty harmonics. Someone with a spectrum analyser could perhaps find out.

I went through the snorting ferrite phase. After going through the 12 meter therapy I no longer use ferrites for anything other than a common mode choke with the coax wound through a large core(s) for 14 turns total (usually). I have used type 43 mix but I prefer type 31 mix for chokes. I have 2 EFHW antennas in use currently (40 and 20). I use coil and cap tank circuits to match them. I understand the desire to have one wire do everything with no ATU. I also understand that no matter how much I desire more money, that simply sticking up a gas station has undesirable losses. Like 15 to 20 in the slammer. When you are limited to flea power for pota/sota/marooned-on-the-ocean I do not understand why one would be happy with dumping watts into infrared. Blue watts matter too. Yes, your pota transformers are cold when you wander out and check them. That is the duty cycle and air cooling, not the efficiency. A wire with some alligator clips for keeping the length in the 3/8 or 5/8 lambda range on each band and a light weight tuner would be my choice. Now, if you are happy with your transformer then I am happy. Sayonara :)

Wow! I found the heating versus loss as well. Running 500W though. Even stacking three 2.5” diameter cores (two #43s around one #42) didn’t help. If I was foolish, I’d put a fan in the box with the cores ;-) . At which point I am tempted to use a random length of wire - maybe 45m to 60m - and my trusty remoted Z-match. Even though tuners are supposed to be lossy, I have yet to see the heating/loss problems of the EFHW + unun set up in my Z-match tuners. Thanks for this very informative video, Manuel. Stay safe & healthy! 73 de W8IJN

Excellent and hard work. Could you please test the hypothesis that the antenna tuner will swallow more energy than the 1:49 transformer? 73 de KD8ZM

Thanks for doing this! Have you explored tight standard winding patterns (tapping like a standard transformer without the crossover)? I have seen general improvements by moving away from the crossover and keeping them tight. Also, take a look at Evil Lair's videos if you get a chance. He has some suggested alternative cores that seem to work better than the standard fare.

Can you recommend or put links to power amplifiers that I can use with your little 5 watt QRP rig? Maybe output to 20 watt or higher.

Great video! Lots of useful info. I want to wind a nice 1:49 with FT114-43 core for use with IC-705 and PA500. That means I may be doing digital modes at 30 watts. What wire gauge is appropriate? Thank you.

Eindrucksvolle Arbeit. Danke dafür, dass du die Erkenntnisse teilst. K6ARK wickelt seine portable endfet anders. Wenn ich das richtig in Erinnerung habe passt er die paar und sechzig Windungen an. Kannst du vielleicht den Hintergrund des anderen Vorgehens erklären? Ich frage mich wie es sich die Dämpfung verhällt wenn man primär oder sekundär schrittweise Windungen erhöht oder erniedrigt... sicher transformiert man dann anders und der Strahler müsste verändert werden. Bekommt man so vielleicht eine Single Band efhw??

It would be very interesting to see what stacking two FT50-43 cores would do to your numbers! 😉

Great information thank you very much Manuel. Your help saves a lot of experimental time for others who want to build similar antennas. I have a few questions: When you spoke of FT140-42, you really meant FT140-52 no ? I real question is that right at 7:50 you spoke of an " FT140-52, 2 turns primary, 14 turns secondary, it works ALL over the bands". Did I hear correct ? I am about to try out this, 52 core type, with 2 turns primary and 14 turns secondary, with a larger core , x2 FT240 size toroids. I will really appreciate your advice in this. I will be buying the toroids, probably from Fairrite shop in the Netherlands. I will also buy some FT240-43 and try them out as well. Thanks, Vincent, 9H5LX

Thank you Manuel! This answers a question I have had for 2 years. The actual situation may be worse, nobody knows exactly the efhw antenna impedance. In a real situation, what is the actual height? Is is probably sloping, and what is the counter poise. Does the coax act as a counter poise? Any grounds? Many thanks!

I see lots of design for higher power that use stacks of toroids or toroids in parallel/series, and I wonder how that might change things a bit. Like, would a stack of 3 37-43s behave more like an 82-43? Or just lose even more energy to heat? Given how much more expensive the larger ones are, it might be worth investigating stacks of smaller ones... Plus, the smaller ones are easier to fit in a box, throw in a backpack and not worry about it breaking in the field.

Hey DL2MAN, the turns ratio looks like it reduces to 1:7, is that right? I've never heard of that ratio.

transformers, baluns, and coax are lossy so the best solution is put your output stage in the centre of your dipole as nature intended

Excellent video. Thanks a lot. Can u just comment is on 80m 2:14 ft140-43 better due to turns # or core size? Actually my point is: I'll use 2 : 82-43 for 40 up and looking for 80 or 80-40 73, kuba sp5nzf

I see FT114-42 on line 9..... was this a typo? I can't find FT114-42 anywhere but want to try using that core if I can find one.

So would the conclusion be - feed the antenna where the impedance is as close as possible to 50Ohm to minimize transformation losses (i.e. 1/4 wave vertical or center fed dipole)? Or use as large as possible core. Another question, would it make a difference if we used 2/14 vs 3/21 in the transformer?

Tuners are not so lousy, but some yes.

Love these videos, but one comment grabbed me, 64:1 work better for elevated feedpoints, I can't find anything on the internet about that. I am want an attic antenna, could this work? How high is elevated, and what affect does that have? Thanks, Ed KC8SBV

Hi, Great content. I trying to figure out what is required to change when increasing the wattage from 10, 15 , 20 watt to 100 watt. This video explains it very well. Could you please provide a document with various Transformer core sizes and capacitors types/uf that can be used with different watt? May be a text/pdf document or point me to a resource that I can read. Thank you very much for the time and effort you put in Nitin

Is it possible that such heat is generated because of different winding? (3 primary turns instead 2 turns) What about toroids with different AL value? (AL = inductance per 1000 turns).

This is surprising..... it makes me wonder how you may modify your future builds on this information. Of course, I can't predict what you should do next, because your IQ is maybe three times greater than my own! So we will watch and learn from you as you modify your approach!

Can use FT82-43 on 10W SSB on efhw antenna . Or need to change to FT114-43

7:20 Cells A9 & A10, typo "...-42" (2 places) should be "...-43". Cheers.

Sorry if i missed it in the show... why use 1kv capacitors? Why not 15kv? I am waiting on these capacitors in the mail in hopes it broadens my bandwidth.

You may have already made a video comparing different mixes, ie. 43, 61, 31, 52. But if you haven't that would be interesting. I think 43 material is not the best material for 80 to 10 meter transformers. Either 31 or 61 may be better. Tests may reveal the best material for each band but testing that would be very time consuming.

Your insertion loss figures on the spreadsheet - Did you half them for the spreadshhet or are the results we're seeing the results that the VNA displayed? I feel like you'd know this already, but the result on the VNA (two transformers) would be double what the antenna sees (single transfoirmer).

Try the FT50B-43. "B" version....This is a taller version of the FT50. Can be hard to find. Wind it as an auto transfomer and I get 91% efficiency. I run mine at 30W SSB. I use 2 as a 1:1 balun, or 4 for a bit more power handling that QRP. Try stacking 3 core of FT82-43. Wind 3 turn primary auto transformer. Similar efficiency. Try winding on FT140-43 as auto transformer and close wind the turns. Close winding gives best flux linking and best efficiency. Spreading out turns reduces linkage and loses efficiency. Running a separate primary and winding as bifilar has no advantage. Try the "standard" of FT240-43 as per "internet wisdom" and you will get.... no surprises from your tests here. SWR testing is interesting only, but lossy winding and builds have better SWR. Loss to SWR in short coax is far less than loss to lossy cores. WELL done on actually building and testing, a refreshing change from repeating designs that have considerable loss - and it is fun! enjoy.

You could try some type 61 and type 52 though I don't think they are supplied in as many sizes as 43. Also could you try a fair-rite 2643625002 core, you can get them thru mouser. Thank you for your efforts and sharing the results. 73/72

It’s unfortunate you gave up further testing of the 3:21 on FT140-43 because of high 10m SWR. As most EFHW transformer builders know, 10 m SWR is most affected by the value of the shunt capacitor. In fact, the cap has little to no effect on 15m and lower. I have tried a variable trimmer cap and found the optimum value for 10m SWR to often be quite different than the typical 100pf, which is just a random value to offset leakage inductance. Every different transformer will require a different value of cap. 73,

All these cores and the ones I found for sale are UNMARKED. How you know the type of the material? How can I identify and validate actual core material?

Not surprising as the turn*amps moves the smaller cores into saturation more. Say 50W=I^2*50Ohm so 1A, 50V RMS.

Warum kann ein Dipol nur auf einem Band betrieben werden? Das ist doch die gleiche Antenne wie eine EFHW, nur in der Mitte gespeisst. Und wenn man sich zs6bkw etc ansieht, das sind auch Multibanddipole. Beim Ringkernwickeln kommt es wohl auch darauf an, ob nur der halbe oder der ganze Kern bewickelt ist, mit der gleichen anzahl an Wicklungen und ob z.B 1:49 mit 2, 3 oder 4 Primärwindungen gewickelt wird:==> kzfaq.info/get/bejne/pMBdepeZr6m4g4E.html Afaik ist der 140-43 damals vom Militär für Multiband genommen worden(die interessieren sich nicht so für effizienz), aber eigentlich im 20m Band am besten aufgehoben. Trotz allem, ein interessantes Video. Irgendwie muss man aus vielen Quellen das passende raussuchen und vergleichen (und selber testen)

As you have two transformers in series, the actual loss per transformer is HALF of the loss that you measured; that is, if you measured the power output from the high impedance output of one transformer, it should be half of your measured loss using two transformers! Thus you would be putting more signal to air than your results indicate!

I can see you're very disappointed in losing 25% of your power! (Yes, I also see that this video is a year old, so probably you've gotten over it.) But, this isn't as bad as you think. A loss of 25% is about -1.25dB. And how much is -1.25dB? On the receivers that are listening to you, it is 1/5 of an S unit. So instead of getting an S7 report, you would get an S6.8 report, but they would just round it up to S7 anyway! I know this seems ridiculous, but even loss of HALF of your power only amounts to 1/2 an S unit at the other end. Every antenna system has losses, and everything your signal goes through that doesn't have gain has loss. Have you looked at how much power you lose in your coax? I think you would be shocked. To avoid this particular loss, you could use a center-fed half-wave dipole. But there are good reasons you want to use an end-fed instead. Like you said, everything is a compromise.

I've done a bit of research on the subject on and off over the last year, and it looks to me like the most efficient set-up across the HF bands is with a Fair-Rite 2643251002 core for up to 100 Watts SSB. Being a relatively large core, and with an efficiency around 90% or more they definitely won't heat up with QRP xmitters. They have a smaller internal diameter, and are taller than the typical doughnut shape you are using in this video. That apparently improves the efficiency. They are about $8 US each, but for the efficiency and power handling capacity I think they're worth it. I plan on getting a couple of them to try out and do my own efficiency tests. The only place I found that had stock here in the USA was Mouser. The Mouser part #: 623-2643251002

Thank you for that awesome video! Conclusion? Use resonant antennas? vy 73 de dl8cy