End Fed Half Wave Transformers.......Are they as good as we think?
Вставка
- Опубліковано 18 лис 2024
- In this video I talk though insertion losses of various transformers with different winding.
Here is a link to the matrix
docs.google.co...
link to Evil Lair Electronics YT
/ @evil_lair_electronics
Here are links to the videos showing one of the transformers in use.
• Saltwater DX on a QRP ...
• Portable QRP Ham Radio...
#fair-rite #FT140-43 #FT240-43 #ferrite #torroid #49:1 #64:1
Colin, this is an extremely impressive example of digging deep and organising your research for the viewer. Efficiency is an interesting concept. One question (which you addressed via your colour coding) is how to benchmark this. It is interesting to note that if we compare your figures to a ground-mounted quarter-wave vertical over typical AVERAGE ground, with 32 quarter-wave ground radials (so that would be 160m worth on 14MHz or 320m worth on 7 MHz), the efficiency of the quarter-wave would be somewhere in the order of 72% (-1.7dB). Most of your designs (with a couple of exceptions) beat or at worst basically equal this. The alternative ferrite (same mass but smaller and chunkier than the 240-43) looks a real winner. The efficiency figures you have with that are excellent. Even a 2:1 feed-point SWR and decent coax will add little to the small loss of that toroid. One other thing - your issues with the SWR on the inverted-L, I can only presume your angle of the L is less then 90 degrees (sloping down?), this may well play its part, but a slightly raised SWR is a good price to pay for an efficient transformer. Well done Colin, this is the best analysis of EFHW design I have seen on YT (including the Steve Ellington ones). Keep up the great work. 73.
Really appreciate your comments and insights getting your opinion on things Tim. Really helps create a complete picture and dismiss all the voo doo going about. For 20m upwards the EFHW is a clear winner. I've yet to prove my hypothesis that stacking the FT240 cores kills high band performance but certainly looks that way based on the 2 stack I checked.
@@MM0OPXFieldRadio Groundbreaking work Colin! hugely impressive research and as Tim says taking Steve Ellington's work to the next level. I have a large quantity of 240-43 & 240-52 cores that I'd be more than happy to loan to you to continue this work as I have not had the time to continue my own experiments. I'll drop you an email so we can discuss this. Speak soon, 73, Tom G2NV.
Hi Tom. That would bring amazing. Look forward to your mail 😀.
@@MM0OPXFieldRadio My pleasure Colin, Email sent!
How can I get in touch with you to support your Channel? I'd like to purchase some of your ununs/baluns. I'm looking for as efficient as possible from 80-10m.@MM0OPXFieldRadio
Hi, nice to see someone else actually measuring loss! When you stack 2 FT240-43 and do a close wound 14/2 you will find that to be very efficient also. I've spent hundreds on ferrite and wound hundreds of different transformers looking for ones that are efficient across all of 80 to 10 meters, and every one of them was 43 mix. You can find some that are 52 or 61 mix that are quite efficient, but not across that many bands. Also cap values different than 100pf will measure lower loss across the usable bands in some cases.
Really appreciate you taking the time to comment and share your experiences. Myself and many others have been inspired by your video's. I certainly wouldn't have went down the rabbit hole otherwise.
After I saw Evil Lair I bought some of the chunky cores. They work great. I was reading about the type 61 and thought they might work better for 20-10 meters with the stainless antenna wire and the fishing real. For portable use. But that will require more testing than I have equipment for.
Excellent and very professional analysis. You have saved me a lot of time messing around. Thank you !
Took me ages, but well worth it.
I followed Steve Ellington's advice and stacked 3 FT240-43 cores, Colin. Distressed beyond belief that this might not be the most efficient way of doing it! I've been down all sorts of rabbit holes, some of the folks your chat with online and on air told me "no mate it's not the cores heating up, it's the wire!", so I have replaced the winding on some of my transformers with teflon coated silver wire (luckily picked some up cheap at a hamfest). Maybe that makes some difference at high power, don't know.
Maybe the EFHW while great for QRP and 100W is not the most efficient at 400W (Aus limit) and even less so at 1000W.
You have inspired me to start measuring myself...
You do high quality work, my transformers never look as good as yours... 73, Tim, VK4QP
Experimenting is the name of the game. Unfortunately there is no magic bullet for QRO.
Dedication for sure. Well done and be watching for your experiments on the bigger version.
Thanks. Just comes down to money for cores. Ones that I may even not use. Perhaps I can get to loan some.
Great video, Colin, and interesting results. There's so much folklore and witchcraft surrounding the EFHW transformers, it's nice to see some actual science being done. 👍
I put up a temporary EFHW for 40m at home about a year and a half ago which uses exactly the same core you ended up with, after reading Owen Duffy's page(s!). It's still up and working very well indeed.
Excellent Ronan. The small or big one? What winding did you do and what antenna orientation?
Best Video about cores on UA-cam.
Amazing is what crap they sell commercially.
Never buy one, just diy.
Now I did one for 160 and 80 and a separate one for the rest of the bands.
For me a core of 31 material worked best for top band and 80…
Thanks for your outstanding work
73 de HB3XBL 👍🇨🇭
Dont know about that but thanks. Yes from my limited knowledge the low bands need one of their own, especially for QRO. The EFHW is far from a silver bullet!
Awesome work and great presentation. Thank you!
Wow, this is wonderful work! Thank you so much for doing all these tests with quantifiable results.
Thanks 👍
I found this to be an excellent video - Your honesty in your descriptions is welcome. Well done.
Thank you
Glad you enjoyed it 😊
Wow, best EFHW transformer myth busting video seen so far and I've seen a lot!!. Thanks for posting and also providing the matrix. 73's from VK3
Your welcome. I'm no expert, just doing my due dillegence for my own piece of mind.
I like this guy's like "my talk, my tea."
What a boss
Thanks, as a new ham I found this quite educational. The whole subject of antenna design is quite fascinating. I appreciate your hard work in making this video.
Thank you 😊
Nice data sheet. Will try to setup my new antenna on that.
Give a good start anyway.
Thanks Colin, you put a lot of effort into this.👍🏻
Appreciated. Certainly took a long time to cook.
This is fantastic work!!! You have saved me a lot of time and money!!! I cannot express how excited I am to see this work, and your presentation is very clear and well documented!!!
Thank you very much. Thats why I do this, i i can help 1 person then its worth it.
@@MM0OPXFieldRadio okay, so, I got the big fat toroid in, the 2643251002, and I want to try and use this (with the 14/2 turns close spaced with 100pf cap) but do you think I can use that with your 1/2 wave square antenna experiment? I ordered the fishing reels, and I think I have some good fishing line and I'll have the wire soon, but is that transformer going to work with your variable/adjustable 1/2 wave idea? The cap isn't here yet, but I want to try this combination!!!
Hi Colin. My experience is with monoband EFHW's. Based on radiation field density tests I have settled on L match matching units with a powered iron core. 400 watts key down for 3 minutes with barely disenable core temperature rise! Efficient yes but the downside is not frequency agile. Recently I have looked at broad band transformers. I am not convinced by the twisted pri/sec cross over winding style. I have had good results both with autotransformer and separate sec with pri wound on top in middle winding method. I also have had good results with using ferrite tubes rather than using torroids. When I do back to back transformer testing (+dummy load) I see SWR changes from when single transformer terminated with a resistor. Then again a resistor is only a so-so simulation for a read antenna! The ultimate test of the various matching possibilities is field density measurements but I have not gone that far yet with my testing. Thanks for the insightful vid
Well done for making the measurements, for a science based hobby it is alarming that so few people do this. 73
Colin, as I've said before I'm a big fan of EFHW aerials. Cores can get warm even when running QRP. Your thoroughness is commendable and I confess I have relied on manufacturers information without backing it up with my own experiments. Just to add to the mix, distance from the ground to the UNUN does make a difference, 0.5 to 1.0 meter is as far as I've tried and does effect the SWR. Only experiments with my own 10 meter Spiderbeam but someone has suggested 2 meters from the ground. PAR EFHW recommends being close to the ground! As usual your content is spot on. Wonderful to have you sharing on UA-cam. 73 de M0AZE Mike
As always Mike. I really appreciate the comments. I too have found that the SWR varies wildly dependant on the height of feedpoint and the orientation. It's important to keep the same for consistency as you don't want to be adjusting in the field. When you have it close to the ground you can get ground losses which can also lower the SWR making you think it's more efficient. I wish manufacturers would publish the results but perhaps they don't even know how to check this? Who knows. Hopefully a few more tests to do.
@@MM0OPXFieldRadio Would just like to double on this that hardly anybody on youtube mentions anything about orientation which plays a HUGE roll in this antenna, my preferred is like yourself an Invert V, feed point about 1 meter above ground straight up a 6 or 8 meter pole with the end sloping down ending up about 1 meter above ground (this is no 40m), I once took the same antenna and tried it as a sloper configuration, feed point 1m above ground the other end 6 m above ground the swr went miles off on all bands, ended up having to re-trim an entire new wire for that configuration. I have tried moving the feed point between 80cm and 150cm and the swr does change much for me in the inverted L setup.
I wish the commercial EFHW transformers were advertised with such attenuation/band info!! Tnx for the video / OZ1DTF
To my knowledge only myantennas.com do this.
Great work colin and I’m lookin forward to trying mine out!
Cheers Chris. Keep me posted.
Fantastic work Colin! It's great to see some proper comparisons like this. The 2643625002 looks like a very good choice indeed. 73 de M3KXZ
Cheers Pete. Both very good choices. Will reply to your email over next few days.
Really great information! I’ve been wondering about which way to make it better and this definitely has me thinking differently!
We should always challenge the "norm"
O man thank you so much !!! please make a video with that box which is under testing (which core/ferit, windings, etc)... i just ordered FT240-43 but now i see that different cores could make a big difference. Thanks you so much and keep going.
I think maybe you have seen the video I do this.
Great video, and thanks for publishing the results of all that testing effort. I can feel a Mouser order coming on for yet more EFHW builds! I suppose one day I should test my main 3xFT240-52 transformer, but I really don’t want to buy another three cores to do so!
Exactly, it's a bit of change just to perform the check but is worth it. Perhaps someone can loan me some. I have a good hunch that stacking the cores makes low band performance great at the expense of 20/15 /10. You can see that when I did the 2 stack of the small cores.
@@MM0OPXFieldRadio Yes, your findings tend to support that, and whilst mine was just a copy of Steve Ellington’s build despite us here in Aus having the same 400w limit as the UK, as opposed 1500w, I was hoping to gain efficiency, and only really care about low bands….I have other antennas for 20-10, but 80m is a problem in a small garden so a shortened/loaded EFHW fixes that.
Yeah its quite easy to make a very efficient transformer for low bands. Lows 90s which is as good as your likely to get.
Colin, I salute you for making this video! - The amount of work you must have put in to make this video, to get the very good and presentable results you did, must have been phenomenal. This sort of research is of benefit to all thinking amateurs who are thinking of making and using an end fed half wave. Regards to you Colin, David (M0XQZ).
Really appreciated. It was months and hundreds of hours of research in the shack and out in the field, pretty much tunnel vision and my wife was almost divorcing me as it was all I did. Was a brilliant learning journey to end up with 2 nice transformers.
Outstanding work. This changes the way I will be winding my torrids. I look forward to more testing.
Glad it helped, best of luck 👍
At least someone posted a normal video about 49:1 64:1 transformers - Thank you MM0OPX for such a great video !!! 4L7DC 73
Thanks, straight up shooter me. A spades a spade!
I just finished building this in 56:1 with your recommended toroid, to replace the “permanent installation” Chameleon Mini Hybrid I’m currently using.
Excellent, please comment again once you have it working for a while and can compare.
Very impressive video. Appreciate the hard work
Your welcome.
Thanks Colin for your hard work in doing these tests and producing a great piece of research for us to see. 73's Andy M6APJ.
A very interesting and useful video. 👍There's certainly nothing wrong with optimising efficiency, especially with QRP when you can't just dial in an extra 40....50.....watts. EFHW antennas are certainly a compromise but if it's it's all you can use due to space, portability or whatever then why not make it as efficient as possible. 👍
Thanks Phil, Absolutely. For some reason people think it's good to have a QRP radio only to use a massively compromised antenna just because its compact.
Great presentation very helpful 😊
Your welcome 👍
Great Stuff ! my ft240/43 runs too hot and it’s clear efficiency is the key to cool running. I’ll try 3 turn primary :) .
Definitely, in time I will get round to checking this too.
You hear A LOT of these guys doing these antenna video's say a dB or two is barely enough to worry about, maybe an S point or so.....what they fail to mention is that lowly S point could be another 1000 kilometres away your sigal could be heard. Efficiency is nice, but distance is nicer. Surely the MOST efficient system is best, as it carries your signal the greatest distance. I'll GLADLY take all these lost S-Points and add them to my signal any day of the week......I used to use a linear to add them, but that requires more power and batteries only have so much power in them.
I agree. It can make the difference between making the QSO or not. In Tesco's words - every little helps.
On every 49:1 I built with an FT140-43 core, the swr would start to rise after about half an hour of intermittent SSB operation at 100 watts. So I build mine now with a stack of 3 FT240-43 cores and have no issues with heating or swr rise even up to 500 watts and they test out to be about 97% efficient !
Interesting. Did you measure the efficiency like this? How did you come to the 97% figure? If you tell me the winding style I can check it out too.
Excellent work Colin. I wonder, though, if you did any experiments with the counterpoise length? In my experience, it can greatly influence both performance and the match. I found the .05 wavelength counterpoise to work well, and using a 1:1 choke at that length of coax works just as well as a separate counterpoise. Frankly, it’s better and easier than a wire counterpoise due to problems with the counterpoise coupling with the coax. Also, you may want to do some experiments with 36:1 ratios. I found it to be excellent. Lastly, I have experimented with loosely coupling the primary by winding it with larger gauge wire over the smaller gauge secondary, both with shared ground and with isolated grounding, though the latter would require a wire counterpoise.
I have done some experimenting but didnt fine any real difference with match. I even split the wires so the counterpoise was separated in an attempt to improve the match but it didnt make any difference. What does make a huge difference is the difference between a choke at the feedpoint or 5% Lambda away from it.
Something on the small side for portable 40m-10m that tested low loss for me was 2661102002 which is a 55 gram 61 mix part, 2t/14t close wound with 120pf.
Awesome, I will pick some up with my next order 👍
The reason I love it so much.. even though I think there are better antennas... is because I have a long backyard and I can run as much wire as I want. That gives me a lot of bands and I don't have to run too much coax. I just placed a link to your video on the cores on a blog article and to the spreadsheet as I think is a great comparison. 73 DE YO6DXE. ( ohhh I forgot ... I also have a 140-43 as thats all I had on hand haha )
Great to see you sharing. There needs to be more awareness on end feds 👍
@@MM0OPXFieldRadio I agree on that. So easy to build and tune.
5 years we consulted with an engineer at Ferrite Corp. His recommendation was to use 3 stacked 240-52 mix cores for the highest efficiency. They are indeed more expensive than 43 mix but the Currie temp is way way higher than 43 and the leakage inductance much lower across the 3 - 30 mhz spectrum. I use 2:14 turns ratio. You can add or remove a turn off the secondary for matching whatever configuration your antenna is. Small cores like 143 are much more lossy which is ironic because people want to reduce weight for QRP activity. Also the crossover idea works best for the higher bands and for the low bands it doesn't matter much. For 160m we use 290-43 cores with 3:21 turns ratio, a stack of 3. Very heavy.
👎🙄
Thanks Steve, hopefully over time I can get my hands on some of these and perform the same checks.
This was a great test. The only part that you missed is the increasing inductive reactance as frequency increases. It seems like efficiency goes up with increasing number of turns, but then inductance from the toroid stops the higher frequencies from passing.
I have a feeling if you tested the efficiency on 160 m, it would play out the results with even higher efficiencies.
I was actually doing research for a potential 49 to 1 transformer for 630 m antenna. It looks like I would need a drastically increased number of turns in order to raise the efficiency.
Thanks, I'm no expert in this. I think your right regarding 160m. On later windings I also checked 160m and it was usually the best (excluding 52 material).
Hi Colin, one of my Viewers pointed your Video out to me. I was planning to do something similar. Thanks for the Video. This gives me already some comparison, and also some new ideas. Well done and documented ! On the other side, I´d like to point out that there might be additional losses involved, when core goes into saturation (while under power). So NanoVNA alone with its mW Output might be only half of the truth. Might.... 73 Manuel; DL2MAN
Hi Manuel, thanks for commenting. I dont even have a fraction of the knowledge you have. I agree, I have read this too. Ideally what I've done is the first step followed by measuring it under load. From my limited understanding loss will increase long before Curie point is reached.
Excellent video, Thanks! I'm wondering if bending the EFHW into an inverted L configuration is dropping its feedpoint impedance, similar to dropping a dipoles legs into inverted V configuration does. IF it does then taking a few turns off the 49:1 to drop its ratio might improve your SWR. 73 mike
In short no for me Mike. From my personal experience on an inverted V the SWR is better again as a 15/2. Different ground conditions will effect this though. Give jt a try though.
Oh goody, another non standard toroid to play with. I have had good results with the 2643625002 for 40-10 EFHW up to 20 watts ssb.
Great video.
Nice video Colin, you always come up with something that gets me watching with interest, well done.
I myself try to use resonant ground mounted verticals, simply rig, coax, choke, ant and radials, hoping this will be efficient enough to make dx contacts zl to eu. I did not want a ant which needed a transformer , simply thinking I would be creating loss. Of course ant near objects of any kind can alter things, propagation is another factor, but I did experiment with a 1/2 wave on 40m with a transformer against a vertical. On the air results between the two with dx stations, some said no difference, other picked one over the other and vice versa, depending on their ant orientation .
Overall I got better results on a resonant ant, I only mainly use 40m or 20m, and did away with the worry of having the transformer connected.
Horses for courses I think.
I was always told the less items between the radio and the ant the better. But on the air, as I say with qsb, etc etc , lots of time getting the efficiency
Results might not mean much, in my mind anyway.
But lots would probably disagree, what do I know, iam just a guy trying to enjoy the hobby and have a natter hi.
Many thanks Colin, still waiting to get you in the log.
73 .
Appreciate all the comments Brian. Its ones like these that keep me making the videos. I too prefer the simplicity, one reason I have held off for so long. For as complicated as adjustiwave looks its really simple with no bits in between coax and antenna and performs brilliant.
Nice, but a word of caution with regards to a statistical of 1. All measurements have errors. Are your measurements really accurate to decimals? Likely not. Try wrapping two identically and check the variation. I would expect wrapping tightness would have a large effect. Pedantic comments aside, good work.
I'm a Six Sigma green belt also do know a bit about statistics. Im proficient with Minitab too. Yes I do agree with you, there will he measurment error but how much really? If i had unlimited time and resource I would prove gage R&R and capability.
It sounds like you need to add another column, beyond insertion loss - actual transformation ratio. Every antenna is going to have a different impedance based on height, length and shape, but I guarantee each core type is going to have a different transformation ratio in reality. And that's before the quality of the windings themselves come into play, and the amount of power impacting real permeability...
Correct. The transformation ratio is specified at the top of each collumn. Just because its efficient doesn't mean it will "antenna". This is the reason I had to go with a 56:1 2:15 winding on the larger chunky core.
You can stack a ft-140 inside an ft-240 which makes it beefier without the extra dimensions (it's still as big as a ft-240) Because it's the same material I expect more efficiency than a single ft-240
I've had this thought too about trying something similar. I think your probably on to something.
Texting to english would be great as even without understanding you language, i got so much good info from here 😁 thank you for the video!
Thanks, whats your native tongue?
@@MM0OPXFieldRadio haha, sorry, this was ment for the previous video about wires, that I think was mainly in Urdu with some English words and English schematics in it (the parts I understood) so no worries my friend, you I understood perfectly 😁😁😁 although my native language is Norwegian
@@MM0OPXFieldRadio when that is said, your videos are also great and helped me out so much!! I'm going to place a order for that thoroid so o can try to make my own EFHW 😁😁
Colin, Thanks for a great video. Liked your methodology. I have to figure out how you measured your losses by checking out more of your videos to look for a clue. This is the first of your videos, maybe the second or third??? that I've seen. Wondered if you have ever wound one of these transformers using Litz wire and if there would be any advantages or disadvantages of using Litz? Now I have to find out if you sell the qrp unun... 73... Rich N3JLR
Theres a video on from this that shows how you measure losses. Haven't tried litz wire, not sure what the advantages would be. I think it may be harder to work with.
Thanks for the doc and info.
gold-dust... many thanks for your hard work here.
Thanks 👍
Hi Colin - I’ve seen articles that mention performance limitations at high power levels due to magnetic saturation of the core and I’m curious to know if you have run into core saturation in any of your testing. - Thanks - Jim
I have Jim. Theres a video I made not long after this. I test an FT140-43, 240-43 and the 1002 core. Interesting results.
Changing the cap will allow you to reduce vswr with the higher turn transformers. You simply need to tune out the Inductive reactance with some capacitance. Would be fine. Also should emphasize swr doesn’t much matter. Efficiency does !
Agreed. I have the larger one ready to test again with a different cap value. As longs SWR is less than 2:1 I will be happy.
Fantastic work, great video Colin 👍
Thanks Ape 👍, no massive differences in the grand scheme of things but it's good to know.
Nice video, you should try a 36:1 end fed. I’ve had really good luck with mine. “OCF Masters” have some really good articles explaining why it might be better.
Excellent work! I have been testing UnUns by measuring the core temperature while connected to the Antenna with 100 watts on CW after 1 minute. A tiny 100k thermistor is epoxied to the middle core of a three core 140-43 2:13 wind. I see a 3 deg C temperature rise on 20,15 and 10M but 7 degrees on 80 and 40M. My Antennae is a 130 foot EFHW with SWR < 1.1 on all the harmonic bands. The thick wall 43 core is my next attempt to reduce losses further.I will publish some of my findings on my QRZ page shortly.
VE3XKN
Likewise. I wonder if the back to back testing and your tests correlate?
I have not been doing back to back tests yet. I find variation in core permeability and slight winding variation to be a potential issue with this test though I intend to use it in the future. I test my winds with a variable capacitor across the primary winding and load resistor on the output, tuning for minimum SWR value on the VNA. I swap the variable cap for a silver mica of the same value when I test with an Antennae. The capacitor can make a large difference in the SWR as measured with a resistor load usually from 14 mHz on up. Some of my winds are optimum with 75 pf others may need up to 220pf. Positioning the primary winding with respect to the secondary can also make a large change in SWR. I have wound ~ 50 Xformers and still have much to glean from different winding configurations. A recent test with the a 3 Core stack of 140 - #43, 3:17, 100 Watts CW for 1 minute produced a temperature rise of only 3.4C on 80M and lower temperatures on the higher bands up to 10M using my 130 foot long EFHW. When I started this work I had temperature rises of 20C or more with some Xformers. Still much to try.
The capacitor is very important if needed to cover 5 or 6 bands, better than 5kv is great. I used the 30kv 100pf cap for FT240-43 whatever stack 2 or 3.
Thank You very much from Germany,73!
Great video! Probably a stupid question… Is it harmful if the coils touch each other on the toroidal core?
Not silly at all. It wont be harmful at all. This is enamelled wire and so insulated.
I'm wondering, maybe try half close wound and cross to the other half close to the other side of the primary, it seems the core is most efficient when the primary and secondary are close together, so the bit of the core to the other end of the primary is being wasted.
The classic crossover is good for choke baluns / common mode rejection as it keeps the input and output spaced to limit capacitance between them
It's worth a try. The standard crossover on the standard cores gives a nice SWR across multiple bands and thats why people like it.
MM0OPX, Please make a video of how you do the efficiency testing! Awesome video! What gauge wire did you use on the Fair-Rite 2643251002 15:2 (56:1) UnUn? WD0DUD
Here you go. I used 1.6mm wire ua-cam.com/video/TZjY2Jim7fY/v-deo.html
@@MM0OPXFieldRadioColin, apologies for more questions. Have you concluded your design and testing of the UnUn? Did you end up using 56:1 or 64:1? I'd like to know if you wound it tight or loose and how it was wound. I want to build one! A picture would be worth 1000 words; a video, priceless!. Best Regards. WD0DUD
Hereby you go for now Gary. I will make a video on this soon once I have done a but more experimenting. twitter.com/MM0OPX/status/1563531288656887809?t=k5lUNCAaCRz2Arp1QCLFbA&s=19
Great to see more of these experiments and measurements. I'm considering an EFHW on 80 and 40 only as there are better antennas for the higher bands. Using a too-long wire for higher bands is a compromise in many ways. So optimizing the transformer for 80/40 makes sense. Note that loss in the transformer can help with SWR, but I'd rather have low losses and deal with the SWR some other way. 73 de w6akb
100% agree. On my 100w version 40m is 1.8 at the dip and less than 2:1 across the band. I'm more than happy as the efficiency more than makes up for it.
If i'm reading the chart correctly the Fair-Rite - 2643251002 would be the best choice for 100 watts on the 10 meter band?
Yes, it's very low loss across all of HF and can handle 100w FT8/FT4 no problem.
Pulling the windings close together increases core saturation and the possibility oF RF Flash over between windings.
From my testing the core always reaches currie point before the wire starts arcing. What important is the insersion loss. Pulling the windings close together helps on the low bands but the upper HF bands suffer. Poor efficiency will the saturate the core quicker as more power is lost as heat. Have you seen the KN5L method? Its recognised as likely being the most efficient broadbanded transformer. www.kn5l.net/kn5lEfhwUnun/
Very interesting video Colin. Well done.
You have probably tested more cores than most of us will see in a lifetime!
Having tested so many variations, have you come up with a few 'one-liners' such as:
"Higher permeability cores and high number of turns = lower efficiency above 20m?" or
"2 turn primary is better for 20m and higher"?
One or two key 'takeaways' are good for folks to commit to memory
Good to hear this investigation will continue 🙂
73
Bruce G4ABX
Thanks Bruce. I'm definitely not the one for the technical terms with this kinda thing, very much discovery from experience trying explain without knowing the real name for it.
Hi Colin,
I’d be really interested how the double stacked 240-43 49:1 compares with the rest of your spreadsheet. Being following your experiments closely and it’s good someone has done an analytical approach to core losses. Fantastic work.
I've still to do a double stack test. Watch this space.
Very interesting and detailed work, thank you. (But measurements to fractions of a dB are meaningless.) Looking forward your definitive conclusion of core types and turns radios for EFHW antenna matching transformers for each band up to say 50 watts. I hope then we can all put the subject to bed
Agreed on the fractions. Not noticeable and there's measurement error too. The stacking of the 2 small cores really did kill upper HF. No one wants 16db of loss.
Excellent information Colin! Many thanks for your assembly and documentation time! Would it be possible to post the wire gauges and type for the 56:1 using 2643251002 and 64:1 2643625002 assemblies? Also, what output power do you think the 56:1 can handle SSB?
73 de N5OJX
I'm using 1.18mm and 0.63mm respectively. SSB rating is 250w.
Excellent work. What your work indicates to me is that the higher the field strength of the magnetic field at at the turns of the output, and the greater the number of turns in the output the greater the efficiency of the transformer. But the greater the number of turns will introduce resistance and therefore heating in the wires, stronger magnetic field will also introduce greater heating in the ferrite. I was wondering if SWR issues in the high efficiency transformers could be resolved with increasing the capacitance? Thank you MM0OPX
Very nice video. I also enjoyed Evil Lair's experiments. What wire did you wind it with 14 or 18 AWG? Thank you for your video!
Nether, it was 16ga. For these NanoVNA tests i find wire size is irrelevant.
@@MM0OPXFieldRadio Interesting... I came here to ask about wire size as well.
Very good work! Excellent.. Thank you , K5NR
Thanks 👍
VERY NICE !!! Can you put measured swr (or atleast R +jX ) on the spreadsheet when tested? Is it possible you can place stubs hanging off the longwire to tune out the reactance on each band whilst using the higher effeciency bead (toroid) ?
No, all of these dont exist now, just the 2 I use. You need to check SWR as an antenna with an antenna wire and I just dont have time. Its pointless checking the SWR is its a lossy transformer. Only check the best and work back.
@@MM0OPXFieldRadio AH !!! Thank You ! Good point !
@@MM0OPXFieldRadio You headed off my next question: A metal film or carbon resister load wouldn't be a good cheat ?
Yes but I've found even a non inductive resistor varys wildly compared to an actual wire fitted.
@@MM0OPXFieldRadio copy that !! Thank You very much ! Just great to watch the video and the work involved.
(for something completely different...) On a side note, I built a 10M Moxon last year for field day thanks to your video.
At the time (time crunched) I couldn't source good spreaders for 20M. I'm looking at bamboo this time around. Any advice ?
I am failing to understand the loss situation. I have seen these efhw w/49-1's set up and the actual rf radiation output measured and then compared to the exact piece of wire configured as a dipole and the amount of output radiation was the same...Can you explain what kind of loss we are talking about here. Your work in interesting. I am questioning it it actually translates to power output at the antenna ??
Yes your spot on. This is comparing is how much of your TX power actually reaches the antenna wire. SWR is not a measure of antenna efficiency which some companies like to quote.
Doug Doty yes exactly! Configure a half wave wire as a centre fed dipole and measure field strength at a given distance for a given power on the design frequency. Then re-configure the same piece of wire as an end fed and you'll find the field strength will be the same as the dipole. I bet you'd be very hard pressed to see the effect of the loss in the Txfr using this method. Regarding loss I have used an FT240-43 49:1 and it was lossy as evidenced by it getting quite warm at 100 watts SSB. I changed it for 2 x FT240-43 and it ran cool as a cucumber. I think much of this stuff is just academic especially if better transformer efficiency results in higher SWR. Any advantage is then quite likely cancelled out by increased coax and antenna tuner loss. de VK2DJP
What was the number for the first one on bottom list all green please if u could please as I’m in need for one lol great video very informative thank you
Its a 2643251002 👍
Excellent work and research!
Why don't you make a switch and multiple cores for the bands, that way you can have the setup with the lowest SWR and best efficiency for each band. How much does it cost in the long run? If its not a portable transformer set, does the size and weight really matter?
I'm looking to make an end fed half wave kite antenna. Sourcing transformer design and antenna tuner design. Interesting measurements. But I'm seeing different gauges of wire used. Will that have an influence? It could be the case. I don't know if I can be conclusive, but some toroid cores are more efficient, it must be their shape and or composition.
Greetings,
Jeff
Yes I could do that. You could even remote switch it.
Useful information, but how was the percent efficiency determined and with such high precision? Theoretically? Deducting heat dissipated by the cores? Field strength measurements from the half wave antenna? Research by others? Some kind of modelling program provided by the manufacturers? I'm at a loss here.
Here you go ua-cam.com/video/TZjY2Jim7fY/v-deo.html
...as pointed out by others, this is a bit of a black art.
I use a single FT 240 43 from Amidon. crossover wound 49:1 with #16 wire, matching a straight 130' EFHW sloper, one end at 65' the feed at about 25-30' (no bends or turns), fed with LMR 400, the coax shield is the counterpoise, (little to no SWR return loss) ; on key down or FM, 100 watts, using a tuner with high sensitivity to any SWR change (don't need the tuner on any band except 10, it's there to measure SWR only), there is not so much of a smidgen of change even after a couple of minutes of key down... except a bit on 10 which is expected. With the 20-30% losses as published, that core/wire should heat up like crazy, the match and SWR would surely change.. but it doesn't...Where does the 20 - 30 watts loss go?
Interesting. How do you measure the temperature during use? So you have some sort of remote temperature sensor? It would be great if you could share your temperature curves vs tx time?
@@MM0OPXFieldRadio Unless you have a lab full of equipment and considering the antenna feed point is 35' in the air, there is no way, but... you can extrapolate from the obvious; if there is VSWR creep on different bands, but not others, ANY change in VSWR = heat and heat = inefficiency.
When you see VSWR creep, and most rigs are very sensitive to VSWR change; so what you are seeing is the core and windings start to get warm. If you see NO VSWR change CW key down for 10 or 15 seconds, watch it close on each band you want to check, it happens pretty quick; but if there is no change then you can consider the UNUN is transferring TX power efficiently to the EFHW. Remember, no heat = efficient transfer of RF energy. Remember too, heat causes thermal expansion of the core winding wires which in turn causes the physical change through expansion.
The first thing you look at is the core mix... there is a rule of thumb, 31 mix is best for 1.8 and maybe 3.5, the 43 mix is best for 3.5 to 21mhz, it'll work OK but efficiency starts to fall off as you go higher up, and the 52 mix cores are best for 7mhz through 28mhz... even up to 6 meters. But these core mixes cannot be perfect for all bands.
Winding the cores fairly close spaced seemed to work better for me... it's a toss up whether the crossover or a continuous wind is better.. it depends.
My next project is to stack a 43 core and a 52core together to try to get the best of both worlds...
@@mortimersnerd801 Have you tried stacking an FT-240-52 with an FT-240-43 yet? Intetesting to know the results! 73
@@mikesmith5139 ...no, and common sense says it would largely take on the characteristics of the core with the highes mu; think about it ...the 800 mu of the 43 material... that's what would be dominant. Wanting to approximate the crossectional mass of a cable core, similar to the Fair Rite 2643251002, but of 52 material, I took a FT 240 52 core and carefully cut it in half with an angle grinder....it's brittle so you score 1 side first, that way it breaks clean when you cut the other. You glue the 2 halves together. Now you have the rough equivilant of a cable core.. it's almost exactly the same, less than 1mm difference....wound it auto-transforemer style 15:2 (you seem to get a better match) with #20 enameled wire. Into a perfect 1:1.04 18M EFHW on my NanoVna... 1 min ...100 watts key down... it doesn't even get warm, maybe a hint, so I know there is an efficient transfer of power from that 50 ohm coax to the 2400-3200 + ohm 1/2 wave piece of wire ....that's good enuff for my pay grade kiddies...oh and yes it works great... Richard VA7AA/XE1
Thank You for the information! I would like to know what stacking 2 240-43. That is what I was thinking of doing. Could stack 3, 240-43 and wind with 14 gage wire. I don't know what that is in your wire sizes. But it is bigger than most people wind them. Thinking of 21:3 winding.I don't have a way to test how well it will work. 73 de W4DES.
I will test a 2 stack also. For this type of testing I've found wire size irrelevant. I will do a follow up video on the FT240 size core. Stay tuned.
John G4MRK
Thanks for the excellent video.Colin. What was the source of the 2643625002 ferrites?
Either Digikey or Mouser.
How's the swr when using an end fed antenna? Tell me, how do you tuna boat?
Real life SWR depends on winding configuration. I've found a happy medium between efficiency and a good SWR.
I learned a lot, thanks. Subscribed.
Thanks David 👍
Hiya, I am looking to wind a 9:1 to make a version of Coastal Wave's "non-resonant" vertical off-centre-fed-dipole for 20w portable use. Do you think the efficiency of a core will be similar as your findings, or is every test scenario completely different (this is what it looks like)?
I know what you mean. No, I've not tried any of these. Yes same method would work. Will I do this and time soon? Probably not, just because 9:1's dont interest me (for now anyway). I said that too about the EFHW and look what happened there.
@@MM0OPXFieldRadio I will try winding one, but I don't have the facilities to check efficiency, so I will just try it and look at swr. I will post what I can.
Hi Colin, do you sell the ununs and how much do the small ununs cost. MM0JZP.
Hi William, I dont. I had a couple extra at the time when making my own.
Great video!
Thanks 👍
Can't help but wonder what a 3:24 build on the 2643625002 would give for efficiency?
What you will find is the low bands will become more efficient >90% but at the detriment of the higher bands. For me the point of these transformers is to have broadband performance. If all your interested in is 160/80/40 then it's worth doing this.
This is a great video, I have been playing with EFHW for a year now using 43s material but mainly as a linked efhw on 17m,20m,40m. Looking at the result sheet I would presume if you stacked 2 or 3 FT140-43s the efficiency would increase? Cheers M0KCB
Thanks. I've only checked stacking one with 2 of the wee QRP cores. Efficiency increased on 80 and 40 but it fell off a cliff after that, serious loss. I suspect you may see some drop off on the higher bands bit I won't know till I check it.
Hi Colin, what a very interesting video you made. My compliments! But, what I don't understand is that you're telling that you ended up with the 64:1 16/2 qrp version, while the 49:1 21/3 on the same core has much less losses. Can you explain that please? And can you also tell how long the endfed radiatorwire is that you normally use? I expect 10,1 meter, a coil of 34 uH and another 2 meters. Thanks again and I will continue to follow you!
It its interesting. Just because you have a more efficient core it does not mean it will "Antenna". I used this configuration and the SWR was over 2:1 and I could not lower this no matter what I did. The 64:1 presents a much better match typically less than 1.5:1 on all bands. For 40m up its close to 20.5m of wire used.
@@MM0OPXFieldRadio Thanks for the explanation. And what fast! :-)
I have made de the big 56:1 with cross over. The swr measurement come on 1:2.8 at 28.500MHz with 2700 Ohm. What goes Wrong?
Nothing is wrong. Connect up an antenna wire and let me know how it goes. Thats a normal reading with a resistor.
It did go wel.
What kind of wire do you use for the qrp toroid? What is the max power handeling on ssb for the qrp toroid?
@PD5RM_Robin I used 0.66mm wire. 20w SSB and 10w CW.
Very interesting video and impressive amount of work Colin!! My humble comment on SWR is that you use a Inverted L configuration which in my opinion may have a different impedance to a normally level endfed LW. Do you have any experience on this? Regards LA3YP Alf
Yes your right. EFHW'S are very fickle and a different orientation gives a different SWR. I needed it to good as an inverted L as that gives best low angle radiation.
Excement work, the right way to test transformers ,,,,,,, i was wondering if you also measured / checked the transformation ratio 49:1 ? Thanks OZ1CGQ
Yes, tried many 49:1. Check out the spreadsheet.
But did you measure the impedance transformation ?
By trying a non inductive resistor yes. Its not that important IMO as it WILL be different when you have an antenna wire connected.
You are the hero we need.
Wouldn't go that far, just hope it can help 1 or 2 people.
@@MM0OPXFieldRadio Humble heros are the best!
Love it when hams experiment. Very good job Colin. I’m not sure how you did the test exactly. Wouldn’t you want lower resonance on the radiator as well? So if the core is 90% efficient but has double the reflection, are we at square one agn with the loss on the coax? Fb brother. Look forward to a follow up.
I will make the video of the test. You measure the pair then half the loss. Keep an eye out.
Hi. Very nice videos. Thank you for sharing. I have searched information about how the wire diameter affect the efficiency of the unun but have not found any information. Have you done some measurements using different wire diameter with the same toroids, ratio and spacing between the turns? Is the largest possible wire diameter the best or is the optimum diameter for low loss somewhere else? 73. SA6SGH
I have done this but not documented it. I've found no difference between wire diameters for testing purposes.
Type 31 mate . is best , specially on 160, 80, 40 , 20 meters .
Core geometry and winding style is also critical.
When you say 16 and 2 turns, is that 16 turns total including the 2, or 18 total. Thanks for all the great info K7WHN
Its 16 total so 2/14. Sorry for any confusion.
@@MM0OPXFieldRadio Thanks
Let me guess, you wind two identical ununs, connect them back to back, i.e. secondary to secondary, antenna side to antenna side, ground to ground, TX at one end, dummy load at the other end, then connect with two power meters one on either side, measure the loss and then halve it because there are two ununs?
Almost, I'm using a NanoVNA. Measuring loss back through S21 then halving it.
Where can I get that set of tables!?!?
Doh' should have read the expanded description. Thanks!
Hi, master Colin. What's the number (AWG) of wire do You use on transformer (20:13 min)? Thanks so much. 73 - PU2AOO
This is 1.18mm diameter wire.
@@MM0OPXFieldRadio As always I'm listen good ideas and shortcuts with your proficient knowledge. Greetings from Brazil. 73 PU2AOO