Link to my spreadsheet of results docs.google.co... Best 100w EFHW • Best Ferrite Core For ... How to test EFHW transformer efficiency • How to measure Insersi... Many thanks to Tom G2NV for thr loan of the cores.
Hi there. Auto transformer interesting. I published a couple of videos some while back, using that method and went a step further by using tinned copper wire which made soldering even easier. Great Channel. Keep up the work. 73 Peter.
I like the Auto TF style of winding because it winds on easier than a 1.2 mm wire twisted and trying to wind it on a ferrite core. Thanks for your efforts.
I think they are mistaken when they say they are close to curie temp when seeing an increase in SWR and are actually nowhere near that hot. If you look at the material specs, 43 permeability is 650@0C and doubles to 1300@75C which is certainly going to drastically change transformer behavior. 52 is a little better with 200@0C and 400@100C but again that's not an unreachable temperature so a drastic change is still going to be seen during normal operation. 61 is much flatter, 125@0C and only gets to 150@325C, but you're not going to cover all of 80 to 10 with a 61 mix transformer.
Just getting back into the hobby after being away 20 years! So much had changed! Bought a few of the 1002 cores from Digikey, some stainless wire and all the bits to make an efhw and others. So much stuff coming in the post and can't wait for it all to get here! Just along the road in Bathgate.
Hi Colin, I just tested triple stacked 240-52 as an autotransformer and this is giving me 89, 91, 90, 88 and 82 percent efficiency on 80,40,20,15- and 10-meter bands. Not as high as your measurements. However, I’m using 1.4mm enamelled copper wire. The wire covers about a third of the outside circumference of the cores. This is so the inner circumference does not have any overlapping wire. So, I might rewind them using 1mm wire and see if there is any difference, as in better efficiency. But for now, a slight difference. I guess any heat will at lease be spread about a little.
Thanks for letting me know Chris. Your numbers are not that far away and there will be measurement error on mine and yours. It wont matter in real use I dont think. Look forward to seeing you getting it up as an antenna and seeing how it goes.
This looks excellent Colin. I would like to suggest PTFE sleeve over the full length of windings to provide insulation for QRO power levels. Merry Christmas OM.
Thanks Colin. Will sure use the autotransformer and 2 stack 240-43 and do the tests when I get the cores. Have a great Christmas and a Happy New Year. 73 de SM3TEK
Got my hands on 4 240-43 from Electrokit in Malmoe and the last 4 pice xxx1002 43 coars from Connectiv. Gone a have a great start on 2024 with winding cores and testing EFHW. Hopefully get some good ressults. Merry Christmas
Superb Colin! This design looks fantastically promising. I'll be knocking up a couple in the coming days. one close wound and one wide spaced, If it replicates the efficiency on 15 & 10 in real world tests we'll all owe you a huge debt of gratitude! To have a 49:1 transformer with such equal level of efficiency across 80-10m will be a real gamechanger for the guy with room for only one antenna. hang on to the 240's for as long as you need em, your your blazing a trail for everyone! 73 & a big big thank-you!! Tom G2NV.
Couldn't have done it Tom without the loan of the cores from you. I knew I had unfinished business. I will drop you a mail re getting them back to you but I would like to see if it will actually "antenna" first by fitting a trio of the 52's into an enclosure and seeing how it goes. Thanks again mate.
@@MM0OPXFieldRadio Spot on, No rush at all mate! test away and we'll compare notes in the new year. I'm excited to try out the AT design and particularly some A/B testing on 10m against my T2LT type 10m monobander. If you've cracked the efficiency issues on ten you'll have done a lot of people a huge service! 73.
Colin, first let me say thank you! Your work is amazing! Thank you! I just had a perverse thought while rewatching every video I could find between you, Evil Lair Electronics, and Survival Comms on core testing. What would be the effect of cocentrically nesting a ferrite inside a wound transformer? That is placing a ferrite rod or smaller ferrite bead inside a wound transformer. I've never seen or heard of it being done. They fill university libraries with what I don't know on the subject, but its my understanding that the ferrite has two functions. One is to increase inductance per turn, and to constrain the elrctromagnetic field. I am not even sure how the righthand and lefthand rule applies to transformers wound on ferrite. I can see such an experiment either being amazingly good or amazingly bad. You already have several wound transformers and ferrites, so if its not too much trouble perhaps you would be gracious enough to empirically answer the question.
Thats an interesting thought. I will keep in mind going forward. Actually I think I will struggle as a dont have as many cores as you may think. I will see what I can rustle.
@@MM0OPXFieldRadioIt has been thought of, sort of. Fair-rite sell ferrite blocks that have little tiny holes drilled into them that you pass the wire through. Think of it as being like peg board but thicker with tiny holes. It another type of EMI suppression core.
Well done sir. I would be very interested to see the measured losses on the "Traditional" version of the transformer (14-2 with crossover and 100 pF cap). I have followed your testing and even built an Evil Lair transformer using the type 61 Mix. Still field testing the various configurations but I have used the "Traditional" 49:1 and it works well for my base station on 40, 20 and 15. Not so well for 10. I will be building a new auto-transformer using your winding recommendation's on stacked 240-52 cores to see if I can improved 10M,
@@MM0OPXFieldRadio Thanks Colin, I missed that. Did you do any analysis of wire gauges and their effect on efficiency? Merry Christmas to you and yours.
Interesting work, I assume that much of the measured loss is stray reactance, although imperfect transformer behaviour it does not automatically lead to core heating. I have been using similar cores for common mode chokes, not transformers, even with something as apparently simple as the chokes it certainly pays to make the measurements! 73
No worries! the second one off the bench has your name on it! I've got to say, Colin's work on this is epic! This design Auto-transformer looks like a real game changer for 80-10m. I think the whole community will owe him a massive thanks for this! I'll be heading over to G6AD-HQ shortly with your Xmas prressie shortly. 73 Tom.
Hello Colin, II really appreciate your videos and I'm a fan of EFHW. do you think that in the future it will be possible add the results obtained with 2 cores? 73 de iw2hus, Franco
52 material has much less Permeability value than 43. Im not sure how this influence on EFHW, but for common mode chokes(dipole), you need more turns windings.
Hi Colin, I'm a bit confused on a few points that span more topics than just this video, but I hope it's ok if I post here. 1. There's more than just the unun efficiency, as when a really efficient unun is tested with an antenna the SWR can be terrible. Is there a way to test the entire system to see how much power is actually making it out of the antenna without relying on WSPR (which doesn't let you compare across multiple days since propagation changes)? How scientific can we get with these measurements? (I'm not talking about modelling the gain or directionally etc of the antenna, just the actual percentage of watts coming out of the radio that actually make it out the antenna). 2. I'm confused as to which antenna setup is the most efficient (EFHW, EFRW, dipole, etc). With a random wire you have a tuner which some people say turns a lot of the energy to heat, but others say that's only if you have a bunch of coax between the tuner and antenna, whereas if you just have a random wire coming straight out of the tuner the reflected energy is just re-reflected back into the antenna again, so eventually makes its way out as signal. If so, could a random wire and tuner be more efficient in that setup? A dipole needs a balun, an EFHW needs and unun and these all absorb some energy. I've also seen an end-fed quarter wave with no unun, but rather using a big counterpoise, but then they say the resistance is only around 30Ω, so wouldn't that mismatch be inefficient? Basically, I'm wanting to do QRP portable, and just want the most of my meager wattage making its way from the radio out the antenna and I am trying to figure out the most efficient overall HF antenna setup that factors in all the losses from all the required bits working together as a complete system, not just independent components that may be efficient on their own but don't play nice together. Or are all these setups basically equal in reality (ie the differences are barely noticeable and it's not worth nitpicking a percentage point or two)? Thanks!
I always mention in these videos that transformer efficiency is only one element, but an important one. Honestly don't fret too much, I don't. Follow good antenna practice and remember Height is might with the ground being able to suck away efficiency.
@MM0OPX I know every test takes significant effort; however, what would happen if you mixed cores? i.e. 2x43 and 1x52 or complete mix 1x31,1x43 and 1x52
I've always made my EFHW transformers using autotransformer windings and wondered why others were essentially putting the ground at the tap. To me what we are trying to accomplish is more like a delta-match or gamma-match setup. You want to put the 50 ohm feedline at the right tap point to match the impedance. I also have recently wondered why I've always used #43 material when the handbooks say that it's great for HF common mode chokes but not good for HF transformers. And since #67 material is good for HF transformers and poor for HF common mode chokes, perhaps it would be a better core material in an EFHW transformer. ? It would take more windings to get the same winding impedance but perhaps there is enough room on the core to accomplish that. Perhaps the #43 in our configuration is a "sweet spot" in being able to both transform and choke. I should mention that the curie temperature is higher for #67 too so it might be more resilient when the situation is not optimal. Thanks for your great videos and I'm looking forward to the investigation for your further optimizations. All really good stuff!
You said you aren't really interested in this antenna for the 10M band because you have better antennas for 10M. Since you clearly know your stuff, what antenna do you use for the 10M band?
Its not terrible on 10m. 10m is one of those bands you can use a small beam like a Moxon or Hexbeam without much space. If you want to go the vertical route the 5/8 is good. Height is might remember.
The whole twisted bifilar winding making no sense if you are just going to use a common ground anyway. I've been using a AT EFHW for a while now in my POTA antenna and haven't had any issues. I've compared a FT140-43 using 2/14 cross over and no crossover, overall found little difference between AT and non-AT. AT no-crossover ~100pf 2/14 ended up the best for me. I found that inverted V ~20m wire, 12ft choked coax gave me great 40/20/15 and good 10m on 2/14, 2/15 and 2/16 tuned low on 40m (around 6.8) and slightly high on 20m (14.4) but good on 15 and 10. I assume this was more of the impedance shifting on 56:1 and 64:1 to being happier closer to the ground vs higher bands. I haven't tried using 2/12 and 2/13, might be worth it to see the effects.
Isn’t the traditional way of wiring a transformer (with the twisted wires) also an auto-transformer? The wiring looks the same, just without twisting the wires and wrapping the primary on the other side instead of parallel with the secondary.
I just tested single core 240-43 with 14/2 auto transformer, closed spaced winding. Here's what I got: 80m 76.21% 40m 80.35% 20m 84.24% 15m 84.92% 10m 84.24%
Efficiency on my stack of 3x 240-52 measures better with larger gauge wire. I started with 18awg, but now I use 12awg for primary and 14awg for secondary because I am a wimp, autotransformer, tight windings. With 2 primaries and 12 more secondaries with 120pf shunt I get 96% efficiency on 80m, nearly that on 40m, and it drops to 89% in the 10m range. Here are my insertion losses (back to back method nearly identical to resister method): 3.9Mhz -0.19db 7.3Mhz -.21db 14.3Mhz -.42db It stayed pretty much the same until the end of 12m band 25Mh -.48db 30Mhz -1.33db To get it 160m-20m, it needed 3 primaries and 19 more secondaries, and 180pf. Similar efficiency measured. Half hour of 800w FT8 couldn’t melt 1.5” of snow on the open design. I’m happy to share pics of design, and nanovna screens. KK7JMH
Thats brilliant, the extra 20pf in shunt will make a difference too. I've been using 100pf. A pic of the install would be great as i need inspiration to build my one. How was SWR dips on all bands. Did they tie in nicely?