Other SOTA sites: SOTAwatch | SOTA Home | Database | Video | Photos | Shop | Mapping | FAQs | Facebook | Contact SOTA

EFHW Transformer Toroid T-6 (Yellow) or T-2 (Red) or FT-43



A loss of 1.5db has no noticeable effect on the efficiency of the antenna. Most qrp antenna tuners suffer similar losses, especially at random long wires.

It should be indifferent in most cases, whether you heard with S5 or S4 3/4.

The decisive advantage for me, especially for Sota, is that I’m qrv without tuning and without changing over on several bands with a feeding point near by the ground.

73 Chris


I just made a test with a HAM about 5 km from my home.

On 40 m the endfed with the 43 material transformer was 1 S-unit less at the receiving station than with the material 61 transformer! On 20m there was no difference in signal strengh although the power measurement results yesterday showed a higher loss on 20 m.
Endfed with 61 material transformer showed the same S-meter reading compared to my linked dipol.
The performance of my transformer construction is ok from 7 MHz to 14 MHz. Currently I only have FT 114-61 available but for 10 W it should be ok. I am using 3:27 turns and 330 pf…

My endfed is a trap-version for 7/10/14 MHz, halfwave length on each band, no harmonic operation.

Results for me: for the moment I will carry my 61 material modified endfed along for my upcoming activations.

Tanks for the links, there is a lot to read and still a lot to learn about.

And thanks for all the replies



The QRP EFHW xfmr that Owen Duffy recommends is wound on a round cable core (type 43), not a toroid. A crucial parameter for power handling is a parameter related to cross sectional area and path length and this is superior for the cable core versus the toroid. This is the core he used in his recommended xfmr. I suspect it might be challenging winding many turns on it.

73 Barry N1EU


yes, thanks for you comments. EFHW is very good antenna for SOTA, I use it myself also.

What I wanted to know is that has some one measured these transformers and now I can see they have. Thanks for their efforts.:+1:

quote “Just wondering if anyone else has measured these famous efhw transformers with diffirent core materials? Oh yes, I know it is only one dB, but thats not the point here.”

73 good luck



thanks for the link Barry, will check it

73 Marko OH9XX OH3XR



Your conclusion regarding the efficiency of the ferrite mixes 43 and 61 seems to me a bit daring and too generalized (not taking into account the frequency range of operation). The main reason is because not both of the compared Ununs have an optimal design which can falsify the results.

Unun FT-140-61 2t primary

  • Core efficiency (along Owen Duffy):
    98% @ 7.1 MHz
    98% @ 10.1 MHz
    96% @ 14.2 MHz
    78% @ 28.5 MHz
    -> optimal design
    -> efficiency drops below 90% at 21 MHz
    -> total variation of core efficiency approx. 20%

Unun FT-140-43 2t primary

  • Core efficiency (along Owen Duffy):
    66% @ 7.1 MHz
    67% @ 10.1 MHz
    70% @ 14.2 MHz
    72% @ 28.5 MHz
    -> inefficient design, could be improved by 3t primary (below)
    -> total variation of core efficiency approx. 6%

Unun FT-140-43 3t primary (and 21t secondary)

  • Core efficiency (along Owen Duffy):
    85% @ 7.1 MHz
    85% @ 10.1 MHz
    86% @ 14.2 MHz
    88% @ 28.5 MHz
    -> good design
    -> total variation of core efficiency approx. 3%

It would be interesting to see if/how the core efficiency (along Owen Duffy) plays in your comparision using a rewound FT-140-43 3t primary and 21t secondary?
Note: The value of the compensation capacitor may need to be adjusted.


Hi Heinz,

My application is only from 7 to 14 MHz, so within this frequency region the FT114-61 transformer efficiency seems to be similar to:


I will try the FT 140-43 with 3 turns primary



Here are efficiency figures for 14MHz per Owen Duffy calculation method for type 43 ferrite and 3T primary. First figure is for single core and second figure is for stacked dual cores:

FT82-43 73% 87%
FT114-43 75% 88%
FT140-43 86% 93%

73 Barry N1EU


That is what I measured yesterday. At 14 MHz FT 140-43 (2 T primary) single core loss 1.75 W, that results in an efficiency of about 83 % or am I wrong?



Peter, Owen Duffy calculates only 68% efficiency with 2T primary. It climbs to 86% with 3T.



Ferrite materials can vary considerably depending on the manufacturer and even the batch. Tolerance for AL is typically +/- 20% so it’s possible that this is just a particularly efficient core. The core could be anywhere from ~60% to ~74% efficient with 2 primary turns and still be within tolerance.

Peter, if you’d like a better idea of what the AL of that core is you can do a quick test by putting a bunch of turns (Fair Rite uses 5) through the toroid, measure the inductance in uH, then do a bit of algebra and work out the final value: uH=(AL*Turns^2)/1000.



As a not so well qualified layman, it is not so clear to me what is being implied here. Are we talking about insertion loss or blocking impedance, or neither? Doesn’t the importance of one over the other depend on what kind of antenna you are building!
Cheers Matt


FWIW I often find this link useful:



I believe the efficiency numbers are losses incurred in the transformer material. Another consideration is SWR losses introduced by too low inductive reactance of the primary winding (which is effectively in parallel with the transformed impedance of the antenna, which should be 50 ohms by design). A primary inductive reactance of 100 ohms results in an SWR of 1.5:1 if the antenna is optimally matched. So you really want at least 100 ohms and hopefully a bit more. This is what makes type 61 ferrite a challenge - for the primary winding you need 4T@14MHz and 5T@7MHz to get this much due to its lower permeability. 3T with type 43 gives far more than 100 ohms (i.e., 200-400 ohms at these freqs) so it’s easy to get low SWR and reasonable number of windings.

Barry N1EU


The original post by AC3B started out with questions about powdered iron toroids, T-2 and T-6, etc., as well as ferrite cores.

Our posts then got into ferrite baluns and various kinds of ferrite cores. This touched a lot of related questions, with some confusion evident among the various posts.

The key point about the powdered iron cores is that they are very different from ferrite cores.

  1. Powdered iron cores have names in the form T-50-2, T-50-6, etc. These names use a code:
    A) The T means toroid
    B) The first number is the diameter (OD) in 100ths of an inch. T-50 is 1/2 inch OD.
    C) The second number is the core material. 2 is red, 6 is yellow, 10 is black, etc.

  2. If you want to know what’s going on with these cores, please go to Micrometals’ website and take advantage of their expertise:


They are the actual manufacturer of many of the top-quality powdered-iron cores we use for RF, no matter who sells them. They have lots of information about Q, power handling, frequency ranges, etc.

  1. Powdered iron cores are mostly used for stable inductors and tuned transformers at HF frequencies. Ferrite cores are mostly used for HF broadband transformers, baluns, and chokes used to suppress EMI.

  2. The many types of applications have a broad area of overlap - so you may see ferrite cores used in tuned circuits, and powdered iron cores used for broadband transformers, baluns, EMI chokes, etc.

  3. For HF SOTA use, we are going to be using mostly type 2 and type 6 material for tuned transformers.
    A) Type 2 has a permeability of 10
    B) Type 6 has a permeability of 8.5

  4. We prefer type 2 mostly for lower frequencies, approximately 2-15 MHz.

  5. We prefer type 6 mostly for higher frequencies, approximately 5-30 MHz.

  6. Both materials will work quite well for tuned transformers for our 40-30-20 meter bands. Their efficient performance overlaps over a broad range of frequencies.

  7. Type 6, yellow, has better stability with temperature than Type 2.

  8. Type 6 has slightly higher Q, in many cases, than Type 2.

  9. Type 6 requires more turns of wire than Type 2, for the same inductance.

  10. Both Type 2 and Type 6 toroids can handle surprisingly large amounts of power when used in tuned circuits. The power level is determined by the application, the loaded Q, temperature, and other variables, so there are no simple guidelines for power handling. A good rule is if the core feels hot to the touch, use a larger core.

  11. Ferrite cores become non-linear at high power - they saturate - so designing with ferrites is very technical. Powdered iron does not saturate as easily, so overheating is the main issue to be avoided.

  12. Some experience tells me that, for general use:
    A) A T-50-2 or T50-6 toroid will conservatively handle 5-10 watts of RF at HF in a filter or tuned circuit.
    B) A T-106-2 or T-106-6 toroid will conservatively handle 50-100 watts of RF at HF in a filter or tuned circuit.

  13. Smaller cores can be used with excellent results. My KX-2 uses cores smaller than T-50 for 10 watts.

  14. The Q’s of the Type 2 and Type 6 cores are so high, that capacitors may contribute more loss in tuners than the powdered-iron toroid inductors.

  15. I use a T106-6 core in my homebrew end-fed tuner I use on most of my SOTA activations. According to Micrometals, this core will deliver:
    A) Q = 380 @ 3.5 MHz, 26t #18, 8.1 uH
    B) Q = 400 @ 7 MHz, 15t #18, 2.8 uH
    C) Q = 365 @ 10 MHz, 10t #18, 1.3 uH
    D) Q = 270 @ 14 MHz, 5t #18, .39 uH
    E) Q = 260 @ 21 MHz, 5t #18, .39 uH

I chose this oversized core to keep my tuner loss low. When I use a reactive match on a non-resonant end-fed antenna, often with a small counterpoise, the tuner is doing a lot of the work to convert the reactive current to a resistive 50-ohm load for the rig. Losses tend to increase with reactive loads.

I often use an end-fed 66-foot wire hung on a fishing pole, inverted-L, and I can easily match it on 5, 7, 10, 14, 18, and 21 Mhz with no traps, no links, and a minimal counterpoise. In order to get this kind of performance, I use a tuned transformer with multiple taps for various impedances. Both the input tuned circuit (primary) and the output tuned circuit (secondary) have polyvaricon capacitors supplemented with mica caps for the lower frequency bands.

My KX2 sees a virtually perfect 50 ohm load on all the bands I use. The overall Q of the system is high enough that I can see the load impedance vary as the wire flies in the wind, when I move from one site to another, with the wire at different heights above ground, etc. These variations are inherent in the field, regardless of how you attempt to deal with them.

My SOTA logs and RBN spots tell the rest of the story.

I also spent a lot of bench time working with type 43 and 61 transformers. Type 43 is so lossy that it’s hard to measure a resonant frequency in a tuned circuit. It’s really designed for broadband transmission line transformers and EMI suppression chokes. It’s not a good choice for conventional transformers. Many of the people posting here don’t seem to know what these terms mean. I would not use type 43 to feed the antenna in my SOTA system. The reason some antennas use it is that its loss hides the mismatch errors in the system! It lowers SWR by adding loss.

Type 61 ferrite has much lower loss and can offer good performance at our HF frequencies. I made some conventional transformers that had low loss with type 61, but I don’t know how to make these perform over a range of frequencies, because of significant reactance variations.

Therefore I chose to use an adjustable high-Q tuned circuit, using powdered-iron toroids, with three sets of variables, so I can match my antenna across many bands. I change bands often as I hunt for S2S contacts, but the band changes just take a few seconds, since I know where my settings are. The KX2 doesn’t seem to care about a 2:1 match, so I can just peak the noise and call if I’m in a hurry to get someone!

Beware the high-Z tuners with just one variable capacitor - they work, they may even work OK for what you want to do, but you need at least 2 or 3 variables to take care of impedance and frequency variations. Just changing from the CW part of the 20M band to the SSB end is often a LARGE change of Z!

Or use an auto-tuner. The Elecraft and many other auto tuners use powdered-iron toroids and high-Q caps, and they will often deliver useful matches for reactive as well as resistive loads. If you want a really nice match for an end-fed half wave, which may be near 3000 ohms and maybe reactive as well, use an adjustable parallel-tuned circuit with powdered iron toroids!

I do a fair bit of chasing as well as activating, and one thing I know: I can’t even hear many SOTA activators! Every db counts when you’re at the noise floor of the chasers!

Leave the lossy ferrite on the bench…likewise small lossy coax feedlines - small lossy loops - etc. With the solar flux at 69, every db matters!