ad: MonitorSens-1

The EFHW Transformer Redefined

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by AF7NX, Jul 22, 2021.

ad: L-HROutlet
ad: l-rl
ad: MessiPaoloni-1
ad: L-MFJ
ad: Left-2
ad: HRDLLC-2
ad: L-Geochron
ad: abrind-2
ad: Left-3
  1. K6BRN

    K6BRN XML Subscriber QRZ Page

    Mike (WA7ARK) is very correct - EFHW transformer heating (loss) is less on higher frequency harmonics, not more. This is demonstrated every time I operate FT8 with an amp for an extended period. Heating becomes an issue near the HW fundamental frequency, regardless of the transformer used and is much less pronounced at higher frequencies.

    For example, the MyAntennas EFHW-4010 and -8010 use essentially the same transformer. But I find heating is an issue on 40M with 63 feet of wire and on 80M with 126 feet of wire. Also, the problem gets worse the lower the antenna is hung. A good explanation of exactly WHY this happens is welcome - I simply note that this is WHAT happens.

    I believe Danny, who designs and sells these antennas, has commented before on flux density issues (nod to post on this above) and winding implementation difficulties at lower freqiencies leading to higher losses and increased core heating.

    Also - I'd think bonding a metal case to the cores would impact their efficiency and response over frequency, even if were a non-magnetic pot metal as the O.P. describes. Guess I should read his article.

    OK ... off to do so now.

    Brian - K6BRN
    KU3X likes this.
  2. KX4OM

    KX4OM Ham Member QRZ Page

    Joe - "Waste of effort aside the engineering exercise for the sake of exercise. Does nothing from a practical standpoint of operating."

    Whatever floats your boat. Mine isn't (just) operating, which I do occasionally. It's the technical stuff.

    Ted, KX4OM
    W6JJZ likes this.
  3. KW4TI

    KW4TI Ham Member QRZ Page

    If it's interesting to you, I made some mathematical models to calculate the efficiency of EFHW transformers based on a circuit model.

    I also came up with a winding design and compared the losses to a conventional autotransformer.

    KU3X likes this.
  4. W6JJZ

    W6JJZ Premium Subscriber QRZ Page

    Thanks, Dan.

    I recall seeing those several years ago. Was it here on QRZ? Not sure.

    Anyway, I'll take another look at them. I suspect AF7NX will get more out of them, however. He'll have an easier time with the math. I do have a Ph.D.----in American history. His (I gather from the bio on his blog) is in either E.E. or applied physics.

    Charlie, W6JJZ
  5. AF7NX

    AF7NX XML Subscriber QRZ Page

    If you read my write up, you will see that I did make those measurements. I don't have a 300W transmitter, but my estimate based upon the temperature measurements I made is that the small box would work at about 50% duty cycle with 300W indefinitely with the temperature rise I measured at lower powers -- staying will with int eh Curie limit - which is rather low (130 C) for type 43 material.
  6. K7JOE

    K7JOE Subscriber QRZ Page

    Fair enough, Ted.

    But remember that's the very reason why enginerds are often overruled by marketing, operations, and finance once a product hits "the real world" !
    There's the practical aspect of why we do things and for me, that has to intersect the technical costs and benefits.

    A half a dB is not a new mousetrap...

    I'm reminded the camel is a great creature technically - but most preferred riding a horse for practical reasons.
    KA0HCP likes this.
  7. AF7NX

    AF7NX XML Subscriber QRZ Page

    However the number of times the open hysteresis loop is traversed doubles when the frequency doubles - effects roughly canceling each other.
  8. AF7NX

    AF7NX XML Subscriber QRZ Page

    Curiously, the nanoVNA showed roughly constant losses versus frequency -- but be aware that that is very low excitation level. The heating expmnts I did showed that there was more loss at the lower frequency and less loss, even less than the nanoVNA showed, on the upper bands. If you think long and hard - you get to the point where you realize that losses are inversely proportional to the primary magnetizing inductance. Then the issue becomes how to have a large primary magnetizing inductance and not sacrifice high frequency performance.

    As to why bother -- well most of us don't like to spend money; big ferrites are expensive. My xformer uses about $4 of ferrite. It is also light enough to ride on the end of the antenna wire if you want it all up in the air.
    Gary AF7NX
    KX4OM likes this.
  9. K6BRN

    K6BRN XML Subscriber QRZ Page

    Hi Gary (AF7NX):

    "As to why bother -- well most of us don't like to spend money; big ferrites are expensive. "

    Hmmm... not a very good reason - but a persistent cliche' - it tends to paint hams as eccentric paupers and misers; some are, most are not and we get enough strange looks as is, so why perpetuate this nonsense.

    A better reason would be for the challenge and for fun. This is a hobby, after all. It's basically techno-entertainment that is sometimes beneficial. If you enjoy experiementing and developing a better EFHW transformer, that's good enough for me. And since it's your hobby, your time, your efforts and your money going into it, I wouldn't pay too much attention to anyone who suggests you're wasting your time. If you're enjoying the journey, that's payback.

    Personally, I'd be more interested in how elegant and useful your solution is, which is what attracted me to resonant "EFHW" antennas in the first place. And I was a big skeptic and had to have a fellow ham demonstrate what he could do with one. THEN I was interested. But thermal dissipation issues have always been a problem.

    Many users have "skeletonized" their transformer enclosures to improve air cooling of the cores while others have even put in a forced-air blower tube and better venting. Both approaches claim significant (but so far unquantified) success. I've toyed with the idea of sealing the SO-239 with silicone and filling the enclosure 2/3 with mineral oil or antifreeze. Both have materials compatibility, leakage and vapor issues and will affect parasitic capacitances. But if successful would improve the thermal path to the case and air, as well as increasing thermal mass which would slow initial heating and make the transformers more forgiving.

    Your quoted 300 watts continuous FT-8 operation for your metallic enclosure is pretty much what I see with the MyAntennas 2-core, plastic-enclosed EFHW-8010-2K on 80M with a 126 foot wire, or on 40M with the 64 foot wire. So its not a surprise to me. But it does tend to indicate that the thermal dissipation path is not improved much in your 1st cut, despite the metal case.

    BTW - good thought regarding isolating the enamel primary and secondary windings with kapton though efficiency might suffer a bit. At high power and insulation failure could easily damage an attached amplifier, and I believe I've seen exactly this happen.

    Curiously, I've found that the cores in the 49:1 transformers heat the most on the half wave resonant frequency of the attached wire, which seems to drive efficiency vs. frequency very significantly. For example, the MyAntennas EFHW-8010 and -4010 transformers are essentially identical yet I seem to have the same power limit on 40M with the shorter wire as I do on 80M with the longer wire using the very same transformer. I.e power limits seem much higher on 40M using the longer wire.

    Also .... as antenna height and clearance for surrounding (conductive) objects increases, transformer heating becomes less of a problem.

    I'd be interested in hearing some speculation, analysis, prognostication (etc. it IS a hobby after all) as to why this happens. But I've yet to see it addressed on any thread. Curious.

    Time for the antenna experts to jump in.

    Best Regards,

    Brian - K6BRN
    KX4O likes this.
  10. K8DO

    K8DO Ham Member QRZ Page

    I am currently playing with EFHW antennas in various configurations. They are fun and do work. I do not like the expensive ferrite dummy load called a transformer. After building and testing a bunch of transformers (20 and 40 meters) by following several published designs I no longer use them. (my amps eat them for breakfast)
    Mike points out a fact of physics that having one end of the horizontal antenna (especially the feed end) down in the dirt has a price to pay. Get the current portion of the wire as high as possible for as far as possible.
    If you have the luxury of monoband antennas, for EFHW let me strongly urge you to use a coil and capacitor in a conventional tank circuit. I currently have two in use for EFHW (20 and 40 respectively)
    The 20 EFHW - which is currently a vertical - covers the entire band at less than 1.2 SWR on a fixed setting of the tuner (dead flat, not even a twitch on the needle at 14.150).
    It is 150 feet from the shack so I am not running out in the rain to retune.

    The 40 EFHW - which is currently a sloper - covers the entire band at less than 1.5 SWR on a fixed setting of the tuner (dead flat at 7150)
    It is closer to the shack but I am still not running out in the dark and the sleet.
    It will become vertical in the next iteration when I can find time to rehang it. This being retired eats up a lot of hours per day.

    The mechanical tuners run cold at 1500 watts - the transformers (that survived) not so much.
    10 seconds of key down then 10 seconds key up, rinse and repeat until either boredom or smoke.
    KX4O likes this.

Share This Page