EFHW Antenna: A Detailed Analysis (NEC4.2)

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by W9XMT, Mar 29, 2018.

ad: l-rl
ad: Left-2
ad: Left-3
ad: Subscribe
ad: abrind-2
ad: L-MFJ
ad: l-BCInc
  1. K4DJM

    K4DJM Ham Member QRZ Page

    Your simulation shows a way of generating a very good counterpoise, but did you do a VSWR sweep with EZNEC? Contrast the 50 ohm plot with an alternate Zo of 2500 ohms...

    Regarding transformer characteristics, there are several people asking about that now, and I will post a link to a related QRZ thread further down the list...

    W3JJW likes this.
  2. K4DJM

    K4DJM Ham Member QRZ Page

    Well Mark, speaking for myself -- If I was free to do so, I'd run ladder line up to a balanced doublet with a balanced tuner, or just deploy a fan dipole. However, that would be too noticeable in my neighborhood. We have restrictions. An EFHW can be deployed in a single direction from my shack back into the woods behind my house. So, I believe that in general MOST OMs would go the balanced route if there were not some factor at their QTH pushing them away from it.

    Regarding the 49:1 transformer: I did some testing last year that initially left me a bit disappointed. My testing involved doing S11 measurements at the coax port and assuming that an "open" at the antenna terminal should give me close to 0dB return loss. I took the measured return loss, divided that by two (round trip to thru conversion) and determined that was an estimate of the tranformer's thru loss. When I measured S11 with a 2500 load, I got reasonable numbers, but I was disappointed with the open-circuit S11 measurements and my computed thru-loss. I started a QRZ discussion.

    Well, the vendor got involved in the discussion and told me that the measurement really should be done using two transformers back-to-back. Well, I agreed that would certainly be a more direct route. I challenged the vendor to loan me a 2nd transformer and he complied. In the end, the thru loss of a pair of transformers was 2dB at 30 MHz, so I concluded that a single transformer would have a loss of about 1dB at 30MHz. The vendor said that he typically measures 0.7dB transformer loss in his setup.

    In both test methods, computed loss at lower frequencies was much better.

    That thread, and my test results can be found here: https://forums.qrz.com/index.php?threads/legal-limit-efhw-transformer.557447/

    WB2UAQ likes this.
  3. WB2UAQ

    WB2UAQ Ham Member QRZ Page

    Having worked with ferrite cores only for transmission line transformers I became interested in the loss of these transformers. I wound a few of them and tested them two ways with good agreement between both methods. I only tested a 1:64 version with a 2 turn primary and 16 turn secondary. At 7 MHz they have a loss of about 1.1 dB. First I measured the impedance looking into the primary and then tuned out the residual inductance with a capacitor of 185 pF. This is optimized for 7 MHz and the resistance was about 48 ohms with a 3280 ohm load. The load is a 1/4 W metal film resistor with a 50 ohm RF voltmeter probe on the ground side to measure the current. If the load consumed all of the power supplied from the source the power drop from in to out (into the 50 ohm RF probe) by 18.3 dB. The power dropped by 18.3 dB + 1.1 dB for the core and wire losses. I then made a second transformer (both BTW on FT-240-43 cores) and did a back to back test with about the same loss (50 ohm source to 50 ohm probe again). The SWR degraded a bit with back to back test but not by much.
    I was also curious about why these transformers needed 12 or 14 AWG wire so I wound one with 21 AWG wire and the loss results were exactly the same. This makes good sense to me because the current in the secondary is pretty low at about 140 to 150 mA when running 100W assuming the 1.1 dB loss in the core.
    I also did some tests with twisted vs untwisted primary winding and I did see any significant improvement with the twisting. I was going to revisit the twisting vs untwisted version but lost interest in going further. I switched over to winding and testing a QRP version based on a binocular core and it has lower loss at about 0.5 dB. The core is a BN-43-202 (Amidon #'s and not Fair-Rite Product #'s the actual mfr). For this version I had to add a series inductance instead of a shunt L (0.6 uH). I didn't get a chance to test it on 40 but I friend tested it into a straight 64 ft wire at a 45 deg angle to the ground. He had very good results running 5 W PEP into Europe. I am looking forward to testing it with my old HW-8 later this month.
    I had a great time testing these transformers and I thought the loss would be more like 3 dB. Still 1 dB is pretty high.....100W in and about 20W going to heat. BTW I did run 100W into the back to back test as well and both FT-240-43 core became warm after a few minutes of sending V's and CQ into a dummy load. The cores were getting warm and not the copper even with the 21 AWG wire. How can you run 1 kW thru these even with few of the cores stacked up? The flux density is pretty high! Almost forgot. I don't see how these transformers are called "auto transformers" . They are conventional prim/sec type transformers. 73, Pete
  4. WB2UAQ

    WB2UAQ Ham Member QRZ Page

    Got similar results....didn't see your post until now! I also did an open circuit test and the loss was like 3 dB based on that so I pressed on with more tests. Ferrite is pretty lossy for a core and why they are best for transmission line transformers where the flux is kept down and not the energy path. 73, Pete
    KD6RF likes this.
  5. K4DJM

    K4DJM Ham Member QRZ Page

    In the transformers I tested, I didn't get up to the 1dB loss numbers until I got to 30MHz. I won't go QRO on 10M if that band ever opens back up, but 100W is plenty there.

    Apparently, core heating is a thing on these transformers for sure. I've got the 2KW version of the E73M transformer, and it is rated ICAS.

  6. G3YRO

    G3YRO Ham Member QRZ Page

    The best way to feed an End Fed Half Wave antenna is not to use a Transformer or coax at all . . . just bring the end of the wire to an ATU, and tune it properly.

    Over the decades I've been licensed I've done this in several QTHs where the shack was in an upstairs room, usually a loft conversion. So the end of the wire came in through the top of the window straight to an ATU sitting on the windowsill.

    As 160m has always been my main interest, this was usually a 260ft end-fed . . . but with the ATU I could obviously tune it on all bands.

    Never used any dedicated counterpoise . . . but here in Britain we usually have Central Heating consisting of copper pipes connecting radiators all over the house, as well as pipes to the various sinks and toilets. This is all connected to the mains earth too, and the gas and water pipes coming into the property underground.

    With this setup, I always felt it worked just as well as a centre-fed Half Wave Dipole (that I used in other QTHs) . . . and even running an Amplifier, never had any "RF" issues.

    I've never understood why more people don't do this ! Lots of people who can't put up "proper" antennas do it over here, often with just 50ft of wire out the window to the back fence.

    Roger G3YRO
  7. W9XMT

    W9XMT Ham Member QRZ Page

    At a minimum, the counterpoise in this situation consists of the "earth" conductor of the a-c mains connection powering the transmitter.

    RF current flowing along that path to reach a buried ground rod at the service entrance produces e-m radiation for the same reason it does so when flowing along the "End Fed Half Wave antenna."
  8. WB2UAQ

    WB2UAQ Ham Member QRZ Page

    Roger, I think you have been pretty lucky not to have issues feeding end-fed wires. In several situations I had no choice and had to use a random wire and each time RF currents crawling around did something in the way of interference. One time I caused a chandelier to eerily glow freaking out a few people or turned on or off some entertainment electronics, etc.. This is way before the use of PVC and other plastic piping thru out the house.
  9. G3YRO

    G3YRO Ham Member QRZ Page

    It just seems strange to me that a lot of US stations say they've had problems using this kind of setup . . .

    Whereas LOTS of people in Britain have done so without any issues !

    I would infer from this that our standard Mains Earth going round all our Ring Mains sockets all over the house, and also always connected to the copper pipes used for central Heating and other utilty pipes seems to work well as a Ground and/or counterpoise for such antennas.

    Roger G3YRO
  10. WB2UAQ

    WB2UAQ Ham Member QRZ Page


    Out of curiosity, do you have ground fault current interrupt outlets and/or breakers in the UK? (GFCI) I have tripped a few of these at times. Every outlet that is even close to a possible connection to earth must have one of these (or one can protect a string of outlets but is seems best to have one for each case).

    In the US there are three connections at our 120 V receptacles that go back to the circuit breaker panel. AC (hot), AC common(or the neutral that goes back to the breaker panel and is attached to ground ) and ground (goes to the breaker panel, neutral and then to the copper water pipes and a ground stake). AC common is the center tap of the 120/240 V secondary of the power transformer. The ground wire is not supposed to carry current and is for safety and used to be of a lower gauge than the other conductors but that has changed. All of these wires can float up and get hot with RF because there really is no such thing as an RF ground. My house was built in '83 and has all copper water pipes but newer houses have a lot of plastic so I guess things could be even more uncontrollable.
    73, Pete

Share This Page