Legal limit EFHW transformer

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by K4DJM, Mar 2, 2017.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: Left-3
ad: Left-2
ad: Subscribe
ad: FBNews-1
  1. K4DJM

    K4DJM Ham Member QRZ Page

    I'm currently running an end-fed line at 154' out of a Palomar 9:1 UNUN. The wire is longer than 1/2 wave at 80M since the 9:1 is not sufficient to feed the large impedance of a true 1/2 wave. The 154' gives me VSWR minima in convenient places without being resonant at those places.

    I've looked with interest on the true EFHW antennas (at 132'), and the relatively good reviews. I wanted to give the approach a shot, so I ordered a MEF-330-2K 2KW ICAS 50:1 transformer. I want to add an amplifier to my setup soon...

    I have not placed the transformer into service yet, but I did some measurements. I recently became disturbed while looking at the VSWR curve of the EFHW-8010. Notice that at high frequencies, the VSWR gets into respectable territory even BETWEEN the resonant points:

    Is the transformer getting LOSSY up there???

    Using a VNA, I measured the return loss of the MEF transformer with the antenna terminals open circuit. In a perfect world, I should get 0 dB return loss across the whole frequency range. I measured it at -1.02dB at 3MHz and -4.01 dB at 30MHz. For comparison, the Palomar device measured -.05dB at 1.8MHz and -1.25dB at 30MHz.

    Since those are round-trip numbers, I estimate the thru loss of the MEF device at 2dB at 30MHz. This means 37% of the power would be burnt up in the transformer!

    I would be interested in hearing if anybody else here is running QRO on a multi-resonant EFHW -- and the matching solution used thereon.

    I am getting discouraged and may move onto experimenting with a G7FEK -- which would eliminate the transformer entirely.

  2. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    Why an end-fed?

    Impossible to run a center-fed or OCFD?
  3. K4DJM

    K4DJM Ham Member QRZ Page

    I live in an HOA property with woods behind it that are not part of the HOA. EFHW was my first impulse since I only had to "stealth" a very small portion before it is off my property (and hence none of the neighborhood's business).

    The G7FEK looks attractive too, since one end is so short... It is really just a special case of an OCFD. As I mentioned, that antenna may be my next move.

    I just re-read the QST review of the EFHW-8010. The ARRL labs has lost some stock with me. The VSWR between bands was not shown in the article (it is telling at the higher frequencies), and loss was measured at a too-low termination impedance of 800 ohms. Doing an S11 sweep from 3 to 30MHz would have cost them just a few more minutes.
  4. KW4TI

    KW4TI Ham Member QRZ Page

    Well, I have been looking into this problem. Look at my page for several pieces I've written on EFHW.

    There are fundamental compromises to be made when designing a transformer to match a very high impedance load. The problem is that these transformers are typically autotransformers wound for low winding capacitance. The winds are widely separated to minimize the capacitance between the winds, and furthermore, a certain minimum number of winds are necessary to have enough choking impedance. Because of this, they have significant leakage inductance. As the ferrite is lossy, part of the leakage inductance is resistive and therefore heats the ferrite core. As the permeability decreases with frequency, leakage flux increases and so do losses.

    I think one of the ways you can tell that the EFHW8010 is lossy from the SWR is that the SWR should not be low between the amateur bands which the transformer should not match. If the 64:1 impedance match was really working, lengths for which the antenna is a quarter wave plus a multiple of a half wavelength long should transform to an even lower impedance which should be poorly matched. Instead, the SWR between the upper ham bands stays low around 2.5:1-4:1. I think this is because of transformer losses dissipating the power. One of the problems with a transformer that matches a high impedance is that even a high shunt resistance of the transformer itself can be matched, which means that you can efficiently deliver power to the ferrite core even if an antenna is not attached.

    Here is an example of the impedance I measured of a 62 ft wire. The theory predicts that the impedance of the wire should remain high at frequencies for which the antenna is a multiple of a half wave, but as you can see here the impedance of the antenna drops as the number of half waves increases. This is commonly seen, and its because a long wire becomes more like a traveling wave antenna than a standing wave antenna, and so the effective Q of the antenna drops as the antenna length increases. The common assumption that the current on the wire is sinusoidal does not hold exactly and the current exponentially decays down the wire as it is radiated. Because the Q of the wire drops at the higher harmonics of the wire and the feed impedance drops as well, a transformer designed to match a high impedance should not be able to match these higher harmonics of the wire, but the transformer produces a low SWR anyways.

    I have deduced partially what is going on through simulation of the transformer using circuit models. In fact, the transformers do have winding capacitance. When the wire is slightly less than a multiple of a half wavelength, the wire is both resistive and inductive. The winding capacitance (or extra added capacitance) of the transformer parallel resonates the inductance of the antenna and drives up the feed impedance of the wire. Then the transformer is able to match the high impedance more easily. It actually may be necessary for the transformer to have some capacitance for the transformer to match the upper bands by using this parallel resonance effect. This effect is used similarly in a Fuchs circuit, except an external LC parallel resonator is used rather than the inductance of the wire itself and a capacitor.

    I have an idea which may help the efficiency of these transformers. Because leakage inductance produces loss in the ferrite and winding capacitance is generally much less lossy than the ferrite, the transformer should be wound for less leakage inductance and more winding capacitance, and the antenna should be cut slightly short to resonate the winding capacitance with the antenna wire inductance. I have been trying to find a different winding technique that achieves a significantly lower leakage inductance with a tolerable winding capacitance to drive an end fed half wave.

    I just posted this document yesterday

    regarding my latest attempt to do this. These end fed half wave antennas are always going to be compromises, but perhaps they can be improved. I hope this might explain some of the issues and you can look at my page for links to more documents.

    OE2RPL likes this.
  5. AF7ON

    AF7ON Ham Member QRZ Page

    You can only properly analyze this problem if you specify the whole antenna and feed system, including counterpoise (or lack thereof) and feedline. As many have previously pointed out, you cannot just stuff current into the end of a wire - the current must return to the source. With a high impedance feed point, currents are low, but still have to go somewhere! Short counterpoises can work well, as can a direct connection to earth, but in many cases some common-mode current will also flow on the coax feedline so the length and routing of this affects the antenna performance and matching.

  6. G3TXQ

    G3TXQ Ham Member QRZ Page

    The G7FEK is not a special case of an OCFD.

    It is two Inverted-L antennas fed in parallel at ground-level. The open-wire section is not acting as a transmission line - it's radiating, and is just a convenient way of forming the two parallel vertical sections of the Inverted-Ls. Like any Inverted-L it will benefit from a good ground system.

    That doesn't make it a bad choice, but make no mistake - it's not an OCFD.

    Steve G3TXQ
    WB5YUZ, KD6RF and K7TRF like this.
  7. KW4TI

    KW4TI Ham Member QRZ Page

    Yes, the counterpoise and feedline all matter. However, the counterpoise just needs to be a particular resistance, and the requirements on it are modest, being less than 100 ohms or so. Also, being in contact with earth or other large capacitive object or damping medium helps as well. The feedline shouldn't be that influential, as a choke is typically used at the feedpoint, and using a choke definitely makes the SWR far less sensitive to the feed conditions. I can tell this because if I remove the choke, I can touch the end of the coax shield and the SWR can change a lot, but if the choke is there, touching the coax shield barely matters at all. So yes, all this stuff matters, but in terms of relative importance the impedance of the radiating wire itself is a far bigger consideration when determining if the transformer can match the system than the counterpoise, as the impedance of the radiating wire is in series with the counterpoise and is has a much higher resistance than the counterpoise.


  8. K4DJM

    K4DJM Ham Member QRZ Page

    I sort of knew I was going to be dinged on this a few seconds after I posted. I should have said they were physically similar. In construction, they differ mainly at the feedpoint, and I realize that they work very differently...
  9. K4DJM

    K4DJM Ham Member QRZ Page

    This sort of deviates from the original point of the thread. There are other threads that discuss the details of the EFHW system, but I think that if you start with lossy components, you are sort of screwed from the start. I will very likely never deploy that new transformer I just purchased, because measurements show that too much power will be spent heating it. I was fishing for a better transformer solution. Dan returned with the expected response that this was a difficult problem...
  10. WB2UAQ

    WB2UAQ Ham Member QRZ Page

    Why would ferrite be used for conventional transformers in this application? Conventional meaning a primary winding and secondary winding coupled thru flux linkage in the core. For whatever reason, when this endfed wire antennas started showing up in discussions, I immediately questioned the use of a conventional transformer (auto-tranformers are conventional) designs. I don't understand this whole thing about using 1:9 Z step up to begin with. Low SWRs are discovered for the wrong reasons just like the old article "Low SWR for the wrong reaons". Also VNAs are being used at mW power levels. Has anyone measured the SWRs starting at low power and to see how the SWR is changing as the power is raised? There is something basically wrong with this EFHW or I should say just endfed wires. I see so many questionable aspects about them and the kits and hardware that are being marketed for them. 73, Pete, WB2UAQ

Share This Page