EFHW Antenna: A Detailed Analysis (NEC4.2)

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

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

    K4DJM Ham Member QRZ Page

    I think the way these antennas interact with their counterpoise, be it vestigial, a wire to ground, or the coax sheath itself is very different when fed with a proper transformer. For example, a wire to ground from the feedpoint "common" could have a significant impedance compared to 50 ohms, but be a short ot ground for all practical purposes in a 2500 ohm system. When I saw the way this antenna was being driven in the simulation, I wrote off any possible conclusions as junk.
    AA5CT and KD6RF like this.
  2. W8IXI

    W8IXI Platinum Subscriber Platinum Subscriber QRZ Page

    Well, I resemble that remark. :)

    Not so much QRP, but low power at less than 45 watts. Due to restrictions, I settled upon a 31' piece of wire in the attic years ago and made it work. Luckily the attic is entirely walkable, because there is always a bit of trial and error in final tuning. But the final component to be worked out was the RF ground. It turned out to be a tunable counterpoise with radials.
    Last edited: Mar 31, 2018
    N0TZU likes this.
  3. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    Good, I hope it works well for you.

    When I've been restricted to an "attic" (twice in 53 years) I ran a loop around the interior perimeter of the attics and brought window line down into the shack via holes drilled from attic floor through closet ceilings, and used a balanced line tuner in the shack and ran a kilowatt both times with zero RFI problems, even with neighbors.

    It worked okay considering what it was, and used it on 80 through 10m but it wouldn't hit 160. Luckily in both cases I was up 30-50' above ground in the attic (2-1/2 to 4-level high condos) and that may have helped. In 1999 I was temporarily in a rented condo (town house) with the nearly 50' high attic since it was three levels but the loop did not do well on the lower bands -- it did "great" on the higher bands, and I worked a lot of DX on 10/12/15m. I never really figured out why it did not do well on 30m/40m but the higher bands were open a lot, so at the time it didn't matter much.

    I was waiting for the new house, which we moved into in November 2000, so this was all pretty temporary but it was still fun!
  4. W8IXI

    W8IXI Platinum Subscriber Platinum Subscriber QRZ Page

    Easily worked all states and all continents. No big deal for an average ham like me. Never a contestor or paper hanger. But I like them all. The only thing on the wall in the shack is an ARRL calendar. For some of us, ham radio is not about where we live, but what we have done with that choice. :)
  5. WA7ARK

    WA7ARK Ham Member QRZ Page

    Boy, what an understatement.

    The doubters are focused on the perceived need for common-mode current on the coax feedline to balance the current that flows in the secondary winding of the transformer. They are focused on the common-mode current on the coax, but what happens if there is virtually no coax; i.e. the transmitter is connected directly to the Unun transformer with only about 6 inches of coax...

    I purchased a commercial 1:49 Unun auto-transformer, and posted some measurements on it in my #6 in this forum thread. I also did the following test, which should add some grist to the mill...

    I put up a temporary 76'10" wire antenna using available supports. The wire is PTFE insulated #20awg M1678/4. The insulation is 0.008" thick. It was put up like an upward sloped inverted-L. The bottom end is just 3ft off the earth. The vertical section goes to a height of 18.6', turns the corner and the far end is about 30ft agl. Here is the picture: (click to enlarge)


    The Unun is anchored to the fence with the turnbuckle, but there is no electrical connection. The loop on the top of the plastic box is not-connected to any part of the transformer or coax PL259. There is no connection to the "ground" terminal on the Unun.


    So I connected the AA-600 analyzer to the 6" piece of RG58A coax, and I swept from 1Mhz to 30Mhz to find SWR dips at 5.6, 12.0 and 18.4MHz. When scanning such a wide range, the AA-600 does not find the exact frequencies of the dips. It is necessary to zoom in around each dip to take more detailed data.

    I focused on the dip near 5.6Mhz, which is where the system (wire, Unun, short coax, RigExpert) first resonate. I wanted to compare the SWR scan with no counter-poise (only the 6" RG58 bnc jumper) to another SWR scan where the Unun's ground terminal is connected to the galvanized wire fence that the Unun is hanging on.


    The dark blue trace is a plot of SWR with NO counterpoise (no connection to the fence). The faint blue trace is with the Unun grounded to the fence. First, note that the SWR is very low in either case. Second, note that connecting to the fence, which is hundreds of feet long, shifted the SWR null 140kHz higher , which to me is counter-intuitive.

    Here is the actual data:
    no ground: Lowest SWR=5.527MHz, SWR=1.19, Z=57.5 - j5.6
    fence gnd: Lowest SWR=5.657MHz, SWR=1.22 Z=52.8 - j10

    So, by actual measurement, the SWR is not substantially effected by either grounding the Unun or letting it float. The SWR is very low, even without any physical counterpoise. The frequency at which the lowest SWR occurs unexpectedly shifts higher in frequency when a counterpoise is added. The lowest SWR frequency occurs at a frequency substantially different than that predicted by 468/(wire length) = 468/76.83 = 6.09MHz ( actually near 5.6MHz, so the 76ft wire acts like it was 83ft long).

    I also fed the Unun (with and without a connection to the fence) through 58ft of RG8 (laying on the earth, stretched out straight at right angles to the fence line) to see what that did to the SWR and resonance. Answer, not much. Just like using the fence as the counterpoise, the coax shifted the resonance higher by about the same amount (140kHz). Adding the fence to the counterpoise shifted the resonance another 5kHz higher.

    So how can this work at all with no counterpoise...
  6. WA7ARK

    WA7ARK Ham Member QRZ Page

    Ok, wise one! How should the common-mode current on the coax shield be modeled? How would you simulate the transformer? We are all students here. Enlighten us!
    Last edited: Mar 31, 2018
  7. WA7ARK

    WA7ARK Ham Member QRZ Page

    I forgot to add this to post #35. Here is the input impedance of the Unun when the output side is terminated with a 2424Ohm carbon resistor. This particular transformer is specified to be used from 1MHz to 10MHz at 1.5kW.

    2424/(7^2) = 2424/49 = 49.47 Ohms...

    Last edited: Mar 31, 2018
  8. AI3V

    AI3V Ham Member QRZ Page

    But your wire is not 2424 j0. Over the frequency range of your sweep.

    Why don't you terminate your unun with the same impedance of your wire and measure it's (the unun) characteristics?

  9. SM0XHJ

    SM0XHJ Ham Member QRZ Page

    But a transformer only transforms differential mode currents, not common mode currents. So for common mode currents on the feed line, it doesn't mater if there is a transformer there or not, unless the transfomer is a full transformer with high CM impedance. Even then, the CM current reduction (and CM impedance) will have nothing to do with the transformer ratio.
    SM0GLD and W9XMT like this.
  10. WA7ARK

    WA7ARK Ham Member QRZ Page

    The two-winding transformer which I measured in post#6 and #35 is a three-terminal auto-transformer, ergo no common-mode isolation.

    Top of the three-turn winding goes to the coax center conductor. Top of the twenty-one-turn winding goes to the wire. The bottom of both windings are tied to the coax shield (and "ground" terminal).

Share This Page