Feeding EFHW Vertical antenna

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by VK4WTN, Apr 4, 2021.

ad: L-HROutlet
ad: l-rl
ad: l-gcopper
ad: Left-3
ad: Left-2
ad: l-BCInc
ad: L-MFJ
ad: QSOToday-1
ad: abrind-2
  1. WA7ARK

    WA7ARK Ham Member QRZ Page

    The OP (Wayne) wants a 0.5wl vertical for the 20m band. That makes the vertical radiator wire ~ 33.6ft tall. I suspect he does not want the extra height of a 0.25wl ladder-line matching section that would make the antenna 0.75wl (52.5ft) tall.

    IMHO, the best way to drive ~34ft vertical wire is the method I proposed back in post #12 to this forum thread (transformer and ground rod).

    If you trim it to 14.175MHz without regard to the second harmonic ( falls into10m), the Red bars show the expected Swr plot from 14 to 29.5 MHz:

    If you trim it for best swr at 28.5MHz, then the Blue bars show the expected Swr:

    Last edited: Apr 6, 2021
  2. VK4WTN

    VK4WTN Ham Member QRZ Page

    thanks for the detailed information Michael. I will initially try the 49:1 transformer match with the .05 wavelength counterpoise. Hopefully the nice low angle of radiation will work well on dx
    cheers Wayne VK4WTN
  3. G5TM

    G5TM Premium Subscriber QRZ Page

    Hi .. would this 17/20 arrangement work without a ground rod and just fed by a 49:1 near ground level or just above ground?
  4. WA7ARK

    WA7ARK Ham Member QRZ Page

    No! look at the "Current on Segments" plot in post #12. The 200ma (current on segment 1) from the transformer has to have a place to go. You gotta have a "current sink" to satisfy Kirchhoff!

    The first alternative for this antenna is the ground rod, which creates a Galvanic connection to the dirt under the monopole, and that is where the 200mA goes. However, compared to driving a 0.25wl vertical monopole against a ground rod, which has horrible losses (like half the transmitter power), the current into the ground rod in this design is about 1/5 of what it would be with a quarter-wave monopole, making the power loss in the ground rod about 1/25 of what it would be a quarter-wave monopole! You dont need a CM choke on the coax.

    A second possible alternative would be to elevate the entire antenna by a couple of feet, and then add 2 to 4 short (parallel to the ground) radials in lieu of the ground rod. The short radials act as a "current sink" (counterpoise, other part of the antenna). This will also need a CM choke on the coax.

    A third alternative is use the common-mode current on the coax shield as the current sink, and then place a coax-Common-Mode choke a few feet down the coax from the transformer. Presumably, the coax will run parallel to the ground, but be isolated from the ground, so it becomes just a single "radial" which is as long as the distance along the coax from transformer to choke.

    Running only coax from transformer to rig is asking for trouble...
    Last edited: Jun 9, 2021
    G5TM likes this.
  5. G5TM

    G5TM Premium Subscriber QRZ Page

    Thank you
  6. G5TM

    G5TM Premium Subscriber QRZ Page

    Ah ok just re-read this so option 3 in summary: a cmc a few feet from the 49:1 transformer will allow the coax between 49:1 transformer and cmc to act as a ground radial. And this is without a ground rod? Sorry I’m just clarifying that.
  7. WA7ARK

    WA7ARK Ham Member QRZ Page


    I have posted this before, but it is worth repeating. Suppose we have a free-space EFHW wire antenna that looks like this:
    There are two lengths M and L that cause the wire to act as a resonant dipole (M + L + ~0.5wl at the design frequency) at the same time that the feedpoint impedance is close to 2450 + j0 Ohms (which is amenable to being driven through a transformer with 1:7 turns ratio). This is really just a dipole with the feedpoint offset more than what we would call a OffCenterFed dipole (OCFDs are typically driven through a 1:2 turns ratio, 50:200 Ohms ratio transformer, where L/(L+M) is about 0.2).

    Initially, I create the three-wire model above, (wire 2 is just a short place holder for the source) and use the optimizer in AutoEz to find values for M and L that meets the objectives, above. Using 14.175MHz as a design frequency, #14 awg bare Cu wire, if M=32.77ft and L=3.9ft, we have a resonant dipole with the current distribution as shown by the magenta line, and a feedpoint impedance of 2444 - j20 Ohm, for an swr(2450) = 1.008.

    I know from previous experience that M is a critical length, but that L can vary quite a bit without changing the current or the feedpoint impedance much. Here is a sweep of L from 2 to 40ft, keeping the frequency constant at 14.175MHz and M fixed at 32.77ft:

    Here is the swr relative to 2450 Ohms (note, the swr relative to 50 Ohms looks the same shape):

    If L is around 4ft long, then the Swr is really low. This is the ideal length for the wire counterpoise (single radial) case. Something weird happens if L = 33ft. This is the case when instead of using a real wire or radials for L, we rely on CM current on the outside of the coax shield to act as the L wire and the coax shield is near 33ft long. The x-axis in these plots is L in ft.

    This shows the feedpoint impedance which causes the high swr as L varies around 33ft:
    Everything is fine when L is short (~4ft, Z=~2400 + ~j0), but look what happens when L = 33ft! btw, 33 ft is ~1/2 wl at 14.175MHz. Things get bad when L = 33+33=67ft, too (when coax shield length = 1wl or 1.5wl or 2wl...)

    Here is the current along the wire (or along the coax shield) at two values of L = 4ft and 33ft:
    Note that when L=resonant length=33ft, the peak radiation is off the coax shield; not the antenna wire!

    So how do we constrain L to be 4ft, when the coax happens to be near 33ft or near 67ft?
    By using a choke!

    How far from the transformer do we put the choke?
    ~4ft for a 20m EFHW, ~8ft for a 40m EFHW, etc...

    Here is the current distribution if the 4ft counterpoise wire is replaced with 4ea tiny 1.15ft radials:

    So, you can see that EFHW antennas are just an OCFD where the feedpoint is pushed even further off center.
    KO4OCF and G5TM like this.
  8. KI4MYD

    KI4MYD Ham Member QRZ Page


    Forgive me if you've already answered this. If a 40M EFHW is placed as a horizontal or gentle sloper, is it better to install the choke at the ~8Ft point away from the transformer OR at the point where the coax enters the shack? I've repeatedly read both schools of thought but unsure what is the best approach.

  9. W9IQ

    W9IQ Ham Member QRZ Page

    I advocate a choke at both locations.

    - Glenn W9IQ
  10. WA7ARK

    WA7ARK Ham Member QRZ Page

    I see several possibilities assuming that the EFHW is some distance from the house.

    1. No choke(s), no ground rod, either near the transformer nor is the coax shield bonded to a ground rod before the coax enters the house. Only ground is incidental through the house AC wiring. You are likely to get RF-in-the-shack problems on some band(s). You are likely to receive RFI from switching power supplies in your own house.

    2. No choke, transformer connected to single ground rod with a short wire as in post #12. Problems less likely than #1. RFI from inside the house will be reduced. Note that this is not sufficient to comply with recommendations for lightning protection.

    3. Like 2, but add a choke on the coax between the transformer ground rod and station. Choke position not critically important. Much better than #1, and better than #2 for both RF problems as well as suppressing RFI to receiver. Note that this is not sufficient to comply with recommendations for lightning protection. If doing #2 or #3, you are making the antenna more like a Marconi than a Hertz dipole. Read the differences, here:

    4. Nothing connected to the transformer's ground terminal. Transformer grounded only through coax shield. The coax shield is bonded to a ground rod (could be the one under your electric meter) just before the coax enters the house wall. Here a coax choke should be added ~11% of the EFHW wire length from the transformer ( ~7ft if the EFHW wire is 67ft for 40m). Better than #1 or #2. Not as good as #3. Note that this might be sufficient to comply with recommendations for lightning protection. If doing #4, you are making the antenna more like a Hertz dipole than a grounded Marconi.
    Last edited: Jun 10, 2021

Share This Page