EFHW for 60/30/15 Meters

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by W9KEY, Sep 30, 2020.

Tags:
ad: L-HROutlet
ad: l-rl
ad: MessiPaoloni-1
ad: Left-3
ad: L-Geochron
ad: abrind-2
ad: Left-2
ad: HRDLLC-2
ad: L-MFJ
  1. W9KEY

    W9KEY XML Subscriber QRZ Page

    I'm building a 49:1 EFHW for 60/30/15m in an inverted L configuration with 12ga THHN wire - about 30' vertical and 60' horizontal (40' AGL). Am using a CMC about 8' from the feedpoint.

    It works, but am having a problem getting the SWR to be minimum at 5.37, 10.125 and 21.2 MHz.
     
    KF8RM likes this.
  2. KK4OBI

    KK4OBI Ham Member QRZ Page

    There is a fundamental problem with (near) end-fed antennas, the harmonics are never exact harmonics of the fundamental frequency. In your case the 60 meter fundamental is about 0.67 meters short when the 30 and 15 meter harmonics are correct.

    The reason for this is commonly referred to as “end effect”. The mechanical solution is to switch in the needed length when operating on 60 meters. The electrical solution used by commercial antennas is to add a small inductance 4 or 5 feet from the feed point and a capacitance at the center or in the transformer. This aligns the harmonics to the fundamental.
     
    YO9FMB, WB2UAQ and WB5YUZ like this.
  3. WA7ARK

    WA7ARK Ham Member QRZ Page

    To illustrate this, I start with a simple near-end-fed dipole that is a half wavelength on 60m that is a proxy for W9KEY's eventual EFHW.
    To keep it simple, I model it as three co-linear wires in free space. So that it can be fed from a source impedance of 2450 Ohms (the eventual 1:7 turns ratio transformer), I find the lengths of K and L that makes the antenna resonant at 5.37MHz and has Z=2450 +- j0
    upload_2020-9-30_6-56-59.png

    The optimizer in AutoEz finds that if K=7.4ft and L=86.14ft, then Z=2453+j0.8 for an SWR2450 = 1.001. Those of you who have read my previous posts on this topic will remember that K's length is not particularly critical, so I show what happens to the swr if L remains fixed at L=86.14 ft but K is varied from 2ft to 20ft. Note that when K>5 ft, the SWR2450 remains less than 1.5
    upload_2020-9-30_7-35-38.png

    K=7 ft and L= 86.14 ft is a good prototype for a 60m EFHW. Let us see what it does at its second and fourth harmonic:
    upload_2020-9-30_7-27-33.png

    It does good at 60m, but it is acting as though the wire is too short both at 30m and 15m. If we lengthen the wire to better fit 30/15m, then it will be too long at 60m. There is a simple solution that I will show in the next post in this forum thread...
     
    WB2UAQ and M0AGP like this.
  4. K8JD

    K8JD Ham Member QRZ Page

    I would just make a 3 wire aray with spacers , each a QW for one of your bands. No fancy Xfmr BalUns or anything needed, just have a radial system at the base. ( Your inv L is just a bent over vertical).
     
    N8ZL likes this.
  5. WA7ARK

    WA7ARK Ham Member QRZ Page

    ... I do have a real life, so it will take several posts over time to complete this discussion...

    Dick, (the OBI) suggested a couple of tricks that might help. First thing I would do is instead of resonating the antenna at 5.37MHz (and let 30m and 15m fall where they may), is to resonate the antenna at 10.15MHz, and see what that does:
    upload_2020-9-30_9-24-52.png

    Now we are too long at 60m and 15m, but just right at 30m...

    A trick to shorten the antenna on the fundamental frequency relative to its harmonics is to place a capacitor in series with the main wire. On 60m, there is a current maximum at the dipole center, while on 30m and 15m, current minimums occur there. If you place the capacitor at the dipole"center", then it will effect 60m, but have almost no effect on 30m and 15m, so I will turn the AutoEz optimizer loose on the problem.

    I task the optimizer to find capacitor value (X pF), capacitor position (N, initially set to L/2 ft), and a new wire length (L ft) that simultaneously minimizes SWR2450 at 5.35, and 21.0MHz. K is not critical, so I leave it at 8 ft for the optimization. The prototype antenna now looks like this:
    upload_2020-9-30_10-45-21.png

    The optimizer finds that if L=92.5ft, N=45.4ft, and X=254pF, we get the following SWR2450 minimums:
    upload_2020-9-30_12-17-2.png

    That is about the best I can do by adding a single compensator capacitor at the optimum location at the optimum wire length..

    So this demonstrates what can be done to make the prototype antenna have a usable Swr on three bands... In a future post, I will apply this compensation method to W9KEY's proposed configuration of his antenna.
     
    M0AGP and 2E0CIT like this.
  6. K9UR

    K9UR Subscriber QRZ Page

    Go dog go ! Was just today playing with a 30M dipole and thinking about some options for 15 and 60.
     
  7. WA7ARK

    WA7ARK Ham Member QRZ Page

    W9KEY emailed a sketch of his existing antenna. I used his dimensions to create the following inverted-L EFHW model, which matched his measured data quite well. I added the capacitor, making the total length of the wire, the capacitor position, and the capacitor value the variables of the optimization. Because of the bend, and proximity to earth, I was fully expecting different numbers than for the free-space, straight line, uninsulated wire model, in post #5, but actually not much changed except the capacitor value...

    So after adding the insulation to the wire ( 0.0315 inches of PVC/Nylon for which I used a combined dielectric constant of 3.7) here are the results of the final optimization of W9KEY's 60/30/15m EFHW Inverted-L. The optimized antenna will look like this:
    upload_2020-10-1_15-26-27.png

    The optimization converged close to the free-space values, but I got a net lower SWR50 at the input side of the transformer by changing its turns ratio from 1:7 to 1:6 (Z ratio of 50:1800 Ohms). The best value for the capacitor is 200pF, placed 46.4ft from the transformer (6.4ft from the bend).

    Here are the predicted Swr50 values:
    upload_2020-10-1_15-20-57.png

    Here are the patterns:
    upload_2020-9-30_17-11-18.png upload_2020-9-30_17-7-41.png

    There is a high RF voltage across the choke during transmit, so hopefully the choke will be placed so it will not be touched. It is good that the choke is 2ft agl; if it were lower, losses would increase because of coupling to the lossy earth. It is ok to earth-ground the coax anywhere between the choke and the rig for static drain/lightning protection.

    You can parallel two 100pF capacitors to get 200pF. The capacitor has less than 200Vpp across it on 5.3MHz @ 100W. The RF current through it is about 0.9Arms.

    You should be able to modify your existing (50:2450 Ohm) transformer to 50:1800 Ohms. If the primary has two turns, remove two secondary turns; if the primary has three turns, remove three secondary turns.

    The weak point of the simulation is likely how I modeled the THHN wire insulation. For any final tweaking, I would cut/add a bit of wire at the distal end to achieve the lowest SWR50 at 21.0 MHz.
     
    W9KEY, N7EKU, WB2UAQ and 2 others like this.
  8. WA7ARK

    WA7ARK Ham Member QRZ Page

    No, it is not a Marconi requiring a ground-plane/radials/counterpoise. It is a Hertz dipole, that works entirely independently from the earth below. The part of W9KEY's antenna that does most of the radiating is the horizontal part.
     
  9. W9KEY

    W9KEY XML Subscriber QRZ Page

    Wow Mike - thanks for all the analysis!!! While it continues to rain here, I'll build a PVC tube housing with thimbles to add the in-line capacitance, measure the effect, then work on the suggested transformer alteration -- although a substantial amount of electrical grade RTV is going to be pretty unhappy about that!

    Checking my inventory, I've got:
    • 220 pF 3KV 10% N2500, and
    • 47 pF 6KV 20% R3L (N2200)
    While the cap only sees low voltage (220 Vpp) at 100w on 60m, what happens at full power on 15 meters?
    With above stock parts I can do either 3KV or 6KV (but 188 pF).

    The RF Choke is 12 turns of RG-400 on an FT-240-31 core in a big-box store Carlon 4x4x2" plastic box with SO-239's on opposite ends. While they are properly weatherproofed, the SO-239 flanges are exposed. Does your simulation predict high voltage across the choke? I can add some protection, or build another one without the box using PL-259's that can be fully insulated.

    I could also raise the choke higher AGL, but that will increase the horizontal to vertical ratio of the wire - which sounds like a terrible idea given all the above work!
     
  10. WA7ARK

    WA7ARK Ham Member QRZ Page

    I compare four cases where I use what you have, i.e. 220pF instead of 200pF, and the existing 1:7 transformer instead of hacking it to 1:6. Three of the cases use the wire length and transformer postition as shown in my pencil drawing. The fourth case utilizes 220pF and the 1:7 transformer, but I re-optimize only the wire length and cap position along the wire.
    upload_2020-10-2_7-12-46.png
    The re-optimized wire length is now 91.5 ft (0.1ft shorter), but the cap position is now 40.3ft from the transformer (10.3ft from the corner). This gives you an idea of how sensitive the solutions are to changing one thing at a time....

    Here is what happens to the capacitor on the three bands:
    upload_2020-10-2_7-25-14.png upload_2020-10-2_7-26-22.png upload_2020-10-2_7-27-13.png

    The high voltage would be on the top PL259 skirt; the bottom one would be cold. I would just tape it as you would for rain protection.

    PS: I just noticed that I should add one more situation: optimized with 1:6 transformer and 220pF capacitor. I'll add that when I get some time.

    Here it is: Wire Length =91.4ft and Cap position 39.6ft from the transformer:
    upload_2020-10-2_7-36-31.png
     
    Last edited: Oct 2, 2020

Share This Page