ad: AlphaRF-1

End Fed Half Wave Sloper fed from existing tower

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by WA7ARK, Feb 21, 2021.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: abrind-2
ad: HRDLLC-2
ad: MessiPaoloni-1
ad: Left-3
ad: L-MFJ
ad: L-Geochron
  1. WA7ARK

    WA7ARK Ham Member QRZ Page

    My friends and I have been getting on 60m on weekday afternoons. I use my 160m one-wavelength loop on its third harmonic and that works very well.

    One of my group, Glen, has an existing tower which has a HF tri-band Yagi at 54ft agl, and a VHF array at 59ft agl. The tower is grounded at its base. Glen would like to feed a down-sloping wire against the tower, and asked me for some modeling help before building it. He has been getting on with a random antenna and tuner, but would like something better.

    I had previously done some modeling of "slopers" that are close to 1/4wl long. I found that if fed between the wire and the tower, that induces large currents into the tower itself, some of that current flows into the dirt under the tower, generating losses. Here is a different approach using a Hertz half-wave dipole wire, fed against the tower, where the currents in the tower are minimized.

    I came up with a couple of innovations during this design that I have not seen in the ham literature before. I will point those out later in this posting.

    First, I created a model of Glen's tower with a proxy for the existing HF antennas on the tower. My first cut was to run the wire out from near the tower top (just below the HF Yagi), sloping down to fit Glen's intended routing. If the feedpoint is placed in the wire where it connects to the tower, a ~88ft wire resonates at 5.37MHz, but the feedpoint impedance is very high, like 4000 + j0 Ohms...

    That is a bit of a stretch for a RF transformer, so it occurred to me to first use a Q-section (37ft) of 75 Ohm coax (RG6, VF=0.82) to step 50 Ohms up to 100 Ohms. Then the transformer only has to step 100 Ohms to 4000 Ohms, which makes for a net lower turns ratio: {sqrt(4000/100) = sqrt(40) = 1 : 6.3 }. Doing it without the Q-section would require a 1:9 turns ratio. I have not seen combining a coax Q-section with a transformer before in the context of EFHW antennas...

    Since there is a power limit of 100W on 60m, the RF voltage across the 1:6 transformer secondary is less than 900V even as the current through the secondary winding is only about 150mA. The peak current at the center of the wire is ~1.41A, same as it would be if the wire was center-fed.

    As I was optimizing the design, I noticed something that brings higher efficiency to this antenna. I made the height of the feedpoint (transformer height and where the wire leaves the tower) variable. This effects how current along the tower divides above and below the feedpoint. Note that the current induced in the vertical tower sections comes from two sources; one part comes from the transformer (call it the counterpoise current, about 150mA) and the other part comes from mutual induction between the wire and the tower.

    As the feedpoint position (transformer height) is moved up and down the tower, I found a height that minimizes the current that flows down the lower part of the tower and into the ground rods. Any current that flows into the ground rod resistance is just lost as heat, so the goal is to minimize that current.

    At a critical height of 44ft (0.95% of 0.25wl), the current in the lower tower section is minimized while the current in the tower above where the transformer is connected to the tower is maximized, meaning that the ground losses are reduced, and all of the current that contributes to actually creating radiation remains high up the tower...

    At 60m, that happens only when the height of the transformer is about 95% of 1/4wl. Since the bottom of the tower is grounded, the impedance looking downward along the tower from the transformer is high only when that part of the tower is electrically ~0.25wl long. Change that length either way from 44ft, and the lower current increases.

    That high impedance looking down forces the "counterpoise" current the other direction (upward) along the part of the tower above where the transformer is. This avoids RF current flow into the ground rods, minimizing loss, and improving the radiation pattern.

    This is the other important "innovation" that I have never seen published before...

    Here is a schematic of what I describe, above:
    schematic.png
    This shows the critical 44ft vertical dimension between the earth and the transformer. Notice the currents (magenta lines) and where they occur. Almost a perfect dipole half-Sinosoid along the sloping wire, and almost none in the lower tower section, and a bit in the tower above the transformer...

    Some modeling notes: Note that it is not necessary to model all the details of the HF and VHF arrays. The takeaway is that the 44ft distance from earth to transformer suppresses current along the lower part of the tower, and forces that current upward instead.

    To monitor the current into the simulated ground and to estimate ground losses, I added a 50 Ohm resistor just above the "radials", which are there only to act as a "current sink" in lieu of using the Mininec modeling method of making a Galvanic connection directly to earth. Here is EzNec's calculation of current into the ground (small) and power dissipated in the ground rod (also small):
    Loss.png
    Conclusion: the 44ft FP height trick is doing its job... Using other than 44ft increases the loss.

    Here is the Swr50 looking into the 75 Ohm coax Q-section:
    60mSWR.gif
    Smith chart impedances:
    Smith.gif

    R and jXs:
    Z&SWR.png

    Patterns:
    Patterns.gif

    On a lark, I swept swr over the HF spectrum to see if this is useful on any other ham bands:
    WideSWR.png
    Looks like it might be usable on 30m, 17m, 15m and parts of 10m with a wide range tuner. This was not part of the design, just interesting....
     
    Last edited: Feb 21, 2021
    M0AGP likes this.
  2. AI3V

    AI3V Ham Member QRZ Page

    Well at least you are staring to model (albeit with a awful lot of guesses) a antenna comprised of 2 halves.

    The "q section" and lossy transformer will help the vswr, with the usual s-unit of rf turning into heat.

    He would of course be far better off with a (50 ohm) coax fed quarter wavelength of wire fed agaist the top of the tower (for the added height above ground of the horizontal wire), of course the antennas on top of the tower you are unable to model will more or less degrade the radiaion of the horizontal wire if you get too close.

    Rege

    P.S. after a hundred years or so of people feeding a wire against a tower, there is a reason nobody published your "innovations" ;)
     
  3. WA7ARK

    WA7ARK Ham Member QRZ Page

    I have always modeled EFHW antennas exactly as they are; ~5% of the wireto the left of the transformer, ~95% of the wire to the right of the transformer. Only you thinks it requires more than that to the left of the transformer... In this case, the 5% is connected to the tower, and you do not want it to become 50%!

    Total B.S! Qsection loss at 5MHz is a fraction of a dB. Transformer loss is ~0.5dB (measured).

    The reason I used a simplified model to represent the antennas on top of the tower is that it doesn't matter if you model them in detail or not. If you knew anything about modeling, you would know that. The reason that the half wave wire is better than a quarterwave wire is because the behavior of the EFHW is totally predicable while the short one is a total crap shoot, which you would know if you ever built one, like I have!
     
  4. AI3V

    AI3V Ham Member QRZ Page

    Since when is a quarter wavelength wire a "crapshoot"?

    I must give credit where its due, you have a knack to make a foolproof setup into something that appears to require exceptional detail in construction.

    Like I said, hams have been feeding quarter wavelength wires against towers "forever"

    They work, without the necessity of a "q section", "transformer" , and a critical height of the feedpoint.

    Nor do they require several assumptions and computer models.

    Rege
     
  5. WA7ARK

    WA7ARK Ham Member QRZ Page

    Always!

    You cannot predict how long to make the wire (you have to determine it by cut and try because the total dipole length depends on the tower height).

    The radiation pattern sucks because the V created by the down sloping wire against the tower is typically more acute than 45 degrees.

    The efficiency sucks because the V created by the down sloping wire against the tower is typically more acute than 45 degrees so that the radiation from the two halves of the "inverted V dipole" begin to cancel each other.

    The efficiency sucks because one leg of the "inverted V dipole" is stuck in the ground (at the tower base).

    With ~1/4 wl sloper coming down off a tower, you cannot predict what the feedpoint impedance will be (its never close to 50 Ohms because of the acute angle and unknown tower height, so matching is required).

    OTOH, the Hertz end-fed half-wave sloper or horizontal wire is totally predictable and avoids the problems outlined above.
     
  6. AI3V

    AI3V Ham Member QRZ Page

    What property makes a Hertz antenna any more predictable than a Marconi?

    That's right, there is none.

    In my day job I maintain 2000 or so SCADA links.

    Every single one is a Marconi fed against a random ground plane many wavelengths large.

    The performance is so predictable we do not bother to measure VSWR as it os a complete non issue.

    And, there is not a "common mode choke" to be found anywhere in the millions and millions of idendical systems installed worldwide.

    Your "half wave endfedz" on the other hand...

    Rege
     
  7. N7WR

    N7WR Premium Subscriber QRZ Page

    I agree totally. I have used multiple quarter wave slopers fed at the top of a tower (selectable by a remote switch) with the tower as the "other half" for decades. As you said no "q section or transformer". Used those slopers to work DXCC on 40 and 75. Never modeled them at all....just trimmed for lowest SWR in the 75 meter "DX window" and in the center of the 40 meter phone band. Sometimes I think we "over think" or try to over design something that works just fine without the academic exercise.
     
    KI4POT likes this.
  8. W4BCE

    W4BCE Ham Member QRZ Page

    I also have used a sloper on 20 meters for 15 years against the tower. And did the same- just trimmed and no loops or transformers involved. Worked for me.
     
  9. K4SAV

    K4SAV Ham Member QRZ Page

    A quarter wave wire fed against a tower is known as a half sloper. W8JI used to call these sloppers because they are very unpredictable since most people just put them up with no idea of how they work. Sometime they turn out OK, and other times they are dogs. Most people never know the difference unless they have another antenna for comparison.

    It has been known for ages that for good performance you need most of the current confined to the wire and the top part of the tower. You need minimum current going down the tower. If you have most of the current going to the wire and the bottom half of the tower, then you effectively have an inverted vee with a very acute angle and one end connected to the dirt, You should be able to imagine how bad that will be.

    An end fed half wave (extreme OCF) is the same principle. That's the exercise that WA7ARK went thru and showed the same conclusions.

    Jerry, K4SAV
     
  10. AI3V

    AI3V Ham Member QRZ Page

    Except there was nothing "end fed" about this wire.

    Rege
     

Share This Page