ad: ProAudio-1

Cascaded Hybrid Couplers

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by UA3TW, Nov 11, 2019.

ad: l-rl
ad: Subscribe
ad: l-BCInc
ad: Left-3
ad: L-MFJ
ad: Left-2
ad: abrind-2
  1. UA3TW

    UA3TW Ham Member QRZ Page

    The article on "magic bullet" looks very intriguing but there are too much math and too little physics for a simple electric as myself, I will be hardly able to use it for phasing 6 verticals in a circle for 80 m band.
    Since there is some interest to the thread, I add a picture Fig.4 for 4 different variants of HCs interconnecting and zip EZNEC models for all 4 cases. Also Fig.5 is a partly filled-in table of the TLs for various phase shifts and amplitudes for one of the interconnectoins.
    For example if you want 60 dg phase shift and 0.4 amplitude ratio, see the box in the crossing
    23 meters is the direct line 9.08 meters is the feedback line. Input them in the EZNEC model 90/0 and see the result.
     

    Attached Files:

    Last edited: Nov 12, 2019
  2. AC6LA

    AC6LA Ham Member QRZ Page

    Very clever, Nick! A few minor (very minor) things to consider in order to slightly improve the model accuracy.

    1. I notice you are using ground conductivity of 0.03 S/m. Using EZNEC "Average" ground (0.005 S/m) reduces the gain at 23° elevation by about 2 dBi. But perhaps your ground really is that good.

    2. Adding 5 ohms of simulated ground loss at the base of each vertical element further reduces the gain by about 0.6 dBi.

    3. Adding realistic loss to the transmission lines reduces the gain by about 0.2 dBi. (I used Belden 8213 RG-11 for the 75 and 30 ohm cables and Belden 8267 RG-213 for the 50 ohm cables.)

    4. Using "Aluminum" rather than "Zero" wire loss results in no change to the gain but has a very minor effect on the source impedance.

    With those changes, here's how the 6-element dual hybrid coupler pattern compares to a conventional 4-square using a single hybrid coupler. The focused pattern of the 6-element array produces 1.7 dB greater forward gain at the expense of F/B.

    ComareAzim.gif

    An EZNEC model of the 4-square hybrid coupler used in this example is attached along with a modified version of Nick's model per above suggestions. For more details on the 4-square hybrid coupler model see this page.

    Dan, AC6LA
     

    Attached Files:

    W1VT, WW1WW and UA3TW like this.
  3. UA3TW

    UA3TW Ham Member QRZ Page

    I was going to attach an example for Fig. 5 table but the edit time has expired. So i attach it here.
     

    Attached Files:

  4. UA3TW

    UA3TW Ham Member QRZ Page

    Many thanks, Dan! Now we see - it is worth building, if one has room. Though my purpose was to share the double couplers, not the antenna - it was just an example. And I have other curious ways of feeding phased arrays, and 6 verticals as well. Do not know if it is worth sharing - very few people can afford building so big systems nowadays, seems to be pure academic interest.
    Forgot to mention - I took good ground for the sake of comparison, because it was an argument on a russian forum where some models were compared and they took that ground in the models. So I had to switch for their ground and use it since then.
     
    Last edited: Nov 12, 2019
  5. W1VT

    W1VT Ham Member QRZ Page

    I don't have the room, but I do have unusually good ground. I have a base layer of heavy clay and a topsoil layer of 8" to 12" with plenty of organic matter on top of that. With time release fertilizer to grow roses. There is water ponding in my back yard when it rains. This year it was wet all Spring and the main back yard rose bed wasn't really workable for planting until July.
     
  6. AC6LA

    AC6LA Ham Member QRZ Page

    Nick, in your first post you said,
    In your model "6DIR_2source.EZ" you have a 50 µH load (+j1147 ohms at 3.65 MHz) at the base of wire 9 (front element) and a 100 pF load (-j436 ohms at 3.65 MHz) at the base of wire 10 (rear element).

    Can you explain to me the reasoning for those two loads? Why not just simulate an open circuit at the element feedpoint with either a very large resistance load or by putting a small air gap between the base of the element and ground?

    Thanks.

    Dan, AC6LA
     
  7. UA3TW

    UA3TW Ham Member QRZ Page

    Yes, Dan, so it happened. Because the initial file was without front and rear elements at all, and when I added them, the nulls of the rear lobe became shallow, so I decided to make the added elements play along to get better parameters. I thought this was not important for the main explanation.
    When I attached the second model , with the single source and dual HC, I left omitted the front and rear elements.
    For curious people I desided to attach an example of the array which in known as W0UN system in ON4UN book, and I fed it from a triple hybrid for a change. It gives higher impedance of the feeding harness and the impedances of HCs in this example are 27 Ohm.
    It is just an example, because in my opinion a better way of feeding this system is single conventional HC of 10 Ohm
     

    Attached Files:

    Last edited: Nov 13, 2019
  8. UA3TW

    UA3TW Ham Member QRZ Page

    Dan, I studied your model - it is very difficult in EZNEC to see the details. I guess the red framed loads are compensators of reactance, connected in parallel?
    And fogot to mention - in previous post the hybrid couplers are not transformes, but made of 1/8W coaxes and capacitors.
     

    Attached Files:

  9. AC6LA

    AC6LA Ham Member QRZ Page

    Ah, now I understand. Thank you.

    Not related: For folks who do not have the 5th edition of the ON4UN book and would like to know more about the hybrid coupler model as developed by W7EL, see Demystifying the hybrid coupler.

    Dan, AC6LA
     
  10. AC6LA

    AC6LA Ham Member QRZ Page

    It is easier to understand the various loads in the AutoEZ version of the model as opposed to the EZNEC version. See model "Vertical 4-sq Hybrid sweep.weq" available for download here and as documented here. (Can be viewed using the free AutoEZ Demo program.) In that model the load table looks like this:

    Loads.png

    The two loads in question (rows 54 and 58) are shown as variables BG (Port 2 shunt) and AO (Shunt inductance of phase reversal xfmr) in this diagram:

    Feed System.png

    Their use is as described in sections 3.4.6.6 and 3.4.6.5.7 of Chapter 11 in the ON4UN book.

    Dan, AC6LA
     
    WW1WW and UA3TW like this.

Share This Page

ad: vanity