Eznec Question

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by K5WZI, Jul 23, 2019.

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

    UA3TW Ham Member QRZ Page

    Of course. Though it will take some time. Now i can say that in my post #15 the hybrid coupler is made of coax.
    It may be seen in your #16 zoomed picture - in the very center of it the wires 8-16 make the HC. In MMANA I use Bushero bridges as equivalents of quarter wave lines. Will attach the models later, on a week day - I am not home, at the club station now.
  2. UA3TW

    UA3TW Ham Member QRZ Page

    If we want to make 4SQ and we have MMANA as a tool
    we can do that without any other tool.
    I chose an example for a 50 Ohm hybrid coupler.
    (It depends on the chosen hight of the elements and the quarterwave line waveresistance,
    75 or 50 or any other)
    1) The first way is to connect large resistors of 10000 Ohm in series with the sources.
    See the attachment 1. We get adequate radiation pattern and element impedances.
    We do not get correct Gain but the gain is defined by radiation pattern.
    Then we subtract 10000 from the real parts of impedance numerals and get real
    element impedances for the further calculations.
    They are: front element 51.64+j2.36

    Rear element -0.89-j58.78

    Side elements 34.34-j60.8


    When we connect quarter wave lines to the elements we will have

    48.28-j2.2 for the front element through the line

    -0.64+j42.5 for the rear element through the line

    17.62+j31.27 for each of the side elements through the lines.

    We calculated that by the formula for quarter wave lines.
    Then we have to change polarity for the front element and connect it in parallel
    with the rear element. We have to calculate the resulting complex impedance
    by the complex numerals calculator.
    Also connect in parallel the side elements.
    As result we will have two complex numerals.
    What else we have to do before presenting these impedances to the hybrid coupler
    is to eliminate the imaginary parts of them by connecting some reactances in parallel.
    We will have about 54 real ohms for the front-and-rear pair of the elements
    and about 44 real ohms for the side-and-side pair of the elements.
    We can connect that to the ports of the 50-Ohm hybrid coupler at last. The antenna will work.
    The calculations are tedious but routine.

    2) The second way is to connect the sources to the elements through
    Bushero bridges – they are the equivalents of the quarter wave lines.
    See the attachment 2.
    This way we can use voltage sources, we can get the “through the line” impedances right away and we can see that they are very close

    to those of the first way: 49.64-j0.34 for the front

    -0.94+j44.38 for the rear

    17.33+j32.6 for the sides

    Just compare them with the first way impedances. And we have real Gain value to boot.
    The rest of the procedure is the same as in the first way.
    Since I do not have a complex numerals calculator, I made a small MMANA file of a few short wires in parallel to calculate complex numerals in parallel connections. See the attachment 3.
    The formular for the quarterwave line is in the picture.

    Attached Files:

    Last edited: Sep 3, 2019
  3. AC6LA

    AC6LA Ham Member QRZ Page

    Very interesting. And very clever use of MMANA.

    I've found that SimSmith (http://harriman.ddns.net/Smith_Charts.html) is very handy for doing ad hoc complex number calculations. For example, here is how to use the "Plt" block of SimSmith to calculate Z at the input end of an ideal qtr-wave transmission line (Zqtr) given a specified load (Zload); or to calculate the equivalent impedance (Zpar) given two impedances (Z1 and Z2) in parallel.


    SimSmith can also calculate the input Z for realistic (lossy) transmission lines, as can TLDetails (https://ac6la.com/tldetails1.html). With TLDetails, using Belden RG-213 instead of ideal line the results look like this.


    Or one can use EZNEC in a manner similar to the way Nick has used MMANA. For a qrt-wave ideal 50-ohm transmission line with load 51.64+j2.36:


    And for two impedance values in parallel:


    The attached zip contains a sample SimSmith circuit and two sample EZNEC files.

    For even more ways of feeding a 4-sq array see "Hybrid Coupler Feed System" and "Lahlum/Lewallen Feed System", both in the first section of this page: https://ac6la.com/aecollection8.html

    Also see "Phasing Lines for a 4 Square Array" on this page: https://ac6la.com/aetltu.html

    Dan, AC6LA

    Attached Files:

    UA3TW likes this.
  4. UA3TW

    UA3TW Ham Member QRZ Page

    Dan, what I miss in EZNEC is that I cannot open it twice or more. I can open MMANA many times. It IS convenient.
  5. AC6LA

    AC6LA Ham Member QRZ Page

    I agree. Unfortunately that feature is available only in the Pro/2 and Pro/4 program types, not in standard EZNEC or EZNEC+. In the EZNEC Help see topic "Multiple Instances".

    If you have Microsoft Excel (not an Excel clone) you are more than welcome to use the free Demo version of AutoEZ. The main workbook is named "AutoEZ Demo.xls" or "AutoEZ Demo.xlsm" under newer versions of Excel. If you make a copy of that workbook with a different name, perhaps "AutoEZ Demo - Copy.xlsm", you can open both at the same time with Excel. Both will use the same (single) instance of EZNEC.

    The actual name of the copy does not matter, you just can't open two workbooks with exactly the same name in Excel.

    Main AutoEZ web page: https://ac6la.com/autoez.html

    Demo available here: https://ac6la.com/aepurchase.html

    The demo version is limited to 30 segments for calculations.

    Dan, AC6LA
  6. UA3TW

    UA3TW Ham Member QRZ Page

    I hope Dan will see that. Is it possible to connect 2 TLs with capacitor between them using virtual segments? I use real wires now as in the picture.

    Attached Files:

  7. AC6LA

    AC6LA Ham Member QRZ Page

    Normally when a load is to be used with a virtual segment it must be with Ext Conn = Par. That is, the load is placed in parallel with any other insertion objects on the same virtual segment. However, the EZNEC Help describes a way to connect a load (such as a capacitor) in series with other insertion objects.

    Here's a clip from the Help section "A Lap Around Track 5":


    So the trick is to use an L network insertion object with the capacitor in the series branch of the L network and an open circuit as the shunt branch.

    As an example of what Nick wants to do, suppose T1 is between the source and the cap and T2 is between the cap and the feedpoint or other such end-point load. Further suppose that T1 is 10 ft long, T2 is 20 ft long, both are Belden 9258 RG-8X, the cap is 100 pF with a Q of 1000, and the end-point load is an arbitrary 40+j30 ohms. Like so:

    Source --- T1 --- Cap --- T2 --- 40+j30

    With virtual segments connecting everything:


    Under these assumed conditions the input impedance would be approximately 4.7-j4.3 ohms.

    A circuit simulator such as SimSmith can be used to verify the calculations.


    The very small discrepancy, EZNEC 4.675-j4.273 vs SimSmith 4.670-j4.253, is primarily because SimSmith used more digits for the velocity factor of the transmission lines.

    The attached zip contains the above circuit in EZNEC, AutoEZ, and SimSmith formats.

    Dan, AC6LA

    Attached Files:

    UA3TW likes this.
  8. UA3TW

    UA3TW Ham Member QRZ Page

    Many thanks, Dan. Underrstood all. Will build the model soon.
  9. G8HUL

    G8HUL Ham Member QRZ Page

    The checking is pretty much the same between the two, and based on the guidance in the NEC Manual, which is available free and linked to in the 4nec2 help, along with other guidance on the 4nec2 help.
    It is just as easy to get duff models in EzNec as it is in 4NEC2.

  10. UA3TW

    UA3TW Ham Member QRZ Page

    Very difficult to build the model like in my post #26 because I cannot magnify the HC in the center enough as to see the details - the dimentions of the whole antenna does not allow further magnification.
    Is it possible to do something about that? (I do that by temporarily scaling down the elements and the radius of the antenna, but it is not convenient)

Share This Page