Discussion in 'Antennas, Feedlines, Towers & Rotors' started by K5WZI, Jul 23, 2019.
oh, I understand that. Personally I ptrfer it, but that's s personal choice.
ON4UN did a nice presentation on Hybrid Couplers for Dayton several years ago, available here (22.2 MB). There's also lots of information in Chapter 11 of the 5th edition of Low-Band DXing, including supplemental files on the CD.
Also see the first section of this AutoEZ page (scroll down to the "Hybrid Coupler Feed System" section).
And I was quite pleased with the v5.0+ upgrade, when I got it. The extra segments made complex antennas and vehicle modeling easier, not to mention the added features. Planning to get the v6+ and AutoEZ next!
I did get the software last Friday. Will study it carefully.
Both versions are on the screen at once. v.6 gives me warnings. v.4 gives no warnings. How serious are the warnings i wonder. Do they influence the result essetially?
The model is attached.
When first viewed Nick's model looks like this:
A qtr-wave vertical in the center of the array with 8 satellite qtr-wave verticals, each spaced 34m (~0.41 wl) from the center. So far so good.
Zooming in reveals this:
Now I'll admit I'm not the sharpest tack in the box but I just can't make sense of that. So one approach to simplify things would be to take advantage of two features that were new as of EZNEC v5, virtual segments (virtual wires) and parallel-connected loads.
Other than the 9 qtr-wave verticals all the other wires of the model are very short and seem to be used only as connection points for the 57 (57!) transmission lines and/or as a way to connect a load in parallel with the end of a transmission line. So all (most?) of those short wires could be replaced with virtual segments. Doing that would eliminate many of the segmentation check warnings.
Here's the "Virtual Segments" topic in the EZNEC Help.
Another option would be to change all the short 1-segment wires to 3 segments, then change all the insertion objects to be placed at 50% on those 3-segment wires. The EZNEC "Group Modify" option makes that relatively painless. You'll then see different warnings about segment lengths being too short. In this case those warnings can probably be ignored.
Another great tip Dan. I'll admit I've always created short wire segments for things like transmission line, transformer, or L network terminating points. Virtual segments is a much slicker approach to those things.
Thank you very much, Dan for your thorough analysis and your help. Excuse the bulky example I chose - it was the last I worked on. The reason for so many wires and TLs is that the model contains all the means of "rotating"
the RP of the antenna. This way I can see all the relays and the additional TLs. Here another model is attached where the antenna is switched to the 45 dg position (all 8 positions have their models).
And I will go on studying the EZNEC+ v.6 and applying it to my models, using virtual segments and transformer function for the hybrid coupler building.
I neglected to address your original question about why you get segmentation check errors with v6 but not with v4. From the v6 EZNEC Help:
The limitation in the NEC engines has always been there. EZNEC v6 added the check to highlight that limitation.
And here's another tip for EZNEC+ users. You have two choices for the calculating engine, NEC-2 (mixed precision) and NEC-2D (double precision). To make the double precision engine the default click Options > Calculating Engine > EZCalcD (NEC-2D). Then, important, to make that choice permanent click Options > Save As Default [last choice in list]. That way the next time you start EZNEC you won't have to change the engine choice again. With today's computers there's really no good reason to use the mixed precision engine.
With a multi-directional vertical array it's sometimes nice to see how the azimuth patterns overlap when switching the direction of the array. Using Nick's "0dg" and "45dg" model variations, and taking advantage of EZNEC's "Save Trace As" and "Add Trace" options on the 2D plot window, the depth of the null at 22.5 deg is seen to be 2.13 dB.
Rather than have 8 different models another option is to just rotate all the wires of the array by 45 deg about the Z axis. On the EZNEC Wires window choose Wires > Rotate.
Here's a similar example from AutoEZ using a Hi-Z 8-circle receiving array. In this model the array direction is set using a variable.
See section "Modeling 8-Circle Arrays" on this AutoEZ page.