Monday, November 19, 2012

Antenna Modeling – The Practical Use

What do these software products all have in common:  EZNec, cocoaNEC, NEC 2, and NEC-Win?

  1. They are intimidating, obscure, and un-intelligible software programs somehow related to Amateur Radio.
  2. They each provide a means to use mathematical sets of formulas to simulate antennas in the "real world."
  3. Electrical Engineers, PhD's, Nerds, and the socially unacceptable are the only humans who know the "secret and arcane" art of antenna modeling using these software tools.
  4. These are very expensive software products, well out of the reach of a typical Ham.

No, the answer is not "3"; the correct answer is "2"!  Yes, really - I'm not kidding…

One of my favorite things to do in Amateur Radio is trying to model the "perfect" antenna.  I've "invested" (or wasted depending upon whom you ask!) a lot of time with EZNec (on Windows) and cocoaNEC (on Mac) and I really enjoy them a lot.  I've often asked myself questions like:
  • I wonder what that dipole would look like in my back yard?
  • What does a 3 element, 2-meter yagi do at 8' above the ground?
  • How does a j-pole really radiate?
  • Can I make my own Buckmaster (and save a TON of $$)?
Recently, I was able to put my "investment" (of time) to good use.  I'd decided that I wanted to put up a 40m dipole and remove the doublet that was in its place.  Sure the doublet worked OK, but I'd recently discovered just how inefficient that antenna radiates on it non-harmonic frequencies.  It doesn't really matter what "it" is, but I like things that work efficiently and my doublet surely didn't.

I had to opportunity to "design" my own dipole.  How tricky can that be you ask?  As tricky as you want to make it.  Since I knew how to use EZNec and cocoaNEC, I didn't just want a flattop dipole with it's 70+ Ohm feedpoint and 1.8+ SWR across the frequency band for 40m (see Figure 1).  I used cocoaNEC to model a dipole, sized for 40m, about 45’ above the ground – which was very close to my QTH’s backyard setup – to see what a regular dipole would look like.

Figure 1 - 40m Dipole "Flat Top"

Since the 40m antenna is lower to the ground than the optimal height, I started changing my model by changing the legs of the dipole.  I started off by slightly raising the legs by 10 degrees at the insulator and saw the feedpoint resistance drop a little.  Eventually, I managed to find a nice combination of dipole leg “angle”, based upon the height over ground.  I was also able to adjust the length of the legs to get the lowest SWR across the lower part of the 40m band.  The SWR model is show in Figure 2.

Figure 2 - 40m "V" Dipole

During November of 2012, I managed to find some time to hang my 40m dipole based upon the calculations.  I hung the wires and balun, trimmed, re-hung, trimmed, re-hung, etc. for a few hours.  When I was finished, I used my analyzer to see just how close I was.  Figure 3 shows the graph based upon the numbers my analyzer showed.  The curve is about as close to a match (compare Figure 3 to Figure 2) as I could ever expect!  I was so pleased – it has been a while since a project that I’m working on has turned out so well!

Figure 3 - Actual Analyzer Measurements at my QTH

As I’m writing this, I still can’t believe just how accurate the EZNec and cocoaNEC programs are.  I know it’s not rocket science, but I was able to design an antenna based on my requirements, build the antenna, and finally, measure to see that it matches the antenna design.

Once all the measurements were complete, I got right on the air.  The ARRL was having a Sweepstakes competition and the contacts were rolling in!  Perhaps band conditions were great or my antenna worked even better than I’d dreamed ☺  Either way, there were many contacts to be made!  It was a great way to end a weekend!

If you think that you’d be interested in trying to model your own antenna, RARC will be offering a 3 or 4 week “Introduction to EZNec Antenna Modeling” class in the late winter/early spring.

73 de KJ4WLH


  1. Your final SWR curve is remarably similar to the predicted curve and apparently it works great. Now, you modeled a length that you hung but it needed to be trimmed-rehung-trimmed-rehung etc. How did the final length compare to the modeled length? How did those lengths compare to the rule of thumb 468/freq formula?

  2. The reason I measured, dropped the antenna, cut, raised the antenna, measured, and repeat was that I was just EXTRA cautious of cutting the wire too short :) I’ve cut dipoles too short before and it’s just a maddening thing - know that you just measured incorrectly, cut and then your SWR curve is too high for your target frequency range.

    As far as cutting, I got within a couple of inches of my antenna model target for the dipole legs. I probably should have cut another few inches off just to make it even “closer” to my goal - but I just didn’t have it in me to screw up what was looking pretty good. I know it’s math and science, but somehow I “knew” that if I did “just one more cut”, it would mess it up... So, I just left well enough alone.

    The 468/f didn’t really work for me due to the changes in the antenna model - I wasn't trying to model a flat-top dipole, which is where that formula really kinda works out well. Since the legs were at an angle relative to a flat-top dipole, they were a bit shorter (not much though) than if it was horizontal. I must say that I’ve done more math (geometry) than I ever thought I’d use in school :)