Nothing mentioned on the ground radials eliminating CMC or contributing to CMC.
did you happen to model a feed line and mast also?
Wavrider asked DB a very good question.
IMO, if we model this design over Real Earth without a Mast/FL included we might expect to see noticeable differences in results...on adding the mast and/or feed line back to the model later. I think there are lots of recommendations in the science on antennas that support this idea.
Lets respond to your opinion with a fact, one of the best ways to minimize to eliminate CMCs, even better in most cases than using a choke, is to use a full set of 1/4 wavelength radials. I have no disagreement with seeing noticeable changes when adding masts to models after the fact, however, in my experience models with 1/4 wavelength radials don't seem to have this problem. Sure, their are changes, but not significant changes. Have you modeled this antenna with and without a mast to show its changes between the two sets? (NOTE: I didn't this time around, but I have modeled this antenna both with and without a mast many times in the past, and thus have a very good idea of what to expect.)
I think I would have set these models in Free Space...if the idea was not to include the Mast/FL.
Just because their is no mast added you would limit it to freespace?
I disagree with this. The model is not illegitimate just because it is over an earth and doesn't have a mast. It is possible to isolate antennas from masts (and by extension properly choke feed lines to effectively isolate them as well.)
So, do all antennas in the "real world" have a mast that goes straight down to the earth? The one on my roof right now wouldn't match that idea, so if I modeled it that way, any CMC's on the mast would actually not accurately represent what my antenna is doing. To assume all masts (or with antennas like mine, the electrical connection) goes from the feedpoint of the antenna straight down to the earth is actually a dangerous assumption. Mine has to follow the roof line of my house before the ground wire attached to the mast actually connects to a ground rod. If you have common mode currents, that in and of itself, is enough to change the results of what the antenna will do.
I'm not saying masts are useless, but you are making a quite large assumption when you only model a mast going straight down to the earth.
And that is not even factoring in the very real capability of fully isolating the antenna from the mast.
Doing so, might show us more reliable results for comparison of this idea on slanting radials down in this design, and then we could also check the Average Gain results for each model as we go...in order to see how reliable the models might be.
I have not done this as yet, but I may try the idea and see how it compares.
I actually have made a lot of models with and without masts to compare. I can generally tell you just by looking at a model's layout and its height above the earth if the result from adding a mast will make a large difference or not. For example, if you add a mast to my models above that goes straight down to the earth, I would actually expect them all to have slightly more gain than their non-mast counterparts. The reason is the length of the mast that would be used. If I were to add a few feet to their height, however, I would expect a little less gain at lower angles. However, in both cases I would expect said differences to be very small because their are full sets of 1/4 wavelength radials present. Why do I have these expectations? It is because I modeled these antenna designs in the past and compared having and not having said mast.
This said however, I have come to realize that models having part of the mast physically located inside the radial area of the antenna itself, needs to be carefully considered as to the correct approach in modeling in Free Space.
I have no problem with this, it needs to be considered when over an earth as well. Angled radials don't do quite as good a job at minimizing CMC's as horizontal radials do. The reason is the mast and radials affect each others electrical length to some extent. The more you angle said radials down, the more this change to said radials and mast are. The change this creates causes them to be electrically shorter than their physical length.
Maybe adding a small offset hub for the radials is also a good idea that might help to prevent geometry errors.
I love how you are apparently assuming that I didn't use a hub... We have had discussions on this in the past, I am well aware of the issues acute angles create when using the nec2 engine. Adding a hub is one way to fix this known problem with the nec2 engine, however it isn't the only way. I created a thread on it a while back showing a comparison between using a hub and using another method to solve this problem. My models with and without a hub have near identical results. In this case I used the other method. Its the same method that ghz24 told me about to adjust AGT, and I have tested it extensively within the nec2 environment when it comes to correcting for things like acute angles. Prior testing shows that I won't see any significant difference between either of these methods as the results are virtually identical.
Also, when modeling antennas like this, if you don't have a hub, or use some other means of accounting for the error such as what I used, when angling the radials down nec2 actually over-reports gain as AGT goes up. In my models, which have all been corrected to have the same AGT (and that is exactly 1) you will note that as I angled the radials down gain actually went down, not up. It is your models that shows gain going up as you angle the radials down, although I assume you used a hub as you brought its includion up.
Actually when you posted your models, I was curious what the AGT was for each of your models. I let it stand, however, because I had the same thought as what fourstringburn posted later.
Those numbers the modeling geeks give you will hardly be noticed on the other end.
I didn't see the point of us dominating this discussion with modeling like we have multiple threads in the past. Using modeling to show what they can expect to see is one thing, and it is ok if their are differences between our models (we do use very different techniques after all). However, we don't have to turn every thread that one of us posts some models into a discussion between us on modeling...
The DB
