The I-10K manual goes into something similar. Jay notes for these effects in the 1st PDF file below.
With Eznec I select "source" and a window opens and I enter the wire # and the wire end as a % of the segments for the wire, and the software figures out the segment to place the Feed Point. Not sure, but to me your source setup sounds similar. I though you said earlier you had to assign the location for center conductor and the ground to locate your Feed Point with 4Nec2. Do you have two methods of inserting the source?
Before I learned that Eznec produces unreliable results using taper, I made my models with taper, because I felt it was just more accurate to model the antenna to its actual physical dimensions. After I learned better that taper was no way no I started making my wires without any taper. I started estimating my tubing dimensions based on the real dimensions I was able to find. I try to get close to what that might be for my models, but sometimes I have to just guess.
I've read about software that does make an effort to solve this issue, but I use an old fashion way of trying to determine how close my estimate is...trial and error. Sometimes, if I can find some claims data from a well documented antenna manual, and/or info noted elsewhere as well and I don't need to match the antenna...I can the test the model for match and bandwidth and get close enough for government work.
I also hear as long as you get close in you estimate of the wire size distribution along the length of your wire...it just doesn't make much difference in the real world vs. modeling. This is sorta' like I'm finding, in general, adding the physical match to a model...it just don't matter that much. I think your modeling using the matching features in 4Nec2, also show similar results for the performance indicators, gain, and angle. So when we see a model that does not include the physical match or the mathematical equivalent match in a model...the difference is still minimal.
Eznec is not that slick to determine what tubing diameter the model needs...if this is what you are telling me. Like I said above, I've read about software that does do what I think you describe however.
Well DB, Eznec will produce all of the wires needed to produce a circle or a coil, etc, but I have to enter all the correct data noted for the physical antenna into the screen feature for the software using straight wires to try and duplicate a circle or an arc. I also have to describe the location for such devices such as the Sigma3/Vector ring for the radial cone design attaching right at the top of the 3-4 radials on the antenna. Eznec also works similar and automatically for radials, coils, and a helix form.
Personally, I am thankful I was able to figure out the use of these features using Eznec. Looking back I could say I'm amazed I understand these uses at all. I just haven't figured out the matching feature yet...it is like a mental block, with me maybe.
I figure now you're are talking about the optimizing feature using 4Nec2. Eznec will not do that directly, but there is software available on the Internet...what claims to use Eznec data descriptions, and build macro's for automatic optimizing the model.
I still can't get my I-10K model, with the trombone added, working to produce the match you're getting. I also am not getting a good Average Gain value for my model in Free Space either. I can find 50 ohms on the T1 end of the tuner, and it shows a perfect match at resonance, but the AG is still off by too much.
How say you?
