I took the model from above and made the changes as per your posts. Somehow I completely missed the lengths you had already posted on the first page of this thread...
Anyway, I presumed that if I shorten the upper and lower vertical element lengths that gain will go up, and it did. In addition to shortening said elements to your lengths, I also lengthened the coil from 8 to 9 inches to better match the antenna. One day I may or may not actually put the wires in to model said coil directly, but that is a decision for future me...
These changes netted a gain of nearly 1 dB, which was more than I was expecting, but its there. I should note that the original model under reported gain slightly, and this one over reports gain slightly, so there is that to consider, but even with that I am surprised at this change in results... Anyway, here is the model...
[ATTACH=full]72365[/ATTACH]
This model has a feed point impedance of 52.5 + j 1, which has an SWR of 1.05, so it matches very well. The structure also has a much higher efficiency than the model above, which is where I think most of this gain over that model comes from. It is still 1.1 dB off of the provided G-Max model...
I mentioned that I was using aluminum elements, and not perfect conductors, so to see the difference, removing the "lossy" aluminum elements in favor of "perfect conductors", nets me a whole 0.01 additional dB bringing it up to 5.38 dBi in gain so that isn't the cause of this difference...
Beyond this, I am wondering what is different between their and my models? Like me they seem to be using a lumped inductance, but unlike mine, theirs seems to be much a smaller section that doesn't account for the length of the coil? This can have an effect, but 1.1 dB worth? I highly doubt it.
I guess I can play around with earth settings and see what happens... I'm not sure what else I can to do duplicate their data short of seeing their model data itself and figuring out what else is different?
The DB
