[TowerTalk] Sloper & Tower

[Gene Bigham]

Merry Christmas to all, and to all, antennas at new heights!

I have been playing with EZNEC, which is dangerous since I really do not know what I am doing.  But here is what I have gotten so far.  I have a two section aluma tower 50' crank up that is nested at about the 30 foot level, a rotator on top, a 10 foot mast on which 2 feet above the top of the rotator is mounted a Cushcraft A3S, then at the top of the mast is a Cushcraft vhf/uhf 4 element beam.  At the 23 foot level up the side of the tower I ran a wire down to X= -30 and 5 foot off the ground, there it turns and runs an additional 50 feet along fence toward X= 20.  I actually have a dip in the swr on 160 of 2:1 with this arrangement.  The tower has 10 radials on its base fanned out in a semicircle toward the back fence line, laying just on top the ground, which I am not able to model due to the 20 segment limitation on the demo EZNEC.
When I model this on EZNEC with the source right where the wire attaches to the tower (coax center to the wire, braid bonded to tower), I get the 3D pattern EZNEC makes, BUT what is interesting is raising the attachment point on the tower lowers the gain of the antenna structure, and lowering this point raises the gain on the patterns.
Lowering it makes the pattern more a donut with gain around 13-19dbi, raising the feed point on the tower turns the donut into a bottom flattened sphere and the top of this sphere begins to grow upwards as the feed point is raised lowering the gain drastically.  Looking at the currents on the model shows the entire tower as a vertical with current highest at the base and tapering up the entire tower structure just like you would expect on a vertical, there is also some current distributed on the wire.
I would have expected the opposite results.
Am I doing something wrong?
Thanks
Gene Bigham