[Antennaware] Modelling yagi split DE

[Charlie Ocker via Antennaware]

Hi gang,

I'm trying to verify a YO generated 20m yagi in EZNEC (version 3.x).

I am using tapered tubing.

The first taper on the DE is 36" of 1.25" diameter tubing.  The DE is 
split, and there is 1.0" spacing between the tubing halves.  The DE is 
insulated from the boom.

Here's the deal - when I model the DE as a single wire, say, from -37" 
to +37", and make it 3 segments, and excite the model at the center of 
this wire, the results agree very well with what YO predicts.

But, when I model the DE as two separate wires (1.0" to 37.0"), with the 
actual spacing between, and insert a 2" (from -1.0" to 1.0") wire of #12 
with 1 segment, and connect this wire to the two DE wires, then excite 
the model at the center of this #12 diameter wire, the input impedance 
is way lower than I expect.

What is the accepted practice when modelling split DE's?

Many thanks in advance!

73,
Charlie  N9CO