A good way to check the validity of an EZNEC helix model is to create
the helix by itself, run a wire from one end of the helix to the other
right through the middle of the helix, and put a source in the middle of
the wire. Specify a low enough frequency that the helix will be small in
terms of wavelength. I've found the source reactance to compare quite
favorably with the reactance of the inductance calculated by Reg's
program for the same physical dimensions. The self resonant frequency
comes out quite close, also. The Q should be in the same ballpark,
provided wire loss is included in the model, although radiation will
lower it some in the EZNEC model. (The source resistance with wire loss
set to zero is the radiation resistance. As long as it's much lower than
the resistance with wire loss included, the effect of radiation will be
small. If it's not much lower, reduce the frequency.) As I mentioned
before, EZNEC doesn't model proximity effect (significant only when the
turns are pretty closely spaced) but I don't think Reg's program
includes proximity effect, either.

Roy Lewallen, W7EL