wrote:
Hi, all --

Looking for guidance: I want to do some virtual experimenting
with small transmitting loops before I start welding up a few of
the real things. Questions:

* Is NEC, or one of its children, good for this sort of work?

Yep, providing that if there's more than one turn, the spacing between
turns is at least a wire diameter or two.

* Do I need NEC/4, or will NEC/2 do?

NEC-4 will handle somewhat smaller loops, but double precision or mixed
precision NEC-2 will probably be adequate. NEC-2 is free; NEC-4 is
several hundred dollars.

* Can models handle curved conductors? If I have to enter a
zillion straight segments, OK, but I'd rather not.

Conductors must be straight, but NEC-2 has the provision for
automatically generating segments in a circle if you can figure out how
to do it. EZNEC (my program) fully automates the process.

* Will the model handle lumped impedances, e.g. the matching
capacitor at the far side of the loop?

Yes.

* Does NEC take Q issues (conductor resistance, capacitor
leakage) into account?

Only to the extent that you know them and therefore can tell the program
how large the loss resistance or conductance is.

* Can I usefully model over ground (ideal and real)? I have read
claims that STLs aren't affected by height-above-terrain, and I
rather suspect that it simply hasn't been convenient to hang the
contraptions half a lambda in the air, and the writers are just
whistling past the antenna range.

Yes.

I recommend that you download the free EZNEC demo program from
http://eznec.com. It's adequate to do the modeling you describe,
provided that the loop has only one or two turns. There's no time limit
or restriction on its use. And it includes the full EZNEC manual.

Roy Lewallen, W7EL