Modeling of wires connected at an angle is a problem only with
MININEC-based programs such as AO or ELNEC, not with NEC-2 based
programs like EZNEC or most other modern programs.

The problem with wires connecting at an angle is that MININEC "cuts the
corner" -- it acts like a wire extends between about the middles of the
corner segments, rather than going clear to the corner as they should.
The problem gets less and less as the segment length gets shorter and
shorter near the junction (since the distance "cut" is less), or as the
angle gets more obtuse. There's no magic threshold where the problem
suddenly occurs (same reason). Its seriousness depends entirely on how
serious a small length error is -- it'll just shift the resonance
slightly of a single loop modeled with a moderate number of segments,
but it can seriously alter the pattern of a quad because of the
sensitivity of the parasitic element(s) to small wire length changes.

ELNEC had a way of correcting for this, called "segment length
tapering", where it would automatically convert a wire into a series of
wires with increasing segment lengths. That way you could have a very
short segment at the corner, but not have to use a large number of very
short segments for the whole antenna. That feature is still in EZNEC.
It's no longer needed for this purpose because NEC-2 doesn't have the
problem, but it's occasionally useful at other times.

You can find out more about this phenomenon in "MININEC: The Other Edge
of the Sword", in Feb. 1991 _QST_.

Roy Lewallen, W7EL

art wrote:
My understanding of antenna modeling is that sharp corners or
connecting points should be avoided as they can create errors.
The question I ask is what constitutes a sharp angle ?
I am assuming that it is the capacitance at the connection is the
problem. If so
is the corner of a quad a problem or is a delta feed point on a dipole
a problem ?
I see many instances of quads being modeled and also " V" wire antennas
each without a correcting "gusset" at the corners. So when is an angle
termed to be sharp?
Regards
Art