At 06:08 PM 12/24/2006, Bob Nielsen wrote:
>Of course, NEC4 isn't anywhere close to "free".
True, but the original poster was looking for a comparison. And that
comparison (NEC4 wasn't used in the previous comparison) might
justify forking out the few hundred bucks for the NEC4 engine. Could
be worse.. other codes like HFSS are even pricier, but if you need
it, you need it.
MMANA also (I understand from reading.. I haven't used it) uses the
MININEC modeling code, not the NEC codes.
There are some significant differences, particularly for antennas
with "non thin wire" elements and ones that are "close to the ground".
FWIW, I use 4nec2 as the front/back end, and both NEC2 and NEC4 back
ends, although I've got a fair amount of one-off stuff for specific
applications in VB, C, and Matlab that calls the NEC4 (or NEC2) backend.
>On Dec 24, 2006, at 2:28 PM, Jim Lux wrote:
> > At 09:49 AM 12/24/2006, D. Scott MacKenzie wrote:
> >> I use MMANA quite a bit - it also has an implementation of NEC-2.
> >> It works
> >> well, fast, and gets me in the ballpark. I use it to see whether
> >> an idea
> >> works, or to see what sort of changes I need to make.
> >> The best part of it - is that it is free.
> >> I would love to do a comparison between several different pieces
> >> of software
> >> and see which one models the best, or gets accurate
> >> results....anyone have
> >> any simple and more complicated models that they would like to
> >> compare? It
> >> might make a good article for QST or QEX.....
> > There is already such a comparison. It's in one of the antenna
> > compendiums.
> > Basically, all programs that do method of moments work about the same
> > (many of them actually use the identical NEC engine underneath). The
> > differences are in convenience of user interface, and in some
> > "addons".
> > NEC4 does produce better results than NEC2. It has better arithmetic
> > codes, and, uses a slightly different way of representing the current
> > distribution in each segment which has better numerical conditioning
> > (i.e. roundoff doesn't bite quite so hard).
> > NEC4 also does much better at modeling wires close to ground.
> > You definitely want to use the Sommerfield-Norton ground for either.
>TowerTalk mailing list
TowerTalk mailing list