Tan Boon Phing wrote:
>
> I am using the pc version nec2. while simulating for larger number of
> segments, I switch to the UNIX NEC. However, with the same input card
> that I have created for the PC-NEC, the unix NEC always claim that the
> input geometry has error.
> CE
> GW,41,1,0,0,-0.007495,0,0,0.007495,0.001
> GW,41,2,0,-0.0008994,-0.0112425,0,-0.0008994,-0.007495,0.001
>
> WIRE
> NO. X1 Y1 Z1 X2 Y2 Z2
> 1 0.00001 0.00000 0.00000 -0.00749 0.00000 0.00000
> 2 0.00002 0.00000 -0.00090 NaN -0.01124 NaN
Which implementation of NEC for UNIX are you using? The problem you
describe is more closely connected to the particular version of the
source code, the language it was written in, and the compiler used to
generate the executable than the OS per se.
In any case, your UNIX implementation of NEC is clearly not reading
the entries in the GW card fields in the manner you intend. And it
looks like you have the ITG and NS fields reversed.... See
http://www.dec.tis.net/~richesop/nec/cards/gw.htm
for all the details concerning the GW card.
You might have better luck if you (1) use floating point numbers
(rather than integers) where floating point numbers are specified
(e.g., 0.0 rather than 0) and (2) adhere to the field specifications
for the GW input card a little more rigorously.
-- Dave Michelson AT&T Wireless Services davem_at_ee.ubc.ca Strategic Technology GroupReceived on Wed Dec 10 1997 - 04:16:07 EST
This archive was generated by hypermail 2.2.0 : Sat Oct 02 2010 - 00:10:38 EDT