I'm trying to model a 3-wire dipole over finite ground in NEC2.
However, the SOMNEC output keeps clashing with the data in
my main code (GN card), though I input the same values.
Error is:
ERROR IN GROUND PARAMETERS -
COMPLEX DIELECTRIC CONSTANT FROM FILE IS 2.59287E-09-1.09717E+00
REQUESTED 1.20000E+01-7.00000E-02

Any thoughts?

Ross Tucker
Ariz. State Univ.

"rjtucke" wrote in message
oups.com...
I'm trying to model a 3-wire dipole over finite ground in NEC2.
However, the SOMNEC output keeps clashing with the data in
my main code (GN card), though I input the same values.
Error is:
ERROR IN GROUND PARAMETERS -
COMPLEX DIELECTRIC CONSTANT FROM FILE IS 2.59287E-09-1.09717E+00
REQUESTED 1.20000E+01-7.00000E-02

Any thoughts?

Ross Tucker
Ariz. State Univ.

Have you tried manually calculating the complex dielectric constant? If
there is a
difference greater than 10E-3 from the GN card an error message is
generated.

From your information above the difference is 10E-3. pp 55, 56 NEC 2
Users Manual.

Regards,

Frank

Per emailed request, here is my deck:
CM 40/15M folded dipole model by NS7F
CE
GW 1 11 -10.2761 0.0508 10.0000 10.2761 0.0508 10.0000
0.0016
GW 2 11 -10.2761 -0.0508 10.0000 10.2761 -0.0508 10.0000
0.0016
GW 3 11 -10.2761 0.0000 10.0880 10.2761 0.0000 10.0880
0.0016
GW 4 1 -10.2761 0.0508 10.0000 -10.2761 -0.0508 10.0000
0.0016
GW 5 1 -10.2761 -0.0508 10.0000 -10.2761 0.0000 10.0880
0.0016
GW 6 1 -10.2761 0.0000 10.0880 -10.2761 0.0508 10.0000
0.0016
GW 7 1 10.2761 0.0508 10.0000 10.2761 -0.0508 10.0000
0.0016
GW 8 1 10.2761 -0.0508 10.0000 10.2761 0.0000 10.0880
0.0016
GW 9 1 10.2761 0.0000 10.0880 10.2761 0.0508 10.0000
0.0016
GE 1
EX 0 1 6 00 1.0000 0.0000
FR 0 11 0 0 7.0000 0.0300
GN 2 0 0 0 12.0000 -0.007
RP 0 35 71 1001 0.0000 0.0000 5.0000 5.0000
EN

"rjtucke" wrote in message
ups.com...
Per emailed request, here is my deck:
CM 40/15M folded dipole model by NS7F
CE
GW 1 11 -10.2761 0.0508 10.0000 10.2761 0.0508 10.0000
0.0016
..
..
..
GN 2 0 0 0 12.0000 -0.007
RP 0 35 71 1001 0.0000 0.0000 5.0000 5.0000
EN


My first observation is that your soil conductivity is zero.
Should always be zero (GN card should be: GN 2 0 0 0 12.0000
0.007).
I am certain that this would give you the kind of error messages
you are seeing. The program actually runs under these conditions, using
Nittany's NEC-Win Pro,
but produces erroneous results -- input impedance, radiation patterns etc.

Correcting the conductivity gives much more realistic results: resonant at
6.8 MHz,
Zin 700 ohms, classic dipole pattern.

Your segmentation is probably a bit low for such close proximity conductors,
also
the segment sequence could be improved, but not important unless you want to
graph the conductor currents. I would suggest making the conductor segments
equal to the triangle side lengths. This may change the results a little,
but the
correct sign for conductivity will make a big difference.

Regards,

Frank

rjtucke wrote:

Per emailed request, here is my deck:

etc...

Took a short look at your data. Did not have any (GN) difficulties
running it. (btw the negative GN value is used when running the model
using a freq-loop/sweep). However it showed up you model is largely
unreliable due to the large segment length differences at the end of
you 'dipole'. When running an average-gain test you will also notice a
value of 0.35, indicating a largely unreliable model.

Arie.

"Arie" wrote in message
oups.com...

rjtucke wrote:

Per emailed request, here is my deck:

etc...

Took a short look at your data. Did not have any (GN) difficulties
running it. (btw the negative GN value is used when running the model
using a freq-loop/sweep). However it showed up you model is largely
unreliable due to the large segment length differences at the end of
you 'dipole'. When running an average-gain test you will also notice a
value of 0.35, indicating a largely unreliable model.

Arie.

SIG zero is new to me, but I see it is mentioned in the NEC users
manual. Segmenting tags 1, 2, & 3 at 201 fixes the Average
gain test. Not sure if you can apply segment length tapering to speed up
the program.

Interesting design: VSWR 2.5:1 over 1 MHz, Zin nominal 600 ohms,
fr ~ 6.8 MHz.

Frank