Re: NEC-LIST: Trouble with dipoles

Tiffany

I've compared NEC-2 and FDTD previously for modelling dipole
impedances. I found discrepancies between the two methods because the
wire model with NEC giving the better results. The reason for the
difference was that the FDTD code I was using didnt correctly model
the capacitance of the wire ends, as it didnt use a different
formulation for the ends compared to the middle segments.

I don't know whether this helps

Regards

Michael Bradford

In message <9810290329.AA02446_at_fs1.ece.ubc.ca>, Tiffany Buffaloe
<tkb_at_ee.duke.edu> writes
>Hello,
>
>I'm doing some simple dipole impedance calculations using NEC4. I'm
>modeling a 23.5cm half-wavelength dipole with a radius of .000425 in
>free space. I'm comparing these results to an finite-difference
>time-domain calculation of impedance for the same dipole. A frequency
>of 638 MHz corresponds to a half-wavelength of 23.5 cm. My NEC dipole
>contains 7 segments, and I'm using a frequency step of 22.5MHz in the
>NEC simulation to match the step-size in the FDTD simulation. I get a
>match frequency of 607 MHz from the NEC code, and a match frequency of
>630 MHz from the FDTD simulation. (I have investigated running the
>NEC code with a much smaller frequency step of 1 MHz over the region
>of interest and again found a match at approximately 600 MHz) It would
>seem that the FDTD is giving more accurate results.
>
>I am looking for an explanation for the discrepancy. If anyone can
>give me any insight, suggestions, or corrections, it would be greatly
>appreciated. A listing of my NEC input file is below.
>
>Thanks,
>Tiffany
>
>
>CM LAMBDA/2 DIPOLE
>CE
>GW 0 7 0. 0. -.1175 0. 0. .1175 .000425
>GE
>EX 0 0 4 0 1.
>FR 0 255 0 0 22.5 22.5
>XQ
>EN
>

-- 
Michael Bradford