Gain and current depend on segments problem

Hello,

I have been trying to verify some published results I found for the gain of a 6 element yagi antenna. The antenna design is from a paper by

David Cheng, IEEE Antennas and Propagation, Vol 33, No3, June 1991 p 43.

Cheng simulated the antenna using a 3-term expansion of the currents which is different than the pulse expansion in NEC. The simulated gain that he got using this method is 13.4 dBi.

I wanted to use NEC to compare this antenna to some others , but I'm having troubles deciding how many segments per element should be used to model the yagi. As I increase the number of segments per element, the gain and maximum current go down. Here is what I get:

Segments/element Gain (dBi)
3 13.45
7 12.98
11 12.28
21 11.36

As I continue to increase the number of segments, the gain continues to go down. The currents also continue to decrease as the segments increase. It does not seem that they are converging toward a steady state. I have tried using the extended wire kernel but it didn't help much.

How do you choose the number of segments that will accuarately model the antenna without comparing it to test results?

thanks for any input,
eric
Received on Thu May 23 1996 - 21:45:00 EDT