Yes, i have tried with different mesh size, and there are small differents, but not enough to match the measurement with the simulation results.
I think that it can be a problem with the excitation. I'm using a wire port with a voltage source, and i have seen that results are very sensible to the radius of the wire. The problem is that if I use the real radius, then I have warning or errors due to the fact that the length and width of the segment are more or less the same.
I don't know exactly how to introduce the excitation correctly (especially if the substrate is thin).
If you wish, I could send the the feko file.
KRAIG, Lots of thanks