A problem of result accuracy in HFSS

[ifelix]

Hi,
I've a problem related to HFSS results.
I've simulated a planar antenna as in the pictures in the attached file.
I've made a simulation with lumped port then I've substituted the lumped port with the wave port and I've deembedded till I've obtained the corrispondence with the position of the lumped port.
Then I've compared the results:
S11 diagram is quite similar and also Radiation efficiency values are in good agreement.
The problem arises when comparing Directivity and consequently Gain because between the two models there's a difference of about 1 dB and a quite strong change in 3D polar plot of gain as you can see from the pictures attached.
I think wave port should be more accurate and as a consequence more reliable but I'd like to know if I can assume also lumped port results as reliable as the other ones because I've to modify my model and I can't continue to use wave port. I need to pass to lumped port and so I'd like to know if the related results could be considered reliable.
In my application the radiation pattern and gain values are very important!!
Moreover, could anyone tell me what's the cause of this difference in polar plots behaviour?
Thanks a lot in adv.

 

[costox]

Hi,

Could you please send me the .hfss file to:
sintox9 @ ya hoo com

From your figure, it is because with the lump port, the results are calculated right at the output, whereas with the wave-port, the results are calculated a little further away. Since the extension is not perfectly match, the waveport would have a smaller gain and the side lobe will also be affected because of this missmatch.