Problem with simulating the near-field in HFSS

I'm trying to simulate the near-field of a circular patch antenna driven by 2 probe feeds that are 90 degrees out of phase. Also, the center of the patch antenna is grounded.

The model simulates fine (at 915 MHz), and I'm able to obtain some near-field plots using a sphere via the radiation menu in the project tree. However, the near-field plots tend to look a little strange depending on the size of the near-field sphere I use.

The circular patch has a radius of 61.5 mm, so what would be a good size near-field sphere to use?

Also, the far-field radiation pattern should be symmetric about the patch center, but it turns out slightly askew, what would be the cause of this?

(I can upload my model if need be)

Thanks!

The whole complication with near field analysis is that you have a radial component to your fields rather than just the typical far field "TEM". So, indeed your near field plots will be a function of the size of your near field sphere.

I would use the near field radius that you plan to measure for accurate results.

Feel free to upload your model and I will take a look at your skew issue when I get chance later today.

Upload your file. Let's try to figure out solution together.

Hey guys,

Sorry for the delay in uploading my file, I was out of town for a few days.

Here's the file containing my patch antenna that I'm trying to simulate at 915 MHz. In the current file, I had tried out using a near-field sphere with a radius of 100 mm and a skewed near-field resulted.

I'm still stumped as to why this is happening, but then again, I'm pretty new to HFSS.
Thanks in advance for all your help!

I have looked at and fixed your project. You had some major/minor problems in your model.

1) Your waveports were connecting to coax TL, however the teflon had no cladding. Therefore no energy was getting to your antenna and your Return Loss was bad. I added pec BC to outside face of coax. Also added pec puck to allow for waveports inside solution domain.
2) Your airbox was too small. If you use airbox follow these guidelines...quarter wavelength away from any radiating component for Radiation boundary condition and an airbox that is 1/10 lambda away for PMLs. I used PMLs as I prefer them for antenna applications.
3) Your near field sphere needs to be LARGER than your antenna ;) In your case, your antenna has diameter ~120mm and your sphere radius was 100mm, bad. I increased Near Field radius to 130mm and all is fine. If you need fields closer than this use the near field along a line option.
4) This is an ideal solution for iterative solver, assuming you have multiprocessing
5) I like the use of mixed order tets, as they are more efficient...but different strokes fro different folks on that one
6) I was unsure whether this was to be a finite or infinite ground plane. I have simulated as if finite.

This project solves well now, have fun :)