Ask for help about probe position in HFSS!!

I use probe to feed the MPA, when I set the probe to one position, it died and said:"port refinement，process abc3d died unexpectedly." But when I set the port in another position, it ran normally! I am confused. Who can help me?
Thanks a lot!

if you can post the project file, it is easier for us to help you find out whats wrong. This kind error we always meet, most of them comes from the model building up.
Best Regards,

Thanks a lot! It does not appear every time, so, I think my model is right. Please verify it!

Moreover, I am studying on a dual-frequecy MPA, recently, I am confused with the high modes. I cut the patch into two pieces with L slit, the bigg one is connected with the lowe mode and the small one is connected with high mode. The problem is how to compress the unexpected high modes. Thanks!

LOL. Where is your ground plane of your coaxial feed wave port?
Modified porject is attached, and I get very roughly S11 paramter for your reference.
Best Regards,

in the freuency band you set I got 3 resonance frequency, I am not sure it is you design specification, and I didnt carefully check whether it is physically resonant.
The characteristic impedance of coaxial feed line is 45 Ohms, I didnt optimize it to 50 Ohms, I have set up the parametric analysis, but I didnt run it. Any further updates, feel free for discussion.
Best Regards,

Thanks very very much! But I think the model you make is very very complicated. what I draw is OK, I think. The terminal port must touch the background, so I substract the cylinder from the box.
I design the dual-frequency MPA, which is used in WLAN, resonating in 2.4ghz and 5.2ghz. Unfortunately, it is far from my expectation, so, i dont know what i can do to modify the structure.

Thanks a lot!

[quote="EdmundZheng"]Thanks very very much! But I think the model you make is very very complicated. what I draw is OK, I think. The terminal port must touch the background, so I substract the cylinder from the box.
I design the dual-frequency MPA, which is used in WLAN, resonating in 2.4ghz and 5.2ghz. Unfortunately, it is far from my expectation, so, i dont know what i can do to modify the structure.

Thanks a lot![/quot]
The ground plane in your model is too big, the aspect ratio of this thin plane is very bad, once you need to include the effect of metal loss in your design, it will bring much troubles to mesh generation.
You did connect trace with coaxial pin, but pin touched ground plane, at this point of view, I cant see any signal transimitting. Even you drill a through hole on the ground plane, you assigned a wave port to this pin, but there is no ground reference plane for this pin. Then the solve will treat this port at this way:
an infinitely long cylinder immersed in air, Z direction is the longitudal direction, using 2D eigenvalue solver to solve the eigenmode of this kind of structure, then I also dont know what kind of mode can be supported.
Simulation time of your time is quite short, you can do parametric studying and optimization by HFSS, lots of parameters can be adjusted. :P
Good Luck for your design.
ps, what is MPA? I dont know :( :(
Best Regards

Thanks a lot! I know what your mean. I should draw another cylinder, and make it to be the pin groud plane. however, your modified model was so difficult :) , I must spend more time to study it. Are there any simple methods to draw the coaxial port? ^_^

MPA is short of microstrip patch antenna, ^_^

Anyway, thank you very much!
Have a nice day!