Thank you PlanarMetamaterials for your reply,
In fact it is a microstrip coupler, I use FR4_epoxy with relative permittivity of 4.4 . I see the training manual, which gave an example on 180 Ring Hybrid, so I try to follow the same procedure with changing the geometry.
I tried the driven modal solution, it gives me more reliable result, but still there is a difference with what I should get.
Thank you again
The structure you created does not represent a microstrip device. A microstrip is characterized by a solid conductor backing on one side of a dielectric, with traces on the other. See here for an example of setting up a microstrip device in HFSS.
I re-design the coupler using microstrip structure as in the link, which was very useful. Unfortunately, the S-parameters plot was wrong, is there differences in excitation between driven modal and driven terminal. I appreciate your help.
This file for my new design TisView attachment Quad_hybrid_org.rar
I worked on the three points you mentioned, and also try to simulate using and without meshing, but the plot is almost the same.
Now it looks much better, but I believe there is error(s), can you help me, please?
This is the file after final editing: View attachment Quad_hybrid_org.rar
Your setup now looks correct. I would change the frequency sweep to type "discrete", but other than that I think the rest is just your particular design. This one seems to work OK around 3.6 GHz.