Re: About RF simulators
Hi tjunqueira,
you wrote, that you have to simulate a wave combiner in 3D: I think it is better if you use a expilicite method (e.g. FDTD or FITD; -you simulate explicite in the time domain) as an implicite method like HFSS (FEM,... you have to invert the whole system matrix and this costs you much memory space)!!
So for 32bit system a FEM structure with 1e6 nodes is quiet much, but with with an explicite method like e.g. FDTD you can arrive up to 40-50e6 Yee-cells. In the frequency range of 400MHz you will have no problem with your geometry dimensions and the discretisation step. If you go to lower frequencies (<50MHz??) because of the Courant criterion in FDTD you have to use a Adi-FDTD, and for more lower (<1MHz??)you have to turn back to a implicit method as e.g. FEM.
Maybe this program can do what you need:
SEMCAD-X: **broken link removed**
There is also a "Light Version" available with 1e6 Yee-cells for 99USD, and the full competent and quick support!! (-what can be very helpfull at the beginning... ;-) )
I use this program for my PHD tesis, and have made verry good experience with this.
Hope this helps, best: Ricy