HFSS, semi-infinite dielectric substrate

[icodreanu]

Hello everyone,

I am trying to model a very simple structure in HFSS at 95 GHz. It is a half-wave dipole on a dielectric substrate (εr=28). I would like to model the substrate as extending to -Inf in the z direction and from -Inf to Inf in the x and y directions.

I was hoping that I could accomplish just that by terminating a finite size dielectric box with PMLs. However, the input impedance of the printed dipole depends on the thickness (z direction) of the substrate.

Can anyone provide some advice?

Thank you in advance.

 

[rautio]

If there is any way you can get hold of Agilent Momentum or Zeland IE3D, those will be much better tools for this problem. The infinite dielectric comes for free. No worry about PMLs and you just mesh the surface of the metal of your circuit. Excitation is very easy too, just specify a port and let it analyze. I view IE3D as faster and Momentum as more accurate. Others will of course to some degree have different opinions. I view volume meshing a problem like this very close to being an outright mistake. (I work for Sonnet, we compete with Momentum and IE3D in some areas, but this problem is not really one of those areas.)