through glass via(Ground-Signal-Ground)

- November 13, 2023 at 9:55 am
  
  Yoon Jung
  Subscriber
  
  Hallo Ansys Teams,
  I'm simulating this(through glass via(Ground-Signal-Ground)) now in HFSS.
  The red part in the middle is the signal trace and the two sides are Ground. Cylinders are via.
  
  I'm wondering where the waveport and radiation boundary are located... Should I just specify them in the middle like a coplanar waveguide, or should I specify the port and boundary conditions exactly where the substrate ends like a grounded coplanar waveguide?
- November 13, 2023 at 10:15 pm
  
  Faezeh Ladani
  Ansys Employee
  
  Hi Yoon,
  
  The important point about the size of the waveport is that it should be large enough to capture the field pattern of the transmission line well. But it should not be too large so that higher order modes start propagating. You can once run and plot the fields on the waveport.
  Also, waveport should be away from discontinuty.... Make sure the vias are not closer than lambda/4 to the waveports.. You may need to extend the upper lines and then deembed the waveport to the where you want to read the S parameters.
  Feel free to navigate the waveport topic on online HFSS help as well. You may find it useful.
  You may use lumped port excitation here as well. You can connect the two grounds through a small PEC sheet and create the port between the trace and this little PEC sheet bridging the two grounds.
  Regarding Rad boundary, it might be better to leave some distance from the substrate in x direction, and of course enough distance on z direction.