Properly meshing a port

[dom11990]

Hello,

I have a question about how to apply proper meshing to the lumped port. Currently my mesh looks like the attachment. In the middle of the port on the left, the excitation occurs. Could someone elaborate on what I need to look out for when meshing a port? Should the excitation signal be on a mesh line exactly? If the mesh is a bit larger than the port, as in the picture, will it significantly degrade the 50 ohm characteristic?

thanks!
dominik

[thorsten]

The excitation, (data) probes and such are always snapped to the nearest mesh lines in any case. Therefore it is better to align your mesh with your structures in the first place, otherwise the excitation and such may end up in locations you did not plan them to be.
Therefore, place mesh lines at the edges of your port to be on the save side. Additionally it is a good idea to model a lumped port a thin sheet. In your case zero thickness in x-direction? But one cell should be fine too...

regards
Thorsten

[dom11990]

Hi Thorsten,

Thanks for your reply. As a follow-up, I have a stepped impedance microstrip structure so making use of the thirds rule is not really feasible. I'm also concerned about the impact of having the lines too close to the edge of the metal, as you state on the website that FDTD struggles to correctly calculate the strong fields here. I'm trying to combine what you said with putting mesh lines directly on the edge of the lumped port with maintaining thirds rule for the metal. In my case, the lumped port sheet is the same width (in y) as the msl that it feeds in to. Therefore it would place lines in y that are directly on the edge of the metal. How can I consolidate this issue? Or said differently, what is more important, that the mesh lines are on the edge of the lumped port or that they are NOT on the edge of the metal? Thank you for your insights!

cheers
dom