Is that possible to use meep to calculate PCB stackup impedance?

[ganghuang]

The first step for any RF PCB design is determine the stack up and calculate the trace/gap width for typical impedance, e.g. 50 ohm, 100 ohm.
May I use the meep to verify the calculation like below?
I tried the atlc (https://atlc.sourceforge.net/) and got similar result for single ended co-planar wave guide (50 vs 51 ohm), but for the differential pair, I got ~108.5 ohm instead of 100 ohm as suggested by the vendor.
I want to try some other EM solver to see which one can do better on this kind of job.

[smartalecH]

You typically want a mode solver for these kinds of calculations (meep is a linear solver, useful for analyzing the response of a system given some specific current-source input). (It's also a nonlinear solver, and even has a mode solver wrapped around it using a shift-invert scheme... but that's probably overkill and not really worth it... as described later).

In particular, for this problem, you probably want a variable meshing mode solver that works well with metals. So not MPB.

Maybe check out femwell.

In essence, you would solve for the modal fields and from those fields, compute the characteristic impedance etc.