PCB eletromagnetic simulation

I had watched this video about simulations regarding the PCB geometry, it use a proprietary software](https://www.keysight.com/en/pd-2998595/pathwave-advanced-design-system-ads-2020).
At past I had worked in a high current density / high frequency switching (power electronic inverter), I would like to glimpse this simulation possibility (for a updated optimized version) but Pcbnew doesn’t even export to ODB++ format.

Possible the proprietary software import GERBER, but I would like to ask first:

  1. Someone knows some open (at least free) package software do this “geometrical electric simulation”?
  2. Some approach to suggest me? Convert into FreeCad first, …

Maybe https://openems.de/start/index.php would be interesting for you?

I knew that I already have read this in old topic.
I will some and evaluate the integration.

This topic has been mentioned several times before on this Forum.
Aparently the underlying math is very similar for other matrix oriented calculations such as thermal designs and FEM.

A good magic word seems to be “field solver”:

In this topic about OpenEMS, FreeCAD has been mentioned a few times, and apparently FreeCAD is at least somewhat usable for this, although I’ve never tried it myself.

FreeCad’s FEM capabilities are not really up to the task mentioned. But you can use it to get the .step files, which is the exchange format and I’d imagine that any tool capable of importing OBD++ should easily import .step files. You’d probably need to set the parameters of different items (copper, substrate, mylar, …)

Currently the fcad_pcb macro which imports KiCad layout to FreeCad does not support board stackup (which is available in 5.99). But the author is quite responsive so maybe you could ask him/her to support importing the board stackup.

Is FastMaxwell a relevant product in this case?


This is an open-source, quasi-static and full-wave, 3D, electromagnetic solver for efficient extraction of interconnect, power distribution, RF inductor structures accounting for skin/proximity effects, displacement currents, and substrate effects. The code is accelerated via the discrete complex image method, full-wave integration routines and an iterative pre-corrected fast Fourier transform (pFFT) matrix vector product. The code has been tested on a variety of large scale (up to 300,000 unknowns) structures.

I have no expertise in this field.
A few days ago I just bumped into it via:

which lead to the “spiki” on github, which uses FastHenry to make an estimation of the inductance of a spiral track on a PCB.
After MitjaN’s last post I revisited this thread and noticed that the text above was still in “edit” mode and I missed to click on the “Reply” button…

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.