Convergence Issue with Illumination

maurice1707maurice1707 Member Posts: 1
edited May 16 in Photonics

Hello everybody,

I'm currently simulating dark and illumination currents in Ge diodes in CHARGE. After increasing the absorbtion layer thickness I get an convergence error when simulating. This only occuurs when simulating with Optical Generation Data enabled. I get the following error.

I think the following thread in the old forum could help me : "https://kx.lumerical.com/t/issue-with-converging-under-illumination/24227". Since the migration I can't access it anymore. Could somebody from the Support Staff please give me access to the information of that thread or answer my question?

I'm looking forward to your help!


*EDIT: I think the issue has to do with the size of the Optical Generation Data!

Comments

  • greg_baethgegreg_baethge Posts: 120Ansys Employee

    Hi @maurice1707,

    Thank you for posting your question. The page referred in the error message was moved here. You will find the main points to check when your CHARGE simulation is diverging. Regarding the other page, it was a reply to a problem another user had, so some comments were specific to the simulation settings, and they might not apply to your case:

    Since you need to bias the cell in forward, you are better off switching the solver type to newton which is easier to converge in forward bias. In addition, because the generation values that you have in your simulation are quite low, the simulation is slow to converge and needs more iterations to find the solution. So you might wanna increase the global iteration limit to 100 and also use the default update limits which make the simulation faster to converge.

    Could you give some additional information about the simulation settings you are using? Does the issue occur at a specific bias? Could you try and reduce the "scale factor" in the optical Generation import object? What is the optical power set when you extracted the generation rate from the FDTD simulation?

Sign In or Register to comment.