The program terminated due to an error: Convergence failed and minimum time step reached


#1

Greetings,

I am facing problems in transient simulation of a simple PN junction diode. When I simulated the device using version 4.6.590, everything works fine. However, when I do with the latest version (5.1.871) I get following error.

“C:\Program Files\Microsoft MPI\Bin\mpiexec.exe” -n 1 “C:\Program Files\Lumerical\DEVICE\bin\device-engine-msmpi.exe” “C:/Users/admin/Desktop/DEVICE Exp/Exp1.ldev”

admin-HP(process 0): The program terminated due to an error: Convergence failed and minimum time step reached.
admin-HP(process 0): Error: there was an unknown parallel error. The error code is 9002, the process number is 0"

I have tried to reduce the minimum time step to as low as 200fs and abs lte limit as high as 0.05 but the problem persists. I also reduced the step voltage to 0.1V only but in vain. On the other hand, the same simulation works fine even with 500fs time step, 0.001 abs lte limit, 1.0V voltage step, and much coarse mesh on the older version of DEVICE (4.6.590).

Please advise,

Thanks in advance.


#2

Hi. This is indeed unexpected. Can you please share the file with me? You can either post it here or inbox me.


#3

Dear aalam

I have shared the project file with you on your inbox as the file is larger than 6MB limit. Meanwhile I tweaked “max update” to 1V for both the Poisson and DDS with minimum time step 1ps, abs tol 0.01 (meaning I am no more using the default solver controls), it seems the solver is now able to go beyond 1ps (the minimum time step). Before, it was generating the error I mentioned in my previous post for the first step (i.e. the error was occurring at t=1ps). After tweaking the solver is now able to calculate till t=41.817ps and then generates the following error.

“C:\Program Files\Microsoft MPI\Bin\mpiexec.exe” -n 1 “C:\Program Files\Lumerical\DEVICE\bin\device-engine-msmpi.exe” “C:/Users/admin/Desktop/DEVICE Exp/Exp1.ldev”

admin-HP(process 0): The program terminated due to an error: Numeric overflow error. The simulation has diverged in the charge calculation.
admin-HP(process 0): Error: Process number: 0 had a PML error and had to terminate.

Again, the simulation is running without errors on the older version and is pretty fast as compared to the later version. It took only 40 minutes to complete the simulation on previous version while the same simulation took 5 hours to complete only 50% of the simulation and then generated the above mentioned error.

Please help me out of this.

Thank you in advance.

Best regards.


#4

Hi,Abdurrahman
Did you solve this problem? If you have settled, would like to share with me, as I am puzzled at the same problem with you! Thank you so much!
Kindest regards!


#5

Hi @xmqiu, if you can share your file with us then I will be happy to take a look at it.


#6

Thank you, Dr.Alam, I think this problem has been solved.