HFSS Simulation Help Needed

I am a new user in HFSS v.12. I would like to simulate inductive coupling between two inductance. The problem when I increased the number of turn the program gives errors.

[error] Adaptive solution setup, process hf3d error: Matrix solver exception: out of memory.. (10:22:56 mai 17, 2010)
[error] Unable to create child process: hf3d. (10:22:56 mai 17, 2010)
[error] Simulation completed with execution error on server: Local Machine. (10:22:56 mai 17, 2010)

I think the problem comes from meshing. But I could not find the solution.
I have also uploaded the HFSS file. can you help me on that?

Thanks in advance
Enver

puzzled with the same problem,thanks in advance !

You need more memory for your computer.

egkilinc said:

I am a new user in HFSS v.12. I would like to simulate inductive coupling between two inductance. The problem when I increased the number of turn the program gives errors.

[error] Adaptive solution setup, process hf3d error: Matrix solver exception: out of memory.. (10:22:56 mai 17, 2010)
[error] Unable to create child process: hf3d. (10:22:56 mai 17, 2010)
[error] Simulation completed with execution error on server: Local Machine. (10:22:56 mai 17, 2010)

I think the problem comes from meshing. But I could not find the solution.
I have also uploaded the HFSS file. can you help me on that?

Thanks in advance
Enver

Click to expand...

dear egkilinc
I had same problem with this error :

[error] Adaptive solution setup, process hf3d error: Matrix solver exception: out of memory.. (10:22:56 mai 17, 2010)
[error] Simulation completed with execution error on server: Local Machine. (10:22:56 mai 17, 2010)

I think your radiation boundary condition is too large in dimension
as you know it should be about lambda/4 from radiating surface.
try it and tell the result
I can't simulate your file because my software is hfss11.

The error you are getting is due to you not having enough memory. If you have a 64 bit machine, you can turn up your virtual memory to allow for swapping. As for your simulation, the distance a radiation boundary is from an object is a rule of thumb!!!! If you have a strongly RADIATING problem, lambda over four is necessary. however, if you problem radiates very little (like small loops) you can bring your radiation boundary closer. As an example I get accurate inductances on loops at 30MHz with an airbox that is 25mm away from my coils. The most important thing to look at is the following...

Plot mag E in a plane in a quick, unconverged simulation. Look at the magnitude of E at the boundary as opposed to the structure. If the magnitude on the boundary is ~3 or 4 orders of magnitude BELOW the magnitude around your object, then even if any of the fields are reflected due to poor ABC conditions, the reflected fields are so small that they will be even smaller after reflection and minimally influence your results.

You can check this by parameterizing your airbox. Keep increasing the size of your airbox until the inductances no longer change...then plot mag E in a plane and you will see that the E fields satisfy above or better (for poor radiators).

Have Fun!!