The information you provided is insufficient. Could you mention the source and boundary conditions.?

I suggest you to please share some more details so I can give you the needful suggestions. Beside that, I think you you getting different solution might be variation in the frequency due to the source and boundary conditions.

Regards

Shrish

Thanks Vasant, Mohammad and Shirish for your interest and reply. The study is a time dependent study with Robin boundary condition, followed by the frequency study at the last time step where I use a small sinusoidal perturbation to get Impedance response. Its probably this stationary step that is giving different solutions to my problem. The intial conditions are constant voltage and constant concentration with ohm's law at boundary elements. Source is however a tome dependent term evolving. However even if I make my source time independent, the problem persists. Any help would be useful.

Do you think it might be because of your mesh? You need to have small enough mesh that by decreasing the size of your mesh you still get the same result. 

Hello,

I think frequency domain is some sort of eigenvalue problem. I saw the Eigen value solver of COMSOL some times are not working very well.  My suggestion is to consider a possible minimum for your Eigen frequencies.

Good luck with your research

Hi,

I realize this is an old question so you must got your answer by now, but I faced the same problem until my supervisor helped me with it, so I'll leave an answer for others just in case.

Go to: Study>>Solver Configurations>>Time Dependent Solver>>choose: (Time Stepping), then control (steps taken by solver) which is (free) by default.

I kept it (free) but set a value for (maximum step) and was satisfied with the results.