You'll love this one then...I tried some very small input cap Pi( or CLC ) filters. Started small and worked my way up. I got some errors realted to allowable time step size which went away and re-appeared as I increased the size of the cap. Good old finite-element numeric methods and all the stability problems which can crop up if the proper math and approximations aren't used.
For a system composed of masses and springs and forcing functions, the results are pretty close to what real stuff does in the continious world.
cheers,
Douglas