If it’s missing an event, you may need to decrease it. Can you have it print out integrator.dt in the affect!? I think this may be a case where you have a bit of an adverse effect because the ODE might be simple enough that the error is essentially zero for any step, so the dt always grows, and so it grows so large that numerical errors become easy. In which case, fixing maxdt needs to be done. Show integrator.dt to see if that’s the case.