We apply the P32Tools model for the C100 detector to staring observations. We find that the model does not correctly treat large downward flux steps.