Circuits with abrupt changes in voltage or current levels are often difficult to correctly simulate and often need some help to dullen or accept the 'sharp edges'. I reckon it is the choke to diode node that is causing the issue, as the choke model being used is likely just a pure inductance with some ESR and such an inductor can numerically generate a voltage across the choke that flies off to kV or higher at the instant the diode turns off or on. A real choke has a shunt capacitance, and some capacitance to ground/core, so those are initial model changes that can help. Simulations can also relax the 'accuracy' that they internally use to make iterative calculations, and that is another typical change that can get a simulation working adequately. It's all part of the spice of life (pun intended) - well at least for the last 40-45 years.
Last edited: