Reduced-order averaged modeling of active-clamp converters

Reduced-order averaged modeling and analysis of soft-switching dc-dc converters employing active clamping techniques are presented in this paper. A general modeling method based on time-scale separation of converter state variables is used to develop large-signal averaged models that accurately represent the converter dynamics up to half the switching frequency. The fast-changing resonant variables of the active clamping circuits are eliminated in the modeling process, resulting in reduced-order models that are much simpler than existing full-order models. A method is also presented to account for the effects of switching-frequency ripple on converter dynamics in the low-frequency region. A transformer-coupled active-clamp single-ended primary inductance converter (SEPIC) is studied in detail to demonstrate the application of the method. Numerical simulation and experimental measurement of frequency responses of a experimental SEPIC converter are presented to validate the developed models. Additionally, an active-clamp flyback converter is briefly studied to demonstrate the general applicability of the method to other active-clamp converter topologies.