SPICE compatible sampled-data models for switching regulators

Although switching regulators are nonlinear devices, a generalized linear, time-invariant model can be used to characterize their behavior for small perturbations around an operating point. The state-space averaged technique, introduced by Bello (1980) for use with the circit analysis program SPICE, has been used successfully by several authors. Brown and Middlebrook (1981) introduced sampled-data modeling techniques to extend the accuracy of the state-space averaged model by including the effects of fixed-frequency, discrete sampling within the feedback network. Presented here is a novel method which allows sampled-data modeling of switching regulators on SPICE.