Design and Modelling of Buck Regulators Driving Loads With Complex Parasitics

Abstract

Switching regulators improve power efficiency and reduce cost of electronic devices. Continued advances in the performance of components have outpaced the models used for their analysis and design. Physical limits of the parts and their arrangement introduce nonideal parasitic effects that can no longer be ignored. This article introduces an extensible frequency domain model of buck regulators that can capture parasitic effects of arbitrary complexity. Such models are derived for both continuous and discontinuous conduction modes of peak current mode controllers. The proposed models are shown to be accurate when compared to other methods, while having lower computational cost. Numerical optimization is applied to design systems using the increased complexity models. Cost functions are developed, and shown to produce improved designs when compared to existing published examples. Experimental measurement finds the models accurately predict the performance of implemented designs. Additional parasitic elements are added to the model, showing improved fit to measured results, demonstrating the extensibility and accuracy possible.