A novel parameter-independent digital optimal control algorithm for DC-DC Buck converters based on parabolic curve fitting

Abstract

Recently, optimal control on Buck converter for powering the latest computer central processor units (CPUs) has attracted more and more attention. In this paper, a novel digital control algorithm is presented to achieve the time-optimal response for dc-dc Buck converters without relying on any knowledge of converter design parameters such as output inductance, capacitance and even ESR value. This algorithm is based on the parabolic curve fitting analysis for deriving the algorithm formulas under the step load transients. Furthermore, this algorithm can be extended to adaptive voltage positioning (AVP) applications with simple modifications for a low ESR designed Buck converter. Also, it delivers a practical and cost-effective interface to AVP schemes due to the parameter-independent and current-sensorless detecting mechanism. Finally, simulations and experimental results of a 12 V-1.5 V Buck converter prototype are provided to validate the proposed schemes using digital signal processor (DSP) implementation.