Motivation

Switching mode power supplies (SMPS) are widely used in power management systems due
to the high efficiency that they can achieve. When comparing the two methods of controlling
an SMPS, i.e. voltage-mode and current control, the latter offers an easier compensation loop
implementation and faster response to load changes. However, a current sensor is required, at the
inductor or the input. This sensing scheme increases the overall bill of materials of the system, the
converter footprint, and makes the system more sensitive to several noise sources.
With the constant development of integrated circuits such as microcontrollers, it is possible
nowadays to develop new control techniques in terms of hardware and software synergies. This
provides a shift in performance to higher levels by the deployment of advanced digital signal
processing algorithms, breaking the barriers imposed by technological limits of the today’s semiconductor hardware components.

The present dissertation aims at investigating the implementation of an extended Kalman filter
(EKF) to improve the performance of a synchronous buck converter. The proposed EKF is
based on a mathematical model of the buck converter, which takes into consideration several
parasitics such as the inductor and capacitor equivalent series resistors. The goal is to take the advantages of using an observer type filter as the EKF in order to reduce the impact of noise sources, for instance thermal noise or any other type of Gaussian distributed noise. In the proposed EKF approach, the inductor current is estimated with no additional hardware and used as a parameter for the predictive control.

The EKF algorithm has been implemented on an Infineon XMC microcontroller controlling a synchronous buck converter. Experimental results demonstrate an accurate estimation of the inductor current and reveal fast step responses.