Accurate Modeling and Elimination of Double Vertical Crossing in Multi-Sampled Digital-Controlled Buck Converters

Abstract

The multi-sampling method is an effective approach to reduce delay in digital controlled power converters. However, this method may deteriorate the dynamic performance of converters due to double vertical crossing (DVC) of modulation signal and carrier signal, which makes the closed-loop regulation lost. This phenomenon is influenced by the operating point, parameters of power stage and control system. Therefore, it is significant to predict and eliminate the DVC for avoiding its negative influences. This paper proposes an accurate model of DVC in multi-sampled digital-controlled buck converters, where the output capacitance and the sampling delay is taken into account. Then a carrier phase adjusting algorithm is proposed to eliminate DVC. The proposed method features better dynamic performance than existing methods, and it is easier to realize since it keeps switching frequency constant. Finally, the theoretical analysis is validated by the simulation and the experimental results.