Adaptive Dead-Time and Partial-ZVS Regulation for GaN-Based Active Clamp Flyback Converter With Predictive Hysteresis Current Mode Control

Active clamp flyback (ACF) converter is regarded as a good candidate in high-frequency small-size adapters. With adaptive dead-time optimization method, there is no reverse conduction loss and switching- on loss in power switches, then ACF can realize optimum full-ZVS control at every operating conditions. However, the other losses such as the conduction loss of clamp switch, the primary winding loss of transformer may be increased to counteract this efficiency advantage, but which has not been outlined. Therefore, this paper proposes a higher-efficiency predicted hysteresis current mode control strategy with adaptive dead-time and partial-ZVS regulation for GaN-based ACF converter for the first time. First, the accurate analytical model by considering above losses is derived. Second, the optimal valley current is calculated by the minimum loss, to determine the conduction time of clamp switch and the dead time of main switch. Finally, a digital controlled GaN-based ACF converter is designed to verify the method. The computation speed of needed analytical equations and the magnetizing inductance offset are considered in implementation process. The experimental results show the proposed method can improve efficiency by 0.4% than existing method, which also show a fast dynamic response since the cycle-by-cycle variable-frequency control concept.