Identification of ZVS Points and Bounded Low-Loss Operating Regions in a Class-D Resonant Converter

Abstract

This paper presents an analysis of the different loss modes of the switching devices in a class-D series resonant converter operating with either fixed dead time or fixed duty cycle. A feasible operating region where the FETs in the inverter stage only exhibit reverse conduction losses, with no hard-switching, is identified. Furthermore, the impact of using a fixed dead time versus a fixed duty cycle is investigated. We find that using a fixed dead time is superior to using a fixed duty cycle, as a broader operating range can be achieved for the same losses, or the same operating range can be achieved with lower losses. A reduction in the reverse conduction losses of up to 59% or an expansion of the operational frequency range by 33% when using a fixed dead time is found. The modeling approach is validated on a 1-MHz prototype employing GaN switching devices. Lastly, a design example shows how the presented analysis can be used to determine the optimal fixed dead time/duty cycle for use with frequency modulation control such that the losses are minimized.

A 1-MHz Resonant LED Driver With Charge-Pump-Based Power Factor Correction

“…A half-bridge gate driver (Si8274 by Silicon Labs, Austin, TX, USA) drives the half-bridge switches. The dead time is set to 88 ns using a fixed resistor and adjusted to achieve ZVS with minimal reverse-conduction through the switches [44]. The gate driver package is placed adjacent to the devices' gate terminals in order to control noise coupling from the power loop to the gate-drive loop.…”

Improved ZVS Criterion for Series Resonant Converters

“…stored energy ( ) not determined correctly because of those presumptions. Recently, an enhanced charge-based criterion has been proposed in [5], [6], where the nonlinear characteristic of C oss and the change of i r during the dead time are taken into account and it is given by…”

“…The charge-based criterion works under the principle that the current flow in the circuit should be in such a direction as to sweep out the charge stored in the output capacitor of MOSFET and the charge displacement also should be larger than the total stored charge in the same capacitor. It is formulated by the initial current and the dead time required to discharge the output capacitor of the switches [3]- [6]. However, the charge-based criterion is only partly accurate since the output capacitor of the switches is mostly regarded as being discharged by a constant current source.…”