Event-based Spread Spectrum Voltage and Current Mode Digital PFM Controllers in LLC Converters

Abstract

A pulse frequency modulation (PFM) technique is widely accepted for LLC converters in order to achieve reduced RMS current of the transformer. However, it suffers from varying switching frequency at steady-state under varying power levels, which makes it difficult to design an input filter to meet EMI compliance over a wide load current range. Under PFM, the switching frequency is set by the feedback loop based on voltage/current regulation requirements, and also a variation in the duty ratio is often not recommended from the RMS current point of view. This makes it difficult to identify suitable spread spectrum methods under closed-loop. This paper proposes event-based sampling and periodic perturbation methods in digital PFM control. Both voltage and current mode digital PFM architectures are proposed. By suitably modulating the control signal, spectral peak can spread over a wider frequency range with insignificant effects in RMS current and transient performance. A unified discrete-time (DT) large-signal model is developed, which can be used for different mode configurations of the LLC converter over a wide operating range. DT small-signal models are derived, which can be used for closed-loop stability analysis using the proposed methods. Conventional PFM and the proposed spread spectrum PFM techniques are compared. Hardware prototype of a 2.5 kW LLC converter is developed, and experimental results are presented to demonstrate spread spectrum using the proposed technique.