Power ultrasonic systems are typically operated at their eigenfrequency to achieve the required large vibration amplitudes. In systems employing multiple transducers, overall performance strongly depends on the alignment of their mechanical eigenfrequencies. Even small differences caused by manufacturing tolerances and uncertainties in material properties can lead to undesired energy exchange, excitation of disturbance modes, and instability in the generated acoustic field. Existing resonance adaptation strategies typically require additional components or complex shunt networks. This work presents a compact and hardware-efficient method for active and continuous eigenfrequency adaptation based on a resonant-switching full-bridge converter. The approach integrates the alternation of electrical boundary conditions directly into the driving process of the transducer. A resonantly switched rectangular voltage waveform is applied, and its amplitude fundamental harmonic is adjusted through block width. Introducing controlled open-circuit intervals within each vibration period enables continuous adaptation of the effective eigenfrequency from below the series resonance f s toward the parallel resonance f p . Two complementary models are developed: a harmonically linearized model for rapidly estimating the achievable eigenfrequency adaptation range, and a model with piecewise representation of switching states for accurately describing the system dynamics. Both models are experimentally validated. The proposed method achieves a maximum eigenfrequency shift of 424.77 Hz, which corresponds to 84.7% of the separation f p − f s =504.95 Hz. Further measurements demonstrate that when the total driving intervals exceed approximately 40% of one vibration period, the system can operate without additional sensor hardware, while the converter maintains an energy conversion efficiency above 82%. The proposed eigenfrequency adaptation approach provides an efficient and low-complexity solution for operating multi-transducer ultrasonic systems at a mutual eigenfrequency. • Enables active continuous eigenfrequency adaptation without auxiliary hardware, reducing system complexity. • Provides two complementary models for efficient adaptation range estimation and accurate transient-dynamic analysis. • Shifts the effective eigenfrequency from series to parallel resonance by modifying the switching strategy during operation.
Chen et al. (2026) studied this question.