Bio-inspired acoustic MEMS sensor with tunable resonance frequency

Abstract

A microelectromechanical system (MEMS) – based acoustic sensor with adaptive bandpass characteristics is presented, i.e. its resonance frequency can be dynamically modified by up to 25 % using relatively small bias voltages (<1 V). This allows for a frequency analysis of sound in hardware, for covering larger frequency ranges with a single sensor while retaining high quality factors, adapting the sensor for optimum detection performance under different conditions and compensating for mismatches due to fabrication tolerances. To achieve this, we utilize geometric nonlinearity effects through incorporating a clamped-clamped micromechanical beam. We show that for the fabricated sensors, DC actuation changes pre-deflection of the clamped-clamped beam and this pre-deflection changes the resonance frequency of the sensor. To adjust the design parameters, such as geometry, we developed a model, based on the Duffing oscillator equation, which describes the sensor’s nonlinear dynamics, including hardening/softening behavior due to an applied static deflection. We demonstrate the application of the frequency tunability for compensating fabrication variabilities (here thickness variation) and for analysis of frequency components in an applied sound signal. This principle of dynamic frequency tuning and hardware-based frequency analysis using a single MEMS sensor demonstrates potential for applications in (bio-inspired) acoustic sensing, gas flow sensing, and other sensing tasks requiring adaptive, tunable bandpass characteristics and for (adaptive) sensor (oscillator) networks.