Signal-to-noise ratio enhanced electrode configurations for magnetoelectric cantilever sensors

Abstract

Magnetoelectric cantilevers consisting of strain-coupled magnetostrictive and piezoelectric (PE) layers are applicable to magnetic-field sensing. For the first bending mode, the magnetic field-induced stress distribution is of equal sign along the cantilever length. Thus, a plate-capacitor electrode configuration encompassing the complete PE layer may be used for collecting the strain-induced charge. For higher order modes, stress regions of the opposite sign occur in the cantilever length direction. To prevent charge cancellation and to harvest the piezoelectric induced charge efficiently, segmented electrodes are employed. This study investigates the effect of the electrode configuration on the signal-to-noise ratio (SNR) for higher order bending modes. The charges collected by the electrodes are calculated using a finite element method simulation considering the mechanical, electrical, and magnetic properties of the cantilever. By combination with an analytic noise model, taking into account the sensor and amplifier noise sources, the SNR is obtained. We analyze a 3 mm long, 1 mm wide, and 50 μm thick silicon cantilever with layers of 2 μm magnetostrictive soft amorphous metal (FeCoSiB) and 2 μm piezoelectric aluminum nitride. We demonstrate that an SNR-optimized electrode design yields an SNR improvement by 2.3 dB and 2.4 dB for the second and third bending modes compared to a signal optimized design.