Beamforming Performance of a Stand-Alone Digital Piezoelectric MEMS Microphone Array

Описание: Condition monitoring within the resources industry involves tracking equipment parameters to inform the health of machinery. These parameters are acquired through placement of sensors onto equipment surfaces, creating potential exposure to moving machinery, high temperatures and hazardous electrical contact. Acoustic analysis offers the potential for safer condition monitoring through contactless data acquisition via MEMS microphones. Capacitive MEMS microphones currently dominate the MEMS microphone market. A known issue with the capacitive design is a decline in sound acquisition performance when subject to harsh environmental conditions commonplace within the resources industry. A piezoelectric MEMS microphone constructed with a cantilever diaphragm is a potential solution against failure modes such as water ingress, dust ingress and mechanical shock. However the beamforming capability of these piezoelectric microphones for acoustic condition monitoring applications has not yet been tested.
This work implements a pair of Vesper VM3000 piezoelectric MEMS microphone into a stand-alone digital microphone array to investigate the beamforming performance of the technology. The microphone array prototype is designed to acquire sound data without aliasing at frequencies of 2000 Hz or below.
Testing is conducted in both an anechoic chamber and a non-anechoic laboratory environment to simulate a more realistic acoustic environment. We compare the performance of this microphone array to a simplistic theoretical beamforming model which utilises ideal data. The experimental results obtained in the anechoic chamber compare well with the theoretical model.
This work demonstrates the successful design, build and test of a stand-alone digital piezoelectric MEMS microphone array using the Vesper VM3000 MEMS microphones. The developed prototype exhibits accurate beamforming performance under controlled, anechoic conditions but performs poorly under non-anechoic conditions.