MEMS MIKROFON ARRAY ARCHITEKTUR

Abstract

The Signal to Noise ratio of Arrays can be improved by increasing the number of samples (the number of microphones or the number of averages). Increasing the array sized also allows the direct measurement of the source directivity, without moving the array. In this paper we present a module microphone array with data acquisition system, consisting of 1 m x 2m panels with 800 MEMS microphones, each. This modular architecture allows the time-synchronized measurement of an arbitrary number of panels and thus, aperture size and total number of sensors. The array employs Infineon IM69D130 MEMS sensors. They feature a sensitivity accuracy of ±1 dB and phase precision of ±1◦ at 1 kHz based on their data sheet and thus, offer sufficient precision for aeroacoustic research. The raw PDM bit streams from 200 MEMS microphones are each collected synchronously by a single FPGA, and are then sent together with meta information such as timestamps, status, etc., as User Datagram Protocol (UDP) network packets to a server at a continuous data rate of approximately 620 Mbit s−1. Reference airframe noise measurements are performed at the low-speed wind tunnel DNW-NWB, Braunschweig, Germany. The MEMS array architecture was used with 7200 MEMS microphones with a 6 m x 3 m aperture. A second test used 12.000 MEMS microphones with an 6 m x 5 m aperture on an open wind tunnel test at DNW-LLF Sources identification directivity measurements will be shown. Additional far-field microphones and a classical analog array are employed to validate the beamforming results and directivity analysis.