Developing Next Generation Earable for Health Care

Fu, Haoyu

Student Work

Developing Next Generation Earable for Health Care

Public

In this project, we study the design challenges of developing a microphone array for ear health monitoring. We investigate the effects of sampling rate, the distance between microphones, and the number of microphones on the design. Literature shows that the distance between two microphones needs to be half of the wavelength to have a beneficial phase difference. Therefore, we found that the minimum sampling rate of such a microphone array needs to be at least 8000Hz with a minimum of 1.5cm distance between the microphones. Then we investigate various data reading and processing protocols available in common of the shelf-embedded platforms with the goal of identifying the suitable protocol and platform. We looked into three designs, where we used Raspberry Pi with I2S protocol in the first design, Arduino and Raspberry Pi (with external ADC) with SPI protocol in the second design, and Raspberry Pi with 4-channel ADC (supports I2S interface) in the third design. Our first design achieved the desired sampling rate but failed to support more than two microphones. However, literature shows that having a three-microphone array is important to remove front-back ambiguity. Our second design acquires data from four microphones but fails to achieve the desired sampling rate. In the third design, we used Raspberry Pi and 4-channel ADCs with I2S to support four microphones with a minimum sampling rate of 8000Hz.

This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator