Suppressed drift and low-noise sensor module with a single-axis gold proof-mass MEMS accelerometer for micro muscle sound measurement

This paper describes a highly sensitive microelectromechanical systems (MEMS) accelerometer sensor module to measure micro muscle sounds. Regarding the drift mechanism, we focus on charged particles based on residual water related to MEMS fabrication. Based on temperature programmed desorption analysis and electrical experiments, it is clarified that the drift is caused by negatively charged OH- and is suppressed by vacuum packaging. The noise floor can be reduced by the improved sensitivity of the MEMS accelerometer. The developed MEMS accelerometer can realize the sensitivity of 1157.0 fF/G and Brownian noise of 149.0 nG/root Hz. The developed sensor module consisting of the developed MEMS accelerometer and the capacitance to digital converter can achieve a noise floor of 26.7 mu G/root Hz without any obvious drifts, establishing the target value of less than 50 mu G/root Hz. Therefore, it is confirmed that the developed sensor module can detect micro muscle sounds without contracting a muscle in the frequency range of 20-40 Hz.