Self-Locking Avoidance and Stiffness Compensation of a Three-Axis Micromachined Electrostatically Suspended Accelerometer

• Main Authors: Yonggang Yin, Boqian Sun, Fengtian Han
• Format: Article
• Language: English
• Published: MDPI AG 2016-05-01
• Series: Sensors
• Subjects: MEMS accelerometer; three-component geophone; electrostatic suspension; cross-axis coupling; self-locking; nonlinear compensation; dynamic stiffness; seismic sensing
• Online Access: http://www.mdpi.com/1424-8220/16/5/711

A micromachined electrostatically-suspended accelerometer (MESA) is a kind of three-axis inertial sensor based on fully-contactless electrostatic suspension of the proof mass (PM). It has the potential to offer broad bandwidth, high sensitivity, wide dynamic range and, thus, would be perfectly suited for land seismic acquisition. Previous experiments showed that it is hard to lift up the PM successfully during initial levitation as the mass needs to be levitated simultaneously in all six degrees of freedom (DoFs). By analyzing the coupling electrostatic forces and torques between three lateral axes, it is found there exists a self-locking zone due to the cross-axis coupling effect. To minimize the cross-axis coupling and solve the initial levitation problem, this paper proposes an effective control scheme by delaying the operation of one lateral actuator. The experimental result demonstrates that the PM can be levitated up with six-DoF suspension operation at any initial position. We also propose a feed-forward compensation approach to minimize the negative stiffness effect inherent in electrostatic suspension. The experiment results demonstrate that a more broadband linear amplitude-frequency response and higher suspension stiffness can be achieved, which is crucial to maintain high vector fidelity for potential use as a three-component MEMS geophone. The preliminary performance tests of the three-axis linear accelerometer were conducted under normal atmospheric pressure and room temperature. The main results and noise analysis are presented. It is shown that vacuum packaging of the MEMS sensor is essential to extend the bandwidth and lower the noise floor, especially for low-noise seismic data acquisition.