Vibration-induced errors in MEMS tuning fork gyroscopes

This paper analyzes potential causes of vibration-induced error in ideal MEMS tuning fork gyroscopes. Even though MEMS gyros are known to be highly susceptible to mechanical vibration, the mechanisms responsible for generating vibration-induced errors are not well understood. We focus on the tuning fork gyroscope (TFG) design that is known to be relatively immune to vibration because of its differential operation and common-mode rejection. Our analysis, however, demonstrates that even an ideal TFG cannot completely eliminate vibrations in special situations because of vibration-induced asymmetry and nonlinearity. We identify three major causes of error that arise from (1) capacitive nonlinearity at the sense electrode, (2) asymmetric electrostatic forces along sense direction at the drive electrodes, and (3) asymmetric electrostatic forces (i.e., drive-electrode capacitance) along drive direction at the drive electrodes. The occurrence conditions and characteristics of each cause are analyzed. The effects of the causes on three classes of TFG gyro designs are analyzed and compared both qualitatively and quantitatively using dynamic analysis and simulation. Interestingly, in our simulation conditions, one TFG design (denoted as Type-DD) is less sensitive to vibration (>99% reduction) than the other two TFG designs (denoted as Type-CP and Type-DS). The reason for this difference is that Type-DD gyroscopes are immune to the dominant cause of error afflicting Type-CP and Type-DS designs. Our analysis also demonstrates that the most critical error-generation condition is vibration along with the sense direction of gyroscopes (denoted as sense-direction vibration) because of its contribution to all causes of error. (C) 2012 Elsevier B.V. All rights reserved.Authors especially appreciate Prof. Noel C. Perkins (Department of Mechanical Engineering, University of Michigan) for the thoughtful discussion on vibration analysis and simulation. Authors also thank to Dr. Jaeyoong Cho (University of Michigan) for gyroscope discussion. This work is supported by DARPA's HERMIT Program (Contract #: W31P4Q-04-1-R001). Simulation resources were provided by The Center for Advanced Computing in The University of Michigan