Characterization, Control and Compensation of MEMS Rate and Rate-Integrating Gyroscopes.

Abstract

Inertial sensing has important applications in navigation, safety, and entertainment. Areas of active research include improved device structures, control schemes, tuning methods, and detection paradigms. A powerful and flexible characterization and control system built on commercial programmable hardware is especially needed for studying mode-matched gyroscopes and rate-integrated gyroscopes. A gyroscope can be operated in a mode-matched rate-mode for increased sensitivity or rate-integrating mode for greatly increased dynamic range and bandwidth, however control is challenging and the performance is sensitive to the matching of the modes.

This thesis proposes a system built on open and inexpensive software-defined radio (SDR) hardware and open source software for gyroscope characterization and control. The characterization system measures ring-down of devices with damping times and automatically tunes the vibration modes from over 40 Hz mismatch to better than 100 mHz in 3 minutes. When used for rate-gyroscope operation the system provides an FPGA implementation of rate gyroscope control with amplitude, rate and quadrature closed-loop control in the SDR hardware which demonstrates 400% improvement in noise and stability over open-loop operation. The system also operates in a RIG mode with hybrid software/firmware control and demonstrates continuous operation for several hours, unlike previous systems which are limited by the gyroscope ring-down time. The hybrid mode also has a simulation module for development of advanced gyroscope control algorithms. Advanced controls proposed for RIG operation show over 1000% improvement in effective frequency and damping mismatch in simulation and 25% reduction in drift due to damping mismatch in a test RIG. By tuning the compensation, the drift can be reduced by almost 90%, with worst case drift decreased to -41 deg/s and RMS drift to -21 deg/s. Harmonic analysis of the anisotropy in a rate-integrating gyroscope measured with this control system is presented to guide development of new error models which will further improve performance.