We present a long-term bias drift compensation algorithm for high quality factor (Q-factor) MEMS rate gyroscopes using real-time temperature self-sensing. This approach takes advantage of linear temperature dependence of the drive-mode resonant frequency for self-compensation of temperature-induced output drifts. The approach was validated using a vacuum packaged silicon Quadruple Mass Gyroscope (QMG), with signal-to-noise ratio (SNR) enhanced by isotopic Q-factors of 1.2 million. Owing to the high Q-factors, measured frequency resolution of 0.01 ppm provided a temperature self-sensing precision of 0.0004°C, on par with the state-of-the-art MEMS resonant thermometers. The real-time self-compensation yielded a total bias error of 2°/h and a scale-factor error of 700 ppm over temperature range of 25-55°C. The presented approach enabled repeatable long-term rate measurements required for MEMS gyrocompassing applications with a milliradian azimuth precision. © 2012 Elsevier B.V. All rights reserved.

North-finding based on micromachined gyroscopes is an attractive possibility. This paper analyzes north-finding methods and demonstrates a measured 4 mrad standard deviation azimuth uncertainty using an in-house developed vibratory silicon MEMS quadruple mass gyroscope (QMG). We instrumented a vacuum packaged QMG with isotropic Q-factor of 1.2 million and Allan deviation bias instability of 0.2 °/hr for azimuth detection by measuring the earth's rotation. Continuous rotation (carouseling) produced azimuth datapoints with uncertainty diminishing as the square root of the number of turns. Integration of 100 datapoints with normally distributed errors reduced uncertainty to 4 mrad, beyond the noise of current QMG instrumentation. We also implemented self-calibration methods, including in-situ temperature sensing and discrete ±180° turning (maytagging or two-point gyrocompassing) as potential alternatives to carouseling. While both mechanizations produced similar azimuth uncertainty, we conclude that carouseling is more advantageous as it is robust to bias, scale-factor, and temperature drifts, although it requires a rotary platform providing continuous rotation. Maytagging, on the other hand, can be implemented using a simple turn table, but requires calibration due to temperature-induced drifts. [2012-0378]. © 2013 IEEE.

This paper presents a method of dynamically balancing tuning fork microresonators, enabling maximization of quality factor Q-factor) in structures with imperfections. Nonsymmetric tuning of stiffness in a coupled 2-DOF resonator is completed through the use of the negative electrostatic spring effect. This variable stiffness is shown to be able to adjust the reaction forces of the structure at the anchors, effectively balancing any spring imperfections caused by fabrication imperfections. Balancing the structure through stiffness matching minimizes the loss of energy through the substrate and maximizes Q-factor of the device's antiphase mode. The approach is experimentally demonstrated using a vacuum packaged microelectromechanical tuning fork resonator with operational frequency of 2.2 kHz and antiphase Q-factor of 0.6 million. By electrostatically tuning the reaction force at the anchors caused by fabrication imperfections, anchor loss can be suppressed, increasing the Q-factor to above 0.8 million. The experimentally validated analytical model of substrate dissipation is confirmed to be applicable to Q-factor tuning in antiphase driven resonators and gyroscopes. © 2001-2012 IEEE.