ABSTRACT
Bias stability is one of primary characteristics of precise gyroscopes for inertial navigation. Analysis of various sources of the bias drift in a micromachined electrostatically suspended gyroscope (MESG) indicates that the bias stability is dominated by the temperature-induced drift. The analytical results of temperature drift resulting from the rotor structure and capacitive position sensing electronics are modeled and analyzed to characterize the drift mechanism of the MESG. The experimental results indicate that the bias drift is mainly composed of two components, i.e., rapidly changing temperature drift and slowly changing time drift. Both the short-term and long-term bias drift of the MESG are tested and discussed to achieve online bias compensation. Finally, a neural network based-bias compensation scheme is presented and verified experimentally with improved bias stability of the MESG.
ABSTRACT
Spin rate of a high-speed spinning-rotor gyroscope will make a significant impact on angular rate sensor performances such as the scale factor, resolution, measurement range, and bias stability. This paper presents the spin rate effects on performance indicators of a microelectromechanical systems (MEMS) gyroscope where a free-spinning rotor is electrostatically suspended in an evacuated vacuum cavity and functions as a dual-axis angular rate sensor. Theoretical models of the scale factor and measurement range of such a spinning-rotor gyroscope are derived. The experimental results indicate that the measured scale factors at different settings of the spin rate match well with the theoretical predication. In order to separate the disturbance component of the rotation control loop on the gyroscope output, a testing strategy is proposed by operating the gyroscope at different spin rates. Experimental results on a prototype gyroscope show that the squared drive voltage generated by the rotation control loop is approximately proportional to the noise of the gyroscope output. It was further investigated that an improved performance of such spinning-rotor gyroscopes can be achieved by operating the gyroscope rotor at an optimal spin rate.
ABSTRACT
Air-film damping, which dominates over other losses, plays a significant role in the dynamic response of many micro-fabricated devices with a movable mass suspended by various bearing mechanisms. Modeling the damping characteristics accurately will be greatly helpful to the bearing design, control, and test in various micromotor devices. This paper presents the simulated and experimental squeeze-film air damping results of an electrostatic bearing for use in a rotary high-speed micromotor. It is shown that the boundary condition to solve the three-dimensional Reynolds equation, which governs the squeeze-film damping in the air gap between the rotor and its surrounding stator sealed in a three-layer evacuated cavity, behaves with strong cross-axis coupling characteristics. To accurately characterize the damping effect, a set of multiphysics finite-element simulations are performed by computing both the rotor velocity and the distribution of the viscous damping force acting on the rotor. The damping characteristics varying with several key structure parameters are simulated and discussed to optimize the device structure for desirable rotor dynamics. An electrical measurement method is also proposed and applied to validate the numerical results of the damping coefficients experimentally. Given that the frequency response of the electric bearing is critically dependent on the damping coefficients at atmospheric pressure, a solution to the air-film damping measurement problem is presented by taking approximate curve fitting of multi-axis experimental frequency responses. The measured squeeze-film damping coefficients for the five-axis electric bearing agrees well with the numerical solutions. This indicates that numerical multiphysics simulation is an effective method to accurately examine the air-film damping effect for complex device geometry and arbitrary boundary condition. The accurate damping coefficients obtained by FEM simulation will greatly simplify the design of the five-axis bearing control system and facilitate the initial suspension test of the rotor for various micromotor devices.
ABSTRACT
A micromachined gyroscope in which a high-speed spinning rotor is suspended electrostatically in a vacuum cavity usually functions as a dual-axis angular rate sensor. An inherent coupling error between the two sensing axes exists owing to the angular motion of the spinning rotor being controlled by a torque-rebalance loop. In this paper, a decoupling compensation method is proposed and investigated experimentally based on an electrostatically suspended micromachined gyroscope. In order to eliminate the negative spring effect inherent in the gyroscope dynamics, a stiffness compensation scheme was utilized in design of the decoupled rebalance loop to ensure loop stability and increase suspension stiffness. The experimental results show an overall stiffness increase of 30.3% after compensation. A decoupling method comprised of inner- and outer-loop decoupling compensators is proposed to minimize the cross-axis coupling error. The inner-loop decoupling compensator aims to attenuate the angular position coupling. The experimental frequency response shows a position coupling attenuation by 14.36 dB at 1 Hz. Moreover, the cross-axis coupling between the two angular rate output signals can be attenuated theoretically from -56.2 dB down to -102 dB by further appending the outer-loop decoupling compensator. The proposed dual-loop decoupling compensation algorithm could be applied to other dual-axis spinning-rotor gyroscopes with various suspension solutions.
