TL;DR
This paper tackles the critical issue of quadrature error in dual-mass MEMS gyroscopes, a major contributor to bias drift and reduced accuracy. The authors propose and experimentally validate the Coupling Stiffness Correction (CSC) method as the most effective approach, significantly improving bias stability and Angle Random Walking (ARW) compared to other methods and the uncorrected system.
What Problem Does This Paper Address?
MEMS gyroscopes are increasingly vital for applications requiring precise motion sensing, but their performance is often hampered by bias drift. A significant source of this drift in dual-mass MEMS gyroscopes is the quadrature error, which arises from manufacturing imperfections leading to coupling stiffness between the drive and sense modes. This coupling generates an unwanted force that mimics an angular rate input, degrading the gyroscope's accuracy, especially over long periods and varying temperatures. Existing correction methods have limitations, prompting the need for an optimized solution.
Methods
The paper investigates three primary quadrature error correction methods:
- Charge Injecting Correction (CIC): This method directly compensates for the quadrature error signal in the sense loop. It's simpler but doesn't address the root cause of the error.
- Quadrature Force Correction (QFC): This approach aims to balance the quadrature error force using electrostatic feedback. It's theoretically better than CIC as it addresses the force, but still relies on modulation steps that can be affected by phase errors.
- Coupling Stiffness Correction (CSC): This method directly targets the source of the problem by using specialized combs to generate a negative electrostatic stiffness that counteracts the inherent coupling stiffness. This is presented as the most fundamental and theoretically superior approach.
The authors analyze these methods through control theory, model simulations, and experimental validation. They also emphasize the importance of correcting the left and right masses independently, as their quadrature errors are found to be different.
Datasets
Not discussed in the paper. The paper focuses on experimental results obtained from a custom-built MEMS gyroscope system.
Key Insights and Ideas
- Independent Correction is Crucial: The left and right masses of the dual-mass gyroscope exhibit different quadrature errors, necessitating separate correction strategies for each mass to achieve optimal performance.
- Targeting the Source is Key: The CSC method, which directly addresses the coupling stiffness, proves to be the most effective. This highlights that tackling the fundamental cause of the error yields superior results compared to compensating for its downstream effects (signal or force).
- Temperature Stability Matters: The CSC method demonstrates robust performance across a wide temperature range (-40°C to 60°C), indicating its suitability for real-world applications where environmental conditions vary.
- Noise Reduction Benefits: Beyond bias stability, the CSC method also shows improvements in noise characteristics, further enhancing the overall performance of the gyroscope.
Key Results
The paper presents significant improvements in bias stability and Angle Random Walking (ARW) using the CSC method:
- Bias Stability:
- Without correction: 38 °/h
- CIC method: 48 °/h
- QFC method: 9.9 °/h
- CSC method: 3.7 °/h
- Angle Random Walking (ARW):
- Without correction: 0.66°/√h
- CSC method: 0.21°/√h
The CSC system also demonstrated excellent repeatability across six test groups and maintained stable performance over a wide temperature range.
Limitations and Future Work
- Limitations: While the CSC method is shown to be superior, the paper notes that the "correction together" method (applying the same correction to both masses) can lead to overcorrection in one mass. The paper also mentions that phase errors and noise can still influence CIC and QFC methods, which CSC avoids.
- Future Work: The paper concludes by emphasizing the effectiveness of the CSC method and its potential for improving structure yields in applications requiring high coupling stiffness. Further research could explore integrating CSC with other advanced control techniques or miniaturizing the CSC implementation for even wider adoption.
Summary updated