Mechanical-Thermal Noise in Drive-Mode of a Silicon Micro-Gyroscope

Bo Yang, Shourong Wang, Hongsheng Li, Bailing Zhou2009 · DOI

TL;DR

This paper investigates the impact of mechanical-thermal noise on the drive-mode of silicon micro-gyroscopes (SMGs). They introduce a new closed-loop driving scheme that decouples phase and gain, and use stochastic averaging to model the system's dynamics. The key finding is that while the closed-loop system improves frequency stability, it doesn't reduce the RMS noise amplitude, though it can be reduced by increasing the quality factor and drive amplitude.

What Problem Does This Paper Address?

Silicon Micro-Gyroscopes (SMGs) are crucial inertial sensors, but their accuracy is limited by various noise sources. While mechanical-thermal noise has been studied for the sense-mode of gyroscopes, its effect on the drive-mode has been less explored. This noise, arising from the random motion of atoms within the gyroscope's structure due to temperature, sets a fundamental limit on the precision of the drive-mode's vibration. Accurately accounting for this noise is essential for developing higher-accuracy gyroscopes. Furthermore, achieving stable and precise drive-mode operation is critical for the overall performance of the SMG.

Methods

The authors propose a novel closed-loop drive scheme for SMGs that decouples the phase and gain control of the driving system. This allows for independent optimization of these two critical parameters. To analyze the system's behavior under thermal noise, they employ a stochastic averaging technique. This method simplifies the complex, non-autonomous differential equations governing the gyroscope's dynamics into a set of autonomous equations that describe the "slow" evolution of the drive amplitude and frequency. This approach allows for the derivation of the spectral density of random errors caused by thermal noise and the analysis of both steady-state and transient responses. They also develop a system model for the closed-loop driving, incorporating the effects of thermal noise and the control system's components.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Closed-loop driving doesn't reduce RMS noise amplitude: A significant finding is that, contrary to potential intuition, the closed-loop driving scheme does not reduce the root-mean-square (RMS) noise amplitude caused by mechanical-thermal noise. This amplitude is found to be independent of the quality factor (Qx).
  • Frequency stability improvement: The closed-loop system effectively stabilizes the drive frequency around the natural resonant frequency of the drive-mode, indicating successful phase control.
  • Trade-offs in control: The paper highlights that the gain control branch of the closed-loop system can lose its effectiveness if the reference voltage (Vref) exceeds a certain criterion voltage (Vrefo).
  • Quality factor's role: While the quality factor (Qx) does not affect the steady-state amplitude of the drive displacement, it does influence the output of the integrator in the closed-loop system.
  • Electrical vs. Mechanical Noise: In the tested configuration, the electrical noise from the driving circuitry is found to be larger than the mechanical-thermal noise. However, as the driving mass decreases, the mechanical-thermal noise can become dominant.

Key Results

  • The RMS noise amplitude due to thermal noise in the closed-loop driving is equal to that in open-loop driving and is independent of the quality factor Qx.
  • The RMS noise frequency can be reduced by increasing the quality factor (Qx) and the drive amplitude (ā).
  • Simulation results demonstrate that the closed-loop system achieves stable drive frequency close to the natural resonant frequency.
  • Experimental results show that the closed-loop driving is immunized to changes in the quality factor due to temperature variations, with drive amplitude signal changing by only 0.083% over a -40°C to +40°C range, despite a 2.6-fold decrease in Qx.
  • The standard deviation of the drive frequency drift was 0.0205 Hz (5.0 ppm), and the amplitude drift was 0.0165 mV (14.7 ppm) over 1 hour, indicating high stability.

Limitations and Future Work

  • Electrical Noise Dominance: The current study indicates that electrical noise from the driving circuitry is the dominant noise source in their setup. Future work could focus on further reducing this electrical noise.
  • Driving Mass Dependence: The paper suggests that as the driving mass decreases, mechanical-thermal noise becomes more significant. Further investigation into the behavior of SMGs with very small driving masses would be valuable.
  • Model Simplifications: While stochastic averaging provides a good approximation, it is based on certain assumptions. The paper acknowledges tiny differences between simulation results and the averaged equation due to integrator simplification and truncation errors.
  • Future Work: The authors' findings suggest that further optimization of the closed-loop control parameters, particularly in relation to the reference voltage, could lead to even better performance. Exploring the impact of mechanical-thermal noise on other aspects of gyroscope performance, such as bias stability, would also be a logical next step.

Summary updated

BibTeX
@misc{yang2009mechanical,
  author = {Bo Yang and Shourong Wang and Hongsheng Li and Bailing Zhou},
  title = {{Mechanical-Thermal Noise in Drive-Mode of a Silicon Micro-Gyroscope}},
  year = {2009},
  doi = {10.3390/s90503357},
}

Loading related content...

Figure