硅微振梁式加速度计的温度检测及闭环控制

为实现硅微振梁式加速度计系统芯片级温度测量及系统闭环，本文针对系统的非惯性结构部分提出了微机电系统（MEMS）结构温度的芯片级测量和闭环控制优化方法。与以温控罩的温度作为参考温度的方法不同，该方法提出了供芯片级温度测量的MEMS结构、工艺及配套电路，通过直接测量MEMS结构的温度完成实时补偿，从而提高了测量精度。该方法在闭环控制的前置电路中应用了二极管电容解调电路，与前期使用的跨阻或者跨导方案相比，对器件的要求从pA级降至nA级。运用时域方法求得二极管电路方案的解析解，提出参数优化设计方法，保证了电容测量输入与输出间的线性关系。最后，采用二阶最优模型对闭环控制的后置电路进行参数优化，控制了上电时间。配合硅微振梁式加速度计原理样机进行了实验。实验结果表明，温度补偿后的零偏稳定性为52.0 μg，标度因子稳定性为16.0×10^-6，分辨率为34.9 μg。这些结果验证了本文理论的可行性。

Temperature measurement and close-loop control of silicon resonant accelerometer

To measure the chip-level temperature and to achieve close-loop control for a Micro-Electro-Mechanical(MEMS) resonant accelerometer system, this paper investigates non-inertial parts of the system, proposes a method to measure the temperature of the MEMS structure and optimizes the parameters in the close-loop control. Different from the traditional method that using a temperature from a temperature control cover as the conference, this paper proposes design methods of MEMS structures, processing technology and circuits, and achieves the temperature compensation by measuring the MEMS structure temperature directly to improve the temperature measuring accuracy. Furthermore, a diode pre-circuit is applied to detection of the variation of the tiny capacity instead of the transimpedance amplifier and transconductance amplifier, which reduces the requirement for high-performance components from pA to nA magnitudes. Based on the analysis in the time domain, analytical solution of diode pre-circuit is proposed to optimize the parameters and to guarantee the linear relationship between input and output. Moreover, a second-order optimal mode is applied to control of the settling time of the post-circuit. Experiment shows that after compensation on temperature, the MEMS resonant accelerometer has the performance of bias stability of 52.0 μg, scale factor stability of 16.0×10^-6, resolution of 34.9 μg. The result indicates that the proposed theories can satisfy the requirements of high-performance MEMS resonant accelerometer systems.