Study On Measurement Methods And Key Technologies For Dynamic Characterization Of MEMS Microstructures

It is very important to measure and test dynamics of microstructures in order to develop reliable and marketable Micro-Electro-Mechanical Systems (MEMS) products. However, it is difficult to test dynamic behaviors of MEMS microstructures because of their micro size, high frequency responses upon excitation and complex operation environment. On the basis of such a research background, this dissertation intends to study systematically the denoise algorithms for interference patterns, the measurement methods for in-plane and out-of-plane motions of MEMS microstructures, and some key technologies of measurement system for dynamic characterization of MEMS microstructures are realized.The denoise algorithms for interference patterns, including general filter algorithm, spin filter algorithm and SUSAN filter algorithm, are studied. A new filter algorithm suitable for wrapped phase maps is proposed, which can reduce noise without influence of 2πjumps. After studying and comparing the denoise effect of above mentioned algorithms, three practicable schemes for denoising interference patterns are presented.In order to extract in-plane motions of MEMS microstructures with high resolution and high speed, a new sub-pixel synthetical orientation and matching algorithm is presented, which incorporates the advantages of standardization covariance correlation method, sub-pixel step length correlation algorithm based on cubic spline interpolation, surface-fitting algorithm, sequence similarity detection algorithm, and simplex method. For out-of-plane motions measurement of MEMS microstructures, several typical phase unwrapping algorithms such as traditional unwrapping algorithms, the branch cut algorithm and the quality-guided path following algorithm are studied. A method is proposed to select the unwrapping seed point after identifying the"bad points", which can be avoided to choose the noise points. A time-domain unwrapping algorithm on the unwrapping seed point is presented, which can get the displacement error resulted from the motion of the unwrapping seed point, and compensate the out-of-plane displacement at different moving time.A measurement system for dynamic characterization of MEMS microstructures has been developed, which includes three modules such as base excitation, environment control and motions measurement. Based on the testing principle and the requirements for synchronous– control of the dynamic measurement system for MEMS microstructures, a control method based on computer is proposed to automatically realize vision images or interferograms acquisition for moving device with high frequency. A new file format is defined for saving and reading the float style data of phase map effectively. The error effect to the phase measurement because of the phase shifting error of phase stepper and the non-uniformity of illumination source is discussed. A method is put forward to compensate the system's main error by the system error model, which is jointly established by both curved surfaces fitting based on the least square and testing of a standard plane.A lot of experimental investigations are carried out on several typical MEMS microstructures such as micro gyroscopes, micro mirrors, micro pressure sensors, micro resonator arrays and AFM micro cantilevers. The investigations confirm that such a measurement system is able to measure the static surface topography and deformations as well as the in-plane and out-of-plane motions of microstructures, and is effective and efficient to characterize dynamics of MEMS microstructures with a nanometer resolution. Moreover, it includes an environmental control facility, which can realize base excitation, and generate variable pressure and/or temperature for testing of MEMS microstructures. The system's main performance parameters include the frequency measuring range of 0~250kHz, the base excitation frequency larger than 10kHz, the motion resolution for out-of-plane measurement better than 11.2nm, the motion resolution for in-plane measurement better than 15.2nm, the error indication for static surface topography of±2%, the variation of indication for static surface topography better than 1%. The system is demonstrated to have good repeatability and stability for both static and dynamic testing. The prototype of the measurement system was highly evaluated by the Ministry of Science and Technology of China in September, 2004, and thus received continuous financial support from the National High Technology Research and Development Program of China (863 Program). After optimization of the hardware and software, the system was checked and accepted by Ministry of Science and Technology of China in October, 2005, and has been successfully applied to test different MEMS microstructures from several institutions and national laboratories.