【作者】 郭彤；

【导师】 胡小唐；

【作者基本信息】 天津大学 ， 测试计量技术及仪器， 2004， 博士

【摘要】 微机电系统(MEMS)测试的主要目的是为科学研究与产品开发中的各个环节提供数据反馈,其中一个重要方面就是MEMS器件三维运动特性的表征。一种基于显微干涉术的计算机控制的动态测试系统被开发,用于MEMS可动部件的离面运动特性测量,实现了纳米级分辨力。论文完成了以下主要工作: 1.对干涉条纹自动分析技术进行讨论,重点介绍了时间相移干涉高精度测试方法,针对三个核心算法-相位提取算法、相位展开算法和图像预处理算法展开分析; 对影响相位提取精度的误差源进行详细分析,并提出了相应的解决方案; 2.系统地分析了国内外MEMS静态和动态测试技术,基于虚拟仪器的模块化设计思想构建了集光学、机械、电子、计算机于一体的MEMS动态测试平台,满足了MEMS测试的多种需要,实现了高精度、快速、大量程非接触测量,系统配置灵活,可扩充性好、通用性强、容易操作; 设计并调试了频闪驱动电路和高压放大电路,实现了MEMS器件的频闪照明和高压静电驱动; 设计了LED频闪照明和MEMS器件驱动之间的同步控制方案,通过软件进行延迟补偿,实现了高精度、宽频MEMS运动特性的测量; 3.将相移技术分别应用在单色光干涉测量、时间平均干涉测量以及白光垂直扫描干涉测量中,实现了对干涉条纹相位信息的调制,可以用于MEMS器件运动特性的分析中; 4.提出了利用条纹对比度这个重要参数来进行图像质量分析、相位提取算法分析、质量图导引的局部相位展开算法和加权最小二乘全局相位展开算法分析,并应用对比度变化实现时间平均干涉方法和白光垂直扫描干涉方法对MEMS器件离面运动的测量; 5.提出了MEMS器件离面运动分析中的基于时间轴和空间轴的双向相位展开思想,将MEMS器件运动的空间信息和时间信息有机结合起来,有利于对MEMS器件动态过程进行分析,配合基于模板的全场重建技术,实现了对MEMS器件的整体三维运动表征; 6.提出了MEMS运动特性的差分测量方法,通过选取被测器件上的固定点作为参考点,可以减小随机噪声对MEMS器件运动特性测量的影响,提高了系统测量的重复性和精度。更多还原

【Abstract】 A key purpose for testing MEMS is to provide data feedback of measurements for every stage in the process of scientific research and product development, where MEMS devices’ three-dimensional motions characterization is a very important aspect. A dynamic testing system based on microscopic interferometry and controlled by computer is developed, which is used to test the out-of-plane motions of movable MEMS components with nanometer resolution. This dissertation’s main work is as following: 1. The techniques of interferometric fringe automatic analysis are discussed. The temporal phase-shifting interferometry with high accuracy is introduced and developed, aiming at three core algorithms, including phase-shifting algorithm, phase-unwrapping algorithm and image preprocessing algorithm. The error sources, which influence the phase extracting accuracy, are analyzed in detail. And the solutions are presented respectively. 2. Domestic/foreign static and dynamic testing techniques of MEMS are analyzed systematically. A dynamic testing platform is built up according to the modularized design based on the virtual instruments concept, integrated with optical, mechanical, electronical and computer technology. It has many excellent features, such as high precision, high speed, non-contact measurement in a large range, good expandability, high versatility, easy operation and flexible configuration. A stroboscopic drive circuit and a high-voltage amplifier circuit are designed and debugged, in order to realize stroboscopic illumination and high-voltage electrostatic stimulation of MEMS devices. A method of synchronization control between LED stroboscopic illumination and stimulation of MEMS devices is put forward, where software delay compensation is processed, with the result that the system can achieve high precision and broadband MEMS dynamic testing. 3. The phase-shift technology is applied to monochromatic light interferometry measurement, time-average interferometry measurement and white-light vertical scanning interferometry measurement respectively. By this way, phase information of interferometric fringe can be modulated, which is used in the analysis of MEMS dynamic behaviors. 4. A very important parameter, fringe contrast, is proposed, which can be used in analysis such as image-quality, phase-shifting algorithm, the local phase-unwrapping algorithm and global phase-unwrapping algorithm using weighted least-square method. Appling its variation, the system can measure the out-of-plane motions of MEMS devices with time-average interferometry method and white-light vertical scanning interferometry method. 5. In MEMS out-of-plane motion measurement, the two-direction phase-unwrapping method based on time axis and space axis is demonstrated. And the time and space information of MEMS motions are combined in favor of analyzing MEMS dynamic behaviors. In addition, 3-D motions of MEMS devices can be measured by using the whole field rebuilding technology based on mask. 6. The differential measurement method of MEMS dynamic characterization is put forward. A fixed point on the tested device is chosen as a reference point. It can decrease the influence of the random noise resource on the dynamic measurement, and improve the precision and repeatability of the system.更多还原