Study On In-Plane Dynamic Testing Technique Of MEMS Based On Machine Micro-vision

Measuring and testing technique has actual significance in the course of research, development and industrialization of MEMS. Testing technique based on MEMS mainly includes testing technique of material characterization, dynamic characterization and reliability characterization. Among those, testing technique of dynamic characterization has the most important significance in the course of research and development of MEMS. Firstly, dynamic characterization of MEMS determines the basic performance of MEMS. Secondly, the critical problems on three dimensional motion condition, material characterization, reliability, failure mode, failure mechanism etc. of MEMS can all be solved by dynamic testing technique. Finally, a series of related basic theories can be studied by dynamic testing technique of MEMS. So many research organizations on MEMS pay much attention to dynamic testing technique of MEMS. This paper designs a set of MEMS dynamic testing system based on machine micro-vision. The hardware part of the whole system consists of six modules: high voltage driving circuit module, stroboscopic driving circuit module, microscope module, image acquisition module, data acquisition and driving module and test device controlling module. The software part of the whole system consists of nine modules: image acquisition module, image preprocessing module, image calibrating and standardizing module, image physical dimension measuring module, image searching and patterning module, data acquisition and driving module, device controlling module, dynamic characterization measuring module and test report generating module. Based on dynamic testing system, two testing techniques have been given in this paper on measuring in-plane dynamic characteristics of MEMS: blur image synthetic technique and block matching technique. The course of blur image synthetic technique characterizes as followed. At the condition of continuous lighting blur images of MEMS resonator at certain driving frequency are acquired. Using all kinds of image process techniques such as histogram equalization, space area smoothing and sharpening, wavelet area deleting noise and enhancing etc. Using these techniques and assembling properly can prepare origin image sufficiently. After preparing, amplitude of MEMS resonator is acquired by edge detection technique. Testing result precision achieves sub-pixel because of sub-pixel orientation technique. Combing sweep-frequency technique amplitudes of MEMS resonator at a series of frequencies are acquired. So amplitude frequency response characteristics curve of MEMS resonator is obtained. At last the resonator frequency and quality factor of MEMS resonator are computed. On the block matching technique clear images of high-speed MEMS device are acquired based on stroboscopic imaging principle. Then using block matching technique to analyze these images with different phases and different driving frequencies and to achieve in-plane motion characteristics of MEMS device.