High throughput measurement techniques for wafer level yield inspection of MEMS devices

This paper presents two distinct measurement systems that were custom-built for the parametric and functional yield inspection of MEMS devices on wafer-level. Throughput as well as accuracy was optimized by using automatic feature detection and data segmentation algorithms. Inaccuracies in stage positioning during scanning are compensated for by a grid detection algorithm. The analysis of the measurement data is performed in parallel with the ongoing measurements. The data analysis includes the detection, parameter extraction, analysis of failures or damage of a single device and the final stitching of the results in order to obtain a visual mapping of the measured arrays. The performance of both systems has been demonstrated using arrays of micromirrors as test vehicles.