Enhancement of flying probe tester systems with automated optical inspection

This paper investigates the possibility of enhancing a flying probe tester (FPT) with an automated optical inspection (AOI) module. The goal is to achieve a wider range of defect detection and decrease inspection times. The AOI system utilizes principal component analysis (PCA) implemented in a fashion similar to the eigenface decomposition methods for face recognition. Several sub-images of components were extracted from the global image of a printed circuit board (PCB). System training was performed with a partial set of these component images, which represent different classes like capacitors, resistors and unpopulated slots. Testing was performed on a wide range of component images and the effects of noise, occlusion, position shift, rotation and lighting variation, were studied to characterize system performance. A scheme to use the FPT in conjunction with AOI is also proposed. It makes use of board manufacturing data and FPT result feedback to enhance test coverage