Automatic detection method for BGA defects based on x-ray imaging

Abstract

Ball grid array (BGA) packaging is a high-density surface mount technology with the advantages of small size, good heat dissipation, and electrical properties, and is widely applied in the production of large-scale integrated circuits. With the rapid development of IC integration, devices assembled using BGA technology generally have greater complexity. However, BGA defects can seriously affect device performance and bring difficulties to product quality inspection. More importantly, in the process of BGA defect inspection, the high complexity of the device brings unprecedented challenges to the precise location of defects, which means that corresponding inspection methods should be improved. To this end, this paper proposes an automatic detection method for BGA defects based on x-ray imaging. First, x-ray imaging technology is utilized to achieve non-destructive detection of the BGA area inside the device and generate image data. On this basis, a set of algorithms including threshold separation, detection filling, and closing operation is designed to complete automatic detection of BGA defects. Furthermore, to objectively evaluate the effectiveness and performance of the proposed method, we conduct a series of comparative experiments using simulated and real data, and generate visual outputs. Through these experiments and analyses, we confirm that the proposed method plays an active and effective role and has robust performance in BGA defect detection. In particular, our method shows the expected performance in precisely finding BGA edge defects and subtle defects.