Evaluation of CCGA Solder Pillar Grinding Effect Based on End-Face Imaging Analysis

Due to the advantages of Ceramic Column Grid Array (CCGA) packaging technology such as the good thermal matching and vibration resistance, it is often used for the manufacturing of high-end chips. However, in practical engineering applications, CCGA has many process problems, such as the oxidation of bottom surface of lead column, and the poor coplanarity of welding column end. The CCGA welding column is manually ground for end face flatness processing, it is easy to lead to a poor end face coplanarity, skewing of the welding column and other issues, so that the grinding effect is difficult to be guaranteed. This paper proposes a CCGA column grinding effect evaluation method based on end face imaging analysis, and we use image processing, feature extraction, machine learning algorithms, and other technologies to achieve the binary classification of CCGA column grinding pictures. Compared with other related algorithms, the method proposed in this paper is highly targeted and its effect is stable. It puts forward solution ideas and methods for existing industrial problems.