![Sample our Engineering & Technology journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5933/TOC-Subject-Sample-2021-Engineering-Tech.jpg"})
Abstract
Copper (Cu) wire has become an alternative material for wire bonding in many microelectronic applications due to the high appreciation in the price of gold. However, the Cu wire bonding process is relatively new to integrated circuit (IC) assembly and must be appropriately controlled to save on manufacturing costs without encountering reliability problems or losing quality. This study proposes an adaptive diagnosis system for the control and quality assessment of the Cu wire bonding process using grey relational analysis (GRA) and a neurofuzzy technique. A fractional factorial experimental design is first utilised to collect structured data, and the results are analysed through an integrated GRA and entropy measurement method to convert the multiple quality characteristics of Cu wire bonding into a synthetic performance index. Next, the neurofuzzy data-learning technique is used to establish the essential knowledge bases. The in-process-quality-control (IPQC) data are then clustered and utilised to fine-tune the membership functions and adjust the weights of fuzzy rules through neurofuzzy data-learning. Finally, customised programming codes are generated for fuzzy rule retrieval and for a graphical user interface design to link users and the process diagnostic and quality assessment knowledge bases. The proposed diagnosis system is evaluated using real-world production data collected from an electronics manufacturing service (EMS) provider of IC packaging to verify its prediction accuracy and applicability.
Acknowledgments
The author would like to thank the manufacturing management team from company T for their valuable assistance and the National Science Council of the Republic of China, Taiwan, for partially supporting this research under Contract no.: NSC 101-2410-H-366-003.