Die attach process monitoring through multivariate statistical process control technique

Statistical Process Control (SPC) is a statistical technique to monitor and analyze the variations of process parameters. Traditional SPC can be used under certain assumptions, such as the parameters being IIND (Identically, Independently and Normally Distributed). For the semiconductor industries, the process often contains several procedures and multi-variables. Due to the characteristics of multi-variety, multiple norm and non-normal distribution, multivariate SPC technique plays an important role in monitoring and controlling the process of die attach. In this paper, the problems related to shear force monitoring in die attach process are analysed. Process monitoring methods on the shear force data of a die attach process are presented. The Xbar-S chart through Johnson transform is used to control bonding force. Based on the shear force data acquired from a real production line of an enterprise in China, a multiple index which means real process capability index (PCI) is calculated. The result shows that this multiple index is an appropriate method for evaluating and monitoring high quality product process.