Study On Robustness Parameter Optimization Of Lead - Fit Process

Wire bonding is one of the key processes in the microelectronic package and assembly, and the parameter optimization design of wire bonding is of very importance during the process development. The bond parameter optimization design can not only reduce the development period of the process and products, but also improve the stability of the process and the reliability of products, thus significantly reducing the cost. In this thesis, a multi-responses robust parameter optimization approach based on Surface Response Method (SRM) is presented in terms of the characteristics of gold wire ball bonding parameter optimization problem. The proposed approach built the statistic model between input parameter and the mean and variance value of output response separately, combining the advantage of high efficient modeling of RSM and the concept of process robustness. The optimize goal is to get the minimum variance with the mean value which meets specific requirements. Then, the optimum process parameters sets can be obtained by using multi-response optimize method.The process parameters of fully automatic thermo-sonic gold wire ball bonding on the LTCC pad was optimized by the proposed approach. For the first ball bond, the experiment scheme was designed and undertaken, with the mean and variance value of ball size and shear force as the output responses, and with the3key parameters (ultrasonic level, bonding time and force) as the input. Then the experiment results were analyzed and discussed based on theory of probability statistics. As to the second stitch bond, the experiment design and analysis method are basically the same with ball bond. Instead of the commonly used double bond pull test as the output response, a single bond pull test, which can respond to the bonding strength more efficiently, was used. The bonding character under the optimized parameters was compared with corresponding standards, and all meet the requirement.