Optimal Parameter Design for IC Wire Bonding Process by Using Fuzzy Logic and Taguchi Method

Abstract

Fuzzy logic Taguchi method (FLTM) is used to optimize parameters for wire bonding process. The proposed FLTM integrates orthogonal arrays, signal-to-noise ratios, response tables, analysis of variance (ANOVA), fuzzy logic, and multiple performance characteristics index. The main process parameters for wire bonding are ultrasonication time, ultrasonication power, bond force, bond force time, and search force. The output responses include ball shear and ball size. The orthogonal arrays, signal-to-noise ratios, and response tables reduce the number of experiments needed to find the best factor-level combinations. The significant control factors are determined by ANOVA. In engineering practice, ball size increases as ball shear increases, but an excessively large ball size causes a short circuit, whereas an excessively small ball size cannot provide enough ball shear. Due to the two contradictory output responses, the FLTM is used to find the best process parameters. The experimental results show that the process parameters obtained by the FLTM result in a smaller ball size and a larger ball shear compared with those obtained by statistical methods, artificial intelligence techniques, and previous designs.

Keywords

• Wire bonding
• Taguchi method
• fuzzy logic
• multiple performance characteristics index