A hybrid approach for multi-objective design optimization of ball grid array gold wire bonding process

A hybrid approach for multiple performance characteristics optimization of ball grid array (BGA) gold (Au) wire bonding process is proposed. 8 main process factors of BGA wire bonding technology are selected as the control factors; such as factor A (Seating USG), factor B (TIP Height), factor C (C/V), factor D (USG Current), factor E (USG Bond Time), factor F (Bond Force), factor G (FS Threshold), and factor H (FAB Size). The performance characteristics of the process in the study including wire pull strength, ball shear strength, ball thickness difference, ball size difference, and percentage of the Au-Al intermetallic compound (IMC) are measured. The hybrid approach is to firstly use the Taguchi orthogonal array to carry out experiments for calculating the S/N ratios of the selected performance characteristics. The principal component analysis is then applied to determine the principal components of the S/N ratios, which are transformed via fuzzy logic reasoning into an integrated multiple performance index (MPI) for further analysis of the effect of each control factors on the bonding process. Through the analysis of process factors and analysis of variance (ANOVA) on the MPI, the best process design is determined as A1 (Seating USG = 80mA), B3 (TIP Height = 6mils), C1 (C/V = 0.7mils/ms), D3 (USG Current = 95mA), E3 (USG Bond Time = 25ms), F1 (Bond Force = 16grams), G1 (FS Threshold = 10grams), and H3 (FAB Size = 1.65mils). Experimental results show that the best process design indeed helps improve multiple performance characteristics. Compared with the initial process design, the best design is able to improve the wire pull strength by 0.014db, the wire shear strength by 1.569db, the ball size by 9.326db, and the IMC by 1.088db.