Heavy Aluminum Wire Wedge Bonding Strength Prediction Using a Transducer Driven Current Signal and an Artificial Neural Network

摘要

Heavy aluminum wire wedge bonding is commonly used in the packages of power electronic devices. In this paper, a heavy aluminum wire wedge bonding strength prediction system was developed. In this system, the driving current of the transducer system was recorded by a data acquisition system, decomposed, and analyzed using a wavelet model. The fundamental frequency component was extracted, and seven characteristics (five in the time domain and two in the frequency domain) were obtained from the experimental observations. An artificial neural network using back-propagation as training (using 1440 bonding samples) was used; 2200 bonding test samples were then used to verify the performance of the bonding strength prediction system. Experimental results show that this method can be used to accurately predict the shear strength of heavy aluminum wire wedge bonding.