Fault diagnosis using magnetic image of PCB

With the digital technology and very large scale integrated circuit technology widely used, the structure and function of electronic devices are becoming more and more complex, the defects of contact diagnosis has become increasingly prominent for using the probe or needle bed, it demands people began to in-depth study the diagnostic techniques of PCB non-contact. Based on magnetic image of the PCB fault diagnosis technology is a new non-destructive testing technology of PCB, developed in recent years, it can achieve rapid detection and location of the circuit board failure, improve the reliability of the circuit board. Aimed at the specific PCB, its different failure modes of the magnetic image of the PCB are attained by Ansoft software, are carried out through the magnetic image filtering, image enhancement technology; they are completed on multi-level wavelet decomposition, the construction and extraction of wavelet energy features of the magnetic image. The PCB fault diagnosis based on the magnetic image is completed by using the improved momentum-adaptive rate neural network algorithm. The results show that the magnetic image method is effectively diagnosis method for PCB, and it´s a new fault diagnosis approach of PCB.