Signal Processing Method for Scanning-Acoustic-Tomography Defect Detection based on Correlation between Ultrasound Waveforms

It is common to apply scanning acoustic tomography (SAT) for microelectronic packages inspection because it enables nondestructive imaging of defects. SAT phase inversion imaging is useful tool for detecting defect areas accurately and quickly. Phase inversion imaging is based on the phenomenon of the phase inversion of an ultrasound wave at interfaces in a sample. We propose a signal processing method for phase inversion imaging. Specifically, it is a robust method of finding the phase inversion. With this method, the phase inversion is found by estimating the correlation between a reflected wave signal of a sample and reference wave signal. A negative correlation between these signals indicates the existence of defects in the sample. Defect detection tests with the proposed method were carried out using integrated circuit packages. The tests results indicate that the proposed method improves robustness for detecting defects compared with a conventional method. The tests also demonstrated that small defects on the order of 100 micrometers could be detected using an ultrasound probe with a nominal frequency of 75 MHz.