Fusion of acoustic and X-ray image features for solder joint through-life monitoring during thermal cycling tests

Abstract

Novel solder joint monitoring through fusion of acoustic and x-ray image features was designed, aiming to increase the efficiency of multi-inspection tools in studying solder joint through-life performance during accelerated thermal cycling tests. The condition of solder bonds was monitored through ultrasound and x-ray image features from manufacture to failure. A batch of FR4 test boards with flip chip components were thermally cycled for 96 cycles. The test boards were removed from the test chamber at 8 cycle intervals to perform ultrasound and x-ray imaging. The variation of the image features in both ultrasound and x-ray images were tracked and extracted during the whole sequence of the thermal cycling tests. The extracted features were then analysed to allow solder joint defect detection and monitoring. Due to the intrinsic properties of ultrasound inspection, ultrasound images were affected by voids in the solder bumps. Therefore, void information obtained from x-ray images was gathered and fused with ultrasound image features which were able to increase the accuracy of the ultrasound monitoring. The results have highlighted the potential of using multiple image features in solder joint defect detection and performance monitoring during environmental tests.