An improved method for inspection of solder joints using X-ray laminography and X-ray microtomography

Abstract

This paper describes the application of several imaging technologies available at the Center for solder joint inspection. X-ray laminography was combined with artificial neural networks to classify solder joints. Components with ball grid array, gull-wing and J-lead joints were imaged and several neural network methods were used to identify different classes of defects particularly significant to each type of joint. A novel probabilistic neural network approach for 2-D image classification has been developed which performs as well as or better than a conventional backpropagation network. The smear caused by the laminographic process poses a great challenge to accurate reconstruction and subsequent evaluation of the object. An improved method of accurately reconstructing the solder joint shape from the laminographic images has been developed as part of this research. The method removes artifacts caused by out-of-plane contributions, noise, and smear due to rotation of the source around the object while forming each laminograph, and can be adapted to consider the finite size of the aperture and X-ray scattering. Preliminary application of the method has produced dramatic improvements in the visual quality and signal-to-noise ratio for laminographs of experimental objects. More importantly, the ability to accurately measure the dimensions of the objects being imaged has been made possible by this approach. The possible extension of this work by using more X-ray projections and mathematically intensive routines brings this research into the realm of microtomography, which can help achieve more precise reconstruction at a much smaller scale. A new method of microtomography has been developed that can exceed previous limits in image resolution.