Vibration Analysis Based Modeling and Defect Recognition for Flip-Chip Solder-Joint Inspection

Abstract

A novel approach for flip-chip solder joint quality inspection based on vibration analysis is presented. Traditional solder joint inspection methods have their limitations when applied to flip chips. The vibration detection method is a new approach which has advantages such as being non-contact, non-destructive, accurate, fast and applicable for in-line inspection or during process development. In this technique, a flip chip was modeled as a thick plate supported by solder bumps. Changes in solder joint quality produce different flip chip vibration responses, and also change its natural vibration frequencies. In this paper, the vibration frequencies of a flip chip on a ceramic substrate were calculated using the finite element method. Based on vibration analysis, a laser ultrasound and interferometric system was developed for flip chip solder joint quality inspection. In this system, chips with good solder joints can be distinguished from chips with bad joints using their vibration responses and frequencies. Data analysis and defect recognition methods were developed and tested. Results indicate this approach offers great promise for solder bump inspection in flip chip, BGA, and chip scale packages.