A testing method for assessing solder joint reliability of FCBGA packages

A brittle solder joint at the interface between a solder ball and a substrate metal pad could result in an open joint failure after the solder ball is attached to a motherboard. Existing metrologies, such as ball shear, do not always detect the weak solder joint strength. A peel test for solder joint strength evaluation of substrate was developed. Several problems were observed during the initial stages of the development for this metrology. These included solder bridging and inconsistency of the test results. Solder bridging was reduced by changing the stencil design to have stencil openings that match the solder ball layout and using a jig to align the Coupon and substrate while assembling the peel test samples. The paste printing and peeling methods were modified and the coupon quality was carefully monitored to make the sample preparation and testing more consistent. The impact of the physical aging and reflow on the solder joint strength was investigated. It showed that the peel strength did not decrease with physical aging, but did decrease with an increased number of reflows. The peel test results correlated with the performance of the substrates. The substrate with low peel strength had weak solder joints and failed under use conditions. The substrates with high peel strength demonstrated stronger solder joints and performed well under use conditions. The metrology provides a gross check about the presence of brittle solder joints on flip chip ball grid array substrates and is successfully used in monitoring the solder joint strength. (C) 2003 Elsevier Ltd. All rights reserved.