Abstract

The reliability of organic flip-chip ball grid array (FC-BGA) packages has yet to be understood to the depth needed for widespread acceptance. In this work, first-level (chip to package) interconnect assembly reliability was investigated using thermomechanical modeling and experimental verification. In modeling, the mechanical behavior of the package's substrate, stiffener, die, and underfill were accurately represented. In experimental work, test die from 10.35 mm to 18.5 mm in size, with flip-chip bump counts of 332 to 3828 were evaluated. These die utilized various passivations and bump metallurgies. Die fracture strength was evaluated in three-point bending. By combining this die fracture data with modeling results, quantitative prediction of failure rates under various environmental conditions could be made. These predictions were compared with experimental results from similar die attached to organic FC-BGA substrates. It was found that the understanding the details of the FC-BGA substrate behavior, of the assembly materials, and processes, particularly underfilling and cleaning techniques, were crucial in the development of a reliable assembled packages. The combination of modeling and experimentation has resulted a predictive tool which is proving helpful in the design of new organic FC-BGA packages. Initial results on second level (package to board) interconnect reliability indicate that organic FC-BGA packages do provide a significant advantage over ceramic packages.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call