Validation and reliability assessment of board level drop test of chip-scale-packaging

Abstract

The main objective of this study is to develop a stress-buffer-improved package that is subjected to a board level drop test under a specific-G impact level. In this regard, both the drop test experiment and the ANSYS/LS-DYNA simulations are carried out. Several studies have shown that the solder joints having the brittle intermetallic compound (IMC) layers within the wafer level chip scale packaging (WLCSP) are the weakest part. For the most part, this is due to the large relative motion occurring between the board and the chip. In addition, the stress buffer layer exhibiting a relatively large elongation which reduces the impact on the solder balls. Meanwhile, the novel stress-buffer-improve package's failure mode is different from the convention WLCSP structure which shifts to the trace damage of the chip side. The leading concern between the solder ball and trace damage is the critical region where failure occurred owing to the stress concentration effect. During the drop test experiment, the proposed stress-buffer-improved package is able to survive over 100 drops (most packages survived at above 200 drops). Hence, this drop performance very much surpasses the Joint Electron Device Engineering Council (JEDEC) criterion (drop number is 30 times). Nevertheless, the metal traces which are embedded in the stress buffer layer suffered relatively larger deformation. Generally, the stress concentration occurs at a single position, much like the trace/pad connecting junction in the analysis of detailed stress-buffer-improved package. Finally, the predict result in finite element (FE) analysis is similar to the broken metal trace's failure analysis in the drop test experiment.