Plastic Ball Grid Array (PBGA) packages have gained a substantial market share in the IC packaging industry in the past decade. Although PBGA packages have many advantages such as a larger number of I/Os, robust assembly process, and better thermal / electrical performance, their board level reliability is still a major concern. This is because their solder joints have a relatively short stand-off height especially for the Chip Scale Package (CSP)....[ Read more ]

Plastic Ball Grid Array (PBGA) packages have gained a substantial market share in the IC packaging industry in the past decade. Although PBGA packages have many advantages such as a larger number of I/Os, robust assembly process, and better thermal / electrical performance, their board level reliability is still a major concern. This is because their solder joints have a relatively short stand-off height especially for the Chip Scale Package (CSP).

It is a global trend to replace today's harmful lead-containing solders with lead-free solders in future electronic products. Although there are various lead-free solder systems available in the industry, people are still unsure over the selection of lead-free solders due to the lack of reliability data. When implementing such a large change in a material system, the effect on the reliability is always the first concern, especially at the board level. The implementation of the lead-free solders is a challenging process. It involves lots of study of process control and reliability for the board level interconnects. This brings out the value of the current study on the experimental evaluation of board level solder joint reliability of plastic ball grid array assemblies with eutectic lead-tin and lead-free solders.

Four different reliability tests, namely the accelerated temperature cycling test, three point bending test, random vibration test, and the mechanical drop test were performed to evaluate the reliability of the lead-free board level solder joint. Five DOE testing programs including four lead-free systems with different combinations of solder materials and peak reflow temperatures together with one conventional lead-tin eutectic were arranged for the PBGA samples. In the present study, it is found that the selected lead-free solder candidates had better performances than the eutectic lead-tin solder when subjected to the thermal fatigue load. However, the four lead-free solder systems had poorer performances than the eutectic lead-tin solder when subjected to the mechanical loads. The effect of solder joint size on board level reliability was found in CSP samples. The failure mechanisms of various lead-free solders subjected to different kinds of loading were studied. The failure modes were inspected. The locations that are critical to solder joint reliability were identified. Suggestions of improving on the current study and corresponding future work are made.