Assembly And Reliability Of Preform Underfilmed BGAs In Drop Test And Thermal Cycling
Authors: Andrew Mawer, Thomas Koschmieder, Paul Galles, Randy Anding, George Skevofilax, and Tim Lippe Company: Freescale Semiconductor, Flextronics International Date Published: 10/16/2011
Abstract: When the transition to Pb-free solders for most consumer electronics occurred leading up to the July 2006 WEEE deadline, the industry initially struggled with the relative poor drop test performance of BGAs with Pb-free solder spheres. Much work was carried out to optimize Pb-free alloys to improve drop test performance [1-5]. Many companies have also implemented epoxy underfill to provide further increases in the robustness of the interconnection between the BGA and the printed circuit board. These underfills are typically dispensed around the BGA following SMT assembly and electrical testing. Capillary action then slowly draws the underfill under the BGA. The underfill may be dispensed in sufficient volume to cover the entire BGA underside, the corner areas only or just the package sides. Finally, the underfill is cured in an offline batch oven. While excellent drop test results have been demonstrated with underfill, the process adds cycle time and may involve processes that are not normally performed during printed circuit assembly. Another option for providing greater BGA to PCB mechanical robustness that may provide a more seamless integration into the SMT manufacturing process is the use of a preform underfilm material . The material can typically be provided in reels similar to discretes and is placed around the perimeter of the BGA following solder paste printing and prior to BGA placement. In reflow, the underfilm softens and initially expands such that it contacts both the BGA and the PCB. By the conclusion of reflow, the material fully cures while remaining adhered to both the BGA and PCB. This paper will cover all aspects of the PCB design and SMT process related to the underfilm material. Industry standard drop test and thermal cycling data on the underfilm will be presented and compared to control samples without underfilm. Cross-sectional analysis showing the coverage of the underfilm and the failure modes in drop will also be presented.