Issue link: https://iconnect007.uberflip.com/i/721932
44 SMT Magazine • September 2016 dent upon the component termination geom- etry, confirming the conclusion from the previ- ous study. It is also clear that the assembly pro- cess has a meaningful effect on self-mitigation, which is a new finding. The difference between HASL and OSP pad finish has no significant impact on self-mitiga- tion. Variation in pad size exhibits a statistically significant, but very minor impact on self-miti- gation, for the range of pad sizes investigated in this study. Self-mitigation of the various packages in the study can be broken down into three broad groupings: 1. Packages that exhibit a near certain prob- ability of self-mitigation over the entire range of processes examined in this study (0603 chips, SO14Gs). 2. Packages that exhibit a near certain prob- ability of self-mitigation, but only for one par- ticular assembly process (TFQP100 and LQFP48 for Process G). 3. Packages that exhibit a very low probabil- ity of self-mitigation under all conditions (0612 chip). 4. Packages that exhibit a moderate proba- bility of self-mitigation within the range of 0.6– 0.9 (all other packages). Readers should note that 0603 chip compo- nents are available in a wide variety of compo- nent heights and termination lengths, so the re- sults shown here would only be applicable to chip components with heights and links that were equivalent or smaller than those used in this study. Future Work Plans IPC task group 8-81f plans to continue in- vestigating this hardware over the next year. A comprehensive plan of cross sectioning and SEM/EDS analysis will be developed and implemented. The intent of this is to under- stand the causes for the very low thickness readings and determine which if any of this data is valid. Also, we will validate where the original tin finish has been completely con- sumed, and to cross check the composition measurements. A detailed statistical analysis will be per- formed to identify any meaningful correlations between the various process parameters associ- ated with the assembly processes and self-miti- gation behavior of the various components. Two additional boards have been assem- bled using components that were in uncon- trolled storage for a number of years. The ef- fect of this storage on self-mitigation will be investigated. Table 6. Self-mitigation probabilities with 90% confidence. MITIGATION OF PURE TIN RISK BY TIN-LEAD SMT REFLOW