Issue link: https://iconnect007.uberflip.com/i/1520213
88 SMT007 MAGAZINE I MAY 2024 for the different moisture conditions using the same color scheme as before, with orange and yellow as MSL3 and MSL4, respectively. While little can be concluded from Figures 23 and 24 in terms of sample behavior over warp- age, the trendlines do show a minor pairing with MSL3 and MSL4 slightly separated from the trend of the Control and MSL3 + Bake data. More than anything this helps to justify the validity of the data set and the conclusion that moisture is not having a significant effect on the overall warpage of these samples with higher die-to-mold ratios. Conclusions and Summary Unlike previous studies in the space of pack- age moisture vs. thermal warpage 5 , a tangi- ble correlation between warpage and mois- ture was found for certain package types. is is possibly due to better variable control with increased understanding over time as it relates to different conditions that can affect thermal warpage. For the samples in this study, those with larger mold volumes relative to die size exhibited spe- cifically higher warpage levels near and above reflow temperatures. As expected, the longer MSL3 moisture soak showed higher warpage levels than the MSL4 exposure. Samples with less mold and more substrate and die exhibit no relevant variation in thermal warpage due to moisture exposure. Certainly, the larger samples also showed higher warpage overall, as is to be expected. Samples that were prebaked, subject to MSL3 and then prebaked again, or "reset," cor- related with the control data set which went through a single prebake. is shows the abil- ity for a sample to be exposed to moisture then prebaked again, without playing a tangible role in thermal warpage. Next Steps More samples can, of course, be tested using a similar setup. Further sample types would help to enforce the correlation between die-to- Figure 24: Average relative coplanarity, second-order polynomial fit, 11 x 12 mm.