Issue link: https://iconnect007.uberflip.com/i/1526114
62 SMT007 MAGAZINE I SEPTEMBER 2024 Figure 14 shows multiple instances of the superposition of local PCB thickness profiles and various global PCB curvatures for a few selected samples evaluated in this study. For example, Figure 14 A1-A4 show instances of "frown" global PCB curvatures and the result- ing different differences due to local thickness variation. In general, there was no trend for general bow/twist curvatures by material, fabrica- tor, or quadrant/design. ere were observed panel-to-panel variations between "smiley" or "frown" curvatures within most fabrication build lots. For a given manufacturing panel, each of the quadrants behaved similar and had similar warpage signatures. e primary factors driving the local thick- ness variation at a given component footprint were the delta copper profile of each quad- rant and size of the component footprint. e "hourglass" thickness shape always fol- lowed when the delta copper profile contained a high level of removed copper from the cen- ter. e "bowtie" thickness shape always fol- lowed when local areas within the interior of the component footprint had a region of maxi- mum copper that approached or exceeded the copper outside of the footprint. Figure 15 shows the relationship between local PCB thickness differences by the compo- nent footprints measured, different copper bal- ance strategies, and material selection. ese measurements were obtained via Akrometrix interface analysis and confirmed with physi- cal microsections on a random sampling. e change in PCB thickness locally under each component had the same trends as Figures 11 and 12. e magnitude of the local PCB thick- ness increased with component footprint size and was dependent on the material selection. In addition, there was a very strong correlation by design, with the non-optimized designs hav- ing the highest delta thickness to the optimized copper balance designs having the lowest. Conclusions It was shown in this study that design and material/stackup selection can be the primary Figure 14: Local coplanarity examples of local thickness variations superposition with general PCB bow/twist.