Issue link: https://iconnect007.uberflip.com/i/989774
74 SMT007 MAGAZINE I JUNE 2018 0.5 mm and 0.4 mm BGAs. Each data point on the 0.5 mm chart represents the average of 3780 deposit readings—84 I/Os per device, 3 devices per board, three boards per panel and five panels per test. Each data point on the 0.4 mm BGA chart represents 16,200 measure- ments, as the same number of devices have 360 I/Os each. This quantity of data generates high confidence in the results. Notice the TEs are all slightly higher than 100%. This is not uncommon and can be due to numerous factors relating to gasketing breaches and/or paste pump out due to the 1:1 aperture:pad ratio and the combination of round pad with squircle aperture (Figure 5) [3]. The pads on this device are NSMD (labeled "copper" in our study), and, given this method of definition, shape differences naturally lend themselves to gasketing issues, and the 1:1 aperture:pad ratio in general is highly suscep- tible to positional errors in the stencil or PCB, alignment error in the printer, or slightly under- sized pads or oversized apertures. Figures 3 and 4 show that that all the print results were within our prescribed limits of ≥80% TE and ≤10% CV for both T4 and T5 pastes and coated and uncoated stencils. Comparing the results of T4 and T5, T5 gives very little advantage over T4, with the one slight exception being uncoated stencils on 0.4 mm BGAs with an AR of 0.62. More striking than the powder size compar- ison is the impact of coated stencils on print consistency. In every case, the coated stencil deposited similar or greater amounts of solder paste with half the variation. Cutting CV by half is a profound improvement in process control. To the printing specialist, this data should facilitate the decision between finer powder and nanocoating as means to improve release and the overall solder paste printing process. Nanocoating has a significantly more positive effect on both increasing TE and, most impor- tantly, reducing CV. Print-To-Fail The next area of focus is the TV 'Print to Fail' (PTF) patterns. The PTF patterns combine pad sizes ranging from 6 to 15 mils with shapes including circular, square and rectangular pads, defined with both copper (SMD) and solder mask (NSMD). Each data point repre- sents 480 measurements for each PTF config- uration. Figure 6 illustrates the advantage of SMD defined pads towards improving print Figure 5: "Squircle" aperture design for BGAs offers optimized solder paste release by minimizing paste sticking in corners. Figures 6a and b: SMD pads (top) produce more consistent prints than NSMD pads (bottom).