During mass production of a certain 0.4mm pitch QFN package product, we found that SPI inspection continuously detected low Solder Paste volume and severe contour burrs. After troubleshooting, it was confirmed that the issues were not caused by abnormal solder paste batches, squeegee pressure fluctuations, or PCB pad oxidation, but rather by the fact that the stencil aperture was changed from the original rectangular opening to an optimized version with 0.05mm single-side chamfers. This caused the adhesion between the solder paste and the aperture wall to drop too quickly during the release moment, leading to local solder paste being pulled apart and forming tiny splatters. This phenomenon is particularly prominent when using a 25μm thick stencil paired with Type 4 solder paste.

Whether solder paste can be completely transferred during the stencil printing process depends not only on the aperture area ratio (i.e., the ratio of aperture area to aperture wall surface area), but also is more significantly affected by the geometric characteristics of the aperture sidewalls. For example, when apertures are fabricated using laser etching without electropolishing treatment, the sidewall Ra value often reaches above 0.8μm. At this point, the shear resistance experienced by the solder paste during the separation stage increases significantly, especially in fine-pitch device printing, which easily causes fracture at the top of the solder paste rather than overall peeling. This ultimately results in an asymmetric forming shape on the pad, with a depression in the middle and bulges on both sides.

We once compared and verified stencils with three different aperture methods using the same BGA substrate, namely traditional vertical openings, stepped openings with a 0.1mm step at the bottom, and tapered openings with full aperture wall trapezoidal taper. The results showed that the trapezoidal taper design had the lowest solder paste volume coefficient of variation (CV) of only 6.2% on 0.3mm pitch CSP devices, while the vertical opening had a CV value as high as 11.7%. This indicates that the interfacial separation behavior between the aperture sidewall and the solder paste is not solely determined by the area ratio, but is also closely related to the coupled effects of solder paste yield strength, stencil material elastic modulus, and printing speed.

Before replacing the stencil on site, the actual dimensions of the apertures must be measured, because the heat-affected zone after laser cutting will cause micron-scale melting and re-solidification protrusions in some areas. This structural defect, which is invisible to the naked eye, will be concentratedly exposed as solder paste trailing after continuous printing of 200 pieces. Especially when the solder paste metal content is 89.5% and the viscosity is 650Pa·s, the trigger probability of such defects increases by nearly three times. Therefore, each new stencil must be sampled and inspected for cross-sectional morphology of at least 10 typical apertures using a 500x metallographic microscope before going online.

For large pad devices with a pitch of 0.5mm or above, appropriately increasing the micro-texture treatment of the aperture inner wall can actually help improve the stability of solder paste release. However, this measure is absolutely not applicable to densely arranged pins with a pitch below 0.3mm, because under the latter working conditions, any surface irregularity will amplify the tendency of solder paste edge tearing. This requires process engineers to dynamically define aperture acceptance criteria based on the specific package type, solder paste model, and equipment parameter combination, rather than applying a unified template.