Solder ball defects are particularly prominent in the placement of 0201 and smaller package devices. The root cause lies in Solder Paste being entrained and splashed by Flux vapor during the reflow phase. Unreasonable stencil aperture design, excessively high oxide content or overly broad particle size distribution of the solder paste itself, and uncontrolled heating rate before the liquidus in the reflow profile often occur in combination. We verified on Jia Jinyuan F750 series solder paste that when the stencil thickness was reduced from 0.12 mm to 0.10 mm and the apertures were expanded outward at a 1:1.15 ratio, the solder ball occurrence rate decreased by 63%. The prerequisite is that the squeegee pressure must be adjusted synchronously from 4.8 kg to 3.2 kg; otherwise, edge tearing will instead aggravate splashing.

Stencil apertures must match the rheological properties of the solder paste

We do not adopt a uniform 1:1.05 aperture rule. Instead, based on the measured viscosity decay curve of Jia Jinyuan F750 solder paste, we apply a 1:1.15 rectangular outward expansion to the apertures for 0201 components under a 0.10 mm thick stencil, while polishing the inner wall to Ra ≤ 0.2 μm. This ensures smooth release and avoids solder paste residue on the aperture walls. If a 0.12 mm stencil is still used, even expanding to 1:1.2 will hardly suppress solder balls, because excess solder paste will inevitably overflow to both sides after thermal expansion in the early stage of reflow.

Solder paste selection cannot rely solely on Tg and activity

Jia Jinyuan F750 solder paste has a metal content of 89.2%, an average particle size of 15 μm, and a D90 of 28 μm. This combination maintains a stable printing yield of 99.3% under a 0.10 mm stencil. However, if switched to a competing solder paste with a nominal 88.5% metal content and a D90 of 35 μm, even with an identical reflow profile, the number of solder balls still increases by 2.4 times. This indicates that larger particles are more prone to fracture under shear force and form independent molten spheres, while excessively high metal content further aggravates thixotropy imbalance. Therefore, selection must be based on measured particle size distribution and actual printing height CV values, rather than only referencing the manufacturer's TDS parameters.

The reflow profile must be set in conjunction with the stencil and solder paste combination

We have eliminated generic template-based reflow profiles. For the 0.10 mm stencil + Jia Jinyuan F750 combination, we strictly control the slope of the 150°C to 183°C heating segment at 0.8°C/s, and maintain the soaking zone at 175 ± 2°C for 90 seconds. The purpose is to allow the flux to fully release low-boiling-point solvents without causing violent vaporization. If the slope exceeds 1.0°C/s, the rupture energy of micro-bubbles inside the solder paste will drive liquid tin to spray toward the pad edges. At this point, even lowering the peak temperature will be of no avail, because solder balls have already completed nucleation before the liquidus.

All adjustments must be verified synchronously. Optimizing any single step alone may introduce new defects. For example, only refining the solder paste particle size without reducing the stencil thickness will lead to insufficient printing height and subsequent cold solder joints, while only lowering the reflow heating rate without improving the solder paste oxide content will increase the proportion of poor wetting. These correlations must be confirmed through cross-comparison of AOI and cross-section data after 48 hours of continuous production line running to verify the actual effect.