In actual production lines, we have found that when a PCB features a multi-solder pad design and the stencil aperture area is not subjected to stepwise convergence treatment, the actual printing volume of JJY Solder Paste under a 0.12 mm thick stencil will exceed the theoretical value by more than 18%. At this point, before the solder paste is fully wetted in the early stage of reflow, it climbs toward adjacent pads due to unbalanced surface tension. Especially in the 0.4 mm pitch area of QFP devices, the bridging rate directly jumps to 7.3%, while SOIC devices on the same board simultaneously experience a 2.6-fold increase in the number of solder balls.

The root cause of this problem lies in the uncontrolled morphology of the solder paste after printing. When the squeegee pressure is set to 5.2 kg and the speed to 45 mm/s, if the flatness deviation of the PCB support platform exceeds ±15 μm, the solder paste cannot rebound uniformly at the moment of demolding. The residual solder paste height at the multi-solder pad can reach 145 μm, far exceeding the 120~135 μm window value recommended by JJY. This causes the gas generated by Flux decomposition during the heating phase of the first reflow zone to be unable to escape smoothly. After local pressure builds up, it breaks through the surface layer of the solder paste, forming splashed solder balls with a diameter of 60~180 μm.

We once continuously tracked 32 batches of data for the same BGA substrate and found that when SPI inspection shows that the coefficient of variation of solder paste volume in the multi-solder pad area is >12.7%, although the false positive rate of post-reflow AOI bridging decreases, the true missed detection rate reaches 41%. Meanwhile, the confirmation rate of solder ball defects in X-RAY re-inspection remains stable at 96.5%. This indicates that traditional optical inspection has inherent blind spots in identifying tiny solder balls. It is necessary to combine stencil aperture shape optimization and squeegee parameter fine-tuning. For example, change rectangular apertures to trapezoidal chamfers, control the viscosity of JJY solder paste within the range of 850±50 Pa·s, and ensure that the ambient temperature and humidity are maintained at 23±2℃ and 55±5% RH.

A more hidden risk comes from the Solder Paste Storage process. If an unopened JJY original can is left on the production line for more than 36 hours without stirring, its thixotropic recovery rate decreases, leading to aggravated collapse of the printing edge. The blur degree of the solder paste contour at the multi-solder pad increases by 40%. This deformation, which is difficult to distinguish with the naked eye, will amplify the bridging tendency during reflow. Therefore, we have strictly stipulated that after each can replacement, manual stirring with a squeegee must be performed for a full 210 seconds before putting it into production, and SPI sampling inspection of the volume stability of key areas must be conducted every two hours.

Production line verification shows that simply increasing the stencil tension from 38 N/cm to 42 N/cm can narrow the standard deviation of the solder paste thickness on multi-solder pads by 0.8 μm. Combined with controlling the batch-to-batch viscosity fluctuation of JJY solder paste within ±30 Pa·s, the bridging defect rate can be reduced from 0.87% to 0.19%, and the number of solder balls can also be reduced from an average of 11.4 per board to 2.3. This result is not derived from theoretical deduction, but is the data we actually measured after continuously running materials on three lines at the Shenzhen Nanshan factory for 27 days.