During SMT line debugging for 0.3mm pitch BGA chips, we found that bridging defects detected by AOI were concentrated between the pads of columns 3 to 7, which is highly correlated with uneven Solder Paste thickness caused by squeegee pressure fluctuations. When stencil tension decays below 38N/cm, the bridging rate surges from 1.2% to 9.7%. After JJY replaced the stencil on-site, the first-pass yield of this batch immediately rebounded to 99.1%.

The bridging risk in the solder paste printing process is strongly positively correlated with the dimensional deviation of stencil apertures. When the aperture width is 0.015mm narrower than the pad and the aspect ratio exceeds 1.8, the tendency of solder paste edge collapse intensifies. Combined with a mounting Z-axis downward pressure exceeding 0.08mm, continuous solder bridges will form between adjacent pins of 0.4mm pitch QFN. Such bridges cannot be automatically separated by surface tension after reflow, because the Flux activity is exhausted prematurely before the peak temperature.

For the SP-820 solder paste used by JJY, at 25℃ room temperature, every 500Pa·s increase in viscosity leads to a 0.13mm increase in solder paste stringing length during printing release. If the stencil bottom is not treated with nano-hydrophobic coating, the stringing ends easily bridge to adjacent pads. Such micro-bridges are wrapped by an oxide film at the early stage of reflow, eventually solidifying into irreversible electrical shorts. Local heating with a hot air gun during repair will instead expand the bridging range.

We once divided the same roll of JJY stencil into three sections for testing. Section A had a bridging rate of 4.3% after 120,000 uses; Section B dropped to 1.9% after laser repair of the aperture edges; Section C, replaced with a stepped aperture design, stabilized at 0.6%. This indicates that bridging is not caused by a single factor, but is the result of coupled failure among stencil life, solder paste rheological properties, and mounting dynamic response.

During operation of 0.3mm pitch chips, if SPI detects that the solder paste volume coefficient of variation for a single pad exceeds 12.7%, the subsequent bridging probability breaks through the threshold. At this point, it is necessary to simultaneously check whether the lateral runout of the track conveyor belt is greater than 0.025mm, because tiny vibrations amplify the Y-axis positioning error of the mounting head, causing the solder paste that should land at the pad center to actually shift to the edge, thereby inducing capillary climbing between adjacent pads.

Statistics from JJY's production line show that 73% of bridging defects occur in the first article and the first 5 boards after line changeover. This is related to the stencil temperature rise lagging behind the machine ambient temperature. When the stencil body temperature is 2.3℃ lower than the set value, the shear stress distribution of the solder paste inside the aperture becomes unbalanced, causing the solder extrusion shape to change from rounded to serrated. The serrated tips become the starting point of bridging at the moment of chip mounting, and routine inspections cannot capture this temperature difference at all.