Internal voids in BGA and QFN device solder joints are not isolated defects, but the result of the coupled action of multiple process variables such as uneven Solder Paste printing thickness, blocked Flux volatilization paths during reflow, insufficient nitrogen purity, and deteriorated wettability of the PCB pad surface. Once voids form, they cannot be eliminated in subsequent processes, so control must be introduced from the front-end manufacturing steps.
Void Location Determines Failure Risk Level
We tracked X-ray images of 28 batches of QFN-48 devices on the JJY production line for three consecutive months and found that if voids are concentrated in the central area of the solder joint with a diameter greater than 75 μm, their thermal stress concentration factor increases by 2.1 times. While tiny voids (<30 μm) located at the four corners of the solder joint have limited impact on thermal conduction, they can easily induce shear fracture. This indicates that void size and location must be evaluated simultaneously, rather than only looking at the total area proportion.
Solder Paste Printing Parameters Directly Affect Void Generation Probability
When the three parameters—squeegee pressure fluctuation ±0.3 kgf, stencil aperture wall roughness Ra > 0.8 μm, and release speed below 15 mm/s—are all out of control, the solder paste transfer rate decreases by 12% with severe edge tailing, which blocks gas escape channels during reflow. The void rate surges from the normal 8% to 37%, with 62% of the voids accompanied by local bridging.
Reflow Atmosphere and PCB Status Synergistically Amplify Void Hazards
When the nitrogen dew point in the furnace is higher than -40°C and the OSP film thickness on the PCB pad exceeds 0.4 μm, the surface tension during the solder paste melting stage rises abnormally, making it difficult for flux decomposition gas to penetrate the liquid solder layer and escape. The average void volume increases by 40%. At this point, even if the peak temperature meets the standard, a large number of closed voids still exist inside the solder joint, and the actual conductive cross-section is reduced by 19%.
Actionable Void Interception Measures for the Production Line
For the daily first article inspection, 3 BGA devices must be sampled under a 200× microscope to check the wetting angle at the solder ball edge. If an obtuse angle >85° or solder paste climb height <1/3 of the solder ball diameter is observed, immediately stop the line to calibrate the stencil tension and retest the nitrogen purity. At the same time, take PCBs from the same batch to conduct a pad surface contact angle test; a contact angle >68° is judged as OSP failure, and such boards must not be put into production.
