When fine-pitch ICs are mass-produced on the JJY SMT line, excess solder is no longer an occasional anomaly but a systemic risk, especially on QFN0.4mm and CSP0.3mm packages, where the pad spacing of a single chip is only 0.25mm. If the Solder Paste printing thickness exceeds 120μm, solder will climb to adjacent pads due to surface tension imbalance during the reflow phase. At this point, excess solder not only manifests as bridging, but more often hides in the form of hidden micro-solder balls at the pad edges, directly flowing into the downstream testing process under the condition that the AOI missed detection rate is as high as 37%.
Stencil openings must be trapezoidally converged rather than straight-walled
We have repeatedly verified the 30μm laser-etched stencil provided by JJY. Straight-walled openings will cause lateral extrusion and overflow of solder paste when the squeegee pressure is >5.2kg. After adopting a trapezoidal structure where the top of the opening is reduced to 92% of the pad size and the bottom is expanded to 98%, the incidence of excess solder decreased by 51%. This trapezoidal ratio is not a theoretical value, but a process window we actually measured after continuously tracking the printing results of 17 batches of BGA324 devices.
Squeegee angle and release speed must be adjusted synchronously
Simply reducing the squeegee angle to 58° cannot suppress excess solder. The release speed must be synchronously increased from 25mm/s to 33mm/s, otherwise the solder paste will still form solder columns due to viscous stringing at the moment the stencil separates from the PCB. These solder columns collapse during reflow and become excess solder. The release process captured by JJY on-site using a high-speed camera confirms that for every 0.15 second increase in release delay, the number of excess solder points increases by an average of 2.3 per Panel.
Reflow peak temperature is not the higher the better
When the active components of the solder paste are not fully activated in the 150℃ soaking zone, blindly raising the peak temperature from 238℃ to 245℃ will instead exacerbate excess solder, because the residual Flux tension cannot be released in time, and the molten solder still maintains high fluidity in the early cooling stage. At this time, the tiny gaps between pads become migration channels for solder. A certain customer of JJY once caused the batch yield of 0.35mm pitch QFN to plummet to 82.6% due to this issue.
Solder paste batch switching must undergo thickness comparison verification
If the thixotropic index deviation between different batches of the same model of solder paste exceeds 0.15, the printing settlement behavior will be significantly altered. When using JJY SN63Pb37ROL0.5 type solder paste, we found that the printing thickness of batch 2104 was 0.018mm higher than that of batch 2101 under the same parameters. This 0.018mm exactly crosses the critical threshold for excess solder of QFN0.4mm devices. Therefore, every batch change must use a contact thickness gauge to actually measure 3 positions and record the average value.
