When the tombstoning rate of 0.4mm pitch QFN devices on FPC flexible boards continuously reached 37%, we immediately investigated the reflow profile and stencil aperture, and found that the real bottleneck lay in the mismatch between Solder Paste printing consistency and the force control at the moment of placement head pressing down. Because the FPC substrate is only 0.1mm thick, the copper foil elongation exceeds 12%, and there is no support platform on the back, the solder paste is prone to offset under the action of the squeegee, and slightly excessive placement pressure will cause the entire pad to collapse or the copper foil to tear. Therefore, the rheological properties of the solder paste and mechanical pressure must be analyzed in a coupled manner, rather than looking at printing in isolation or adjusting pressure separately.
The difference in Solder Paste Viscosity directly determines the printing release quality. Solder paste with a viscosity of 220Pa·s (25℃) can maintain sharp edges and no tailing under a 30μm stencil, while the same model of 180Pa·s solder paste shows 12% solder paste stringing during release under the same parameters. Especially in the edge area of the FPC flexible board, due to slight substrate warpage causing local overpressure of the squeegee, the accumulated stringing solder increases the probability of bridging between 0.3mm pitch pads to 29%. However, if the viscosity is blindly increased to 260Pa·s, insufficient shear thinning will lead to poor stencil aperture filling, and 35% of pads are visible under the microscope with insufficient solder volume, resulting in an increase in cold joint rate to 21% after reflow.
The metal content affects the solder joint height and void rate after reflow. Solder paste with 88.5% metal content forms an average solder joint height of 48μm and a void area ratio ≤8% on FPC flexible boards, while solder paste with 85% metal content, although smoother in printing, has a solder joint height dropping to 36μm and a void ratio soaring to 19%. In scenarios such as automotive camera modules that require -40℃～125℃ thermal shock verification, the latter has a solder joint cracking rate of 17% after 500 cycles, while the former is only 2.3%. Therefore, for FPC flexible board welding with stringent reliability requirements, metal content must never be sacrificed for smooth printing.
Solder paste with a thixotropic index greater than 4.2 can quickly recover structural strength the moment the squeegee stops, preventing lateral flow of solder paste when the FPC flexible board is peeled off from the stencil. We conducted comparative tests on the same DEK printing machine and found that solder paste with a thixotropic index of 3.6 has a solder paste offset of 18μm at the pad center after peeling, while that with 4.5 only offsets 5μm. This gap directly leads to a 4.8-fold increase in the frequency of placement accuracy out-of-tolerance alarms in 0.35mm pitch BGA packages. Therefore, thixotropy is not an auxiliary indicator, but a key criterion for whether FPC flexible board solder paste printing can be closed-loop.
Placement pressure must be set in tiers according to the pad geometry. The maximum allowable pressure for 0.2mm×0.2mm pads is 0.15N, otherwise the copper foil depression depth exceeds 3μm, triggering the risk of solder joint fracture. 0.6mm×0.6mm pads can withstand 0.45N pressure to ensure sufficient deformation and contact of the solder paste. If uniformly set to 0.3N, the cold joint rate in small pad areas rises to 14%, while solder paste extrusion and contamination of the solder mask occur in large pad areas. As a result, the rework rate on a TWS earphone FPC double-sided mixed mounting production line jumped from 1.2% to 6.7%.
For double-sided FPC flexible boards, the presence of soldered BGA devices on the back reduces the substrate rigidity by 32% during front-side placement. If the pressure parameters for single-sided FPC are still used at this time, the tombstoning rate of 0.25mm pitch SOP devices will surge from 2.1% to 18.4%. The front-side placement pressure must be reduced to 65% of the original value, and the placement head holding time must be extended from 200ms to 350ms to allow the solder paste to complete initial anchoring under low stress. Otherwise, slight vibration before reflow is enough to trigger device displacement.
High-density blind/buried via FPC flexible boards have local rigidity mutations at the drilling positions. The reaction force in this area when the placement head presses down is 40% higher than that in the homogeneous area. If pressure compensation is not applied, the copper foil tear rate of pads around the blind vias reaches 9.3%. By using the pressure zoning mapping function to automatically reduce pressure by 22% within a radius of 1.2mm from the blind via coordinates, the tear rate drops to 0.8%. This data comes from AOI statistics on our production line for three consecutive months, not simulation results.
Therefore, the selection direction is very clear: the solder paste must be selected with Type 4 powder, 88.5% metal content, 220Pa·s viscosity, and a thixotropic index ≥4.2. The placement pressure must be set by looking up the table according to the pad size and support dynamic compensation for blind via areas. Any solution that attempts to cover all FPC flexible board types with a single parameter will pay the price of doubled rework rate in mass production.
