Rework soldering is not a simple process of reheating and recooling, but a comprehensive response to the original PCB's thermal stress distribution, the thickness of intermetallic compounds on the pads, and the heat capacity differences of component packages. We have handled over 2300 8-layer board BGA rework cases in the JJY SMT workshop, and found that 92% of cold solder joint defects stem from pad warpage caused by a preheating ramp-up rate exceeding 3.2℃/s.

Hot air system parameters must be dynamically matched to PCB thickness

When using the JJY JRY-850A rework station for 6-layer boards, the upper hot air nozzle should be set to 280℃/45L/min, and the lower preheating platform set to 145℃. For 12-layer thick copper boards, the lower temperature must be increased to 165℃ and the preheating time extended to 180 seconds. Otherwise, when the BGA bottom solder balls detach, excessive temperature difference between the top and bottom will cause pad peeling. This failure manifests as a sudden increase in void rate on one side of the solder joints by over 47% under X-ray inspection.

Stencil aperture dimensional error directly determines Solder Paste filling consistency

We insist on using a 0.12mm thick laser stencil with the same layout as the original SMT production line, whose aperture size is enlarged by 8% relative to the BGA solder ball diameter. If the stencil is arbitrarily changed to a 12μm electroformed stencil or the aperture is expanded to +12%, the probability of solder bridging after rework will increase by 3.8 times. Especially on 0.4mm pitch QFN devices, bridging is often accompanied by bottom Flux residue, leading to subsequent ion contamination test failures.

Flux paste activity grade must closely match the removal temperature window

When removing 0.3mm pitch CSP chips, ROL0 level no-clean paste is selected, because its activation temperature peak falls in the 215℃ to 228℃ range, which exactly covers the critical wetting window in the last 3 seconds before hot air withdrawal. If ROL1 paste is mistakenly used, it will not fully activate until above 240℃, at which point the solder balls have already started to collapse and deform, ultimately causing solder joint height deviation exceeding the ±0.03mm tolerance band.

Post-soldering AOI and X-ray inspection cannot replace manual microscopic re-evaluation

AOI has a recognition rate of less than 61% for BGA bottom voids. We must use the JJY ZM-2000 microscope with a 40x bright field + polarized light combination to inspect each solder joint one by one, focusing on whether the solder ball edge presents continuous metallic luster, whether there are microcracks extending to the pad interface, and whether the IMC layer shows irregular jagged growth. These details are directly related to the failure rate of this PCBA in high-temperature and high-humidity aging tests.