Let's start with the observed phenomena. Less than half a minute after Solder Paste printing, the edges of the pads begin to slump, the solder paste profile under fine-pitch QFNs looks blurred, slight bridging appears between 0.4mm pitch SOP pads, and height measurement with a 2D vision system shows the thickness is 18%–25% lower than the theoretical value. This is not a solder paste batch issue; at least three links in the process chain have loosened simultaneously.

Preparation is not just for show. The constant temperature and humidity room must be controlled at 22±2℃, 45%±5%RH. The thermo-hygrometer must not be placed against the wall; it must be suspended 30cm directly above the printer feed inlet for real-time readings. After being taken out of the refrigerator, the solder paste must be allowed to warm up for 4 hours. This is not a vague "natural warming" statement; you must use an infrared temperature gun to measure the center of the paste and confirm it is 20–24℃ before opening. Once opened, it must be used up within 2 hours. Even if it is only 13 minutes over time, the viscosity will drop by more than 8%.

Among the core parameters, viscosity is the most critical. Lead-free solder paste has a nominal viscosity of 750–950 Pa·s, but actual measurement must be performed using a Brookfield LVDV-II+ with an SC4-21 spindle at 20rpm. If it is below 780, stop using it directly. Whether to choose a stencil thickness of 0.12mm or 0.15mm does not depend on the BOM, but on the pad aspect ratio. Use 0.12mm for 0805 and above; for 0402 and below, you must use 0.10mm, otherwise the shear force during demolding will pull the paste apart instantly. Apertures cannot be copied 1:1 from the pads. For QFP types, shrink by 0.015mm; for BGA pads, you must add a 0.008mm chamfer compensation.

In the operating steps, the squeegee angle is the hidden killer. 55° is the bottom line, 62° is the red line, and 58°±0.5° is the truly safe zone. For every 1° deviation in angle, the solder deposition volume fluctuates by 3.7%. Pressure, on the other hand, does not need to be obsessively fixed to a value; after scraping the back of the stencil, it is only qualified if there is no residual paste, no stringing, and no reflective oil film on the mesh surface. The demolding speed must be strictly controlled at 25–40 mm/s. Below 25mm/s, demolding stringing occurs; above 40mm/s, the bottom of the solder paste will be torn, and a sawtooth-shaped fracture at the root of the paste can be seen under a microscope.

Do not beat around the bush when troubleshooting abnormalities. First take three photos: immediately after printing, after 30 seconds, and after 60 seconds. If slumping occurs at 30 seconds, focus on checking viscosity and temperature/humidity; if slumping only occurs at 60 seconds, it is most likely due to insufficient stencil support or squeegee wear. Inspect the squeegee edge with a 30x magnifying glass; if there are continuous burrs longer than 2mm, replace it. Do not believe the claim that "it can last another two days." The stencil tension must be ≥35N/cm². Use a tension meter to measure at the four corners and the center. If the difference exceeds 2N/cm², re-tension immediately.

Experience and techniques are all bought with blood. At 3 AM in winter, when the workshop air conditioning is dormant, the surface temperature of the warming solder paste is 21℃, but the core is only 17.3℃. If you force printing at this time, the slump rate will soar to 67%. The solution is to soak the solder paste jar in a 22℃ water bath for 15 minutes, then re-measure the core temperature. There is also a counterintuitive point: the solder paste mixing time is not the longer the better. Set the planetary mixer to 30 seconds, pause for 10 seconds, then another 30 seconds, for a total maximum of 70 seconds. Overtime will generate bubbles, which rupture during printing shear and directly cause local slumping.

Let me add one more hard indicator. For the same jar of solder paste, shake it well before printing and let it stand for 2 minutes. Measure the viscosity once each at the surface, middle, and bottom layers. If the difference among the three is >40 Pa·s, it means the paste has separated, and this jar must be scrapped. Do not save this material; the subsequent losses from component placement defects will be enough to buy three new jars of paste. One final reminder: slumping is not an independent failure. It often coexists with solder balls and insufficient solder. When you see slumping, stop the line immediately; do not wait for the AOI alarm before taking action.