Every day, we change three types of Solder Paste, two types of Solder Wire, and one type of Solder Bar on the SMT line. We do not rely on supplier inspection reports to decide whether to place an order; instead, we check whether more than 3 out of the 23 batches of solder paste delivered by them last month showed a viscosity attenuation of over 15% after temperature recovery. This attenuation directly causes the printing release defect rate to jump from 0.8% to 3.2%. If the diameter tolerance of the solder wire exceeds ±0.02mm for two consecutive batches, the automatic wire feeding mechanism will frequently jam, forcing operators to manually intervene, and the average downtime per shift will increase by 17 minutes.

The melting point of solder bars must be measured rather than only relying on nominal values. For example, a batch of Sn63Pb37 solder bars purchased once had a nominal melting point of 183℃, but the DSC curve showed that the actual eutectic peak was as wide as 4.3℃. As a result, a 0.5℃ fine adjustment of the preheating zone temperature in wave soldering caused a large number of bridges. It was later found that the measured lead content was 36.8% instead of the standard 37%. This 0.2% deviation, combined with the difference between 99.95% and 99.99% tin purity, expanded the dispersion of the IMC layer thickness at solder joints by 2.1 times after continuous soldering of 5000 boards.

The rosin softening point of soldering Flux cannot only rely on the 120℃ data provided by the manufacturer. We must use a hot stage microscope to observe whether it starts to flow significantly after being held at 115℃ for 30 seconds. Otherwise, it is easy to accumulate at the bottom of the PCB during wave soldering, forming amber residues that are difficult to clean after cooling. A batch problem once occurred because the nominal acid value of the soldering flux was 1.8mgKOH/g but the measured value reached 2.3, resulting in micro-corrosion on the copper surface of OSP boards 48 hours after soldering, and the AOI missed detection rate reached 61%.

All solder paste must provide the measured D50 particle size distribution value for each batch; vague descriptions such as "compliant with J-STD-005" are not acceptable. We once had a 12% fluctuation in printing yield for the same model of solder paste in different months because the supplier mixed 25μm and 35μm powder. The wetting angle test of solder wire must be measured on a bare copper plate using the IPC-J-STD-002C method. Any manufacturer that claims "excellent wettability" but refuses to conduct this test will be eliminated.

The nitrogen replacement rate of the inner packaging bag of solder bars must be spot-checked on site. Last time, a certain factory claimed 99.9% but the measured value was only 92.4%, resulting in a daily increase of 0.8kg of oxide slag on the surface of the melting furnace. The excessive oxygen content in the molten tin increased the void rate of QFN device solder joints from 8% to 29%. Solder paste must be temperature-controlled throughout transportation. Once, ice packs were not used for a summer shipment, and a small amount of crystalline metal particles were found precipitated at the bottom of the can during the temperature recovery process after arrival. It was confirmed under a microscope to be SnO? agglomerates, and the entire batch was directly scrapped.