The application discloses a manufacturing process for improving the
solder mask bridge falling problem of a high-layer circuit board based on an AI
server, relates to the field of PCB manufacturing technologies, and comprises a pretreatment process, a silk-
screen printing process, a pre-baking process, an
exposure process, a developing process and a high-temperature curing process. The pretreatment process adopts
volcano ash plate
grinding and super-rough cleaning. The silk-
screen printing process adopts secondary printing of oil, and the first
solder mask bridge position is blocked from oil. The
exposure process reduces the gap between the
solder mask bridge and the design PAD and
etching PAD. The developing process maintains the pH value of developing liquid at a constant value by calculating the ink usage of a single PCB and adopting the drag tank method or the production green oil
plate method, so that the problems of the prior art, such as the easy falling of the solder
mask bridge and the high defect rate, are solved.