After implementing line improvements, tiny bright spots appeared on the surface of newly made circuit boards. What was the cause? Mysteries of Science
Transcript
A factory produced many kinds of printed circuits requiring dozens of process lines. Management was instituting a new initiative to boost quality and productivity. Each department was assigned a team to determine if they were functioning efficiently.
One by one each department addressed the inefficiencies within their processes. The first process for the imaging department was the application of resist. By speeding up the coater they improved yield, although the coating did look a little rougher. The next step was hardening. The team was able to shed a second off of the exposure time. To make up for this modest improvement the team reduced the oven bake time.
These small improvements went on through each department.
The following week the line was down for a day as a chemical tank needed replenishing. Once the line started up again the inspector noticed tiny bright spots where there should be nothing but green circuit board.
What was causing these spots, how did the line improvements affect the product?
Here's the rest of the story.
Close inspection showed that the tiny spots were solder. Closer examination revealed that the solder spots were sitting on top of spots of copper foil that should have been etched away. There was only one way that solder could have gotten there. It must have come from the plating line, and it had to be getting through the resist.
The initial conclusion was that a different brand of additive in the newly made plating bath was the culprit. It must have been stronger and was seeping through weak spots in the resist. Everyone agreed except a chemist who insisted that the new additive was chemically identical to the old brand.
Upon investigation the chemist learned that they had changed a few things in the process, including reducing the resist bake time. He knew a change in the bake time would change the finish resist properties. The chemist realized that the resist was not fully hardened.
So instead of holding back the plating bath, it softened and allowed plating to occur through the resist, probably where there were small surface bubbles that made it thinner.
The processes were returned to the old settings. Everything worked just fine, and the 8% productivity savings turned out to be the real phantom. Reducing the process time spelled disaster, not productivity.
