Productivity and solder voids

The news about the first decline in U.S. productivity came at an interesting time for me. The afternoon before the government announced a 0.9 percent drop in productivity, I had an interesting discussion with David Hillman of Rockwell Collins.

Hillman was discussing the difficulties he faced while researching solder voids in BGAs. He and his team were trying to see how large voids could be before they created problems, so they set out to produce a run of boards that had solder voids.

Production people initially thought he was a bit off his rocker to want solder voids, but they finally understood the project’s goals and gave it their best effort. Unfortunately, I guess, they couldn’t make boards that had faults in the solder connections. “In a run of 229,000 solder joints, we only got 10 voids,” said Hillman, who will be discussing solder voids during IPC Electronics Midwest in Rosemont, Ill., Sept. 28-30.

That seemed to provide intriguing insight into the productivity decline announced by the Labor Dept. Productivity gains have been so steady that failures are difficult to find. Companies with one or two part per million defect levels for high volume parts like semiconductors are running manufacturing lines at near perfection.

When workers can’t figure out how to produce a void, it’s obvious that they’re not spending a lot of time trying to prevent that defect. That’s a good sign for reliability and quality control.

There are undoubtedly a lot of fields where there’s still plenty of room for improvement, particularly some of the service industries I’ve recently interacted with. But here in electronics, improvements are becoming hard to come by. That’s the good news, unless of course you’re trying to do research on potential problem areas like voids.

2 Comments

  1. C. Blair
    Posted August 23, 2010 at 11:09 am | Permalink

    Interesting. JEDEC has been working on a pre-SMT Solder Void standard to complement the post-SMT criteria of IPC. The task group, led by Intel, did some trials that produced intentional voids in the BGAs. The TG did not have difficulty producing parts with voids. The work also included a comparison of void sizes post-SMT.

    • Kim Sterling
      Posted August 26, 2010 at 9:42 am | Permalink

      Thanks Christine. Appreciate the input.


Post a Comment

Required fields are marked *

*
*

%d bloggers like this: