Estimated reading time: 4 minutes

Maggie Benson's Journey: The Power of Pareto

Editor’s note: Indium Corporation’s Ron Lasky continues this series of columns about Maggie Benson, a fictional character, to demonstrate continuous improvement and education in SMT assembly. 

Why Was Patty Scowling?
Patty had just received a phone call from Mike Madigan, CEO of ACME Electronics and her former employer.

She still consulted with them. Madigan had previously expressed to Patty that he suspected a vendor was averaging Cpks from two partial lots to get a higher overall Cpk. It seemed to be happening again. Mike remembered that the scheme looked like this:

The vendor had many components that did not meet the required Cpk, for example, 1.0. The vendor would sort some components and form a partial lot that had a resulting Cpk of 1.5 and the remaining parts in the original lot might have a Cpk of 0.8. The vendor would then recombine the components and claim the Cpk was 1.15 [(0.8+1.5)/2 = 1.15]. Mike remembered that this approach was wrong, but couldn’t remember why. Patty had written a report¹ on this type of dishonest scheme and was now looking it up on her computer to send to Mike. Bottom line: you can’t average Cpks. Doing so is meaningless.

Two weeks later…

Patty was waiting expectantly in her spacious office to meet with the Benson Electronics (BE) team. She knew they were working hard on improving productivity and first pass yields, and she was anxious to see the results. Maggie, John, Chuck Tower, and Frank Emory entered the office. 

“Pro-,” Maggie started, then said, enthusiastically, “Patty, I mean, it’s so good to see you.”

“I’ve been on pins and needles waiting to hear how you made out with all of your improvement plans,” Patty said to the group.

“Let Chuck explain,” John said.

“Well, the improved uptime continued at a little over 45%,” Chuck began, “and nearly everyone is involved. Not only is uptime high, but the morale and camaraderie are remarkable.”

“Profitability has skyrocketed even though we have been the low bidder on several contracts. Maggie and John are even considering a late mid-year bonus in addition to the Christmas bonus,” Frank chimed in.

“I understand that you have been working on first pass yields too,” Patty mentioned.

“Let me take that one on,” John said. “For years, we repaired the boards coming off the line before tallying first pass yields. Because of this, we never developed continuous improvement² action plans to improve yields. In the past six weeks or so, we have collected the failure modes before repairing the boards. The Pareto chart shows the yield fallout by failure mode type,” he continued. 

“Many people don’t recognize how powerful using a Pareto chart can be in understanding the data. It shows where to focus,” Patty commented.

“Once we saw what the failure modes were, we developed action plans to improve the yields. First pass yield was 95% and now is up to 98%,” Maggie added.

“We were all surprised at how many head-in-pillow defects there were. After chatting with our solder paste vendor, they suggested a different paste that was developed to reduce this defect. The new solder paste virtually eliminated the head-in-pillow defects,” Chuck pointed out.

“In addition, a solder defects book³ created by Indium Corporation really helped us to reduce the defects. It’s free and can be obtained at I-Connect007. Indium Corporation developed webinars on some of the topics, and even offered to come onsite for in-person workshops, if requested,” John said.

“We also found the solder defects book’s advice on reducing “insufficients” to be helpful,” Maggie said. “I’m a little embarrassed to say that many of our stencils had BGA area ratios⁴ of less than 0.66. Apparently, in the past, BE used stencils that were quite thick, and didn’t decrease the thickness on some as lead spacings became closer,” Maggie said.

There was a lull and it appeared as though discussions were wrapping up, until Frank chuckled.

“As an MBA candidate, I was embarrassed to find out that I was the only one of us that didn’t know that the 80/20 rule came from Pareto and his charts,” he said, jokingly.

“As soon as I understood “the rule,” I ran to the sales department and asked Charlie Becker (head of sales, the editor) if it applied to BE sales,” Frank said. “Charlie said that almost exactly 80% of BE’s cash sales come from 20% of BE’s customers. A perfect example.”

Chuck was feeling more comfortable with the executive team of BE. So, out of character, he said, “You know, Frank, I just have a two-year tech degree, you are almost an MBA, and even I knew the 80/20 rule.”

Everyone burst into laughter. John even tapped Chuck’s shoulder as a sign of acceptance.

With this moment of levity, it appeared the meeting was over. But then Maggie added, “One more thing. We are planning on buying a competitor, and we would like any suggestions you might offer, Patty.” Maggie was feeling felt quite proud that she had finally remembered to call her former professor by her first name.

“As long as it is not Ivy Electronics, you should be OK,” Patty replied.

“But it is Ivy Electronics,” John said dejectedly.

Why is Patty down on Ivy Electronics?

Should Maggie and John not buy it?

And what does a Pareto Chart of your first pass yields show? What is your resulting continuous improvement action plan?

References

1. “It’s Wrong to Average or Add Cp and Cpk,” by Ronald C. Lasky, Dec. 10, 2010.
2. “Continual improvement process,” Wikipedia.com.
3. The Printed Circuit Assembler’s Guide to… Solder Defects, by Christopher Nash and Ronald C. Lasky.
4. “Process Guidelines to Ensure Optimal SMT Electronics Assembly,” by Ed Briggs and Ronald C. Lasky, Indium Corporation.

This column originally appeared in the December 2021 issue of SMT007 Magazine.