Issue link: https://iconnect007.uberflip.com/i/553274
18 The PCB Design Magazine • August 2015 feature of years at IGT in Reno, Mark Taylor. He's a four-star general in the war on failure. He introduced our PCB design workgroup to a book by Franklin Covey called "The 4 Disci- plines of Execution," or the 4DX Methodol- ogy. Our group had a really extensive check- list that was so long and time-consuming it was rarely being used. Needless to say, design problems found their way onto our manufac- turing assembly line. For a good year, our as- sembly folks tracked these errors and many more. But without an ally, the data was use- less. Mark suggested we solve our ineffective checklist problem by formulating a new one based on the data collected by the assembly group. His leadership and 4DX skills helped us to win this battle. As the designers and as- semblers allied together to eliminate designer induced failure, they found only five major show-stoppers that were slipping through the design process. The designers revamped their checklist to check for only those five things, kept score with the assembly team, and met each week to report on progress. Victory against failure was won in this case, by great leadership and team alliance. Shaughnessy: is it all about process control? is that the key? Thompson: Yes, at all levels. Process control at the design and engineering phases all the way through to fabrication and assembly. As a designer or a company going to fabrication, are you sometimes giving them too much? Is there too much for them to deal with? Are there too many caveats? Is everything 3x3, and does it need to be? Were you just ingress - ing and egressing out of a fine pitch part and then, all of a sudden, you decided to run the remainder of the board at 3x3? That's where the proof is in the pudding. That's when, ultimately, your design goes into the product and they do the initial prod- uct testing and you have to go through your first set of revisions. That's when you're really starting to see what sorts of things you need to address through that process. Certainly, there are process issues on the fabrication lev- el. They happen all the time, anything from, as you said, solder failures at an assembly due to low copper barrel cracking, heat consider- ations, etc. Any number of things can happen in a fabrication environment that could af- fect the actual physical board. One example that we talk about in our fa- cility tours is how various fabricators could do the customer a disservice. Let's say you've got an impedance control job and you're at the initial testing phase and you're running high. You're running at 55 or 56 ohms for 50 ohms, so you're just out of tolerance. Now, your solder mask guy is going to say, "Give me those boards back. I can throw another coat on that and I can take it back down to 50 ohms." That is true. Now, the board comes through as another rev, it's the same imped - ance, and you build it right this time. You have no etching issues and you don't have to put a second coat on it. The customer has two sets of boards in their hands. They both meet 50 ohms per- formance, but they have slightly different performance characteristics. Why? Because the shop played that little trick. Again, you need to be diligent about that kind of stuff. How do we do that? Mainly through site