Issue link: https://iconnect007.uberflip.com/i/1337116
FEBRUARY 2021 I DESIGN007 MAGAZINE 21 Watson: If you do, I would go back to the five whys. Matties: But the odds of missing it have to be reduced. Watson: Yes, it has now decreased consider- ably. What you're pulling out of this is that your "minus one" is risk. You're pulling risk out. You're not just creating a component and throwing it against the wall and hoping it sticks. at's the real concern with that. e other thing is that with some of the comments I've heard about components, for example, we get replies from people and they say, "It's close enough, the component, the model. It's pretty good." But is it correct? Just because you're able to put a compo- nent in a spot on the board does not mean the footprint was correct. at just means you got lucky. ere are situations where footprints have been wrong and aer a time those prob- lems show up. Let's say you had a wrong pad size that caused a short or less solder. Aer a time, now you're seeing problems introduced and issues that come up with that footprint. But a lot of times, the mentality is, "We got it. We made it." Matties: But it's still a science and not an art, correct? Watson: I would think so, yes. Matties: Again, if you follow the process. Watson: If you follow the process, yes. Happy Holden: I was involved in design with a company that had been doing it for 50 or 60 years like I was, and nobody was a librarian out of 300 or 400 PC designers, and that's because they learned the hard way that components are so important and that components have a life. It was better to have corporate engineer- ing create all the libraries worldwide, includ- ing the expensive models and things like that, but also to be tied in very tightly with corpo- rate quality and field service. Once a compo- nent was approved, and the model and the footprint were generated, it was tied to the purchase of that part and put away in a catalog so that design engineers could select from the catalog, and specifically in that catalog was the cost and the field service quality. Bad parts that were coming back from the field, not due to manufacturing errors, but the component itself were highlighted in those lists to stay away from these "known bad actors." Watson: Right. As I said, the librarian is the most important person on my team. Matties: John, this has been a very informative conversation. Any final thoughts to this con- versation? Watson: I would say that whenever there's a problem or issue like wrong footprints, you're going to have to take a step back, take a deep breath, look at it in a more objective way and ask yourself this one question: What went wrong and why? What happened here? Why did it happen? Be honest with yourself. I mean, people should be able to say, "Yes, the buck stops with me. I made a mistake." And then deal with those mistakes and issues and move forward. And watch out with connectors. Your biggest problem will be with connectors, because the datasheet does not tell you how the pin num- berings are, whether they are one down the row or they're alternating, and that actually will cause major problems when you get into fabrication. Shaughnessy: is was really good. anks, John. Watson: Guys, thank you very much. DESIGN007