Issue link: https://iconnect007.uberflip.com/i/873992
20 The PCB Design Magazine • September 2017 things that we couldn't manufacture or manu- facture with low yield, that we were defeating the purpose. I asked when design was going to embrace total quality manufacturing. The vice president said, "Don't hold your breath on that." But we didn't have that problem in integrated circuits because we talked to the integrated circuit man- ufacturer. He said, "We do all the integrated cir- cuit design. We do not allow anybody except us to design an integrated circuit that goes into our manufacturing." They create the schematics and all the specs and things like that. We do the physical IC lay- out. He said, "I never could understand, since we make all the circuit boards, why we aren't designing the circuit boards, so I'm going to promote you to be the design services manag- er," which is not what I was complaining about. He told me to fix the problem. I said, "Well, the first thing is that our printed circuit design- ers aren't paid very well. They're not trained. They don't have the right software tool. They don't have the ability to do research." He said, "Then give me the budget. I'll ap- prove all of it." The VP said, "And I'd like to see the printed circuit designers put on the same salary level as the IC designers so we can rotate them." I told the VP how much PCB designers typi- cally made, and he said, "That would essentially double their salary." I said that was great. If I doubled their salary, then they were going to complain a lot less when I demanded that they go back to school. I also demanded that they all use the same design process so that we had a process that we could constantly improve. If each person has his own design process, there's no way to create tools and get synergy out of it. Unlike other people, I got the budget and the salary authorization to elevate printed cir- cuit design to where it should be and the bud- gets to get the best tools and write my own scripts and my own software and standardize as per Six Sigma or quality manufacturing pro- cess. They all complained about using the same process, since they were inventing the process by consensus, like this figure of merit. We also brainstormed and came up with these figures. How do I measure your performance? I'm your boss now. How do I know that you deserve a raise? What am I going to measure with? Right now, you say schedule, but that doesn't talk about quality. I insisted on a quality metric that they can put on each board they designed, so I had a way to compare it. Who needed more training? Who needed more skill? Matties: It's like making your designers the su- perstars in the company pays off. Holden: In the HP product division, everybody in R&D was ranked according to their contri- bution. You had the PhD gurus who invented wiz bang products in HP that get ranked high. Somebody had to be ranked low and it was al- ways the PC layout people. If you're paying on a bell-shaped curve and everything, that means that PC designers are always the tail of this. When we moved printed circuit design to manufacturing they became the most important engineers we had in manufacturing, because they determined our delivery, our profitability, our yields, our efficiency. They went from the bottom tail to the top tail of the distribution of contributions, because we valued them more than anything else. This is something we learned from our Japa- nese partners. In Japan we were surprised to find that a lot of the PC boards are being designed by print circuit manufacturers. You had to work in manufacturing for at least five years before you could get the opportunity to move up into design. They didn't want anybody designing a PREDICTIVE ENGINEERING: HAPPY HOLDEN DISCUSSES TRUE DFM " Well, the first thing is that our printed circuit designers aren't paid very well. They're not trained. They don't have the right software tool. They don't have the ability to do research. "