Issue link: https://iconnect007.uberflip.com/i/1011746
20 PCB007 MAGAZINE I AUGUST 2018 fications that you're getting from your custom- ers, and how does that compare to what they should be specifying to get the reliability that's needed? And it could be either way. It could be that they're specifying too high a quality to make something that's a 15- or 20-mil line and space battery charger, or where they're not specifying enough quality to meet the 2-mil line and space for an 8K monitor. Osborn: Right off the bat, I'm going to say com- munication; unless you know what those cir- cuit boards are going into and what the prod- uct is, you have no idea about reliability other than you have the spec that you built it to. And that goes back to the lack of communication. This information is not generally shared, par- ticularly with the contract manufacturers. They prefer not to tell you who the end customer is. They throw up all kinds of roadblocks and ex- cuses for why they are unable to assist once you discover that their files are corrupt, in- complete, or poorly designed. Perhaps a better word to explain it is Trust, or lack of! Colonial does have a few customers who are contract manufacturers that are willing to work closely with us to take care of the customer. Krick: I've asked certain designers and custom- ers what these parts go in and Mark is cor- rect. Literally it's a "Well if I tell you, I'd have to kill you" type of an environment. We also know here that we deal with mostly military and it could be guidance systems for missiles or going up in space. If I was an end user that was putting my signature on the approval for this part in a $35 million satellite that got off the launch pad and blew up, I'd scrutinize it pretty heavily too. This isn't 10 x 10 going in a transistor radio. This is pretty high-end stuff that has people's lives and potentially innocent people's lives resting in your hands to provide a reliable product. Can it go too far? We deal with some space flight folks where you stand back and say, "Re- ally? A tenth of a mil in etch factor and you're going to shoot me down for that?" If this is going up in a shuttle or something similar, I would like to know that the shuttle is going to make it to destination, not necessarily turn around and come right back down to earth. Knapp: I have a lot of thoughts on this from the quality point of view. The first one is, and I've had this conversation with Randy Cherry from the IPC, in many cases the soft specs have not kept up with board technology. IPC-6012 is si- lent or very vague on several things that are not middle-of-the-road technology. Another issue that we see, and I am sure every oth- er board shop sees too, is what the print notes say and what the engineer meant may not be quite the same thing. There needs to be standardization in terminology. A recent example is a board that wanted 0.062 plus or minus 10%. We kind of judged that over laminate. He judged it overall. Neither one of us was probably 100% right or 100% wrong. We didn't talk to each other, which goes back to Mark's point. We're going to have that corrected now, but we both learned something. One of the other things that occurs is there's not necessari- ly universal buy-in to using the Figure 3: Colonial's newest electrical tester, an atg A7a, utilizes eight flying probe test heads for a wide variety of advanced, high-speed electrical test techniques, with an auto-load/unload system for true "lights-out testing."