Issue link: https://iconnect007.uberflip.com/i/1011746
22 PCB007 MAGAZINE I AUGUST 2018 soft specs as they're currently written, and the best example is that there's a lot of work that we do for space flight. We're up to IPC-6012DS, with an S for space flight addendum. But I'd say close to half of the space flight work that we do is specified to 6012B. That's about two generations back. And they're brand new de- signs. There's obviously some discomfort out there with the specs, with the requirements, and with the designs. Ultimately, quality is: what does the customer need? But conveying exactly what he needs may not be universal here. Osborn: Just to add to what Kevin said, even though the via fill process has been around for several years, not all design- ers are on the same page. We still see new designs speci- fy via fill holes with no men- tion of cap plating. In the past that is exactly how we manu- factured the boards until the customer experienced mas- sive fallout during assembly and suddenly we're open to discussion. It is amazing how much progress can be made when you sit down and talk. Feinberg: Mark, that brings up a very good point. Standards and specifications, gee, don't we have the IPC for that? Do we need a standard or specification on things like filling the via? Goldman: Remember that IPC is just a facilita- tor for standards; it's the people from compa- nies in those meetings and who are develop- ing the standards. But the lag time in specifi- cations is not a new or unrecognized problem. The trouble with industry standards, though, is they take time to develop and review, and in the end, everybody must agree, there needs to be a consensus. Feinberg: I have two other basic questions for you guys. We're seeing a tremendous advance- ment in computerization of process control, computerization of manufacturing with quan- tum level computing coming. Are you starting to see any higher role in computerization of process control at your own company? Knapp: Computerization makes process control a whole lot easier in the fact that usually there is a key or a password that the guy setting up the process set it up that way and nobody else has a chance to change anything. In days gone by, everything had knobs and toggle switch- es and you had to rely on the operator to be enough of a craftsman and have enough integ- rity to do things correctly. You don't have a lot of those people anymore. You must basically use computerization to lock out the happy fin- gers which allows you to have process control. It doesn't necessarily enhance it. It just allows you to have it. It's the consistency. Back in our younger days, I remem- ber if you had three shifts, you had three etchers. They would all give you an ade- quate product, but they would all go about it just a little dif- ferent. If you had three shifts, you had three copper platers. They all gave you a decent product but they all may have gone about it a little differently. But there was a craftsmanship there. It isn't here anymore. You must maintain that consistency. And what we considered consistent 20 years ago is going to give you nothing but scrap today. It's just so much tighter, but you must have that, and I don't think you can rely on that from an indi- vidual point of view anymore. Krick: I think that for the volume that we do here and the small fortune we avoid in having processes and/or equipment that requires that level of technology, I'd rather have 25 well- trained craftsmen to build it. We're not a bath- tub shop, by any means, but we pull off what we pull off because of time and tenure from the people on the floor. Even our plating line is controlled by a program, but they still need Kevin Knapp