Issue link: https://iconnect007.uberflip.com/i/857644
20 The PCB Magazine • August 2017 lem, and it takes a little longer, and then you have some panels to work on. So far everything is correctable. Holden: Does the WIP system currently read barcodes? Stepinski: In the current system, when we re- lease a job to the floor, we print stickers. We don't have travelers here; stickers are coming out of a sticker printer with barcodes on them, and it tells the person at drill how many are in the stack, what material type it is, and then they build the stack and put the sticker on the stack. The drill machine reads the sticker, and then it drills the material codes into the frame. Holden: How many additional readers are throughout the system? Stepinski: We have six total—in process, in- put, and things like that. The part-specific data is at AOI at each imaging step. Then, right at the end, as we build the board, we make anoth- er barcode. If you look at finished boards you'll see there's a barcode made by a solder mask and legend, in combination. That gets read by the score machine and the router. Goldman: Nobody programs anything in; each machine reads the barcode and knows what to do. Stepinski: Yes, it's a paperless factory. Holden: What happens if something breaks? Stepinski: It's like at any other factory. If some- thing breaks, the first thing that happens is that the panels in the previous step get offloaded to the loader/unloader, and when that one fills up it goes to the next loader/unloader. It stops feeding and it cascades all the way back. Goldman: It must be difficult then, because you lose that advantage of things not sitting. Stepinski: Usually we can recover very quick- ly when we have a downtime issue. At a regu- lar shop, what happens is it just gets shoved in the corner somewhere, and then at the produc- WHELEN ENGINEERING, TWO YEARS LATER tion meeting, someone gets notified that there's a big problem. Things don't resolve themselves automatically. Usually, someone's got to push to make sure someone shows up to fix it, or changes somebody's priority to go fix it. Here, there are different line segments and I think we have seven total line segments broken up, and for each segment, all the critical alarms go into a big megaphone that we make. Whelen makes the megaphones—an emergency warn- ing megaphone. You can have it sound like a British police car, a U.S. fire truck, or whatever. All these sirens. We have seven different sounds, and we've all memorized the sounds. You hear that sound, everybody in the factory goes there. That's how we've functioned, and it's the same way in the new process. Goldman: I presume you have some sort of pre- ventive maintenance schedule so that doesn't happen very often. Stepinski: Yes, we have a very detailed preven- tative maintenance schedule. But, it's not just for breaking that we have sirens. If someone has to change a filter, we get notified as well. If the resist strip drum needs to have the resist taken out of it, the alarms go off too. All these things are what we call critical issues that require la- bor. The equipment is considered fully autono- mous, and when an alarm goes off, that means it needs some labor. Because then your process is more predictable, you've got less downtime, and your customers are happier. We can afford to do things that way, which is great. Holden: You selected copper ammonium sulfate etchant over the traditional ammonium chlo- ride. Was that specific or simply because that was what was recommended? Stepinski: This process came from a European company that built a number of these types of machines. They relocated to China, and we pur- chased from them. As received, the equipment was not what you see today. There were a lot more peripherals, and we did a lot of develop- ment work on the tool, and arrived at the current process. As received it did not have fume recy- cling, and the rinse recycling wasn't effective. We