Issue link: https://iconnect007.uberflip.com/i/1284035
38 SMT007 MAGAZINE I SEPTEMBER 2020 ations that could be done the same way if there's really no compelling reason to do a given thing differently. Holden: A couple of times, you mentioned transferring Gerber files and things like that. Gerber is a 60-year-old artifact of our indus- try that many of us wish would go away, but we can't seem to kill it. What's your opinion about data transfer and Gerber versus ODB++ and IPC-2581? Steiner: One of the things that we maintain is a separation between our fabrication require- ments and our assembly requirements. And when we generate packages for our suppliers, we generate everything to do with fabrication, which goes directly to the fabricator. We don't put the same information mixed in with our assembly stuff, so we maintain some security and privacy around trade secrets and things, such as pushing all of that information in well- documented formats that are easily understood and controlled, like Gerber. G-code is hard to mess up. If I do something more complicated like IPC-2581, I need to separate the fab from assembly. I understand I can do it, but I can separate my assembly portions and push those out to the EMS provider, and then the same is true with the fabricator. But being that we're automotive, all of our industry suppliers and customers under- stand Gerber. They don't all understand IPC- 2581; even though it's really cool, it's still not widely viewed as mainstream. And for me, I don't get to see IPC-2581 outputs in the same way I can separate that stuff and ana - lyze each output the way we've been doing in some industries, like automotive, to main- tain that known good comfort level and stan- dard. And for us, the data exchange is work- ing out fine using the old stuff. You mentioned it is a 60-year-old artifact, but I'm a 62-year- old artifact, and I'm doing fine using that old stuff! Holden: Does your organization have any road- maps regarding the smart factory and Industry 4.0 looking at the future? Steiner: CASCO is not really focused on the leading edge or Industry 4.0. That's sort of thing is left up to our EMS providers to pursue on their own. Johnson: From your perspective, do you see that situation as enabling better design for manufac- turing, or could that potentially get in the way? Steiner: I don't know because I really don't know how much our suppliers are focused on the new factory automation and controls and communications. I'm more results-oriented at the product level than the business level. I couldn't comment on how it's going or if we're suffering from not having it. Holden: One interesting thing about power inverters is that temperature and thermal is a big design issue. Steiner: There's heat everywhere. There are spots on a lot of these products that handle so much current, efficiency losses, and thermal management must be handled well, or you can end up with an area you could cook on. Our engineers work hard to produce designs that can be operated cool enough to extend that lifetime as much as you can. In designing high current power inverters, you're often up against many limitations of components or physics, in general. Locating robust components with ade- quacy in availability, cost, and junction tem- peratures can really be a challenge. Holden: A lot of those new power devices were bottom termination, especially the lead-free, that induced the whole new set of how to clean with lead-free and not have voids. That's why, a lot of times, they use the bottom termination. Steiner: And it's a challenge sometimes when we use surface mount devices. There are pros and cons. Thermal management is something that we spend a lot of our time on when we're doing a layout. I focus on making sure that we have design rules that are adequate, as well as the production capabilities that our EMS providers are going to be able to produce that