Issue link: https://iconnect007.uberflip.com/i/1481876
OCTOBER 2022 I PCB007 MAGAZINE 13 But you have to pay attention to what you're printing along too. Shaughnessy: Do you still use a yellow room? Davidson: Yes, for inner layer and outer layer all the print areas are yellow room because the resist reactivity—the millijoules to polymer- ize that resist—does not allow white light. We tried to do filtered white light a few years back, but as we went to a faster and faster resist, we had to convert to yellow. Shaughnessy: Imaging went from screening to liquid photoresist to dry film, and then to aque- ous dry film over 50 years or so. How would you describe imaging over the last five to 10 years? Davidson: When I started 35 years ago, it was all liquid printing for inner layer imaging. Until 10 years ago, we still had a liquid line here at TTM for inner layers. Liquid resist has been around a long time. e problem is the handling mech- anisms for liquid coated are very difficult, so dry film has an advantage with the Mylar cover sheet and over the last few years I've seen the increases in speed and resolution. I actually just took a call earlier today from a manufac- turer with a faster, better-resolution resist. ere's a certain point where it doesn't buy TTM Chippewa Falls anything; we don't need that speed, because of the cost differ- ence between that speed vs. what we need for our resolution here. Is it worthwhile? We have our advanced technology center about a mile down the road, and they're printing in the much finer category. ey're down into the 1 mil trace and space realm. Now their resists are vastly different than ours because of that requirement. Right now, most of our custom- ers are really hanging around 3/3 lines and spaces technology. We see a few of them push- ing a little bit under, maybe a 3-mil trace with a 2.5-mil space. You're reducing spacing, which makes wet processing much more difficult. But the traces are generally staying right around 3 mil or larger depending on the product. Most of that is for signal integrity. Most of what we build are high-speed traces. If your trace is too small, your signal gets too rough. Shaughnessy: Is LDI as automated as the old- school printers? Davidson: Right. at's one of the drawbacks of LDI. ere is some automation available, but it's not cohesive yet; there isn't just one manu- facturer making the printer and the supporting equipment, so you have to combine different manufacturers together, and get them to work with each other. But it is doable. It's just trying to get everyone to work together and play nice, so to speak. Shaughnessy: ere's not a turnkey situation where you can get the soware and printer and supporting equipment from one company. Davidson: For the most part, they work with each other relatively well, but it would be nice to have one system. Right now, there are several auto- mation suppliers that will work with Orbotech, and Orbotech will work with them. But so far I haven't seen anyone say, "Hey, this is the best match." We're a little cautious sometimes. Gen- erally, TTM won't be the first to jump into some- thing like that. ey'll be second or third. Shaughnessy: Are customers worried about geometries of the side walls? What are your thoughts on that? Davidson: Yes. We have quite a few custom- ers asking if we can improve how the traces look. roughout my years in the industry, the traces actually have a trapezoid shape, where they're a little smaller at top and a little larger on the bottom from the wet processing. For signal speed, if you can get those square, you'll get a smoother signal, less interference. On the thinner coppers, it doesn't buy you as much, and most of our customers are only