Issue link: https://iconnect007.uberflip.com/i/1116895
MAY 2019 I PCB007 MAGAZINE 67 roads is in RF, which have been predominately dominated by PTFE. We're seeing two camps: 25-GHz applications for the side radar, and 77–79-GHz applications for forward radars. In the past, that has been dominated by the PTFE marketplace because of the consistency of the dielectric constant. In addition, the packaging in the market- place has shrunk down from what used to be a patch antenna, two-sided product, up to an 8-layer product. Due to the pitch, we're look- ing at some 0.8-millimeter and 0.65-millime- ter pitch products with multilayer PTFEs. From what we're hearing from a number of OEMs, the yield is not very good. So, they're looking for a more thermal set product that can meet the dielectric constant consistency so that they can hit their numbers for frequency drift over the temperature range. We have a number of product offerings that fit into that category. We've been making a lot of traction in that ar- ea as well. Johnson: My recent coverage of the industry shows that autonomous vehicle sensor manu- facturers are moving computational chips onto the sensors themselves to distribute the com- puter power rather than try to have a central computer do it all. Do you see the autonomous vehicle sector as a major driver for market needs? Cochrane: Cost is a big driver in that. This is a tough marketplace because they want the low- est cost and the highest performance all in one. At first, only high-end vehicles had all of the collision avoidance, lane assist, etc. But entry- level cars have them now. The price point of that product offering has dropped consider- ably, hence the move away from PTFE due to the cost of the base materials versus a ther- mal set epoxy type of material. When we look at it from a dimensional stability standpoint over the temperature range, we're having to test from -40°C to 140°C, and we have to look at what is the frequency drift throughout that whole temperature range. This is also true with humidity. We have to go from 50% to 95% humidity. Then, there's aging, which is two times what the norm used to be. It used to be 1,000 hours of aging, but now it's up to 2,000 hours. And some interesting things tend to happen to some materials through that aging process at either 80°C or 125°C. Some materials don't maintain the consistency from an electrical standpoint. I think the market is growing in that area considerably. A lot of different things are going on in the automotive industry, and they want low-modulus and high-modulus materials. They want high-modulus materials because they want the reliability for the lower Z-axis expansion, but they want low-modulus materials because they want to do a gasket sealing in something like a hermetically-sealed area so that they can prevent the degradation of the new chip-on-board applications. Then, because so many of these products are a direct chip attach, we have the XY movement of the substrate or the board that the IC is attached to, seeking less movement to match the silicon in that area. Additionally, because of the small pitch, we look at the surface profile of the thinner glass- es that we have to use. For example, TUC just put in three new treaters in a new factory in Taiwan. The focus for all three of those treaters is super low-loss materials and thin glass. Johnson: We've just touched on a number of constraints and factors that need to be weighed Alan Cochrane