PCB007 Magazine


Issue link: http://iconnect007.uberflip.com/i/1116895

Contents of this Issue


Page 35 of 131

36 PCB007 MAGAZINE I MAY 2019 Johnson: Earlier, we touched on trends like batteries. Automotive is a big driver in battery technology right now, creating an immense ap- petite for copper. Batteries offer a higher po- tential for a profit margin for copper producers than foil does. At the same time, those same automotive producers are trying to put a whole bunch of electronics in the car, which is driv- ing the demand for electronics. Senese: Copper foil has been the most volatile thing that we buy over the last five years. Be- cause we are not as big as the largest suppli- ers in terms of square feet, especially the low- end kind of material, our challenge has been to drive our products toward thinner coppers. You can do that by picking the type of device that your customers make and trying to supply them. For example, almost all of the iPhones (8, 9, and 10) use a buildup process with 3-micron copper, which comes on a 2-mil sacrificial cop- per. That's how it's shipped out. However, we do not sell copper. All of the people that have to buy 3-micron copper for everything but the inner two layers of an iPhone are buying it from a copper supplier. Essentially, they're buying three ounces of copper to use a few tenths for this modified semi-additive process. The copper that they sacrifice all goes right back to the copper supplier and is recycled and used again. That dynamic has driven cost down in some- thing as high-volume as mobile phones. It has the added advantage of being easier to use for a buildup process that has very fine line spac- ing. However, when they buy that copper, it's 10 times more expensive than one-half-ounce copper. Swallowing that pill and knowing that it's going to pay off later is not the way our supply chain has been trained. This is a value proposition that has to be sold from the start when the designers are building something. In the cellphone business, it's easy to do because it's a one-year cycle. If you make a mistake for one year, you can fix it the next year. In IT and automotive businesses, you don't get a chance to make a mistake like that. Johnson: No, it would have to have a much longer life than that. Senese: Yes. You have a long development cy- cle, and then you have a longer business adop- tion cycle since everybody is more risk-averse; they don't get to reset every year. This cre- ates an issue for us. We could easily try to sell people on the idea, "If you want to save some money on copper, which is the most volatile thing, start adopting a semi-additive process." That's great if you're doing a buildup or an "any layer via" kind of thing, but if you're do- ing a traditional one or two lamination multi- layer, it doesn't work. For example, most of the IT business is looking at HDI for some of their outer layers of the line cards and backplanes. It's a possibility, and it certainly will happen at some point in time on some level of tech- nology. It's going to be part of what drives the cost of 5G devices. But we still have to have the backbone of the network built out on tra- ditional servers, routers, and switches. They can't take advantage of that technology from a fine-line spacing or cost standpoint right away, but they're all learning about this at the same time. There will be some movement over the next 10 years toward not being so highly lever- aged on a subtractive process where most of the copper is going into waste treatment. Johnson: There are a lot of decisions being made there, and those designers have a sur- prisingly large influence on everything that happens downstream. What can a designer or design team do to help with this?

Articles in this issue

Archives of this issue

view archives of PCB007 Magazine - PCB007-May2019