Issue link: https://iconnect007.uberflip.com/i/834021
June 2017 • The PCB Magazine 55 years ago, say at 25 ohms per square cm, would be 0.4 microns thick; now it's down to 0.2 mi- crons and on its way down to 0.15 microns. The only difference is the surfaces are very smooth and it's driving down resistivity. At the same time the question is: how do you im- prove the bond strength? Because when you get these very smooth surfaces, the bond strength starts dropping and then you're subjecting the PCB to thermal shock and so forth. The low- er bond strength then turns into a reliability is- sue, so these two worlds are kind of colliding. The high-speed digital is getting faster and so is the RF. Basically the bottom line is we all know about Moore's Law, but there's also this frequen- cy thing. If frequency is increasing, it is inevita- bly driving down the thicknesses of these em- bedded layers because of the need for smoother and smoother conductors. And that's where we are now. We're look- ing at copper foils that we're going to be plat- ing on and we're looking at chemical solutions, because it's not going to be mechanical, for im- proving bonds so that we can use these very smooth conductors. We have an R&D program now directed at that, and it's what we think is the next big thing coming along. Meanwhile we are very busy; 2015 was our record year. There was a slight downtick last year, but this year we think we'll beat that. Certainly, year-to-date, we're in record territo- ry in terms of business, mostly driven by the combination of aerospace defense increases and the commercial has also increased. It's not that they're going to sell that many more cell- phones, but we're in the China market as well as the U.S. market, and they're using more micro- phones per phone. Plus they're using more lay- ers, like going from one to two, but to go from one to two layers per PCB doubles it right there. We're really seeing a huge upswing in these ap- plications. Matties: For the increase, if I'm hearing you right, you're saying that people are coming in and doubling up on your material in the same application for the same board? Brandler: Yes, two things are happening: more microphones for noise cancellation to improve sound fidelity; and second, more layers for the same reason of resistive layers and capacitors. And we're still pushing on the combined prod- uct I mentioned, although most of the capaci- tive layers and resistive layers are separate. But what you must have is both capacitors and re- sistors to create a filter and part of the problem of miniaturization is, if you look at a laptop, the high-speed end of it is rather far away from whatever audio stuff is going on, and the digi- tal is in the middle. But as you then squeeze it down to the cellphone size, things get a lot clos- er together, and if you squeeze it down further to an ear bud, now they're right on top of each other. And all that RF interference is going right through the digital circuits. You must have a way of dealing with that and one of the ways is to put the capacitive layer right below the resistive layer to elimi- nate noise. This is where signal integrity, this is where the fact that things are all on top of each other at very high frequencies, there's all this crosstalk and parasitics going on that must be dealt with. As a result, the inevitable drive for higher densities and higher frequencies has caused the need for more applications of these embedded paths and layers, both resistive and conductive and capacitive. So it's really a design thing. I would say the driver for us now is high- er frequency, smoother copper. There are also some issues in the supply chain. The standard coppers are available in the States, but some of these ultra-smooth coppers are not made here. We have to go to Asia to get some of these su- " This is where signal integrity, this is where the fact that things are all on top of each other at very high frequencies, there's all this crosstalk and parasitics going on that must be dealt with. " A DEEP LOOK INTO EMBEDDED TECHNOLOGY