Issue link: https://iconnect007.uberflip.com/i/1213413
84 PCB007 MAGAZINE I FEBRUARY 2020 Active components, in particular, bring in a whole new cost and yield barrier that has to be met to make that product feasible. where it is important; this is not a "try it out" strategy. They only embed because it's needed. Ryder: From an OEM perspective, the decision- making process to lead you to an embedded component board or package is going to be ex- tremely well thought out. It's a big decision to embed a component. There has to be a lot at stake for this to happen, which is a testament to how young the technology still is, despite decades of activity. Active components, in par- ticular, bring in a whole new cost and yield barrier that has to be met to make that product feasible. I would argue that the use of embedded components is predominantly driven by minia- turization, reliability, or performance. In terms of miniaturization, the key benefit is newly won real estate for the PCB or substrate sur- face. Reliability may improve in some cases, as the component is not directly exposed to outer assembly stress, and the component's connec- tivity is not limited to the success of a solder joint. Regarding potential performance im- provements, one may benefit from inner layer shielding effects, short trace/wire routing, and electrical signal benefits that you get by hav- ing a package directly connect with memory, for example, versus having to be redistributed throughout the substrate layers. Johnson: The criteria that would force you to embed actives and get serious about this tech- nique, with everything else we talked about so far, screams smartphones and handheld devic- es that are going to be manufactured in very large quantities and with relatively short, use- ful lives. Cellphones don't work the way the military expects them to work, which seems to be a very natural fit. Ryder: Mobile communications is a very com- petitive market. One of the major hurdles get- ting embedded components, and especially active components, into a smartphone is the potential new cost paradigm. As long as there's still room to breathe in terms of performance and real estate within the PCB, substrate, and package design, more inexpensive options will be preferred. Hence, for proper embedded ac- tives, the cost is largely inhibitive for wide- spread adoption as yet. To my knowledge, however, there are vari- ous efforts ongoing to embed particular ASICs and MEMS, usually in the form of a packaged module. We're not talking about CPUs here, as such components can be immensely difficult to embed, due to, for example, a high level of registration and accuracy required for a high I/O count and potential thermal management issues stemming from the lack of heat channel- ing away from the chip. Johnson: In what applications are active em- bedded components being used? Ryder: It's not to say cellphones don't make use of it. Again, there are likely package modules in cellphones that contain embedded compo- nents. Johnson: It's at the point where design teams start to build up the subsystems. Ryder: Correct. I'd say that's where things are right now. From a packaging perspective, tech- nology development is rather conservative. To move crucial functionality to a fully embedded system would require a lot. That doesn't mean it's not happening. The development is there, and some of it is more mature than others, but the degree to which we see embedded actives in the market is fairly low compared to pas- sives. Eight or nine years ago, the prognoses showed a broadband ramp-up by 2020; we'd see X million embedded actives in the market