Kevin Parmenter, Field Applications Manger, Taiwan Semiconductor
When it comes to power supplies the requirements keep going up for efficiency, performance, reliability, conformance to standards – safety and compliance, RohS, Weee, Conflict materials, operation at 5000 meters altitude with no derating in safety or performance, volumetric efficiency / power density, flexibility -configurability (the desire to buy one power supply and use it in multiple products) so economies of scale and not having to start from scratch to qualify -requalify the power system in the product. EMI- EMC performance for both conducted and radiated emissions as well as susceptibility to EMC and real-world perturbations on the incoming line with 3rd world applications needing to operate reliably from AC mains that surge and sag and have transients worse than ever before. With all this the end customer wants to pay almost nothing for the power supply – even though much if not most of the product risk – safety, EMC and so forth related to the product success relies on the power supply system itself. This leads up to what is really the impediment to optimizing all these often mutually exclusive tradeoffs? Billions have been spent on semiconductors and optimizing the switches both silicon and wide bandgap devices. What if these are at the point of diminishing returns? The suppliers of these devices often state that they are “waiting for the breakthroughs in magnetics” if they only had that the payoff for the money spent on wide bandgap device development could be realized. Yet Magnetics have been with us since the 1830’s and much of the principles were well documented by Bell labs during the vacuum tube era and the explosion in electronic technology from 1920s and during WW2. So, what is awaiting “discovery” and “breakthroughs”? Is there an undiscovered “Moore’s law” in magnetics design and materials? If so, are we really waiting on it? Or has everything about magnetics materials already been know and we are waiting for semiconductors switches, to “catch up”. When there is progress to be made, one of my favorite lines from a movie “The Right Stuff” about pilots and astronauts – “no bucks, no buck rogers” implying that management always wants to do amazing things without committing the necessary resources to making it happen – not much has changed there. If there are big breakthroughs in magnetics needed – who’s going to pay for it? The VC community won’t get excited about investment in magnetics. The Semiconductor companies are not likely to share some of the R&D dollars on ecology enablement. Magnetics companies can’t afford to. Universities do not have the industry oriented “it has to meet a timeline and work in the application with high production yields and reliability”. This is such a proximate and important topic that it will be a topic of a RAP session at APEC 2019 this year in Anaheim. Who’s going to do the work? Is magnetics really the constraint? This article poses more questions than answers and if you want to participate in the exploratory process come and participate in the RAP session at APEC https://www.apec-conf.org/rapsessions. Regardless of the outcome, the expectations for a power supply and the advancement to meet requirements are not letting up any time soon.