Issue link: https://iconnect007.uberflip.com/i/1441485
JANUARY 2022 I DESIGN007 MAGAZINE 17 high voltages and currents, and in conjunction you're going along with higher switching fre- quencies. But all of that is still, I wouldn't say in its infancy, but in a growth stage that hasn't attained the level of maturity as something like silicon-based MOSFETs. I would say it's prob- ably weak points. Maybe less so with passives, but not 100% sure; they're probably a little bet- ter suited. Obviously, your mileage may vary. Matties: When you're designing your rules, what communication do you have with mate- rial suppliers, the base materials; are they a part of your engineering and discussions? Buja: When it comes to materials, we're pri- marily concerned about just the standard loss tangent, dielectric constant, thickness, and weight. We're not really partnered with any direct manufacturers to get any inside informa- tion on that. For example, about 10 years ago, we had high-speed materials. We had to make sure we had special routing tools that would route on a diagonal, just so we didn't have any alignment going through the weave of the material. What happened? at material was no longer needed. Everything was that way. It was a short-lived manufacturing crisis that a lot of people were very concerned about, and Zuken developed unique routing func- tions just to deal with that type of frequency requirement. Happy Holden: I worked with high-voltage boards at HP in 1971 or '72. Back then, we had boards that operated at 100,000 volts because we made the field emission X-ray machines, and field emission is a totally different prin- ciple. Some of the machines go up to 1 mil- lion volts, but the thing is, aer you get past a couple hundred volts on the circuit board, mechanically created transformers and devices like that never see another circuit board again. en it's a matter of installing the cables, etc. Interestingly enough, we couldn't use FR-4. For those boards, we had to design multilayer and double-sided using phenolic. In the early 1970s, the epoxy and the glass would break down, but phenolic wouldn't, and that was terrible because paper phenolic-type circuit boards were just a bugger to drill. Unfortunately, automobiles will have five- and six-year warranties while we expect a 15-year lifecycle. I don't know what happens at those voltages with conditions of automobiles for a 15-year life. It may exceed anything that we know of in the epoxy fiberglass, and it'll have to be a little more exotic. Buja: It's interesting you bring that up, Happy. What about the longevity of EVs? I am not the type of person who likes to buy brand new cars, and when it comes to electric vehicles, what's the used car market going to look like? Who will be responsible for giving you that creature comfort of getting a fresh battery when you buy a used car? Warrier: We must have a market which involves replacement of the battery, right? ere are companies in the UK running the Nissan Leaf for a while and they've been opening the bat- tery pack and taking the ones that are in good condition, because it all comes down to bat- tery management. ey've taken the batteries that are at half-life, putting that together, mak- ing a new battery pack and it's costing maybe $5,000 less. Andy [Buja] mentioned earlier that the sep- aration needs to happen where you'll know I am not the type of person who likes to buy brand new cars, and when it comes to electric vehicles, what's the used car market going to look like?