Electronics

Dr. Alex Lidow: The Mind Behind Power MOSFETs And The Rise Of GaN

Guests Dr. Alex Lidow | Uploaded : 12/12/2023


The EEcosystem Podcast

Dr. Alex Lidow: The Mind Behind Power MOSFETs And The Rise Of GaN

In this episode of The EEcosystem Podcast, our guest is Dr. Alex Lidow, CEO and co-founder of Efficient Power Conversion and the co-inventor of the HEXFET, power MOSFET, and eGaN. Lidow holds degrees from Caltech and Stanford. He earned a Ph.D. in applied physics from Stanford University in 1977 as a Hertz Foundation Fellow. In this episode, we will find out more about Alex and learn about MOSFETs and the rise of GaN. How did MOSFETs grow so quickly? What technologies are driving GaN adoptions and why. We will also discuss the potential obstacles to GaN adoption.

Links & Resources

EPC Website

 

GaN Design Support

 

GaN talk support forum

 

Text books and Publications

 

Connect with Alex Lidow on LinkedIn

Connect with Judy Warner on LinkedIn

 

Visit The EEcosystem Website¬†For a Free Download of a sample chapter of Ken Wyatt‚Äôs EMC ¬†troubleshooting and receive a free 90-day subscription to all of Eric Bogatin’s training at The Signal Integrity Academy

Sponsors Resources

Transcript

								 Hi Alex. Thanks so much for joining today. I'm so excited to learn from you in this conversation.

Dr. Alex Lidow: Thanks for having me.

Judy Warner: So first, before we get started, What is up with that? It looks like a Peter Bilt truck behind you. I don't see that in a MOSFET GaN guy's office. What's that about?

Dr. Alex Lidow: It's just a reminder of what's coming at me every day.

So just like to keep running faster and faster to stay ahead.

Judy Warner: I see, the speed of technology bearing down on you!

Dr. Alex Lidow: And the force of a Peter Bilt

Judy Warner: yeah, exactly. So if you don't outrun it, you're gonna get run over. Alex, it's so nice to meet you. You and I share a friend in common who is Steve Sandler, and I appreciate him introducing me to you and learning all about you.

I've done a little research ahead of our conversation, but for our audience's sake, can you please give an overview? [00:02:00] Who you are, your educational and professional background, and just give us a snapshot and then just a thumbnail of EPC, your company.

Dr. Alex Lidow: Okay. My name is Alex Lidow and I am c e o and co-founder of Efficient Power Conversion, and we make gallium nitride devices to make energy efficiently and improve energy efficiency

I got my undergraduate degree in 1975 from Caltech and my graduate degree PhD from Stanford in 1977. And started at International Rectifier in 1977 and my first project was to develop a better transistor and colleague of mine from Stanford and I developed the early power MOSFET

Judy Warner: Yes, So no wonder you're outrunning that truck still at this age.

Dr. Alex Lidow: Still at this age. Wait a minute...

Judy Warner: But you CAN still outrun it. I'm impressed. So let's talk about that, the MOSFET, like what led you,[00:03:00] what led you to that place? And I know traditionally up till that time it was a bipolar transistor. So what drove the technology and what drove your interest in that field?

Dr. Alex Lidow: So when I was a graduate student--fairly soon before I graduated a colleague of mine who was a professor there came down to my lab and we were just chatting about various things and he took off his glasses and he said, do you know what makes these glasses cost what they cost?

And I said I don't know. And he said what makes them what they cost is the buildup of energy it took to bring them to this state. Oh. And he said that energy has six elements to it. It has the cost of generating the energy, distributing it, storing it, converting it, consuming it, and cleaning up after it.

He said that that's the foundation of all of our global standard living. He says, now, geopolitics can change the [00:04:00] price that we pay for those things, but the cost of those things is very fundamental to that. So if you wanna improve the world's standard, Improve the cost of energy. Yep. And that set me on a lifelong quest to improve energy efficiency.

And the MOSFET was frankly the thing that I could see at the time had the greatest opportunity. And in fact, over time, I think it's been proved to have saved about 30% of the world's energy. So it's quite a big savings. That's

Judy Warner: amazing. So before, Too much further. Almost everybody that will be li listening or watching to this conversation is familiar with mosfet.

So can you unpack first of all, the acronym, because I didn't know it until you taught me. And and what that did going from, what that did practically to the cost of.

Dr. Alex Lidow: So MOSFET stands for Metal Oxide Semiconductor Field Effect Transistor. It's a [00:05:00] long acronym. Yeah. That's why we use the acronym.

And it basically is a transistor that turns on and off the flow of electrons or electricity. And it does it at a very high speed and that's, Because before that there were things like thyristors and bipolar transistors, and those were the switches that we used to modulate current on and off.

Now in the seventies there was a I'll say a push towards a different kind of power conversion switching power conversion as opposed to linear power conversion. And in switching power conversion, you basically chop the current and voltage into small chunks of. And then reassemble them after a transformer or an abductor into the form that you want.

Usually going from a high voltage, a low voltage would be most common form. And if you do that in a very I'll say slow device, like a Thyristor or bipolar transistor you tend to have a lot of. And the analog I use is if you had a small cup of [00:06:00] water and you wanted to fill it from a fire hydrant, you have this big valve, you turn the valve on, you have the water comes out, turn the valve off, you glasses full, but you got water all over the street, right?

That's what happens with low switch. And with a fast switch you can actually chop that water into individual droplets without impeding the flow and fill up the glasses and stop. I. And so that was something that needed to have a much faster transistor to be practical and. The key motivation behind developing a power moss vett, which was, a hundred times faster than the bipolar transistors of the era.

And that, that started that trend I see towards switching power conversion, both for power supplies and motor drives. And that's where the savings came from. I see.

Judy Warner: Okay. That's a good, that works for me. The picture of the fire hose. So again, I mentioned it at the top that we were introduced by Steve Sandler and when he told me about you and your relationship over the years together, [00:07:00] he said early on you sent him some mofits and he.

Stomped around, like in a pet's child and said, Alex, it'll never work. He goes, turns out I was right just 40 years later.

So he's a prophet in his own mind. But let's talk about, obviously for 40 years MAs have served, the world really in a really impactful way. So where are we? From the seventies till now. Where are we today with MOSFETs?

Dr. Alex Lidow: MOSFETs, by the way, spawned a derivative called the I G B T, which is really a combination of mosfet, bipolar transistor.

And between the two of them, they created, tens of billions of dollars of marketplace. But right around the turn of the millennium it really became evident that Moss vets had hit their theoretical limit. There's something called the Belga Curve that tells you exactly how much you could get theoretically out of a transistor.

And we knew that we were [00:08:00] at that limit. Now, when we started, the very first MOSFETs were 400 x away from that theoretical, right? So over a period of about 22 years, we made this journey of improvement and. But once you hit a limit with a device a lot of people say but those limits, but this in a power device was pretty fundamental.

So you really have to think about something radically different to do. And in the case of power devices, that really meant a different material. And it was right around that time that I read about some work done in Japan growing. Device grade gallium nitride on top of standard silicon wafers, huh?

And this became a sort of a light bulb moment. I said if you can just put a micron of gallium nitride on top of a standard silicon wafer, you could probably process that in a standard silicon factory. And if you can do that, then you've got something that is potentially thousands of times better than silicon with, without a whole lot of cost.[00:09:00]

And I think, turns out 23 later, 23 years later we are demonstrating that's true.

Judy Warner: So what does that limit look like? Is it just a speed of signal? Is it capacity? Is it both? And then how did you pivot at that point and start moving towards can technology.

Dr. Alex Lidow: So first of all, that limit is the resistance of the device for a given area and a given voltage, okay?

That you wanna block. So on. If you think about this graph, it's usually the vertical axis is the resistance of a one square millimeter device. And on the horizontal axis it's a voltage and there are all these diagonal lines for different materials and silicon's diagonal line, which you can buy devices all along that diagonal line is actually 6,000 times higher than the diagonal line for gallium nitride about, it's about a thousand times higher than the diagonal line for silicon [00:10:00] carbide for.

That was I knew that, yeah, I knew that for my graduate school days, it just, there was no way to really make good gallium nitride economically in those days. So this thing about putting it on top of silicon was really crucial. Now, at the time I was CEO of International Rectify. And we were the largest producer MOSFETs in the world. And I realized that I'd better do something. Cause when something hits a fundamental theoretical limit the fun is out of it. That's all. There's no more fun. It's just who has the most money to build the biggest factories.

And is willing to take the most things.

Then

Judy Warner: the commodity thing happens, and then it, yeah. All the innovation stops. It's no fun. And

Dr. Alex Lidow: I like to invent things, I was always interested in technology so we started working on it at international Rectifier and actually bought a small company called GaN Rose and started, I'll say tinkering in the early days with it.

The good news is I got fired from International Rectifier and what. You. [00:11:00] Yeah. There, so now I'm all of a sudden kind of loose ends. What the heck do I do with myself? And I, I still had ringing in my ears that thing about energy. Yeah. So I thought of a couple of things and I remember standing in the shower thinking I, there are two things I can do.

There are two things that I want to do. One of 'em is I want to create a carbon exchange because I believe that a carbon exchange is the best way to deal with global warming. Because

Judy Warner: we all stand around in the shower and contemplate these things. Alex, you're so funny.

Dr. Alex Lidow: Especially when you've just been tired.

And the, and then the second thing was gallium nitride. And I thought to myself as the shampoo was running down my face and my eyes that maybe I better do gallium nitride cause I know how to do that a little bit better. Better. And so I started a company called. Fishy power conversion with a couple of co-founders.

One of 'em was the guy whose company I bought a few years earlier. Bob Beach is the [00:12:00] name. And another one is a fellow that I hired right out of graduate school named Joe Cal and the three of us. And they were still working International Rect Bar, so they really took a A leap since they were in good standing there and I wasn't, they took a leap of faith and and we started efficient power conversion on the thesis that we could make power devices that were both more efficient and lower cost than s.

And realized that the two of those together was something that had never been done before. Nobody had ever dared compete on a cost basis with silicon. That's just like crazy stuff. And and, but we succeeded at that. We can build again, devices now at a lower cost of silicon.

And then the second part of that was that we would then use the knowledge of the markets that we get with discrete devices in order to develop integrated service. And the reason that is that we understood that with gallium nitride, it was easier to make power integrated circuits than in [00:13:00] s.

Oh, interesting. So you had an advantage with Dan and then you had another advantage with integrating ga. So that became our long range trajectory. And we've been following that

Judy Warner: well in my career. I remember a while back I was just because I've been around the block and I remember when everyone started talking about.

And it was like again it's gonna take over the world, but it's been one thing Steve had mentioned to me, as a fellow historian of power devices, he said when Moss Fitt hit the market, it was instantaneous. Like the industry adoption, he said was at break net speed and he made a point.

GaN hasn't enjoyed that same, vertical, industry-wide adoption. So where are we today in regards to [00:14:00] Moss Fitt GaN, like now in 2023, where are we in that journey?

Dr. Alex Lidow: In, in most new technologies, there's a crossing the chasm moment. There's something that, that takes off that, that gives you the volume that, that you need to crank down your costs and then, it's self-reinforcing for it.

In the mosfet it wasn't quite as vertical as Steve. That's how he remembers it. Yeah, good for Steve cuz I live that and every day it was why this darn thing so slow, but there was a very big crossing the chasm moment. Okay. And that occurred around 1980, where when I, ibm decided to come out with a pc.

And. Now before that, there was already a PC out in the market, if you remember it. The Apple two E. And I re, I remember getting a call from a guy named Steve not Sandler. It was one of the two Steves. It was Steve Wosniak who wanted to use [00:15:00] who wanted to include the power supply for the computer inside the computer.

So the whole thing would be self-contained on your desk. And he did a MOSFET to do that. And we had developed an app note. Converting from AC to dc, a universal power supply. We had a very talented apps engineer who was, a godfather of all topologies. His name was Brian pe.

And and he thought that would be a great thing to incorporate. He did. That was relatively low volume, but the minute that I, IBM hit. The market it really was high volume. So now we had a volume application. I see. That was helping us cross the chasm. We could build a small factory, then we built a big factory and then the cost came down even more.

And then, itself reinforces. So the question is, what about Gant? Yeah. We introduced the first production version of GaN in March of 2010 and about three months later a gentleman by the name of Dave Hall, a very brilliant guy, founder of Dyne Velodyne[00:16:00] corporation. And he said, I think that GaN would be a good.

For a 3D sensor. And he wanted to build a 3D sensor for a Google mapping function. Goodness. Google Maps at the time had GPS and a camera screen. I remember when GPS hit. Yeah. And so this is, this is like 20, 20 10 late 2010. And so he and he said that, If if we can increase the speed that we can fire lasers by an order of magnitude, then we can create a long distance sensor that has very high resolution and will be much better than radar.

Wow. That was his thesis. And that's where we're really the first commercial lidar sensors came from was Dave Hall and Valenti. And they used Argan devices. Now that evolved into, thousands of 3D sensing applications and they all use GaN devices. Pretty much all of 'em do. And so [00:17:00] that created a volume base that we could grow on, and also learning base that we could improve our devices.

We then got a vehicle company to put our stuff on into headlamps. And then we started to get higher volume applications in in, in things like power supplies for servers and power supplies for servers became a another step function, bigger business. And then now solar has adopted GaN in a big way.

So we have another step function with solar. In addition to all that, in parallel our gallium nitride devices turned out to be extremely radiation. Not by coincidence, but by design. Okay. I was like, wait, the space applications started taking off. Oh goodness. And we got the space. And in parallel with that, then we all want to do these last mile e-bikes, ees, scooters and stuff like that.

And we developed a bunch of ICS for that and those start taking off. So there's a bunch of these applications. As people become more [00:18:00] familiar with GaN that, that are very high volume, they're kicking in right now.

Judy Warner: So do you think. And Moss fit still fit for a myriad of applications, but now you're opening the doors to all these like space and lidar and all these advanced technologies that have taken off.

So that's really interesting. So both have their place is what I hear you saying, but GaN is really opening doors.

Dr. Alex Lidow: So I'd say both have their place, but one is in the morgue.

Judy Warner: Okay. That's interesting because, okay. What is your sale? What is the adoption though? Industry adoption for both.

That's what I want to ask you. Being the devil's advocate here. Yeah

Dr. Alex Lidow: And it's a fair question, and I'm exaggerating only this far. So back in the early days, again, back in the mosfet, the competition was the bipolar transistor. And if you look at it today, 40 years later, the bipolar transistor still exists.

And as a matter of fact, the size of the [00:19:00] bipolar market is about the same as it was. But what happened was all the new applications went to MOSFETs and IGBTs, so the growth came from that. And of course there's a lot of growth in our industry over these years. Yep, you bet. Now, if you look at what's happening with GaN is starting to attract or is starting to secure.

The a very high fraction, maybe even now, a majority of brand new sockets. When somebody upgrades from something, they are more inclined to do with GaN I see. Than with another mosfet. That is most true in higher performance applications or applications where you need very small size or very lightweight or some very high speed for some reason.

But it's occurring more and more frequent. So with this, and so the growth curve of GaN is gonna be very high and the growth curve of MOSFETs is starting to roll over and we'll flatten

Judy Warner: it. I see. So when you talk about socket, does that help in that you're not res spinning a [00:20:00] board or that's just a good way to adopt?

What is that socket piece about?

Dr. Alex Lidow: So when I say a new socket, so let's say you're a power supply designer. And you are designing a power supply for a server. Good application right now. Server architecture used to be predominantly an AC input and you'd have an AC power split lives in your rack, and then you'd have a DC output, which is 48 volts, and then you'd have a 48 volt to 12 volt converter, usually in the rack.

At each of the each of the racks individual. But on the others, on the rack side of the equation. And then on the server board, you'd have point load converters would take 12 volts down to 3.3, 2.5, 1.8 and stuff. In the last few years, particularly at the high end, starting with ai, cloud servers, gaming machines, things like that, they found that the amount of current it took to power the GPUs, the.

Processing, processing units was [00:21:00] much higher than the CPUs that they were using in prior generations. And so the boards were becoming more and more power consumptive. So it went from like a 500 watt server board, then to a one kilowatt, then to a two kilowatt, then a five kilowatt, and then to, on and on.

And that, that became a problem. So they started to bring the 48 volts onto the, I see. And the minute you do that, you go from, very inexpensive real estate on a server rack to very expensive server real estate on a board. Yeah. And so there's a huge premium paid for power density.

Now if you just compare what's happened here this is actually this is a GaN device. I'll show you a GaN power supply for servers. Okay. But this is a thousand WA device, right? Okay. In silicon, this was about 330 watts. So now g gets you up to a thousand watts like this. But now in the new generation of servers of power supplies, this is a

Judy Warner: [00:22:00] thousand.

Oh my goodness. Yeah. So it's easy to see. Now we've gone from it's 30 per percent, 30% expensive real estate differential there. Yeah.

Dr. Alex Lidow: Yeah. It's almost four times. It's less than one. This. Yes. Yeah. And of course your efficiency's higher as well, so you gain there, but this is perfect for a server board because it doesn't take up lot of space, has a lot of power.

Got it. So that just gives you an idea of the evolution. Now you can't do any of that with a mosfet, on a mosfet you better put it on the rack instead of on the board because it just, it doesn't fit. I. So that those, as those servers convert to 48 volts, now cars are converting to 48 volts.

I dunno if you just saw, for example, the cyber truck a big announcement. Cyber truck's gonna be 48 volts. There are a lot of 48 volt cars out there, by the way. And doin you have a 48 old system. By far your best choice is gm because GM is also lower [00:23:00] cost to produce on a component basis, plus you have a much, much smaller, more efficient system out of it.

So we're seeing car systems convert to G based 48 volts. We're seeing as new car platforms come out, Not as the old cars, you're not gonna convert, 1967 Chevy to 48 volts, but your brand new, Porsche Cayennes and your cyber trucks. Yeah. And your, all the mild hybrids, they're all coming out 48 volts.

So that's a, an an example where something that used to be a mosfet realm at 12 volts in the car now is a 40 volt GaN thing. And it's becoming big business.

Judy Warner: What I hear you say is you one you love from your shower moment is you created a blue ocean, right? You got out of the highly competitive moss vett business.

And you said, it's such a blue ocean to say I can make it better and faster and cheaper, whoever gets to do that. And and it's interesting. I saw Elon Musk [00:24:00] not long ago. He was unveiling their robots and he was saying our cars are already robots. People just don't realize it, and again the compute speed on all of that is, our minds don't get wrapped around that.

But I think it's really a truism. So

Dr. Alex Lidow: interesting. By the way you mentioned that cuz all the humanoid robots use. Of course they do.

Judy Warner: And and it's, and now AI machine learning seems to me, Alex, you're in a, all that's,

Dr. Alex Lidow: again, based based stuff. It all that new stuff is where GaN is dominant.

And Silicon can't touch it anymore. It's

Judy Warner: really interesting. I don't know, month or so ago, month and a half ago, I was podcasting from DesignCon where. Nominated, by the way, engineer of the Year. I remember that. But I, he won the engineer. I know. We're rubbing elbows with the man.

Anyways, he I got to interview just actually a [00:25:00] moment of it was just fun and I was in the right place right time. But I got to interview one of the lead electrical engineers from Boston Dynamic. And he talked a lot about their distributed power systems. Like it was power. And so I'm sure you're in, like you said, in all those technologies.

So when you introduce scan though, I call my podcast the ecosystem with two E's because what I'm noticing that kind of high level very high level. We used to get in these little well guarded silos with our IP and all of that, but if we're not working together, nothing, we can't build anything.

So there has to be this ecosystem. So when you brought on GaN, was the ecosystem in place to support that technology? Like I'm thinking material science, all the things that need to come together. Was that an issue for you at that?[00:26:00]

Dr. Alex Lidow: So I'd say that there are two things at pay scan's adoption.

One is ecosystem, and the other one is engineers that are experienced working with high speed components. Because, now we're going to a whole different level of speed. And so those two things really pace it. In the early days, 2010 of course the ecosystem was much, much more primitive.

When I say ecosystem, He switches, 10 to a hundred times faster than MOSFETs. So there weren't drivers for that. I see. There weren't controllers for that. There wasn't magnetics that could handle that. So the good news was that, a la firing a laser requires no mag or very limited magnetics and very limited ecosystem.

So that became the ideal cross the chasm kind of thing. And then we were able to motivate National Semiconductor, which became ti. To develop a driver for us. Oh, great. And they did. It's an excellent driver. And it's it, and it's, it's progeny are still out there and very popular.

And many people make that now. And then [00:27:00] the issue became controllers because now that you have a very high speed device, controllers couldn't keep up with it, and we're still fighting that. Of getting enough ubiquity in controllers that any function that you want to do, you can do with, again, device.

For example, motor drives. We have now, you say three years ago was only microchip. Now you have ti, you have Veras, you have St. You have on semiconductor. They all make controllers that work well with GaN in motor functions.

Judy Warner: From your seat in the house at the front of the, at the front of the pack here, so to speak.

How are researchers and engineers, and particularly, we've talked about different technologies and so are there any obstacles in their path? Besides these ecosystem pieces? Because I come from the board, I'm thinking Yeah. What'd that do to the board? Does it build up heat? Like I'm just thinking about the, is [00:28:00] does that introduce new problems for engineers specifically, but also researchers?

Dr. Alex Lidow: No it certainly does. You've gotta sharpen up your skills a lot. The analogy I use there is if you'd been your whole life driving a stick ship VW Bug and somebody gave you a Ferrari and said you had to drive it as fast as. You'd probably crack. And, GaN devices are very high speed devices.

And so people using their sort of low frequency techniques very often get into trouble with that. And so there's a lot of education involved and there's a learning curve and I think that was, it was certainly more than I expected. And we put in a whole infrastructure. To try to support that.

As a matter of fact, we wrote textbooks that are used in universities to help with that. But so that is hard. Now you mentioned PC boards, and I know you have a specific background of PC boards. That was another channel. Yeah. Because all of a sudden you have these devices that are 10 times smaller than Moss Fatts.

Yes.[00:29:00] So if your PC board doesn't have a higher resolution of traces Yeah. You can't really take much advantage of that. That's right. Now all of a sudden, you needed two ounce copper on 200 micron pitches and very few PCB manufacturers could do that. Now a whole lot more can.

So you know that evolved. People who, who were familiar with making cell phones never had problems, but people who were familiar with making motor drives and power supplies. That, that became a high hurdle cuz none of their supply chain supported it. I

Judy Warner: think about is, could that also be as people hit that wall and it's I don't wanna get in that Ferrari.

It's scary. I know. How does I'll just drive my VW faster, does that. Does that affect the adoption or the willingness to try to give? Because we all get in our lane and we wanna stay where we know how to do things. Yeah.

Dr. Alex Lidow: But more and more those people driving the VW are left in the dust.

Yeah. Like

Judy Warner: [00:30:00] that truck, they get ran over by that truck.

Dr. Alex Lidow: It's true. And, look, one piece of advice I'd give to any student in interested in this field is, get into either GaN or Silicon Carb. Don't waste your time with silicon because there's plenty of support for silicon engineering and there's not much for GaN and silicon carbide, and that's where the future is.

And also say as sure as the sun comes up GaN will replace MOSFETs below 650 volts and as sure as the sun comes up. So carbide will replace silicon above 650 volts. The only variables are when it happens and who's gonna

Judy Warner: benefit. It's start running or get run over by the truck I worked with.

Excuse me. I'll cut this

Dr. Alex Lidow: out.

Judy Warner: While I was with all Tim and I was engaged with not only board layout [00:31:00] engineer. But I dealt with universities and I dealt with an organization specifically that was trying to bridge the gaps between what's taught at university and what happens in the real world. And I'm like, you guys aren't missing your, you're educating to research dollars is not what your day-to-day job's gonna be.

And I know that you sit on the board of Caltech now, and I'm sure you see this. And are an advocate and it's a tricky problem to solve. But I saw a presentation by North Rip Grumman and saying, Hey, when you come on, we want you to be able to understand all this whole ecosystem and plus be able to do a good PowerPoint.

And so it's designing some chips, designing some boards, and most east think they're gonna go design chips and maybe on. And they don't think they're ever gonna touch boards aga, again, or, and it, and then they jump into the workplace and the students I notice that get the most [00:32:00] uptick is the ones that jump on these really amazing engineering teams, cuz it throws 'em into the deep end, into the ocean.

What do you see? Like I said, you're on the board, so how do you speak into that demographic? Especially being on the board. What's your thoughts there? Because I think it's a really important point as we look forward.

Dr. Alex Lidow: First of all, I rolled off the board of Caltech after 25 years. Oh, you did?

Okay. Yeah. But just a couple months ago, so I'm still familiar with the story. It's but Caltech is certainly a leading edge research institute. But you look at schools like Virginia Tech university of Texas, and Austin uc, San Diego, uc, Irvine, they all have very vital programs in power manage.

University of Wisconsin, university of Waterloo i, I t in, in Mumbai. And they crank out some very practically fluent engineers. Some very impressive engineers. As a matter of fact, just yesterday we had [00:33:00] a PhD out of UCLA start here. And he's, he is, he knows all this stuff.

That's great. It's not. So I think we do have those education institutions, the, they may just not be the ones that specialize in, in, very advanced reasons. I

Judy Warner: see. So in that case that the ones that are heavy in research, cuz on the board laying out a board that is, seems ugh, I'm never gonna touch that.

But then they end up having to touch it cuz you have to cross the chasm between, it's the board effects, it'll kill you. You're making these great things. Yeah

Dr. Alex Lidow: it's even more than that when you get to this, yeah. Everything interacts. Yeah. So this, the black thing is actually a transformer and you see it's integrated into the assembly.

And the very specifics of the magnetic field in that transformer is very crucial to the proper operation. So you can no longer be a PC board person. You have to understand thermals, you have to understand the signal pass, you have to understand the magnetics and the passes, all that stuff together.

It's a very difficult [00:34:00] job today.

Judy Warner: Yes, it is. And how to, the older generation is the old drafters that became really good board layout guys. And then these kids are coming up outta university and I'm like, Ooh, how do I help them know what's going on at the board level and that everything is radiating?

We're putting all this power and they all become, Steve's taught me, they all become little antenna. It's not really good.

Dr. Alex Lidow: E exactly. And now there's all sorts of software for simulating it all. Yeah. And you can't just simulate it from a signal point of view. You have to simulate it from an emmi point of view.

And from a thermal point of view and a thermal mechanical point of view. And all these things that, we just were talking yesterday with a customer where we were doing all sorts of thermal mechanical modeling on a powerway very similar to this. To make sure that it all squeezed and fit and torque

Judy Warner: together.

Yeah. One of our sponsors for the podcast is Keysight, and one of the things I love about them is that they can do that end to. And they can see, yeah, they can see the part and [00:35:00] they can see the overall, and they do it at the ed e d a level and the measurement. So it's a cohesive whole. And so they've been a good partner to help deliver that message.

And also, of course, they're saying, here's our tools. So

Dr. Alex Lidow: yeah they're fantastic instrumentation company and also now they make really good

Judy Warner: software. Yeah, their software is really impressive. I always like to tease out if I can Alex some case studies. Are there, can you discuss any current initiatives that you've been involved in at E P C and specifically re related to, victories are failures around Moss fitt.

Dr. Alex Lidow: Yeah, I, there's some things that are interesting revolutions. And I think GA pay plays a big role in the revolution itself. So for example, just take motors. Brushes, DC Motors replaced brush DC Motors many years ago. [00:36:00] And so we're all used to these three phase brushes. DC motors, you have 'em in your washing machines, you have 'em in e-bikes, you have 'em in drones, you have 'em in everywhere.

You have 'em in motor, in your power tools. And we got used to them. And they all run at 20 kilohertz. So there's controllers, there's MOSFETs, there's all sorts of things. There are little, multichip modules for all that kind of thing. What turns out that we all got myopic about that.

And as a result we missed a huge opportunity. So for example, if you go to higher frequencies, let's say a hundred kilohertz, all of a sudden the dynamics get much. The motor can deliver a whole lot more or the motor control can deliver a whole lot more power to the shaft. There's a lot lost in the shaft because of that low frequency. There's a sixth harmonic that gets you, that also makes a lot of audible noise. And now the entire system can shrink enormous. Into just a tiny little thing. I'll show you one of these. This is a drive system for an [00:37:00] e-bike.

See those three things? Yeah, those are three I of in GaN and those three plus a controller. Make an entire e-bike system, motor drive system. So by going to higher frequencies, you get an order of magnitude shrinkage, but you also get rid of the most gnarly component, which is an electrolytic pasture.

You get rid of your electrolytic faster. So now all of a sudden things start unwinding in terms of cost and reliability. I see everything going, in the right direction. Yeah. And because it's so small, it's real easy to shield it. So the emmi problems go away. There are these.

I won't call 'em mini revolutions cuz there's, a few hundred million hand. Power hand tools that are battery operated, they're made every year. And that is a big market that is converting to GaN. There are bunches of electric bikes. Although it's a new market, there are still millions of them.

And those are all going to e-bikes and then scooters, same thing. And it's, they're all taking advantage of both a higher frequency capability, smaller size and lower noise. [00:38:00] And that's just one example. It takes a few years. For all those companies to get through their design cycles and, obsolete, their old design and all that kinda stuff.

And for the market

Judy Warner: it will happen. Yeah. And those markets mature, as you said. I have, my husband and I have two amazing e-bikes. I love those things and it's amazing. I love mine. That's, I love mine. And it's like around where I live here. You and I talked about it. I'm in the Temecula area.

It's. So to have a pedal bike in your sixties is not that fun. It's I wanna power up to get up that hill and, and it makes for I could ride hours now and, but what amazes me because I have been electronics, is the, it's not a big, bulky thing, right? That motor is just in the back rim and battery's a little bulky, at least in mine, but, It's just, it's pretty miraculous, right?

[00:39:00] We just hop on and write 'em and whatever, but it's, the footprint's pretty amazing. So that size piece is no small thing. Yeah.

Dr. Alex Lidow: Yeah, I have the same thing. I live in a Hillary area and, I couldn't do it otherwise. That's

Judy Warner: fantastic. Yeah. I tried, I almost gave myself a heart attack. I did too.

And I'm like, I have a fancy

Dr. Alex Lidow: bike to prove it. Yeah. That

Judy Warner: I don't use. Exactly. I know. And we haven't tried it yet, but we keep saying we're gonna go. Wine country up Rancho. And I'm like, I don't know if that's very smart, unless we have an Uber pick us up at the end, but, or maybe you go to one winery, I don't know.

So any thoughts on, I think we've covered a lot. By the way, for our audience, I wanna hop in here and Alex, there, the EPC website is an absolute gold mine. There's books, there's resources. It is an absolute goldmine. And so I'm gonna put the link to that in the show notes, but make sure you go check out epc, it's [00:40:00] epc.com.

Dr. Alex Lidow: EPC co.com. There you go.

Judy Warner: So please go over and check it out. There's, you can learn so much more than we've had the time to talk about, but you can get amazing resource sources. Learn so, so much, and. Alex where else would you recommend that our listeners go learn about adoption? What's available?

Maybe if they haven't been paying explicit attention to maybe making the swap?

Dr. Alex Lidow: I it's maybe a little self-serving, but I'd probably start with the textbooks that are available and, the, this one from, this is actually a peer reviewed textbook by John Wiley. Not a commercial thing, but GaN transistors for efficient power conversion gives you both a background in, how GaN.

Works as a semiconductor. And then it also proceeds to tell you about ac about applications. Now Wiley publishes these, this third edition, and Wiley publishes these on five [00:41:00] year centers. So we're gonna do a fourth edition probably in about three more years. Cause that came out two years ago.

But that's a problem because the technology's moving much faster than that. So we publish an intermediate. Which is, again, power devices and applications. That it was basically everything since that book published. I see. So after two years we published this, and not only that, but we put QR codes at the beginning of every chapter.

So if you click on the QR code, you can actually get an update to the minute. Oh goodness. That's great. So yeah, we're trying to have a, I'll say a reference library that actually keeps pace with technology develop. So yeah, I mean that those QR codes are updated practically daily.

And we found that universities are adopting these books. Wonderful. So we have a lot of students that, that use them in classes and stuff like that. And I think that if you want to have a general education, that's what they were designed for. A general education in the advanced power conversion technologies, [00:42:00] specifically using GaN.

And I noticed

Judy Warner: on your website too, you have a forum where engineers can talk to other engineers or probably some of your subject matter experts.

Dr. Alex Lidow: Yeah. Yeah. We are all over that. It's a fun place. There's some fun controversies

Judy Warner: as there always are in engineering. You guys fight amongst yourselves.

Like I try to be like, I'm Switzerland, is stay in the middle. This has been so valuable and Alice, I really thank you for taking time out. Busy schedule and teaching us all about the history of the applications. I think this is really insightful and our audience is really gonna enjoy it. So I appreciate your time.

Dr. Alex Lidow: Thanks very much, Judy. It's been a lot of fun. Anytime I can talk about this all

Judy Warner: day. I know. It's fun for me too cause I love learning. I'll say farewell now and for our audience please go. I don't care if you're working out or driving your car. Take time to go over to the description and the show notes, and go to those links.

There's [00:43:00] really rich resources for you there. Thank you for joining us today. I hope you've enjoyed this conversation with myself and Alex Lido. We'll see you next time. Until then, remember to always stay connected to the ecosystem.