On Episode 72 of The Edge of Innovation, we’re talking with entrepreneur Simon Wainwright, president of Freebird Semiconductor, about Gallium Nitride technology and the future of the space industry.

Show Notes

Freebird Semiconductor’s Website
Contact Freebird Semiconductor
Find Simon Wainwright on LinkedIn
What is GaN?
What is Moore’s Law?
How2 Cut The Power Cord: Wireless Power Is Ready For Prime Time
SPWG — Space Parts Working Group Conference 2018
Freebird Semiconductor to attend and present at 2018 Space Parts Working Group
The Aerospace Corporation
Link to SaviorLabs Assessment

Sections

What is Gallium Nitride?
How To Build Gallium Nitride
Computing Technology
The Future of Gallium Nitride Technology
The Space World Today
The Beginning of Freebird Semiconductors
How To Convince the Space Industry to Adopt New Technologies
How to Do Accelerated Life Testing

Gallium Nitride Semiconductors: The Future of the Space Industry

What is Gallium Nitride?

Paul: So now it’s gallium nitride.

Simon: That’s correct.

Paul: From my science in school, I’ve seen gallium, and nitride is…what is nitride?

Simon: It’s nitrogen. Basically, it’s gallium and nitrogen.

Paul: So you put them together.

Simon: Yeah. Put them together. One of each.

How To Build Gallium Nitride

Paul: So what do you do? Go to the store and buy a bucket of gallium? Or what is it? Is it a metal?

Simon: Well, no. They exhibit semiconductor properties when you join them together. So essentially you would still use a silicon handle wafer, which is basically just the base. Consider you’re building a house, let’s say. So you would use silicon wafer as the foundation, which basically does nothing, but it makes everything strong. So it has no functional role as you build onto that. And you gradually grow it. So you have a reactor, and you grow by atomic layer by atomic layer, and you grow the structure.

Paul: You do this with tweezers?

Simon: We do this at very high temperatures. So we basically grow it and we insert different gases into this chamber, and they react, and their natural state is to form gallium nitride. We put in dopants of different kinds to make, to change…

Paul: So you’ve figured out the process.

Simon: No. But we’ve figured out how to modify the process. So the process was figured out by EPC. So we’ve figured out how to modify that process.

Paul: People have all seen the sort of semiconductor circles with all the chips not cut out. So you take a wafer like that, and you’re collecting this by using gas. Is it diffusion?

Simon: Well, you basically grow different layers. So, if you can imagine, you’re building your house onto your silicon foundation.

Paul: Atomic layers.

Simon: Yeah, layer by layer.

Paul: So, one atom thick of gallium or nitride or is it together?

Simon: You start introducing different concentrations. And you gradually go from a pure silicon wafer to like a pure gallium nitride layer. So you gradually introduce it. There’s obviously a transition, a buffer region. But the real gallium nitride, pure gallium nitride layer, which where all the action of the transistor, is a couple of layers of atoms thick.

Paul: Wow. So it’s more like a peanut butter sandwich.

Simon: Absolutely.

Paul: I mean, the house is good, but it’s got to have no basement.

Simon: There you go.

Paul: So we got the bread, and we start putting peanut butter on. But we’re really putting peanut butter and jelly. And by the time we get to a certain thickness, it’s perfect mixture of peanut butter and jelly. So you’re really in the sandwich-making business.

Simon: They wouldn’t fill you up. They’re very thin.

Paul: They’re very thin. But how can something so thin switch electrons. Do you do any other things to them?

Simon: Well basically, this gets really technical. We confine the layer of gallium nitride to be so thin that you form what’s called a quantum well.

Paul: Sounds cool.

Simon: Sounds really good. So if you go into atom-sized dimensions of everything, then you get quantum physics starts kicking in and you confine a load of carriers into a very very small space, and you increase the mobility of those carriers. So that way, they can travel through the semiconductor a lot quicker. And our components are actually called HEMTs — high electron mobility transistors because of that.

Paul: Alright. And so then what’s the next step? So you’ve got these wafers, and you’ve succeeded in putting how many atomic layers are there?

Simon: I couldn’t really say that.

Paul: Okay, so that’s a secret.

Simon: Somebody would kill me. I’m not sure who. But somebody would kill me if I said that.

Paul: So it’s more than one and less than a billion or whatever.

Simon: There you go.

Paul: I don’t know. A billion wouldn’t probably even show up. But it’s an atomic layer, so you’ve got this sandwich. So then what do you do? Slice these up and put them in packages?

Simon: Basically, you need to put the third electrodes. So at either end of this very thin layer, you have a source and drain. That’s where the current flows between, in and out. And then you have to have a control contact, which, in this case, is called a gate. When you open the gate, you allow the electrons to flow from in to out. And essentially that is a transistor. So the Jell-O, if you like, on the top, is the gate. The technology with that is, there’s a lot of physics involved. There’s a lot of technology involved to enable that to work correctly, so to speak.

Paul: Sort of make it all happen. So then the application of power to that gate can be faster, switched faster. So and we’re talking very small amounts of time here, even in a regular transistor. So if you take a silicon transistor and you apply power to the gate, what’s the switching time?

Simon: I mean, it varies. There’s lot of different configurations but I’ll give you the limitations of the switching time. So the switching time is determined by charge. You have charge on the gate and charge on the drain and the source. So the more charge you have to move during your switching operation. So the lower the gate charge or the drain charge or whatever, the better, the faster you can move it, because you have less things to move. So that’s basically what determines the switching time of a transistor, any transistor. So if we can compare apples with apples, a radiation-hardened silicon MOSFET, which is the silicon way of implementing this, to an enhancement-mode GaN, HEMT, our gate charge is an order of magnitude lower. An order of magnitude lower.

Computing Technology

Paul: So now does this have any application in actually computing technology?

Simon: Absolutely.

Paul: Because that’s the point, we’ve got to get things to switch quickly. So that’s cool. Is there a projection in somebody’s mind out there for the impact of computers being faster because of this?

Simon: Absolutely. I mean, you’ve heard of Moore’s Law, where I think it’s every 18 months the size of electronics reduces by half. So this will actually permit that to continue because silicon has really gotten to…

Paul: We’ve squeezed as much as we can out of it.

The Future of Gallium Nitride Technology

Simon: Yeah, to its fundamental limits. On this is more on the commercial side, not related as much to our product. But certainly more on the commercial side, the founder of EPC, Dr. Alex Lidow, has predicted that Moore’s Law will continue. Some of us like to now call it Lidow’s Law.

Paul: Interesting. So does that mean, and again, I am not holding you to this. Is this five years from now I’ll see computers doing this? When am I going to go to the store and buy a computer that’s a magnitude faster because of this technology?

Simon: At this point, I’m not able to tell you that because my world is the power world rather than the digital world. So, I don’t really know how fast it’s going to be adopted in the digital world.

Paul: How about in the power supply world?

Simon: The power supply world, it’s here already. You’ll see new products coming out, to put it directly into people’s lives. You’ll see that you can actually cut the cord. You can throw away wires because you can remote charge most things. You will be able to remote charge most things.

Paul: So it’s not wishful thinking.

Simon: Oh, no. It’s actually happening.

Paul: Because we’ve heard a lot about wireless charging and all that, but it doesn’t work all that well, and it’s sort of working, but it’s not. So you’re saying it’s prime for market betterment.

Simon: Absolutely. I mean, I have a Samsung. I have the pad. I replaced the Samsung. So gallium nitride is not actually used in the Samsung or even the Apple remote charging things at the moment. But it will be in the future. It will have to be incorporated.

Paul: And what does that make it? Does it make it charge faster?

Simon: Makes it charge faster.

Paul: Further away?

Simon: I’m not sure about that. I’m not as familiar with that technology to give you stats and distances and things like that. But it’s certainly faster. It’s more efficient, and, you know, it would enable you to charge higher powers rather than just a phone. You can actually run a laptop on a desk. You’d have a charger pad underneath it. You just put the laptop or whatever.

The Space World Today

Paul: Alright. So now in the space world, there’s all these people putting satellites. Is it just satellites? I mean, there’s a few missions outside of our planet, I would imagine. But the majority of it is satellites, or is there other stuff in the space world?

Simon: I would say there’s satellites. There’s space exploration vehicles, the ones that go to different planets.

Paul: Have you made it into any space exploration vehicles yet that you can talk about?

Simon: We’re working on one. Let’s leave it at that.

Paul: And do you work with any aliens yet?

Simon: No. Some of the guys back at the office.

The Beginning of Freebird Semiconductors

Paul: The market there is huge, I tell you. That’s just incredible! So you guys started this and you’re on the North Shore here in Massachusetts. What does that look like over the next three years? How does your company grow? Are you commercialized? Are you shipping? What are the next sort of milestones?

Simon: Okay. So let me go back to when we founded it. So we spent a year basically developing our product portfolio to making sure. We had to do a bunch of testing, do radiation testing. We do electrical testing. We do temperature testing. We do a plethora of different kinds of tests. So we spent a year, 18 months getting to that point. And that never ends. We have to continue to test, continue to push the boundaries of the technology so that we know where it fails, why it fails, how it fails.

Paul: And then how to fix it.

Simon: And how to fix it. If you know that, then you can determine the lifetime of that. But the bulk of that work was done in the first 18 months. Then we sort of came out of the closet, so to speak, and we went public. We came out, out of hiding, so to speak, after year one, essentially. At the end of year one. And we presented to the industry at a conference called SPWG — Space Parts Working Group over in California. This is sponsored by the aerospace corporation, which is one of the, I would say, like a regulatory body sort of thing. And people there were NASA, the European Space Agency, the Japanese Space Agency, and then all of the guys that build satellites — so Boeing, Lockheed Martin, Northrop, all of the prime contractors are there. So we came out at that.

Paul: Was it a surprise to them?

Simon: There was a lot of interest. Let’s just put it that way. There was a lot of interest at that point.

Paul: So I’m a designer in the satellite world. You’ve just given me new tools.

Simon: I’ve just given you a new solution.

Paul: So this is like, “Okay. I’ve got to redesign all my power supplies.”

How To Convince the Space Industry to Adopt New Technologies

Simon: Well firstly, there’s a lot more work that has to be done before anybody that is remotely involved in space will actually adopt your technology. Firstly, you have to convince them. Bear in mind that the technology for space has not changed in 30 years, since the lunar landings and all the Apollo missions. So you have to break down resistance to change first in a lot of these companies. And the only way to do that in this space industry, which is an extremely cautious industry, the only way to do that is with data.

So we had to go through our portfolio, and we had to test everything. Every single device that we ship goes through an individual screening program. Some parts get tested for 2,000 hours, for instance.

Paul: This is a part you’re going to ship.

Simon: Oh, yeah.

Paul: So it’s not just one sample.

Simon: Oh, no, no. Everything that we ship has been tested, 100%, at various different levels of stringency. So our second major goal was to break down the barriers of acceptance on this new technology into a world that had been dominated by silicon.

Paul: So it’s really marketing. I mean, it’s marketing with backup.

Simon: It’s marketing and engineering. Yeah.

Paul: But you’re talking to an engineer, and an engineer isn’t going to take that risk without compelling evidence.

Simon: Yeah. Absolutely.

Paul: Just like you’re not going to buy the car unless you like it. So you’re breaking down those barriers to entry or barriers to integration, I guess.

Simon: And essentially, you, you go through all of the data that they would require, and then you show them, once they’re satisfied that you’ve gotten to a point of reliability that they need, then you have to show them that the performance is worth it. So they’re not going to put anything in there that’s going to break after five years. So then you’ve got to show them. Then the sales end starts. Then you have to differentiate your product with switching performance or the losses or whatever somebody is specifically interested in for their specific design.

So at that point, then the sales effort starts to communicate all of those differences. So you have to have in your back pocket, one, a bunch of radiation testing, two, a bunch of life testing, reliability testing, and then — only then — once they’ve seen that data and believed that data can you then start trying to sell the product. So there’s a lot of upfront work, and there’s a lot of barriers to entry into this market.

Paul: Yeah, I could imagine. So do you give them samples?

Simon: Wherever we can, we try to sell them samples.

Paul: Well, okay. Alright.

Simon: But yes. We’ve been known to give a few away.

Paul: So they’re actually trying it and, and playing with it. It’s not like just a piece of paper.

Simon: No, no, no. Most, most of the major satellite companies in the world have Freebird parts that they are testing at this point.

How to Do Accelerated Life Testing

Paul: So now, you talked about radiation testing and life testing. So how do you do life testing? Just for the average person, you’re not going to be alive in 90 years or 100 years, how do you tell if this is going to—

Simon: So we do accelerated testing. So basically, what we do is, we increase temperatures or increase voltages — whatever is sensitive during the lifetime of a component — and we put more of that than it would normally see. So we try and accelerate the aging process. So, for instance, a very easy example to understand is temperature. So we would test our parts for a thousand hours at the temperature of 150 degrees. Okay?

Paul: Fahrenheit or Celsius?

Simon: Celsius.

Paul: Okay. So that’s pretty warm.

Simon: You’re going to have to convert that into F.

Paul: Yeah, sorry. Okay.

Simon: It’s been a while since I did that. So, you’d leave that on with a bias, or you have your in and your out, your source and your drain, so you bias the drain at 80% of its rated voltage, and you leave it on test, continuously energized for a thousand hours, which is eight weeks, more or less. But the fact that you’ve done that at temperature, allows you, with statistics, to predict an accelerated aging, so to speak. So you get a lot into statistics.

Paul: It’s burning. You’d burn your fingers.

Simon: It’s 320 maybe.

Paul: Okay. So you’d burn your fingers. But isn’t space cold?

Simon: Space is cold, but we’re not trying to simulate space. We’re trying to accelerate the aging process.

Paul: I see. So basically, you’re stressing the technology. What about freezing tests?

Simon: Well, when you say, “Is space cold?” it depends where you are in space. If you have a direct line to the sun or, so are you on the bright side or the dark side of the moon, so to speak. When you’re in the dark side, you’re at minus 50-something C. If you’re on the bright side, you may be at 80 degrees C.

So we also go through thermal cycles. We have a chamber which has an elevator, basically a small elevator. It goes between an oven and a fridge.

Paul: Oh really? Oh, that’s cool.

Simon: It’s great.

Paul: You can put a soda in there, and you can cool it off really fast.

Simon: When we have office parties, we put the pizza in the warmness and the beer in the cold.

Paul: There you go. So you’re doing this, and you’re doing it from, I guess, a compliance level where you’re actually testing it and certifying it and making sure that it’s true so that people can track that all back.

More Episodes:

You’ve been listening to Part 2 of our conversation with Simon Wainwright! If you missed Part 1, you can find it here! To listen to Part 3, you can find it here!

Also published on Medium.