Powering NewSpace Missions Navigating the Cost vs. Reliability Challenge

Explore additional resources by viewing the Powering NewSpace Missions slide deck and answered Q&A from the webinar.

Tom Freeman:

Hello and welcome to today's webinar, Powering NewSpace Missions: Navigating the Cost Vs. Reliability Challenge. I'm Tom Freeman, Product Manager at Solsta, a distributor in the UK for VPT, and joining me for this webinar is Jeremy Ferrell from VPT.

Jeremy Ferrell:

Yeah, thanks Tom. Thanks for having us and thank you all for joining. My name's Jeremy Ferrell, I'm the Director of Engineering here at VPT. I've been with VPT for about 22 years now. During that time, I've developed, or I've designed, or I’ve been in charge of the development for many new products. 

Military, aerospace, several different space applications, from geosynchronous deep space missions to Low Earth Orbit (LEO) missions. And that's kind of what we're going to talk about a little bit today, are some of the new products that VPT has to offer. 

Tom Freeman:

Thank you very much. So VPT is an industry leading manufacturer of space-grade DC-DC converters with experience of flight systems for more than 30 years. Details for who to contact in your country can be found in the contact area of the VPT website. You’ll find an extensive network of technically knowledgeable distributors to provide support in more than 30 countries around the world. 

VPT distributors around Europe, we have AVE Added Value Electronics in Luxembourg, the Netherlands and Belgium; Expando AB in Sweden, Denmark, Norway, and Finland; Incomtech in Ukraine; Milexia in France and Italy; PROTEC in Switzerland, Austria and Germany; Semicon in Poland; Solsta in Ireland and the United Kingdom; and Venco Electronics in Portugal and Spain. 

If you have any questions, please hold them until the Q&A session at the end of the program, or you can submit them via the onscreen Ask a Question box, and we'll address as many as possible at the end.

Also note that there is a live audience here today, so if you can please hold your questions as well until the end, and we'll address when we can at the end of the program. A PDF of today's slides will be available for download under the event resources. 

To start: the changing landscape in space. It's just under 70 years since the first human-made satellite was launched in 1957, and we have just gone past the 60 year anniversary of the very first commercial satellite, Intelsat I, at the turn of the millennium. There were roughly 750 active satellites in orbit. 

In 2010, there were just over a thousand, so a 30% increase in 10 years. In 2020, there were approximately 3,400 satellites, so a 340% increase in 10 years. In 2024, we're already in excess of 10,000 active satellites, and by the end of this decade, we may well be over 40,000 active satellite systems.

It’s not just that the number of satellites is increasing, but the rate by which the number of satellites is increasing, is increasing. This huge growth in satellite and space systems is dominated by LEO satellite systems constituting roughly 93% of all active satellites. 

A more developed understanding of this LEO environment and increased commercial pressure for high volume, low cost, and shorter lifetime missions has seen a re-evaluation of what is necessary at a component level for these systems, resulting in a new approach to satellite design, which is cost-performance optimized and generally placed under the category of NewSpace. 

Jeremy, why should we listen to VPT regarding NewSpace? 

Jeremy Ferrell:

That's a great question. So, VPT has been in business for about 30 years. We have an extensive product line, a lot of it in space. We have traditionally, most of our space market is thick-film hybrids, and we've had thousands of missions that are very successful throughout that time. 

At VPT, we concentrate on high reliability. That is our market. We don't make super high-volume parts for automotive or for cell phones. We only look in the high reliability market. We have a lot of history, we have a lot of testing knowledge, we have a lot of manufacturing knowledge in all these different areas. 

We're also a subsidiary of a larger company called HEICO, and that has allowed us to make investments and buy other companies, one of which that we're going to talk about today is VPT Rad, which is a radiation test company. 

This new line that we're going to talk about today for LEO applications, or shorter mission applications, is really just an extension of the larger product line that we have. We’re trying to do it at a lower cost.

Tom Freeman:

Traditional space power systems were rad-hard, hermetically sealed, thick film hybrids. Why are those not ideal for NewSpace applications? How is VPT addressing this changed market that is NewSpace?

Jeremy Ferrell:

One of the things we did was, we looked at screening. Typically when you think of a space application for our hybrids, you think of Class K screening, space level screening, and it's pretty extensive just to look at the components that we buy. 

On average, we're seeing a 38-week lead time for the components that we use to build the parts, just because of the screening of those components to put in it. So those components, because of that long lead time, because of that screening, they're very expensive, and so it causes a lot of delays on when we can develop and when we can send parts to the customer. 

We also do an extensive amount of testing for that screening at the overall burn-in level also. And if you look at the amount of wire bond pull tests, every single wire bond has to be pulled.

We do temperature cycling, we do constant acceleration, we do PIND testing. We do a 320-hour burn-in, we do hermeticity, radiography and all the visual inspections throughout the process.

Each of these steps not only adds cost, it adds time. What we're seeing, not only for these LEO type or NewSpace applications, it's not only cost, it is also the design/development time is much, much shorter than we've seen for the deep space applications. So we have to reduce costs, we have to reduce lead time. 

What we came up with is a different screening method, what we call ES+. We went through, and we looked at where we are seeing failures and what can we do, what can we eliminate, to try to get the customer a high reliability component at the shortest lead time and lowest cost. 

So we reduced the amount of testing. We still do temperature cycling, we're still selling a high reliability part. We do temperature cycling, we do a burn-in, it's just less time. We still do visual inspections throughout the overall time, and all of that adds up to reduce costs, to reduce lead time, but it still is not giving the customer a commercial part; it's still giving them a high reliability part that is screened.

Tom Freeman:

So, the VSC is not a hermetic product – is that not important for LEO missions?

Jeremy Ferrell:

We don't think so. The hermetic hybrid is a great technology. We've used it for many, many years. It has a lot of history and it is great for many, many applications. It's expensive. 

Just to give an example, the package will probably make up 50% of our BoM costs alone, if we were to use that. This is our solution: the VSC product line. It’s more what we like to call COTS, or Commercial Off-the-Shelf type product. 

We're using plastic parts that we've spent years identifying what these components are. Then we take those components, and we screen them; every component that goes into the VSC is screened. 

We look for radiation shifts. Not every component, but a sample from every lot is tested for radiation. The overall converter is tested up to 42 MeV/mg/cm², we characterize at 30 MeV/mg/cm² for single event effects (SEE) testing.

We test for total dose 40 kilorads (krads) and then we guarantee to 30 krads to give some margin. We also do ELDRS (Enhanced Low Dose Rate Sensitivity) testing. Everything is tested before we put it in, and then we fully screen it also. 

So you're right, this is not a hermetic part. It’s an epoxy type packaging; it is a very rugged structure that's designed to handle the high levels of shock and vibration that it would see during launch. And it still offers a wide temperature range, not quite as wide as the hermetic hybrids, but it does provide some other advantages like double-sided cooling. 

Giving it a completely enclosed structure, the customer can put the heat sink on the top, or they can put the heat sink on the bottom. It gives a lot of system-level design. 

This packaging also has been tested for low outgassing, if that's a requirement for certain applications. It is low outgassing to not cause any problems with any kind of optics, but it still gives very wide temperature range. It gives a wide input voltage range. We still, for our isolated products, we're offering 500 volt isolation, and there's still fixed-frequency parts.

This design is using lower costs, but it still uses a lot of our topologies that we've used for years; it still uses a lot of our design techniques and circuits that we've developed over many years and have been very successful with. 

Tom Freeman:

Is screening level, something more appropriate to NewSpace like ES, enough to meet this cost challenge?

Jeremy Ferrell:

It is one aspect of meeting it. The other aspect of the low cost is just using lower cost components. We've reduced the BoM cost, we've reduced the screening cost, and we're always looking to make sure that we're reaching that optimal point of reliability versus cost. 

This is a new market for a lot of people. There are a lot of new companies that are getting into this market. The launch costs are getting cheaper and cheaper it seems like every year. And as more parts are launched, the cheaper they're going to get continuously. It's a lot easier for new customers to get into this market than it ever has been before.

Traditionally, there are just really big companies that have enough money to launch that could get into it. Now we're seeing a lot of smaller companies. With that, it’s creating a lot of new dynamics that we haven't seen in this market in a while, because there are a lot of various levels. 

Every company now has a variation of their opinion of reliability versus cost. And at VPT, we're always navigating that and we're trying to adjust and be dynamic as possible to make sure that we're meeting all the customers’ needs.

Tom Freeman:

So, what does VPT have? What is a VSC, and what's in the range?

Jeremy Ferrell:

For those of you who can see the slides, it's a little bit of an eye chart, but it gives a broad overview of what components we have released today. We are continuously looking at expanding this market, expanding this product line. 

But what we have today is: our isolated line is 5 Watts (W), 15W, 30W, and 100W. When we say a VSC5, that's referring to the output power. 

It's nominally a 28 Volt (V) input bus, so that's where “28” comes from. And then we have a lot of different output voltages. So 3.3V, 5V, 12V, 15V, are kind of our standard output voltages, and those are offered both in single and dual output. 

We also have our first point-of-load (PoL) converter.

Tom Freeman:

That is going to be your first PoL converter in the range, not your first PoL converter ever? 

Jeremy Ferrell:

Yes, thank you. We do have a lot of PoL for space applications and for non-space applications. This is the first one in our VSC product line, and we plan on expanding that as customer needs arise, but that's up to 50W of output power. 

It has a wide input range up to 13.2V, and it's really designed for 12V or 5V input, adjustable output, to meet some of the more digital demands, lower cost demands like FPGAs. You can have one part for the core and another small part for I/O and things like that. 

We also offer EMI (electromagnetic interference) filters. It’s in the same packaging as our converters, still designed for low outgassing, but with a ruggedized structure. The way that we spec those is through the current rating. We say VSCF1, that's for 1 Amp (A). We offer a 1A filter, a 3A filter, 10A filter, and a 20A filter. 

Those levels are chosen to meet a specific number of converters, which is highlighted in their data sheets. They're designed and guaranteed to meet conducted emissions for MIL-STD-461. You can see that here's just one example of a MIL-STD-461 measurement that we've taken.

All our filters are passive, so we don't worry about radiation hardening for the filters, but they are in a unique package for the outgassing, yeah, same low outgassing epoxy. 

Tom Freeman:

They’re compatible across the range as well, is that right? If I wanted a 5W and a 30W converter and I put them together, no problem.

Jeremy Ferrell:

Yes, you just have to make sure that you're still meeting the current requirements.

Tom Freeman:

So, it's great you've reduced the cost, but why shouldn't I just use a commercial part or an automotive part? In principle, I could get those screened. 

Jeremy Ferrell:

Automotive parts, they have a lot of testing. It's really amazing what you can get for the costs for the automotive screening. It's a lot of testing, but it's a lot of environmental testing that doesn't include radiation. 

They have a lot of life testing, temperature testing, and that's great. That's great really for what you get, especially for the cost. But the problem when you get out of the atmosphere is radiation. That is the biggest challenge that we've always faced when launching anything into space, whether it's for LEO or GEO (geosynchronous equatorial orbit) deep space, it is all related to radiation testing, and that's what we have a lot of experience with. 

That's what we've done for many years, and that's what we continue to do. 

This is an example of a radiation test report, or some excerpts from our radiation test report, that we offer for all of our products – not just the VSC, but for all our products.

What you see here is just an example of SEE testing and what the transients might look like, and then how many times the transients occur – as you hit it with a lot of different ions, what's the probability that you're going to get a larger transient or smaller transient? 

We also characterize that for both capacitive load or non-capacitive load. If you look at our data sheets, we have a max capacitive load. So we test the single event transients with and without that. So if you need to reduce the transients for some reason, you can look at it with and without capacitors to see what it looks like. 

We do all of our testing for the overall converter once a year. So we do a total dose once a year and SEE testing once a year to make sure there are no kinds of shifts. But the components that we put in there are also tested on a lot-by-lot basis. 

We do a sample from each lot of components that we put in there, and we look at that for total dose. We will test up to 40 krad (Si), both for the overall converter and for the individual components. 

We do ELDRS testing at the design point, we do an extensive amount of ELDRS testing and look for any kind of shifts we have in that. We said earlier, we characterize that 30 MeV/mg/cm²; we test up to 42 MeV/mg/cm² for any kind of latch-up or destructive events. 

The radiation test reports, again, are available for all of our parts, including the VSC line.

Tom Freeman:

If I understand you correctly, it's the screening and the radiation together; it's a combination of those two things that makes the difference. Even an aerospace-grade part, which goes through a very similar screening process, why couldn't I use that – it’s because of the radiation, is that right?

Jeremy Ferrell:

Yes. The aerospace will be a lot of testing, but again, we're talking mechanical, life, temperature testing, and not radiation.

And that is the difference, whether you're using anything that's non-radiation-hardened, it takes a lot of testing that we've done over many years to develop this. We're very fortunate to have our own radiation testing company, VPT Rad, that does this. 

So, we can send our parts up to them, they have our full test system in-house; we can send them parts, they can test it, and we can see any kind of shifts in performance almost instantly, no waiting. 

SEE testing – anybody that's had to schedule SEE testing, knows it takes a huge amount of time. And we have so many radiation-hardened, radiation-tolerant components that we're going, it seems like, almost every month to test parts. 

So, we have a very good relationship with an SEE test place. And then our radiation test house has their own source that we use extensively to make sure that we have the right component, and that there are no shifts in the component over the lifecycle.

Tom Freeman:

So we've eliminated commercial and automotive and aerospace grade parts. But if I wanted to design my own power system, my own DC-DC, what's the process? What are the difficulties, the challenges, that you go through?

Jeremy Ferrell:

Yeah, we've gone through a lot of difficulties over time, and it is finding the component that you're going to get. I mean, if you go and you develop a part with radiation-hardened Integrated Circuits (ICs), then you don't have to worry about it as much. 

The problem is you're never going to meet the cost targets. You'll be two orders of magnitude off in costs, and you still need to test the overall part to see what the transients are. So we've spent years developing that. 

I think we were a pretty early adopter in looking at this NewSpace market and really trying to find a set of active components that will work. There were a lot of parts that didn't. There were definitely a lot more parts that didn't work than did; it is taking us years to build up that active component list that will work for us.

Tom Freeman:

We've gone through how radiation testing has worked, and you've explained that you're doing all of that in-house pretty much through VPT Rad, but what about the converter as a whole? 

You can't just test the components and hope that that works. How does the whole DC-DC work for radiation testing? 

Jeremy Ferrell:

We go once a year and we look at total dose testing, and we look at SEE testing, and we go and we compare that data to previous test results to make sure that we're in line and there's no large shifts. We do sometimes see stuff and we adjust as necessary.

Tom Freeman:

You mentioned class K is just not cost effective for NewSpace. How did you determine that? How did you determine the ES was appropriate? 

Jeremy Ferrell:

We knew that given the cost targets that we were at, that Class K type screening was not going to be viable. We would have to pass that cost onto the customer. So we came up with ES+ screening, but we wanted to make sure that they did have a reliable part. 

We want to be giving the customer the most reliable part for the price. What you see here is just a sample of a test report that our automated test system has, and the details aren't important of what every number is, but we have an automated test system that looks at every part that we sell. 

We test at three temps, and we go through every line of the data sheet and run that test, then we look at the results and make sure it's within the limits over the entire temperature range, and then we save all that data, we put that in a database, and we do analysis on that to make sure that we're not seeing any shifts over time, so we can catch a problem before it's a problem.

We also do a lot of visual screening throughout the whole process at a sub-assembly level, at a component level, and then an overall converter level. We still do a burn-in, so we run the part at maximum temperature for 96 hours. 

We’re looking for any kind of infant mortality type that will screen out components. The Class K of 320 hours – it is too long. That all adds costs and lead time to the part. We picked 96 to try to give the customer a sufficient reliability bell curve, while reducing cost. 

We still do temperature cycling, internal visual, three-temperature testing, burn-in, and overall visual testing to make sure that it still meets all the requirements in the data sheet. And we're always open to listening to customers on what they need and adjust as necessary. VPT is always a dynamic company that's willing to move as needed.

Tom Freeman:

You mentioned earlier that you characterized the radiation performance at 30 krad, 30 MeV. How did you get to those numbers? Where does that come from, and why is it appropriate for NewSpace?

Jeremy Ferrell:

The number one is to answer the customer's needs. We want to make sure that we're developing a part that the customer is going to buy. In the beginning, we did a lot of component screening, a lot of component testing, to see what our capabilities are. 

Then we did a lot of interaction with as many customers as we could to figure out what their needs are, making sure that we're placing the part in the right cost versus testing. So ultimately, it is up to the system engineers out there that are deciding, for that specific company, the reliability or the cost vs. reliability curve that they're on. 

Here's one example of probability of a higher energy particle hitting the component. And we see there's a drop off around 30 MeV. At the beginning of the design cycle, we needed to be kind of above that to try to give it the most reliable part, the most probability that it's going to survive throughout the mission.

Tom Freeman:

So you've got 30 MeV, but you test it at 42 MeV, is that right?

Jeremy Ferrell:

Yes.

Tom Freeman:

That's guaranteeing the part performance per the data sheet at 30 MeV, but then 42 MeV, it's not going to fail, it’s non-destructive, but you might go out of regulation, is that right?

Jeremy Ferrell:

What we see is there'll probably be some larger transients; the transients will kind of increase as the energy increases, but the big thing that we wanted to do is make sure that it didn't fail. There's no latch-up, no destructive event. 

And so that was the reason for the higher testing versus characterization. Most customers want to see how the part is going to perform under 30 MeV, which is a higher probability that they're going to see that level, that energy.

Tom Freeman:

Of course, being the power system, if it does fail, it takes the whole system with it, right?

Jeremy Ferrell:

Yeah, we do have a little bit of a higher pressure, that if we're the first to fail, then the satellite is not going to operate. So it's always a challenge.

Tom Freeman:

Okay. So where would VSC not be appropriate? We've gone through why you've picked it, why you've selected the specifications you have, and why it's appropriate for NewSpace applications. It sounds great, but where is it not going to work?

Jeremy Ferrell:

Every mission will have a different spot on the reliability versus cost curve. The VSC is on the lower cost portion of that, but VPT – we've been developing parts for many, many years. We have extensive amounts of successful missions for higher level screened parts. 

We kind of concentrate, for the VSC, on around 5 year type missions. But if you're looking for 10–15 year type missions, the total amount of radiation is probably going to be on the low side. 

If you have a manned mission where you're going to move more towards higher reliability and cost isn't as big of a concern, you're probably going to lean away and move more to our hermetic hybrid type parts. 

VPT has an extensive range of products, and this is another addition to it, and we think it's going to be really applicable to this NewSpace type market that are shorter missions, that are higher volume, that are lower cost, that are shorter development cycles.

But if that isn't what you're looking for, we do have other products. Our SVR product line, that's our highest level of radiation, our highest level of screening, and that's a 100 krad part, that's an 85 MeV part. So that's for your 10–15 year type missions. 

So that one is designed around the aerospace tour for deratings, for component screening, things like that. 

We have the SVL series, which is 60 krads, 85 MeV. All of these higher-level screen parts are hermetic hybrids or hermetic. And then we have the SV series, so that's 60 krads, 44 MeV. 

It gives kind of a wide range of available products. And with that, it's costs too. The higher level of screening, the higher level of radiation, our BoM costs go up, our screening costs go up, the part is more expensive. 

So our goal was to have a large range to make sure that we have a part that's offered for every application.

Tom Freeman:

But suitable for every application. That makes sense. So, thank you very much; I believe you are at an exhibition next week, is that right?

Jeremy Ferrell:

Yes. We're going to be at Small Sat in Amsterdam next week. We're going to be at Booth #212. Please ask any questions through the ask question box today, but if not, and you're going to be at the show, please stop by, and ask in person. We'd love to talk to you and answer any additional questions that you have. 

Tom Freeman:

Then you're at RADECS and Space Tech Expo later on this year? 

Jeremy Ferrell:

Yes, somebody from VPT will be there. I don't believe that'll be me, but somebody will be there. So yeah, September, we have the end of September, beginning of October, is RADECS, and November is Space Tech Expo. 

We'll be at both of those, and somebody will be available to make sure that they can answer any kind of questions or if you need help figuring out what level of product you need for your application, we're more than happy to help. So please reach out and let us know.

Tom Freeman:

Thank you very much. Now we've got a couple of extra slides that we've included in the presentation for the download. We're not going to go through them directly now, but we thought you might like to take a look at some of them. I thought they might be helpful, hopefully they are. 

Thank you very much to everyone. A reminder to the audience, you can still submit your questions using the Ask a Question box, and we're now going to open up to the Q & A session. Just give me one second and I apologize, I'm going to have to log back into my machine. 

Okay, here's our first question: What sort of NewSpace missions are a good fit for VSC series?

Jeremy Ferrell:

What we're concentrating on are unmanned applications that are looking at around a 5-year lifecycle. There are still a lot of military operations that, even though they're in an LEO application, they're still concerned more about reliability of the part. So each system is going to be a little bit different. 

The VSC is targeting the lowest cost that you can have, while still having a radiation guarantee and screening. If cost is not as much of a concern and reliability is, you probably want to go with one of our higher-level screen parts to make sure that you have enough margin. 

So it really depends on what the mission is, what the system requirements are, and again, we're happy to help. Please reach out and we can help guide you if you do have any questions.

Tom Freeman:

So short duration, LEO ComSats, great; GEO? Maybe not. 

Jeremy Ferrell:

Maybe not. Yeah. 

Tom Freeman:

Okay, another question here: Can you list some famous programs and vehicles that VPT parts have been used on? Also, can you share what kind of bicycle you're riding these days? I think this person knows you quite well. 

Jeremy Ferrell:

As far as famous missions, wow, we have thousands of missions that we've been on. I believe we've been to every planet, including Pluto, asteroids, so there's a lot of different missions. I don't know if I can single any couple out, but many different missions, a lot of history there. 

Tom Freeman:

You used to have a flight heritage page on the website. I'm not sure if that's carried through to the new one, but maybe they could look on there.

Jeremy Ferrell:

Yeah, yeah, that's a really great point. I think there's definitely a section on the website with some of our bigger customers and flight heritage. That's a good point. So next question. I'm going to bypass the bike one for now. We might come back to that later. 

Tom Freeman:

Can I learn what is the test level for random vibration in gRMS (root mean square acceleration)?

Jeremy Ferrell:

Yes, we can provide all of that to you. I do not have it off the top of my head for the VSC specifically. It's different levels for hybrids and VSC, but yeah, we can definitely get all that for you. 

Tom Freeman:

Okay, so we’ll go back directly with some more information later on. Next question: Are the parts molded or lidded and filled?

Jeremy Ferrell:

It is an epoxy mold, so it's not a structure that is filled. It’s a very rigid epoxy designed for very rigid structure for shock and vibration requirements.

Tom Freeman:

How did you define the minimum screening needed for ES+ for this component? What is the reliability of ES+ versus K?

Jeremy Ferrell:

Oh, that's a good one. How we came up with ES+ was really identifying every step in the process and where it is. So first of all, Class K is a hybrid spec. 

Some of the stuff doesn't apply at all for obvious reasons, but if you look at it just from environmental type screening – temperature, vibration, things like that – what we try to do is identify where we see failures, if we can associate each one with a cost, and how to pass any kind of savings on to the customer. 

So ES+ is what we feel is kind of a compromise to that. We're still offering burn-in, we're still offering temperature cycling, but we're doing it at a reduced rate to try to reduce lead time to try to reduce costs. 

As far as reliability, the more screening that you have, any kind of failures are going to be screened down. So there is higher reliability with the Class K than the ES+, but we were trying to meet a balance of cost and reliability. And let us know if you don't think that we've met that, because we want to make sure that we're passing on the best part that we can to the customer. 

Tom Freeman:

I believe you've got some flight heritage on ES parts as well, historically over time. It’s not something you've gone into cold, it's something that you have seen works. 

Jeremy Ferrell:

Yes. 

Tom Freeman:

Question: in electrical point of view, specifications and components, is there any difference between ES+ and qualified space versions? I think that's conflating two issues slightly there.

Jeremy Ferrell:

Yeah, so we are not picking a radiation-hardened component to put in there. To make sure that we're very clear, this is not a radiation-hardened component. So yes, there are going to be differences, there are going to be electrical differences that we have to design around from an individual component that we use. 

Now as far as, if you look at our overall converter as a component, it will be similar electrical performance as our radiation-hardened line when we talk about regulation points and transients and things like that, it will be similar. 

The hybrid will be a little bit tighter on a lot of the tolerances, just because of the way it's built. But from an overall converter point of view, it's pretty similar.

Tom Freeman:

Do you contemplate, as part of the lot screening, radiation tests for SEE and total ionizing dose (TID) for ensuring that potential modifications in the foundry process did not modify the radiation tolerance observed during qualification?

Jeremy Ferrell:

Yeah, so that's the reason that we test total dose on a component level, is to look for those foundry changes. That's a really good point of view, really good question that we have to address. Someone that is developing this part on their own, if the foundry changes, the radiation performance can change. 

And that's why we do component-level screening. Each lot of component that we put in, we look at a sample and make sure that the parts don't shift, and that's why we have yearly screening on the overall converter to make sure that that hasn't shifted.

Tom Freeman:

As an addition on that same question: Do you think for COTS, where there is no traceability of the foundry process, modifications between lots is necessary? Is it necessary, this type of screening?

Jeremy Ferrell:

It kind of goes back to where we considered reliability versus cost, and that kind of ultimately is what it comes down to. Would we like to know the foundry changed? Yes, we would. And if the manufacturer releases a change notice, and we are made aware of it, then yes, we do go do additional testing to make sure. 

But a lot of times with COTS components, you don't have that information, and that's why we have to do as much work as we can to make sure that we're minimizing and mitigating those risks. 

Tom Freeman:

Can you talk about the expected reliability of these devices? For example, what is the expected life for these devices? How did you decide 96 hours was sufficient burn-in to eliminate infant mortality, and have you tested these parts to failure to determine activation energy? 

Jeremy Ferrell:

Wow, that was a lot of questions. 

Tom Freeman:

Should we go with the first one first? Let's break that up a little bit. Reliability of the device: what's the expected reliability of this part? Can we justify it?

Jeremy Ferrell:

Yeah, so with this, we do full MIL-HDBK-217 calculations that look at mean time before failures (MTBF). We do extensive life tests, qualification tests during the qualification process to try to determine length of reliability. 

Tom Freeman:

So, what is expected life for these devices? I believe that the MTBF data is in the data sheets, for each of the parts. And then how did you decide 96 hours was sufficient burn-in to eliminate infant mortality?

Jeremy Ferrell:

That's a great question. What we did is, we looked at all the history of burning-in components, where do we see failures? And where we saw really the highest failure rate is in the first 24 hours. We knew that we needed some type of burn-in to screen that out. 

What is the right number? That’s a good question. We did come up with 96 – should it be 96 or 48? These are good questions. I'm not sure that we're right. 

Looking at our data, we said this is going to be the most reliability for the cost. We could burn-in more, and that's kind of feedback that we love to get. And so we're navigating this along with many of the other customers, making sure that we're trying to get the right screening of the part.

Tom Freeman:

The final part from that question was: Have you tested these parts to failure to determine the activation energy?

Jeremy Ferrell:

Yes. When we look at SEE type testing, when we look at failure rate for total dose, originally when we went and developed the part, yes, we tested everything to failure. We don't test that on an annual or lot type basis. 

When we were originally doing the development and figuring out what the capabilities of each component were, to figure out what kind of capabilities we could have, we did that type of testing, but we don't repeat that on any kind of annual or lot type basis.

Tom Freeman:

Okay. Next question. For VSC, is there a subgroup that receives testing over a higher number of temperature cycles? I think the answer to that is no.

Jeremy Ferrell:

Yeah. Part of the cost reduction was to not have the different screening levels, to have standard parts, so everything can move through the manufacturing process as seamlessly as possible and as quickly as possible. 

For our hybrids, we have a lot of different screening levels. That means this order has to be screened at one part, the next order has to be screened at another part, which causes a lot of delays in the manufacturing process. So only having one screening level streamlines that. And again, that all relates to cost savings and time.

Tom Freeman:

But you're still open to have that conversation with customers? Obviously it depends on the situation, right? 

Jeremy Ferrell:

Yeah, sure. 

Tom Freeman:

If a customer requires DPA (Destructive Physical Analysis), it can add a lot of cost to the buying of a power supply versus me building my own. What info can you give me to reduce the perceived need for DPA?

Jeremy Ferrell:

So, first of all, when we talk specifically about the VSC series, we can't do DPA. That is a hybrid, and that's really a requirement that the systems engineers have as far as making sure that the part is what it says. 

And so any kind of destructive part of course adds destructive testing, adds a lot of costs. They have to end up throwing it away. But we do a lot of testing internally to make sure that we are selling a high reliability part.

Tom Freeman:

Sorry, we are getting quite a lot of questions here. Unfortunately, we're not going to be able to get through all of them, but we're going to try and go as quickly as we can. 

Jeremy Ferrell:

I'm glad people are asking questions. 

Tom Freeman:

Next question: Converters are closed loop systems. How does someone assess the performance of the converter in a system via analysis, apart from testing the final implementation? Test the whole circuit, right?

Jeremy Ferrell:

Yeah, they need to test their overall system for stability. Unfortunately, we do not offer SPICE (Simulation Program with Integrated Circuit Emphasis) models. A lot of people ask for some kind of simulation model. 

But a DC-DC converter is a very complex part, and we develop SPICE models internally, but it wouldn't be applicable for customer use. Everybody's system is different. Everybody's load is different, everybody's operating system and operating conditions are different. Really the best way is to test the part in the system that's going to be operated, and to really ensure that they're getting the performance that they need.

Tom Freeman:

This is a multi-part question here: Please describe how VPT decided VSC is a reliable product. What kind of testing and what sample sizes were used, and what are the results, eg. FIT (Failure in Time) rate, and MTTF? 

What manufacturing controls does VPT have to avoid workmanship and manufacturing issues? What is the life guarantee on VSC at temperature extremes and LEO radiation? Does VPD provide de-rating guidelines for the VSC product? 

I think we'll have to split that up. Too many questions in one go. What kind of testing and sample sizes were used, I think is the first.

Jeremy Ferrell:

Okay. When we talk about individual component radiation testing, we pick five pieces from each lot and that's for radiation testing.

Tom Freeman:

Next question was workmanship and manufacturing issues: how do we avoid them?

Jeremy Ferrell:

We have very extensive internal procedures that highlight and document the part through every part of the process. And then we have travelers that follow it through to make sure everything is done. 

We have visual inspection at several points to make sure that everything passes, and then we have overall inspections to make sure it's good. All of these are internal processes that we have for manufacturing, and they all stem from our high reliability line. 

We're building high reliability parts. We've been doing it for 30 years, many, many successful missions. So it's the same. We're using the same manufacturing processes. It is different components, but our philosophy has not changed. I think there was a de-rating question. 

Tom Freeman:

There is, yeah. Do you provide de-rating guidelines? 

Jeremy Ferrell:

Yes, we have de-rating guidelines. Our de-rating guidelines for the VSC is an internal document. 

Tom Freeman:

And that could be, under NDA, potentially shared?

Jeremy Ferrell:

Yes. 

Tom Freeman:

Okay, then the last question was about temperature extremes and radiation. How do we guarantee the lifecycle under temperature extremes?

Jeremy Ferrell:

Part of that, during temperatures extremes, we do temperature cycling. During the qualification process, we run it at maximum temperature for 1,000 hours, make sure that it's still running. We do extensive life type testing for the parts. That’s kind of all part of our qualification process. 

Tom Freeman:

Quick question here. Can you connect the VSC modules in parallel and/or series? Yes, you can.

Jeremy Ferrell:

Yes. It's actually in the data sheet. 

Tom Freeman:

There's a guideline on the website as well, isn't there?

Jeremy Ferrell:

Yes. So for parallel operation, we only allow the 100W to be paralleled. Anything below, you would just use the next size up converter. But for parallel operation, it needs a share-type connection, which VSC100, they have that. Now when you say series, I like to call it stacked. You take two 5V converters to make a 10V converter. 

And yeah, all of our isolated products can do that, but they have to be isolated in order to do it. Anything that's isolated, you can stack the outputs to create different output voltage configurations.

Tom Freeman:

Okay. Probably one more question. What about launch vehicles? Launch vehicles are, simply put, a mission that cannot fail, making reliability king. Are VSC parts applicable for these types of missions? 

Would you see an SVL or SVR series product as a better match for a critical launch application?

Jeremy Ferrell:

I think that SVL type or SVR type series, they’re more designed for that kind of application. VSC is more driven for the cost-sensitive components, cost-sensitive designs, higher volume satellites, things like that. So if reliability is your number one objective, you should choose the highest reliability part that we offer.

Tom Freeman:

This one's too good. I've got to ask this one, sorry. Do the VSC series regulators have monotonic voltage rise during turn on transient

Jeremy Ferrell:

Yes. Yes. So that's something all of our parts require, but yes, very good question. That can cause a lot of digital instability. 

Tom Freeman:

Okay. So that's it. Thank you very much everyone. Any questions we haven't addressed, we'll reply back to you directly. I'm sorry we couldn't get to all of them. 

As a reminder, there's a survey that will pop up at the end of the program. Three quick questions. If you can please answer that, we’d really appreciate it. You'll be emailed the link of the recording. 

And once again, thank you very much. We really appreciate you tuning in and listening in today. 

Jeremy Ferrell:

Yeah, thanks Tom. Thank you very much.