Today’s spacecraft and satellite requirements giving COTS a fresh look
The fast growth of small satellites and their lower life cycle and launch costs has created a sizable market for low-cost electronic components that have radiation-resistant characteristics. Designers are forced to get creative with their design processes and business models to meet these demand
Opportunities for designers of electronics intended for use in the radiation-filled space environment are greater than ever with the growth of small satellites, the push toward mega constellations, and increased funding for military and NASA programs. An agile approach is necessary on electronics suppliers’ part, as end users want their spacecraft – manned or unmanned – faster, with greater performance, having more stringent radiation protection, and at times at an 80% to 90% reduction in cost, which means more commercial off-the-shelf (COTS) components.
“Increased funding from the Department of Defense (DoD) is a positive for the military space market, as brand-new DoD satellites are being defined,” says Tony Jordan, Senior Director of Business Development, Cobham Advanced Electronic Solutions (Colorado Springs, Colorado). “We’ve also got a new threat in hypersonics that we can’t identify and track using traditional systems. Industry and the DoD are developing new methods for defeating those. There is also the promise of growth for NASA with the current administration setting a goal to return to the moon.
“Commercial-wise it’s all about constellations that are being built to gather more images of the earth and enable faster broadband communication,” he continues. “Small satellites are driving a paradigm shift in the business model because they are relatively inexpensive to launch, and inherently redundant because when one comes down you can just throw up another satellite. This is the commercial model of replacement as opposed to the model of the exquisite spacecraft that can last for 20 years without a repairman. The customer base understands the risk and some are willing to embrace it while others are more reluctant. It really depends on the mission.”
While those missions – typically classified – still exist, more and more programs are looking for ways to reduce costs to embrace new technology more quickly. Space electronics development times take years to prepare and ensure the components can survive in space.
“One of the reasons rad-hard circuits lag the commercial components by almost a decade is that the process technologies available do not have inherent capabilities to ensure the part can survive and operate in extreme radiation environments,” says Anton Quiroz, CEO of Apogee Semiconductor (Dallas, Texas).
“As commercial satellites get more and more complex, we are seeing a general trend of customers wanting more integration and higher-speed devices in a smaller form factor which challenges density and efficiency at the component level,” says Eli Kawam, Business Development Manager in Microchip’s Aerospace & Defense business unit (Chandler, Arizona). “We are also seeing our customers customize the satellites to the specific mission (i.e. ... orbit, timeline, etc.) and by doing this they are able to require less margin. This can result in lower radiation or quality requirements. These programs are also more sensitive to acquisition cost, service entry dates, and a technology refresh plan, so they are attempting to reduce costs by using components with lower levels of qualification and screening. Typically, military rad-hard applications require very high levels of qualification and screening, up to QML [qualified manufacturer’s list] class V for ICs, JANS for discretes, and class K for hybrids. In general, it’s all about trade-offs between performance, cost, operating lifetime, and time to market.”
Time to market is important, especially if you want to embrace state-of-the-art tech. “That pace is so fast today that if you have to make money with 15-life-year satellites you’re going to be bypassed,” says Jerry Festa, Senior Product Line Manager, Space Segment, Curtiss-Wright Defense Solutions (C-W’s Newtown, Pennsylvania facility). “Cost is what’s driving space requirements to embrace more commercial components, that and the growth of the constellation and small-sat markets, which have much lower price points.”
COTS in space
The procurement term “COTS” has often been considered taboo among space electronics designers, as the commercial part of the phrase connotes low quality and/or low reliability. Yet many COTS solutions are not low-quality at all, and in demand for critical space applications. This demand is forcing traditional high-reliability (high-rel) designers to meet the cost constraints while remaining radiation-tolerant.
“Space COTS has been discussed and debated for the last 15 years and it’s finally getting to the point where the customer base is becoming more comfortable with using COTS,” Festa says. “Some RFPs mention COTS for space and I think it’s a trend gaining momentum. The military is usually the last to take hold of it because of its entrenched business model, but I think it everybody is realizing it’s the way to go.
“With COTS you still have to mitigate radiation effects on a limited budget and you can’t compromise on mission-safety requirements,” he adds. “There are various satellite designs and launch vehicle designs trying to compromise and balance those requirements.”
“While generic plastic-packaged COTS components offer intrinsically high levels of quality and reliability, it is important to note that COTS components are not designed with the space radiation environment in mind, nor are they built to withstand the rigors of launch into space,” Kawam says. “Additional engineering and logistical work is necessary to characterize their radiation behavior and mechanical characteristics, and to evaluate workarounds for any weaknesses.”
The challenge is that the requirements for cost reductions are extreme in some cases.
“In the past our customers would require a $3 million box for geosynchronous (GEO) applications; now for the constellations they want a low-earth orbit (LEO) box for only $300,000,” Jordan says. “For these requests, reliability requirements may be changed – in some cases down from 20 years, and more in the 7- to 10-year range. They want an 80% to 90% reduction in costs while maintaining or getting near the radiation performance of previous requests.”
“Rad-hard electronics themselves trend about five times more expensive than typical COTS products,” Festa notes. “There is a bit of ‘buy and fly’ when it comes to COTS electronics, but then you run into reliability issues.”
A spectrum of reliability
There are multiple levels of reliability when it comes to electronic components for space, Jordan says. “When it comes to supplying electronics for these satellites, we look at it as a spectrum from pure COTS parts to industrial parts to automotive-grade to high-rel parts.
“People ask me why not use an automotive-grade part in a satellite if it can be upscreened and is used in cars, which have stringent reliability standards for safety reasons,” Jordan continues. “I always answer: Because I don’t know about the radiation performance. The automotive industry puts in multiple controls to reduce risk of supply-chain distribution, but because the volumes are so high they have to use multiple fabs for the same parts, retooling often. Every time you retool, risk is presented in terms of radiation performance.
“My customer wants that radiation performance and the support that we can provide for parts used in critical radiation environments,” he adds. “They won’t get the same from the automotive-grade supplier when it comes to traceability, pedigree of design, radiation performance, etc.”
To compete, the traditional rad-hard designers have to rethink their entire development process. “Some space electronics manufacturers catering to the commercial-constellation and small-satellite market – in order to cut costs – are stripping down the traditional high-rel component design process, Quiroz says. “They are using plastic instead of ceramic packaging, eliminating extra quality production processes, all while using the same base die then releasing it at a low price. For rad-hard testing they see if the die has rad-hard legs to it and then upscreen from there.”
“The military and government customers want those price and time-to-market advantages, but not at the cost of eliminating the radiation tolerance of the components,” Quiroz notes. “The challenge for high-rel component designers is to maintain reliability – or some measure of it – while moving to a commercial production flow.”
It comes down to acceptable risk. “We anticipate that space programs may face situations where accidental damage could happen to components that are intended for, or already integrated into, flight hardware,” Kawam says. “For space programs attempting to acquire lower cost systems using COTS components, the need for technical support from component suppliers to meet the mission assurance objectives should be top of mind.”
Solution example: TalRad
“At Apogee Semiconductor we developed specialized IP for high-rel space electronics applications, called TalRad,” Quiroz says. “The long form is Transistor Adjusted Layout for Radiation. It all starts at the transistor-level design. We partnered with TSI Semiconductors to provide the TalRad process design kit (PDK) in TSI’s 180 nm high-voltage process. (Figure 1.)
“We take the commercial process for a standard transistor and reduce the total ionizing dose (TID)-induced leakage with minimal size impact compared to annular transistors. It becomes more of a rad-hard transistor that looks and feels like a standard cell,” he adds. “With this IP we are enabling customers to do their own IC [integrated circuit] design. Using the TalRad IP cuts the design time in half, improving time to market. As opposed to designing IP from scratch manually, we are now designing with known foundational components in rad-hard designs. Right now TalRad can operate in LEO applications and we can take it even higher.”
Solution example: Lean Rel
Cobham developed a new product line it calls Lean Rel. “We’ve basically gone through and optimized assembly flows, instituted lean manufacturing practices, removed a lot of value-added screens, etc.,” Jordan says. (Figure 2.)
Lean Rel is comprised of microprocessors and microcontrollers, as well as memory and interface ICs, with radiation-hardened, QML-level reliability. Cobham will release a suite of products during 2019, with an Arm microcontroller and controller area network transceiver available now. The LEON microprocessor, low-voltage differential signaling receiver and driver, synchronous dynamic random-access memory, NOR flash memory, voltage supervisor, and additional product releases are planned throughout the year. Depending on quantities, pricing for LeanREL products is less than 60% of QML variants.
Solution example: Smart Backplane
“We’ve taken a different approach to enabling COTS use in space through our Smart Backplane design for data acquisition systems,” Festa says.
“It is based on our COTS equipment designed with standard components that are susceptible to single-event latchups (SEL). The Smart Backplane overcomes the latchup by recycling the power. The backplane contains rad-hard components and senses the current coming into each individual module. When a higher current indicates that a SEL has occurred, the backplane recycles power to only the impacted module which clears the latchup. It is essentially rad-hard equipment protecting COTS components and providing radiation tolerance. Our signal conditioning modules are made with commercial parts. These modules are configured to meet customer requirements in a data acquisition unit – our KAM-500 – that is then submitted to a series of space-related environmental acceptance testing.”
“The Smart Backplane has been radiation-tested by NASA, surviving 100 latchups without degradation,with a total dose equivalent to five years on the International Space Station,” Festa notes. “It is currently being used in numerous space applications including the Boeing CST-100 manned spacecraft program (see lead photo), launch vehicles, ISS experiments, etc.” This is important, as heritage is so crucial to the space market, he adds.
Solution example: Microcontrollers for space
Microchip’s most popular microcontrollers, Kawam says, are the ATmegaS128 and the SAMV71, an 8-bit AVR and 32-bit ARM microcontroller, respectively. The ATmegaS128 is currently designed into a large constellation, to be launched in the near future. It provides a common remote terminal interface for the on-board computer communications to each subsystem. The SAMV71 is being used in a broad array of applications from star trackers to optical communications.