Design News is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Embedded Hardware in Space. It’s Hard.

Embedded Hardware in Space. It’s Hard.
Putting things into space has long been a human fascination, but it hasn’t always been easy. Learn more during talks by Eli Hughes and Peter Ateshian at ESC Silicon Valley next week.

Putting things into space has long been a human fascination, but it hasn’t always been easy. Indeed, sending hardware into orbit can be challenging to say the least. “It’s really hard,” said Eli Hughes, a research engineer at Penn State ARL, who will be delivering his talk at ESC Silicon Valley this year on how engineers can go from a base idea to getting something prototyped and ready to go to space.

The first challenge, said Hughes, is finding parts to build a prototype with. “Your component selection goes down from millions of parts to not many parts,” he explained, noting that anytime one is designing something for quick operation, state-of-the-art goes out the window. “You’re always bumped back five, 10, 15 years. That’s the first challenge.”

Adding to that challenge, of course, is then how to implement an algorithm on the parts you do have available, while keeping the entire project within budget, because using space-qualified parts can be expensive. That means finding the best available proxy parts that would work in space to test your theory before building your actual device with proper space parts.

ESC logoTake a Journey with "Mr. X."  Designing embedded systems for space applications is both costly and difficult. A particular challenge is developing lower-cost hardware to serve as a development model before designing fully qualified flight-ready hardware. This session, at ESC Silicon Valley, Dec. 6-8, 2016 in San Jose, Calif., will tear down hardware used for the development of a high-speed X-ray event detection processor. Register here for the event, hosted by Design News ’ parent company, UBM.

“You have to cobble together your own design tools, to prove the concept and get the investment money,” Hughes explained, noting that engineers had to pull together a development system representative of parts they have available just to prove that the algorithm or measurement system will work. Once that hurdle is overcome, it’s easier to get money to fund the next step, building flight-ready engineering models.

“We have an idea, we have an algorithm, but then to move it into hardware is really tough.”

Hughes’ team should know. They have been developing a new circuit to do high-speed X-ray event detection in space, which the team has fondly dubbed “Mr. X.”

“A good way to describe it is that what’s out there now is standard definition video and we’re trying to move it to a high-speed high-definition video.”

The "Mr. X" design incorporates a Xilinx Virtex-5 FPGA, high-speed QDR-II SRAM, and a ARM Cortex M0-based microcontroller to implement the core processing architecture. One of the constraints of this design was to use components that offered a clear path to rad-hard, flight-ready hardware. 

The "Mr. X" board allowed the engineering team to demonstrate the core functionality of the processing concept while being able to significantly improve the technology readiness level of the instrumentation.

The space-qualified FPGA Hughes’ team wanted to use, made by Xilinx, was based on a commercial part roughly 10 years old.

Unable to simply go out and buy the right development hardware, or source bits and bobs from eBay, the team found hardware that was fairly close and decided to use that. “But halfway through the process of proving our logic, we needed to wire in some additional hardware, and that was just impossible, there was just no way to get the circuits we needed with the development hardware, so we decided to use the funds we had to build our own development system, so we could prove that what we had with commercial parts worked,” explained Hughes. Every part on the current board has a path to a rad-hard equivalent, so the team can get as close as they can with the money they have, and should be able to make a pretty good case to investors.

“With ‘Mr. X’ we have a pretty good representational model. We’ve gone a couple of steps up from what we call technology readiness level (TRL).”

When designing for space, Hughes noted one is automatically working with expensive parts, even the commercial equivalent parts. “I probably have $12k in parts on one of these boards here,” he said, adding that one had to have a great respect for the design process and a lot of discipline going into a board design, because typically teams only have enough money to get one shot.

“The original budget didn’t include designing this hardware, so we had to get it right the first time. We had to really sweat the small stuff. Looking over every detail we possibly could, running as many simulations as we could, having as many design reviews as we could,” he said.

The result in Hughes’ case is a 20-layer board that uses a fairly exotic substrate and cost quite a bit to get made. Hughes said he couldn’t stress enough how important it was to have a good relationship with an assembler to help engineers through the process. “I literally spent a week at the assembler,” he said, noting that when his team received the boards, they had the assembler make up partial sections and ensure they were there to make any fixes and get the process right. 

“Make sure you have a very methodical, well-planned, and disciplined approach,” he re-iterated.

“Of course, spending a lot of time on it costs a lot, but it costs a lot more if you don’t get it right the first time,” he added.

The ESC talk, said Hughes, will be particularly valuable for engineers who are interested in FPGAs, especially FPGAs for going into space, as well as people looking to learn about how hard it is to prototype for space.

“A lot of the things you might run into in space are not necessarily the things you might think of; for example, one of the biggest challenges is dealing with heat. A lot of people think space is cold, but it’s not, it’s a vacuum, so it’s really hard to get heat to dispel from anything that’s generating heat,” he explained, adding that engineers really had to think about things like that as well as their general software development cycle.

“In space, things can’t crash. They have to be rock solid,” he said.

Meanwhile, another ESC speaker looking to the final frontier for embedded hardware is Peter Ateshian, Faculty Research Associate Lecturer at the Naval Postgraduate School.

Ateshian believes the Internet of Space has some resounding benefits, including the ability to connect anywhere, anytime, without infrastructure or power.

His team at the NPS has been developing Femto satellites, powered by onboard solar cells and which transmit with CDMA frequency so the signals can be received with a standard cell phone. Each satellite orbits a point about three times a day, so Ateshian notes that with around 60 of them, one could get near continuous coverage.

ESC logoInternet of Space.  Internet of Space (IoS) is an embedded device application with a 1" x 1" UV stabilized PCB containing a CDMA radio transceiver, MEMS magnetometer, gyroscope, inertial measurement unit (IMU), switched ECC RAM, and thermal sensors. This embedded device platform is called the Femto Satellite. Learn more at ESC Silicon Valley, Dec. 6-8, 2016 in San Jose, Calif., will tear down hardware used for the development of a high-speed X-ray event detection processor. Register here for the event, hosted by Design News ’ parent company, UBM.

The Femto satellite program is an evolution from the days of the Cube-Sat, some 15 odd years ago. Cube-Sats were cool, but a bit on the large side. People had also started trying to add various sensors and capabilities to them, so NASA and Cornell University teamed up to see how they could improve them, resulting in the Femto Satellite.

The NPS is currently working on the next generation to those developed by Cornell and NASA, applying Moore’s law to satellite technology and shrinking it down to a 1” x 1” PC board dimensional thickness for the entire device.

The 1" x 1" UV-stabilized PCB containing a CDMA radio transceiver, MEMS magnetometer, gyroscope, inertial measurement unit (IMU), switched ECC RAM, and thermal sensors costs less than $50 to make, including its $10 software-defined radio (SDR) dongle which maps the live frequency to your cell phone frequency (because you’re not allowed to transmit at cell phone frequency for obvious reasons).

The CC430 SoC is at the core of the Femto Satellite, providing all computing and communication capabilities. It combines an MSP430 microcontroller, which is clocked at 8 MHz and provides 4 kB of RAM and 32 kB of Flash memory, with a very flexible CC1101 UHF transceiver capable of output powers up to 10 mW and data rates up to 500 kbit/s. Both the MSP430 and CC1101 have flight heritage on CubeSat missions. An Arduino-based development environment, known as Energia, has also been ported to the CC430 to facilitate rapid code development and prototyping.

The Femto Satellite part of the Internet of Space means that effectively, a cell phone, tablet, or notebook can become a ground station via the low-cost SDR. It can also be useful for missions like asteroid detection, true random number generator (TRNG), protected CDMA communications, solar weather or CME monitoring, earthquake and tsunami detection, and radiation or cosmic particle detection.

“There are hundreds of industrial, commercial, and agricultural applications” said Ateshian, noting that if other simple pH, chemical, and salinity sensors were added to this IoS and IoT platform, it would increase that number even more.

“Not only can the Femto Satellites be deployed in space, they can also, with a little coat of shipping container primer, be deployed at sea and float on the ocean and operate as a sea sensor. That’s one of the very interesting IoT applications,” he added.

“They can also be used on the ground as a GPS navigation system, so if we lost our GPS satellites you could use a swarm of these to be your navigation, though it wouldn’t be quite as accurate,” he noted.

In space, mission time is usually six to eight weeks, and the satellites burn out on re-entry.

Interested in learning more about embedded technology in so far as it relates to Femto Satellites and the Internet of Space? Ateshian’s ESC talk should have you covered, targeted at anyone interested in the Internet of Space, or anywhere, anytime connectivity in power grid-less and infrastructure-less environments, as well as those interested in Earthquake, tsunami, radiation, and cosmic particle detection, or TRNG for cryptographic and communications applications.

Both talks sound out of this world to us!

A regular speaker on the tech conference circuit and a Senior Director at FTI Consulting, Sylvie Barak is an authority on the electronics space, social media in a b2b context, digital content creation and distribution. She has a passion for gadgets, electronics, and science fiction.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.