Atmel Introduces Rad Hard Microcontrollers

The Internet is full of extremely clever people, and most of the time they don’t realize how stupid they actually are. Every time there’s a rocket launch, there’s usually a few cubesats tucked away under a fairing. These cubesats were designed and built by university students around the globe, so whenever a few of these cubesats go up, Internet armchair EEs inevitably cut these students down: “That microcontroller isn’t going to last in space. There’s too much radiation. It’ll be dead in a day,” they say. This argument disregards the fact that iPods work for months aboard the space station, Thinkpads work for years, and the fact that putting commercial-grade microcontrollers in low earth orbit has been done thousands of times before with mountains of data to back up the practice.

For every problem, imagined or not, there’s a solution. Now, finally, Atmel has released a rad tolerant AVR for space applications. It’s the ATmegaS128, the space-grade version of the ‘mega128. This chip is in a 64-lead ceramic package, has all the features you would expect from the ATmega128 and is, like any ‘mega128, Arduino compatible.

Atmel has an oddly large space-rated rad-hard portfolio, with space-grade FPGAs, memories, communications ICs, ASICs, memories, and now microcontrollers in their lineup.

While microcontrollers that aren’t radiation tolerant have gone up in cubesats and larger commercial birds over the years, the commercial-grade stuff is usually reserved for low Earth orbit stuff. For venturing more than a few hundred miles above the Earth, into the range of GPS satellites and to geosynchronous orbit 25,000 miles above, radiation shielding is needed.

Will you ever need a space-grade, rad-hard Arduino? Probably not. This new announcement is rather cool, though, and we can’t wait for the first space grade Arduino clone to show up in the Hackaday tips line.

88 thoughts on “Atmel Introduces Rad Hard Microcontrollers

    1. iPhones and ThinkPads will likely survive in the ISS just similar to on the ground. The ISS itself is radiation hardened for the warm bodies that inhabit it. Low earth orbit, although it has more radiation then on the ground is still not extremely high compared to open space where geosynchronous satellites are.

      Take that iPhone outside the ISS or further into space and it will have a lesser lifetime.

      There are other ways to radiation harden things though like extreme shielding so that the contents of a satellite can be protected similar to how the inhabitants of the ISS are not cooking.

      1. “The ISS itself is radiation hardened for the warm bodies that inhabit it”
        it is well documented that they have not solved how to block the radiation. cosmic radiation can penetrate 10 foot of steal. please show some specs on how they are radiation hardened as applied to the iss.

        1. Well, you are right, there is still some cosmic radiations. But those are just flashes (several astronauts experienced it, google it) and could maybe reboot a few laptops, nothing bad I believe.
          You can notice that somme old onboard video cameras have a LOT of dead pixels too.

    2. No, no ,no. You’ve got it all wrong friend. They ARE faking something, but not space travel. You see, there is no radiation in space. Radiation is the result of man made chemicals whose sole purpose is to perpetuate the “space is dangerous” spiel of theirs and to control the population numbers. This way, they can keep inter-galactic travel to the Uber rich. After all, there are only so many Tarantulan females to go around…

      Also, if you’re going to walk around posting half-assed quotes of my work, you should provide proper attributions, or maybe I should call the Copyright office, they do exist, right?

    3. Uhhhh. Realize, these devices are generally inside spacecraft most of the time. Like the crew, they spend a limited time in actual “space” and are for the most part inside shielded space craft, net getting bombarded by charge particles from the sun and the cosmos. So comparing iPhones and iPads to hardened circuits designed to be in orbit for 20yrs or more is a bit of a mistake for certain. My only beef with consumer grade electronics in orbit is that it will, by its nature, become space junk in some short number of years and is susceptible to unpredictable behavior and therefore a hazard to navigation (See the Planetery Society’s solar sailer for an example of unexpected behavior, although, since they are smart people, it *was* on a decaying orbit on purpose). So if its not on a planned decaying orbit to reenter the atmosphere in a reasonable amount of time, you are better off having something that can at least send a transponder signal so that everyone can know where it is.

      I am thinking, Atmel’s RAD hard ATmega128 is just *friggin* awesome.

    1. That’s the big question. I was thrilled just to see native -55 tolerant AVRs (TI is the only company that makes reasonable -55 tolerant MCUs) but since they’re rad-hard, they’ll probably be crazy-priced.

  1. Any chance that the iPods and Thinkpads last is because they are:
    1) well within Earths magnetosphere
    2) INSIDE the ISS, which provides a fair bit of additional shielding against charged particles.

      1. http://blogs.esa.int/promisse/2012/06/13/radiation-on-the-iss-mapping-in-3d-testing-materials-for-future-exploration/

        “The radiation environment inside the ISS is much milder than outside the ISS. That is because of the shielding effects from the complete structure of the ISS. The ISS was never built to provide radiation shielding as such, it was built as a mechanical structure that, almost as a side-effect, offered shielding to the crew.”

        According to wikipedia, the crew is exposed to 50-2000 mSV during a 6 month mission.

        The radiation in free space at the orbit of the ISS is approximately 100 mSv per day, so at least one order of magnitude greater than inside the ISS. The radiation consists of mostly high-energy particles, which are quite effectively stopped by the ISS hull.

  2. The ISS actually provides some shielding from radiation, though the shell of any satellite would also provide some shielding from radiation. The ISS probably has more, though, due to having human occupants aboard.

    1. The ISS has more due to being much much bigger. The modules are also filled with stuff, racks and racks full of equipment and supplies carried on-board by the ton with each resupply flight, whereas a small 10x10x10 cm cubesat only offers a cardboard-thin barrier against fast particles.

      1. The original ipod came out in 2001 and had a 6gb hard drive (and soon was available in 20gb). Flash memory wasn’t anywhere near cheap and small enough for applications like that for another 5 years or so.

      2. My iPod Color (yup “Color was in the model name because the previous ones were a monochrome display) was originally 40GB. The 1.8″ Hard Drives weren’t really known for longevity, so It currently houses an old unused 8GB CF card in a CF -> 1.8” HD adapter.

        The iPod Color was also one of the last dual power iPods too (Firewire or USB).

        With the CF card in it, I have often wondered how large of a battery I could stuff in there, since I have the wider case for that 40GB HD (I’m guessing it was a dual platter).

  3. Not all transistors are created equal. Radiation in the form of highly energized particles strike the gates of transistors and cause them to toggle state. For many processors, they compute the wrong thing for one cycle then go about their business. For many other processors, though, the transistor toggles and doesn’t toggle back until a power cycle. The problem with that is CMOS technology, where you have a pull up and a pull down transistor. If both transistors turn on, you get a short! If you’re lucky this crashes your processor’s power supply and it reboots. If you’re unlucky this melts the transistors and your processor is fried.

    AVR’s microcontrollers unfortunately tend to do the latter. This explains why they need a rad hard product line. Other manufacturers use different techniques and you can find non-rad hard chips that don’t blow themselves up when blasted with radiation. That’s why thinkpads and iPads and all your other pads can last for years in space. I guarantee they crash and reboot every few days though!

    The tricky part with using a non-qualified part for a student project is that, unless you have a proton beam on campus, there’s no way to know whether or not your processor will blow up or walk it off.

      1. And I bet you can improve the latch up-performance even further by under-volting the Atmega. Get the supply voltage below 1.5 volts and I bet the parasitic thyristor from the CMOS process won’t turn on no-matter what you do.

    1. Umm, not to be a tool, but…

      Wouldn’t 100% depletion force decay? I thought the general use of “depleted” is meant to indicate that a material has been used to a point that it is no longer viable for its intended use, but its still pretty bad? Kinda like a depleted battery still has juice, but not enough to power most circuits?

      1. No. Depletion simpy means removal of the unstable isotopes from the naturally occurring mixture of the element.

        The boron-10 isotope is unstable, and upon being hit by cosmic rays decays itself, causing secondary emission of radiation that produces local effects that are much stronger than the original particle that hit it. Depleted boron is purified boron-11 isotope which is stable and resistant to radiation damage. It basically “absorbs” the radiation without causing secondary emission.

          1. Naturally occurring boron is a mixture of multiple isotopes. Mainly boron-10 and boron-11. The stable isotopes are 10 and 11, which means that it takes something like the lifetime of the universe for them to decay spontaneously.

            There are other isotopes of boron, but they are so unstable that they exist only for fractions of a second, so they aren’t found in nature except as products of the spontaneous decay of other radioactive elements, or because of neutron capture such as boron-11 becoming boron-12 which then exists for about 50 milliseconds, releases an electron and transmutes into carbon, or an electron and a helium particle, and becomes beryllium in a more rare case.

            The transition of boron-10 into boron-11 by neutron capture instead often results in the isotope splitting into a charged lithium and helium particles instead, which fly around at high speed like shrapnel from a grenade, which is what causes the troubles. This property is used in cancer treatments by injecting a tumor with boron-10 and then irradiating it.

            There is about 20% of Boron-10 in naturally occurring boron, and when you remove it you get what is called depleted boron. The word “depleted” simply refers to the fact that something has been removed. If we were more interested in having boron-10 and instead removed the boron-11, we would call pure boron-10 “depleted”.

  4. Am I the only one here to have actually done the work to assess COTS silicon lifetimes in LEO? The work is relatively straightforward, needing some basic numbers:

    1. What is the radiation flux in the operating regimen? This is collectively known as “space weather”, and includes EM spectrum and intensity; gamma and beta spectrum, intensity and direction; and the ever-present cosmic rays. The earth’s magnetic field is a factor, which makes the intended orbit a factor (which specific LEO orbit).

    2. What is silicon’s sensitivity (aka, “low cross-section of absorption”) to the above fluxes? Much radiation simply passes right through silicon, leaving the chip unaffected. Other radiation can be countered (but never completely stopped) with simple, inexpensive, lightweight shielding. (Tip: Most shielding converts radiation of one kind into another, such as high-energy beta into X-Rays, to which silicon may be less sensitive.)

    3. Once you know what’s flying about, and what’s there to get hit by it, you take your device size (size of the silicon chip) and crunch the statistics to determine how often it will get hit (the probability distribution function). Then you look at the volume within the chip to determine how those interactions will affect the circuits on/in the silicon (another PDF).

    By far and away the most common radiation affect is local ionization. This can cause bits to flip, latches to unlatch, and a host of other glitches. Fortunately, these are often easy to recover from, by first trying a whole-chip reset, and then a full power cycle (if needed).

    However, sometimes the ionized path can connect a power rail to ground, in which case it is vital to immediately cut power long enough for the ionized path to neutralize (the “recombination time”). A quick-acting power monitoring circuit is a very common feature in nearly all space electronics.

    On occasion, a hit can dislodge atoms in the silicon matrix, which also contains a relatively small number of dopant atoms. When a silicon atom is dislodged it is seldom a problem. But when enough dopant atoms are dislodged in a small area, a transistor can permanently stop functioning. It is the accumulation of permanent damage that ultimately determines the useful lifetime of a chip in orbit.

    One surprise is that a significant (but minor) problem comes not from silicon or its dopants, but from the aluminum metalization used to interconnect the circuits on the chip. Aluminum atoms wind up going everywhere, as if they were being sprayed around, increasing the odds of open conductors and shorts between aluminum conductors. Some rad-hard chips use other metals for the metalization layers.

    The important thing to understand is that it’s all statistics: Two iPads on the ISS, sitting right next to each other, could experience radically different actual lifetimes.

    So, given all the above, what can be done to make electronics more robust in space?

    1. As mentioned above, the first step is to ensure quick power cycling when current spikes occur. This in turn implies having fast boot times to ensure minimum down time.

    2. Use multiple redundant processors. Given enough processors doing the same thing in parallel, only the voting logic that connects them needs to be rad-hard! This is the path many systems take, but it does multiply the total amount of electronics and the power used.

    3. Add selective shielding. Shielding increases weight (and launch cost), but it is often the case that most “upsets” are caused by a relatively narrow portion of the incident radiation flux: Knocking out this small part can greatly extend the expected lifetime.

    4. Use rad-hard devices. What makes a circuit rad-hard? There are silicon chip layout designs that minimize the risk of shorts to ground, and selective doping of the base layers can also inhibit other classes of short circuits. Using larger transistors means more atoms in each one, meaning a harder hit is needed to affect any single transistor, This is why many rad-hard circuits are made using older production technologies. On-chip self-monitoring and protection circuits can also be added to mitigate radiation effects and damage.

    These can all be bundled together into a single hybrid device packaged in shielding. Which generally only the military can afford. But even taking only the simplest and least expensive measures (current-spike detection and redundancy) can yield an extremely robust space system.

    However, there is only so much that predictive calculations can accomplish: At some point you need to test your system and MEASURE how it performs in various fluxes.

    In one system I worked on, we couldn’t get much information about the internals of the inexpensive commercial processor we wanted to use, so we decided to combine a live test with some reverse engineering. We were most concerned about the impact of cosmic rays, and fortunately it is possible to simulate these affects by slamming high-energy ions into the chip. Selecting different ions give you various mass and charge, and then how hard they are accelerated gives you the rest of the energy. Varying the angle between the beam and the chip changes the total silicon thickness.

    Now, these ions don’t do well in the atmosphere, so the experiment needs to be carried out in a vacuum. And the ions also don’t do well with a chip’s epoxy encapsulation, so we had to “de-lid” several devices to get a few that worked after their lids were removed.

    An excellent source of accelerated ions is the Tandem Van de Graaff heavy ion accelerator at Brookhaven National Labs on Long Island, New York (https://www.bnl.gov/science/accelerators/TVdG.php). A larger company had a series of test runs scheduled, and we were able to piggyback on their down time to run our own tests.

    We didn’t get all the data we had hoped for, but the data we did get indicated that we were exceedingly lucky: Our selected inexpensive commercial processor was surprisingly rad-hard! It was so good we decided to eliminate redundant processors and shielding, and rely only on our current trip circuits, the watchdog timer, and careful software design (including fast reboot). In fact, the reboot was so fast that we decided to power-cycle the processor every so often, just in case there were radiation effects present that we weren’t detecting.

    Unfortunately, like so many other small space experiments, our mission never flew. Launches are always overbooked, and it is difficult for a low-priority experiment to get to the head of the line.

        1. No, I’m an embedded/real-time systems software engineer who has had the chance to work with lots of truly brilliant scientists, researchers and investigators doing amazing things. I like working on systems in the “real world”, especially when that world is harsh or extreme. Which means I’m also a test engineer, trying my best to break everything I make.

          I’ve been doing this (R&D) for over 30 years, so I have lots of stories to tell, and love leaving the longest comment in a thread. However, my favorite story is also my simplest: Being a curious engineer is the best career I can think of.

          But there are downsides: More than once I’ve been given the “up or out” choice, an employer thinking I was overpaid as an engineer, which somehow magically made me better qualified to be a manager. I’ve also recently been seeing age discrimination (today is my 59th birthday), but that’s largely handled by doing good job interviews. I also find I easily get bored once a product enters maintenance, so I’m often taking (temporary) pay cuts to work on the new hotness. I’ve spent over a third of my career on my own as a solo contractor, which is more often a form of “self (un)employment”.

          And, true to form, I start a new job next week! Can’t wait to see what new stories I’ll accumulate.

          I am not looking forward to retirement. Not one bit.

          1. “But there are downsides: More than once I’ve been given the “up or out” choice…”

            Man, what a terrible thing to read. I’m 24 and just recently entered into my first real job after getting a degree in computer engineering. I went into the field because I love everything about it, and I have a huge addiction to learning. Part of me wishes I could go back and get degrees in almost every field of engineering out there. So needless to say, I feel some similarity here and I’ve thought about this exact topic before. My wife has half-jokingly suggested before that I better move into management to make more money one day. I can’t stand the thought and I know I’ll never do that, but it’s a shame that some employers really do resort to that. Just out of curiosity, can you give more info about this? Does it happen often or has it happened to you often? Is it something that the majority of employers would be likely to do? And so really, one can expect a fast-approaching salary ceiling if steadfastly avoiding management?

          2. My degree is also Computer Engineering, but the meaning of that varies between schools. When I got my degree, I had to do a year of analog electronics (mainly S-domain stuff), and two years of digital (transistors to gates to adders to full-up computers), then all the rest was CS.

            My ticket to career freedom has been that I ALWAYS keep a good savings balance, enough to support me for at least a year (generally 2). So I never had to stay in a bad or boring job because of financial insecurity. In fact, I worked my way through college, and not only graduated debt-free, but with $10K more in savings than I started with. I always live within my means, and pay off my credit card balance every month.

            It is that financial freedom that empowered my curiosity. Which I’ve followed like it was the Pied Piper of Hamlin.

            But it isn’t all work and no play! I’ve been a scuba diver (for a decade), paraglider pilot (also for a decade), and now I’m a triathlete (for the rest of my life). All of these physical hobbies can also have technical sides, but they get me out of the house and make it easy to stay fit. And fitness is a great way to both get rid of and tolerate stress. I do have other tech projects at home, but they’re mainly for rainy days or the flu.

      1. Unfortunately, that work was under an NDA (as is so much commercial research).

        We found that particular processor because that company had several peripheral chips had a reputation for radiation tolerance (that’s “sorta-kinda rad-hard, but with only anecdotal evidence”). We found what production lines were used for those products, then looked at what else was made on those same lines. And lo and behold, we found a microcontroller that was perfect for our needs.

        But there are no shortcuts here: When using an uncertified processor, you still have to do the design and analysis work (and the testing) to show that it will meet the mission requirements. We were lucky to be very friendly with a local company that made satellite electronics.

        1. Bummer. Thanks for taking the time to reply though. I guess I have some work for a friend in the radiation business. Has to be done either way; I expect even something simple like a 555 based watchdog can fail in unexpected ways..

        2. Hmmm. I worked at Atmel before 2008 and many of their ATmega were made in their (former Honeywell VHSIC) Colo Spgs 0.35u fab. Not sure if ATmegas in industrial /automotive reliability / quality are still fabbed there. Atmel did make RadHard Sparc and 68xxx processors thru their Atmel Grenoble group so they knew deep space design needs back when LHC was in construction.

  5. Hi Guys, I am a space engineer and my company actually builds satellites and we also have some equipment on the ISS. None of the stuff actually is space rated just space qualified. So we take commercial grade electronics and throw all kinds of tests on them – if they survive they will fly (and live) in space and if they dont then well we have to change another component. As you already pointed out ISS has a very low orbit that helps with radiation and the inclination is 50 something degree; which means that it does not fly over the north pole which also reduces radiation levels quite significantly. The more you get to an equatorial orbit the lower will be the radiation. A typical polar LEO satellite @650km orbit will accumulate about 3 krad per year behind 2mm Aluminium shield, on the ISS (which also has 2mm AL shield) its only a few hundred rad per year. In an equatorial LEO orbit such a the one where our next satellite will go to you barely have 100 rads per year behind the same shield. You can actually calculate the ratiation in each orbit using the radiation calculation tool Spenvis (https://www.spenvis.oma.be/), its the stuff that we also use to calculate the radiation profile for our missions. Since your off the mill processors will aproximately last for 10krad (the ones that are also in ipods and notebooks, the 30krad of the AVR is no big deal I know processors that last for 50krad easily, its more interesting that they dont latch) there is no problem using the stuff in space. We have done that so; have other companies and literally hundreds of universities. Beyond the total dose there are radiation induced latch ups which might kill your electronics. Here a cosmic particle crashes through your chip and creates a short cut in your transistor which will exist until you switch off the device and if you dont the short kills the circuit. For this you have to build latch up protection (basically a fast resetable fuse in front of the chip that you want to safeguard) good things is though that some of todays processors (especially those with low voltages and silicon on insulator process) are latch up immune by design (they simply dont latch). Other effects are much more likely to kill your cubesat in space but if you are careful you can get away with quite a bit; some people have even flown phones (or arduinos) ;)

      1. I wouldn’t mind being privvy to this too! I work in research for nuclear decommissioning robotics, my research group was pretty eager to find a supplier for these atmel micros as judging from the documentation on atmels website they’re pretty cost effective!

  6. Quote: “… has all the features you would expect from the ATmega128 and is, like any ‘mega128, Arduino compatible”

    And no this is not ‘negative feedback’ to [Benchoff]. I am hoping someone can point out what I may be missing.

    I just looked in my Arduino IDE and there is no board that has a ATmega128 so I don’t get were Arduino compatible comes from.

    Sure the toolchain supports a lot, if not most ATmega’s. I have added ATmega664 and ATmega1284 to my IDE and I have over 30 board configs to select from but all of these need a bootloader to be called ‘Arduino Compatible’.

    Technically – yes you can use USBasp programming but that is the exception and definitely NOT the convention.

  7. Hey Brian,

    Thanks again for sharing your detailed knowledge about the radiation effects on microcontrollers. I don’t know how you have the time to write all these awesome and of so informative articles here AND have a full time job in the aerospace industry working on rad hard electronics.
    I love your opening paragraph. Lets break some of that down eh?…

    “Internet armchair EEs inevitably cut these students down: “That microcontroller isn’t going to last in space. There’s too much radiation. It’ll be dead in a day,” they say. This argument disregards the fact that iPods work for months aboard the space station, Thinkpads work for years, and the fact that putting commercial-grade microcontrollers in low earth orbit has been done thousands of times before with mountains of data to back up the practice.”

    1) Good shot at the “arm chair EEs”. I forgot and maybe you can remind me… What school did you get your EE degree from? You at least have a masters right? PHD? Someone must have hacked your HAD profile because you are the only one with ZERO qualifications listed.
    2) “iPods work for months aboard the space station”… If I was going to spend millions of dollars on an instrument that was going to be launched into space, I would want it to last more than “months” but I guess that’s why I’m not in the aerospace industry huh…. Also maybe you can comment about why the guys on the space station keep their IPODs OUTSIDE the station all the time and not bring them inside where they are actually pretty well protected by the station itself? I guess I just don’t understand space much huh since I would keep my IPOD inside.
    3) “the fact that putting commercial-grade microcontrollers in low earth orbit has been done thousands of times before with mountains of data to back up the practice”….. So I’m not a professional “journalist” but I always figured that if you were presenting “facts” as you stated above that you might want to list your references there. With your crazy busy schedule I’m sure you just forgot. It shouldn’t be hard to list some of the “Mountains of data to back up the practice”. No hurry backing up your seemingly arbitrary and uninformed opinions though… . Whenever you have a free minute.

    Keep up the good work,
    AC

    1. “the fact that putting commercial-grade microcontrollers in low earth orbit has been done thousands of times before with mountains of data to back up the practice”….. So I’m not a professional “journalist” but I always figured that if you were presenting “facts” as you stated above that you might want to list your references there.

      I can only speak about the satellites that I know (its not thousands but none of them uses space grade micro controlers)

      – TUBSAT A build by TU Berlin, in Berlin Germany launched in 1991 lived for 16 years (died of battery faillure) and used multiple Hitachi H8
      – TUBSAT B launched in 1994 lived for 42 days and used multiple H8 (died of malfunctioning ham radio)
      – TUBSAT N & TUBSAT N1 launched in 1997 lived for 2 and 2.5 years until reentry used multiple H8
      – DLR TUBSAT launched in 1999 lived for 9 years (died of battery faillure) multiple H8
      – Maroc TUBSAT launched in 2001 lived for 9 years (died of battery faillure) multiple H8
      – LAPAN TUBSAT launched in 2007 lived for 8 years, not yet (fully dead) multiple Hitachi 32 bit processors
      – LAPAN A2 launched in 2015 still living uses multiple 32 bit Hitachi processors
      – Beesat 1 Cubesat launched in 2009 uses 60 MIPS Phillips processor
      – Beesat 2 Cubesat launched in 2013 uses 60 MIPS Phillips processor

      Basically all satellites from the small satellite market leader SSTL used an industrial version of the intel 386. They just recently changed to an ARM based processor in their SSTL X-50 platform. I dont want to write them all done but its tons of satellites in the last 30 years. Here is a source:

      http://www.sstl.co.uk/getdoc/d638251e-4395-455c-b0c8-5935f85a599a

      In the last 30 years universities and companies have sent small satellites into space. Increasingly many in the last decade. My personal guess is that it was around 20 to 30 in the 2000 to 2010 period and we are now having more than 100 satellites launched per year since 2010. Euroconsult says that about 600 small satellites have been launched in the last 10 years.

      http://www.euroconsult-ec.com/shop/space-industry/64-prospects-for-the-small-satellite-market.html

      And its growing

      http://www.satellitetoday.com/technology/2015/06/16/nsr-sub-100-kilogram-satellites-to-exceed-2500-next-decade/

      Each of these satellites (whith the exception of less than 10% for military or large space agency missions since the others simply dont have the budget) easily has on average 10 micro controlers on board (all of them commercial grade) which in my eyes seems to back the statement that already thousands of commercial grade MCUs have flown into space.

  8. So we have RoHs compliant chips now, lead free, so they rebrand that ol sh.. and claim it’s space compatible is this just marketing? Does environmental legislation not apply to space? Cubesats are coming back down eventually.

    1. RoHS solder does not deal well with low temperatures. IT will become brittle and break, which is not desirable in satellite electronics.
      And in any case the amount of lead in electronics is miniscule, bullets and stuff like lead acid batteries used in cars and UPS’es have significantly more lead and those don’t always get recycled properly.
      Not to mention bullets.

      But the main reason is reliability. Aerospace, military and medical equipment are free from RoHS for a very good reason. They need high reliability, which RoHS currently cannot offer.

      Sn/Pb all day long :)

    1. You will need the same kind of latch up protection but depending on how close you actually are to the blast you also need much much more total dose resitant. There are processors out there that can withstand 1 Megarad and are totally latch up immune. In space this stuff is used for deep space missions say to Jupiter

    2. I don’t think rad-hard devices will automatically survive a possible EMP from a nuclear explosion. You’d want to build a Faraday’s cage around your circuit and include some kind of overvoltage protection.
      EMP: short-term. Gamma radiation: long-term

  9. Don’t forget that there are reasons other than being in space to need radiation tolerance, there are lots of radiation generating environments here on earth. I worked for a company for a while that made particle accelerators for radiation cancer treatment. Electronics getting whacked due to radiation exposure was a real problem in some instances.

  10. Sweet! Now we can explore Chernobyl and Fukushimi with our homebuilt drones! How many amp hours do I need to get my quad-copter across an ocean?

    Assuming these are more expensive I don’t think I would bother building them into cube sats though. Why worry about weeks of radiation exposure killing your electronics when it’s going to burn up in re-entry in a matter of days anyway?

    If you want your cube sat to last longer you need to get it into a higher orbit. To do that… your most imediate concern is the billions of dollars you need to raise. When your local hackerspace, school or ham club manages that task I’m sure radiation hardening your electronics will seem like childs play.

Leave a Reply to meCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.