Advertisement

Blog

Jupiter: The IC Danger Zone, Part 1

Jupiter, Juno, and Why Space Qualification Still Doesn’t Cut It

I have always been fascinated with space exploration. The endless possibilities of discovery captured my imagination as a kid. It wasn’t until I began working with space development that I could fully appreciate what goes into the development of a new space craft. Watching the Juno probe perform its Jupiter Orbit Insertion maneuver brought tears to my eyes and inspired me to learn more about the probe and Earth’s biggest neighbor.

There are many things which engineers at NASA, Space X, and at a semiconductor company like ADI must consider when developing products for use in space that anyone else wouldn’t have to worry about for earth-bound applications, most notably of which is radiation (of course unless you’re working on a nuclear power plant). Radiation can wreak havoc on any CMOS device, causing glitches and even damage beyond repair in many cases. We’ll cover these failure modes in a little bit.

For those with limited experience working in the space market, any device which is targeted to companies like NASA must adhere to a set of standards set by the Defense Logistics Agency (DLA). Among these specifications are different radiation tolerance levels which a “rad-hard” space qualified device must pass. This typically entails a standard level of radiation testing up to 100Krads.

While 100Krads may sound like a lot of radiation, even a standard “rad-hard” device as tested to DLA standards wouldn’t stand a chance around Jupiter. Jupiter has an enormous magnetic field, and it’s extremely strong. If we could see Jupiter’s magnetosphere from earth, it might look something like this. See Figure 1.

Figure 1

Jupiter's Magnetosphere if it could be seen from earth; Photo Credit: NASA

Jupiter’s Magnetosphere if it could be seen from earth; Photo Credit: NASA

Just like Earth’s, Jupiter’s magnetic field is distorted by the solar winds. The tail shown here extends all the way past Saturn’s orbit! This strong magnetic field traps solar wind particles, as well as those ejected by the volcanic moon Io, and flings them around the planet at ridiculous speeds. It’s these energetic particles that Juno’s electronics will need to worry about. The radiation levels seen around Jupiter can reach upwards of 20 million rads. To put that into perspective, that’s the equivalent of being blasted by 100 million dental x-rays at once! Juno’s orbit will thankfully keep it outside of the worst areas of radiation, but it still must punch through these zones in order to take the data it needs. In its closest approaches, Juno will be just about 3,100 miles (5,000 km) from Jupiter’s cloud tops.

With all of that radiation, even a space qualified device wouldn’t last long without extra shielding. Juno’s sensitive and most important electronics are housed inside a heavy titanium box to block as much radiation from reaching the inside.

Let’s take a look at a few aspects of radiation testing that any DLA space device must go through.

<>

There is an inherent amount of radiation present in our solar system. Some of it comes from our own sun in the form of solar flares and solar wind, and some comes from other cosmic rays from other distant stars.

Imagine for a minute that we are testing a waterproof fabric. It would clearly have to be able to withstand a light mist that lasts all day as well as a sudden storm that rains cats and dogs and then goes away. This kind of basic resistance testing would be the same as the Total Ionizing Dose, or TID test.

An IC is exposed to gamma rays from a cobalt-60 source over a certain period of time and at a certain rate depending on the testing being performed. There are two categories of TID testing: High Dose Rate (HDR) and Low Dose Rate (LDR) and each simulates a different situation. HDR simulates the “raining cats and dogs” example, such as a blast from a solar flare. LDR ensures that the device can survive drifting through cosmic space radiation, or a “light mist of rain all day”. Devices are then tested to ensure they still pass production testing post-irradiation.

Typically, the radiation dose rate is somewhere between 50 to 300 rad(Si)/s, though smaller rates are sometimes used depending on the intended application environment. Since these rays are charged particles, any that get stuck inside the crystal structure can alter device performance. This can cause any number of hard and soft failures, though this is rare and these kinds of failures are more common with Heavy Ion Radiation than it is with stray electrons darting through the cosmos.

Displacement Damage

Let’s return to our waterproof fabric analogy. Any waterproof fabric should be able to withstand a rainstorm or a quick dunk in a pool of water without damage or degradation, otherwise why call it waterproof! DD, or Displacement Damage (NOT Designated Driver), helps simulate this kind of scenario.

Neutrons and protons have a significant amount of mass compared to electrons, so they are more likely to cause damage to the actual atomic structure of the IC than to cause other performance degradation. Also, radiation in the form of these particles tends to cause more destructive damage than gamma rays, which just excite otherwise immobile charges into an excited state before they settle back down. A frequent failure mode seen from Neutron Irradiation is actual shifting of aluminum traces, creating voids and migration of atoms. It’s also possible for neutron radiation to actually damage the silicon crystalline structure, though this is rare.

It is possible to combine TID and Displacement Damage testing into one test by using protons, which have significant mass and a charge associated with them. Though on some occasions it’s more favorable to do these separately so it’s possible to determine which effect caused the failure in initial studies.

That’s not to say it’s impossible to determine which effect caused the failure. One method is to use different biasing conditions to separate TID effects from DD effects.

We’ve covered the basics of radiation testing. Next time, we’ll dive into detail on Heavy Ion radiation and different transient effects it can cause. What do you think Juno will find out about Jupiter?

Radiation Testing Documentation

Analog Devices offers radiation reports for many devices in our space portfolio. A good example is the ADA4077-2S, a dual low offset and drift, high precision amplifier that was recently released. For examples of the HDR and LDR reports, check out the product page here.

And be sure to take a look at the full space products portfolio

7 comments on “Jupiter: The IC Danger Zone, Part 1

  1. dick_freebird
    August 10, 2016

    Talking about DLA qualification really has nothing much to do with parts for Jovian missions. DLA does not set a radiation hardness requirement for any part, but lets the manufacturer rate it (you can find QML-V parts with 10KRad or 1MRad HDR specs). The protons will kill bipolar parts and DLA has no care for protons really. Parts with neutron ratings may get “special attention” viz ITAR and hardly anyone rates those as a result (ditto high gamma dose rate tolerance). DLA also has little care for single event effects outside of latchup.

    I've never seen a part datasheet that says 20MRad (or ten, or for that matter anything above 1MRad and those are few).

    And of course at about one flight per decade, there's a sizzling market for you. Especially if your product is used one per system.

     

  2. Steve Taranovich
    August 11, 2016

    @dick_freebird, I thank you for your insights, but your last comment seems a bit bitter to me. 

    “And of course at about one flight per decade, there's a sizzling market for you. Especially if your product is used one per system.”


    Have you had bad experiences with space-qualified parts or are you against space exploration?

    Think about how much electronics are on board Juno to get it to fly over 500 million km to Jupiter, enter a precise orbit pattern in order to not burn up, then orbit that Planet Jupiter for more than 18 months while employing numerous electronic systems onboard to study and explore the planet……That's a lot of electronics!! I'll bet that ADI has lots more than just one IC on Juno, even if you just consider the redundancy of ICs needed on a mission like this.

    Plus, it's not always all about the money semi companies like ADI make on a rad hard IC. or selling NASA lots of space qualified, or in this case very Rad Hard ICs. It's about being a part of the adventure of exploration.

    In case you are concerned with the money we spend in space, NASA/JPL Caltech recently retired director, Charles Elachi, recently told me that NASA does not spend money in space, it spends money on Earth with the equipment and ICs it buys for its programs, as well as employing the people that NASA have on projects such as Juno, and the money it pays for contractors to help develop amazing technological engineering designs, ………. That's a great deal of investment right here on our planet.

  3. dick_freebird
    August 11, 2016

    Since you ask, I've been an IC designer doing product and technology development for rad hard parts all of my 30+ year professional career, and I've got a pretty good grasp of the environment in all its aspects. Especially the hard slog of introducing a new part with a narrow demand base and how often the market plan, not the part, fails.

  4. Steve Taranovich
    August 11, 2016

    @dick_freebird, I value audience members like yourself and your inouts with such a solid tech background. I get where you are coming from in your comments now—thanks

  5. kristen.villemez
    August 11, 2016

    @dick_freebird, These are all excellent points and I appreciate the great input. One of the points I was attempting to get across in this post is no matter how qualified or even over-qualified a device is for earth orbit, most ICs don't stand much of a chance once completely outside of the earth's atmosphere without extra protection. Certainly not in such close proximity to Jupiter. The standards of radiation testing that are outlined for QML-V devices does provide a great springboard for discussion with those with limited experience.

    I think that if a device doesn't pass some half-decent minimum that not many customers will buy the product anyway unless it offers something which nothing else on the market does and is valuable. Of course DLA doesn't care much about single event effects, but many customers do, if only to be aware that they exist. I do find SEFI effects to be the most intriguing of all the radiation effects, since solutions to the problems they create are not solely tied to process technology and IC design.

    I've seen a few times where the device release has missed the design-in window, so the loss can definitely go both ways. Though with more and more private companies such as SpaceX engaging in space exploration that narrow window is beginning to be less and less of an issue. It is certainly one of the more unique aspects of the space industry.

  6. Andy_I
    August 17, 2016

    > than gamma rays, which consists of just electrons.

    Gee, when I was in school, they taught us that Gamma rays were pure radiation (in the E-M spectrum) which would make them photons, and did not consist of electrons.

    OTOH, a flow of electrons from radioactive decay was what they called Beta “rays” (or Beta particles).

    Were they wrong, all this time?

  7. kristen.villemez
    August 19, 2016

    This was my mistake, thank you for pointing it out. I mixed up Beta and Gamma rays, but the implications remain the same. In both cases, there are extra charges present which affect device performance. The photons excite electrons to a higher energy state, making them more mobile (and leaving holes behind in the process).

Leave a Reply

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