32C3: So You Want To Build A Satellite?

[INCO] gave this extremely informative talk on building a CubeSat. CubeSats are small satellites that piggyback on the launches of larger satellites, and although getting a 10 cm3 brick into orbit is cheap, making it functional takes an amazing attention to detail and redundant design.

[INCO] somehow talks through the entire hour-long presentation at a tremendous speed, all the while remaining intelligible. At the end of the talk, you’ve got a good appreciation for the myriad pitfalls that go along with designing a satellite, and a lot of this material is relevant, although often in a simpler form, for high altitude balloon experiments.

satellite_2-shot0002CubeSats must be powered down during launch, with no radio emissions or anything else that might interfere with the rocket that’s carrying them. The satellites are then packed into a box with a spring, and you never see or hear from them again until the hatch is opened and they’re pushed out into space.

[INCO] said that 50% of CubeSats fail on deployment, and to avoid being one of the statistics, you need to thoroughly test your deployment mechanisms. Test after shaking, being heated and cooled, subject to low battery levels, and in a vacuum. Communication with the satellite is of course crucial, and [INCO] suggests sending out a beacon shortly after launch to help you locate the satellite at all.

satellite_2-shot0003Because your satellite is floating out in space, even tiny little forces can throw it off course. Examples include radiation pressure from the sun, and anything magnetic in your satellite that will create a torque with respect to the Earth’s magnetic field. And of course, the deployment itself may leave your satellite tumbling slightly, so you’re going to need to control your satellite’s attitude.

Power is of course crucial, and in space that means solar cells. Managing solar cells, charging lithium batteries, and smoothing out the power cycles as the satellite enters the earth’s shadow or tumbles around out of control in space. Frequent charging and discharging of the battery is tough on it, so you’ll want to keep your charge/discharge cycles under 20% of the battery’s nominal capacity.

mpv-shot0001In outer space, your satellite will be bombarded by heavy ions that can short-circuit the transistors inside any IC. Sometimes, these transistors get stuck shorted, and the only way to fix the latch-up condition is to kill power for a little bit. For that reason, you’ll want to include latch-up detectors in the power supply to reset the satellite automatically when this happens. But this means that your code needs to expect occasional unscheduled resets, which in turn means that you need to think about how to save state and re-synchronize your timing, etc.

In short, there are a ridiculous amount of details that you have to attend to and think through before building your own CubeSat. We’ve just scratched the surface of [INCO]’s advice, but if we had to put the talk in a Tweet, we’d write “test everything, and have a plan B whenever possible”. This is, after all, rocket science.

46 thoughts on “32C3: So You Want To Build A Satellite?

      1. Works for me using Firefox. Make sure you have scripting fully enabled, as this is one of those very badly designed sites that have many layers of nested script-heavy sites that you need to peel like an onion before it actually works. I think I had to allow like eight nested sites before the video would stream.

      1. If I had the money for the launch, I would absolutely make my own cubesat. I wouldn’t even have any use for one, It could be blasting out Rick Astley’s Never gonna give you up, for all I care.

    1. Holy hell. It’s fartface! Oh, man, I thought you left us! You just changed your name, I see. Wow. Long time no see.

      I was just reading this comment and wondering, ‘surely this guy posts on hackaday under different pseudonyms. This type of trolling is just too advanced for your usual random internet commenter.’ It turns out, if you look at the IPs and email domains, Yep, this guy is FartFace.

      Wow. Amazing. I thought we lost you. It’s like meeting a long-lost friend. Spectacular.

      1. Still don’t get it, huh. My real name also isn’t fartface, good luck asking my ISP for it.
        Also no cubesat has 10cm^3 volume, they are usually 1000cm^3 in volume, which is only a difference of 2 orders in magnitude

        1. OK you’re right. The cubesats are 1 liter volume. That’s 1000 cm³ but I still don’t get why you used the parentheticals.Also Brian and Elliot have access to the HaD server logs. They compared FF’s IP address to yours and I think Brian is saying it is an exact match. He included email domains too. The coincidence is too weird.

          1. The brackets indicate ten centimetres cubed, ie a cube with sides of ten centimetres and volume 1000 cubic centimetres, as opposed to the ten cubic centimetres stated in the article. The article should be corrected, and ridicule of the person pointing this out is uncalled for IMO, fart face or not.

          2. steves – I’m sorry but I don’t remember ridiculing JustMe. Not sure Brian’s intent was either. I am just asking questions as I’m learning stuff here at HaD as I am a autodidact. If your saying brackets or parenthesis mean to cube the contents then I’ll have to take your word as it’s not intuitive and may be a SI nuance I was not familiar with. I was trying not to be dogmatic or absolute with my comments. I tend not to insert IMO a lot – maybe I should do it more.However, we still have not addressed the server log coincidence unless you’re accusing Brian of mendacity. You’re not doing that right?

            Yes I agree EW should correct the article to reflect 1 liter for cubesat volume.

          3. sonofthunderboanerges – Your guess about the parentheses is correct, everything inside the parentheses is cubed. That is not an SI convention, but just a regular convention of mathematics. And steves is right, the article should be changed to either read (10cm)^3 or 1000 mL or 1 liter, regardless of who points out the error.

    2. Why express it as a formula? What’s wrong with the SI STANDARD 10 cm³? The cm is not a variable; it is a standard metric measurement called centimeters. IOW it’s not 10 times CM to the power of 3. It’s just 10 cubic centimeters. Some how I fell you already knew this… ʘʖʘ

    1. For myself not really an experience wothy of bold all csops. . While this program was a damn good overview, I have to believe many in his audience had already heard or read much of what he said.

  1. Elliot Williams – Excellent article as usual EW!

    I remember once ordering some dark gray weather balloons (WB) on EBAY (or Amazon) and purchased some Walmart Helium. I got some light-weight heavy-test fishing line from sporting goods dept. I went to a vacant cul-de-sac nowhere near an airport (but near an industrial park on a weekend). I filled the balloon with the He. Tied it off and sent it aloft tethered. Boy did I learn about how not to do this WB operation!

    It took off like a rocket… the fishing line tether held it in check just fine; HOWEVER the thin line tether was murder to try and control with the wind and the balloon’s rise-rate. It could have taken a payload of a small simplex repeater and FRS XCVR radio but I did not do that fearing loosing the entire radio rig.

    Suffice it to say it was not a complete fail but food for thought. It was almost impossible to reel back in so I would need to go back to drawing board on that. And even though it did not attract any noticeable attention I could easily see how it could from just about anywhere as it rose about 200-300 feet, and was angling from the wind. And since it was an overcast day dark gray was a poor color choice,

    I got it back down, deflated it, and stowed it in the SUV just fine. Imagine using this as an emergency communications system when in a remote area. The bottle was only $20 usd and could maybe inflate 1-2 large weather balloons. A motorized reel with some sort of nylon para-cord would probably be better. It needs a gondola for the electronics. The gondola also needs a parachute in case of catastrophic failure (someone shooting it down). Doing this on a high mountain peak would be cool.

    Not me…

    This is the poor-man’s CubeSat sans NASA :-)


    1. Actually, using a balloon for radio communications is a thing. Though usually they only hual up the antenna and keep the radio on the ground. Easier to replace a wire and a balloon than the whole radio. And also no need to worry about anything falling on you if the balloon fails.

      1. I thought of that but the coax cable has to be very thin to control weight payload (or the envelope needs to be larger). The thinner the coax the more lossey it is. I was thinking cheap expendable equipment like a Simplex Repeater (Arduino?) and a FRS/GMRS radio from Walmart? HAMs would want to use some sort of cheap QRP rig. I was thinking of a cheap parachute made from a thin cloth to cover falls. The whole thing would be tethered so it doesn’t have to be tracked by RDF equipment later. I wonder what the FAA thinks about free floating balloons.

        1. I found some very lightweight para-cord at Walmart that can handle 500 lbs (226 Kg). It’s only 50 foot (15 m) though. It is about $6 USD (4 UK). They have black and dark gray. You have to use good knotting techniques to make hundreds of feet of it. Some sort of plastic spool from an old spool of cable would be good. If you could put it on a frame with a dc motor with some sort of braking would be good.14.9 cubic feet of Walmart helium is $49 USD. 8 foot diameter weather latex balloons (white only) are available for $20 USD at American Science & Surplus website.

  2. I’m just gonna drop this one here, an old quote from none other than the JPL regarding software:

    “Test what you fly, fly what you test.”

    Unproven, untested software has no place on a device you will never lay hands on again, let alone be able to troubleshoot remotely if it goes into an unknown state and goes dark.

    Many moons ago I worked for a company (now absorbed by a much bigger company) that used to make weather instruments designed for remote installations. We had an “isolation room” for final testing that had two neet attributes:
    1. Once set in place for final testing, the room was locked and no one was allowed “hands on” inspection, repair, “whatthefuckisthecomLEDdoing” type inspection.
    2. Although on mains power, one evil bastard had access to things like coms noise, machine voltage (they were designed for solar power too) and could even force a reset outside normal coms. The programmer was expected to test all of the machine’s abilities in this environment, and was never given a “heads up” on the OTHER test script, which was to throw as much sh!t at the unit as possible to trip him up.

    It was always fun to institute a forced reset and hear a muffled scream from the other room…

  3. Much of the information is what most interested in the topic have read or hear previously. The program is good refresher course for those individual, and great to point those first learning and gaining an interest in amateur satellites. The elbow jab in the ribs to we US citizens about the metric system is not unearned, but I truly hope there is no need for him to teach US students how to use the metric system, the mat used to put metric units to work is the same math use to put our the measurement units we now use to work. In the use the idea that the metric system is confusing, has become like faith doctrine generation push onto subsequent generations.

    From my read of the history of OSCAR the builders weren’t reinventing the wheel with each communications satellite, IMO that’s an unfair characterization. Each satellite had it’s own using the components available to them at the time. Components that where going through an evolutionary process, so it shouldn’t be surprise that the amateur radio satellites where going through an evolutionary process. Amateur radio satellites design to be communications satellites are such isn’t a surprise, but from what I read many of the cubesats become amateur radio satellites because to be able to use amateur radio allocations during the communications satellites primary mission it to be able to default to a communications satellite when the primary mission is complete.

    My take on the cubesat standard st that isn’t driven by amateur satellites, but commercial satellite interests interests. IMO opinion amateur satellites builders better get on the ball to get more cubsats on the air using amateur radio allocations, because commercial cubesat interests will get amateur radio spectrum if it’s seen to be idle.. Terrestrial use of that spectrum could also keep it from being idle. Much less expensive that satellites, would be regular high altitude balloon launches that cross continents and oceans.

  4. Here’s another SQTB BRAINSTORM. It is a CubeSat based on the new NASA OSCD CubeSat only extremely parsed down to the poor man’s version using ideas from modulatedlight dot org. It’s in IMGUR and I can’t figure out how to make it private with a password. It’s being looked at by others and some have given up posts. They don’t even know what it is (duh!). So anyway NO FLAMES I already said someone BTDT so no need to say that. It’s just a innovation on existing VLC and CubeSat technology: http://imgur.com/gallery/zLDYVP6/new

    Constructive comments welcome. It’s called LITtE or “light” not “leetey”. :-)


Leave a 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.