CubeSat For Under $1000?

Want to build your own CubeSat but have been put off by the price? There may be a solution in the works — [RG Sat] has challenged himself to design and build one for less than $1,000. (Video, embedded below.)

He begins by doing a survey of available low-cost options in the first video, and finds there isn’t a complete package for less than $10,000. By the time you added all necessary “options”, the final tally would probably be well over $20,000.

His idea isn’t just a pipe dream, either. In the the fifteen months since he began the project, [RG Sat] has designed and built the avionics and electrical power system circuit boards, and is currently testing his sun tracker design. Software is written in Rust, just because he wants to learn something new. You can check out the hardware and software design files on the project’s GitHub repositories, if you are inclined to build one yourself.

[RG Sat] lays out a compelling case, but we wonder if there’s a major gotcha lurking in the dark somewhere. In fact, [RG Sat] himself asks the question, “where do these high costs come from?” Our first instinct is to point the finger at qualifying parts for space and/or testing. But if you don’t care about satellite longevity or failure rates, then maybe [RG Sat] is onto something here.

Stepping back and looking at the big picture, however, the price of a CubeSat can be a drop in the bucket when compared to the launch costs, unless you’ve got a free ride. Is hardware the best place to focus cost reduction efforts?  Regardless, [RG Sat]’s project is bound to provide interesting and useful results whether he succeeds in his goal or confirms that indeed you need $10,000 to build a CubeSat. We’ll be following his progress with interest.

We’ve written about open source CubeSats before, and also a port-mortem analysis of a failed mission that contains some good lessons. Thanks to [Jeremy Grosser] for the tip.


19 thoughts on “CubeSat For Under $1000?

  1. The $50 sat project back in 2013 was successful (although the parts were more like $250):, launching on the same rocket as our first FUNcube (AO-73). Unfortunately I have no indication of the launch fee charged by GAUSS for integrating their PocketQube, I can say that the cost for a 1U CubeSat 10 years ago was of magnitude 1e5, but that is reducing significantly now that we have competition in the market and reusable launch vehicles.

    1. I thought I remembered reading about people making cube sats for far less than $1,000 already. Thanks!

      When I saw the headline I thought he couldn’t be talking about just building the satellite at such a price because that’s old news. But it couldn’t be about launching one for that price because.. reality.

      Beyond doing it to say you can is there much point to making such a cheap satellite?

      When you are spending at least tens of thousands of dollars on a launch what is it to spend mere single-digit thousands more on getting hardware that is made to withstand the difficult environment up there.

  2. Do you want it to work when it gets deployed? Space is a hostile environment. Testing. Thermal, shock/vibration, vacuum. Launch is a traumatic event. 20g vibration, 2000g shock. Temperature +/- 100C if you have decent thermal design. Radiation environment. Tolerance to single event upset. This kind of testing mostly isn’t something you can do at home. That’s where a lot of your costs come from. That’s why people use a qualified platform rather than building their own from scratch. You can throw a pi in a box for cheap. Ensuring that it works on delivery and for some period after that is a totally different world. And, that doesn’t count the actual launch costs.

    1. I came to the comments looking for the one person who nailed the answer. First, the rocket ride is incredible. If you can build your satellite and have it sustain being chucked against a wall, super heating and rapid chilling right away–and on top of all that data fortification from all the damn radiation that is constantly eating the silicon away….then maybe you have something here. I’m looking at this and yea, it’s cute. This won’t survive space, and i’d be surprised if it makes the initial launch. Rocket’s don’t handle like boats. Those off-the-shelf solar cells will likely crack. They need to be thicker. There doesn’t appear to be any kind of thermal considerations. So this thing will likely be in a constant state of solder reflow and rapid chilling. Oh, and yea the vacuum. If you don’t have perfect, defect free silicon–forget it. The die could crack, or the epoxy package could fracture exposing the die to space. The reason this is such an expensive field is because all the groundwork that came before this moment. Lots of folks figured out you can’t send breadboard level components into space and have them last. They tried! It cost unbelievable amounts of cash for a ride up, just to fail within days or weeks. My tip, is to build your satellite to the same spec’s as the rocket. Space is hard.

      1. And furthermore space is unique in that you can’t test in the environment you’re designing for. As you’ve said, this is where the experience comes in, and why components that are space rated are very much not like the ones you can buy on banggood or aliexpress.

      2. Thermal management is definitely a concern with any reasonable satellite build. Many CubeSats use large milled aluminum enclosures for some of their PCBs to act as both a minor rad shield and to smooth out the thermal cycle.

        That said, it is a bit of a stretch to argue it will be pushing reflow temperatures. The total amount of surface area exposed to the heating effects of the sun (and the earth) is quite small for a CubeSat and so long as you keep a reasonable emissivity for your external surface materials and have some decent thermal mass you are likely to be fine for a reasonable mission lifetime with consideration to the space hardiness of your electronics/batteries.

  3. If the launch costs are high, then you’d want the satellite to be reliable and long-lived, even if there are significant costs for that. Also, in space you can’t afford multiple iterations to get things right, so there’s a big benefit in buying a proven solution from a supplier.

  4. OSCAR 1 in 1961, didn’t cost much, but it was basically a transmitter sending two letters in morse code, and the speed varying as a form of telemetry.

    But it was a first, it didn’t have to serve another purpose.I

    The goal now isn’t to build a satellite, but to actually do something. This “$1000 satellite” seems more like an end in itself, and ultimately a waste of launch resource

  5. Do not forget that when you pull a vacuum the materials the payload are made from will start to out-gas. Not only does this cause your payload issues but could contaminate any co-launched payloads. That is why flight qualified parts are carefully chosen for the materials used.

  6. Aside from environmental concerns (temp, vibe, shock, radiation) which for a short lived satellite are probably doable “by cookbook”, there’s an enormous regulatory infrastructure that one has to deal with. You need to either have the skill set to do them yourself, or you need to *pay* someone to do it, and that’s way more than $1000 (which is about 1 day’s fully burdened pay for someone who knows what to do).

    Here you go:
    The FAA/FCC are going to want an Orbital Debris Analysis Report (ODAR).
    The range (through your launch provider) is going to want a Missile System Prelaunch Safety Package (MSPSP) which gives them some assurance that you’re not going to spontaneously combust or explode on the pad.
    You’ll need a radio license from the FCC (if in US), and it is *unlikely* that you will be able to use an amateur license for this – so you’re looking at Part 5 experimental.
    Your launch provider will want some form of material usage list – to make sure that your spacecraft won’t burp out things that hurt the other things on the launch.

    Then we come to “proof of environmental tests and bake out”

    If you’re getting a free ride from NASA (e.g. CubeSat Launch Initiative or similar) – you’ll need a bunch of paperwork to *prove* that you’ll meet the environmental requirements – yeah, a cookbook design should meet it, but doesn’t prove it. You could probably build your own vibe test stand and TVAC in your garage. Probably not for $1000, but if you scrounge and use surplus it would probably be “less than a new car”. You’d need to have some awesome convincing and presentation skills to get NASA to buy off on that, though.
    If you’re paying the “several hundred K” it costs to launch a cubesat – the paperwork burden is somewhat less – the launch provider will help with a lot of it. But they still might want to know if your cubesat will turn into rubble from the launch loads.

    Bear in mind that the launch provider has a dispenser into which your cubesat will go. And those dispensers aren’t cheap – so you’re going to have a hard time finding a paid launch under, say, $50k.

    It’s been a long time since Mir offered “chuck it out the door” small sat deployments at very low cost (<$10k, as I recall)

  7. +1 on all the above. We built stuff to withstand 10g loads,, deep space vacuum, lightning strikes, salt water, and +400f on one side,-400 on the other. The joke used to be that Radioshack bought silicon that failed the milspec but were good enough for earth applications. Good luck.

  8. Components for under 1000 $ doesnt sound that unrealistic these days, but cost of getting cubesat tested, space rated and getting it actually to space is still a huge beast in comparison…

  9. Don’t know if I want to skimp on something that costs at least 10k to launch into space.. But it’s a cool thought experiment.

    You could build a cubesat for a couple of dollars if all you just make it blink a LED.

  10. The testing for launch is definitely a real cost. Shake and Vibe table time was about $2k, each time, and we got a really good deal. The solar cells would not last through that ordeal without epoxy coating anyway.

    Thermal Vac testing can be done relatively cheaply, but that doesn’t include day and night illumination heating. Pretty much
    need to attempt to calculate that and hope.

    The satellite bus itself can be done for <$1k, By that I mean the physical framework, separation springs, deployment switches, card mounts, and wall panels. I think the video says Pumkin has a kit for 1200. With commercial markup I bet that is pretty bare.

    Wall panels will want to be customized for solar, radio, or sensors. The internal cards will need to be made to fit the mounts. We used something semi standard, similar to PC104. (Go KiCad / OSH Park!)

    getting S-Level is much more expensive. These have had considerable testing and tough time from the manufacturer.
    Rad hard is even worse, with different manufacturing processes. An example is the AtMega128 vs AtMegaS128. something like $3.80 for the former, $2700.00 for the latter, if you could actually get one (political problems)

    Daughter boards are useful, Dropping in that raspberry pi zero could work, I used a radio board. but you will want to
    layout and have fabricated the Arduino equivalent computer to fit the Cubesat board format. Mine included the battery,
    power, actuator drivers, processor, storage, deployment switches, radio board socket, connectors,,…. basically everything.

    Engineering time is everything. I had some prototype parts for the frame built for about $500. A machine shop built the
    final products for a very small fraction of that (each). Less than $1k in parts is completely reasonable. Laying out, testing,
    assembly, paperwork, etc was several orders of magnitude more.
    The launch was about 250k. (3U)
    The bird about 500k. (I didn't do the main payload that sat inside my sat)
    The project $1.5M ( lawyers, fundraising, concept, vigorish :)

    1. I don’t know details, but apparently AMSAT will supply radio equipment so long as it can be used for ham radio purposes after the main use is over.

      So if a university is doing an experiment, the radio is there for telemetry (and control?), but once tye exoeriment is over, the gets some ham purpose.

      Obviously lots of detail for someone actually at that stage.

  11. For cubesat no radiation hardening really necessary. Vibration – normal PCB-s, extra support against resonances is not needed, screws locked with thread lock are enough. Launch switches, antenna burn wire deploy, batteries and solar panels (non-gassing viscoelastic tape) are easy. Just ballpark link budget and use freezer, owen and impact drill for tests and it is a go. Considering, what is up there already and working at at few hundred mW, it is quite easy to get something working on LEO, as long someone pays the launch bill. Seriously, it is way easier than one might think.

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.