3

I'm working on a small little project involving the Raspberry Pi 3 and the LSM9DS1 Breakout board from Adafruit. We plan to place the Pi and the sensor board in a high radiation environment (but protected by a lead box), and the kit will be contained within a vacuum. I was wondering what would be a good way to cool down the Pi? Due to the lead box, we're stuck using Ethernet, and we can't use a fan due to being in a vacuum. According to some other questions, the Pi's Ethernet controller can only go up to 70 degrees Celcius. Using pipes like a laptop CPU seems to be an obvious choice, but I'm not sure if this is necessarily the best way.

EDIT: Someone mentioned that it was a bad idea to place the Pi in a vacuum due to the capacitors. Capacitors in a vacuum have the potential of exploding. I can see that happening with electrolytic capacitors, but what about ceramic caps? Looking at the Pi 3, I don't see any electrolytic caps that could explode. Also, in this regard, wouldn't any computer in a vacuum be a bad idea?

Thinking about the question in context, I guess I'm not exactly familiar with the conditions of a vacuum. To me, a vacuum assumes little to no pressure with no air. So, if there's no air, the environment will get hot quickly as things heat up, but am I fundamentally wrong about what goes on in a vacuum? Also, would it help if the Pi were enclosed in a box or something so that it doesn't get subjected to the conditions of the vacuum out in the open?

3

Liquid cooling comes to mind as an alternative to heat pipes.

There is a project called The Wet Pi:

The Wet Pi is a fully functional water cooled Raspberry Pi, fitted with a micro pump, water block and reservoir.

It is attached to the BCM CPU/GPU, the Ethernet/USB controller, and the power regulator, as these are the components that generate the most heat. In this particular use case (vacuum) it is however necessary to consider vapour pressure of the tubing material depending on the required quality of the vacuum. (Assuming the coolant does not leak of course.)

Another approach could be to entirely rely on thermal conduction, i.e. find a way to contact these three chips to the metal housing. This will likely require a set of height adjusted spacers or one spacer milled to fit all three components of a thermal conductor such as copper or aluminum.

  • Thanks for the comment. Liquid cooling seems interesting, but given how the setup will be in a vacuum (as well as being bombarded with radiation), I'm not sure if we can use it within our design. We were considering using copper pipes to lay over the CPU and Ethernet/USB controller, just like how laptop CPUs are, and have them be connected to the case housing to release heat that way, but just relying on conduction, I'm not sure how cool the Pi will get. – user101402 Feb 28 '18 at 20:43
  • Although it may be besides the point (and perhaps you know it already anyway), those copper tubes in laptops, etc. contain fluid and a wicking material: en.wikipedia.org/wiki/… It works via an evaporation cycle. I would guess pound for pound they are actually more efficient than pumping water around in a loop, but the advantage of a pump is you can increase the effective scale (i.e. "more pounds"). – goldilocks Feb 28 '18 at 20:48
  • Yeah heatpipes contain a liquid and a vapour of the same substance, that is the idea. I would argue that is a little easier to DIY a liquid cooling system than a to DIY a heatpipe fitting to the Pi (unless of course there is a vendor for such a thing), but hey, there are a lot of posts on the web to DIY a heat pipe, intriguing. But then again I'd give the solid state solution a try first - heat conduction ftw. Just use the maximum cross section that is possible for the heat conducting elements and keep the thickness to the minimum that is necessary due to the connectors and stuff. – Ghanima Feb 28 '18 at 20:52
1

The RPi in a vacuum is not recommended. The capacitors are made under +/- sea level pressure.

I know that inside industrial 160-300kV X-ray tube heads a voltage multiplier is used. The capacitors in that circuit are strong closed containers. After reparing the tube head is connected to vacuum pump for sometimes 4 days to get a clean deep vacuum. After that the tube head is filled with an isolation gas up to 6 Bar. To put it simply: a capacitor in vacuum can explode. In high pressure it can implode.

Sorry for my writing... Regards: Jan

An other thing is the Rpi under a weather balloon. It easy to reach 20 kilometre attitude. What is the air pressure at that attitude? The casing used for Rpi in space ? Sorry no answers....

  • 1
    Thanks for the comment. I never considered that. If that's the case, then wouldn't any computer be at risk in a vacuum? If the computer was taken out of the vacuum, that's a different story, but if we can't do that, then what kind of computer should we look for? – user101402 Mar 7 '18 at 14:38
  • Also, based on what you said, does this apply to all capacitors? The RPI3 seems to mainly have ceramic capacitors, so I'm not sure how well it will handle in a vacuum. Then again, all I really know about a vacuum is that there's little to no pressure, so it'll get hot. I'm not familiar with how electronics behave in a vacuum, so I'm just trying to learn. – user101402 Mar 7 '18 at 15:52
  • But wait a minute, after doing some research, a vacuum usually implies a low pressure environment, right? So, the capacitors aren't under high pressure. Wouldn't they not explode then? – user101402 Mar 7 '18 at 16:10
  • @user101402 may I suggest to add this detail to your question? – Fabian Mar 9 '18 at 16:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.