Why do microcontrollers fail in a Vacuum?

[Tech2025]

I was trying to log data inside a vacuum chamber, tried an arduino and a rasperry pi. When reaching -30inHg the devices just stopped working, they powered off, no data in or out and would not work ever again. Why? And no, they did not overheat.

 

[berkeman]

I was trying to log data inside a vacuum chamber, tried an arduino and a rasperry pi. When reaching -30inHg the devices just stopped working, they powered off, no data in or out and would not work ever again. Why? And no, they did not overheat.

When you pulled them out to analyze the failure, what did you find? Are there any electrolytic caps on the board?

 

[Tech2025]

When you pulled them out to analyze the failure, what did you find? Are there any electrolytic caps on the board?

Yes, but those seem fine, were not bloated and no signs of leakage.

 

[jedishrfu]

From what I read, if there's no air to dissipate the heat then the unit will overheat and fail.

I was thinking that maybe you capacitors would fail too. Someone mentioned that you should replace then with tantalium ones:

https://arduino.stackexchange.com/questions/48937/how-well-will-the-arduino-uno-survive-in-a-vacuum

 

[berkeman]

Yes, but those seem fine, were not bloated and no signs of leakage.

It sounded like you were saying the power failed. What do you measure when you try to power it up?

 

[Tech2025]

It sounded like you were saying the power failed. What do you measure when you try to power it up?

I haven't tried measure that yet, but the power supply is not the issue

 

[berkeman]

Also, can you post the schematic for anything that was in the vacuum?

 

[berkeman]

but the power supply is not the issue

the devices just stopped working, they powered off,

What do you mean by that?

 

[Tech2025]

Also, can you post the schematic for anything that was in the vacuum?

Arduino - https://www.arduino.cc/en/uploads/Main/arduino-uno-schematic.pdf

rPi2 - https://www.raspberrypi.org/app/uploads/2012/04/Raspberry-Pi-Schematics-R1.0.pdf

 

[Tech2025]

What do you mean by that?

Well exactly what I mean, at -30inHg they just powered off and would not reboot

 

[berkeman]

No, I was asking about the "powered off" part. How can the board power itself off and there not be a power supply problem? You said it still doesn't work out of the chamber, so presumably it is still not powering up, no?

 

[berkeman]

rPi2 - https://www.raspberrypi.org/app/uploads/2012/04/Raspberry-Pi-Schematics-R1.0.pdf

Is the 1 amp input fuse F3 still intact?

 

[Vanadium 50]

And no, they did not overheat.

How do you know?

 

[cjl]

I would guess that some component overheated. Berekeman's fuse suggestion seems like a good thought as well, since the fuse could easily blow far below rated with no cooling.

 

[davenn]

And no, they did not overheat.

How do you know?

I would guess that some component overheated. Berekeman's fuse suggestion seems like a good thought as well, since the fuse could easily blow far below rated with no cooling.

exactly ... with no air to aid in the convection/conduction of heat away ... components could easily overheat and fail

 

[jedishrfu]

Don't let your preconceptions or intuitions rule your view. The unit failed. Software is immune to a vacuum (unless the programmer forgot to write it but that's another issue - software joke :-) ) but hardware is not. Hardware will fail in extreme conditions of high pressure or low pressure, of high temperature or low temperature. In your case, the extreme condition is no pressure, no air and no means to conduct the heat away and so it failed.

In high pressure situations, like deep ocean sensing equipment, you can get a loss of power from your batteries as they get compressed under the weight of the water and from the coldness of the environment.

 

[synch]

As well as the overall heating, some ICs have defined heating rate limits. Ordinary humidity can lead to moisture absorbed in
minute gaps near the pins, heating too fast can lead to failure as the vapor exerts pressure before it can exit. It may seem odd but it is routine for ICs to be packaged in controlled-humidity packages and the soldering /heating ramp rates to have critical limits defined by the manufacturers.

 

[CWatters]

I would power up the Pi and check the power rails. Might be the regulators are the first to overheat?

 

[CWatters]

Discussion here..

https://www.raspberrypi.org/forums/viewtopic.php?t=102367

You cannot count on radiative cooling of electronics in a hard vacuum.

At 85C the RPi's SoC radiates just 145mW, and gets 65mW back from the walls of the vacuum chamber (at 20C). Compare that with 300-600mW power consumption. In order to shed 600mW the chip must reach 250C!

A conductive path is needed to cool the chip. In the case of the RPi conduction to the PCB may be enough, but I wouldn't count on it.
.

If it is overheating perhaps you can modify your software so the Pi only needs to be powered up briefly when a sample is required and powered down between samples?