What is the storage temperature range of the bcm2835?

[SentekdB]

Hi,

I have an application which needs the single board PC board, such as Raspberry Pi Zero, to be stored at as low as -60C. The operating temperature can be higher, such as -25C. So I would like to know the storage temperature range of the BCM2835, because the on board PAM2306 has storage temperature from -65 to 125C.

Thanks,

Bo

[jamesh]

I don't think we will have tested to that low a temperature. So I doubt figures would be available. I'm currently hoping to get some official labs figures at some point, will pots them when they turn up. I doubt they will go that low though.

I'd test it to see if one survives.

[Heater]

SentekdB,

-60C ? !!!

Do you happen to live in the Soviet Vostok Station in Antarctica ?

I mean, who else would want a guarantee of such a low storage temperature of anything?

Do you imagine that a super cheap, consumer device like the Pi comes with such a specification?

Even if the Pi survives that, which I'm pretty sure it would, I doubt the Pi Foundation would want to write such extremes into their specifications and guarantee it.

[davidcoton]

Consider the cost of purchasing a Pi equivalent which is guaranteed for storage at -60C.
Then consider the cost of hiring (and feeding) someone to keep your stored Pis warm.
Which is the cheaper option?

(BTW "Heater" might be a good person to keep Pis warm?)

[W. H. Heydt]

Considering that Pis have been operated in an ambient temperature below -40 (Antarctica, counting penguins), I wouldn't all all be surprised if it could survive -60C. You'd probably want that to be non-condensing, though.

[Heater]

davidcoton,

Consider the cost of purchasing a Pi equivalent which is guaranteed for storage at -60C.
Then consider the cost of hiring (and feeding) someone to keep your stored Pis warm....(BTW "Heater" might be a good person to keep Pis warm?)

Yes, if anyone has Pi in dangerously cold conditions they should consider sending them to me for warm storage.

No food required.

@W. H. Heydt

I'm sure you can freeze the nuts of the PI and many other consumer devices without failure. I don't expect the Pi Foundation to specify such extremes and guarantee them.

[tpylkko]

BTW. don't these high altitude balloon systems operate in like -50 °C... of course that still isn't long term storage at -60....

[broe23]

I don't think we will have tested to that low a temperature. So I doubt figures would be available. I'm currently hoping to get some official labs figures at some point, will pots them when they turn up. I doubt they will go that low though.

I'd test it to see if one survives.

I say send some to Oymyakon, Russia. Their winters run around -50 at most times. The colder the unit is, the dryer the air, the unit could be hammered with a lot of data and should be able to perform faster than expected, since the internals are not heating up. You just do not want too much moisture in the air, that the heat generated by the board could cause some form of corrosion on the pads.

[broe23]

Considering that Pis have been operated in an ambient temperature below -40 (Antarctica, counting penguins), I wouldn't all all be surprised if it could survive -60C. You'd probably want that to be non-condensing, though.

The great thing about Antarctica is that it is considered a Desert, since it does not get much moisture.

[jamesh]

I suspect that the requirement here is something to do with freezers, rather than ambient temperature. Or how about being part of the LHC, which gets really cold.

[hippy]

Deleted.

[Heater]

hippy,

Why not?

Because consumer devices, especially cheap ones, do not normally come with such extreme storage or operating temperature specifications. If they have any such spec at all. Do you know of any?

Because writing that into the spec. would require extra expensive testing during development and during production. It will require all the components to spec'ed to those extremes, which might be more expensive.

I looked through a random selection of chip datasheets and most of their storage ratings were down to around -60C. It therefore seems reasonable to suspect that might equally apply to most chips on the Pi.

If one needs such a specification then "reasonable to suspect" is not good enough. All the components would have to be qualified to meet the specification. Down to the last capacitor and resistor.

That is not the end of the story. It's not just about the components. What about the PCB and it's assembly? Things crack and break under thermal shock, differential expansion and contraction and all that.

Ceramic capacitors, for example, are notorious for cracking when stressed. Such stress can occur even without such extreme temperature changes.

Boards have to be designed and qualified to handle these things.

And he was simply only asking.

And I was only replying.

I wish SentekdB would come back and say exactly what his application is and why it needs such an extreme storage temperature. And how rigorously or important it might be to have such a spec. Then we could talk about it with more focus.

Having said all that, I am no way implying that because it's cheap the Pi is lacking in quality. I'm not saying the Pi won't work very well in such extremes. Only that it is not reasonable to expect a consumer device to come with specifications more typical of automotive or military products.

I understand the distributors offer custom Pi services. Might be an idea to approach them and ask if they can provide Pi with such specs and guarantees.