Lower operating temperature specs for single-board computers

I'm curious... what's the difference between your run-of-the-mill single-board computer (e.g., those based on Atoms or ARMs or whatever that people are so fond of running Linux on) when they're spec'd to operate down to -20C (or even -40C) vs. the far-more-common 0C? It's obvious that you need to add additional cooling so that nothing fries as you want a higher and higher ambient temperature operating range, but what happens at -20C that isn't already happening at 0C?

Thanks,

---Joel

Reply to
Joel Koltner
Loading thread data ...

Freezing, mostly. For example, depending on the construction electrolytics reach their limits.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

Ah, thanks.

Most that operate down to 0C will let you store them down to -20C, interestingly enough.

---Joel

Reply to
Joel Koltner

I am not familiar with those particular boards; however I tried what happens with other MCU boards at low temperatures. The #1 likely problem is getting stuck in reset state because supervisory circuits and voltage references are running away. ESR of the elecrolytics increases dramatically. Flash memory reliability and endurance is much worse at low temperatures.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky

At least you should check that the oscillators start up reliably at low temperatures.

When you get the oscillators running, it may need some failed restart attempts, until the heat produced by the logic gates running at the clock frequency has heated sufficiently some marginal components.

However, if the main clock does not start reliably, you are out of luck.

With non-coated PCB's you may end up with various condensation problems.

Reply to
Paul Keinanen

single-board

Crystals can have a hard time, just like engines do. Russian truckers sometimes solve it but gathering some wood, making a little fire underneath and having a cigarette in the meantime.

May need a lower value "kicker resistor", or a real kicker circuit.

That problem can also hit well above 0C. There were days in Puerto Rico where it was nice and warm but condensation was so bad that water was running down the windows and making puddles on the ground.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

And after you get things so that your oscillators start, and your power supplies regulate correctly, and the socketed parts don't pull out (or off) of their designated spot, all of the digital electronics will go faster -- by different amounts.

So latent race conditions that weren't a problem at all at 0C will suddenly _be_ a problem.

(Note: this is much more likely to be an issue at high temperatures when CMOS gets slow, but hey -- it can happen at cold, too).

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
Reply to
Tim Wescott

In medieval Europe the digital designer would have been dunked into a moat three times, in front of all the villagers :-)

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

single-board

Now that's an OCXO...

Reply to
Spehro Pefhany

And if he died then he wasn't a witch, and if he lived then he was a witch and needed killing?

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
Reply to
Tim Wescott

No, this was meant as an "ethics enhancement" process. They used it on bakers who made undersized rolls but sold them at full price. They usually tried that sort of business practice only once ...

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

Witch hunts were a Renaissance phenomenon, not a medieval one.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net
Reply to
Phil Hobbs

Or they do not shut down the engine during the winter.

During the 1939-40 Winter war (at temperatures down to -40 C) when the Finnish planes returned from the last mission for the day in the evening, the oil (and other fluids) were taking out immediately after landing. In the early hours in the morning, the liquids were heated with open fire and the liquids were poured into the plane, before the first mission in the morning.

Regarding electronics, if possible, keep the system powered up all the time, even some power save mode may keep the electronics warm enough.

Fluorescent lights have a hard time starting at cold temperatures and if started, the light output is very low at low temperatures. Keeping the light on constantly during the winter and mounting it into a closed enclosure to keep the operating temperature sufficiently high.

Tropicalisation is needed in those conditions.

If possible, run the system constantly with sufficient dissipation and it will keep the condensation away. Use some enclosure to keep the water dropping from above away from the electronics.

Reply to
Paul Keinanen

That was even done in the northern parts of the US. When dad came home the drill was to let the fluid out into a large pot and that pot was put in the back of the banked stove. AFAIK you can still buy the little lever valves, like a faucet so you could do this job without getting your hands dirty.

Yup. The little CFL in the "dog poop area" needs several minutes to go from a dim reddish glow to useful light output. Must be turned on at least five minutes ahead of time. So far for the much-touted savings ...

But you can't rely on that. There can always be power outages and it must survive those.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

Wouldn't a block heater have worked just as well and have been far less hassle?

Indeed... although some brands/designs seem noticeably better than others at this. I remember from reading Jim Williams' notes that it's largely a trade off with bulb life -- you can certainly use a feedback loop and pump enough power into the tube to have it at full-brightness immediately even at low temperatures (assuming your have enough margin in the power supply), but it seriously degrades the lifespan.

I'm just happy that "instant on" CFLs are now ubiquitous compared to the early versions that flickered for a second or two at turn-on -- the WAF there was pretty low.

---Joel

Reply to
Joel Koltner

Sure but in those days I doubt they existed. Plus many remote dwellings still had oil lamps. Or if you were really lucky one circuit. My student quarters were still sort of like that. One circuit 230V/6A for a whole apartment. My roommate and I had to make sure only one TV was turned on, and not while cooking on the two burner electric range. Of which only one burner could be used, the 2nd for maybe 2-3 minutes (if you turned the lights off). Else it was a trip down three flights of stairs. If you had a spare fuse, that is ...

Engineers could do that decades ago, as evidenced by the fluorescents in our garage. Instantly on, at well over 50% brightness even when it's almost freezing in there. Same tubes since 13 years, judging from the cob webs after buying the house I'd say they are much older. Yet they get turned on and off several times per day.

I guess the CFL guys must have cheaped out on some parts of their designs.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.