16

The Arduino Uno Page doesn't say the temperatures that it can operate in. I'm thinking about placing it outdoors. How can I make sure my Arduino Uno is safe in weather that can reach -20 to 105 degrees Fahrenheit? (-26 to 40 degrees Celsius).

Peter Mortensen
  • 435
  • 3
  • 12
Anonymous Penguin
  • 6,365
  • 10
  • 34
  • 62

5 Answers5

15

It's not that big of deal. The ATmega 328p datasheet states this:

Temperature range: -40 to 85 degrees celsius.

The same goes for the USB chip on the Uno (ATmega 16u2 for UNO R3).

That's inside your limits. It probably could go a bit colder than mentioned, but it'll shorten the length of the board a little bit.

However, there are some things that may go wrong:

  • EEPROM might not be stored as long in extreme temperatures. Keep this in mind if you're storing critical data.
  • The voltage regulator may not perform as well in hot conditions
  • The crystal oscillator may not produce exact values. However, I would imagine that a few hertz more or less wouldn't affect a 16 MHz processor. The tolerance is actually a little less than 1%. You may have some issues with serial (baud rate not correct). I would look into any communications like I2C also. (I don't know exactly how the clock line works... it might be fine for I2C.)
  • Resistors/capacitors may not produce the exact values. I'd imagine that the tolerance won't be more than 8% on resistors: most resistors are rated at 5% for normal temperatures. It does depend on the manufacturer. Capacitors do have a bigger tolerance, but their main purpose is to "smooth" a signal.
  • Extreme cooling/warming may cause minor expansion problems. (Note: it's fine once in a while, but not on an hourly basis of dropping 30 degrees.)
  • Other components (LCDs, etc.) should be kept in mind also when looking at the viability of keeping it outdoors.

So as long as all the other components not on the board will happily run in the temperatures that you need, you should be fine. Also, as with all engineering, values often have added "padding" built in. (i.e. 5% tolerance is often 3-4%, maximum 12V it can run on 12.5V, etc.)*

*What I mean by that is your Arduino won't explode when it's -41 degrees C. It's not great for it, but most likely you should be fine as long is it isn't a regular occurrence.

Anonymous Penguin
  • 6,365
  • 10
  • 34
  • 62
4

Like everyone mentions, as long as you're in the shade, hot temperature probably doesn't matter too much as it's within the limits of the components.

I'm more worried about condensation in the mornings. Water vapors will condense on electronics just like it does on grass. You could try electrical epoxy to cover the circuit. The Arduino doesn't run very hot, so the epoxy doesn't do much in terms of preventing it from cooling down. But epoxy does keep water vapor condensation from being a problem.

user851
  • 111
  • 3
1

I agree with the advice to read the data sheets, but here is a personal answer to the question.

I installed a Raspberry Pi in an enclosure in which the existing Arduino survived last summer.

Even though they have the same overall temperature limits (except the communications section) it was the Pi that stopped working first.

The good news is that bringing it inside it booted right back up.

The temp/humidity went up to 140 F (like a hot car in Phoenix).

So in the end, the data sheet was right in terms of survival. But I would suggest a more conservative approach, such as placing them into a white-painted enclosure to minimize the effects of the sun.

After removing the pi, the Arduino woke up like nothing happened, still outside.

These things are very resilient.

SDsolar
  • 1,175
  • 3
  • 11
  • 35
1

If you must place your device outside, I would suggest a diecast box. The main heat generating component is probably the regulator (Think about it - at 12V, the reg drops 7V, where the micro runs on 5V or maybe 3V3). So running the lowest voltage into the arduino is recommended, I think its good to 7V (for a 5V unit). If you can connect a heat pathway from the chip surface to the case, good (Use heavy guage - at least 2mm thick). Take care you do not connect to the reg tab - use mica or thin mylar and heatsink paste (Avoid galvanic interations). Normal finned heatsinks on the outside of the box actually do the job of shedding heat to the atmosphere. This whole thing should be in a slatted wooden white painted container (A stevenson screen) so direct sun (and rain/dew) dont hit the containing box. This would be a solution for extreme environments. Remember that any heat build from the board must get to the box internal environment - using trapped air you get a very poor thermal contact. Then it has to get through the box and into the air again. Dont forget that current sinks you use on the chip generate some (small) heat in the process.

1

For hot temperature, just follow the datasheet.

For low temperature, I remember someone last year trying to overclock a UNO with liquid nitrogen so I guess you won't ever meet any problems with low temps :-)

In his blog, that person shows that he could run his UNO at 65 MHz by reducing temperature to -196°C.

Of course, the process was more complex than just reducing the temperature and checking what happens: a lot of improvements were performed on the board.

The blog explains very well how various components can react to cryogenic temperatures; the major issues seemed to be capacitors which capacitance dramatically reduces under low temperatures.

jfpoilpret
  • 9,162
  • 7
  • 38
  • 54