Go Down

Topic: About 2 years (Read 2192 times) previous topic - next topic


We're beggining a project which could be for 2 years or more.

What about the life time of an arduino in years ? Has something a feedback on this kind of projects ?



Well I've never done a project like this, but if used correctly I think an Arduino can last for 2 years.
I have Arduinos that are 2 years old and still work fine, but these weren't used continuously.

Jack Christensen

Should be no different than any other modern electronic device, assuming it is used within specifications. I would expect far more than two years. Since it has relatively few components, and therefore fewer parts to fail, I might even expect an Arduino to outlast common consumer electronics like phones, TVs, cameras, etc.


iv got one thats been going for 5 months now its still running fine. And at work we have a couple that have been constantly running for a year and a half. So as long as your not doing anything horrible to it, 2 years should be fine.


Solid state electronics don't "wear out".   Failures are very infrequent and basically random (assuming no electrical, mechanical, or thermal abuse).  On average, 2 years should be no problem.   But if the design is critical, such as life-support, or if it's going into space where it can't be repaired/replaced, you have to take extra precaustions, because failures are random and there is a slim possibility that you get a failure after 2 days, or 2 weeks... 

Take a look at this.    (That's just for the chip.    Of course, there are more parts on the Arduino board, and you might be adding some parts.  If you use 100 parts, life-expectancy (MTBF) is 1/100th.)

I do have an Arduino that "went bad" after a couple of weeks...   The chip is running, but the bootloader doesn't work.  (I don't know if the bootloader somehow got corrupted, or if the serial-chip "went bad" or somehow got burned-out.)

I built a few projects with a different microcontroller about 15 years ago, and they still work fine.   One is a car alarm that's been running 24/7 for more than 15 years... even when it's not armed.  The only time the microcontroler & software-loop haven't been running is when the battery was dead!   

Mechanical parts (switches, relays, motors) and connectors can wear-out and connections can become corroded.  Electrolytic capacitors can have lifetimes less than most other "electronics".   And of course, vacuum tubes age, deteriorate, and wear-out. ;)


Also, the flash data retention is quoted 100 years at room temp, or 20 years at ridiculously high temp, IIRC.  So, 2 years (assuming it's not in a pot of boiling water) should be cake.

For long-term, unattended operation, I would encourage the use of a watchdog timer to reset the chip if the code crashes or locks up for whatever reason.  This is built-in to the AVR, and there are examples in the playground.


Data point:

I have 14 '328 based boxes (mix of promini's and standalone designs) that have been powered up 3-4 times a week and run for 2-5 hours each time, have been doing that since Dec 2010. 1 promini failure - the box it was connected to via serial interface also died, no way to say which one caused the other to go. 
Designing & building electrical circuits for over 25 years.  Screw Shield for Mega/Due/Uno,  Bobuino with ATMega1284P, & other '328P & '1284P creations & offerings at  my website.


As pointed out electronics can suffer from a random failure at any time.
I did build a ham radio repeater controller that worked for 17 years 24/7 before they turned if off when they moved site.

The most likely to fail part is the power supply, cheap ones have capacitors that have a short life. Large electrolytic capacitors unlike most other bits of electronics do have a wear out mechanism and the cheaper ones will last about 2 years, so don't economies on the power supply.


We're beggining a project which could be for 2 years or more.

What about the life time of an arduino in years ? Has something a feedback on this kind of projects ?

You are going to find a shorter bathtub curve in components like power supplies and so on. It is usually passive components like capacitors that fail before semiconductors. Keep things cool, and over-design so things are well within spec.

(And there we have it. I should have read all the way to the bottom of the thread.)
I yield() for co-routines.


Great answers,

I'm confident, now.


a thing worth mensioning imo is that when u use timers beware of the overflow i've red that for millis() this happens once in about 50 days and for micros() it happens once in about 70 minutes

Go Up