Hi,
I want to ask you: do you someone has experience with the nonstop working arduino in normal indoor or outdoor conditions?. How durable is arduino in nonstop work?
(that means read 4 analog inputs, count somehow this data and send PWM and trigering LOW HIGH to digital outputs - 10).
Is there some risk of the unit "time-damaging" ?
How about long time working in warmer places ( 50°C)?
Some people spend there whole lives working on this sort of issue.
There are three problems with regard to long time reliable working:-
Disturbance through interference both radiated, emitted and electro static discharge (ESD).
Component failure.
Component ware out.
The Arduino is basically a learning platform and has not been ruggidised, the same goes for your inputs sensors. Better decoupling of supplies, and interference suppression on the inputs would be required for long term reliability. You would need to put ESD protection on all inputs and outputs. The nature of this probably means that you have to include them on the PCB for best results, so that means a new layout. Also a watchdog timer can allow you to recover from interferance that affects the processor's execution of the code.
Nothing much you can do about this. All components could suffer from sudden failure. The best protection is a burn in period of a few days as there is a peak of failures when parts are new.
Most capacitors ware out, that is they dry out inside and drop in value over time. Elevated temperature accelerates this and so the cooler you can get things to run the better. Having said that then 50C is no great temperature, most components are rated up to 70C. If you want higher then there is military spec devices but they cost a lot more. So 50C is a good ambient to aim at.
So in short while it is not bad it is far from bomb proof, it depends on the environment and what you ask it to do.
Most capacitors ware out, that is they dry out inside and drop in value over time. Elevated temperature accelerates this and so the cooler you can get things to run the better. Having said that then 50C is no great temperature, most components are rated up to 70C. If you want higher then there is military spec devices but they cost a lot more. So 50C is a good ambient to aim at.
This is true for electrolytics, but not the other types that I've seen in 30+ years of electronics. Also, the time for this to happen is usually measured in years and is a function (generally!) of the quality of the caps: I've got some equipment that uses electrolytics that is nearly 50 years old and runs fine, I've also had cheap knockoff gear fail after just a couple of years because of dried out low-priced caps. That old saying about "You get what you pay for" still seems to be true
Some people spend there whole lives working on this sort of issue.
There are three problems with regard to long time reliable working:-
... Also a watchdog timer can allow you to recover from interferance that affects the processor's execution of the code.
The Atmel has a watchdog timer, is there a way to access it (i.e. configure it and feed it) from within a sketch?