What is the approximate working life of a Microcontroller ?

No, it's not a rhetorical question. I'm looking to use a PICAXE or
similar Micro controller in an application and I'm trying to find out
an approximate working life to see if the application will be
reliable. There won't be anything much going on in terms of running
its program. It will operate four solenoids, one at a time for
approximately 4 seconds each, in a sequence separated by about 30
seconds before operating the next. ie the basic program will simply be
output to a certain pin briefly, followed by a 30s pause and so on.

In between this basic cycle I'm looking to provide a brief flurry of
approximately 10 seconds of pulses to drive a device drawing
approximately 300 mamps. However the pulses will be approximately 40
per second and the device will be running continuously at these
sequences for about 12 hours a day.ie in any given day there will be
approximately 1500 of these pulse sequences and also about the same
number of operations of the solenoids.

Has anyone had any exposure to running these cheap Micro controllers
for continuous periods to know approximately how long they may last
and continue to be reliable ie able to continuously execute the
program without errors? They are designed for a hobby and educational
applications but are tthey able to be applied reliably in a more
practical sense?

 

As long as the power supplies are good then the things keep on running.
Critical applications may care to facilitate themselves of an auto-restart
circuit ('watchdog'), but general reliability is same as (or better) than
standard logic components.
The 'Pioneer' probe electronics are still at it after maybe 30 years of
unattended operation.
Heat will kill micros. transients will kill micros. Design 'em out and
forget about any 'working life' aspects
regards
john

 

Take a close look at the driving and operation of the solenoids. These will
be the weak points of the system and the ones that will fail if anything is
going to.

The electronics, based on a PIC should work for a lot longer than you will
be around. My own experience is that PICs have a life of greater than 18000
hours. This is beacuse the first batch we used and installed have been
running continually for 2 years or so. Ask me again one year from now and
I'll be able to tell you the life is greater than 27000 hours. I am
conviced that the electronics will outlast the mechanical parts of your
assembly by a huge margin.

 

just do not run the outputs to max current all the time.
run no more than 50% of rated currents and keep it in a some what
normal temp and it should at last you!

 

Ever seen a video game arcade?

The power levels and switching rate of the load(s) shouldn't
have any effect at all on the microprocessor; if they do,
then the circuit is poorly designed. As long as the micro
is run within spec, it should last on average the published
MTBF, and theoretically indefinitely, or until the dopants
diffuse through the chips so bad that they won't transist
any more.

Has anybody _ever_ seen any solid state electronics just "wear
out?"

Thanks,
Rich

 

The power levels and switching rate of the load(s) shouldn't

I was taught that any solid state device would last approximately 115
years before the dopants became to diffused.