ESP32 life-time

[mbratch]

My program was running in an endless-loop with no delay, so the CPU-load was 100% all the time. Maybe this caused the problem.

That's certainly an issue. What is the loop doing on each iteration? FreeRTOS is a cooperative real-time system. You should have at least a small delay (using `vTaskDelay`), or regularly have a call to wait for an event. With a dead loop, if there are any lower priority tasks in the system, or if you have software timers, they won't run. But until you rectify that and run more tests, you won't know if it is the primary cause of your problem. So fix that first, then see what happens.

[svenbieg]

It's a thermal solar controller. I had a simple loop reading analog values, setting the output corresponingly and starting over. I have changed the code already, now the main task is a in a waiting state until a timer is triggered. There are two timers now, a 100ms-timer for analog samples and a 1s-timer for controlling the output. Unfortunately i can't test my controller anymore because my father has bought a new plant. I've got no idea for another application at the moment, but i'm going to post an update as soon as i know more.

[mbratch]

Perhaps you could run the unit on its own and just feed it fixed, artificial input from a breadboard or something? That way you can determine if your original problem is resolved. Even if the problem changes a bit with a different plant, there would be a good principle to discover.

[svenbieg]

Thats right! I could do this, but it's not that interesting for me anymore. The new plant has a new controller and should work many years. Maybe i find another application.

[vonnieda]

For what it's worth, and to answer your original question, my company sells a number of ESP32 based products and we have had no longevity issues at all. We have tens of thousands of ESP32 devices in the field, some of them having been running 24/7 for 4+ years.

Jason

[mbratch]

Thats right! I could do this, but it's not that interesting for me anymore. The new plant has a new controller and should work many years. Maybe i find another application.

I was thinking that the lesson learned from the old application in determining the root cause of your original problem could be useful for future applications.

[svenbieg]

Thats right! I could do this, but it's not that interesting for me anymore. The new plant has a new controller and should work many years. Maybe i find another application.

I was thinking that the lesson learned from the old application in determining the root cause of your original problem could be useful for future applications.

Maybe i track the temperature and humidity of my living-room and see how long it works.