Temperature cycling effects on electronics durability

[mrspeedybob]

I just went out to my car to retrieve my laptop. My car is at about 1 degree C, my house is at about 20 degrees C. I soon noticed condensation forming on the screen and it occurred to me that condensation was probably forming inside as well on the electrical components. The combination of moisture, electricity, and dis-similar conductors seems like a perfect recipe for corrosion.

I'm clearly not doing my laptop any favors by treating it this way, I'm just curious, how bad is this in reality?
Do Electronics manufacturers take measures to mitigate this kind of damage?
Is this sort of thing one of the primary causes of electronics failures?
Given that any piece of computer equipment is going to be obsolete in 3 to 5 years anyway, is this something I should worry about? Or will obsolescence come before thermal cycle or corrosion damage anyway?

 

[anorlunda]

My former employer demanded that all employees take their laptops home with them. Exception, if you rode a bicycle or a motorcycle or if you carried your stuff in the bed of a pickup truck, they demanded that you do not take the laptop home and they bought you a second laptop to keep at home. I think they cited damaged screens as the reason.

 

[berkeman]

Different electronic devices are designed for different environmental conditions. Laptops (except for ToughBook type laptops) are not designed for moderate shock/vibration or temperature cycling and high humidity environments. Also, you bring up a good point about the quick obsolescence of many laptops and similar devices, but I still use some 10 year old laptops in some of my lab setups at work.

For most of the devices I design, we use Accelerated Life Tests (ALT) and STRIFE tests to be sure to catch any design issues that could shorten the life of our products. ALT testing is 85C at 85% relative humidity for 1000 hours while operating. That catches many contamination and corrosion issues that you may have. STRIFE testing is fast ramp temperature cycling of the product with 10C/minute ramps and power cycling several times per temperature cycle. The temperature limits are typically 10C to 15C outside the published operating temperature limits of the product. The life cycle of our products is typically 10-15 years, and we have definitely found and fixed issues with these tests in some of our products before they were introduced to the public.

 

[dlgoff]

Different electronic devices are designed for different environmental conditions. Laptops (except for ToughBook type laptops) are not designed for moderate shock/vibration or temperature cycling and high humidity environments. Also, you bring up a good point about the quick obsolescence of many laptops and similar devices, but I still use some 10 year old laptops in some of my lab setups at work.

For most of the devices I design, we use Accelerated Life Tests (ALT) and STRIFE tests to be sure to catch any design issues that could shorten the life of our products. ALT testing is 85C at 85% relative humidity for 1000 hours while operating. That catches many contamination and corrosion issues that you may have. STRIFE testing is fast ramp temperature cycling of the product with 10C/minute ramps and power cycling several times per temperature cycle. The temperature limits are typically 10C to 15C outside the published operating temperature limits of the product. The life cycle of our products is typically 10-15 years, and we have definitely found and fixed issues with these tests in some of our products before they were introduced to the public.

Aren't these test done while the device is powered up? When I've had something get damp/wet, I just make sure that it is within a "normal"(dry) environment before applying power. Heck, back when wave sloldering double sided circuit boards, they went from 1,1,1-trichlorethane to hot soapy water. Sure brightened them up.

 

[berkeman]

Aren't these test done while the device is powered up? When I've had something get damp/wet, I just make sure that it is within a "normal"(dry) environment before applying power. Heck, back when wave sloldering double sided circuit boards, they went from 1,1,1-trichlorethane to hot soapy water. Sure brightened them up.

Good point. Our devices are continuously powered, typically. A different worst case would be to be subjected to condensation in a cold environment and then powered up...

 

[dlgoff]

Good point. Our devices are continuously powered, typically. A different worst case would be to be subjected to condensation in a cold environment and then powered up...

Guessing a higher failure rate with this kind of test on a device?

 

[berkeman]

Could be, but it should depend on the type of circuits in the device. If it just has vanilla logic at low speeds, then the added capacitance of the water on the PCBAs won't make much difference. Extra water on the PCBA of your smart phone (with lots of RF circuit traces) will hinder operation pretty badly.

 

[meBigGuy]

Consumer product life testing that involves temperature cycling always specifies "non-condensing".
Devices that require high reliability in condensing environments require environmental coatings and other special treatments (anti fungal, for example). For example remote supervisory nodes located in Louisiana swamps (I've tried to repair pcb board returns from such).

I've never actually considered this problem with respect to my laptop, phone, ipod touch, tablet, etc. Could be just as bad as going through a washing machine (which my ipod touch survived). Pretty important not to leave them in the car for extended periods.

Consumer product PCB cleaning is sketchy, at best, and condensation could easily leave corrosive byproducts and alter sensitive circuit operation.

 

[dlgoff]

... condensation could easily leave corrosive byproducts and alter sensitive circuit operation.

Hence the need for hot soapy water.