[Video Idea] Long Term Experiment: Thermal Paste Degradation

[Ridley303]

I personally would be really interested in seeing how important it is to reapply thermal paste after some time.

Similar to the Airflow Dust Experiment you could test different brands and compare it to the Graphite Pad.

The results would probably not be as spectacular as a PC filled with dust but we could learn something new and interesting.

[tp95112]

Thats going to be a good 5 year wait

[Enderman]

Yeah and maybe they can do a proper test this time instead of some inconsistent and subjective test.

Not something like "Oh we had the thermal paste in this PC for a year, our temps are 5 degrees hotter than before so this thermal paste is bad, let's just ignore all the other 200  variables that could affect the temperature."

 

 

Place a controlled amount of each thermal paste between a resistive heating pad and a heatsink where the heating element receives constant power (for consistency across each thermal paste test setup) and turns on and off every 12 hours or something to simulate temperature cycling.

 

Then after a few months compare the delta T between the heating element and heatsink from before and after the test to determine any performance degradation of the compound.

 

Since the device doesn't need to be disassembled to determine the delta T it is possible to take readings without disturbing the test, so you could take readings every 3 or 6 or 12 months and plot the performance over time.

[Ridley303]

7 minutes ago, tp95112 said:

Thats going to be a good 5 year wait

Thats possible, I don't know how long the lifespan of thermal paste normally is.

I guess another option would be getting cheap old desktops from craigslist or making another collab with that company that recycles PC hardware (forgot the name), just testing the systems and then retesting with new thermal paste.

Theres no way to know which paste it was to begin with tho

Edited April 30, 2018 by Ridley303
Typo

[LogicalDrm]

9 hours ago, Ridley303 said:

Thats possible, I don't know how long the lifespan of thermal paste normally is.

I guess another option would be getting cheap old desktops from craigslist or making another collab with that company that recycles PC hardware (forgot the name), just testing the systems and then retesting with new thermal paste.

Theres no way to know which paste it was to begin with tho

Premium cooler brand recommends changing paste every 3 years. To get results which aren't in margin of error, test would need to last 5-10 years. Considering that first questions about when air pressure test would finish were made 3 months after it started... And that most people change their system in every 3-5 years anyway. I don't think they will have resources for this.

[WoodenMarker]

1 hour ago, VegetableStu said:

is manually drying out the paste possible? like heat cycling it with a torch indirectly or something

Yes, that can be done but I don't see how you can get any relevant data from doing so.

[LogicalDrm]

15 hours ago, VegetableStu said:

is manually drying out the paste possible? like heat cycling it with a torch indirectly or something

They could as well hire chemist to run numbers and go test different paste's. Thats how manufacturers R&D pastes.