Premature HP notebook heat pipe failure due to overheating.[quick update]

[Refrigerator]

Just fixed some water damage on a G7-2246NR and immediately after power on noticed the APU area getting warm and the fan going pretty fast.
After some struggle i reinstalled the OS and installed speedfan to have a look at the temps while i play some 1080p video.
APU got hot and fan went nuts but the air blowing out was barely warm, video was stuttery.
Before i did anything with the notebook i cleaned the dust from the heatsink and added new thermal paste, so that it's out of the way.
The amount of dust indicates that the laptop had been running hot, the cooked old thermal paste confirms that it must have been very hot.
Opening the error log shows multiple overheat events.
Taking into account all this and some info from the internet i made the conclusion that overheating causes premature heat pipe failure in HP laptops from around 2012-2013.
While i'm not 100% certain of this i think it's a something worth taking note of.
If i find a replacement heat pipe i might do some experiments with the old one, maybe add some water as a working fluid and seal it. I'd do it now if i had a spare laptop to use in case the experiment failed.

[Armadillo]

Yes, I agree with your findings.
Once the liquid and media deteriorated, it's a goner, probably at best 10 - 20% heat extraction ability, I reckon.
I've seen these faulty heatpipes floated the ebay and market scenes to the lesser informed and innocence buyers, holes in pipes, visible re-solder works etc...
If you heat the heatsink, the heat would quickly dissipate to the fins is a good indication of good heatpipe.
I would be eager to see your experiment to rejuvenate the faulty heatpipe. IPA, 134a etc..
The RX24 Style Aluminum Housing Resistor say 25 to 30 watts form a good heat source for testing.
thanks.

 

Edit: Oh yeah, I think you need to vacuum the heatpipe for it to do its magic.

[Refrigerator]

Yes, I agree with your findings.
Once the liquid and media deteriorated, it's a goner, probably at best 10 - 20% heat extraction ability, I reckon.
I've seen these faulty heatpipes floated the ebay and market scenes to the lesser informed and innocence buyers, holes in pipes, visible re-solder works etc...
If you heat the heatsink, the heat would quickly dissipate to the fins is a good indication of good heatpipe.
I would be eager to see your experiment to rejuvenate the faulty heatpipe. IPA, 134a etc..
The RX24 Style Aluminum Housing Resistor say 25 to 30 watts form a good heat source for testing.
thanks.

 

Edit: Oh yeah, I think you need to vacuum the heatpipe for it to do its magic.

Actually making a water filled heatpipe is very easy since it doesn't require a vacuum pump.
The way it's done is some amount of water is put in and the heatpipe is brought to a boil.
The water vapor fills the whole heatpipe and if you seal it while it's hot the water will remain at it's vapor pressure thus allowing it to function as a working fluid.
And it's working temperature range of 20^oC to 150^oC makes it perfect for a laptop.
Of course measuring out the right amount of water ir very important since there has to be enough fluid to flow but not flood the heatpipe.
I might as well just do it next weekend, i kind of don't have time now.

[Armadillo]

Ease of homebrew, I think there is definite merit without a doubt there, except that, it has higher wall temperature at the evaporator and that's where Junction Temperature of the semiconductor may be of the concerned there trading off amongst many other factors like speed, cost$, space etc...
I wonder why commercial cpu cooler doesn't spec their thermal heat resistance performance and wall temperature in their spec sheets. But I think, most people won't bother anyway as long as the search engine runs everyday.... 

[Refrigerator]

Took me quite a while to get this update here since i didn't have a second laptop to use in case my experiments with my current one failed.
But anyways i'm here to say that the heatpipe was not leaky, in fact it was very healthy as indicated by my flame test.
I gathered some other heatpipes of similar pipe width to compare and had a flame hit the CPU heat spreader while i held the fin part with my hand.
Of the three heatpipes that i compared the the suspect heatpipe came in about in the middle in terms of heat transfer.
"So what made your CPU overheat" i hear you ask. It actually was the poor contact between the CPU and the heat spreader.
As soon as i took the heatpipe off i noticed just how thick the layer of thermal paste was on the CPU, it was apparent at that point that the heat spreader was not pressing on the CPU the way it should have.
Another thing to note is that the surface of the heat spreader was nasty and had a couple of pits.
My first attempt to remedy the situation was to bend the stamped sheet metal heat spreader bracket up a little to make the screws apply more pressure to the CPU, this did not help the high temps, however.
My second try involved a 0.5mm thermal pad and some fresh thermal paste, needless to say this did not help at all, as a matter of fact it made things much worse, enough to make the PC shut down and reset the BIOS presets in the process.
After that went in and ripped off the soldered on metal bracket off the heat spreader and grabbed some 2000 grit wet&dry to polish up the surface.
I then epoxied the metal bracket back on making sure the heat spreader was protruding 0.5-0.7mm more than before.
After installing this back in with some cheapo white thermal paste since i was out of the CoolerMaster stuff i had before the improvement was apparent.
Idle temps dropped by 10-15^oC and load temps went down by about 10^oC.
Valley benchmark results were double that of what they were before the fix, the air coming out feels much warmer than before.

TL;DR - the metal bracket on the heat spreader was poorly designed resulting in the heat spreader not applying the right pressure to the CPU causing a thick layer of thermal paste between the CPU and the heat spreader.