Heat pipe coolers lose effectiveness over time?

[Ugly n Grey]

I just recently (this week) dug out a couple systems I hadn't touched in just over two years.

Sys specs...
DFI Ultra D 939 with x2 Athlon 4200
TT Big Typhoon
OCZ powerstream
I changed nothing in the box except the video card and that went DOWN to an old ATI PCI card that was in a box of junk I bought.

Now, I used to run this thing 24/7 with a non divided DDR speed of 260MHz and pushing quite a few volts though the cores for my OC. Now, it never got over 46c back in the day and I do have the original notes and screenies.

Few days ago I dig it out of the box and put the PCI card into it and then fired it up AT STOCK. All seemed OK, fans all turning, PS keeping the rails up and all that good stuff. The fan on the BT was changing speed just fine as cores heated up etc. I install the F@H client (basic single core version). I open up the system monitor... 55 C and it's going to shut off at 60... I grunt to myself and figure the HS is loose or something or the paste is too old... fine. I tear it down and put some new AS on the (naked - no IHS) core and slap back on the BT. No love. Stressing a single core makes the temps sky rocket. Changed the fan on the BT. No love.

I scratched my butt for awhile and figured... no biggie, the BT must have bitten the dust or got bounced too hard. I dig around and sure enough , the stock cooler is still in the original retail box along with the IHS. I change the mounting hardware back to stock ( yeah I have an idiot proof method for strapping the IHS back on) and put on the stock (two heatpipe) cooler (it was never used even once before this).... well temps are better but the darn thing is noisy as hell. Then I watch it running 50% (one core loaded) and the temps rise some more.... 49 C. At stock.

At this point, I think I'm a total idiot and daren't ask in case it gets confirmed by too many sources. Back to the Big Ugly Box of Junk. Yippee... Zalman copper silent rig . I stick it on. Temps.. 36c single core loaded. Load the other core, 41c. (all at stock mind you with ambients around 23c).

So I don't really know what to think. I have two radically different heat pipe based coolers that won't work worth a... well they won't work. After two years of sitting absolutely still.

I dunno. Thoughts?
I was resurrecting this rig for folding, so it's not a panic. I'm just ... puzzled.

[alexio]

It sounds like the BT and stock cooler weren't seated properly. Although this may sound unlikely it sounds more unlikely to me that 2 different heatpipe coolers developed leaks while sitting still.

[Ugly n Grey]

Good thoughts... but... I mean... three times in a row I blew it on the seating? Cripes.

[Kasparz]

Wow, UnG is BACK. Glad to see you're fine

Vapor inside heatpipes might be leaking or damaged somehow. Heatpipes don't like abnormal temperatures that can also damage them.

[Ugly n Grey]

Wow, UnG is BACK. Glad to see you're fine

Vapor inside heatpipes might be leaking or damaged somehow. Heatpipes don't like abnormal temperatures that can also damage them.

Hiya Kasparz
BINGO!
Boxes were stored in unheated storage in a geography that hits -40c at times. I didn't think that vapor would have a problem with that....

Learn something new every day... crud

[Demo]

woah -40c? where is that place ?!!

[Ugly n Grey]

woah -40c? where is that place ?!!

Central Ontario, near Ottawa is where it was stored.

[Morais]

OOO thats intriguing...Did you check the contact of the heatpipes in the copper base and the bigger heatsink?

[cegras]

Well, the heatpipes ARE soldered shut .. perhaps the constant cold and reheating causes expansion and contraction?

[Ugly n Grey]

Well, the heatpipes ARE soldered shut .. perhaps the constant cold and reheating causes expansion and contraction?

I wouldn't expect the solder to let go on it's own. That's well within the thermal limits of standard solder. But maybe.... the vapor hitting some serious cold patches caused some thing weird to happen.... oxidization with the copper? who knows, because I don't know what's in the contents. I only have the evidence that neither functions at this point. Not that it's a huge deal, coolers are cheap and some of the new ones are quite good.

[cegras]

Could be that without continuous recycling and extreme temperatures, the vapour inside somehow reacted with the copper? But that's an extremely far fetched hypothesis.

Instead of looking at the heatpipes, check to see if the fin connection / base connection to the heatpipe is OK. Perhaps those are where the problems lie.

[alexio]

If the heatpipes are indeed damaged it's most likely the wick material inside the heatpipes that is deformed or has cracked. The wick material can be a metal or a ceramic type material. The wick has a capillary function in which it makes the condensed vapor travel back to the hot side of the heatpipe.

In case of a ceramic material the wick may have cracked, in case of a metal (presumably copper or an alloy containing copper) the wick may have been deformed during freezing of the coolant. If the coolant is demineralized water it may have greatly expended when it solified to ice breaking or bending the wick structure.

[cozwin]

cut it open and look? :P

[Ugly n Grey]

Serious thanks for the nice diagram Alexio, it looks cool and will explain some neat stuff.... I was aware of the construction, my lack of knowledge centers around the chemistry... anyone know what kind of goop is in there?

cut it open and look? :P

And not knowing what I might get on me is exactly why the dremel will stay in the tool box (for the moment)

[cegras]

I would either say water at (low?) pressure (boils quicker?) or some sort of alcohol / mixture.

[alexio]

Ammonia, methanol, ethanol, water or a combination of these (ammonia already contains water of course).

Pure (demineralized) water without any additives under low pressure gives best performance of these coolants, so that's probably what is used in most IC coolers.