Coollaboratory Ultra - Does it dry up?

[Vario]

Delidded my 3570k a few months ago. It was running at 4.4 ghz @ 1.3v (bad chip) with high temps around 80*C. With the summer heat, I reduced my overclock. Just checked my HWMonitor a couple weeks later and now its running 4.1 ghz @ 1.21v and high temps of 93*C Prime95!

I thought it was the summer heat but I cranked my apartments AC setting to 70*C and its still high temps. My ambient is presently 77*C . Does CLU fail over time?

I am guessing that the CLU dried up? I moved the case a couple times and it might have dried up and cracked? The Ultra is at my parents house, not at my apartment so I can't remount it for a few more days. Has anyone had experience with CLU and gradually higher temperatures since installation? It will be about a week before I redo it and I am not really worried about this chip. I am not on a mission to kill it, but if it does die I am going to just go with a Xeon 1230v2 and not overclock. I'd rather have my processor under 80*C

 

[McSteel]

I'd also like to know this. It shouldn't degrade over time, though, at least not according to it's stated chemical composition...

I remember the s478 days, and how some of the stock heatsinks used to have what looked like a thin sheet of very pliable metal in place of the usual TIM paste... I believe CLP/CLU is a similar type of material, meaning it shouldn't degrade by an appreciable degree anytime soon. Besides, what good would a TIM that fails within a month be? I'm pretty sure forums all over the net would be filled with "CLU is crap" threads by now...

Perhaps you just have that much good/bad luck, and have obtained a heavy-leakage chip, perfect for liquid nitrogen, bad for everything else. Maybe try trading it for a more air-friendly one with someone who'd want such a leaky chip..?

 

[Vario]

Yeah its been a bad overclocker since I got it. I got it to 4.3 before I delidded it, then I delidded and got good idle temperatures at ambient (19-25*C with 22*C ambient) and 4.4 ghz as the max stable overclock running a massive 1.3 maybe 1.35v, ibt temps around 85*C. Since I delidded it, its not like I can sell it for very much, maybe $120 at the most. I'll open her up in a few days and report back if the ultra dried up or degraded in a visual way. Why would this chip appeal to ln2 users?

 

[McSteel]

High current leakage chips (the ones that run hot and require more voltage to be stable) are interesting because their internal resistance drops more slowly with lowering temperature.

I could go all physicist on this thread if need be, but to keep it simple, you want a higher resistance because it limits the flow of current through the chip. Too much current over the lines (and higher clock means more current per unit time), and you have crosstalk (thus errors in computing) at best, and a fried chip at worst. Higher leakage means higher temperatures, but at -180°C, who cares?

 

[Vario]

Thats really interesting, thanks. I had heard that low asic equals good ln2 but didn't understand it without your explanation. Im not a engineer (econ and stat background) but I am interested in a lot of it.

 

[cheesy999]

cranked my apartments AC setting to 70*C and its still high temps. My ambient is presently 77*C .

Is your apartment on fire?

Maybe the heatsink mounting isn't putting as much pressure on the processor any more

 

[Vario]

Is your apartment on fire?

Maybe the heatsink mounting isn't putting as much pressure on the processor any more

haha sorry Fahrenheit.

 

[Aquinus]

Higher leakage means higher temperatures, but at -180°C, who cares?

Those people who do LN2 do. If a chip doesn't have enough leakage at low temperatures it will cold bug and won't function at all, which is why leaky chips are preferred for LN2 clocking.

 

[McSteel]

Those people who do LN2 do. If a chip doesn't have enough leakage at low temperatures it will cold bug and won't function at all, which is why leaky chips are preferred for LN2 clocking.

And how is that any different from what I already said? The "who cares?" part refers to the higher heat dissipation/max temperature of "leaky" chips, which doesn't matter thanks to liquid nitrogen cooling.

 

[Vario]

Is delid an issue for LN2?

 

[McSteel]

No, it isn't. At such low temperatures, the TIM and the heatspreader on top of the chip don't make much of a difference, so it's not exactly needed, but it can't hurt. That is, provided the overclocker is skilled in mounting the LN2 pot, and/or re-lidding the CPU.

 

[Vario]

Well Ill try posting it in a few classifieds, maybe go for a Xeon 1230v2, periodically 190 @ microcenter... Ill remount it Friday or Saturday and see if temps go down, will also wash out rad,might be packed with dust

 

[Vario]

Just an update on this, the CLU was glued to the H100i's copper. It sucked removing it. It came apart easily but the stuff was totally dried up into a thin white crust that resembled bird shit. I ended up using toothpaste on a coffee filter as an abrasive to polish it off. Under the lid was still wet liquid metal. Don't use this stuff on bare copper guys.

 

[erocker]

Just an update on this, the CLU was glued to the H100i's copper. It sucked removing it. It came apart easily but the stuff was totally dried up into a thin white crust that resembled bird shit. I ended up using toothpaste on a coffee filter as an abrasive to polish it off. Under the lid was still wet liquid metal. Don't use this stuff on bare copper guys.

It is supposed to harden and form a bond.

 

[Deleted member 6693]

From "Manual: Coollaboratory Liquid Ultra"

"Warning: Please note that you can’t use the Coollaboratory Liquid Ultra with a contact area of aluminum, as it corrodes aluminum. The heatspreader of actual processors consists of nickel-plated copper, not of aluminum. Advice: The application on copper surfaces is a bit easier than on nickel-plated ones, but possible is both."

Sounds not right .... I use it on my Antec Kühler 920 with copper - works like a charm...maybe it is the delidding that is the faul apple??

 

[Vario]

No the stuff was powdery white crud and wouldnt come off the copper