i7 920: idle @ 99!?

R

ryanrhee90

n00b
Joined
Jan 20, 2010
Messages
26
Hi all, I'm new to this forum.

Here's my system set up. It's less than 6 months old, though I don't remember exactly how old it is.

Gigabyte GA-EX58-UD5 mobo
EVGA nVidia 8800GT 512mb
Intel i7 920 CPU @ stock 2.66Ghz (133Mhzx20), cpu v-core: 1.3v
Spire Thermax 2, with 2x Nexus 120mm PWM fans, one on each side. Hacksawed off the top inch or so of the heatsink to make clearance for case. (Crimped shut the open end of the copper pipes so that the ends of the pipes are shut and are conductive.)
2 x 2Gb G-Skill Ripjaw DDR3 1600Mhz 9-9-9-24-2
LIAN LI Black Aluminum PC-C32B Case (Rackmount)
Highpoint Rocketraid 4322 2x eSAS RAID card (dual 1.2 intel IO Processor)
Sans Digital EliteStor 4x SATA drive bay (1U)
Presonus FP10 (1U, firewire audio device)

That temp is ridiculous. Also, that's the temp from the BIOS reading.

I'm pretty skeptical of that temp reading. Honestly though, I don't think I ever measured the temperature of my CPU when I first built the PC, so I'm not sure if this has been going on for a while or if it just happened. However, earlier today my PC started shutting off for no reason, which is why I suspected over heating. It's not a new system, so I did vacuum out the case, reseat the heatsink, etc, etc. No help.

I don't know what's wrong. Surely, it isn't that bad to saw off the tips of a heatsink, right? I'm not even overclocking. Everything is at stock.

Best Regards,
-Ryan
 
ryanrhee90 said:
Spire Thermax 2, with 2x Nexus 120mm PWM fans, one on each side. Hacksawed off the top inch or so of the heatsink to make clearance for case. (Crimped shut the open end of the copper pipes so that the ends of the pipes are shut and are conductive.)
Click to expand...

there's your problem. those are heatpipes.

gl hf
 
By sawing off the tips it's possible you allowed the fluid inside the copper heatpipes to escape. Without the fluid inside, the heatsink loses much of its conductive cooling capability.

The only other explanation would be way too much or no thermal paste. I'd strongly suggest you switch to the stock cooler (you're on stock clocks anyways, the stock hsf is more than sufficient) while you decide what to do.
 
Do you have pics of the heatsink?

If you cut the heatpipes, that could indeed hurt the performance of the heatsink.
 
Oh man. That's very risky, cutting heat pipes..
 
Last edited:
IMHO, you need a new heatsink. Cutting through the heatpipes pretty much ruined it.

On a side note, how come your Vcore is so high? Is that the default setting? I've got a 920 D0 that runs @ 3.2ghz just fine with 1.05V with the stock intel heatsink. Just have to get around to installing the noctua to take it higher... I would try decreasing the Vcore to see if you can get temps stable.

On a side note, I suspect the cpu cooler is really screwed up. The TJ for the i7 is 100oC, so that is why you are probably getting readings of 99oC (the cpu temp would probably go higher, but the cpu is throttling itself to avoid permanent damage).
 
LMAO.

Seriously, don't try to fix your pipes by putting something like water into them and sealing them up with JB weld, or other adhesive. If you do, get ready to party with shrapnel. You've just made a bomb. If the liquid boils, that is, it will explode.

http://en.wikipedia.org/wiki/Heat_pipe

Read that so you will understand heat pipe theory. Easy to read.

I've often wondered why we don't see liquid helium in heat pipes for our coolers.?
 
You answered your own question when you took out the hacksaw.

Heatsink + hacksaw = bad.

Like runnin' the 24 car at Talladega with no coolant in the radiator.
 
According to the wiki page, a heatpipe without working fluid (or an overheated heatpipe) only has about 1/80 of the original conduction capacity. That fact coupled with your 1.3v vCore is what's likely causing your incredibly high temperatures.
 
Heatpipes also usually have neative pressure in them - vacuum - to help the liquid boil at a lower temperature.

Most heatpipes also use a wicking material in them to help with the movement of the hot and cooled liquid.

Cutting a heatpipe ruins it unless you have the equipment and knowledge in order to restore it to a working condition.
 
Like I said, just fill the pipes about 1/4 full with gasoline, then seal them with JB weld.

Get a really long extension cord, like 100 feet.

Let Line pack rip!

This will be the coolest video on Youtube:

"Overclocked 920 blows up!"
 
Wow and i just posted this in my thread right below last night....some guy on new egg gave a heatsink a 1 egg review after he cut off the top of the heatpipes and it didn't work.

When in the world did it become normal to cut things off instead of measure before you buy?
 
It's good that people are trying to mod things, but some things need to be researched before messing with them. Thermodynamics is a complex area. On the other hand, you could probably hacksaw your CPU in half for only half the Mhz, and use the other half to build another machine. Just saying.
 
Last edited:
Surely this is just a really good trolling attempt. I don't see how someone with access to that kind of hardware could have done something as dumb as that.
 
you could try and recharge it and jb weld--seriously.

not going to work great, but might... Trick is you have to seal the sucker while you're finishing boiling off the steam of the water you introduce.

Non trivial, this.
 
ryanrhee90 said:
No thanks, guys: going for a new HSF.
Click to expand...

Now this makes sense.

Also, you could not really refill it. It takes a very precise measurement of liquid to pipe length + diameter to get the effect. Also, if it boils and you have too much liquid, BOOM!@(%&%
 
Derfnofred said:
you could try and recharge it and jb weld--seriously.

not going to work great, but might... Trick is you have to seal the sucker while you're finishing boiling off the steam of the water you introduce.

Non trivial, this.
Click to expand...

Please don't try to give advice about stuff you don't know about.

For one... heatpipes don't use water in them.
 
Ah,

I actually DO know what I'm talking about, unfortunately. So be careful who you chide. Admittedly I didn't give the OP the entire bit of information, but I didn't lie either. The thermodynamics aren't that complex, and with the right setup it's doable to heat to boil and get it to work.

It just requires a bit more than the OP has in store--I was thinking about the lab equipment available to me. A foodsaver could almost get there ;-)

Heatpipes can very definitely run water, just depends on the goal temperature range of the system. It's too high of a boiling point at atmosphere, but under a mild vacuum (27 in water, give or take), you'd start getting very good vaporization around, oh 42C or so. Just look up the boiling point of water at various pressures and you'll get a good idea. Acetone for us hacks at atmosphere is not out of the question--just tricky.

Edit for grammar fail.
 
Last edited:
My old electronics instructor used to have a joke about applying too much voltage to an IC.
He used to say you knew you applied too much when you let the smoke out....

Now figure out why that came to mind when I read this post....

:)
 
Derfnofred said:
Ah,

I actually DO know what I'm talking about, unfortunately. So be careful who you chide. Admittedly I didn't give the OP the entire bit of information, but I didn't lie either. The thermodynamics aren't that complex, and with the right setup it's doable to heat to boil and get it to work.

It just requires a bit more than the OP has in store--I was thinking about the lab equipment available to me. A foodsaver could almost get there ;-)

Heatpipes can very definitely run water, just depends on the goal temperature range of the system. It's too high of a boiling point at atmosphere, but under a mild vacuum (27 in water, give or take), you'd start getting very good vaporization around, oh 42C or so. Just look up the boiling point of water at various pressures and you'll get a good idea. Acetone for us hacks at atmosphere is not out of the question--just tricky.

Edit for grammar fail.
Click to expand...


Water has a bad habit of oxidizing/corroding the metal though... Sure it may work for a while until that starts to happen... but then performance would tank after that.

An automobile venturi style A/C evacuation unit can easily get to 27 inches of vacuum... last time I used it, I was at 28-29 according to my A/C guages, so techinically I have equipment that would work.
 
LOL my sarcasm detector was off, but I see so many " is this temp OK ? " questions I forget some of you guys do know what the limits are. So apologies to Doug but toss a "roll eyes" in next time so this dense half asleep (or half crocked) mod gets a clue. :D
 
Hey guys!

:) I know you've been bashing me for being dumb about cutting heat pipes, and yes, it was pretty stupid.
But I took your advice, got a new heatsink (CNPS10x Flex), and my idle temps are now at 48±2 ºC.
This definitely makes me feel much safer, but, is it still too high? I have two 120mm fans pushing air through the heatsink.

Any input is greatly appreciated!

-Ryan
 
Cyclone--automotive ac vacuum stuff is probably pretty close to the system they use for this type of application.

If all the water touched was copper, some would leach out of the pipes until a certain saturation level is achieved. I don't know the exact kinetics off the top of my head. If there's no other metal that the water touches, then there's no possibility of establishing a galvanic cell.

Anyhow, given the cost of a new heatsink versus the speculative work we describe, it makes total sense for the OP to take his lumps for cutting the heatpipes, buy a new heatsink and get on with it. ;-)

Which seems like he did. If this is the dumbest thing you ever do in your life--you're doing great.
 
ryanrhee90 said:
I took your advice, got a new heatsink and my idle temps are now at 48±2 ºC. This definitely makes me feel much safer, but, is it still too high?
Click to expand...

Yes, they are, absolutely. Since you have a 920 though, go ahead and cut it in to 4 pieces. Each piece will only give off a quarter of the therms. Only use as many pieces of your 920 as you think you'll need each day. Drop all four in for heavy loads. :D

Your temps are fine. You've got some overclocking room!
 
You must log in or register to reply here.
Back
Top