Sign in with
Sign up | Sign in
Your question

Long-term Sub-zero processor cooling

Last response: in CPUs
Share
March 8, 2002 5:59:06 AM

So who else here has ever tried sub-zero cooling techniques? What were your results? Would anyone care to have a discussion on long-term (will work for years) sub-zero processor cooling techniques?

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
March 8, 2002 11:27:07 AM

I've never done it myself, but I had a friend who did a few years ago.

He drilled some holes into an old freezer, ran the cables and an extended power switch, and two air lines out through the holes. Then he caulked the whole thing air-tight and hooked up a dry-air pump to the two air lines and ran that until he felt safe in trusting that there was virtually no humidity inside. Then he'd cap them off and turn the freezer on and let it cool down. Then violla, sub-zero air cooling.

All that just to overclock one of the early Celerons. You could probably have gotten better performance just from a water cooler, but hey. He wanted to do it, so he did it. It was a giant monstrosity, it was power-hungry, and it was noisy. It made him happy though, so I guess that's all that matters. Heh heh.

<pre><font color=green>//error-proof coding</font color=green>
<font color=blue>void</font color=blue> main(){<font color=blue>return</font color=blue>;}</pre><p>
March 8, 2002 2:01:04 PM

My Vapochill lost cooling ability after about 3 months of use. I have yet to rebuild it. I don’t know if I was just unlucky but it cost me a ton to fix and it just wasn’t cost effective. Not to mention we have some power problems here in CA especially during winter with electric heat. Not like it really got that cold anyways. I still have it but I could not tell you what long-term effect of super cooling is. I hope talking about it will inspire me to reactivate it.

All errors are undocumented features waiting to be discovered.
Related resources
March 8, 2002 3:25:38 PM

I recall once some group made a project to make a deep freezed 486 33MHZ, OCed up to 200MHZ!
They began to OC to 50MHZ, claiming now Doom 2 was playable. Then up to 75MHZ, where Duke Nukem 3D was playable. They kept up and up, connecting it inside a freezer, adding liquor bottles if needed. Eventually they reached an extremly huge OC of 200MHZ for a few instants, before it gave in, and blew up! That's about a 640% OC!!
Not exactly long term but it did prove how sub zero makes it all worth it when properly used.

--
For the first time, Hookers are hooked on Phonics!!
March 8, 2002 4:36:33 PM

I remember the same article. That article was a joke man. ;) 

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
March 8, 2002 4:40:04 PM

What caused it to lose its cooling ability? Was it leaking refridgerant? How long did you keep it on before determining it was not going to lower in temperature? If you do not use it for a few weeks straight, the oil will soak up the refridgerant. It will then take about 10 minutes for the refridgerant to be released the next time you turn on the system. Up until that time it will appear not to be working.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
March 8, 2002 5:33:54 PM

It’s been a while. I had been using it constantly for about a week and surprisingly was receiving great cooling the night before it seemed to loose the refrigerant (R134a). I went through a testing procedure with Asetek. The unit would report temperatures well below (-26c) the recommended idle temps of (-19c). This was a sign that coolant had been lost. I spent a couple days looking around San Francisco for a place that would check and repair the unit. Every place I went to wanted $200+ dollars to fix it or wouldn’t touch it at all. So I sent it back to Asetek, which cost $50. When I got it back I tested it and it worked fine. However while testing with an Athlon 1.4 I used a crappy HS before hooking up the evaporator. During this process, with the cheep HS, I cracked the core because I was rushing the process. I have yet to get back into super cooling. I know its sad but sometimes life just isn’t that cool.

PS Intel_Inside, Meltdown, or anyone else who may comment on the cracking of the core. Don’t!!!. I don’t want to hear it and I specifically stated that you shouldn’t do this in advance. Grow up get a job and make some more crappy music.

All errors are undocumented features waiting to be discovered.
March 8, 2002 5:47:42 PM

I can maintain sub-zero but having an ice cube next to my legs is unconfy for long periods of time. My Antec case sits on my salvaged AC (from a fire damaged cray) unit that <A HREF="http://fugger.netfirms.com/reportxp.htm" target="_new">produces -4c ambient temps.</A>

With -4c ambient temps I am in <A HREF="http://gamershq.madonion.com/compare2k1.shtml?2947838" target="_new">6th place on madonion.com</A> by overclocing my video card to insane speeds without the need for any other mods to the card. currently #1 in fills with 350/810 overclock.

Ambient temps also help cool everything else too, RAM is also greatly effected by this method.

Using silicon compound on the outide of the fins on a heatsik also works well (as long as you dont make a mess). Ok, hold your finger in the air with a breeze, no wet your finger and hold it in the air. silicon doesnt evaporate =P.

Another trick is to "checker" the surface of your heat sink adding surface area, now if the checker is deep enough you will create holes in the heatsink (like a cross-cut french fry). You will find checkering on a gun grip, those little pyramids add surface area.

Proving once again that <A HREF="http://www.zombo.com" target="_new">anything is possible</A>.
March 8, 2002 7:03:23 PM

I was going to do a project but i am unsure if i will now...

with my current hardware, or the hardware that i buy soon, i was going to change my watercooler over to a closed system - pump, resevoir, radiator, etc

then i was going to get a freezer or fridge or something and do what pheonix's friend did sort of, but with the whole watercooler setup in the freezer (with some additive to keep the water from turning to ice, of course). I was wondering if this would work good, and I think it would be a cool project to try.

<font color=red><pre>i so good i jealous of me</pre><p></font color=red>
March 8, 2002 7:28:57 PM

Wow. That sounds like fun Grizely1. If you do it, be sure to let us know how it goes every step of the way. :) 

If I had the money (yeah, haven't we all said something like that before) I'd probably mess with a contained-environment refrigerated dry-air cabinet and a 'water' liquid nitrogen well cooler. (A big tank of LN2 that the warmed-up LN2 dumps back into.) The only problem though would be making sure that I always had a big tank of LN2 sitting around in case I ever needed to flush the well.

Of course, the flaw would be that the cooling system would cost more than the computer itself, which is why I don't have the money. It'd be cool to see just what kind of specs I could come out with on a system like that though.

<pre><font color=green>//error-proof coding</font color=green>
<font color=blue>void</font color=blue> main(){<font color=blue>return</font color=blue>;}</pre><p>
March 8, 2002 7:44:41 PM

Good cooling systems usually do cost a great deal. :) 

I am going to throw a 2.4GHz (400MHz FSB) Pentium 4 into my Vapochill soon to see how high I can take it through FSB overclocking.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
March 8, 2002 7:52:31 PM

Do you have a Vapochill I or II?

All errors are undocumented features waiting to be discovered.
March 8, 2002 7:59:20 PM

I have a socket-based Vapochill designed for the socket 478 Pentium 4. I believe Vapochill 1 was for slots and Vapochill 2 is for sockets. Is this correct?

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
March 8, 2002 8:01:57 PM

i will take pictures/videos and make an article to (possibly) submit to hardware sites. :smile:

<font color=red><pre>i so good i jealous of me</pre><p></font color=red>
March 8, 2002 8:04:09 PM

That's the Vapo II it has a longer hose and a better retention clip. Since mine got repaired I have the longer hose but the old retention clip I guess I have an 1.5.

All errors are undocumented features waiting to be discovered.
March 8, 2002 11:43:02 PM

Joke?
You sure? So the entire thing was fake?

--
For the first time, Hookers are hooked on Phonics!!
March 9, 2002 12:10:06 AM

Yes, it was fake.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
March 9, 2002 12:40:40 AM

Quote:
Using silicon compound on the outide of the fins on a heatsik also works well (as long as you dont make a mess). Ok, hold your finger in the air with a breeze, no wet your finger and hold it in the air. silicon doesnt evaporate =P.



OK, It seems that you are saying, that wet things are cooler than dry things, and that silicon dosent evaporate, thus it will remain wet constantly. Therefore cooling the hsf constantly.


Is that what your saying fugger?

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
March 9, 2002 12:42:15 AM

Quote:
If I had the money (yeah, haven't we all said something like that before) I'd probably mess with a contained-environment refrigerated dry-air cabinet and a 'water' liquid nitrogen well cooler. (A big tank of LN2 that the warmed-up LN2 dumps back into.) The only problem though would be making sure that I always had a big tank of LN2 sitting around in case I ever needed to flush the well.


Wouldnt the warmed up ln2 evaporate?

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
March 9, 2002 1:05:01 AM

Quote:
OK, It seems that you are saying, that wet things are cooler than dry things, and that silicon dosent evaporate, thus it will remain wet constantly. Therefore cooling the hsf constantly.

Fugger, unfortunately it simply does not work that way. It is the act of evaporation that cools you down when wet. The heat causes some of the water to turn to gas and drift away. This strips you of some heat. Sweat operates on the same principle.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
March 9, 2002 3:14:12 AM

Ray!!!!!

I was asking him to clarify so I could "educate" him, lol, you totally ruined it mr helpful!!!

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
March 9, 2002 7:56:47 AM

In fact it does work for my scenario. I shouldnt have even suggested it unless you were going with a setup like mine. as this will not work with a vapochill.

Take my video card for example, with just the heat sinks alone on the ram I reach 790'ish memory, a little silicon on the high edges and that is enough to push me 800+ argue with results?

You would have to do your own testing with a laser temp gun, the edges of the heat sink that is lightly covered will drop in temp faster and stay colder than the oncovered area, Im not saying to cover the whole sink or even attempt this with poor circulation at room temps. placement is key, farthest point from heat on high edge works best for me.

Its a trick, if done right it does work.

Im doing benchmarks now and Ill do a sandra report when im done, -3c mobo, 8c CPU @ 2.6Ghz, no volt mods.
<A HREF="http://gamershq.madonion.com/compare2k1.shtml?2954721" target="_new">New high score of 13k</A> Im #1 in fills period, with a huge lead.

The vapochill cannot cool the rest of the compnents in the case like I do.



Proving once again that <A HREF="http://www.zombo.com" target="_new">anything is possible</A>.
March 9, 2002 8:39:37 AM

Quote:
You would have to do your own testing with a laser temp gun, the edges of the heat sink that is lightly covered will drop in temp faster and stay colder than the oncovered area, Im not saying to cover the whole sink or even attempt this with poor circulation at room temps. placement is key, farthest point from heat on high edge works best for me.


That may or may not be true, but whatever is causing it has NOTHING to do with your wet finger analogy.

The wet finger is cold BECAUSE of evaporation.


Now putting silicone on a hsf may lower temps(higher surface area maybe?) but it is NOT due to the same reason a wet finger is cooler.


Simple physics fug.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
March 9, 2002 8:50:21 AM

Dumb question:

How 'bout sticking like 170w peltier or so to vapochill system?
This would bring temps down nicely?
March 9, 2002 2:17:23 PM

The Vapo dissipates approximately 100w of heat. There is a resister that controls how hard the compressor runs. A 170w peltier produces way too much heat for the Vapo’s evaporator. They do use peltiers to test the units since they produce a measurable amount of heat. I doubt it would work. You could use a 100w pelt but you would be doing an equal amount of work but use a lot more power.

All errors are undocumented features waiting to be discovered.
March 9, 2002 3:10:34 PM

Quote:
I can maintain sub-zero but having an ice cube next to my legs is unconfy for long periods of time. My Antec case sits on my salvaged AC (from a fire damaged cray) unit that produces -4c ambient temps.

How high are your power bills? Do you need a high current circuit?

All errors are undocumented features waiting to be discovered.
March 10, 2002 8:26:11 AM

IT costs about 4 bucks per month to run if I ran it constantly. It runs on 110v AC and the thermistat is digital, and its about the size of a 12 pack of soda. The blower fan is quiet, my volcano 7+ is louder.

For all those keeping score, I decided to run my ac till my machine turned into a ice cube and kicked it up to 2.7ghz to squeeze out <A HREF="http://gamershq.madonion.com/compare2k1.shtml?2966509" target="_new">this score</A>
No volt mods, no watercooling, no pelts, just custom heat sinks and cold air. I beefed up the chipset and made a plastic channel for air thru my ram. My ram was at full speed 540'ish.

Ill switch over to the 1.6a and do the same, I should be able to break 166FSB now.

Proving once again that <A HREF="http://www.zombo.com" target="_new">anything is possible</A>.
!