Overclock.net banner

1 - 20 of 42 Posts

·
Registered
Joined
·
43 Posts
Discussion Starter #1
Hi guys,
my friend from school and I are planning to do a project for school. We'll take a Duron 600Mhz and a Geforce Ti 200 and cool it with a mixture of liquid nitrogen and dry ice! We'll install a bowl around a passive cooler for CPU, one around the Graca and one around the ram. We'll freeze everything with some nitrogen, and then fill the bowls with dry ice! I was thinking of reaching about 1.6 Ghz with Duron, and 320/620 with Geforce Ti 200! The SDRAM will just be overclocked a bit! Don't forget that the temperature of liquid nitrogen is about -196°C/-384°F and dry ice is -80°C/-176°F !!!!! So that's kind of an effective way of cooling something ;-)

Will that work?
 

·
Registered
Joined
·
698 Posts
yes it will work. By not doing total submersion cooling you arent going to murder the capacitors or the battery (the dielectrics in the caps and the battery crystallize at that low of temperature). You need to proof the mobo from condensation however. You also need to understand that you will need to constantly replinish the LN2 because it evaporates very quickly since its boiling point is around -172C if I remember correctly.
 

·
Premium Member
Joined
·
3,990 Posts
I'm confused, are you doing direct die cooling on the CPU? ^yes, and what Denmor said is right, you NEED to cool those capacitors, and for condesation proofing, well.. wait for NoAffinity or Plauge to post, they know Hardcore details into condensation proofing.
 

·
Registered
Joined
·
698 Posts
If you only use Liquid Nitrogen in the beginning it will melt very quickly and dry ice wont be enough unless you can find a way to pack it down onto the CPU so tightly that there isnt any airspace, which you cant :p Your best bet is to put LN2 and put Dry Ice in it. That will add some longevity to the Dry Ice and keep it from flash boiling, but just using LN2 at the beginning wont be adequate.

To restate the capacitor thing. Yes you need to cool them, but not with Dry Ice or LN2 because the dielectric in the capacitors will crystallise causing the capacitors to discharge on the Shorted Circut, and kill the caps and probably the motherboard. I would say hit them up with a fan, you'll have enough cold air that they will get nice and cool.

Actually, I have a website for a similar project, but they used full submersion cooling and it explains what happened to the mobo..

Heres the link:

http://www.octools.com/index.cgi?cal...bmersion2.html
 

·
Registered
Joined
·
43 Posts
Discussion Starter #7
Nonono, you didn't understand what I said. I leave a passiv cooler on the CPU, then I put a bowl ,with a whole at the buttom of it, on/around it, and then the passive cooler will be frozen with nitrogen and always be sourrounded by dry ice! Won't that work?
 

·
Registered
Joined
·
43 Posts
Discussion Starter #9
Could you please explain to me what this homepage should tell me? As far as I can see, all these guys drown their mobos in liquid nitrogen, which is not what I was gonna do!!! Plus, I don't know how to get so much nitrogen into school!!!
 

·
Registered
Joined
·
698 Posts
It wont work because metal is a good conductor of heat and therefore would melt the LN2 in a matter of minuites and the Dry Ice wouldnt help much. Your best bet is to fill the container with LN2 around your heatsink and put dry ice into it to slow the evaporation of the LN2 and top it off as necessicary.

Your idea would work fine if it was a moble processor and not producing over 50 Watts of heat to disappate.

And LN2 isnt that hard to get. We have a huge container of it outside one of the buildings and its vented on a regular basis.
 

·
Registered
Joined
·
43 Posts
Discussion Starter #11
Why don't you think a passive cooler surrounded by dry ice will be enough? I could also install a huge headpipe and put dry ice on that! Would that be better?
 

·
Registered
Joined
·
43 Posts
Discussion Starter #13
As I already said: I wanna do this at school, and I'm 15 years old.
I WILL N O T BE ABLE TO BUY TONS OF NITROGEN!!!!
 

·
Registered
Joined
·
3 Posts
Quote:

Originally Posted by Xstream
Nonono, you didn't understand what I said. I leave a passiv cooler on the CPU, then I put a bowl ,with a whole at the buttom of it, on/around it, and then the passive cooler will be frozen with nitrogen and always be sourrounded by dry ice! Won't that work?
This idea will never work . Maybe for a few hours but only with the computer stopped. Because a processor dissipates around 40 Watts (Duron) (let`s say 20 W in these temperature condition) i`d like to know where this power will dissipate if there is no exchage of temperatures with the enviroment(without reminding the heat produced by memory & video...).
The second problem is a big one also : There will be a delimitation area of a few milimeters between the cool-box(processor & memory) and the rest of the motherboard (with capacitors & other components that does`nt support low temperatures).In this area of a few milimeters will shurely appear condensate of water that will shortcircuite the motherboard ...Usualy big capacitors are located near the CPU....I`m curious how will you isolate them from the CPU...
The third problem is that circuits of silicon may not work well at very low temperatures (as anybody knows SEMIconductors act well at room temperature & lower but may become conductors at very low temperatures).
 

·
Registered
Joined
·
698 Posts
Actually you see a ramping of CMOS efficency as the temperature gets lower, so the processor at least will run better at a lower temperature.

As far as the delimination area, if he properly proofs his motherboard from condensation, he will not have a problem with it.

The only real problem I have with this project xStream is your method of cooling the CPU.

Here is why it wont work, the passive cooler "Frozen" by liquid nitrogen will melt very quickly due to heat from the processor, its just that simple. The temperatures we are talking about are way below room temperature, not even considering the heat that the processor is adding to it. A Heatpipe would be marginally better, but still not good enough. Your best bet is to create a coolbox, fill it with LN2 and float Dry Ice in it because the Dry Ice will break up the surface area of the LN2 and it wont evaporate as fast.

As far as aquiring the LN2, you should talk to your Teacher. You cant even buy LN2 since you are a minor
. Your teacher should be willing to help you out.
 

·
Registered
Joined
·
8 Posts
You don't need tons of nitrogen to do this properly, i'd think a few quarts would be enough.

However, you're looking at doing a ghetto-cooling system for cheap. As remarked by several people in this thread, it's not going to work as is, and even if you fix it up, you run a huge danger of frying your motherboard due to condensation issues.

I also don't see why you're using dry ice to begin with. LN2 is colder than dry ice and will ultimately work better.

The other problem is that you have no clue as to what your processor will hit with such cooling. You'll have to slowly raise your overclocks, and that will require a good bit of LN2 to do properly. You can take a shot in the dark and guess what it will overclock to, but i doubt that will be a good technique.

If you insist on doing thing as is, i'll give you 3 bucks shipped for your processor afterwards. I wanna make me a keychain
 

·
Registered
Joined
·
43 Posts
Discussion Starter #17
I saw pictures a graca totally frozen, there was a layer of ice, about 1cm thick, on it. It worked!!!
BTW: Everybody is talking about low temperatures. I want to overclock it till I get the temperature of a standart-cooled duron again!!! I don't know how much it is (just know the die temp is 90°C). So it depends on the temp when I'll stop overclocking!!!
 

·
Registered
Joined
·
698 Posts
yeah, and I assure you that it was Condensation proofed under that Ice. The Insulation can freeze all it wants, as long as it dosent hit the electronic components.

Man, we have told you what you need to do to make this work, and we have explained why your origional idea wont work. I dont know what else to say.

Actually I just thought of something. Your overclocking adventure will stop very much before you hit a standard Duron temperature again due to the fact that you arent cooling the northbridge and the clock generators. To be honest, you could do the northbridge, but the Clock Gens are so small and there are so many of them that you wont be able to cool all of them well at all. So that goal is unrealistic, But I do think you could get a fair overclock if you did it correctly.
 

·
Registered
Joined
·
8 Posts
There is no doubt in my mind that your chip will be unable to overclock any further long, long, long (etc.) before it reaches the temperature of around 90 degrees celc. Furthermore, since you don't have a flowing supply of LN2, your temperatures are slowly going to increase until your chip fries and you have a shiny new keychain.

You need to re-think your approach on this one. Denmor's also right about the dieletrical of the caps having difficulties dealing with the cold temperatures. I suppose in theory, you could replace all of your caps with film-based ones, but this would be tedious and expensive
 

·
Registered
Joined
·
698 Posts
And, as you previously said to me Ebo, he would have to convert the capacitor values from dieletrical to film, which If there is a ratio for, I nor you dont know it.
 
1 - 20 of 42 Posts
Top