Direct Die Cooling.. NO Heatsink....

D

dracos

Supreme [H]ardness
Joined
Jan 8, 2005
Messages
4,354
 
No not water.....


non-conductive fluid.... it is actually thinner than water but yet heavier...


HFE Fluid..
 
Very cool. (no pun intended) :D

/me guesses at the dielectric fluid.

I'm a little curious why you wouldn't use water with that setup.Insulating the top traces on the chip would be easy enough , and the heat transfer would be better.

I can respect the dielectric fluid tho. ;)
 
And Here's the OC

upped Ram voltage and CPU core is 1.85



External Temp sensor

http://img.photobucket.com/albums/v670/dampeal/S3010009.jpg
S3010009.jpg


THE OC

http://img.photobucket.com/albums/v670/dampeal/S3010004.jpg
S3010004.jpg


Mobo Monitor Note the Case Temp and the Above External TEmp... VERY CLOSE


http://img.photobucket.com/albums/v670/dampeal/S301001111.jpg

S301001111.jpg
 
I need to do something like thsi for a science fair project, can you tell me how you did it and are you using a pump?
 
You cant use water because eventually the water "soaks" into the silicon, destroying the core. I'm not sure if the stuff he is using does that.
 
c'mon man, 2ghz is nothing on a 2600. with such low temps try pushing it straight to 2.2 then work up from there. 2.3 shouldnt be a problem either
 
not minereal oil... 3M Hfe


and I can't go any farther with the OC... Ram won't let me..

I loosened timings,upped the voltage etc... only got to 191fsb... after that I get to windows but then BSOD... it's pc2700 ram.. not bad 191 from 166 (PNY Cheap Ram)

I do have some pc4000 in my main rig, which I may stick in there and try some more later..

I'm already maxed on the vcore of the cpu as is.... 1.85...

we'll see
 
could u not try setting up a ram divider to see how far the cpu will go without the ram limiting u. U may not get the additional bandwidth but u will see how far u can push ur cpu
 
I think the Cpu is at it's max...

I stuck my Pc4000 in there and it did no different but 10 extra mhz...

and on Air cooling I could only get this to an Athlon 2800+ rating

I have the vcore all the way to Max.. maybe if there was a way to up the vcore some more, but the mobo only let;s me go to 1.85v

I played with all the setting I have in the bios and nothing doing...

Like I said I think the cpu is at it max OC....

Some are good, some are not so good... that has been proven time and time again, that the same model cpu oc's different for each one,
 
the snake said:
how did you seal that to the mobo ?
Click to expand...

Pressure and Closed Cell Neoprene

it's not to the mobo, it is to the cpu directly.. I got a 2x2x2 block, becuase the cpu is 2x2
 
That is cool as hell bro. Balls to put that on there like that...I would stress on leaks.
 
sweet idea, and it would probably be able to work even better on sealed procs like the A64's or pentiums because the cover they have over them would allow you to use water (right?) for better thermal conductivity and at a lot less than the dielectric fluid...i'm getting ideas now, how much did it cost you to make the block?
 
wow, now come over to my place and do it to my computer.

Whats with the last 2 pics?
 
The last two pics, I threw in for fun.. My cats... Hokey and Pokey...

Anyway the actual block were $3.20 each, and my time invested in them...

as for thermal conductivity, this fluid is more thermally conductive than water, so it wouldn't make much sense to use water... and if/when I do the a64 cooling, the heatshield will come off for better heat transfer...

as for the o/c... My mobo is holding me back..

I ordered an Abit nf7-s and another stick of PC4000 just for this... should be here today or tomorrow
 
Direct die is just cool :)

I haven't been too impressed with the o/c ability or the technique however.
 
well i'm sorry to disagree with you but it is my understanding that water has the highest thermal conductivity of any liquid so unless you can provide me with documentation that says otherwise i'm going to haft to assume you're mistaken.
 
YOUR CRAZY BUT IT WORKS RIGHT :D nice job

I have thought of that before... being a high end water cooling user now having 2 systems running on water I get thinking and thought what would that do well now I know and I did not have to do it... he he good work
 
GothicWombat said:
well i'm sorry to disagree with you but it is my understanding that water has the highest thermal conductivity of any liquid so unless you can provide me with documentation that says otherwise i'm going to haft to assume you're mistaken.
Click to expand...

The concept of thermal conductivity is quite vast, and too often gets over simplified.

Thermal conductivity of a fluid varies according to temperature range in which it is operating. HFE (the material being used here) has superior thermal conductivity in low temperature situations. Clearly water is no longer a viable option as your fluid approaches 0*C for obvious reasons. So there are a few options at that point:
#1) Water + de-freezing additives. EG: Salt, alchohol, etc - Most things water soluable will have an effect on water's freezing point, but they also begin to introduce problems like corrosion, toxicity, etc.
#2) Use a medium with a lower freezing point- In this case, HFE.

Other options that are quite popular are mineral oil and derivitives of such. Though they are far more viscous than HFE and are slightly more conductive than HFE.

If you are interested in looking at the thermal performance of HFE, check out:
http://st-support-cooling-electronics.web.cern.ch/st-support-cooling-electronics/CWGWeb/TABLE8-5.PDF

3m has several flavors of HFE that they have developed for thermal conductivity without any of the typical side effects of water/solute mixtures.

Downside? Cost. HFE is typically $220/Gal.
 
R!P13y said:
YOUR CRAZY BUT IT WORKS RIGHT :D nice job

I have thought of that before... being a high end water cooling user now having 2 systems running on water I get thinking and thought what would that do well now I know and I did not have to do it... he he good work
Click to expand...
well, in watercooling it's always a balance between surface area, and close contact between the hot core and the coolant.

direct die takes close contact to the max, but the surface area available for thermal transfer is teh suck.

as you ramp up an OC and start making more heat at the core, direct die temps go up more quickly than a well optimized waterblock wit lots of surface area.

direct die is cool. direct die works. it has different strengths and weeknesses than using a waterblock.
 
thank you for the information but I must point out that it proves my point that water would have better heat transfer abilities at the temps your system is running right now. So as far as I can see unless you plan to add a chiller to get the liquid down below -5C then a simple mix of water and ethylene glycol (antifreeze) would transfer heat better than the HFE that they tested in the study. I am aware that water was not an option for you because your proc it does not have the heat spreader that covers the circuitry on the newer procs but for those of us who do have the heat spreaders on our procs I think water is a better choice.


On a different note I was showing my dad your creation and he brought up an interesting point, what happens when you want to take it off? My guess is you would haft to try and drain it as best you could but I think it would still have the potential to be a very messy endeavor.
 
DFI Daishi said:
well, in watercooling it's always a balance between surface area, and close contact between the hot core and the coolant.

direct die takes close contact to the max, but the surface area available for thermal transfer is teh suck.

as you ramp up an OC and start making more heat at the core, direct die temps go up more quickly than a well optimized waterblock wit lots of surface area.

direct die is cool. direct die works. it has different strengths and weeknesses than using a waterblock.
Click to expand...

Thanks for telling me all that again BUT WHY DID YOU QUOTE ME? I knew all that and where did i ask a question?
 
R!P13y said:
Thanks for telling me all that again BUT WHY DID YOU QUOTE ME? I knew all that and where did i ask a question?
Click to expand...
you were talking about your past interest in direct die..........
 
if you had a mobile 2600+ in there you'd be able to kiss 3Ghz with direct die cooling....
 
CPU shouldn't max out there. My mobile 2600 is at 2500 Ghz on normal water, so you should have a little more head room.

Mind posting some details on how you set that up? :D - Drool -
 
You must log in or register to reply here.
Back
Top