• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Idea on doing direct die cooling

F

Fark_Maniac

2[H]4U
Joined
Feb 21, 2002
Messages
2,438
I've had this idea for quite a while now, and am finally getting close to starting on my direct die cooling project. Before, I had done a little research, and one major problem always came up. AMD procs are porus and water always seems to slip through the processor. I have an idea to put a small layer of 2 part epoxy over the surface of the processor...minus the core slug. What do the rest of you think of such an idea with the epoxy?
Next, I haven't yet gotten into watercooling...so I have no idea on how much water to run through it to keep an OCd proc cooled. The test proc will be my old Duron 700 hardwired for 900.
Any hints? tricks? misgivings?
 
The misgiving is that if you short out the bridges at all, it'll cause problems. Also, direct die cooling doesn't really have any advantage over traditional block cooling, with the added risk of getting your chip wet.
 
the layer of epoxy will provide the barrier between the water and the bridges. I prefer to be the one not going with mainstream trends. There are lots of peeps out there doing block WCing. If I can get this to work on a P.O.S. old duron, it should work for other procs too...and hopefully get some interest in it in the [H]ard community, and not the [M]ainstream community.
 
It isint as good as it sounds. Its a huge mess and I've heard a white water gets better temps.
If you want to be different order some copper up and make your own block.
 
Its old news. Its been done, and from what I read didn't work very well. The article I read they used a sunny D bottle. Try a direct die phase change system.
 
A direct die phase change project will cost quite alot and take alot of time working and reading if your clueless and toolless. You need:
compressor
evap
condensor
tubing
cap tube
flex suction line
filter/drier
torch with MAPP gas
brazing rods and flux
pipe bender
pipe cutter
vaccume pump or compressor to mod to one
flaring tool
access valves
gauges
refrigerent

Totals to like 700 dollars or more...
 
I never completely understood how a phase change system works really (the die changes water to vapor..but that'd require the die to get too hot and wear away the die)...and that's not what I'm after. Just wanting to run water over the die/slug, which should be very cheap.
Alias - what exactly didn't work well?
 
Direct die cooling didn't work so well.

Who says you use water in a phase change system? You use something like freeon (or substitute)
 
Your idea has merit, and would be worth trying. A few things to keep in mind:
  • Your pump dies, you are instantly, and irrevocably, done.
  • Make sure that the flow from the pump is aimed directly at the top of the die, in order to maximize turbulence, and therefore heat transfer.
  • Pick a pump with high flow rate. It will be very critical to maintain good flow to the die, since you have to move the same amount of heat with a small fraction of the surface area of a conventional waterblock.
 
Instead of water or liquid, why couldn't you seal it with the epoxy and run some sort of refrigerant over the top of the die? It seems like running a gas over the die wouldn't harm it.
 
Originally posted by z3r0-
Um with -40C refrigerent on the CPU it wouldent last long...
Click to expand...

Why not? Does silicone have properties that won't allow it to?
 
Originally posted by md262626
Instead of water or liquid, why couldn't you seal it with the epoxy and run some sort of refrigerant over the top of the die? It seems like running a gas over the die wouldn't harm it.
Click to expand...

Refrigerant is used in phase change systems, a water cooling loop is not a phase change system. Two completely different things here.
 
Originally posted by zer0signal667
Refrigerant is used in phase change systems, a water cooling loop is not a phase change system. Two completely different things here.
Click to expand...

I understand that. But what would happen if you ran a refrigerated gas over the top of a sealed processor core vs. water thru a waterblock over a core? Have you ever tried? If not, how would you know? Got to think outside of the box sometimes.............
 
Its impossible. The chip would break. Anyway how would you get the block to stay on? I hope you understand the pressures involved in phasechange...
 
Originally posted by Fark_Maniac
I never completely understood how a phase change system works really (the die changes water to vapor..but that'd require the die to get too hot and wear away the die)...and that's not what I'm after. Just wanting to run water over the die/slug, which should be very cheap.
Alias - what exactly didn't work well?
Click to expand...

Keep in mind the new organic packaged chips are porus.

http://www.spodesabode.com/content/article/directdie2
http://www.sysopt.com/articles/Direct_Die/index6.html
 
Originally posted by md262626
I understand that. But what would happen if you ran a refrigerated gas over the top of a sealed processor core vs. water thru a waterblock over a core? Have you ever tried? If not, how would you know? Got to think outside of the box sometimes.............
Click to expand...

You could run a refrigerant over the core, but not in a traditional loop, which is what it seemed like you were suggesting. Do you know how phase change cooling works? The refrigerant has to be compressed into liquid form and when it is heated and allowed to change volume, that's when the "cooling" occurs. You're not just flowing refrigerant through some tubing and a chamber.
And there's a difference between thinking outside of the box and thinking outside of reason.
 
Originally posted by zer0signal667
You could run a refrigerant over the core, but not in a traditional loop, which is what it seemed like you were suggesting. Do you know how phase change cooling works? The refrigerant has to be compressed into liquid form and when it is heated and allowed to change volume, that's when the "cooling" occurs. You're not just flowing refrigerant through some tubing and a chamber.
And there's a difference between thinking outside of the box and thinking outside of reason.
Click to expand...

Glad Thomas Edison,, Marconi and the rest didn't think like you:D
 
Originally posted by Fark_Maniac
Thanks for the links...I'll have to look into them a bit further.
The older socket A procs are porus as well as they are ceramic based. That's why I'll be putting a small layer of epoxy over the entire top of the proc...minus the die.
Click to expand...

Really? I didn't think they were. In the end of one of those he shows the core, boy was it ugly! I guess there was't alot of problems. Oh well the sunny d one is 2 years old.
 
You must log in or register to reply here.
Back
Top