let the madness begin...

Why decrease the efficiency with a water cooling loop?

Why not just remove the heat collector and replace it with a block mounted directly to the CPU? This is what phase-change systems do after all.

You are adding multiple layers of complexity as well as hazards to the design that diminish the end result and efficiency (a water pump adds all of its wattage as heat to the line).
 
undertheradar said:
Why decrease the efficiency with a water cooling loop?

Why not just remove the heat collector and replace it with a block mounted directly to the CPU? This is what phase-change systems do after all.

You are adding multiple layers of complexity as well as hazards to the design that diminish the end result and efficiency (a water pump adds all of its wattage as heat to the line).
Click to expand...

the 'heat collector' as you called it it called an evaporator

simple setup, no need to break open the a/c unit

lol, the water pump heat doesnt matter in this case

I think it worked out pretty well actually
 
undertheradar said:
Why decrease the efficiency with a water cooling loop?

Why not just remove the heat collector and replace it with a block mounted directly to the CPU? This is what phase-change systems do after all.

You are adding multiple layers of complexity as well as hazards to the design that diminish the end result and efficiency (a water pump adds all of its wattage as heat to the line).
Click to expand...

it is actually a very simple setup

final temps where great, Phenom II 965 c3 @ 4.6GHz

idle temps measured at around -25c, load temps 1c or 2c using LinX

as for the way it is setup, it took less than 2 hours to build, take an old window A/C apart and pull the evaporator away from the unit, place the evaporator into a large cooler filled with a fluid that can handle the temps(i.e. well below freezing)

then attach the coolant tubes and basically just hook up an otherwise normal water cooling system

so while this does not perform quite as well as a full blown phase change unit, it is cheap and simple, and WAY outperforms anything you can do with water cooling
 
I agree that is pretty awesome. I use a Boreas TEC from time to time and it won't quiet match that lol. You can really crank on those video cards at those temps...:) :)
 
EricFX1984 said:
the 'heat collector' as you called it it called an evaporator

simple setup, no need to break open the a/c unit

lol, the water pump heat doesnt matter in this case

I think it worked out pretty well actually
Click to expand...

It prevents you from reaching the lowest possible cooling temps, and adds wattage that is otherwise wasted.

"heat collector" vs. "evaporator"... do you note when someone calls a "cold plate" a "water block"? Lol.
 
EricFX1984 said:
it is actually a very simple setup

final temps where great, Phenom II 965 c3 @ 4.6GHz

idle temps measured at around -25c, load temps 1c or 2c using LinX

as for the way it is setup, it took less than 2 hours to build, take an old window A/C apart and pull the evaporator away from the unit, place the evaporator into a large cooler filled with a fluid that can handle the temps(i.e. well below freezing)

then attach the coolant tubes and basically just hook up an otherwise normal water cooling system

so while this does not perform quite as well as a full blown phase change unit, it is cheap and simple, and WAY outperforms anything you can do with water cooling
Click to expand...

The downside of smaller nm lithography is starting to show... as the chips get smaller and smaller in their process, overclocking headroom is reduced. Higher currents mean little when the capacitance in the transistors is already so low. Perhaps the overclocker community could get Intel or AMD to start a line of chips (based on existing architecture) with a process designed to require phase-change cooling to operate: otherwise as we get smaller and smaller overclocking will be meaningless... the chips themselves will have to start being designed for it (which will also make them impossible to use with normal cooling as well).
 
undertheradar said:
It prevents you from reaching the lowest possible cooling temps, and adds wattage that is otherwise wasted.

"heat collector" vs. "evaporator"... do you note when someone calls a "cold plate" a "water block"? Lol.
Click to expand...

not really you are not going to do much colder on a single stage phase cooler
 
EricFX1984 said:
not really you are not going to do much colder on a single stage phase cooler
Click to expand...

If that is a 20 watt pump, its like trying to cool a processor that is generating an additional 20 watts of heat... it undermines your potential greatly... could have been -30C otherwise.

Not that it matters... all that super cooling means less and less as the mfg. process gets smaller and smaller.
 
undertheradar said:
If that is a 20 watt pump, its like trying to cool a processor that is generating an additional 20 watts of heat... it undermines your potential greatly... could have been -30C otherwise.

Not that it matters... all that super cooling means less and less as the mfg. process gets smaller and smaller.
Click to expand...

Can you cite some of the particular cpu's you are referring to?
 
undertheradar said:
If that is a 20 watt pump, its like trying to cool a processor that is generating an additional 20 watts of heat... it undermines your potential greatly... could have been -30C otherwise.

Not that it matters... all that super cooling means less and less as the mfg. process gets smaller and smaller.
Click to expand...

lol, my cooling temp with nothing attached is -25c, even with the Phenom and pump it is still -25c at full load...

it is such a low heat load, not to mention the heatload of the pump is next to nothing when compared to the CPU... in short you are talking out of your bum... ;)
 
Very nice work.

What did you use for coolant?
And what were your final overclock results?

I actually have the same block and have been tossing around the idea of building a water chiller...
 
LOL when I read through I saw P2 and assumed pentium 2...
Phenom 2 that makes more sense!
 
Hi, I know this is an older thread and maybe you are not using this cooling anymore. I had an Idea on the thermal paste problem. What if you ran at lower clocks/voltages on standard water cooling for a few days with it running to cure the thermal compound some. Its quite possible that the thermal compound needs some time a higher temperature to set up properly. If you have already solved that problem then by all means do share how you did it!
 
i run a chiller at -10f on my 955be and i have never had a problem whit my paste, i used Shin-Etsu and right now my cpu cores are reading 0C and my cpu temp is reading 3C running folding at home @ 4.35ghz 1.675vcore and 3 ghz on the nb, If the op ever sees this what vcore did you run to hit 4.6? AS5 is just a really bad paste for running cold theirs much better pastes that will run in -f all day long
 
I was just wondering if it needed to set up at normal temperatures before going cold. I was thinking that once it was set up it would be able to go sub zero. What pastes work well when it gets chilly?
 
You must log in or register to reply here.
Back
Top