Copper Plate w/ Ultra 120

[enaberif]

If you had a chunk of copper plate say about 4" square and about 1/4" thick do you think it could or would do any good to put it between a cpu and a Ultra 120?

I know copper is good for pulling heat so I was thinking by putting this plate between the cpu and cooler that you'd give it a bit more to be able to absorb heat or would the plate stop the Ultra 120 from cooling properly?

[CMetaphor]

In general, I'd say the more heat transfers you have to do the less efficient the system becomes. For your solution you're talking Cpu-> T.Paste -> Copper Block -> T.Paste -> Ultra 120... that many heat transfers would probably work worse than just Cpu -> T.Paste -> ultra 120. If anything should go between a cpu and your ultra 120 it should be a TEC cooler (peltier) that can more than make up for the loss of thermal conductive efficiency. Wordy, but I fairly certain on this.

[enaberif]

Originally Posted by CMetaphor

In general, I'd say the more heat transfers you have to do the less efficient the system becomes. For your solution you're talking Cpu-> T.Paste -> Copper Block -> T.Paste -> Ultra 120... that many heat transfers would probably work worse than just Cpu -> T.Paste -> ultra 120. If anything should go between a cpu and your ultra 120 it should be a TEC cooler (peltier) that can more than make up for the loss of thermal conductive efficiency. Wordy, but I fairly certain on this.

Yah its what I figured I just saw the copper sitting beside me and got thinking haha.

[MpG]

Originally Posted by CMetaphor

In general, I'd say the more heat transfers you have to do the less efficient the system becomes.

To a point. After all, water cooling involves several instances of thermal transfer, before the heat finally reaches the air.

For the copper plate, I'd give it a thumbs down. With TEC's, the copper plate isn't so much to improve the cooling as it is to let a device with a larger surface area (the TEC) interface with a smaller surface area (the CPU). Since the U120's surface area is basically the same as the processor's IHS, there wouldn't be any gain to be had in this case.

[CMetaphor]

Originally Posted by MpG

To a point. After all, water cooling involves several instances of thermal transfer, before the heat finally reaches the air.

Not quite. If you think of thermal transfers like electrical circuits then what enaberif was planning more of a series-circuit type setup, with heat being exchanged as i showed before Cpu -> paste -> copper -> paste -> ultra 120.

On the other hand, a water cooling setup is more like a parallel circuit in terms of thermal transfers. True the water does flow to every component one in a row but the water has some chance to cool in the tubing before going to the next component and so on. Also, a watercooling setup can be branched in a "Y" system that even more resembles a parallel circuit.
Ex: Water -\---> Cpu --------------> NB -------/> Rad
.................\-->Video card 1 -> Video card 2--/

[MpG]

Heh, thinking about it further, I was about to take back my point. Since water-cooling involves mechanically relocating the heat (via water flow), it's actually not the same thing as what's happening here, which is just conduction from one material to the next.

I'm afraid that you lost me on the electrical example, though.

[3.0charlie]

I had this copper plate (Intel LGA 775 Spacer - Water cooling systems, pc liquid cooling kit, cpu, video card, hard drive) part of my Koolance kit for a long time. It managed to cool my then oc'ed Prescott 4.2 (3.2 stock) to an acceptable 52C.

Never had any issues with it.