Water cooling questions

Lube Oil is a poor idea, it's specific heat capacity and conductance are far lower than water. Further, it has a capacitive effect if it leaks onto components and gets between pins. Finally, dedicated liquid cooling pumps are designed for water, not oil.

As far as milling your own waterblock, there are a number of considerations. First, the size. The bigger it is the more potential water/metal interface area you will have, but at the same time it must fit the motherboard mounting area. Impingent flow is the most efficient, but probably impossible to mill yourself. Take a look at Koolance and koolance.com, they have the manufacturing drawings of some of their waterblocks on the site.

Sure, you could use ethanol as a coolant, but again, it would be far less efficient than water.

Basically, use water. You can purchase water specifically for liquid cooling systems on newegg or similar sites. Basically it's just water/glycol with some anti-foaming additives and microbial growth suppressants. It is really all you need.

I was refering to lub oil LEAK. I am worrying that any leaked oil would form a coat on the cooling fins, thus cooling performance would collapse.

That is a valid concern.
I also advise you to buy instead of mill a water block, but if you insist on it; take a look at the swiftech apogee. It is one of the top performing blocks and I believe it would be one of the easier blocks to duplicate. For a liquid check out MCT-5 and MCT-40, they are a non-conductive liquid.

Oil does lubricate but it does not transfer heat much better than water. Ethanol/alcohol is simply not usable at all don't even try it. It will only damage the components like the pump and radiator due to it drying/corrosive chemical properties. Not to mention flammable. Water is the most common coolant for pc water cooling. Specifically 90-95% pure water (deionized water or bottled water) and 5-10% car anti-freeze or glycol for anti bacterial and anti algae growth. Lubricating oil may not short the system in terms of leaking out to the sensitive components but like I said will not result in lower cpu temp like water, probably the same as high-performance air cooler like Tuniq and Ultra. Also oil is thicker/denser than water so it will have a hard time circulating in the loop giving the pump a hard time and probably will be turned up high just to have a good flow resulting in more noise.

If you want the best safe coolant then look for the non-electrical conduction coolants. It the same coolant I used for my system and it's almost completely safe. I have used Innovatek Protect 1, PrimoChill ICE and Fluid XP+ Extreme on my liquid cooling rig. And yes I have tested it with my old pc, by pouring some into the video and network card and not shorted out. I 'm not saying you should do this on your system and no not definitely on the cpu and cpu socket.

For the design, :? , I'm not a machinist to begin with but copper is the best duh! And I think you can copy them designs from water block manufacturers, right? As long as you're not selling it for your own profit. That would be nice if I have the tools to machine my own parts and I'll make some simple water blocks.

Don't forget to post some pics with that custom water block.

De-ionized water is also non-conductive, at least at first. Water itself is an insulator, what you have to worry about is the crap that it picks up.

Ethanol can neither remove nor absorb heat faster than water. You're right though, it eats plastics and rubber very quickly. Further, it absorbs water. That's why cars have to be specially modified to run ethanol, the fuel tank has to be stainless to prevent rust, and the fuel lines need to be teflon to not get eaten.

Pumps made for liquid cooling don't need lube. Sealed bearings.

You're contradicting yourself mate, you said yourself that water has a higher heat capacity. It also has higher heat conductivity. Therefore it is better that ethanol, less likely to get saturated, and so on.

It's a moot point though, you can't use ethanol in a coolant loop without some expensive modifications and a special pump.

Of course they're different! Mmmm, how to explain what I'm trying to say.... If the heat conductivity of water is higher, then the thermal resistance of ethanol is higher. So it's less likely to accept that excited particle you mentioned earlier.

You weren't wrong at the beginning, it CAN absorb and dissipate heat quickly, just not as quickly as water.

Like everyone says, it will cost more to make your own waterblock unless you have a CNC mill in your garage. Typical machine rates of $150/hr with a 2 hour min means your waterblock will cost $300 plus CNC setup time. You also have a lot more than a waterblock to cut. You will need a top plate, inlets, etc. If you have access to a vertical CNC mill, a high quality water block design would be a cinch to write and the copper would cut like butter. You're only talking about maybe a meter of total cutting distance, so once your mill was set up you could run through these at about 30/hr. Making a top plate that will seal well is the hard part, unless you also happen to have a precision 10-ton stamp in your garage. Going to an in-cut design with a flat top plate would work, but you would probably double or triple machine time, and you would be going from a 2d cut to a 3d cut. Then you get to your port design. You could weld ports onto a drilled top plate, or you could turn ports if you had a lathe but typical designs would be hard to set up on a CNC lathe without custom tooling.

Like everyone says, water is the best thermal fluid, just make sure you use stabilized coolant or dissimilar metal components will not play well together.

A minor oil leak will not cause a problem if you're only talking about a few ml. If you want you can use a detergent solution in your water to scavenge any leaking oil. Glycol blends naturally have some detergent qualities.

Ethanol or any other working fluid would probably be fine. There are technical challenges with pumps, gaskets, etc. and they don't have the same thermal qualities, but they don't aid corrosion like water does. You're only talking maybe 100-200 watts of heat to dissipate so almost any liquid would work.

If you use water, use a coolant blend so your don't corrode the various metal parts of you system (radiator, coolant block, pump parts, etc.).

You bring up a very good point. Using ethanol for a phase change cooling loop is quite possible, but it comes with its own little bag of worms. Vapor lock would be the worst. The efficiency of the condensor would have to be quite high to prevent this. And the challenge of depressurizing the system would be pretty darn interesting too.

I can't wait to see OCZ's phase change unit hit the market.