Liquid chiller blocks..

so the magnitude of diffence in energy absorbtion is the diffence in the need for more mass in the material above the die?
In a chiller R134a could very well be cooling the Ethylene Glycol (at the evap of the chiller).. this is were I belive I'm getting confused..

The metal is only getting in the way by resisting heat flow. The problem is that an interface median is required. The metal serves as that median but at the cost of loss of performance. The better the heat flow, the closer the coolant will approch the latent heat temperature of the r134a. A phase change unit is designed to put out continous cooling. That means a thermal ballast is not needed.

Say you are overclocked a FX57. (cold bugged) Bazx mentioned that his would run at -105 C for example. If you are pouring in LN2 you want more mass. More mass means more specific heat. This means you can pour more LN2 into the mousepot at a time and take a longer break while keeping the same temperature or you can keep the temperature within a smaller range.

Non cold bugged chips are easy to work with. Just fill the mousepot up and your ready to go. Cold bugged chips will crash if you change the temperature much or get it too low.

Cold plates are a different story. Thicker metal is more resistive through it's thickness but spreads heat better. The balance between heat spread and resistance through it's thickness will yeild best results. If the CPU die was the exact same size as the Processor, that the best case coldplate would be 0. (none) The bigger the evap/pelt to the size the the Processor, the thicker that optimal point is. Thicker is more expensive but too thin and you can't fasten it down properly..

Very informative posts, GigaForce!

yes.. good show GigaForce310

I feel like Im beating this subject to death.. (by asking questions) hopefully its cool..

assuming you have a non cold bugged chip..

and by size diffence from the processor I'm assuming you ment the die of the processor (or the IHS)..

It sounds like you may want faster flow and or less mass to the block according to what your saying..

I guess I'm comming away from this missing something.. when it comes to this particular question (chiller loops, blocks, dynamics like flow), I still feel like somthing should be different in the case of a chiller then a water cooled system..

are they really that much alike that everything could be held constant and once one verible is changed they would perform equaly by a persentage diffence? (this is might be a bad question unless someone has actually tested something like this)

I feel like Im rambling.

There are only slight differences between room temperature water and a chiller. The big difference is that the liquid is cooler therefore must be able to perform under such conditions. Cars operate on a very wide temperature ranges from -40 C up to 100 C. Ethylene glycol is used but it gels at low temperatures and water obviously freezes. A mix between water and ethylene glycol gives a good enough viscosity at very low temperatures and is pressurised to about 1 bar to prevent from boiling at it's peak temperature.

Conditions don't change much between high and low temperatures. The ideals behind water cooling still apply. The higher viscosity will make turbulance harder to come by with the same pumps. If I were to use a chiller, I would use water/denatured alcohol mix. This has a lower viscosity than water/EG mix therefore performs closer to room temperature water. I would use the most amount of alcohol and keep it pretty much non flammable. Then again, I have about 20 gallons of that stuff right now. The water blocks job is very simple, transfer heat from the processor and dump it into the liquid. Higher waterflow, more turbulance, ect. still apply in relitive performace. The difference to the most part is your now approching a much smaller temperature than before. Insulation is required to prevent condensation.

What I'm saying is that lower temperature will require different chemicals as water freezes at 0 C. The design of the ideal waterblock in both cases is pretty much the same. The liquid is what will have the problems. The goal is to find a liquid/mix that performs at -30 C at or better than water at +30 C. Pure ethylene glycol would result the same at -30 as putting jello in at +30, Pure water freezes. Water and ethylene glycol mix is fairly good but not the best. It is however very cheap and easy to get. I guess experimentation is the most viable way to find the best performing liquid in your temperature range. Other considerations such as damage to your system with certain chemicals should be appied also.