DIY watercooling ( here in the philippines )

[irangejuice]

---

3.the tubing is too big for the configuration and its making the water circulate fast and not giving it enough time to cool off before a full circle.
4.the pump is too strong.added the push of a strong pump with a large ID will make the water circulate faster before it gets cooled by the radiator.(plus no resevior to let the water sit for an added equation to further cool off the water before it does a full circle.

I agree with some of the points you pointed out. I know there are several concepts in the net about 3 and 4. However, and I'm talking physics here, does the circulation speed really matter that much? I'm having a difficult time trying to understand the relevance of too fast a circulation.

Doesn't a faster circulation translate to more surface area exposure between the water, the blocks, and the radiator? It's the very same principle behind stirring the ice around in a glass of warm water. Stirring helps the surface area exposure between the ice and the water. Not doing so, will result in a slower overall cooling of the liquid in the glass.

As with watercooling, the goal is to transfer heat to the water as fast as you can while the radiator takes heat from the water as fast as it can--all of which benefit from increased surface area for maximum heat transfer.

[DD]

your concept is true but the relevance of fast water transfer to water circulation is the fact that the water is not getting enoughstop over between each device to fully cool down.the radiator thats suppose to cool down the water isnt staying long enough in the radiator to be properly cooled off however,a slower pump will produce slower circulation giving the water to cool off as it pass the radiator and by the time it gets back to the CPU block it will be fully cooled off.
water that goes thru the CPU block will come out warmer and if not given enough time to cool off,itll still be warm by the time it gets back to the CPU block.the same water will be alot warmer as it goes in the radiator and since the pressure is still the same,the radiator will now have to cool down a much warmer water and itll just keeps doubling up as it circulate.

[EarlZ]

@DD

theres a possibility also that he is using the wrong type of blocks for his system as you mentioned, some blocks are design for slow flow/pressure while others are designed for the strongest and fastest pumps around.. same goes with the radiator.. but the Thermochill PA series has changed all that, as for the waterblock swiftech likes high pressure dunno about his GPU block.

with the current designs of watercooling blocks and radiators i think they are now designed for faster overall flow to get better temps.

also from what ive researched single radiator solutions are designed for a single target cooling, adding block on the loop will put to much taxing on the radiator but it can still function better than aircooling, 120.2 is the real way to go when cooling 2~3 GPU/CPU on the entire loop. it gives the real water cooling performance.

regarding his videocard temps,ive read somewhere that a 9700Pro consumes an equal power compared to a 7900GT which should generate similar heat levels.. with a zalman vf900cu the 7900gt can idle at 46c and load at 56c

[irangejuice]

@weed: dude, i'm officially gonna start the house plumbing experiment..well if you wanna try it..

[irangejuice]

anyone got a waterblock they wanna sell for cheap? all in the name of science?

[blade4638]

water that goes thru the CPU block will come out warmer and if not given enough time to cool off,itll still be warm by the time it gets back to the CPU block.the same water will be alot warmer as it goes in the radiator and since the pressure is still the same,the radiator will now have to cool down a much warmer water and itll just keeps doubling up as it circulate.

this is actually why i think my gpu block is still on the warm side... water from the cpu block isn't cooled before it enters the gpu block (around 20cm of tubing in between). add to that a small gpu block, which upon opening i found out that it doesn't have any pins/fins, just a square block of copper.

looks like you nailed it guys c",)

one thing though... about flow rate. there's this discussion in the procooling forums which goes like this...
2 setups with everything the same except flow rate. lets say 1 setup has 2x the flow rate of the other. given that the same distance is travelled by the coolant, flow rate shoudn't matter. why? say the setup wit the slower flow rate coolant passes thru the block once per minute. for the setup with the faster flow rate (2x the slower), coolant passes through the loop twice in 1 minute. the supposition is that the same amount of heat is picked up by the coolant from the block, therefore flow rate isn't a major factor. the question is how efficient are the blocks in transferring heat, both from heat source to itself and from itself to the coolant. this is where block designs play a very important role (afaik, the best designs i know are cathar's blocks that are based on water impingement)

for my setup, i think there are 2 major factors contributing to the "high" gpu temps:
1. heated water goes through the gpu block from the cpu block (pointed out by DD)
2. transfer of heat from the gpu block to the coolant is not as effective as what i'd hoped for

i tossed out my res in favor of a t-line because being externally located, it took up some real estate (i could've
fastened it to my casing but i was too busy (i.e. lazy hehehe!).

i also tried a rad-less setup, but my res was huge --> a 5-gallon mineral water container. temps were a bit high for my taste... and the setup was too bulky.

@coolguy: some insights from you would be appreciated c",) c",)

[EarlZ]

Mineral water? those things like mess ur rads up

[blade4638]

@EarlZ: mineral water containers bro... for a res (the unused ones sold at makro) c",) mineral water would cause white deposits (calcification??) to appear over time... something we would be quite happy without c",)

[irangejuice]

@blade: bro, still got a spare cpu block to sell?

[blade4638]

@Zerone_null: very sorry bro still in use... i've already disposed all my hsfs so i'm afraid i'm stuck with all this wc stuff.

problem with our local stores is that most of them don't have these parts on stock (that is, if they even sell it). if only the parts were available and at reasonable prices.