• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Whats the order of a water cooling setup?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

glass19

Registered
Joined
Sep 9, 2004
Most people I see have it going:

pump>block>radiator>reseviour>back to pump


I was thinking of

Pump>block>resviour>radiator>back to pump.....

What is the best order...and WHY!?

i never really quite understood the order
 
For the most part, I don't think it matters too much. It would be just as important to keep your tubing short as anything else. If you havn't already puchased the parts yet, you can get away without even using a reseviour, though I have never used a T-line myself.

As for why alot of people put the resivour right before the pump, I think it has to do with getting unrestricted flow to the pump. I also see that alot of people put the rad after the pump, but in the end, it doesn't really matter too much.
 
----Myth Busted---------

As a practical matter, the order of components does not matter. The temp of the water varies less than 1C as it travels through the loop.

The temp of the water is usually 5-10C (or even 15-20C for non-performance systems) above the air tempurate at the radiator. Higher performance water rigs have lower water-air temp deltas and even less water temp variation through the loop.

You want to keep the tubing runs short (as practical) and strive for gradual bends in the tubing.
 
the only thing that matters is that you have the resevoir at the suction of your pump. if its anywhere else you will have MAJOR headloss.
 
ya it really doesnt matter which order they do in. the temp difference of the water better the blocks is so minimal you will see NO performance difference. i HATE when people are like put the cpu first because cold water comes in and hot water comes out and you dont want hot water going to your cpu. :mad:
 
The_Dark_Hacker said:
ya it really doesnt matter which order they do in. the temp difference of the water better the blocks is so minimal you will see NO performance difference. i HATE when people are like put the cpu first because cold water comes in and hot water comes out and you dont want hot water going to your cpu. :mad:

i would still recommend the cpu first. the most demanding waterblock will be the cpu, most require high pressures to perform.

above all, it depends on how it will fit into your case.
 
This is good to know :)

I was starting to get worried about what order I should run mine in when, but now I'll just run it whichever way will be shortest :D
 
SomaGaze said:
i would still recommend the cpu first. the most demanding waterblock will be the cpu, most require high pressures to perform.

That isn't how it works.

For a given loop of components, the pressure drop in the cpu block will be the same no matter where it is in the loop. Route the tubing for the easiest bends and shortest overall length used.
 
While we are talking about this does does it matter if the pump is above or below the res? Does it have to be lower than the res?
 
My order is

Pump>Block>Radiator>Resviour>back to pump....

I may take the Resviour out of the loop but I am undecided. The resion I go to the Block directly is because the WW block has been shown to do best with higher pressure because of its design. If I didnt' have a WW I would have gone with

Pump>Radiator>block>Resvior>back to pump
 
Last edited:
Mine is going to go:

pump>rad>res>block>back to pump

I feel it would work better if I did it this way though:

pump>res>rad>block>back to pump

Because then the water coming out of the rad right before the cpu block would be coolest rather then if it came out of the rad, and then went somewhere else, then went to to block... Doesnt it make more sence to have the rad directly before any blocks?
 
gungeek said:
That isn't how it works.

For a given loop of components, the pressure drop in the cpu block will be the same no matter where it is in the loop. Route the tubing for the easiest bends and shortest overall length used.
true, but the earlier you have it in the loop, the higher the initial pressure will be
 
SomaGaze said:
true, but the earlier you have it in the loop, the higher the initial pressure will be
FALSE

The pressure drop across each component will be the same for that component no matter where it is in the loop. A few seconds after the pump has been turned on, the flowrate will be steady through the loop. If the flowrate is steady, then the pressure drop accross each block, rad, etc. is steady. The order of the loop does NOT matter.

If you are going to continue to insist otherwise, then proof would be nice.
 
JPetrillo said:
Because then the water coming out of the rad right before the cpu block would be coolest rather then if it came out of the rad, and then went somewhere else, then went to to block... Doesnt it make more sence to have the rad directly before any blocks?

Unless you are measuring component and water temp down to 0.1C resolution, you will not see any difference with loop order. The water temp varies less than 0.5C in a good water system.

What you may not understand is that the radiator power dissipation is based on the temp difference (delta) between air at the radiator and the water temp. When you first turn on the computer, the water temp will rise until the radiator can dissipate the heat at whatever the air and water flow rates are in your system. Using a small rad means the temp delta has to be higher than if using a large rad. Doubling the temp delta doubles the power dissipation.
 
Just to butt in for one second... The water in the loop is coldest the second it comes out of the rad, right? So wouldnt it make sence for any blocks to be directly after the rad? Because the longer the water is out of the rad the warmer it will get, no? Meaning if it went pump>rad>res>block it would come out of the rad, start warming up from the rad to the res, warm up even more in the res, and then keep getting more warm from the res to the block... I would think you want the rad to be the last pass before any blocks... or does it not make that much of a difference?

Edit: Oops sorry, we posted at the same time heh... so nvm that^ Question answered.
 
SomaGaze said:
true, but the earlier you have it in the loop, the higher the initial pressure will be

I would agree with you in a system with a resivoir, but if you have a have a closed loop without a res then I think the pressure would be equal throughout the whole loop.
 
gungeek said:
FALSE

The pressure drop across each component will be the same for that component no matter where it is in the loop. A few seconds after the pump has been turned on, the flowrate will be steady through the loop. If the flowrate is steady, then the pressure drop accross each block, rad, etc. is steady. The order of the loop does NOT matter.

If you are going to continue to insist otherwise, then proof would be nice.

first of all, do you understand what i am saying? the INITIAL pressure will be higher for the first water block in the loop, subsequent the water block's inital pressure will only be as high as the FINAL pressure from the outlet of the first waterblock (unless you have another pump to raise head)

lets get away from using the word pressure. instead, i'll use a more accurate term: head. each component in the loop will generate head loss (a term used to refer to the pressure loss of a fluid flowing through a component or tubing). each water block, radiator, and length of tubing will generate headloss. where the energy of the fluid (pressure) is lost due to friction...therefore pressure goes down. further along the path through a system you go, the lower the pressure will be.

so, to increase head, we have a pump (which causes flow through the system). the discharge of the pump will have the highest head (pressure), so, the first water block after the pump will perform the best (assuming all waterblocks are the same).

now about flow: flowrate will be the same througout the loop. i have never said to the contrary...so i don't know why you are arguing about that. flowrate and pressure a VERY different, and should not be intertwined.

oh...about the proof. anything here or here will answer your questions.
 
Void425 said:
I would agree with you in a system with a resivoir, but if you have a have a closed loop without a res then I think the pressure would be equal throughout the whole loop.

not true. only STATIC (no flow) CLOSED (not open to atm) systems will have equal pressures at any point in the system.

if pressure was equal throughout the loop, how would flow occur?
 
At steady state conditions (the only thing that counts here), the total pressure output of the pump will equal the sum of the pressure drops along each component in the system.

Think of it this way, you have 5 lightbulb sockets in series and you have five equal bulbs. Each bulb is going to be the same brightness when powered by a voltage. If what you were saying is true, then the first bulb would be the brightest on down to the last bulb being the dimmest. The first bulb sees the full voltage so it should be the brightest. right?? wrong. The sum of the voltage drops across each bulb is the total voltage (pressure head?) applied to the system. It works the same way with fluid systems; pressure = voltage, flow = current, pressure drop (head loss) = resistance.
 
gungeek said:
At steady state conditions (the only thing that counts here), the total pressure output of the pump will equal the sum of the pressure drops along each component in the system.

Think of it this way, you have 5 lightbulb sockets in series and you have five equal bulbs. Each bulb is going to be the same brightness when powered by a voltage. If what you were saying is true, then the first bulb would be the brightest on down to the last bulb being the dimmest. The first bulb sees the full voltage so it should be the brightest. right?? wrong. The sum of the voltage drops across each bulb is the total voltage (pressure head?) applied to the system. It works the same way with fluid systems; pressure = voltage, flow = current, pressure drop (head loss) = resistance.
Nothing to say but, "yep" ;)

Plumb it the way thats EASIEST for you. Simple as that.
 
Back