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

How to setup watercooling??

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

Elluzion

Member
Joined
Jul 3, 2007
Location
[NorCal]
Ok well once I have all of the parts, I need to know how the basic setup will go.

If anyone knows a post where this is already answered just link me to it. I looked around a bit but couldn't find much.

1. What order should the loop go in for best cooling

RAD => CPU BLOCK => RESERVIOR => PUMP => RAD (<good setup?>)

2. How do I know how much water to use in my loop

3. How to bleed/ how long does this take

4. How long to do testing to make sure it works

I will add any additional questions later.

Thank you!

just wnt to make sure I have all the info to do it right. Also if their is a website that shows this or anything like that it would be great. Thanks a lot

also, how are these fans for my Tripple rad? http://petrastechshop.com/12yalod1blle.html OR http://jab-tech.com/YATE-LOON-120mm-Clear-Case-Fan-with-four-blue-leds-D12SL-12-pr-3625.html

E LL U Z IO N
 
Last edited:
1. That way tends to be best, but it's not worth going out of your way and adding a ton of extra tubing if there's a way to do it with less. The extra lines may end up hurting your waterflow.

2. You don't need to know, just fill it until it's full...and keep it topped up until it's bled thoroughly.

3. Just start by running the WC system without the system on, it will pick up most of the bubbles itself. Once it seems to have cleared out the bubbles, gently tilt your case in different directions. To completely bleed the system it will probably take a few days if you're using a tline, probably less than an hour if you're using a reservoir. Just repeat the process of letting it run for a while, then tilting until it stops pulling air out of corners.

4. People usually say at least 24 hours. If it's your first time you should probably go that long, but once you've done it a few times you learn to feel a comfort with how your setup is and you can test for shorter times, focusing on certain points. Make sure you're feeling any possible location for a leak.
 
Ok well once I have all of the parts, I need to know how the basic setup will go.

If anyone knows a post where this is already answered just link me to it. I looked around a bit but couldn't find much.

1. What order should the loop go in for best cooling

RAD => CPU BLOCK => RESERVIOR => PUMP => RAD (<good setup?>)

2. How do I know how much water to use in my loop

3. How to bleed/ how long does this take

4. How long to do testing to make sure it works

I will add any additional questions later.

Thank you!

just wnt to make sure I have all the info to do it right. Also if their is a website that shows this or anything like that it would be great. Thanks a lot

also, how are these fans for my Tripple rad? http://petrastechshop.com/12yalod1blle.html OR http://jab-tech.com/YATE-LOON-120mm-Clear-Case-Fan-with-four-blue-leds-D12SL-12-pr-3625.html

E LL U Z IO N

Usually the leaking is because you didn't tighten something enough. I tested mine for about 5 hours, and all was fine.

Also, I have 3 of the Yate Loon silent 120's. They work awesome and are OEM so you get them for a fraction of the price. Good luck!!
 
1. That way tends to be best, but it's not worth going out of your way and adding a ton of extra tubing if there's a way to do it with less. The extra lines may end up hurting your waterflow.

2. You don't need to know, just fill it until it's full...and keep it topped up until it's bled thoroughly.

3. Just start by running the WC system without the system on, it will pick up most of the bubbles itself. Once it seems to have cleared out the bubbles, gently tilt your case in different directions. To completely bleed the system it will probably take a few days if you're using a tline, probably less than an hour if you're using a reservoir. Just repeat the process of letting it run for a while, then tilting until it stops pulling air out of corners.

4. People usually say at least 24 hours. If it's your first time you should probably go that long, but once you've done it a few times you learn to feel a comfort with how your setup is and you can test for shorter times, focusing on certain points. Make sure you're feeling any possible location for a leak.

Thanks for the info...

anything else??? I think im gnna go with a reservoir...


should I clean my radiator before I hook it up to my watercooling, or just use it as it comes???
 
any other info that could help me out for a first timer. or any problems you experienced when first setting up your stuff??
 
I like to use white vinegar in the rad for at least 15-20 min, rinse well with tap water, then flush with distilled water before it dries. Except for the vinegar, I like to do that with every component before I put it into a loop. No telling what got in there during processing, packing, and shipping. I found a small piece of foam in a block once and a piece of something black in a rad once. You can't be too careful ... ;)
 
To me it makes more sense to put:

pump > CPU block > rad > res >pump

That way you have maximum pressure on the CPU block.

The way you have it would mean there is a large pressure drop by the time the water gets to the block. Well in my head it seems so anyway...
 
Last edited:
The way you have it setup is best IMO. That way the coolest possible water is going to your CPU. Unless you are going to use a LOT more tubing, it's not a big deal to use a little extra to set it up that way.
 
pump > CPU block > rad > res >pump
That way you have maximum pressure on the CPU block.
The way you have it would mean there is a large pressure drop by the time the water gets to the block.
Not to jump on you, but....
Pressure has no effect on cooling. One PSI inside the block or 20 PSI in the block will cool the same at the same flow rate. Yes, each component adds it's pressure drop to the loop, but the pressure difference between the pump inlet and pump outlet determines the flow rate. And assuming the same tubing length, the order of components in the loop has no effet on flow rate.
 
Not to jump on you, but....
Pressure has no effect on cooling. One PSI inside the block or 20 PSI in the block will cool the same at the same flow rate. Yes, each component adds it's pressure drop to the loop, but the pressure difference between the pump inlet and pump outlet determines the flow rate. And assuming the same tubing length, the order of components in the loop has no effet on flow rate.

But I thought that the whole idea of flow nozzles and stuff like that is to get maximum pressure on the CPU block.

Sorry if I have the wrong idea, I have never even seen a watercooling kit haha!
 
But I thought that the whole idea of flow nozzles and stuff like that is to get maximum pressure on the CPU block.

Sorry if I have the wrong idea, I have never even seen a watercooling kit haha!
Oh boy! No, a nozzle in a water block is just a smaller diameter tube that the water has to flow through. This causes the water to flow faster through the smaller tube and yes, there is some "pressure loss", it's called restriction. In water blocks like the Storm, the RBX and TDX this higher water velocity hitting the hot (well, warm) part of the block increaces the efficiency of the heat transfer. Heat transfer only occurs in a very thm layer of water that is closest to the warm part of the block. So a "nozzle" makes more water flow in this thin layer
 

Attachments

  • Fast Slow 2.jpg
    Fast Slow 2.jpg
    19.1 KB · Views: 101
Oh boy! No, a nozzle in a water block is just a smaller diameter tube that the water has to flow through. This causes the water to flow faster through the smaller tube and yes, there is some "pressure loss", it's called restriction. In water blocks like the Storm, the RBX and TDX this higher water velocity hitting the hot (well, warm) part of the block increaces the efficiency of the heat transfer. Heat transfer only occurs in a very thm layer of water that is closest to the warm part of the block. So a "nozzle" makes more water flow in this thin layer

Thanks bill!

I get it now :beer:
 
Oh boy! No, a nozzle in a water block is just a smaller diameter tube that the water has to flow through. This causes the water to flow faster through the smaller tube and yes, there is some "pressure loss", it's called restriction. In water blocks like the Storm, the RBX and TDX this higher water velocity hitting the hot (well, warm) part of the block increaces the efficiency of the heat transfer. Heat transfer only occurs in a very thm layer of water that is closest to the warm part of the block. So a "nozzle" makes more water flow in this thin layer
I even noticed this behavior when I went from 3/8" tubing to 1/2" tubing on an older Apogee. The 3/8" tubing did a slightly better job of cooling with the Apogee then the 1/2" (~0.5°C but consistent over six tests). Since the 3/8" loop had a higher water velocity than the 1/2" loop I theorized the 3/8" barbs and tubing, being smaller, created more of a "jet" effect inside the block.

As a side note the Storm cooled better (~1°C) with 1/2" tubing and the subsequent increase of ~0.2 GPM (~20%) to the loop ...
 
Last edited:
You changed something other that tubing?
My bad - post edited. *redface* That's what I get for posting just after I wake up - that second cup of coffee makes a BIG difference. ;)


No other changes except tubing and removing two reducers (1/2" to 3/8", one up & one down) from the loop. I was using an MCP655 on both P5 and P3 - I never did go back and try the stock MCP355 with 1/2" tubing. In hindsight that was an error but I had so little (over-all) change with the MCP655 on both settings I decided the other check was useless. To me it was obvious the Apogee benefited from 3/8" and the Storm suffered from it. Since most of my loops will be cooling lidless K8s and using Storms I had my answer. The only Apogees I have will be used with older machines and 3/8" tubing, since they really don't need the extra muscle of a Storm ...
 
Last edited:
Back