When using Dual Radiators

W

WipeoutFTW

Limp Gawd
Joined
Jul 17, 2006
Messages
357
When using dual radiaitors should I put them right next to eachother or between my CPU and GPU block to cool the water before it gets to the GPU?
 
There is lots of arguments with loop order but it doesnt really matter in the end. The water will reach a certain equilibrium throughout the loop and the change of putting a rad in a certain place is so small you'll be lucky if you see a difference +/- 1c. So basically put it wherever it will work best in terms of physical layout.
 
If it reached an equilibrium then the radiator wouldn't be doing a very good job, right? I typically see a 1.5 - 2 degree C drop in temps between the intake and output of my radiator. But to answer your question I don't think it matters too much.
 
oh, another question, would it be efficient to get 2 splitters and have the radiators split into the GPU and CPU and then join back? I think that if both components get freshly cooled water it would be better. What do you guys think?
 
If anything, it would be most efficient to split between the two radiators (i.e. put them in parallel) then go back.

The problem with splitting between CPU and GPU is you lose flow in each branch. Granted, your overall water flow increased due to less restriction, but this is offset by the fact that you are dividing the water in half between the branches. Now, I'd imagine that by using a wye splitter between the blocks you will slow down the flow of water if you use the same size tubing. You want a lot of flow through your blocks, so it would seem counterproductive to do such. This is, however, exactly what you want in your radiators - slower water flow. This way, the water is in contact with the fins longer before the two sections join back together to the main flow line. Note that overall flow (in the main line) doesn't decrease, but slightly increases, thus your system performs even better.

I haven't done this though, so I don't know what kind of performance boost you could expect from putting identical radiators in parallel. If I am incorrect here, please let me know.
 
Top Nurse said:
If it reached an equilibrium then the radiator wouldn't be doing a very good job, right? I typically see a 1.5 - 2 degree C drop in temps between the intake and output of my radiator. But to answer your question I don't think it matters too much.
Click to expand...

Interesting, the temperature difference between my heater cores isn't nearly that high. Then again, I'm guessing I use a higher-flow system than you are :). I think it all comes down to your flow design and your type of heat exchanger. I am not bashing Aqua-Computer parts, by the way - nothing is wrong with using their products. They simply have a different design methodology.
 
DarkenReaper57 said:
Interesting, the temperature difference between my heater cores isn't nearly that high. Then again, I'm guessing I use a higher-flow system than you are :). I think it all comes down to your flow design and your type of heat exchanger. I am not bashing Aqua-Computer parts, by the way - nothing is wrong with using their products. They simply have a different design methodology.
Click to expand...

I think we have a misunderstanding here. He's saying he gets a 1-2C diff between the intake and output of his dual radiator, you're saying you dont see that high a difference between your two single radiators.

Correct?
 
oneil said:
I think we have a misunderstanding here. He's saying he gets a 1-2C diff between the intake and output of his dual radiator, you're saying you dont see that high a difference between your two single radiators.

Correct?
Click to expand...

To clarify, I have a 2 heatercores in series. One is a dual heatercore (240mm) and one is a single (120mm). So, I essentially have a triple heatercore system. Using temperature probes at the inlet of the initial radiator and the outlet of the final radiator, I don't see this high of a difference for normal PC operation - maybe half a degree celsius. Once I set my system up again (it is being revamped) I'll do the same when the CPU is under a stress test. Perhaps that would make a difference.
 
So something like this would be best to max the flow through the blocks and minimize the flow through the radiators? Also which direction would be best for the water to flow?
Code: 
       _____||____ 
      |  Resivoir | 
      \-----||----/
           //\\
          //  \\
         //    \\
 _______//_    _\\_______
| Radiator |  | Radiator |
\----------/  \----------/
         \\    //
          \\  //
           \\//
            ||
       _____||____
      | CPU Block | 
      \-----------/
            ||
       _____||____ 
      | GPU Block | 
      \-----------/
            ||
       _____||____ 
      |    Pump   | 
      \-----------/
            ||
       _____||____ 
      |  Resivoir | 
      \-----||----/
 
DarkenReaper57 said:
Interesting, the temperature difference between my heater cores isn't nearly that high. Then again, I'm guessing I use a higher-flow system than you are :). I think it all comes down to your flow design and your type of heat exchanger. I am not bashing Aqua-Computer parts, by the way - nothing is wrong with using their products. They simply have a different design methodology.
Click to expand...

How many blocks do you have in your loop?
 
I believe at XS they decided in series is best. This way you will get equal amount distributed to both radiators and you will keep your pressure from having a drop.
 
DarkenReaper57 said:
To clarify, I have a 2 heatercores in series. One is a dual heatercore (240mm) and one is a single (120mm). So, I essentially have a triple heatercore system. Using temperature probes at the inlet of the initial radiator and the outlet of the final radiator, I don't see this high of a difference for normal PC operation - maybe half a degree celsius. Once I set my system up again (it is being revamped) I'll do the same when the CPU is under a stress test. Perhaps that would make a difference.
Click to expand...

Any temp difference will be less than 1c for water temps at any two random points in the loop. Anything that reports otherwise is either due to fault temp gauge equipment or people misreporting their temps.
 
i have 2 blocks cpu and gpu. A Thermaltake CPU block (i know, im going to replace it) and a Danger Den GPU block for the 8800GTX (Gets here on saturday). Im just kind of planning ahead before I get the parts.
 
96redformula said:
I believe at XS they decided in series is best. This way you will get equal amount distributed to both radiators and you will keep your pressure from having a drop.
Click to expand...

Ran through various scenarios, and it doesn't really matter (to within ~0.1C) whether the radiators are in liquid-flow series or parallel. Depending on specific circumstances it can swing one way or the other, but rarely does it differ by more than 0.1C either way. Just route it however is easiest, and typically that means in liquid flow series. Very high-flow low-restriction setups, parallel nudges ahead. For lower-flow higher-restriction setups, series nudges ahead. For anywhere in the middle of those two extremes, it doesn't really matter.

One thing that is extremely important is to have the radiators in air-flow parallel. That is, you must never have the air-flow exhaust of one radiator leading into the intake of the second radiator. At best, it'll be no better than just a single radiator by itself. At worst, your temps will actually be worse than with just a single radiator. It's the single best way to royally screw up the use of multiple radiators.
 
If you have two items that are identical (GPUs or RADs) and wye them together with tubing that is about the same length you will see better overall flow because you are halfing your restrictions.

However, using mismatched components or tubing that is very long on one and very short on the other the fluid will take the path of least resistance and one of the components will get little to no flow.

If you have low restriction components (Dtek Fuzion, Thermochill Rads) you will see no increase in performance. I have two Maze 4 GPUs wyed together and it works great.

Flow Path: Shortest possible as long as the reservoir is attached the intake of the pump.
 
Ranker said:
Any temp difference will be less than 1c for water temps at any two random points in the loop. Anything that reports otherwise is either due to fault temp gauge equipment or people misreporting their temps.
Click to expand...

picsstfu.gif

An amazing summation of more FUD from Ranker. You must learn it in the board rooms of LBM. I see you have back flipped on your 0.5 C claim from earlier posts. Since you claim to have all this high tech water cooling gear why don't you show us your screen shots?

On the question of the OP I would say that you might get more cooling from the series configuration. The cooled water of the first radiator is cooled even more as it goes through the second radiator.
 
Top Nurse said:
On the question of the OP I would say that you might get more cooling from the series configuration. The cooled water of the first radiator is cooled even more as it goes through the second radiator.
Click to expand...

No, thats not how it works.

As Cathar said, it doesn't really matter which way you do it.
 
You must log in or register to reply here.
Back
Top