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

Which way to flow?

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

nixie

Member
Joined
Jun 20, 2001
Location
Southampton, ENGLAND
I'm trying to decide what order to set my components up in, what’s the consensus on the following?

Parts

Cooling

  • Eheim 1048
  • DD RBX
  • DD Radeon block (to be modified to fit a Ti4600
  • Mini Metro Heater Core (150mm x 140mm x 30mm)
  • 1/2" DD tubing
  • DD bay res, one output, 2 inputs (not the clear one, the chemical resistant one, can't find a link to it at the moment)
  • 2x 2-port manifolds
  • Y-splitter
  • 2x 120mm fans pushing approx 60cfm to 80cfm each

PC

  • Abit KD7
  • XP 1600+ (currently at 2gig on air with no volt increase)
  • Creative Ti4600
  • 512mb Crucial pc3200 DDR
  • SB Audigy

Options

All setups are to fit in a midi tower case with a box underneath with the same foot print as the case large enough to take the radiator and pump. Plan is to have pump and rad in the box underneath with the 2 120mm fans pulling air over the rad (which will be in a sealed compartment).

Rad is shown at start and end of loops to show joining point.

1

Rad --> Pump --> GPU --> CPU --> Res (using both res inputs for the return from the RBX for less restriction) --> Rad​

2

Rad --> Pump --> CPU --> Y-split --> GPU --> Res (using one input) --> Rad​

3

Rad --> Pump --> CPU --> One RBX output direct to Res , one to GPU then Res --> Rad​

4

Rad --> Pump --> Y-split or manifold --> one to CPU, one too GPU --> gpu into one res input cpu into other using manifold or Y to recombine before res --> Rad​

My opinion is that option 4 is overly complicated so should be discarded. Out of option 1 or 2 I think that 1 would be better as it as less overall resistance (due to using the unrestricted inputs to the res instead of a Y piece). Option 4 is the one I'm uncertain of, I think I'm correct in saying this setup would have the least overall resistance. Having the GPU on one of the RBX inputs would allow the RBX to have the best pressure into its input. I understand that due to the uneven resistance on the RBX outputs flow rate would be better on the side without the GPU block but am uncertain to whether this causes problems or not. I know that it would probably be better to feed the pump from the res, but due to case size and layout having the res in with the rad and pump over complicates the tubing (lots of corners).

Apologies for no pictures but I don't have a good camera, will try to borrow one, or take some shots with my camera phone.

Thanks in advance.
 
I've always been told that the CPU should get the first water from the RAD as it's the coolest. So that may eliminate #1. I would think most people would agree that #2 is the typical setup with your current water block scheme.
 
Thanks for your input guys.

I'm thinking that #3 would suit the case layout/size better so will probably use that unless someone tells me that it will have a detremental effect. I understand that overall the system temperature will even out so layout isn't a major concern in that respect, its more minimising restriction on the system (and hence maximising flow rate and pressure) that I think is going to be important. From other threads it seems that opinion is that a 1048 is capable of providing enough flow and presure for a RBX, but all examples given were with just a rad, pump and block.

I'm aiming to start a project log if I can sort out pictures. Preliminary work is already underway with the rad and fans holes cut in the air box. Once my exams are over and I've moved flat in 2 weeks I'll be getting stuck into finishing the build.
 
Just do everything in series, the easiest way possible. DO not split off lines to differrent components.
 
capneedle said:
I've always been told that the CPU should get the first water from the RAD as it's the coolest. So that may eliminate #1. I would think most people would agree that #2 is the typical setup with your current water block scheme.

You have been informed wrong unless you are measuring in 0.1C increments. As a practical matter, the order that is best is the one that is the most direct with the shortest amount of tubing and gentle turns. The temperature change of the water flowing through the loop is less than 1C. I know it is not intuitive, but it is the truth.
 
Back