Overclock.net › Forums › Cooling › Water Cooling › Question on loop splitting
New Posts  All Forums:Forum Nav:

Question on loop splitting

post #1 of 8
Thread Starter 
So I have been following the Big Budget Boomer Box build by cpachris, and saw that Electrocutor had provided an interesting idea for his build that allowed him to redirect flow to different components selectively.

I thought that looked interesting, but I had a different loop in mind for my build and so tried to overlay that idea on mine.

However, when I now re-look at that plan I have to think I'm really just causing water to push back on itself based on how I did this.

Before I over analyze this to death, I wanted to solicit feedback on the loop below.

1) Will it actually do what I want it to do?
2) Is it really worth it?

Pandamonium
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7-3930K ASUS Rampage IV Extreme EVGA GTX 670 4GB SC x3 Samsung 4 x 4G 
Hard DriveHard DriveCoolingCooling
OCZ Vertex 4 256 x 2 Seagate Barracuda 3TB x4 EK-Supremacy CPU Water Block Heatkiller 670 Waterblocks x3 
CoolingCoolingCoolingCooling
XSPC Rampage IV Extreme Blocks Swiftech MCP35X2 Pump x2 Danger Den Rad Res x2 Alphacool NexXxoS Monsta Quad 120mm x4 Rads 
CoolingOSPowerCase
Alphacool NexXxoS Monsta Triple 120mm Rad Windows 7 Professional SP1 64-bit CORSAIR AX1200 & AX650 Caselabs TH10 with ped 
  hide details  
Reply
Pandamonium
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7-3930K ASUS Rampage IV Extreme EVGA GTX 670 4GB SC x3 Samsung 4 x 4G 
Hard DriveHard DriveCoolingCooling
OCZ Vertex 4 256 x 2 Seagate Barracuda 3TB x4 EK-Supremacy CPU Water Block Heatkiller 670 Waterblocks x3 
CoolingCoolingCoolingCooling
XSPC Rampage IV Extreme Blocks Swiftech MCP35X2 Pump x2 Danger Den Rad Res x2 Alphacool NexXxoS Monsta Quad 120mm x4 Rads 
CoolingOSPowerCase
Alphacool NexXxoS Monsta Triple 120mm Rad Windows 7 Professional SP1 64-bit CORSAIR AX1200 & AX650 Caselabs TH10 with ped 
  hide details  
Reply
post #2 of 8
Seems like a good idea, however im not sure it would work..
First Pc
(12 items)
 
  
CPUMotherboardGraphicsRAM
Amd Athlon 64 X2 6000+ ASUS CROSSHAIR NVIDIA GeForce GTX 280 GeIL Black Dragon 4GB (2X2GB) X2 
Hard DriveHard DriveOptical DriveOS
Seagate Barracuda 500GB Seagate  Sony DVD drive Windows 7 Professional 
MonitorKeyboardCaseMouse
Dell E152FP Microsoft keyboard Antec  logitech wireless mouse 
  hide details  
Reply
First Pc
(12 items)
 
  
CPUMotherboardGraphicsRAM
Amd Athlon 64 X2 6000+ ASUS CROSSHAIR NVIDIA GeForce GTX 280 GeIL Black Dragon 4GB (2X2GB) X2 
Hard DriveHard DriveOptical DriveOS
Seagate Barracuda 500GB Seagate  Sony DVD drive Windows 7 Professional 
MonitorKeyboardCaseMouse
Dell E152FP Microsoft keyboard Antec  logitech wireless mouse 
  hide details  
Reply
post #3 of 8
This is the closest thing i've done to loop-splitting (it's an older photo, sorry)

6Lvly.jpg

the lines have t-fittings in at the GPU, so the pumps output line goes directly to the CPU and GPU simultaneously, and the return lines both work the same way.. I'll just tell you, that when I compared the temps (from doing just a straight res>pump>cpu>gpu>radiator>res line, compared to what that is).. there was practically no difference in temps..

now.. im not quite sure im reading your chart right.. your valves, where you're splitting the water off.. if you close the valves, more water is being pushed across the CPU/GPU etc, but if it's open, the water is being recycled for cooling and NOT going to the CPU/GPU? is that right? if so, then what is the benefit of having the water being cooled if it's not actually being pushed across the things it's supposed to be cooling?
post #4 of 8
Just way too many rads. Get a 420 and or a 360 with a 2 x 355's with a after market top/res and be done with it. That is just too much fittings, leaks and time invested for no gains. Thanks
post #5 of 8
im not sure why you would ever want to skip part of a loop... unless that valve is meant to allow some of the water to bypass some of the components in an effort to deliver more cool water to the components later in the loop. if that is the case, it is completely useless because the water doesnt change temp that much. in one loop the most it will change is like one degree. its not worth the extra effort.
post #6 of 8
Quote:
Originally Posted by Psyenyde View Post

im not quite sure im reading your chart right.. your valves, where you're splitting the water off.. if you close the valves, more water is being pushed across the CPU/GPU etc, but if it's open, the water is being recycled for cooling and NOT going to the CPU/GPU? is that right?

No, the valves are on bypass tubes that run in parallel to the cooling sections so they do not stop flow through the cooling components.

This is a very interesting idea, but I too am doubtful whether you would get any significant benefit from this since I don't think there will be any effect on temps at all. The thermal load is unaffected since that is generated by the electronic components. So no matter what volume of water carries that load away from the components per unit of time (less if the bypasses are open; more if closed), the total heat that the rads will have to dissipate is unchanged. The only thing that it seems to me that this setup will affect is flow rate across the cooling components. And the only effect that I can see of opening a valve will be to lower flow rate on the components that are parallel to the opened bypass, and I don't see why you would ever want to do that.

I also think there's a logical error in your notes on the flow diagram:
Quote:
the more [a valve] is open the less water gets to those blocks but the more water gets recirculated in the rads.

All the parallel sections come together before hitting the rads and there's no bypass around the rads. Therefore the flow rate through the rads will be unchanged no matter how the valves are set. At the point where the parallel sections all join up, the water will all mix together again and as I see it should be exactly the same net temperature as if there were no bypasses.
Edited by threephi - 8/4/12 at 3:29pm
Blinky
(17 items)
 
  
CPUMotherboardGraphicsRAM
i7-950 @4.034 GHz Asus Rampage III Formula XFX HD-5850 Black Edition x2 6GB Crucial Ballistix 1600MHz 
Hard DriveCoolingCoolingCooling
120GB OCZ Vertex 2 EK Supreme HF (full nickel) EK FC-5850 (nickel/acetal) x2 Danger Den Monsoon Dual-Bay Reservoir/Pump 
OSMonitorKeyboardPower
Win 7 Pro 64 Dell U2410 Rosewill RK-9000 Seasonic X-650 
CaseOther
Cosmos 1000 Lamptron FC5-V2 
  hide details  
Reply
Blinky
(17 items)
 
  
CPUMotherboardGraphicsRAM
i7-950 @4.034 GHz Asus Rampage III Formula XFX HD-5850 Black Edition x2 6GB Crucial Ballistix 1600MHz 
Hard DriveCoolingCoolingCooling
120GB OCZ Vertex 2 EK Supreme HF (full nickel) EK FC-5850 (nickel/acetal) x2 Danger Den Monsoon Dual-Bay Reservoir/Pump 
OSMonitorKeyboardPower
Win 7 Pro 64 Dell U2410 Rosewill RK-9000 Seasonic X-650 
CaseOther
Cosmos 1000 Lamptron FC5-V2 
  hide details  
Reply
post #7 of 8
Thread Starter 
Thanks everyone for the input.

To clarify my comment on open valves, what I should have said is that by pushing less water through the components overall the water should theoretically be cooler, and thus another pass through the rads should cause a smaller delta between water temp and ambient.

My feeling though is in the end it won't make a difference.

Anyone else any new thoughts on this? So far I am leaning to just a normal single loop with the 2 pumps.
Pandamonium
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7-3930K ASUS Rampage IV Extreme EVGA GTX 670 4GB SC x3 Samsung 4 x 4G 
Hard DriveHard DriveCoolingCooling
OCZ Vertex 4 256 x 2 Seagate Barracuda 3TB x4 EK-Supremacy CPU Water Block Heatkiller 670 Waterblocks x3 
CoolingCoolingCoolingCooling
XSPC Rampage IV Extreme Blocks Swiftech MCP35X2 Pump x2 Danger Den Rad Res x2 Alphacool NexXxoS Monsta Quad 120mm x4 Rads 
CoolingOSPowerCase
Alphacool NexXxoS Monsta Triple 120mm Rad Windows 7 Professional SP1 64-bit CORSAIR AX1200 & AX650 Caselabs TH10 with ped 
  hide details  
Reply
Pandamonium
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7-3930K ASUS Rampage IV Extreme EVGA GTX 670 4GB SC x3 Samsung 4 x 4G 
Hard DriveHard DriveCoolingCooling
OCZ Vertex 4 256 x 2 Seagate Barracuda 3TB x4 EK-Supremacy CPU Water Block Heatkiller 670 Waterblocks x3 
CoolingCoolingCoolingCooling
XSPC Rampage IV Extreme Blocks Swiftech MCP35X2 Pump x2 Danger Den Rad Res x2 Alphacool NexXxoS Monsta Quad 120mm x4 Rads 
CoolingOSPowerCase
Alphacool NexXxoS Monsta Triple 120mm Rad Windows 7 Professional SP1 64-bit CORSAIR AX1200 & AX650 Caselabs TH10 with ped 
  hide details  
Reply
post #8 of 8
I think you should do the valve setup and see what happens. If you do the valves right....you can set it so that if they are closed...then it's a normal single serial loop. If they are open...then you get the recirculation subloop that should lower your delta temps at least . Since you'll have some fancy monitoring equipment, you can prove or disprove any of the theories. That is what I plan to do.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Water Cooling
Overclock.net › Forums › Cooling › Water Cooling › Question on loop splitting