Overclock.net › Forums › Cooling › Water Cooling › Radiator size advice
New Posts  All Forums:Forum Nav:

Radiator size advice

post #1 of 10
Thread Starter 
Hi,

Im on a dilemma at the moment on getting a new case whether is going to be a Corsair 650D or 600T? Honestly, i prefer the looks of the 600T but I found some issue of it on water cooling setup due to the limited spacing of the top of the case.

So the setup that I'm planning to do is not really a total high end. Thinking of just overcloking an i5 2500k and get a single gpu or perhaps two if there's sufficient room. I'm wondering with this kind of hardware cooling, how big is the radiator that I should use? Let just say enough to have a stable system with 30-35% overclocking.

I'm a total newbie by the way in watercooling by the way wink.gif
post #2 of 10
I would say a 3 x 120mm radiator minimum if you are putting the GPU in the loop with the CPU.
Low Tide
(12 items)
 
  
CPUMotherboardGraphicsRAM
i5-3570K @ 4.7GHz Asus P8Z77-V Pro EVGA GTX 670 FTW 8GB G. Skill 2133MHz 
Hard DriveCoolingOSMonitor
Samsung 850 Pro 256GB Prolimatech Megahalems Rev. C Windows X Samsung T240HD 
KeyboardPowerCaseAudio
Ducky DK9008G2 Red Corsair TX850W V1 Corsair Carbide 300R X-Fi XtremeGamer 
  hide details  
Reply
Low Tide
(12 items)
 
  
CPUMotherboardGraphicsRAM
i5-3570K @ 4.7GHz Asus P8Z77-V Pro EVGA GTX 670 FTW 8GB G. Skill 2133MHz 
Hard DriveCoolingOSMonitor
Samsung 850 Pro 256GB Prolimatech Megahalems Rev. C Windows X Samsung T240HD 
KeyboardPowerCaseAudio
Ducky DK9008G2 Red Corsair TX850W V1 Corsair Carbide 300R X-Fi XtremeGamer 
  hide details  
Reply
post #3 of 10
Thread Starter 
I see, if that's the case, seems like it's not really worth doing water cooling in this mid case. Maybe I just stick with air cooling with this. Still wanna get the 600t, really love the design and a h100 should do the job well already. Thanks
post #4 of 10
Quote:
Originally Posted by azrul View Post

I see, if that's the case, seems like it's not really worth doing water cooling in this mid case. Maybe I just stick with air cooling with this. Still wanna get the 600t, really love the design and a h100 should do the job well already. Thanks

You can fit a 200mm Phobya rad or a 180 magicool rad up front.

If you do the math:

120.3
432cm2 =(12*3)*12

200.1
400cm2=20*20

180.1
324cm2=18*18

So a 200mm rad is almost equal to a 360 (120.3) radiator. Now if you put that 200 up front and a 120.2 up top, you've got enough rad to support two GPU's and a CPU with OC's.

Don't give up, keep asking questions.

Edit: I've also seen a Phobya 480 mounted to the side mesh panel insert on 600T's. Granted you lose the view of the inards, and you need some QDC's on the 480 to help with access in the future. But then again, you'd be good to have a GPU and CPU with OC's within a 650D or 600T.
Edited by Rognin - 3/20/12 at 12:07pm
Synthetic Anatomy
(14 items)
 
NAT-inator
(11 items)
 
 
CPUMotherboardGraphicsGraphics
Intel Core i7 2700K Z68X-UD7-B3 NVIDIA GeForce GTX 580 NVIDIA GeForce GTX 580 
GraphicsRAMRAMRAM
NVIDIA GeForce GTX 580 Corsair  Corsair  Corsair  
RAMHard DriveCoolingPower
Corsair  2x Mushkin 240Gb SSD Raid 0  A lot. ST1500 
CaseAudio
FT02 Xonar ST 
CPUMotherboardGraphicsRAM
i5-2500K P8P67 WS Revolution Sparkle Calibre GTX 560 G.Skill 
Hard DriveHard DriveOptical DriveCooling
Corsair Force SSD WD Caviar Black BD combo Liquid 
OSPowerCase
Win 7 Ultimate AX650 500R 
  hide details  
Reply
Synthetic Anatomy
(14 items)
 
NAT-inator
(11 items)
 
 
CPUMotherboardGraphicsGraphics
Intel Core i7 2700K Z68X-UD7-B3 NVIDIA GeForce GTX 580 NVIDIA GeForce GTX 580 
GraphicsRAMRAMRAM
NVIDIA GeForce GTX 580 Corsair  Corsair  Corsair  
RAMHard DriveCoolingPower
Corsair  2x Mushkin 240Gb SSD Raid 0  A lot. ST1500 
CaseAudio
FT02 Xonar ST 
CPUMotherboardGraphicsRAM
i5-2500K P8P67 WS Revolution Sparkle Calibre GTX 560 G.Skill 
Hard DriveHard DriveOptical DriveCooling
Corsair Force SSD WD Caviar Black BD combo Liquid 
OSPowerCase
Win 7 Ultimate AX650 500R 
  hide details  
Reply
post #5 of 10
If you know your estimated heat load (wattage), you can roughly calculate how much radiator you will need, based on the C/W performance for your radiator. Skinneelabs.com or martinsliquidlab.org have this info for most radiators. Info for a few other radiators are produced by guys like Bundymania who posts his reviews in various forums (Search google).

Basically, you find your radiators c/w value at the fan speed you want (pay attention to model of fans they use in the review), and then you can calculate how much heat can be removed for your desired level of Delta T (difference in water temp vs ambient air temp).

For example, say I want a delta T of 5C and want to use 1000RPM fans (Super quiet) on a XSPC RX360 radiator. I can go to Skinnelabs and see their latest test on this radiator reporting a C/W of 0.02763289 at 1000RPM (note Yate Loon fans). C/W decreases as fan speed and/or radiator capacity increases. But, for this example we're dead set on 1000RPM fans and a desired delta T. So, the only variable left is to find out how much radiator is required.

Now, lets say my CPU has a heat load of 300watts when overclocked (just a number from the sky. Don't know what a 2500k really puts out). I take my desired delta T divided by my heat load to give me the desired C/W rating. In this example it's 5C / 300watts = a desired C/W of 0.0166666667.

Now take the 0.02763289 / 0.0166666667 and you find it will take 1.65 RX360 radiators to dump that heat with 1000RPM fans and a D/T of 5C. This roughly equates to needing a RX360 rad + a RX240 rad in the same loop.

If we go with faster 1800RPM yate loons, our C/W on a single RX360 would drop to 0.01694725. So, punching the calculator again, we take 0.01694725 / (5 / 300) = 1.016 radiators. So, increasing your fan speed will get you the performance you need from a single radiator. Albeit, with more fan noise.

Note, most tests are done with fans in push-only. Push-pull fans will drop your CW by 20-30% (according to MLL spreadsheet). As well, these numbers are almost never rock solid accurate. There's always variables that differ from the test environments, like ambient temperatures, pump heat-dump, etc. So, when it doubt, err for more radiator, if you can.

Hope this helps. PS, I'm not an expert either. So, someone may have some corrections for my figures. But, If my thinking cap is screwed on right, this way will work.
Edited by Stalker1456 - 3/21/12 at 7:02am
post #6 of 10
Quote:
Originally Posted by Stalker1456 View Post

If you know your estimated heat load (wattage), you can roughly calculate how much radiator you will need, based on the C/W performance for your radiator. Skinneelabs.com or martinsliquidlab.org have this info for most radiators. Info for a few other radiators is produced by guys like Bundymania who posts his reviews in various forums (Search google).
Basically, you find your radiators c/w value at the fan speed you want (pay attention to model of fans they use in the review), and then you can calculate how much heat can be removed for your desired level of Delta T (difference in water temp vs ambient air temp).
For example, say I want a delta T of 5C and want to use 1000RPM fans (Super quiet) on a XSPC RX360 radiator. I can go to Skinnelabs and see their latest test on this radiator reporting a C/W of 0.02763289 at 1000RPM (note Yate Loon fans). C/W decreases as fan speed and/or radiator capacity increases. But, for this example we're dead set on 1000RPM fans and a desired delta T. So, the only variable left is to find out how much radiator is required.
Now, lets say my CPU has a heat load of 300watts when overclocked (just a number from the sky. Don't know what a 2500k really puts out). I take my desired delta T divided by my heat load to give me the desired C/W rating. In this example it's 5C / 300watts = a desired C/W of 0.0166666667.
Now take the 0.02763289 / 0.0166666667 and you find it will take 1.65 RX360 radiators to dump that heat with 1000RPM fans and a D/T of 5C. This roughly equates to needing a RX360 rad + a RX240 rad in the same loop.
If we go with faster 1800RPM yate loons, our C/W on a single RX360 would drop to 0.01694725. So, punching the calculator again, we take 0.01694725 / (5*300) = 1.016 radiators. So, increasing your fan speed will get you the performance you need from a single radiator. Albeit, with more fan noise.
Note, most tests are done with fans in push-only. Push-pull fans will drop your CW by 20-30% (according to MLL spreadsheet). As well, these numbers are almost never rock solid accurate. There's always variables that differ from the test environments, like ambient temperatures, pump heat-dump, etc. So, when it doubt, err for more radiator, if you can.
Hope this helps. PS, I'm not an expert either. So, someone may have some corrections for my figures. But, If my thinking cap is screwed on right, this way will work.

Great information there matey. Will have to try and use this to see if my delta t of 10-11C matches up to what skinneelabs calculations say they should.
My Gaming Rig
(16 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770K Z97X-SOC Force MSI GTX 1070 FE G.Skill TridentX 2400Mhz CL9 
Hard DriveCoolingCoolingCooling
2 Crucial M4 256GB XSPC RX360 + RX240 + RX120 XSPC Raystorm EK 1070FE WB 
OSMonitorKeyboardPower
Win 10 64 Qnix QX2710 96Hz CM Storm Rapid Coolermaster V Series 850w 
CaseMouseMouse PadAudio
NZXT Switch 810 Mionix Mionix Onboard Sound / HD449s 
  hide details  
Reply
My Gaming Rig
(16 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770K Z97X-SOC Force MSI GTX 1070 FE G.Skill TridentX 2400Mhz CL9 
Hard DriveCoolingCoolingCooling
2 Crucial M4 256GB XSPC RX360 + RX240 + RX120 XSPC Raystorm EK 1070FE WB 
OSMonitorKeyboardPower
Win 10 64 Qnix QX2710 96Hz CM Storm Rapid Coolermaster V Series 850w 
CaseMouseMouse PadAudio
NZXT Switch 810 Mionix Mionix Onboard Sound / HD449s 
  hide details  
Reply
post #7 of 10
Quote:
Originally Posted by Rognin View Post

You can fit a 200mm Phobya rad or a 180 magicool rad up front.
If you do the math:
120.3
432cm2 =(12*3)*12
200.1
400cm2=20*20
180.1
324cm2=18*18
So a 200mm rad is almost equal to a 360 (120.3) radiator. Now if you put that 200 up front and a 120.2 up top, you've got enough rad to support two GPU's and a CPU with OC's.
Don't give up, keep asking questions.
Edit: I've also seen a Phobya 480 mounted to the side mesh panel insert on 600T's. Granted you lose the view of the inards, and you need some QDC's on the 480 to help with access in the future. But then again, you'd be good to have a GPU and CPU with OC's within a 650D or 600T.

You also have to acct for the dead zone caused by the center of the fan, which the 200 rad proportionally has less of than the rest.
 
Project Atlantis
(16 items)
 
 
CPUMotherboardGraphicsRAM
i7 3960X w/ EK Supreme HF  Asus Rampage IV Extreme w/ EK full cover Tri SLi EVGA GTX TITAN SC w/ EK full cover & Ba... 64 GB Dominator w/ MIPS Ram Cooler 
Hard DriveHard DriveCoolingCooling
1TB Samsung EVO 850 2x 2TB Western Digital Caviar Black 1x Hardware Labs Black Ice GTX 560  Bitspower Pump Upgrade Kit 
CoolingCoolingCoolingCooling
2x Hardware Labs Black Ice GTX 280 Hardware Labs Black Ice GTX 480 Dual Uber MCP655 Noiseblocker PK-3  
CoolingCoolingCoolingOS
Bitspower Compression Fittings Lamptron FC-8 Noiseblocker PL-2 Windows 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Asus PB287Q 4K Steelseries 7G Custum wired EVGA NEX 1500 Caselabs SMA8 
MouseMouse Pad
Rat 7 Razer Goliathus XL 
CPUMotherboardGraphicsRAM
i7 920 c0 @ 4.2 Ghz w/ 1.4v max 4.6ghz @1.55v w... Asus P6T6 WS Revolution w/ koolance nb/sb+mosfe... EVGA GTX 480 x3 w/ Koolance Vid-NX480 12gb Corsair Dominator 1600mhz w/ MIPS Ramcooler 6 
Hard DriveOptical DriveCoolingOS
1.5 Tb Seagate barracuda x3 LG Bluray Burner 2x SR1 360, RX 360, SR1 240, SR1 120, 2x MCP355... Win 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Samsung t260 x3 Razer Black Widow Ultimate Silverstone st1200 Mountain Mods U2UFO w/ MM Pedestal Addition 
MouseMouse PadAudioAudio
Razer Mamba Razer Destructor X-Fi Fata1ity Logitech Z-5500 
CPUGraphicsRAMHard Drive
i7 3840QM 650ti 8GB 1333mhz DDR3 256gb ssd 
OSMonitorPowerCase
OSX Mountain Lion Retina Display 85w Magsafe Power Adapter Aluminum Unibody 
  hide details  
Reply
 
Project Atlantis
(16 items)
 
 
CPUMotherboardGraphicsRAM
i7 3960X w/ EK Supreme HF  Asus Rampage IV Extreme w/ EK full cover Tri SLi EVGA GTX TITAN SC w/ EK full cover & Ba... 64 GB Dominator w/ MIPS Ram Cooler 
Hard DriveHard DriveCoolingCooling
1TB Samsung EVO 850 2x 2TB Western Digital Caviar Black 1x Hardware Labs Black Ice GTX 560  Bitspower Pump Upgrade Kit 
CoolingCoolingCoolingCooling
2x Hardware Labs Black Ice GTX 280 Hardware Labs Black Ice GTX 480 Dual Uber MCP655 Noiseblocker PK-3  
CoolingCoolingCoolingOS
Bitspower Compression Fittings Lamptron FC-8 Noiseblocker PL-2 Windows 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Asus PB287Q 4K Steelseries 7G Custum wired EVGA NEX 1500 Caselabs SMA8 
MouseMouse Pad
Rat 7 Razer Goliathus XL 
CPUMotherboardGraphicsRAM
i7 920 c0 @ 4.2 Ghz w/ 1.4v max 4.6ghz @1.55v w... Asus P6T6 WS Revolution w/ koolance nb/sb+mosfe... EVGA GTX 480 x3 w/ Koolance Vid-NX480 12gb Corsair Dominator 1600mhz w/ MIPS Ramcooler 6 
Hard DriveOptical DriveCoolingOS
1.5 Tb Seagate barracuda x3 LG Bluray Burner 2x SR1 360, RX 360, SR1 240, SR1 120, 2x MCP355... Win 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Samsung t260 x3 Razer Black Widow Ultimate Silverstone st1200 Mountain Mods U2UFO w/ MM Pedestal Addition 
MouseMouse PadAudioAudio
Razer Mamba Razer Destructor X-Fi Fata1ity Logitech Z-5500 
CPUGraphicsRAMHard Drive
i7 3840QM 650ti 8GB 1333mhz DDR3 256gb ssd 
OSMonitorPowerCase
OSX Mountain Lion Retina Display 85w Magsafe Power Adapter Aluminum Unibody 
  hide details  
Reply
post #8 of 10
But big fans have less pressure and cfm compared do a couple of 120mm fans. So i think its roughly the same to 120/140 setups beating bigger fans rads
Sovereign
(15 items)
 
  
CPUMotherboardGraphicsRAM
i7 3770k Maximus V Gene HIS IceQ 7950 16gb Samsung  
Hard DriveCoolingOSMonitor
Samsung 840 120gb H80i Windows 7 x64 Benq XL2420T 
PowerCaseMouseMouse Pad
Seasonic 1000 Platinum TJ08e Mionix Naos 5000 Propus 380 
Audio
Creative ZxR 
  hide details  
Reply
Sovereign
(15 items)
 
  
CPUMotherboardGraphicsRAM
i7 3770k Maximus V Gene HIS IceQ 7950 16gb Samsung  
Hard DriveCoolingOSMonitor
Samsung 840 120gb H80i Windows 7 x64 Benq XL2420T 
PowerCaseMouseMouse Pad
Seasonic 1000 Platinum TJ08e Mionix Naos 5000 Propus 380 
Audio
Creative ZxR 
  hide details  
Reply
post #9 of 10
Thread Starter 
Wow.. thanks for the info there. I'd almost given up already and was planning to spent more on getting i7. But then looking at the amount of people saying that this could work and by theory it should. I surely will watercool this build and perhaps maybe logging as well if I have the time. wink.gif

However, I do prefer having a quieter system and from the math, seems like this will need a higher cfm fan in order to maximize the performance. So by theory, by having a 1000rpm or 2000rpm fans, how much temp increase would it cause? Well I may can just have a try and see as fans doesn't really cost too much but hey, its good to have some experts advice and can lower down the budget a bit wink.gif

Another dilemma is of course which type of components or manufactures that can deliver most from this build. There's quite a lot out there and seriously I really don't have any idea yet which to choose from lol. Maybe someone can help too? I'm from Malaysia by the way and note that not all products are easily found here. wink.gif
Edited by azrul - 3/21/12 at 8:11am
post #10 of 10
Thread Starter 
Quote:
Originally Posted by Stalker1456 View Post

If you know your estimated heat load (wattage), you can roughly calculate how much radiator you will need, based on the C/W performance for your radiator. Skinneelabs.com or martinsliquidlab.org have this info for most radiators. Info for a few other radiators are produced by guys like Bundymania who posts his reviews in various forums (Search google).
Basically, you find your radiators c/w value at the fan speed you want (pay attention to model of fans they use in the review), and then you can calculate how much heat can be removed for your desired level of Delta T (difference in water temp vs ambient air temp).
For example, say I want a delta T of 5C and want to use 1000RPM fans (Super quiet) on a XSPC RX360 radiator. I can go to Skinnelabs and see their latest test on this radiator reporting a C/W of 0.02763289 at 1000RPM (note Yate Loon fans). C/W decreases as fan speed and/or radiator capacity increases. But, for this example we're dead set on 1000RPM fans and a desired delta T. So, the only variable left is to find out how much radiator is required.
Now, lets say my CPU has a heat load of 300watts when overclocked (just a number from the sky. Don't know what a 2500k really puts out). I take my desired delta T divided by my heat load to give me the desired C/W rating. In this example it's 5C / 300watts = a desired C/W of 0.0166666667.
Now take the 0.02763289 / 0.0166666667 and you find it will take 1.65 RX360 radiators to dump that heat with 1000RPM fans and a D/T of 5C. This roughly equates to needing a RX360 rad + a RX240 rad in the same loop.
If we go with faster 1800RPM yate loons, our C/W on a single RX360 would drop to 0.01694725. So, punching the calculator again, we take 0.01694725 / (5 / 300) = 1.016 radiators. So, increasing your fan speed will get you the performance you need from a single radiator. Albeit, with more fan noise.
Note, most tests are done with fans in push-only. Push-pull fans will drop your CW by 20-30% (according to MLL spreadsheet). As well, these numbers are almost never rock solid accurate. There's always variables that differ from the test environments, like ambient temperatures, pump heat-dump, etc. So, when it doubt, err for more radiator, if you can.
Hope this helps. PS, I'm not an expert either. So, someone may have some corrections for my figures. But, If my thinking cap is screwed on right, this way will work.

Honestly, this is really helping me out and many2 thanks for the information. wink.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Water Cooling
Overclock.net › Forums › Cooling › Water Cooling › Radiator size advice