Overclock.net › Forums › Cooling › Cooling Experiments › Silver Water Block
New Posts  All Forums:Forum Nav:

Silver Water Block - Page 5

post #41 of 50
Quote:
Originally Posted by ramicio View Post

Whatever happened to direct die cooling water cooling? Even direct heat-spreader cooling would be awesome.

I'm hoping to do some of that with my next project build....direct die cooling not with water but submerged in liquified refrigerant...all just thoughts and plans atm....but this will be coming together over the next few months...can't wait to start it..smile.gif
post #42 of 50
Nevermind. Metals have a much better heat conductivity than your liquids. You need this factor to spread out the heat. I've been reading about direct-die experiments and it just doesn't work.
 
Quiet Desktop
(15 items)
 
 
CPUMotherboardGraphicsRAM
i7-970 MSI X-58 Pro-E 8600GT 6 GB (3 x 2GB) Corsair DDR3 1333 
Hard DriveHard DriveCoolingOS
Hitachi Deskstar 2TB 32GB Patriot SSD Scythe Big Shuriken 2 Rev. B Ubuntu Server 
PowerCaseOtherOther
COOLMAX 700 watt modular Norco RPC-3216 Areca ARC-1222 w/ BBU 4 SFF-8087 Cables 
CPUMotherboardRAMHard Drive
i5-2500k ASUS P8Z68-M Pro 16 GB (4 x 4GB) Mushkin Enhanced Silverline DDR... 64GB Kingston SSD 
Optical DriveCoolingOSMonitor
Blu-ray reader and DVD/CD burner combo drive Scythe Big Shuriken 2 Rev. B Windows 7 Ultimate x64 1440x900 old LG LCD with anti-glare film peeled... 
KeyboardPowerCaseMouse
Dusty old Dell PS/2 Seasonic 560 watt modular Lian-li PC-V354B Some generic one from Best Buy 
Mouse PadAudioAudio
Plain black ASUS Xonar Essence STX Bottlehead Crack w/ Speedball 
  hide details  
Reply
 
Quiet Desktop
(15 items)
 
 
CPUMotherboardGraphicsRAM
i7-970 MSI X-58 Pro-E 8600GT 6 GB (3 x 2GB) Corsair DDR3 1333 
Hard DriveHard DriveCoolingOS
Hitachi Deskstar 2TB 32GB Patriot SSD Scythe Big Shuriken 2 Rev. B Ubuntu Server 
PowerCaseOtherOther
COOLMAX 700 watt modular Norco RPC-3216 Areca ARC-1222 w/ BBU 4 SFF-8087 Cables 
CPUMotherboardRAMHard Drive
i5-2500k ASUS P8Z68-M Pro 16 GB (4 x 4GB) Mushkin Enhanced Silverline DDR... 64GB Kingston SSD 
Optical DriveCoolingOSMonitor
Blu-ray reader and DVD/CD burner combo drive Scythe Big Shuriken 2 Rev. B Windows 7 Ultimate x64 1440x900 old LG LCD with anti-glare film peeled... 
KeyboardPowerCaseMouse
Dusty old Dell PS/2 Seasonic 560 watt modular Lian-li PC-V354B Some generic one from Best Buy 
Mouse PadAudioAudio
Plain black ASUS Xonar Essence STX Bottlehead Crack w/ Speedball 
  hide details  
Reply
post #43 of 50
Quote:
Originally Posted by ramicio View Post

Nevermind. Metals have a much better heat conductivity than your liquids.

Better conduction yes.......but the liquid will phase change to gas.....that's many many fold better than conduction.......PS......I will be using a heat spreader of sorts.......a sheet of Panasonic pyrolitic graphite directly adhered to the die surface with artic silver thermal adhesive (4-6X times more conductive than copper)...and the surface of the graphite will have diamond dust adhered to it again with thermal adhesive to promote nucleation/phase change of the liquid.

The fact that I'm using graphite and diamond dust may seem contrary to all I've been saying......but guess what.....the liquid I'm using is hydrophobic/non-polar in nature....thumb.gif
post #44 of 50
What liquid?
 
Quiet Desktop
(15 items)
 
 
CPUMotherboardGraphicsRAM
i7-970 MSI X-58 Pro-E 8600GT 6 GB (3 x 2GB) Corsair DDR3 1333 
Hard DriveHard DriveCoolingOS
Hitachi Deskstar 2TB 32GB Patriot SSD Scythe Big Shuriken 2 Rev. B Ubuntu Server 
PowerCaseOtherOther
COOLMAX 700 watt modular Norco RPC-3216 Areca ARC-1222 w/ BBU 4 SFF-8087 Cables 
CPUMotherboardRAMHard Drive
i5-2500k ASUS P8Z68-M Pro 16 GB (4 x 4GB) Mushkin Enhanced Silverline DDR... 64GB Kingston SSD 
Optical DriveCoolingOSMonitor
Blu-ray reader and DVD/CD burner combo drive Scythe Big Shuriken 2 Rev. B Windows 7 Ultimate x64 1440x900 old LG LCD with anti-glare film peeled... 
KeyboardPowerCaseMouse
Dusty old Dell PS/2 Seasonic 560 watt modular Lian-li PC-V354B Some generic one from Best Buy 
Mouse PadAudioAudio
Plain black ASUS Xonar Essence STX Bottlehead Crack w/ Speedball 
  hide details  
Reply
 
Quiet Desktop
(15 items)
 
 
CPUMotherboardGraphicsRAM
i7-970 MSI X-58 Pro-E 8600GT 6 GB (3 x 2GB) Corsair DDR3 1333 
Hard DriveHard DriveCoolingOS
Hitachi Deskstar 2TB 32GB Patriot SSD Scythe Big Shuriken 2 Rev. B Ubuntu Server 
PowerCaseOtherOther
COOLMAX 700 watt modular Norco RPC-3216 Areca ARC-1222 w/ BBU 4 SFF-8087 Cables 
CPUMotherboardRAMHard Drive
i5-2500k ASUS P8Z68-M Pro 16 GB (4 x 4GB) Mushkin Enhanced Silverline DDR... 64GB Kingston SSD 
Optical DriveCoolingOSMonitor
Blu-ray reader and DVD/CD burner combo drive Scythe Big Shuriken 2 Rev. B Windows 7 Ultimate x64 1440x900 old LG LCD with anti-glare film peeled... 
KeyboardPowerCaseMouse
Dusty old Dell PS/2 Seasonic 560 watt modular Lian-li PC-V354B Some generic one from Best Buy 
Mouse PadAudioAudio
Plain black ASUS Xonar Essence STX Bottlehead Crack w/ Speedball 
  hide details  
Reply
post #45 of 50
Not 100% decided yet due to availability....but if i can get it.....HFC227ea refrigerant.....also known as FM-200....chemical name Heptafluoropropane. Boiling point -16c. biggrin.gif

Energy absorbed by phase change 200 joules/ml of liquid.......water energy absorbed 4.2 joules/ml/deg c........see the potential?
Edited by technogiant - 3/27/13 at 12:02pm
post #46 of 50
Quote:
Originally Posted by technogiant View Post

There is much more to solid/liquid heat transfer than just considering the type of flow turbulent or otherwise.

It's highly complex stuff but here's a link which highlights what I've been trying to explain.

http://www.ltnt.ethz.ch/publications/Journal/pubimg/2011_Hu3.pdf

Top and bottom of the report is that by increasing the hydrophillic nature of a solid surface at a water/solid interface they increase heat transfer by a factor of 3X....so these are far from minor inconsequential considerations.

That paper describes the surface treatment that effectively increases the surface area using nano pillars, this is an off shoot research of hydrophobic materials when you make the nano pillars to close together for the water molecules to get at the surface of the material. The lab next to me just got in sheets of manufactured graphene so I will see if they are hydrophobic or just really smooth. They are growing Buckyball on the surface so that would be very similar to your in effect to that paper you quoted.

Good luck with the FM-200, be careful with the pressure in your loop if it boils that low. I want to see some pics of this set up, it sounds pretty bad a$$
All Around
(19 items)
 
  
CPUMotherboardGraphicsGraphics
1100t ASUS M5A99X EVO EVGA GTX 570 HD 2.5GB EVGA GTX 570 HD 2.5GB 
RAMHard DriveHard DriveCooling
Corsair Vengance  750 GB WD Caviar Black Intel 330  Corsair H60  
CoolingCoolingOSMonitor
Zalman CNPS20LQ CPU Water Cooling Thermaltake Water 2.0 Performer CPU Water Cooler Windows 7 Ultimate Dell 21", Samsung 19", Norcent 17", VIZIO 32" 
KeyboardPowerCaseMouse
Logitech G110 Corsair AX750 Cooler Master HAF 922 R.A.T. 7 
Mouse PadAudioOther
Steel SeriesNP+ Optical Toslink to SONY receiver to BOSE System  Jawbone ERA for calls and game chat 
  hide details  
Reply
All Around
(19 items)
 
  
CPUMotherboardGraphicsGraphics
1100t ASUS M5A99X EVO EVGA GTX 570 HD 2.5GB EVGA GTX 570 HD 2.5GB 
RAMHard DriveHard DriveCooling
Corsair Vengance  750 GB WD Caviar Black Intel 330  Corsair H60  
CoolingCoolingOSMonitor
Zalman CNPS20LQ CPU Water Cooling Thermaltake Water 2.0 Performer CPU Water Cooler Windows 7 Ultimate Dell 21", Samsung 19", Norcent 17", VIZIO 32" 
KeyboardPowerCaseMouse
Logitech G110 Corsair AX750 Cooler Master HAF 922 R.A.T. 7 
Mouse PadAudioOther
Steel SeriesNP+ Optical Toslink to SONY receiver to BOSE System  Jawbone ERA for calls and game chat 
  hide details  
Reply
post #47 of 50
Quote:
Originally Posted by ramicio View Post

I've been reading about direct-die experiments and it just doesn't work.

In relation to this, this is a very interesting read:-

http://detector-cooling.web.cern.ch/detector-cooling/data/3M_FAQ_Fluorinert.pdf#page=24

It discusses the heat density that a submersion cooled pc can cope with in "still" phase change coolant before a critical point is reached where the component becomes blanketed by a layer of gas and so no longer cooled.

They have a conservative working limit on the heat density of 6-10 watts/square cm. You can see how direct die cooling could easily go beyond that limit.
It is also mentioned that forcing coolant flow over the component can greatly increase the heat density that can be dealt with......interesting.
Edited by technogiant - 3/28/13 at 12:18am
post #48 of 50
Quote:
Originally Posted by SDH500 View Post

That paper describes the surface treatment that effectively increases the surface area using nano pillars, this is an off shoot research of hydrophobic materials when you make the nano pillars to close together for the water molecules to get at the surface of the material. The lab next to me just got in sheets of manufactured graphene so I will see if they are hydrophobic or just really smooth. They are growing Buckyball on the surface so that would be very similar to your in effect to that paper you quoted.

Good luck with the FM-200, be careful with the pressure in your loop if it boils that low. I want to see some pics of this set up, it sounds pretty bad a$$

That will be interesting to know about the graphene......as regards my next build......I'm going to make a video log when I finally get going....but first I've to get an out house to do it in...as you say could be dangerous if things went wrong so won't do it in the house.


PS...regards the paper I linked to.....its describing several surface modifications only one of which was relating to nano pillars......it also describes modifying the hydrophillic nature of the solid surface.....here's a direct lift from the conclusion.

"The surface recondition with hydrophilic
groups leads to an increase in the thermal conductance of the
interface by a factor varying from 1.8 to 3.2 times"
Edited by technogiant - 3/27/13 at 3:28pm
post #49 of 50
And even those Novec submerged systems still have some tiny heatsink to spread the heat. I can't wait to see what you do with this refrigerant and graphene plan.
 
Quiet Desktop
(15 items)
 
 
CPUMotherboardGraphicsRAM
i7-970 MSI X-58 Pro-E 8600GT 6 GB (3 x 2GB) Corsair DDR3 1333 
Hard DriveHard DriveCoolingOS
Hitachi Deskstar 2TB 32GB Patriot SSD Scythe Big Shuriken 2 Rev. B Ubuntu Server 
PowerCaseOtherOther
COOLMAX 700 watt modular Norco RPC-3216 Areca ARC-1222 w/ BBU 4 SFF-8087 Cables 
CPUMotherboardRAMHard Drive
i5-2500k ASUS P8Z68-M Pro 16 GB (4 x 4GB) Mushkin Enhanced Silverline DDR... 64GB Kingston SSD 
Optical DriveCoolingOSMonitor
Blu-ray reader and DVD/CD burner combo drive Scythe Big Shuriken 2 Rev. B Windows 7 Ultimate x64 1440x900 old LG LCD with anti-glare film peeled... 
KeyboardPowerCaseMouse
Dusty old Dell PS/2 Seasonic 560 watt modular Lian-li PC-V354B Some generic one from Best Buy 
Mouse PadAudioAudio
Plain black ASUS Xonar Essence STX Bottlehead Crack w/ Speedball 
  hide details  
Reply
 
Quiet Desktop
(15 items)
 
 
CPUMotherboardGraphicsRAM
i7-970 MSI X-58 Pro-E 8600GT 6 GB (3 x 2GB) Corsair DDR3 1333 
Hard DriveHard DriveCoolingOS
Hitachi Deskstar 2TB 32GB Patriot SSD Scythe Big Shuriken 2 Rev. B Ubuntu Server 
PowerCaseOtherOther
COOLMAX 700 watt modular Norco RPC-3216 Areca ARC-1222 w/ BBU 4 SFF-8087 Cables 
CPUMotherboardRAMHard Drive
i5-2500k ASUS P8Z68-M Pro 16 GB (4 x 4GB) Mushkin Enhanced Silverline DDR... 64GB Kingston SSD 
Optical DriveCoolingOSMonitor
Blu-ray reader and DVD/CD burner combo drive Scythe Big Shuriken 2 Rev. B Windows 7 Ultimate x64 1440x900 old LG LCD with anti-glare film peeled... 
KeyboardPowerCaseMouse
Dusty old Dell PS/2 Seasonic 560 watt modular Lian-li PC-V354B Some generic one from Best Buy 
Mouse PadAudioAudio
Plain black ASUS Xonar Essence STX Bottlehead Crack w/ Speedball 
  hide details  
Reply
post #50 of 50
Yeah found that with the gpu as the heat density is greater they used a boiler plate with small 1mm pins on it ....but nothing extreme......the most important feature of their system is the coating they use on the heatspreaders/boiler plates.......they use sub 10 micron diamond dust baked on at high temps with some kind of an organic adhesive....this increases cooling efficiency by 15X as I recall due to the nucleation effect.

I've no idea what adhesive they have used but can get diamond dust of similar grit size.......just hope I can emulate their 15X efficiency factor with what I can source.

Having said that about their extremely efficient coating.........it won't prevent them running into the same problem at the critical point where the component is blanketed by the gas.......I think my idea to spread the heat laterally with the graphite sheet may well be the better one...just a 10cm X 10cm sheet would using their heat density max figures be able to cool 600w - 1000w before critical point was reached (in fact more than that as coolant can access the upper and lower side of the sheet) ....even if my coating layer isn't as efficient as theirs I'll have a larger surface area so it will even out and I won't hit the critical point due to the bigger surface area.

Looking at the figures even on the cpu they must be approaching max heat density if they heavily OC......my current 2700K draws approx 140watts at 5.5Ghz package size 3.75 X 3.75 cm - 14sq cm (that's the package size not the heat spreader size) so that's up at their max figure of 10 watts/sq cm.

Having said all that about the critical point though it must be remembered that this applies to a "still" submersion model....I'm intending to also submerge a 2 pumps into the tank, one will circulate liquid over the evap and the other will force jets of fluid directed at the cpu/gpu's so the critical heat density will be much higher.
Edited by technogiant - 3/28/13 at 1:24am
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Cooling Experiments
Overclock.net › Forums › Cooling › Cooling Experiments › Silver Water Block