Overclock.net › Forums › Cooling › Water Cooling › The Quest for the Perfect Waterblock
New Posts  All Forums:Forum Nav:

The Quest for the Perfect Waterblock - Page 5

post #41 of 95
I know its not meant for TEC use but thats what it would be good for

you dont need to cool the edge of the CPU cause your just cooling stuff thats already cool lol
    
CPUMotherboardGraphicsRAM
Q6600/E6600/P4 515/Celery 352 ASUS P5K Premium Black Pearl PNY 8800GTX Gskill HZ 1000s D9GKX 
Hard DriveOptical DriveOSMonitor
2x WD Raptors 36gb Raid 0 Lite On Sata DVD-RW Custom XP Pro Dual 22" Westys WS 
KeyboardPowerCaseMouse
A4 Tech Slim Tagan 1100w U95 TurboJet Gigabyte 3D Aurora The Original G5 
Mouse Pad
Giant Rubber / Felt Pad (awesome) 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Q6600/E6600/P4 515/Celery 352 ASUS P5K Premium Black Pearl PNY 8800GTX Gskill HZ 1000s D9GKX 
Hard DriveOptical DriveOSMonitor
2x WD Raptors 36gb Raid 0 Lite On Sata DVD-RW Custom XP Pro Dual 22" Westys WS 
KeyboardPowerCaseMouse
A4 Tech Slim Tagan 1100w U95 TurboJet Gigabyte 3D Aurora The Original G5 
Mouse Pad
Giant Rubber / Felt Pad (awesome) 
  hide details  
Reply
post #42 of 95
Thread Starter 
if i were to cool direct die, yes. but the IHS is called an IHS for a reason. so the entire IHS is covered. the barbs are all located outside that area, to have the IHS completely covered with pins, without leaving the corners without much flow.
as a result, the block indeed cools parts that could do without. but as heat travels trough the copper horizontally as well, it wont hurt.

thats my logic behind it anyways.
post #43 of 95
I will have to look for the direct link from the other "quad-block" posts ... ( a lot to look through ) but there is a very useful link somebody posted for us that gives exact specifications for "no-fly zones" as well as component height limitations within certain distances from the processor. Great info for block design. Maybe somebody else can beat me to it ?
post #44 of 95
Prolly was mentioned but I think the first step would be to make the mating surface of the block as THIN as possible. I mean if you could make it paper thin and durable the do it. That would mean less heat is slowly absorbed by the copper and it gets cooled faster.

Next thing I would do is do something about the design of water entering the block. They are always at a 90 degree angle of entrance and leaving. How about making the barbs enter at a lower angle? I mean thinking about it it'd be like having a fan blow on your face when you are hot. Less flow means you'd be hotter while more flow means you'd be cooler. So I think a faster flow would be better. Basically make it more sreamlined. Then to top it off I'd do like the Swiftech block with the diamonds in the middle of it all to creat some turbulence.

Problem is it'd be big. In order to do a side entrance and get even flow you'd have to do like a car's intercooler and have end tanks on it. These little ends would have some baffling in them to distribute the flow correctly so water would spread out across width of the block. The baffling could be heatsink thin so that you'd keep sizes down.
The Core
(5 items)
 
  
CPUMotherboardGraphicsGraphics
Intel Core i5 Z77 Extreme4 Intel(R) HD Graphics 4000 NVIDIA GeForce GTX 770 
RAM
32GB 
  hide details  
Reply
The Core
(5 items)
 
  
CPUMotherboardGraphicsGraphics
Intel Core i5 Z77 Extreme4 Intel(R) HD Graphics 4000 NVIDIA GeForce GTX 770 
RAM
32GB 
  hide details  
Reply
post #45 of 95
Thread Starter 
the inlet barb would still be positioned in a 90* angle. this will make the water flow in all directions equally, and create a lot of turbulence, which is good.
i agree with you if were talking about the exits.
good you thought of it, 'cuz i didnt.
post #46 of 95
You mean something like these old Swifty blocks? That design has been around for years..


....SWIFTECH MCW5000-PX COPPER BASE CPU WATER COOLING BLOCK FOR XEON
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
E8600 4545Mhz*9 w/1.29V DFI DK P45 XFX 8800 GS D9GKX & D9GMH 
Hard DriveOptical DriveOSPower
2x Seagate 160 Perp. RD-1 2x Asus Vista Premium Corsair HX 620 
Case
Open Bench 
  hide details  
Reply
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
E8600 4545Mhz*9 w/1.29V DFI DK P45 XFX 8800 GS D9GKX & D9GMH 
Hard DriveOptical DriveOSPower
2x Seagate 160 Perp. RD-1 2x Asus Vista Premium Corsair HX 620 
Case
Open Bench 
  hide details  
Reply
post #47 of 95
Thread Starter 
something like that, but i had something else in mind. i'll show tomorrow.
post #48 of 95
Quote:
Originally Posted by ChielScape View Post
the inlet barb would still be positioned in a 90* angle. this will make the water flow in all directions equally, and create a lot of turbulence, which is good.
i agree with you if were talking about the exits.
good you thought of it, 'cuz i didnt.
You really should read through the other post for quad-core block design ... most of this was already discussed there.

Anyway, here is the Intel processor specs ... good info here ... I'll see about finding the other .pdf concerning height limitations on LGA775 motherboards.

ftp://download.intel.com/design/proc...s/31559202.pdf

A couple images from the other post ...





Hope this helps !
post #49 of 95
coming out of nowhere without any knowledge lol.

but how about a spiral design?

Inlet at 90* angle in the center of the WB then water spirals on the surface of the cpu and exit on the side, as in parrallel to the motherboard.

the side outlet would not work i suppose..

maybe that designs hasn't enough flow..
post #50 of 95
More SA is the way to go. People say "look at the storm"...yeah look at it, its no longer the top performer. Look at the fuzion and gtx is more like it. They have a lot of SA and reasonable flow. I have a design for excellent flow and tons of SA but dont want my idea to get out in the open. Within the next couple months it will become a reality and we will see what happens.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Water Cooling
Overclock.net › Forums › Cooling › Water Cooling › The Quest for the Perfect Waterblock