![]() | ||
|
|
Water Block Design / Construction Building your own block? Need info on designing one? Heres where to do it |
![]() |
Thread Tools |
![]() |
#1 |
Cooling Savant
Join Date: May 2003
Location: Utah
Posts: 160
|
![]()
I was just wondering. I always see blocks that sit on top of the processor die, and nothing else. Why not mill out the block a little and make it fit, like a glove, over the Die? If you could mill it out and get it to fit tight, I think you could get better cooling, because no heat would dissapate to the air between the block and the area around the Die.
If I don't make sense, just ask and I'll draw a pretty 3d image to show you what I'm asking about. |
![]() |
![]() |
![]() |
#2 |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
![]()
The problem is fitting it.
The core measurements are available, within a +/- 0.01 mm margin, but isn't there a glue leak around it? There is also an involved difficulty in making something so precise, which will involve tools that few of us have. Even a CNC mill can't cut those corners square. |
![]() |
![]() |
![]() |
#3 |
Cooling Neophyte
Join Date: Dec 2002
Location: us
Posts: 75
|
![]()
i could be wrong but dont some amc cpu's have connections on top. if it fit like a glove then this would have a posibility of causing a short which would be a bad thing also.
|
![]() |
![]() |
![]() |
#4 |
Cooling Neophyte
Join Date: Sep 2001
Location: Houston, TX
Posts: 55
|
![]()
There are several problems with having a recess machined into the block to fit a die the way you described.
To get proper pressure on the die would require more mounting force to compress the four bumper pads and increased pressure on the corners. I don't think that is a good idea. The surface mounted devices (SMDs) around the die limit how deep a recess can be made. SMDs are not designed to support any kind of cooling unit wieght or pressure. AMD dies are notoriously fragile so any misalignment during mounting will chip the die, especially with the close tolerances needed to accomplish the "glove effect". Finally, if you are able to successfully get it all put together with no damage and it all works, you have created a time bomb! Due to the glove-like fitment of the block and die, any bump of the block will likely break the core, especially any accidental side or twisting motion. I'll stick with the current flat design and give up any possible miniscule temp reduction improvement . |
![]() |
![]() |
![]() |
#5 |
Cooling Neophyte
Join Date: Apr 2003
Location: USA
Posts: 96
|
![]()
I have actually thought of something simillar....kinda like building in a "shim" that are/were so popular with HSF units.
|
![]() |
![]() |
![]() |
#6 |
Thermophile
Join Date: Aug 2002
Location: Just shut up ;) ...
Posts: 1,068
|
![]()
I don't think it would'nt cool any better, there's not much mor surface area, the sides of my AMD chips are'nt flat or symetrical/straight either, they have a 'snapped off' look to them. You could get the same effect by putting a closed cell foam gasket around it...
|
![]() |
![]() |
![]() |
#7 | |
Cooling Savant
Join Date: Feb 2002
Location: Oxford University, UK
Posts: 452
|
![]() Quote:
![]() Surely heat dissipating to the air is a GOOD thing, since it is still heat travelling AWAY from the cpu. 8-ball
__________________
For those who believe that water needs to travel slowly through the radiator for optimum performance, read the following thread. READ ALL OF THIS!!!! |
|
![]() |
![]() |
![]() |
#8 |
Cooling Neophyte
Join Date: Mar 2003
Location: Scotland
Posts: 32
|
![]()
Could you not get away with having a recess a bit bigger than the core, and bridging the gap with a decent thermal compound?
Just tell me to shut up if im wrong ![]() |
![]() |
![]() |
![]() |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|