Water-cooled heatsink suggestion
 

I have to admit that I am new within this area, but there is a lot of information already researched. I am in the market to buy a water-cooled heatsink, but I wanted to know what design would be best for my present AMD beast and future processors.

I found an article that provides a history of many different water-cooled designs, e.g. Hitachi (M880) and IBM’s Thermal Conduction Module (TCM). Also, the design of future components was in question. The present day 2-D chips may change to a 3-D chip, if there is limiting physical space. Currently, this is not a problem therefore cooling is not a serious issue. However, in the near future too much heat might be given off from the entire system to keep everything cool and stable, e.g. CRAY. Imagine something like a G4 cube for your computer box.

The one thing that caught my eye was deionized water in direct contact with a micro finned chip. It stated that this set-up was able to remove as much as 790 W/cm2 of heat. [The source is dated back in 1981!] Basically, the design is nothing more than an encapsulated air-cooled (micro-finned) heatsink built on the CPU. I am curious if someone has (could) designed an Alpha heatsink using this principle. Supposedly, the higher surface area will be more effective at water-cooling than the simple copper or aluminum blocks sold on the market. BTW, the heatsink fins on the larger heatsinks might be overkill. Shorter, finely etched, heatsink fins might work just as well.

Note: I saw that overclock-watercool.com has used something close to this principle with a 486 heatsink. Any comments on this subject? Perhaps, this is already old news, but it is interesting nonetheless.

-phosphor

All the waterblocks I have put together are based on either K6 heatsinks or (1) P100 Heatsink. The CPU stays under 100 F under heavy load. But yeah, you could just go ahead and wrap up some heatsink(an milled down copper inlayed Alpha would proably be best for the newer AMDs) and build a casing around it. You'd proably want to mill it down to a size the size of the Socket or slight larger/smaller. And proably a lower height as well.

------------------
Watercooled K6-2 450 at 600

mfpmax,
I had a quick look at your website (nice setup), and I am curious about the efficiency of the larger heatsink on the last waterblock. Is it any better than the other waterblocks that you have made? I am thinking about the route(s) that you took for making a cooling block, but I am still trying to decide what heatsink to use.

Perhaps, the heatsink on the Abit BX chipset would be just as good as a larger K6 S7 heatsink. I noticed that BX heatsink has a lot of grooves and openings on the surface. Possibilities?

I'm using a weird really thin finned aluminium heatsink encased in wood... not operating yet, but tests are promising...

Well the first block was built with a heatsink with shorter fins than the new one. Can't really say much cause the first block was also in a the first watercooler setup. When I went with the FREAKISHLY larger K6 heatsink(which was cut down to fit in a 4" PVC Cap), I couldn't really tell ya any performence difference. But I can tell you that the heatsink itself did 82F idle while the waterblock can do 77F idle. Now I did see a waterblock with small raised bumps on the inside of it. This may do a good job, the thing is, using a heatsink just forces the water everywhere, while a block you buy(standard ones) are channeled, now having small fins in the channeled ones would proably work well.

------------------
Watercooled K6-2 450 at 600

Channeling water, which will reach every part of a multi-finned heatsink, would be an excellent cooling block. I think that the engineering end of this contraption is going to be difficult and time consuming. I have to wait until I have enough time to construct one.

Thanks for the responses.

homemade waterblocks give me a chubby.

http://people.ne.mediaone.net/sullyfish/chubby.jpg

By the way I have made a few waterblocks and none of mine have internal fins.

Just a chunk of aluminum with a hole drilled through it.

All have kept the P3 internal diode within a few degrees of room temp.

The one in the picture was the worst. It's hole was to narrow and flow was restricted after a while.

Water transfers heat to metal very very well, so I dont think internal fins are essential. http://www.procooling.com/ubb/smile.gif

Its been forever and a day since I saw that image.

------------------
Watercooled K6-2 450 at 600

Thinking about it more, I believe that the silver polyblock will be an interesting prospect for my water cooling project.

Cu later, that is one heck of a polished side of a block.

thanks for the response