View Full Version : check out my "homeade" water block and mount
dozier768
09-09-01, 08:09 PM
finaly got my pictures of my watercooling project developed and scaned, the actual project only took about 4 hrs of machine work and my first mounting bracket was not a great design but i evolved it into a 4 corner spring loaded mount and i have just finished my latest block wich incorperates the mount into the block so its all one peice and instead of 5 passes the new one is six. ill have pic's of this and my cooling tower in about a week i think.
dozier768
09-09-01, 08:09 PM
another shot on my mill
dozier768
09-09-01, 08:10 PM
dirilling holes for tapping
dozier768
09-09-01, 08:10 PM
finished block
dozier768
09-09-01, 08:11 PM
block relay and pump
dimmreaper
09-09-01, 08:11 PM
Looks cool. I'd of ponyed up to the cost of copper myself, but aluminum will do OK. :)
dozier768
09-09-01, 08:12 PM
.
dozier768
09-09-01, 08:13 PM
mounted on my soyo k7vta-b with the bad mount
dozier768
09-09-01, 08:13 PM
closeup of my new style mount
dozier768
09-09-01, 08:15 PM
one last close up, the new one uses 3/8 lines this one was 1/4 inch and held the temps at 41c full load, the new one holds em at 36 full load
dozier768
09-09-01, 08:16 PM
ya reaper my next one will be all copper i just have a ton of alluminum laying around so i used it. and it works pretty well
dozier768
09-09-01, 08:21 PM
oh ya heres a shot of my messy work area
dimmreaper
09-09-01, 08:24 PM
How thick did you make the base Dozier?
dozier768
09-09-01, 08:42 PM
1/16
Warlord2
09-09-01, 09:46 PM
nice block
I didnt think aluminum would of done that good of job myself
tell us your temps after you make the copper one to compare=]
dozier768
09-09-01, 10:02 PM
i surely will, does any one know where to get thick copper stock, i can only find thin stock local
dozier768
09-09-01, 11:26 PM
^BUMP
dimmreaper
09-10-01, 12:15 AM
Originally posted by dozier768
1/16 That's too thin I'm afraid.
According to a few inside industry sources of mine, the optimal thickness for waterblock bases, and peltier coldplates, is 3/16.
I'm interested in seeing how well your block performs with the 1/16" base though. So do let us all know how it turns out.
dozier768
09-10-01, 12:56 AM
im sorry upon going through my schematic's its 1/8 and i am running it now and it is running at idle the delta T is 10c of ambient and 15c of abient at full load after 1 hr
the pump is a maxijet 1200 and the head is only about 1 inch all the line is 3/8 and the radiator is a automotive heatercore and it is a closed circut cooling system %11 glycol/89% distilled ionized water, the fans are dual 12v 87 cfm radioshak specials
i will try 3/16 with the copper one though if that will yeid better performance
dimmreaper
09-10-01, 01:13 AM
What is your water temp, and what is the CPU temp? Also what speed and voltage are you running the CPU at? This would be the most valuable information to us, as we could then calculate your waterblocks C/W rating.
dozier768
09-10-01, 03:32 AM
i am currently running at 1.82v 105x10.5
i will have to go pick up my thermal probe from my buddy's house and make some sort of inline water temp tester ill p.m. you when i'm done and let you know, probably a day or 2
r0ckstarbob
09-10-01, 11:04 AM
Originally posted by dimmreaper
That's too thin I'm afraid.
According to a few inside industry sources of mine, the optimal thickness for waterblock bases, and peltier coldplates, is 3/16.
I'm interested in seeing how well your block performs with the 1/16" base though. So do let us all know how it turns out.
hm. well that brings up a question. when watercooling, are we cooling the water block and letting the waterblock cool the CPU or are we (ideally) trying to actively cool the core? wouldn't trying to make that base as thin as possible be benneficial so your fluid could effect maximum heat-transfer levels and hit as much of the core heat at it's source as it could? it seems to me that trying to cool the core directly and letting the block take whatever residual heat the liquid doesn't get would be a much more efficient technique, no? am i off base here? this is just from an outside point of view....
*wheels a turnin... wheels a turnin*
well done by the way on the water block. very nice job.
Why would not a thinner block cool better as there is less metal to heat up and the water will carry away the CPU heat far more efficently than the metal.
dimmreaper
09-10-01, 03:44 PM
Originally posted by r0ckstarbob
hm. well that brings up a question. when watercooling, are we cooling the water block and letting the waterblock cool the CPU or are we (ideally) trying to actively cool the core? wouldn't trying to make that base as thin as possible be benneficial so your fluid could effect maximum heat-transfer levels and hit as much of the core heat at it's source as it could? it seems to me that trying to cool the core directly and letting the block take whatever residual heat the liquid doesn't get would be a much more efficient technique, no? am i off base here? this is just from an outside point of view....
*wheels a turnin... wheels a turnin*
well done by the way on the water block. very nice job. Thin bases, or no bases (as in direct die cooling), are only acceptible for CPUs generating less than ~40W. After that point you need a thick base to conduct heat away from the core. With a thin base, you just won't have enough effective surface area (area that is conducting heat from the core, to the water) to cool hotter CPUs. This is why both BeCooling waterjackets, and DangerDen waterblocks use a 3/16" base. And believe me, both places have tried a lot of combonations to get the best performing, and most cost effect, watercooling device.
First let me say that I am no expert.....but wouldn't any water block that only can carry away 90% or the generated heat continue to heat up and up (until you reached the equalization point between the ablitity of the CPU to generate heat and the water block to remove heat). The higher heat will cause the water block the heat up the case air temp which will in turn make the CPU run hot etc.....almost to the point of thermal runnaway. No matter what you use for a cooling device (air cooled or water cooled) you are faced with the "design fault" of the Tbird/Duron CPU....the small die size.
Just my thoughts.
Owenator
09-10-01, 08:35 PM
Dozier768
I have bought lots of copper barstock from http://www.onlinemetals.com and they are just up the road in Seattle! Sign up for a user id and you get 10% off. They sell small quantities too, my biggest order was around $12 even with shipping! I am from Portland OR originally but now live in DE :(. I am trying my to get back to the NW soon. Nice vertical mill too, is it yours or from work? I have a cheap drill press and a cross vice so my results aren't as pretty but the cooling is good!
Nice work!
O
dozier768
09-10-01, 10:36 PM
Originally posted by Owenator
Dozier768
I have bought lots of copper barstock from http://www.onlinemetals.com and they are just up the road in Seattle! Sign up for a user id and you get 10% off. They sell small quantities too, my biggest order was around $12 even with shipping! I am from Portland OR originally but now live in DE :(. I am trying my to get back to the NW soon. Nice vertical mill too, is it yours or from work? I have a cheap drill press and a cross vice so my results aren't as pretty but the cooling is good!
Nice work!
O
thanks for the info, and it is my own i also am the proud owner of a central machine big green lathe and various other FUN stuff, beats the heck out of a hacksaw lol
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.