• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

waterblock construction thread/worklog...

rogue_jedi

Supreme [H]ardness
Joined
Sep 6, 2002
Messages
4,821
i'll be putting pics and info up about my 3 waterblocks i'm making, along with results and all that. more later....


pics are up a few posts down. heh. (up - down... get it? cuz i don't)
 
Pretty cool. I'm interested in the materials and design of the waterblocks. I myself have built 2 for my GPU and NB out of a 1U skivved copper heatsink and a block of UHMW polyproplyene. They turned out nicely. I am building a whitewater style CPU block the same way for a friend as soon as I get the heatsink in. It is gonna be sweet :D
 
i'm using 3 2x3x1/4 copper plates for each block, plus other stuff. i'm using compression fittings for the hoses for two reasons: they don't leak if you teflon tape all the threads, and they were the only fittings i could find at the first hardware store i went to, so i got them instead of going looking for hose barbs however, i'm going to a 5/8ID loop when i get around to finishing all the blocks. o yeah, i'm doing these for science fair, which is why i'm not switching now.
 
Do you already have all of the measurements for the GPU and NB and CPU dimensions? Also, what kind of structures are you going to use inside the waterblock to create the most turbulence? Also, what is the advantage of 5/8" tubes? I doubt you will be able to see a difference in flow as the waterblocks will be the limiting factor in your system and will eat up all the flow.
 
i have a pretty big pump, flow isn't really an issue. i'm using 3 designs:
1) drilled in a hexagonal pattern (5/32 bit) and cut between the holes with a dremel (cheap and easy but it works)
2) "Maze" design - basically a tub with channels in it
3) jet impingement design - hardest to make 5/8, but whatever. impinger with channels dremeled into the base (i love my dremel)

NOTES ON CONSTRUCTION:
drill press ++
knowing machinist ++
not planning ahead --
screwing up mounting holes --
tapping holes ++/-- (hard to do but you have to do it)
blowtorching ++++++++
 
Hmm, what will you be cooling? Is there a TEC in the future? I think a #rotor block would work fine, the jet impingement thing would be tough to do unless you knew someone with a slotter. The maze is easy, but performance isnt great.

You should consider doing a lumpy block design, like some of jaydee116's work. Basically, it is a #rotor block but with channels directing the water instead of pins in an open space. With this design, you can make the base plate very thin because the structure will be much more rigid.
 
pic time (~200kb each):
block mounting:
block a mounted.jpg

block a (internal):
inside block a.jpg

hose routing (no good, but what can i do?):
hose routing.jpg

inside block b:
inside block b.jpg

reservoir/pump (yeah yeah the water is really dirty. bite me):
pump and reservoir.jpg



the reason there are 6 holes in the block b is, quite simply, i screwed up the mounting holes. oops:eek:
i'm rebuilding the best block anyway, so its not like it matters much, it just doesn't look good.

<porky>
that's all folks!
</porky>
 
You really need to widen the slots between the holes on your #rotor block, I dont see how your flow could be good at all. Also, when you upgrade to 5/8" tubes you will not see a perfomance increase.
 
yeah widening the channels was suggested... but these blocks are kind of soldered shut, so it wouldn't be much fun to open it up to change it. and it would cost another $4 (for a tube of silver solder)

the 5/8" tubing is mainly for cool factor + less flow restriction - my pump is designed for 3/4" out, 1" in, or submersed. i think that 5/8 will help, at least a little. i can also switch to hose barbs then (i only have the 5/8" tap:eek:)

more after i finish my homework and get some work done:/
 
got the "maze" block to seal, just need to get enough teflon tape on the fittings so it won't leak:mad: stupid me tapped the holes too far:( grrr...

o well, nothing a bit of teflon can't fix:D
 
Originally posted by rogue_jedi
got the "maze" block to seal, just need to get enough teflon tape on the fittings so it won't leak:mad: stupid me tapped the holes too far:( grrr...

o well, nothing a bit of teflon can't fix:D

i did a similar thing on my air chiller setup, (tapping holes "too far") but on purpose, I threaded the fittings on, then dremeled and sanded the excess off before putting block togeter. works great, and you don't see all the threads on top.

from that thread:

I drilled and tapped for the fittings.


Set the fitting with good o’l JB Weld.


Here I cut off the excess sticking through the bottom, and then lapped it flat.


Now bind the top on, again with JB. Clamped it tight with screws, and let it set. After it set I lapped it flat as well.
 
more pics:
block b mounted - my brother got them to stop leaking. w00t:D
block b mounted.jpg

my testing setup:
setup.jpg


topcat: as i stated, i got them fixed my own way:) but thanks for the suggestion - i'll have to try that for a more permanent solution.

and so far this block is working a LOT better then the previous one - 3C under load better:eek: and it isn't lapped yet:p

later.
 
yeah i know but i had to have some airflow over my stuff and my disk was heating up pretty bad. i also didn't feel like having my fan controller sitting on my desk as well. (too much stuff + small desk = bad)

it's 70mm because my brother broke a fan blade off my 80mm one by being an idiot. now he can't use his computer, as it is his cpu fan. i also took his psu before i started testing, so he can't have it back until i'm done. i have his 9000AIW too (for now)

and his DVDburner:D
 
Originally posted by rogue_jedi
.......topcat: as i stated, i got them fixed my own way:) but thanks for the suggestion - i'll have to try that for a more permanent solution........

cool. I could have just as easily threaded the correct depth, but like i siad, i didn't want to see threads on top of the block. Not that it makes any real differance, but I'm kinda anal on stuff like that:p

PS- good work on the block. keep us up to date.
 
Originally posted by retardedchicken
nice job man, that smartfan2 you got sitting your HDD=ghetto

That's the problem with buying a cheap desk, not enough pre-cut fan mounts... :D
 
my dad made the desk for me:eek:

i'm lapping the block now (taking a break to get my internet fix for the day) but it has some nice scratches right smack dab in the center. grrr:mad::mad:

so it'll be awhile til i'm done (i'm using 320grit sandpaper currently)
temps went up a few degrees after the water warmed up (it started cold) but the block still worked slightly better than the first one even with two disadvantages:
1)it wasn't lapped yet
2)my brother put it on, not me, and he doesn't know how to use the mounting system as well as i do.

so things are looking good for block b!
 
Originally posted by rogue_jedi
my dad made the desk for me:eek:

i'm lapping the block now (taking a break to get my internet fix for the day) but it has some nice scratches right smack dab in the center. grrr:mad::mad:

so it'll be awhile til i'm done (i'm using 320grit sandpaper currently)
temps went up a few degrees after the water warmed up (it started cold) but the block still worked slightly better than the first one even with two disadvantages:
1)it wasn't lapped yet
2)my brother put it on, not me, and he doesn't know how to use the mounting system as well as i do.

so things are looking good for block b!

I was just saying... none of the components are in a case, how were you supposed to mount that fan without it being ghetto?
 
bumpity because i'm pissed off again. can't mount the tube for the impinger properly. grr...

ideas? it's a piece of 1/2" copper tubing with the end crimped.

later.

o yeah, also bumpity because i got a linux box set up in my room. that makes two machines... getting better:p
 
hehe i do this a lot... say i'm pissed off, then actually try to fix it again and lo and behold it works. maybe i should try that first next time:p... nah, too easy:D

i had to slightly simplify the design to make it work, but whatever. leak testing now, lapping tomorrow. and hopefully i will be able to convince/bribe/???? my teacher to let me turn in the paper late so i can finish the freaking thing. i can do it with three (i think) but it'd be nice to get all three in there.

pics tomorrow because it is now 11:45PM, i have school tomorrow, nd my homework isn't done yet... life is very very busy. yah.

later (again)
 
Originally posted by rogue_jedi
hehe i do this a lot... say i'm pissed off, then actually try to fix it again and lo and behold it works. maybe i should try that first next time:p... nah, too easy:D

i had to slightly simplify the design to make it work, but whatever. leak testing now, lapping tomorrow. and hopefully i will be able to convince/bribe/???? my teacher to let me turn in the paper late so i can finish the freaking thing. i can do it with three (i think) but it'd be nice to get all three in there.

pics tomorrow because it is now 11:45PM, i have school tomorrow, nd my homework isn't done yet... life is very very busy. yah.

later (again)

hehe, the simplest way is often the best way... good luck, keep us updated
 
god i'm really pissed at my watercooling right now. the stupid freaking waterblock started leaking, after leak-testing it for ~24hours, and after running on my computer for an two-three hours (running prime95 or ut2004 demo just about the whole time)

finished a trial, started recording results, then the computer locked up. look at the machine, saw water on the block. unplugged pump DAMN fast. tried to turn off computer (no power switch... heh) then realized i was 1 pin off. moved it over, shut off computer. dried up the water on the mobo :eek::eek::eek: with a qtip. god this sucks. i'm not at all happy right now. i have to remember to stay away from sharp objects (killing lots of people makes jesus cry)

so yeah, back to work when i get home.:(
 
yah me my mobo / proc isn't dead. it scared me when i turned it on, then plugged in my speakers and it rebooted itself, but i proceeded to play ut2k4 for an hour or three when it came back up (spectated for a while, let the bots have fun) and it's still up now.

still need to test the jet block. stupid thing had practically a curved base, but i had flattened out the core area hoping it would be enough. guess what? it wasn't enough. the contact was not so good. which probably explains why it got a 19C rise over case (ambient - same thing) temp as opposed to the maze block i built getting a 17C-18C rise.

more later. pics if it works, if not i don't sleep til tuesday (night, hopefully;))
 
still no pics... haven't gotten around to getting them out of the video.been too busy working on this block. i lapped the bottom for a long time with 220 grit. i don't know exactly how long, but i started before i started watching george of the jungle 2 (funny ass movie, you should see it if you haven't already) and ended a few minuted after the movie did. the surface is closer to flat (close enough, dammit)

i goe the pipe to stay soldered on. found out/figured out that i'd been soldering it wrong, as i was applying direct heat to the solder. it wants indirect heat (i.e. heat the surface it is on) so it's working now. and damned well, i might add. it's testing about 2-3C cooler than the other blocks, and "responds" to ambient changes faster as well. me happy once more.:) i'll try to have the pics of it up by tomorrow night (but no assembled/attached pics yet... need to steal my brother's AIW to get those:D)

later
 
Damn, that's alot of sanding. Are you using higher grits too?
And yes, the easiest way to solder is to heat the pieces you're trying to join. Even if you melt the solder, it's not going to work if it falls onto a cold surface. If you have problems getting the solder to flow, you can put some flux on the pieces too, it sometimes helps alot when the parts are dirty or oxidized.
 
yeah, i use higher grits. i start with 220 to get out the scratches from belt sanding the base (blowtorching it on cement wreaks havoc with the base, can't imagine why... :p) then i go to 320-400-600-high. high = 1500 or 1000/2000. then i sand it on the back of some sandpaper.

i can read my core stepping in the base of my waterblock (and my alpha, but that's sitting in my closet until i get another athlon to use it on. or until i put it on a video card... hmm)
 
rogue_jedi

they look like some nice wc blocks youv mage there, im impressed. been comtemplating making some of my own to try and save some money but for now it's looking like im gonna start my watercooling experience with an all in one kit
 
yeah... i hate to say it, but a kit might be a good idea. this is cool/l33t/etc., but it's costed me about $120 so far and a lot of time (i mean like 80hours+ probably) and i still have another ~$100 to sink in it still. but it is really cool, and this 3rd block is working like a charm now that i finally got it right:D.
 
But there's something so satisfying about saying "I built this myself". I started out with all prebuilt (but bought separately) components myself, and then later modified stuff and built a couple of my own blocks. I think that's a pretty good way to go if you don't want to dive in headfirst right off the bat.
 
Back
Top