1st waterblock = fail

[terramir]

:rofl: I finished a water block project I started like 4 years ago. and well so far a colossal failure.
I had the baseplate dremeled out well I put the plexi plate and the pipes in and well, it leaks not just from the pipes but also from the all the sides :oops

well check it out and that leaking occurred with a 2 liter per minute windshield wiper pump well look at it and gimme a few hints. http://www.ephotobay.com/image/first-waterblock.jpg (http://www.ephotobay.com/share/first-waterblock.html)
While this is only a learning project and will most likely never end-up in one of my computers it's sort of important for me to figure this out before I start the next one.
help
terramir

[B_Boy]

Some clear silicone around the edges before you bolt it down, would be the simple solution...

Nice idea though :P

[Serpentarius]

you need O-Rings ...

and try mill the waterblock with passion ... passion will help you create better blocks

[terramir]

ehh yeah but I figured It wouldn't leak that much water drops building everywhere LOL I might have to lap the surface first. then try again also need to figure out how to seal the pipes for good. need more tips and tricks. the eventual water pump will either be a bought one or an old fuel pump (those are like rated for 14.4V max and about 60L/minute ergo 15 gallons because the injection systems need that sort of pressure.
terramir

edit: o rings I can do they sell a set at a 99 cents store I know :p tons of sizes

[JasonDTM]

Why u be trollin?

[Sparky]

No seals = leaks.

Also, just a word of warning - looks like you have an aluminum base, with copper pipes. You'll get corrosion that way. Now for testing and experimentation I wouldn't be too worried about it, but for a final product to actually be used you'll want to not mix metals.

[relttem]

JB Weld..

[terramir]

Why u be trollin?
someone translate what he saying cause, cause me = american born but grown up in europe so some figures of speech just lost on me :confused:

No seals = leaks.

Also, just a word of warning - looks like you have an aluminum base, with copper pipes. You'll get corrosion that way. Now for testing and experimentation I wouldn't be too worried about it, but for a final product to actually be used you'll want to not mix metals.
Yeah it leaks alright however as of right now I haven't figured just how to put the o-rings on as for the rest of surface a gasket sealer enough?
As for the galvanic effect if I should be ever as misfortunate of having to use this prototype I go and bury a zinc block in a flower pot with a wire leading to the aluminum block just like they do with metal bridges that way the zinc corrodes first :rofl: Just kidding although that would work, I just didn't see myself spending 20 bucks on a copper plate until I had the design right

JB Weld..
yeah but then if it leaks I'm up sheets creek cause then I wouldn't be able to fix it.
thanks to all of you any more suggestions?
terramir

[terramir]

another question where can I get a copper plate cheap, cause I really want to start on another bottom plate when I get the kinks worked out, BTW need some more input about what I should do with this.


Have another design in mind as well but I have to wait for some tools to arrive from england, Tap and die 9/32" (very rare size) which will allow me to thread the outside of this piping cause I came up with another design that would involve 4 or 5 intakes and outputs with this 1/4 inch copper piping. However this design would be completely in-block so there would be no window and no possibility of leaks. However that is for a future post once I get the tools an aluminum model will be only a week away :D

terramir

[NaeKuh]

Why u be trollin?

seems to me like he wants to learn how to make blocks.

He made a simple mistake of not having an Oring.

We teach first, then laugh later if they dont listen to what we teach.
Never the other way around.

[Sparky]

Yeah it leaks alright however as of right now I haven't figured just how to put the o-rings on as for the rest of surface a gasket sealer enough?
As for the galvanic effect if I should be ever as misfortunate of having to use this prototype I go and bury a zinc block in a flower pot with a wire leading to the aluminum block just like they do with metal bridges that way the zinc corrodes first :rofl: Just kidding although that would work, I just didn't see myself spending 20 bucks on a copper plate until I had the design right

Maybe putting a thin bead of silicone sealant all the way around near the edge (on the inside of the bolts, not outside) before putting the top plate on would work. True, every time you take the block apart you'd have to reapply, but it would seal it.

And no worries about the material - like I mentioned, using the cheaper aluminum is fine for design work, just wanted to warn you about long-term effects :)

[terramir]

seems to me like he wants to learn how to make blocks.

He made a simple mistake of not having an Oring.

We teach first, then laugh later if they dont listen to what we teach.
Never the other way around.
so that guy was insulting me :( well let him know I have oc'ed since the amd k6 processors, till now I have not had a computer sans laptops on stock speed, but I hit a wall on silent oc'ing on air which is what my specialty is. so I decided to learn more about water cooling, the pump noise is what shy'ed me away until now. now it's time to tackle that issue.


Maybe putting a thin bead of silicone sealant all the way around near the edge (on the inside of the bolts, not outside) before putting the top plate on would work. True, every time you take the block apart you'd have to reapply, but it would seal it.

And no worries about the material - like I mentioned, using the cheaper aluminum is fine for design work, just wanted to warn you about long-term effects :)
I doubt that it would work at the present state I got a sealing problem on the sides due to adding o-rings now it's sealed on top but gushing from the side especially the side where the outlet pipe is closest too.
more hints and tricks please.
terramir

[asura]

When I first started looking at LC (about six months ago and I still haven't got it all together) I figured I'd build my own blocks... I might still have to do it for GPU, but I went ahead and bought a CUP block. However I did quite a bit of basic planning and researching about what to do and how to do it.

For milling out a channel for an O-ring I'd planned on using something like this (http://www.cooksongold.com/Jewellery-Tools/Burrbusch-1-Round-2.1mm-prcode-984-110) and it doesn't matter whether you mill out of the base or the top, just don't go too deep and keep a (mostly) uniform depth otherwise it won't make a good seal.

G1/4" BSP has become so universal in LC (even ThermoChill have acquiesced with their new TA's) that I saw little or no point in working in anything else so a drill and tap are essential, though no need for a die as I'd planned on using standard fittings that have O-rings on them as standard, removing another possible leak.

I'd also decided that delrin was the way to go for tops, I'd be much less likely to crack it or blemish it, and if i did blemish it it wouldn't show as much as on a clear acrylic.

Ooh, and I'd have the base as just the thermal interface, not doing any structural work, either have the top as the hold down, e.g. EK supreme le, or have a separate hold down plate e.g. Ybris black sun.

[NaeKuh]

recess the oring location like so:
http://i125.photobucket.com/albums/p73/aigomorla/Water%20Blocks/IMG_1105.jpg

[Waterlogged]

G1/4" BSP has become so universal in LC (even ThermoChill have acquiesced with their new TA's) that I saw little or no point in working in anything else so a drill and tap are essential, though no need for a die as I'd planned on using standard fittings that have O-rings on them as standard, removing another possible leak.

+1

Your going to need a 11.8mm drill bit (http://www.mcmaster.com/#2958a157/=8rddf9) and 1/4BSPP tap (http://www.mcmaster.com/#8328a22/=8rddsc).

[terramir]

For milling out a channel for an O-ring I'd planned on using something like this (http://www.cooksongold.com/Jewellery-Tools/Burrbusch-1-Round-2.1mm-prcode-984-110) and it doesn't matter whether you mill out of the base or the top, just don't go too deep and keep a (mostly) uniform depth otherwise it won't make a good seal.

G1/4" BSP has become so universal in LC (even ThermoChill have acquiesced with their new TA's) that I saw little or no point in working in anything else so a drill and tap are essential, though no need for a die as I'd planned on using standard fittings that have O-rings on them as standard, removing another possible leak.

I'd also decided that delrin was the way to go for tops, I'd be much less likely to crack it or blemish it, and if i did blemish it it wouldn't show as much as on a clear acrylic.

Ooh, and I'd have the base as just the thermal interface, not doing any structural work, either have the top as the hold down, e.g. EK supreme le, or have a separate hold down plate e.g. Ybris black sun.
Ehh I was designing in fact in a way that I wanted to eliminate sources of leaks, discovered yes I'm gonna need an o-ring as for that tool it's a standard dremel tool I got like a half a dozen somewhere. as for the base not doing structural work what exactly do you mean, cause something has to hold the top to the bottom and somehow I also need something holding it to the cpu.


recess the oring location like so:
(photo does need to be quoted)
question in that block how are the tap's o-rings placed that is something I need to know can I get a pic of that please

+1

Your going to need a 11.8mm drill bit (http://www.mcmaster.com/#2958a157/=8rddf9) and 1/4BSPP tap (http://www.mcmaster.com/#8328a22/=8rddsc).
ehhh Actually I'm going with 9/32 and a 6mm drill to do direct 1/4" copper pipe connections, however I still need to look at how to do the o-rings for the input output cause I screwed that up. next aluminum test block will take that into account
Need to find the right size o-ring before I continue designing though.
Thanks for all the help
grateful for any more hints
terramir

[NaeKuh]

Well since u have a recessed oring.

When you clamp on a flat top, it will make a perfect seal, unless u got water going though the recess.

Then you would need whats called a three piece block.

Like a three tier cake, a three piece block has 3 levels.

[Sadasius]

I like to give people the benefit of the doubt and give them an 'A' for effort in trying anything out of their element in life. I am going to go out on a limb here and suggest buying a block instead of making one. :yepp:

[Waterlogged]

ehhh Actually I'm going with 9/32 and a 6mm drill to do direct 1/4" copper pipe connections, however I still need to look at how to do the o-rings for the input output cause I screwed that up. next aluminum test block will take that into account
Need to find the right size o-ring before I continue designing though.
Thanks for all the help
grateful for any more hints
terramir

It'll be much safer to use something already tested and in wide use than what you have in mind. It'll even offer better flow rate compared to that small diameter stuff, which translates to better temps.


Now having said all that, I do have to echo what Sadasius said.

I like to give people the benefit of the doubt and give them an 'A' for effort in trying anything out of their element in life. I am going to go out on a limb here and suggest buying a block instead of making one. :yepp:

+1

[Martinm210]

Nice!

For testing purposes, you could just use a thin layer of clear silicone as well. I did that with one of my old CPU blocks (http://www.xtremesystems.org/forums/showthread.php?t=162372&highlight=martin+thin+line) I made...worked fine.

[terramir]

Nice!

For testing purposes, you could just use a thin layer of clear silicone as well. I did that with one of my old CPU blocks (http://www.xtremesystems.org/forums/showthread.php?t=162372&highlight=martin+thin+line) I made...worked fine.
Hey got any of those still laying around that would fit a 775 I'd take it off your hands, would cool my qx9650 nicely :p:

Oh, well ran some tests with o-rings and well also silicone for the intake and outlet and silicone all around for sealing it. have a big o-ring as well but haven't dremeled out the channel for it yet. as for things that come after the fact it's gonna be a close one. squeezing it between that one pressure screw and the outlet pipe is gonna be close and the channel is gonna look like an inkblot just to make it fit :rofl:.
But here's my progress, the first picture is the test bench with a 2L/min windshield wiper pump.
http://www.ephotobay.com/image/1-03test.jpg (http://www.ephotobay.com/share/1-03test.html)
the second is of the block up-close:
http://www.ephotobay.com/image/1-03close.jpg (http://www.ephotobay.com/share/1-03close.html)
it's not leaking as far as I can tell but the real test will be after I channel it out and use a much stronger pump to test it, gonna get a fuel pump on sunday from a buddy to test this with.
any more thought also suggestions are always welcome.
And yes the tools are sort of expensive; however I want to learn so yes I'm willing to spend some money for tools I will use over and over again because those tools can also be used for heat pipe design which is another project I want to take on after I just saw the inside of a dell heatpipe - copperwires as wicks not mesh hmm I can do that. However whatever ends up in my pc in the end I dunno yet.
terramir