watercooling advice needed

[artoodeeto]

ok guys, so as you've probably seen in my worklog, I'm about to create the world's first watercooled sandcrawler. But...a question first.
Does it completely defeat the purpose of 24-hour leak testing to drain the setup after the test so I can detach the waterblocks to make them easier to install on the motherboard/video card? Or is leak testing primarily to make sure the fittings all..well...fit, and don't leak?
I'm just thinking that it'll be awkward to try and put everything together if it's all connected by various hoses which restrict relative motion.
I consider myself a newbie at this; I once had another watercooling setup, but that was around 6 or 7 years ago, and I don't recall much from it.
Thanks!

[Trace]

I normally attach first. Then I fill, hook it up to a test psu (or jump the green wire on the 20/24 pin to a black wire to start it), let it run for a couple of hours with paper towel around fittings (underneath, etc). Just don't turn on the power of the actual PC, just the pump. It can get wet without damaging anything if the actual computer isn't on.

Any other questions, feel free to PM me.

Anybody else, if you do something different, or if I am wrong in some way, please, feel free to correct me!

[mDust]

I normally attach first. Then I fill, hook it up to a test psu (or jump the green wire on the 20/24 pin to a black wire to start it), let it run for a couple of hours with paper towel around fittings (underneath, etc). Just don't turn on the power of the actual PC, just the pump. It can get wet without damaging anything if the actual computer isn't on.

Any other questions, feel free to PM me.

Anybody else, if you do something different, or if I am wrong in some way, please, feel free to correct me!

That's how I do mine as well. I've had leaks...well, torrents...spill on a brand new graphics card with no ill effects since there was no power to it. It wasn't my fault! There was a tiny slice in the hose above it that wasn't really visible to the eye before the leak. I have no idea how it got sliced, but it wasn't by me. Anyway, if the computer had power at the time, it would have been a major FAIL.

If water does get spilled on anything, dry it off the best you can and then grab an air compressor and take your time blowing water out from under the various components. Then set it in front of a fan for a few hours just to make sure. It's better to be safe than sorry and computer-less.

If this makes you uncomfortable, then set it all up while unattached to the computer first. This is probably safer for a first time, but it's more of a PITA since your hoses may not be "twisted" properly and will have to be adjusted after attached anyway, defeating the purpose of leak-testing.

[artoodeeto]

That's how I do mine as well. I've had leaks...well, torrents...spill on a brand new graphics card with no ill effects since there was no power to it. It wasn't my fault! There was a tiny slice in the hose above it that wasn't really visible to the eye before the leak. I have no idea how it got sliced, but it wasn't by me. Anyway, if the computer had power at the time, it would have been a major FAIL.

If water does get spilled on anything, dry it off the best you can and then grab an air compressor and take your time blowing water out from under the various components. Then set it in front of a fan for a few hours just to make sure. It's better to be safe than sorry and computer-less.

If this makes you uncomfortable, then set it all up while unattached to the computer first. This is probably safer for a first time, but it's more of a PITA since your hoses may not be "twisted" properly and will have to be adjusted after attached anyway, defeating the purpose of leak-testing.

cool, thanks! My original water cooling system, I did dump a fair bit of water on the motherboard...it started leaking as I filled it (I was using barbed fittings, and thought the hoses were in all the way...they weren't). And yep, I used the air compressor to blow water out and paper towels to sop it all up. this time though I'm using a non-conducting Feser coolant, although I still want to avoid spilling it all over the computer.
I don't know if I'll have time to work on this today, but definitely will tomorrow. I think I've figured out how to mount everything, including dual reservoirs over the rad :)

[Spawn-Inc]

cool, thanks! My original water cooling system, I did dump a fair bit of water on the motherboard...it started leaking as I filled it (I was using barbed fittings, and thought the hoses were in all the way...they weren't). And yep, I used the air compressor to blow water out and paper towels to sop it all up. this time though I'm using a non-conducting Feser coolant, although I still want to avoid spilling it all over the computer.
I don't know if I'll have time to work on this today, but definitely will tomorrow. I think I've figured out how to mount everything, including dual reservoirs over the rad :)

just so you know that so called non-conductive coolant WILL become conductive within a week of being in your system. it will pick up copper ions and other metals in your loop over time and become conductive. that's purely a marketing scam if you will. as well be prepared to rip apart the rig in 6 months to a year and have a nice cloggy mess. distilled water and a biocide like .999 silver plated fittings or a silver kill coil or some pt nuke are your best bet provided you are not using any aluminum in your loop.


and to answer your first question, though it's been answered, yes leak test in your case hooked up. with no power to anything but the pump. if you disconnect the tubing you need to re test as that's the primary spot for leaks.

[artoodeeto]

just so you know that so called non-conductive coolant WILL become conductive within a week of being in your system. it will pick up copper ions and other metals in your loop over time and become conductive. that's purely a marketing scam if you will. as well be prepared to rip apart the rig in 6 months to a year and have a nice cloggy mess. distilled water and a biocide like .999 silver plated fittings or a silver kill coil or some pt nuke are your best bet provided you are not using any aluminum in your loop.


and to answer your first question, though it's been answered, yes leak test in your case hooked up. with no power to anything but the pump. if you disconnect the tubing you need to re test as that's the primary spot for leaks.

Good to know, thanks! I didn't pay anything for the coolant, so I'm not out any $$ on it. I'll have to get a silver coil ordered at some point here...

[Spawn-Inc]

thought i posted links to the silver kill coils. here are some links
silver kill coil (http://www.petrastechshop.com/sikibyia.html)

Silver Strip (http://www.jab-tech.com/Antimicrobial-Silver-Strip-pr-4530.html)

[artoodeeto]

I'm using side-by side dual danger den RAD reservoirs in my case, but instead of vertically mounting them the way they were designed, I have to mount them horizontally:

http://i597.photobucket.com/albums/tt53/artoodeeto/crawler%20update/th_10--04-04DSC_0008.jpg (http://s597.photobucket.com/albums/tt53/artoodeeto/crawler%20update/?action=view&current=10--04-04DSC_0008.jpg)

This is part of my case, with the roof removed, the reservoirs are replacing those two gray pipes in the picture:
http://i597.photobucket.com/albums/tt53/artoodeeto/crawler%20update/th_10--03-19DSC_0001.jpg (http://s597.photobucket.com/albums/tt53/artoodeeto/crawler%20update/?action=view&current=10--03-19DSC_0001.jpg)

and a shot of the mounted radiator in the only place it can go (not enough airflow/vertical space to cut a hole under the motherboard and mount it down there):
http://i597.photobucket.com/albums/tt53/artoodeeto/crawler%20update/th_10--03-21DSC_0005.jpg (http://s597.photobucket.com/albums/tt53/artoodeeto/crawler%20update/?action=view&current=10--03-21DSC_0005.jpg)
The loop will be pump - CPU - GPU - radiator - res - res - and back to the pump. I made a custom mount that has the reservoirs slightly angled with the in/out fittings on the down angle, and I can angle them more if need be. I plan on having the pump on the same level as the radiator, although I could mount it below, on the same level as the CPU, if it made a significant functionality difference.

What's the best way to fill the system, how much air space should the reservoirs have at the "top", and should I change the slight angle (probably around 5-10 degrees or so) to a steeper one? I'm not sure how steep I can go before I run out of room. I'm thinking that because I'm using 2 res's, I should fill one most of the way and one all the way? Heck, will any of this even work? Unfortunately I'm working within the limitation of a custom-built case that I didn't design for water cooling at all... hence the weirdness. :) any comments/suggestions/advice are appreciated! Thanks!!

[x88x]

Ideally, you want to not have any air at all in the reservoirs, especially in the horizontal positioning you have them in. I would recommend fitting in a fillport somewhere higher than any other part of the loop. If that's not possible, you could have it 'higher' if the whole thing were, say, turned on end?

[artoodeeto]

Yeah, the more I think about it, standing it up on end for filling sounds like a good idea. my first thought was since I have lots of extra tubing (the green stuff in my worklog), to use that and a plug I have for it, and just attach a short length to the end of one res and use it as a fillport. But, I don't really want to have a bit of hose sticking off the end of the res, it'd look silly.

Good to know about no air in the reservoirs, I've seen setups with air before and seen comments about how you should have some air in there so air bubbles would have a place to go. But if the air can be completely eliminated from the loop, and the res's filled all the way...then there shouldn't BE any air bubbles in need of a place to go right? :)

+rep to X88X for being incredibly helpful on my build!

[x88x]

Well, the ideal is always to have no air. That being said, it is best to have the highest point in the loop be a res or fillport so that as the system runs, air bubbles can be pushed through and eventually end up in a spot where it is easy to see them and fill the empty space. I know with my latest build I had air bubbles coming up for almost a month.

One thing I would recommend is to have the pump running when you are topping off the loop. The reason for this is that fluid moving through the loop will press out more empty space, so when it stops (if there is an opening) it will raise the fluid level, making it seem like the loop is more full than it actually is.

[artoodeeto]

good to know, thanks! wish me luck...gonna be filling and leak testing tonight come hell or high water...

*blink blink*

I probably could have picked a better phrase LOL

[Trace]

I was just going to say, watch what you wish for!