I voted slow Block leak.

My only real problem that I have had was over the course of 1 1/2 years of moving my pump around while still connected to the water block (bad idea, I know) the poly top of my TC-4 cracked and began a slow, unnoticeable drip onto the video card. On several different occasions, I would be in the middle of a game and the whole system would just lock and would refuse to POST. I would shut the system down and pull the video card out to discover that the whole AGP socket was filled with water and one component would be corroded. Clean out socket, clean up card, scrub off corrosion, dry out socket, and put everything back together and it was good to go for another 2 months or so.

Last time that happend though, the corrosion finally ate through a connector on the back of my R9700Pro and the card wouldnt work anymore. :( SO, ordered all the parts of my PCI-e system. Luckly, while I was waiting for the new video card to come, I was able to resolder the missing connector and the video card still works (I got extreamly bored :D )

Moral of the story: will never buy a poly top water block ever again :dome:

voted block slow leak.
I had a small leak in my homemade WB that over time left a green/blue crust ( antifreeze )
on my NB-heatzink and on top of my Geforce3, but no apparent damages to the system.
I just tighted the connection upp and wiped of the graphicsboard.
All still up and running :)

Had one small leak that I caused on my Eheim 1048, and once I had my 50Z stop working because of a loose molex, causing some local boiling in my CPU block.

Voted Other problem/not described.

Got green algae buildup this summer, with 5 months old distilled & WW3% mixture.
Must have been contaminated during refilling of T-line, even though I used a bit of cloth
soaked with antialgae toxins as a plug. This stuff took about 2½ weeks to grow. :eek:

All lines were coated with a thin green algae-film, so they went into the trashcan.
I didn't even try to manually rinse the hoses, as I would have to clean out both
the algae and the WaterWetter staining, too much job.
Couldn't detect any algae buildup in pump/block/rad, but poured in some of the algaetoxins
anyways.

Cut out some new hoses, rinsed every component with strong disinfectants to kill any existing germs/algae, took a shower, used disinfectants on myself :) , in order to get the whole assembly process as sterile as possible.

Now my coolant consists of distilled water, 2,5%WW and a few drops of Simazine 0,5%.
I also have got some waterbed anti-algae stashed away, in case I need to use something stronger... It's active compound is Alkylbenzyldimethylammoniumchloride... Extra strong mixture the bottle reads...

Water Fun!
 

I voted multiple incidents!

1.Pump blockage
I replaced an Old Aquaclear powerhead/big res. with a modded Mcp350/small res. (6800U is hooge!) My temps were abnormally high, I opened the res. to find a trickle flow.... Mcp350 got clogged by bits of Radiator crusties that the Aquaclear readily pumped!!

2.Hosebarb leak
I've had very minor drips from "not tight enuff" on both 9500np and the new 6800Ultra!!! drying cured both cards.

3.Radiator fan
One of my YL 120's stopped spinning, peeled-up the sticker, put more lube in, she's spinning like a champ! I was so suprised to find my shamefully cheap fans to be popular for WCing :)

I finally cracked my WCing open after 21/2 years of running.... Bio nasties, crazy galvanic corrosion re: WW Alu top, and chromed hose barbs. It had no leaks! Thanks to the O-ring sealed barbs.

Had them all and L-O-V-E it! Test until failure is really cool!

As for the home PC's, it's not fun at all. :) I'm all for WC'ing, but only use external WC'ing solutions. Less chances for leaks. :)

Stev

Well, I've been running my first watercooled PC for roughly 24 hours now and am starting to suspect I had my first disaster on assembly!
On first assembly of the system, I had a pretty major leak onto the video card (6600GT). Since the system only contained deionised water at the time and I dried the card out thoroughly before using it I wasn't expecting this to be a problem. However, when running using the NVidia drivers (it's a 6600GT) I get white or pink snow in various places on screen. This doesn't appear when using windows standard drivers (although these are chronically slow at refresh rate), and didn't appear before I installed the watercooling. Doesn't appear in Bios either. I've tried running an earlier version of the drivers (66.93) which is supposed to fix this problem, but I either can't get them to work or they don't fix the problem :(

When I doused a motherboard with coolant, after switching the machine off, I rinsed it with isopropyl alcohol (try your chemist) to drive as much of the water out as possible. I then dried the board thoroughly with a hair dryer for several hours, and then left it for 24 hours in a very hot room.

Without heat, you would need to let that card dry for a good number of weeks since PCBs are pretty absorbent and water gets in underneath sm components and chips and just won't come out.

That bad? Quite surprised. I've got tons of IPA (isopropyl alcohol) at work so can easily wash it in the stuff there no problem (even filtered IPA - it's surprising the amount of crud you get in normal IPA).
Edit: thinking about it, is having operated it while the PCB is slightly damp likely to have caused any long-term damage? Can't think of anything myself, but I could be wrong...

first problem happened last week
cracked a 2-342 rad connection, Iwaki MD-30 pump drained system onto & blew a DH800 mobo, 6800GT vid & PC P&C power supply

I never had a problem, 4 years and going strong, although I replaced my Eheim pump with a CSP-MAG. It never had any problems, I just wanted to go all DC-powered. It is beyond me how so many people has had problems with watercooling. To me it seems more reliable and maintenence-free even compared to standard aircooling. I cannot even imagine how watercooling can possibly have so many problems, thinking it is due to a lot of random circumstances rather than the idea of watercooling being inherently unreliable. In theory, and in my practice, watercooling if done right can be very reliable, quiet, and maintenence-free. In fact I didn't like the idea of refilling my res, so I built a humidity condenser to keep the res filled via moisture from the room air. Good stuff, you guys shouldn't have so many problems, really.

cotd, we are seeing only the problems in this thread, not the many happy watercooling people.

I've had ALL of the above... but all long ago. Back then stuff wasn't as reliable and refined as it is now. Bugs then have been ironed out in the same product now. Lessons have been learned. All the causes of my problems were either choosing cheap products, or not paying attention, or simply not knowing what I was doing... I'll take it from the top...

Eheim 1250 - slow leak at inlet due to incorrect thread used on a fitting of my choice - my fault.
Hydor L30 - complete failure - running on a TEC Rig submersed - Hydors don't like high operating temps. Lesson learned, wrong product choice.
NexCool GT3 - new product, testing sample, leaked from every seal. Returned.
DangerDen Maze3 - bad block top, quickly RMA'd and resolved.
Innovatek Rev3 - bad seal, quickly RMA'd and resolved.
DangerDen Maze3 GPU - but I'd removed the barbs prior and not put `em back right - my fault.
Swiftech MCW462 - not a clue what happened. Seal gave way. Had never been disassembled at all.
D-Tek LRWW - Screws weren't tight holding the layers together... hadn't checked before fitting, hadn't leak tested. Block had never been disassembled - I should have leak tested. Tightened up screws, all fixed and resolved. My own fault really. Never assume...
Crap hoseclamps (I hate nylon hoseclamps)
Never use zipties as hose clamps - lesson learned.
Maxxxpert Rad, Aquacoil rad, Senfu Rad... had all with leaks, but back in the later days of Maze2 era.
Yep - nut dropped off back of mobo. Again my own fault for not putting it together properly.
Frequently due to a faulty fanbus... replaced fanbus, problem solved.
Tube resting on ramchip of graphics card - hadn't realised. Slowly melted thru the tube and flooded the case, card etc whilst all powered up. Fuse blew in PSU. Left all hardware to dry out for a week in the airing cupboard - all worked fine. Lesson learned. Ensure tube is routed properly.
Senfu Reservoir - shattered once WaterWetter was added - lesson learned - senfu = crap res anyways, replaced res, made sure didn't use too much WaterWetter.

In all leak cases, soon as I noticed I powered down system. If leak occured when I wasn't around, fuse always blew in PSU. I tend not to flap under this kind of stress. Just kill the mains, mopup, dismantle system, leave what needs drying out to dry, replace what needs replacing, repair what need repairing. Always had a fully working system again within 7 days.


Nowadays however, I haven't had ANY problems like these... but I also only use relied-upon products with the reputation for not having problems.

Laing pumps, Swiftech & DangerDen blocks, Steel Jubilee Clamps, check every screw and fixing is tight on everything - block tops, barbs etc. Leak test out of case before mounting to any components. Mount in case on components but with no power other than to pump and bl00roll over common drip areas (back of VGA card for instance) so any leaks are fairly instantly and easily spottable. If all is clear, wire it up, fire it up.

I have a blacklist of brands which I'll never recommend to anyone, based on personal experience, either from a poor QC standpoint or just from a poor performance standpoint. Tend to find those not on that blacklist are the usual suspects that everyone already knows and respects for quality & performance.

just a few remarks.

I always leak test (also have a Dtek WW), no leaks
but, that is Dtek's quality issue, not your fault.
Buy a car, check every bolt?

Er. Been using them for years. No issues.
however i use very large ones, given the swiftech shipped as reference,
mine are about 4x larger.
More "permanent" areas get the zipties,
else gets those double snap-grip plastic ones, no issues yet
althou they don't really look that safe.

... can we see that blacklist in full?
should be interesting. :)

I voted pump failure, but the pump was only the immediate cause of failure.
I had a massive dose of some kind of very fine white precipitate form in my system one week. It fouled the water blocks, rad and pump something fierce. The pump eventually died ( I leave the pump on 24/7) when I wasn't around.
I still have no idea what the precipitate was, and it has not occurred since.....

No damage except to the pump, which was terminal.