Sprung a leak

Associate
Joined
19 Jul 2011
Posts
2,370
Rigs been unchanged for a year. Built 3 years ago at least, since then its only had a 2nd gfx card added.

P1020492.jpg

(It's nowhere near that clean now!)

Noticed yesterday that the level in the reservoir was low. Low enough to draw bubbles into the loop which made the rig noisy. That's when I spotted the puddle sat on top of gfx card 1 and another smaller one on gfx card 2.

Powered off immediately. Grabbed kitchen roll and mopped everything up. Checked all the hose fittings etc. - all nice and tight.

Powered up briefly and swept the inside with the torch to spot the leak. Top radiator is dripping from the middle. There is a benefit to never cleaning the inside - all the dust makes it easy to spot a clean patch!

So for now, everything is powered off and is staying that way until the weekend.

Guess I'll take everything out of the case and test the loop without powering up any components. I've got a PSU bridge thing to get the pump online.
I can survive on one radiator for a while.

So fellow watercoolers - should I rinse down the gfx cards with deionised water, or just dry them out as they are?
 
I'm just running de-ionised water as my coolant. Been in there a while mind you. I have a new load ready to put in.

Good call mate
just leave them too dry
what gpus ?
Do you have a spare gpu

2x 6950's.
One is a VTX with a generic gfx block, the other is with a full cover block and a backplate. I believe the backplate has probably protected that card from any water issues.

I have a 7990 to put in their place, but I'd rather have the full compliment of radiators working before I do.

Radiator sprung a leak?

It looks like its the rad. The drip is in the centre, well away from the compressions etc. I could be wrong, but 10pm after a night out wasn't the time to do a full on inspection.

I'd be looking to take it out if that's the case!
Oh don't worry - I will be taking it out.
For now its going to sit unplugged til I can look at it this weekend.

Bit of a shame as I was really getting into my XCOM:LongWar :)
 
If I find it is, I'll post a pic. I'll admit I'm confused too. Not like I've been putting fans in and out and caught it with a self-tapping screw recently.

Course, I could be completely bananas, and one of the O-rings on a compression has failed or something.
 
i would take it out and put an end cap on on of the g1/4 and then a fitting plus tube on the other. put it under water in a bowl and blow through the tube. a bit like finding a bike inner tube puncture really.
 
I think it looks like a loose fitting mate
While you're in my house would you mind hanging out the washing.

On a more serious note, I'd like it to be a loose fitting as that will mean I don't need to strip down or swap out a radiator.
 
i would take it out and put an end cap on on of the g1/4 and then a fitting plus tube on the other. put it under water in a bowl and blow through the tube. a bit like finding a bike inner tube puncture really.

I like this idea. It'll give the missus something funny to watch too.
 
hmmmm, when you can figure out where the leak is coming from, grab a picture please, circling the area of the leak, ive just seen similar happen (not to me), but i want to see if its in the same place, if so il forward it to XSPC :)
 
i wouldnt like to say, ive never had a problem with them, but its possible i guess! i wouldnt like to scare people if there isnt, XSPC has been fantastic to me and ive not once had a failure and ive used 20+ xspc rads, i have 3 sat next to me as i type this lol.

its just i was reading a thread right before this one that had a similar problem, so thought id see if there was any correlation
 
If your radiator has not been completely full, it may be corrosion, oxygen and a warm damp atmosphere even with deionised water over time can cause corrosion particularly at soldered joints. Copper salt deposition will create a weak acid and if left will eat through thin copper. Poor defluxing may also create problems.

Try some PH testing strips in the reservoir and see if it changes from PH7 which is neutral.

However if it is an O-ring, water may be travelling along the side rail to drip onto the card.
 
Last edited:
I wouldn't say its a bad batch - the rad has been fine for 3 years now. Besides I don't know for sure the rad is faulty, that's just my initial inexpert diagnosis. It could be anything in the loop at this point.

Paul at XSPC has always given me really good customer service before - he repaired a pump I borked free of charge. If the rad has gone, I'd email him for advice, but after 3 years, I think I've had my moneys worth.
 
If your radiator has not been completely full, it may be corrosion, oxygen and a warm damp atmosphere even with deionised water over time can cause corrosion particularly at soldered joints. Copper salt deposition will create a weak acid and if left will eat through thin copper. Poor defluxing may also create problems.

Try some PH testing strips in the reservoir and see if it changes from PH7 which is neutral.

I'll pop into boots and see if they have them then.
Certainly an interesting idea.

However if it is an O-ring, water may be travelling along the side rail to drip onto the card.

Still prefer its this though :)
 
Would be interested to see if you can trace it, never seen a radiator just spring a leak!

Three years of running on nothing but DI is more than enough time for corrosion to eat it's way out the rad, the tin/lead flux used to weld them together is almost as susceptible to galvanic corrosion as aluminium.

This article is over two years old but explains it all very well: http://martinsliquidlab.org/2012/01/24/corrosion-explored/
 
Last edited:
Yeah, they have been great with me too :D


the other rads were 3 years old as well, which is what made me think, tbh, if my rads last me 3years + id be happy!!!

aye, its probably not the same thing, but just a coincidence 2 threads in 2 days both with a similar story :D
 
Back
Top Bottom