Apogee GT leaked? Or did it?

My friend recently had his CPU lost to the water. He had an Apogee GT on it and upon removing it, discovered there were really small "puddles" of water on the CPU and on the socket itself. After carefully inspecting the Apogee block, no leaks were found. The barb area was dry, the copper base was dry, no leaks at all.

The block was disassembled to be bowed by the way, but was assembled back in a proper fashion. I am totally stumped. How could the CPU and socket be wet, but the Apogee does not seem to be leaking.

Any thoughts?

Was the radiator in sub-ambient temperatures? If the water was far enough under ambient to cool the CPU area under ambient, then there could have been condensation. I highly doubt that happened though.

Were there air bubbles in the loop? If there were then it was definitely a leak.

The only thing I can think of is that the barb might have leaked either where it screws into the block or possibly at the barb/tubing junction. I had that problem before with silicone tubing; if you didn't crank down on the hose clamps on it you would get some seepage past the tubing/barb junction.

SolidxSnake said:

Was the radiator in sub-ambient temperatures? If the water was far enough under ambient to cool the CPU area under ambient, then there could have been condensation. I highly doubt that happened though.

Were there air bubbles in the loop? If there were then it was definitely a leak.

Click to expand...

No the radiator was not chilled.

He said there were air bubbles in the beginning but after he bled it out he is fairly certain no air bubbles surfaced again. Also, I am just baffled how it could leak if everything on the block is dry. This happened about 2-3 weeks ago and the Apogee today is still dry. If it leaked before, shouldn't it still be leaking now?

muddocktor said:

The only thing I can think of is that the barb might have leaked either where it screws into the block or possibly at the barb/tubing junction. I had that problem before with silicone tubing; if you didn't crank down on the hose clamps on it you would get some seepage past the tubing/barb junction.

Click to expand...

Well again, the barb area was dry. The tubing used was 7/16'' ID tubing on 1/2'' barbs so normally they would already form a tight fit with each other. For added security though, clamps were installed to the last click. If it leaked there, how would it get into the socket anyway?


EDIT: sorry for double post, meant to edit the top one instead.

If your buddy is using 7/16 tubing along with clamps, then I would rule out leakage at the barb/tubing junction for sure.. You might try to see if the barb/ block area is leaking by turning the pump on and then put a little pressure on the side of the barbs to see if you get any leakage past the o-rings on the barbs (try to wiggle the barbs in the waterblock). But I'm just taking a guess there as I don't own an Apogee GT to see if this could be happening. As for water getting into the socket, it could possibly be running down the side of the mounting bracket and then just kind of following the metal retention bracket that holds the proc in place. Or, if one of the barbs isn't snugged down tight, you could be getting an occasional small jet of water out of the threads that shoots upwards and towards the board and it then runs down to the socket. I would think your buddy would have to be using a pretty powerful pump to have enough pressure to that though.

Are there any watercooling parts higher than the CPU block? If so, something higher up could have leaked and dripped down, of even ran down the tubing to the CPU block/socket.

water parts leaks more when cold
parts get warm expand no leak
water leaves marks stains
look good with lite for stains

muddocktor said:

If your buddy is using 7/16 tubing along with clamps, then I would rule out leakage at the barb/tubing junction for sure.. You might try to see if the barb/ block area is leaking by turning the pump on and then put a little pressure on the side of the barbs to see if you get any leakage past the o-rings on the barbs (try to wiggle the barbs in the waterblock). But I'm just taking a guess there as I don't own an Apogee GT to see if this could be happening. As for water getting into the socket, it could possibly be running down the side of the mounting bracket and then just kind of following the metal retention bracket that holds the proc in place. Or, if one of the barbs isn't snugged down tight, you could be getting an occasional small jet of water out of the threads that shoots upwards and towards the board and it then runs down to the socket. I would think your buddy would have to be using a pretty powerful pump to have enough pressure to that though.

Click to expand...

I don't know, a complete leak test was performed upon setting up the loop and there was a leak on the Apogee but that was fixed. Another 24 hour leak test was performed and there were no signs of leaks. It would be pretty scary for me, and fellow watercoolers I imagine, if even after thorough leak tests, popular and reliable blocks such as the Apogee still have the chance to leak and destroy our expensive parts.

The pump is also a D5, I doubt the geyser effect you mentioned occurred.

Daemonfly said:

Are there any watercooling parts higher than the CPU block? If so, something higher up could have leaked and dripped down, of even ran down the tubing to the CPU block/socket.

Click to expand...

Well yes, the radiator was above the socket area. I wish I could provide a pic but I don't have one anymore. The radiator is mounted externally on top of the case but it could have leaked through the fan guard it mounted on top of I guess. That is unlikely though because again, a thorough leak test was performed. And for the radiator water to actually get into the socket area, it would have to leak through a fan guard, through the fan itself, hit just the right spot on the Apogee, and fall into the socket? I mean, it wasn't just one drop, like I said there were small puddles.

Also about running down the tubing, it would have to run down about 1.5 feet, than run up another 1.5 feet to reach the Apogee barb area from the radiator. Gravity defying AFAIK.

johnm said:

water parts leaks more when cold
parts get warm expand no leak
water leaves marks stains
look good with lite for stains

Click to expand...

I'll tell him to look closer but upon initial inspection no obvious stains were to be sighted. I will tell him to observe it closer though.


Thanks for the replies everyone. It isn't really a big deal now I guess, since the CPU is already dead, but I just find it baffling and a little scary. Hopefully I will discover what is the cause of this so I can possibly put my frights to rest.

An update, my friend observed the Apogee GT very carefully, and no stains are to be found.

The loop has been running (pump on) for a week now and no leaks at all. The radiator has been inspected (fins + barb area) and no leaks. This is really really baffling.

Maybe he spilled some coolant while filling the loop and it got into the socket area. The only other explanation would be that his loop was subambient and caused some condensation somehow.

muddocktor said:

Maybe he spilled some coolant while filling the loop and it got into the socket area.

Click to expand...

That one gets my vote. You can always count on humans to make mistakes.

If anyone's worked with plumbing a lot, especially compression fittings, you'll know that newly installed parts will sometimes weep slowly, and seal up after a day or two. I would find it hard to believe that this would happen a while after being installed though.

Only other thing might be a pinhole in the tubing higher up, and it somehow ran down the tube and WB perfectly to collect in the socket. BTW did he install it wet or dry...filled w/coolant before or after mounting?