- Joined
- Nov 22, 2008
- Messages
- 1,506
- Points
- 48
In 2 words: Pump failure.
I've been running a water-cooling setup for quite some time now, about a year or so. Recently however, my pump had been tripping the "cpu fan speed" warning on my motherboard, but if I checked in the hardware monitor, it all appeared to be fine and the CPU temperature was low, so I ignored it (take note: this is a mistake!). I came back to my computer a few hours ago to find the screen blank and not responding at all. After a couple of times cycling the power, I got the side off, to find my graphics card dripping, and a nice pool of coolant on the case floor.
I don't use a fill-cap - the top of my loop is pretty much "open" inside the case. What I think has happened is that the pump must have jammed, and the processor started boiling the water in the block. My system is pretty close to being a convective loop, so the heat spread to the entire top half of the loop, causing the coolant to expand. This caused it to overflow, giving my graphics card a bath. Luckily, my cooling fluid is non-conductive, so with a bit of luck the graphics card should have survived. The pool also just missed my power supply and only splashed a tiny bit on the top of the hard disks, again very lucky!
What I am worried about, however, is why the thermal protection on the CPU didn't kick in. The only reasons I can think of, is that it didn't reach the 85 or 135 degrees necessary because of the coolant already in the block, or that the chip has now committed suicide. The thermal paste (AS5) appeared to have run a small amount towards the bottom, but hasn't burned, and when I got to the water-block, it was too hot to touch, but not enough to scald me. Neither the CPU nor block show signs of the heat "lines" you see on soldering irons, though the CPU top was lapped down to copper only.
What do people think my chances are of the processor having survived? I'll be able to test it tomorrow, and I was planning to upgrade anyway, but it'd be a real kick in the teeth if I can't sell it to contribute to the new parts
Pics will follow tomorrow, under PHR-light since Feser One (coolant) is UV-reactive as well as hopefully non-conductive.
I've been running a water-cooling setup for quite some time now, about a year or so. Recently however, my pump had been tripping the "cpu fan speed" warning on my motherboard, but if I checked in the hardware monitor, it all appeared to be fine and the CPU temperature was low, so I ignored it (take note: this is a mistake!). I came back to my computer a few hours ago to find the screen blank and not responding at all. After a couple of times cycling the power, I got the side off, to find my graphics card dripping, and a nice pool of coolant on the case floor.
I don't use a fill-cap - the top of my loop is pretty much "open" inside the case. What I think has happened is that the pump must have jammed, and the processor started boiling the water in the block. My system is pretty close to being a convective loop, so the heat spread to the entire top half of the loop, causing the coolant to expand. This caused it to overflow, giving my graphics card a bath. Luckily, my cooling fluid is non-conductive, so with a bit of luck the graphics card should have survived. The pool also just missed my power supply and only splashed a tiny bit on the top of the hard disks, again very lucky!
What I am worried about, however, is why the thermal protection on the CPU didn't kick in. The only reasons I can think of, is that it didn't reach the 85 or 135 degrees necessary because of the coolant already in the block, or that the chip has now committed suicide. The thermal paste (AS5) appeared to have run a small amount towards the bottom, but hasn't burned, and when I got to the water-block, it was too hot to touch, but not enough to scald me. Neither the CPU nor block show signs of the heat "lines" you see on soldering irons, though the CPU top was lapped down to copper only.
What do people think my chances are of the processor having survived? I'll be able to test it tomorrow, and I was planning to upgrade anyway, but it'd be a real kick in the teeth if I can't sell it to contribute to the new parts
Pics will follow tomorrow, under PHR-light since Feser One (coolant) is UV-reactive as well as hopefully non-conductive.