awsome story

M

Modder man

[H]ard|Gawd
Joined
May 13, 2009
Messages
1,770
So today while working on my computer my pump decided to stop working for a short period of time. I noticed when the monitor shut off but the computer stayed on. I looked over an realized there was no water moving in the reservoir. I then procedded to feel the waterblock and it was so hot it burned me. Right after this the pump decided it wanted to come back on and push all the hot water into the reservoir. Little thing I know about hot water is that it likes to expand. Problem with that is that this caused me reservoir to explode and spill hot water all over my motherboard and ram. So in short my whole computer fried itself today. :mad:
 
Big time bummer.

Could this have been avoided?
 
Lets see some pics. What you describe doesnt make sense to me. ..
 
MrGoat said:
Lets see some pics. What you describe doesnt make sense to me. ..
Click to expand...



The part about him burning himself on the cpu block sounds crazy. I would think the MB bios would shut down the system at an unsafe temp to help prevent this from happening.
 
SonDa5 said:
The part about him burning himself on the cpu block sounds crazy. I would think the MB bios would shut down the system at an unsafe temp to help prevent this from happening.
Click to expand...

You can honestly get burned with just the normal high temps on some things.

Had the pleasure of touching a heat pipe on one of my GTX460s while it was running hot folding.

Left a long red line. :(
 
it wasnt like a hardcore burn but it definitely burned me. I have one of the EK tube reservoirs. I dont really have a good way to take pics and there is not a whole lot to see.....just fried hardware
 
SonDa5 said:
The part about him burning himself on the cpu block sounds crazy. I would think the MB bios would shut down the system at an unsafe temp to help prevent this from happening.
Click to expand...

thats what I would have thought and is what really makes me mad. it should have shut itself down long before it got to that point
 
Last edited:
It sounds like your loop was faulty to begin with if the expanding hot water was able to burst the reservoir. I don't know what you did wrong, but this should never happen. Everyone running a water cooling setup has to deal with the water going from room temperature to increased temperatures and I've never heard of a reservoir bursting.

Did you disable the safety features of your motherboard as well? I've had a pump failure in the past and within seconds of powering on the computer it shuts itself down if the water isn't moving.
 
this is true this is what should happen but did not....there is also a huge difference in the temp of my water than a normal loop at that point...my water is usually just a few degrees over room temp. yesterday it was boiling so fact that u say we all have to deal with chaing water temps is true but, this is not the reason my res exploded
 
One theory would be that while the water was in the cpu block, it superheated, and due to pressure, it didn't boil. When the pump kicked up and the water reached the reservoir, it met reduced pressure, boiled, increased pressure on the reservoir and caused it to explode. However, there are several problems with this theory. First, pressure should have been equal over the entire system, due to water's inability to be compressed. Second, this would mean the processor had to be running near boiling point or above it (100°C or higher), which doesn't seem very likely.

My guess is that the reservoir wasn't able to cope with the sudden change in temperature, and shattered. When the hot water came in, it heated one part of the reservoir rapidly. The resulting expansion caused stress on the reservoir as one part expanded and the other parts didn't, and when the stress became high enough, it shattered. Or it simply wasn't built to tolerate high temperatures.
 
Sorry to hear about your lose.

someone getting a hold of EK?

Maybe to much water in the system? You should always leave room, no reason to fill it to the very tip top
 
I had an issue with a bad sata cable, and during the troubleshooting had somehow left my pump molex disconnected. CPU survived, but one of my HD4870s didn't make it. Started melting the block (maze 4) and leaked onto the bottom card (which still worked after cleaning) so yeah things can get pretty hot!!
 
the processor had to have been over 100c for the water to boil.....also the res did not explode as into a million pieces, it just burst at all the joints where it is screwed together or otherwise
 
Last edited:
I had this happen to a generic (but sturdy - I tested it to 30+ PSI, and it had seen heavy use) acrylic tube reservoir (just testing with HOT HOT fluids outside of the case). I was actually trying to get a superheated fluid in there...

That tube reservoir failed - the seams leaked, and the top of the cylinder developed a small crack. The water would leak only very slowly... I've ordered another acrylic tube.

When I purchased the new tube, the guy on the phone at a local plastic-supply shop thought I was building a penis pump, apparently because I wanted a threaded hole at the top :confused: - funny... must have some strange customers :eek:

Sorry for your loss :(
 
well got a response from AMD finally, got a big F you. your on your own have a nice day

so i guess im just out a few hundred in hardware
 
when computers overheat they are supposed to shut down before they have a meltdown......mine did not
 
So you drown your PC with water, and suddenly its AMD's fault? Where's a ROFLCOPTER when you need one?


I've never owned a motherboard that would shut down if overheated. IIRC, CPU's simply throttle clock speeds and voltages when trying to reduce temps, but they never actually turn off. Even with reduced voltages, your CPU would still be resting at an insanely high temp if no water was running through the block, especially on modern blocks that hold very little water volume.
I have heard of cases where pumps failed, but the system still managed to work fine due to a convection current developing inside the water loop...which created just enough flow to keep things below the danger level. That said, I've never heard of this happening with modern impingement style water blocks that are really restrictive.

As for the plausibility of this story, I find it hard to believe...but I have no reason to doubt you. The res would fail if it was heated, especially if a bit of pressure was involved. This would be compounded if the res was overfilled...i.e, no air pocket on top to allow for thermal expansion. It sucks either way, but it was a term you agreed to when you chose to cool your PC with water.

If tomorrow a leak takes out my entire computer, the first person I'm going to blame is myself...because I'M the one who put 16oz of water inside my computer, not Intel, Gigabyte, or any of the other brands found inside my case. Perhaps the only exception would be liability on the part of whoever made the failed water cooling part. That said, how do you prove it was their shitty part, and not simply user error?
 
it was dead before the water even hit it....i am willing to take the risk of the water, it was long gone before then, and yes almost all computers have a thermal shutdown
 
Which pump/top was this that decided to stop working and then come back to life again? :mad: Some people like to hook their pump rpm cable to their board or monitoring device (aquaero) so if the pump stops the comp will shut down automatically.
 
Modder man said:
when computers overheat they are supposed to shut down before they have a meltdown......mine did not
Click to expand...


UMMMM WHAT? You are talking about a AMD here :p

But either way even if it did a company isn't going to warranty a board that has been abuse or had water dumped on it ... PERIOD! and I don't blame them

Bad ConNecTioN said:
So you drown your PC with water, and suddenly its AMD's fault? Where's a ROFLCOPTER when you need one?


I've never owned a motherboard that would shut down if overheated. IIRC, CPU's simply throttle clock speeds and voltages when trying to reduce temps, but they never actually turn off. Even with reduced voltages, your CPU would still be resting at an insanely high temp if no water was running through the block, especially on modern blocks that hold very little water volume.
I have heard of cases where pumps failed, but the system still managed to work fine due to a convection current developing inside the water loop...which created just enough flow to keep things below the danger level. That said, I've never heard of this happening with modern impingement style water blocks that are really restrictive.

As for the plausibility of this story, I find it hard to believe...but I have no reason to doubt you. The res would fail if it was heated, especially if a bit of pressure was involved. This would be compounded if the res was overfilled...i.e, no air pocket on top to allow for thermal expansion. It sucks either way, but it was a term you agreed to when you chose to cool your PC with water.

If tomorrow a leak takes out my entire computer, the first person I'm going to blame is myself...because I'M the one who put 16oz of water inside my computer, not Intel, Gigabyte, or any of the other brands found inside my case. Perhaps the only exception would be liability on the part of whoever made the failed water cooling part. That said, how do you prove it was their shitty part, and not simply user error?
Click to expand...

I'm with this guy.

Modder man said:
it was dead before the water even hit it....i am willing to take the risk of the water, it was long gone before then, and yes almost all computers have a thermal shutdown
Click to expand...

you didn't have it setup in BIOS then or you would not have this problem, or it failed, either way you took on a risk and it failed and you paid.

Sorry for your lose, I am not trying to be a jerk but I am trying to make you realize it is YOUR FAULT... no one else and I think everyone here agrees and does still feel bad about your situation you put YOURSELF into.

Time to move on and start over... next time take more time and make sure this doesn't happen again if you do happen to go the same route. I have ran 4 completely different computers on water cooling and never had an issue, lucky me? no just take extra caution.

The pump failed too right? why didn't you blame swiftech?
 
there was no extra caution to be taken....and what good is blaming swiftech going to do. yes there was an issue with thier hardware but it is now working so i cant prove that to them. nor do i feel i that i have any chance in convincing them to pay for the other parts
 
Last edited:
Modder man said:
it was dead before the water even hit it....i am willing to take the risk of the water, it was long gone before then, and yes almost all computers have a thermal shutdown
Click to expand...

Where did you ever say the computer died before the water hit it? All you said was "the monitor shut off but the computer kept running". Computer kept running huh? Sure sounds like it failed...

AMD lists the maximum safe temperatures for the phenom 2 chips at 62c. They gave you a heatsink that they designed and tested to keep it cool. It was your choice to forgo using their cooling solution. At this point, any responsibility after a heat related failure is on you, not them. YOUR pump failed, and it was YOUR pump that was responsible for cooling the chip. Not AMD's problem...its yours.
There is a good chance that had you shut down the system, let it cool, it would have booted right back up. But before that happened, it got doused in water...which pretty much guaranteed a hardware failure.
Additionally, regardless of the presence of thermal control, there is free software that would allow you to monitor temperatures right on your desktop, and if necessary, sound an alarm. It seems reasonable that if you're going to use a cooling solution not approved by AMD, that such a monitoring program might come in handy.

Swiftech isn't responsible for you're hardware...just the pump. They warranty the pump, so if it fails...they replace it. But nowhere in their warranty do they state anything about covering your parts...actually, I'm pretty sure "use at your own risk" is probably printed in several places. So unless you can prove they were criminally negligent in the design and construction of the pump...chances are you'll get squat, but we seem to agree on that point.

As for extra caution...lets see. You throw away the recommended and TESTED cooling solution from AMD in favor of your own. You then proceed to circulate WATER in a small box filled with highly sensitive electrical equipment, even going so far as to place a reservoir FILLED WITH WATER right above your motherboard.

Ya, I'd think extra caution is in order. So lets see what you could have done to avoid this issue.
  • Research thermal throttling on AMD chips, and how to adjust it on supported motherboards.
  • Go into your bios and enable the CPU fan failure warning option (lots of boards have them).
  • Use free software that can monitor system temps and alert you upon a cooling failure.
  • Place the reservoir at the bottom of your case
  • Leave an air-gap in the reservoir to accommodate for thermal expansion...ya, you heard that right. You need a little air in a closed cooling system to allow for expansion.


You wanted to get all of the benefit from watercooling, yet don't seem able to accept the risk. Sorry dude, but it doesn't work that way. You play with water and you get...well, you know...
 
Did you use tap water? I've leaked distilled all over the CPU/GPU before and nothing after a resevoir bulkhead proved leaky. I cleaned it off and restarted. Proper distilled water/coolant will not short out parts.

AMD boards don't have thermal shutdown, which is why i've never considered them as safe for overclocking (plus have a much lower thermal max). The CPU_fan header should always be used as on many boards it stopping will cause the board to shutdown.
 
Are there any temperature monitoring programs that can auto shutdown the computer past a set temperature threshold? I have a MCP655 so it shouldn't happen but it wouldn't hurt.
 
I dunno, my Rampage II GENE does shut down based on 4 sensors (NB, SB, two external sensor wires you can buy) and temps you set (default 100C, but can go down to some 60C, iirc).
 
Speedfan has a tempreture you can set to cause an automatic shutdown. But this isn't so much needed with intel boards, as they have one built in to stop themselves frying. AMD CPUs are famous for their lack of themal cutoff and will usually just cook themselves, I haven't had one for a while, but this might still be true. (watch old video here for some AMD fire) Speedfans tempreture readings are a little screwy however, so you might have to adjust for the low tempretures it gives.

The Rampage II GENE has an AMD CPU version? I would suspect higher end "overclocking" boards to all have this kind of feature.
 
DeathPrincess said:
Speedfan has a tempreture you can set to cause an automatic shutdown. But this isn't so much needed with intel boards, as they have one built in to stop themselves frying. AMD CPUs are famous for their lack of themal cutoff and will usually just cook themselves, I haven't had one for a while, but this might still be true. (watch old video here for some AMD fire) Speedfans tempreture readings are a little screwy however, so you might have to adjust for the low tempretures it gives.

The Rampage II GENE has an AMD CPU version? I would suspect higher end "overclocking" boards to all have this kind of feature.
Click to expand...

Nah, the closest thing is the M4A88TD-M/USB3, but it isn't remotely near the Rampage II GENE. AMD has had thermal throttling for quite some time. My Quad Opty 8530 setup switches off at 60C or so. And it's a 2007 system with 4 AMD quad cores, lol.
 
First off, AMD doesn't manufacture boards - they manufacture chipsets. Even if this accident was the motherboard's fault that would mean MSI in your case, not AMD. Every board I have used from Socket754 to AM3 has at least had a shutdown temperature setting in the BIOS based on the CPU socket temperature.

Everybody needs to check this setting when they overclock.
Sometimes it isn't enabled by default.:mad:
Sometimes it is enabled but set too high.:mad:

I always set mine at 60°C or 65°C for my AMD systems.
 
Were you using a pump top with the 655? If not I'd contact Swiftech about this and talk to EK about the res breaking under such high water temps. I've had an MCP355 pump failure and it basically warped the shit out of my tubing, but my res was fine (XSPC 5.25").

Do you have pictures of how the loop was setup?
 
Yea, if you dont have automatic shutoff and you run only 1 pump, it's only a matter of time before the pump fails. Whether its 6months, or 6 years, they dont last forever.
 
sok0 said:
Yea, if you dont have automatic shutoff and you run only 1 pump, it's only a matter of time before the pump fails. Whether its 6months, or 6 years, they dont last forever.
Click to expand...

same can be said about fans ;)

I have two fans in my case right now that sound like death, I will be replacing this weekend, they also hardly push any air and when new they were amazing, or so I thought
 
You must log in or register to reply here.
Back
Top