This is what happens when a D5 pump fails...

hakank68 · 12 Nov 2013 at 22:08

Repost= deleted

Addicted · 12 Nov 2013 at 22:17

I went the lazy route, but thanks for the reminder - have an Aquaero 5 XT - so just setup a continuous buzzer warning if my CPU hits 80C.

Should be enough imo.

AndyOcUK · 12 Nov 2013 at 22:21

Addicted said:
I went the lazy route, but thanks for the reminder - have an Aquaero 5 XT - so just setup a continuous buzzer warning if my CPU hits 80C.

Should be enough imo.

same

wazza300 · 12 Nov 2013 at 22:27

here,free too

Jasseh · 12 Nov 2013 at 22:32

The remote chance of this happening (and the cost) is what puts me off going for a custom WC loop

. I would like to try it at some point but don't like the idea of something going wrong and killing other parts of my PC.
Not really needed on my current setup but I'll be switching to Z87X-OC with i7-4770k next week and upgrading to either a 280x or 290 in the new year so it would be more useful then.

Ace Modder · 12 Nov 2013 at 22:39

wazza300 said:
[IMG.http://i40.tinypic.com/242bs68.jpg[/IMG]

here,free too

You sir, are my new best friend!

murah · 12 Nov 2013 at 22:40

He ran fine for 2 years, if he had some sort of protection as mentioned he would have probably been fine. its just a bad lesson but a lesson learnt is never a loss on your part

hakank68 · 12 Nov 2013 at 22:48

Ace Modder said:
You sir, are my new best friend!

+1, Great idea wazza300, i sure will be doing this

JonJ678 · 12 Nov 2013 at 22:49

I wonder if an acetal block would have survived. I use one of the EK supreme blocks (black acetal since breaking an acrylic one by over tightening the screws) and the cpu ran hot enough to switch the computer off. Repeatedly actually as I didn't think to check the watercooled system for overheating. No sign of harm to block or O ring.

Panos · 12 Nov 2013 at 23:19

wazza300 said:
Coretemp or realtemp has a built in shutdown pc option once temps reach whatever you set in it,handy thing to have

Indeed, so are the warning settings in the BIOS since forever. And they cannot fail.

(sort off given my UD3 issues

)

tipes · 12 Nov 2013 at 23:32

Bad luck there Ace - but seems it could have been a lot worse. My D5 is molex too, but is there any other option of powering these pumps.I would prefer to run it off a controller, but how?????? In the meantime will set up coretemp to shutdown as precaution - a very useful tip that.

Mark

hakank68 · 13 Nov 2013 at 00:05

Will ek replace the block?

kaku · 13 Nov 2013 at 07:14

hakank68 said:
Will ek replace the block?

I would very much doubt it. I don't think block manufacturers intend for you to boil the coolant inside their blocks!

Unfortunate, but glad nothing expensive was damaged - and a valuable lesson for us all.

esoteric · 13 Nov 2013 at 07:52

Thanks for posting. I want to go for a custom loop with next years upgrade, and this is a good pre-lesson in the dangers and pre-emptive measures.

ubersonic · 13 Nov 2013 at 08:46

JonJ678 said:
I wonder if an acetal block would have survived.

Doubt it, if I'm seeing that correct the heat has caused the metal fittings to expand and crack the plexi, would pretty much happen with any non metal top.

rjk · 13 Nov 2013 at 09:23

let me tell you a little bit about pump redundancy.

whilst having a pair of pumps running an parallel seems like a fantastic idea, it does nothing more than add noise.

i ran a pair of 18w DDC for a while in a single loop configuration and whilst it provided me with peace of mind, it didnt do a great deal for the whole 'low noise' aspect of my watercooling configuration.

my advice to you Lee would be to use this pump failure as a lesson.

this has been relatively small in the grand scheme of things but it could have been a lot worse.

since all decent watercooling pumps now have in-built tachometer readouts, it is probably wise to invest in some form of smart controller.

an aquareo for example can be programmed to turn off the computer if it detects a pump failure

for me, this is the perfect solution. it removes the need for pump redundancy on high impingement single loops and puts a focus on keeping your hardware safe.

for the sake of £50 for the basic aquaero model which has this feature. it is a perfect insurance policy for your hardware, not to mention a valuable tool for system monitoring, fan control and system cooling accessories.

rjk · 13 Nov 2013 at 09:24

also, your new project is going to be significantly more complicated than this single loop solution.

I would consider dual loops given the amount of blocks and rads you will have.

rjk · 13 Nov 2013 at 09:34

ScottiB said:
good job you had non conductive fluid

this doesnt have any impact on the situation whatsoever. his fluid is two years old. fluid that has been in contact with metals and sat in a loop for two years is going to be conductive mo matter what happens. being non conductive would at best grant you a grace period before your hardware eventually died.

i have covered a pair of GTX 480s in fluid before and they worked afterwards. would i have tried to power them on when they were wet? no. of course not.

a spill like this is easy to rectify by using a sensible drying procedure and not being too hasty with turning your hardware back on again.

conductive or non conductive fluid has absolutely nothing to do with it.

JonJ678 said:
I wonder if an acetal block would have survived.

yes it would have done without a doubt. pressure cracks on plexi are all to common in my opinion. repeated heat cycles etc make plexi a liability for long term usage rigs in my opinion. acetal is a far more sensible option. i have never seen acetal blocks suffer from this.

in fact i remember asking MIPS to make me a custom plexi topped memory cooler once and they essentially replied with a lecture on why they refused to work with the material.

the fact is, that crazing around the port on the block was probably there long before the pump even failed, the extra heat would have only exacerbated it.

hakank68 said:
Will ek replace the block?

doubtful really given the circumstances of the failure. but it has been known.

EK are very customer service focused and generally do what they can to make things right by the customer to enhance their own reputation.

ubersonic · 13 Nov 2013 at 09:39

rjkoneill said:
let me tell you a little bit about pump redundancy.

whilst having a pair of pumps running an parallel seems like a fantastic idea, it does nothing more than add noise.

Do people still run pumps in parallel? :confused:

obviously it was useful for some loops in the old days when pumps didn't have a fantastic flow rate but modern pumps don't really have that issue.

Ace Modder · 13 Nov 2013 at 09:43

rjkoneill said:
this doesnt have any impact on the situation whatsoever. his fluid is two years old. fluid that has been in contact with metals and sat in a loop for two years is going to be conductive mo matter what happens. being non conductive would at best grant you a grace period before your hardware eventually died.

Well the leak was spotted with my PC switched on. Anyway I quickly turned it off, and cleaned the fluid off, then put the GPU's into my airing cupboard for about 4 hrs and turned the boiler up to full lol. Once dry I plugged the cards back in, and working fine. At the end of the day, fluid is not as dangerous as people believe if handled quickly and correctly. Of course it always has the potential to kill something though.