Pump failed during first leak testing

J

javisaman

Limp Gawd
Joined
Apr 3, 2007
Messages
500
This is my first watercooling setup.
I think my MCP350 failed during leak testing. It could be a variety of things that caused it.
Here are the details:

The Loop:
Res --> pump --> MCR320 --> Apogee cpu block --> Res. I don't think the loop is too restrictive. It was filled with distilled water and 5% antifreeze. The pump has a petra's top on it (which I notice now is leaking, even though the screws are as tight as I can make them) and the fluid isn't coming from the barbs.

There was A LOT of debris in the loop (white gunk) this time. I don't know why because, I cleaned out the radiator with vinegar (two washes about a month ago) before running it. Also it didn't look like much initally, but I left it for 5min and the res was full of it and the pump died (burning plastic smell). I opened up the pump, there was no debris in it, but the circuitry on the back was fried none the less; the green pcb melted exposing some copper which separated. I'm still not 100% it was the debris that caused it.

I think the thing that killed it was the fact that the res ports were probably partially blocked by this "gunk" preventing it from filling the pump. But I guess do to my inexperience I don't really know if this is the issue either.

I'm guessing the pump did run fine for around 5 min before it died.

I'm hoping one of you could help me narrow down the problem so that I don't make the same mistake again.

I am considering a MCP655 pump as a replacement (w/o the controller because I'm low on dough), but I'm not sure how well I can mount it in a already cramped Antec P180. I was also considering a MCP355, because I already have a petra's top, but I heard they have a high pitched whine. Right now both pumps cost pretty much the same.

Thanks
 
Arcygenical said:
These pumps need to be FILLED with water before being turned on...
Click to expand...

Definitely the crucial point.

Anytime you're sucking in air you're damaging the impeller of the pump. Also if your top is leaking, it's probably because your O-ring inside of the pump is either missing or not set correctly. Also make sure that the top is completely tightened.

Next time, spend more time flushing your radiator/tubing/blocks immediately prior to setting up the loop. I once left my "flushed" radiator out for a few weeks. When I slapped it altogether, it had all this dust, bugs, and what not that gathered inside during the periods of non use, even after I flushed it thoroughly.

Send your pump back for an RMA.
 
Lol yeah I figured something of that sort. They looooove dark small places like a Rad
 
Ranker said:
Send your pump back for an RMA.
Click to expand...

just because you cant read the directions and you break it doesnt mean you should get a new one for free. lots of people come into my store complaining their flowers died very quickly and they want a refund. when i ask if they watered the flowers the customer usualy replies unassumingly "no" or "i didnt know i had to water it". if you buy a puppy from petco, bring it home and dont feed it until it dies, bring the corpse back to the store and see what they say. you would also be suprised how often people take a pie out of one of my deep freezers and ask if its frozen.
 
LOL. Preaching to the choir here. However, everyone deserves a second chance so he should send it back for an RMA.
 
Thanks for all the suggestions!

Yea I wasn't planning on RMAing it, but I'll see if I can get a new one... I mean money doesn't grow on trees and hopefully swiftech is nice enough.
 
javisaman said:
The pump has a petra's top on it (which I notice now is leaking, even though the screws are as tight as I can make them) and the fluid isn't coming from the barbs.
Click to expand...

...did you remember to make sure that the pump's o-ring (which seals the impeller housing) stayed in place during the top swap? If not, then that would explain the leaking and, possibly, the dead pump.

Also, debris==bad. If any of that junk managed to stop the pump's impeller, the DDC would have just sat there, trying to start, until the pump burnt up (at least, I'm pretty sure that's what a DDC-2 would have done :rolleyes: ).
 
I had an instance when the circuits fryed, after i disassembled the pump and reassembled it. One of the solder joints broke on the 12v wire, and it shorted on the pcb of the pump. Most likely not a common problem, but it happened to me.
 
I had an instance when the circuits fryed, after i disassembled the pump and reassembled it. One of the solder joints broke on the 12v wire, and it shorted on the pcb of the pump. Most likely not a common problem, but it happened to me.
Click to expand...

Actually thats exactly what happened to me. A copper wire separated from the PCB, not sure if its the 12v wire though.

...did you remember to make sure that the pump's o-ring (which seals the impeller housing) stayed in place during the top swap? If not, then that would explain the leaking and, possibly, the dead pump.
Click to expand...


The O-ring was not properly set.... :( , not sure if that caused the break.

Also, debris==bad. If any of that junk managed to stop the pump's impeller, the DDC would have just sat there, trying to start, until the pump burnt up (at least, I'm pretty sure that's what a DDC-2 would have done ).
Click to expand...

Well I looked inside the pump itself and it had absolutely no debris inside of it. If anything the debris most likely prevented the filling of the pump. The pump was running before it died. It smelled like burning plastic (PCB melted). I didn't realize these things could get so hot.

Thanks for all the responses.
 
The water keeps it cool and provides lubrication for the bearings. Also they have a limit on what temp water can flow through them make sure your loop has the pump after the radiator so cool water hits the pump.

CPU -> RAD -> Res -> Pump ->
 
Gyrfalcon said:
Also they have a limit on what temp water can flow through them make sure your loop has the pump after the radiator so cool water hits the pump.
Click to expand...

As I recall, the fluid temperature limit for the DDC series pumps is 60*C... and if your coolant is getting that hot, then I think that you just might have more to worry about than your pump frying. :rolleyes:
 
The water keeps it cool and provides lubrication for the bearings. Also they have a limit on what temp water can flow through them make sure your loop has the pump after the radiator so cool water hits the pump.
Click to expand...

Yea the fluid was not 60C for sure. Remember, I was just leak testing; the PC itself was NOT on (I used the jumper technique on the PSU), the only thing that could be heating the fluid would be the pump itself...

I always thought that the best placement was radiator before blocks, so that the devices get the coolest water possible. The water temp in general should never get so high that it would damage the pump.
 
You must log in or register to reply here.
Back
Top