Did I just burn my pump?

[White]

Hey everyone

Just today I took the time out to gather my new rig with a friend. Didn't go exactly as planned.

I am using an XSPC D5 Dual bay Reservoir and an Alphacool VP655 Single edition. When we were filling the loop to do the leak test, it was not behaving as I expected. It would be really slow in filling the loop, the water would just trickle in. It seems like the pump was having a hard time getting anything into the loop?

Sometimes it would do more than just trickle however, as we did eventually get the loop full of water albeit with some air left in it. The videos I have seen and some of the guides describe it as the process going REALLY fast and with this powerful a pump and L/hour that it can manage ... a trickle is way too little.

So ... how exactly did I mount the pump the wrong way? Or what might I have done wrong to cause this?

It seems to me it's quite straight forward, especially since XSPC does not bother to describe how to mount the pump into the res other than put the o-ring into the groove, put the pump in and screw the fastener tight to ensure a good fit.

What noise can I expect when filling the loop?

PS: The pump does smell a bit burnt :|

 

[airdeano]

Tom's last update on Orca was having a similar issue, until he opened a fitting
to release the air-lock. then the coolant filled quickly.

 

[AndyM95]

You have to be pretty quick when filling your loop or do what I do. I fill up the res, turn the pump on until the res is emptied (only a couple of seconds) then fill up the res and so on... If there is air in the pump for extended periods of time it will burn out. If there is no air in your loop the pump should be pretty much silent.

 

[White]

Ah, so it might have been airlock somewhere near a fitting? How do you spot those and fix them? Sorry for the newbie question, I just want to be really really certain.

However, my entire loop was filled at the time when my pump seemed to have burned out. The loop was full, but when I turned the system off some air would dump into the first part of the loop. The first part of the loop goes from my pump/res into a 240mm rad in the bottom of the case.

At one point, the pump even seemed to have gotten a proper flow into the loop. I put it into the 5 setting and air was being rushed through the loop (not via the pump, just the air already in there). It seemed really good. Then I'd switch it off and on, then it'd start having trouble pumping anything into the loop. The first 5 inches of the loop would be air and a slow trickle of water coming from the pump/res.

I'm quite baffled. Most guides seem to be a bit sparse on what can go wrong and such

 

[White]

You have to be pretty quick when filling your loop or do what I do. I fill up the res, turn the pump on until the res is emptied (only a couple of seconds) then fill up the res and so on... If there is air in the pump for extended periods of time it will burn out. If there is no air in your loop the pump should be pretty much silent.

There is going to be plenty of air in the loop when filling it.

My problem was that the loop was not filling up quick, mostly it'd just be a trickle and sometimes it'd dump 20 cl into the loop right at the start and then sorta just slow down to a trickle again.

I never once saw the res go low enough that I'd worry about air in the pump. Unless it got in there some other way? I am worrying about whether or not my pump and res was assembled properly, but it just seems so idiot proof. Or maybe not for this particular one.

 

[airdeano]

just depends on complexity, device locations and routing.
there is no science to how to "burp" the system. if the pump inlet is constantly
submurged in coolant, chances are it is fine. but the other bugger is allowing the
air to purge from the system to allow more coolant to be added. usually the
highest point is the best to start as air rises and "blocks" fluid from entering.
basically an air-lock or fluid lock. just make sure you have enough towel to
protect electronics.

 

[White]

Well, I'm doing it in a Switch 810 case. I have a 360 rad in top, a d5 pump attached to a d5 dual bay reservoir and then a 240 rad in the bottom before it goes to the dual graphics cards, the cpu and finally into the mb vram pipe before returning to the top 360 rad.

So, in essence, the absolutely highest point in my system is the 360 rad which is an alphacool ut60 so it does have that one port in the "far end" of it. However, the tubing from the rad to the res has its lowest point (still the top side of the tubing tho) lower than the inlet on the res so in that way for the res the highest point is the inlet port.

I read in another thread, another forum as well, that a guy had tested his pump by just attaching a straight tube from his pump/res into a container just to see if that part was working properly. I am probably going to do this, as I have currently bled the system anyway. Just have the hose from the outlet of the res/pump go straight into the sink and see if it works or not.

However, if that doesn't work either my pump is burnt or I am completely at a loss as to what could be wrong right now. If my pump is burnt I can just order a new one but I would still be entirely cluess on how to attach the pump when I get it. Seeing as I probably just failed miserably at attaching the pump to the res in the first place.

Oh and is there any sure sign that my pump is burnt?

Well, I pulled the system apart and tried to test the pump and res separately. Good news is the pump still works, but I'm not sure on how to find the airlock.

My plan now is to build the loop one component at a time, then see which component has the airlock. I'll just use a piece of tubing that goes straight into the sink at the end of the loop while doing this.

 

[Hooded]

You might need to do some Rock'n Roll with your case to get the air pocket to move. You can try squeezing the tubing several times as well as that works like a diaphram pump to help move the airlock.
Rads are the usual location for an airlock to form but not the only place.

I like to bottom fill my systems. If you have a port in the top of your top 360 rad, then that port should be open to allow air to escapelike a vent while you fill.

Not having a picture of your layout we can only guess at what's causing your airlock issue.

 

[White]

Would you need a picture?

The best I can do right now is this. http://www.youtube.com/watch?v=-LHjJ8sPe_8 go to 24:30 or so and Tom's got a good view of the case.

The components are, starting from the top radiator and moving along the loop.

1. 360 rad in the top
2. Pump / Dual bay res combo
3. 240 rad in the bottom
4. 2 x GPU blocks
5. 1 x CPU block
6. Asus ROG Thermo Fusion pipe (mb vr cooling)

My only problem with the res is that I don't want to tilt the case and then accidently have the pump run on air. The pump intake is in the middle of the reservoirs back wall. So that means I can't really tilt it that many ways before I run the chance of having the pump run on air.

How long can the pump run safely if there's an airlock preventing it from running properly?

 

[Hooded]

Move the case to a table where it can be rolled around.

If the air lock is in the actual pump it won't be too long before you smell it cooking.
If the airlock is in the rads, then the pump will be fine as it is in fluid it's getting the lubrication that it requires.

You won't fix the airlock issue without rolling your case around . This will allow gravity to move the airlock. Just bump your pump in short bursts until you get enough fluid in.

A picture of your layout is what we would need to help give pointers. there are a lot of things that can be seen in a picture that we can only guess at for now. Rad orientations, tubing route, fill, vent, and drain ports can all come into play on how well your loop will fill or drain.

 

[Deejeta]

Did you flush the radiators first before installing?

 

[White]

Did you flush the radiators first before installing?

Yeah, I did not actually do that ... If there was any gunk/residue from the manufacturing I wonder where that went now :|

I just pulled the entire thing apart again, so I'm going to flush it now and hope it's not a waste of time.

 

[B NEGATIVE]

Yeah, I did not actually do that ... If there was any gunk/residue from the manufacturing I wonder where that went now :|

I just pulled the entire thing apart again, so I'm going to flush it now and hope it's not a waste of time.

Is there any residue in the res?
That is the first place flux residue will end up,normally a whitish ring at fluid surface level.

If you really are struggling to get the air out,add a single drop of washing up liquid,this breaks down the surface tension of the water in the loop and will aid bleeding.

 

[White]

Is there any residue in the res?
That is the first place flux residue will end up,normally a whitish ring at fluid surface level.

If you really are struggling to get the air out,add a single drop of washing up liquid,this breaks down the surface tension of the water in the loop and will aid bleeding.

Nope, I did not see any residue in the res.

When I flushed my rads, I used the pump to do it. I simply put a hose into the "out" plug in the rad to shoot the water into the sink and ran 0,5L or so through each, then drained the rads and the pump afterwards as well. And neither times did I see any residue in the reservoir.

Oh and nice tip, thanks!

 

[White]

I just noticed something this time around, after having disassembled the pump after the first go at it. The water will now flow past the pump when I fill the loop, which means I can actually fill the bottom rad straight up! Practical. This might also have been a clue that something was wrong the first time around? :|

I am wondering, since I am now filling the loop from the top rad in the loop via the two extra ports on top would it not be possible to fill the entire loop with water right from the start?

I'd have to worry less about cycling the pump and so on, I could just add all the water needed right from the start pretty much.

 

[airdeano]

umm, yeah. way back at post #2. you had an air-lock.

 

[White]

umm, yeah. way back at post #2. you had an air-lock.

Not quite following.

 

[White]

If it's an airlock, it seems to be located inside the GPU blocks. As soon as I attach the GPU blocks to the loop, the pump can no longer pump water through.

 

[Ironhand]

If it's an airlock, it seems to be located inside the GPU blocks. As soon as I attach the GPU blocks to the loop, the pump can no longer pump water through.

Try turning the computer upside down or taking the GPU out of the PCI slot and shaking it around alittle bit.