Loop flow back pressure won’t release

[AJ847.63]

Hi all,

just wanted to see if I can get some advise on a issue I’m having with a new loop I’ve put together. Issue is there seems to be extremely low flow to the point where I’m having trouble filling the whole loop up. When I turn the pump on you can see the levels try to move than slosh back in place so it seems like there’s a extreme amount of back pressure that the pump can’t overcome. The pump is definitely working as I can see it spinning and hear it vibrating as well.

Loop layout is: res to cpu, cpu to ek rad, ek rad to GPU, GPU to Corsair rad and Corsair rad back to ek res. According to the ek manual for both the res and cpu block are directional so I have hooked them up matching their manuals. The GPU block and both rads did not mention so I assumed it did not matter.

ive tried removing a hose and trying to fill the radiator directly but again there seems to be extreme back pressure so it just sprays back out. I’ve got one of the rad fittings open and the fill port in the res open to try give a path for the air to come out but the pressure just will not release. I’ve tried opening every fitting in the system but nothing will release the pressure.

im thinking I need to go buy a dedicated more powerful D5 pump but I don’t want to go buy something and it not solve the issue. Not to mention no one in Sydney stocks em and being Christmas shipping I’d be lucky to get it by New Years

components are: Corsair 420 rad, ek 360 rad, ek quantum kinetic 360 res and pump and I’m using soft tubing thank god. Wanted to use hard but talked myself out of it luckily otherwise this would of been even more of a nightmare!

thanks for your assistance.

 

[Nuckles56]

I suggest trying to reverse the flow directions into the rads and GPU blocks and see if that solves the issue as that seems like the most likely issue that I can see for it not wanting to flow (do it one a time and see) as I can't see why you would have so much back pressure that you can't fill the system. The only other thing I can think of is that there's a blockage somewhere in the loop that's causing the issue.

 

[crispysilicon]

Take the res/pump out and put it on top of the case until it's bled out. You need it at the highest point.

 

[AJ847.63]

Thanks for the advise. It’s not a blockage I’m able to blow through the loop myself to get a full rotation through the loop so it seems like a back pressure/weak pump issue. Dunno if the tubes will be long enough but I’ll try take the res out if I can and pop it on top of the case. I should of thought of that you should always have the pump at the top of the loop and in this case it’s closer to the bottom.

 

[micropage7]

too crowded, why you not make it in line, from processor to gpu then to rad and res
and i agree you need to test to fill it up from the highest point or you need to place fill in port at highest point

 

[OneMoar]

your rads are upside down

inlet\outlet on the bottom
you have created a airlock you need to flip the rads so the column of water works for you and not against you
with the inlet/outlet at the top you are constantly fighting gravity with it on the bottom the pressure will equalize and once bled the pump will only need to work half as hard

 

[thesmokingman]

your rads are upside down

inlet\outlet on the bottom
you have created a airlock you need to flip the rads so the column of water works for you and not against you
with the inlet/outlet at the top you are constantly fighting gravity with it on the bottom the pressure will equalize and once bled the pump will only need to work half as hard

Rad ports can be top or bottom, that orientation doesn't really matter. Gravity isn't a big concern as whatever goes up must come down, and in a loop it is equalized on the way down. Now just pushing it up will require one has enough pump and with that mess of routing... it looks like not enough pump. Fyi, I'm running the ports on my 480s on the top just like that and I don't have his issues.



And I would cut out all that extra tubing. Run the rads in series one after the other, ie. plumb them together so they form one unit. The way you have it setup creates a lot of extra complexity for negative gain.

 

[nguyen]

does the pump power connector plugged into the motherboard or a modex to 4pin power ?

 

[OneMoar]

I am a fan of having as much of the Colum of water as you can get feeding the pump a
all things being equal if there is air in the system id rather it not be trapped in the bottom but at the top where it wont do as much harm
and yea his tubing is a cluster

 

[micropage7]

oh one more, he needs to fill the highest point first
so if he wants to keep the setup he needs to fill the rad first

 

[OneMoar]

or just turn it the other way and then the only thing he needs to bleed is the res which is likely going to bleed easier

either way he needs to fix that plumbing

 

[AJ847.63]

Rad ports can be top or bottom, that orientation doesn't really matter. Gravity isn't a big concern as whatever goes up must come down, and in a loop it is equalized on the way down. Now just pushing it up will require one has enough pump and with that mess of routing... it looks like not enough pump. Fyi, I'm running the ports on my 480s on the top just like that and I don't have his issues.



And I would cut out all that extra tubing. Run the rads in series one after the other, ie. plumb them together so they form one unit. The way you have it setup creates a lot of extra complexity for negative gain.

View attachment 179347

View attachment 179348

thanks. I’ll try flip the rads since I think it is the pump struggling to fight the air lock so that might help give it some oomph. Otherwise I’ll go get a dedicated pump.

Moving the res on top of the case has not helped unfortunately. I’ll try flipping the rads and if that doesn’t fix it, which I doubt it will, I’ll go get another pump. Though I doubt that will help as well because I already did this with the bitspower zxy res/pump combo. I really need a dual seperate pump with high head pressure but as usual no one sells em here so I’d have to import them.

 

[micropage7]

have you tried to test the pump, just make sure the pump is ok

what about try to fill it outside the case so you can move it to remove the air then plug it back to the case

 

[AJ847.63]

have you tried to test the pump, just make sure the pump is ok

what about try to fill it outside the case so you can move it to remove the air then plug it back to the case

Yeah I have. Pumps ok as I can see it spinning. I’ve removed everything from the case so I can move things around and inverted both res as recommended but it’s still not flowing through much. Dunno if it’s moved much more or not. Both tubes coming out of the GPU and into the rads are filled on both ends but only a third of the way. The rest of those tubes are empty. The cpu block is also maybe 5% full but it’s tubes are full.

 

[Caring1]

Too much air in the system, try laying the case on it's side, just remember to close any fill ports.

 

[DeathtoGnomes]

Too much air in the system, try laying the case on it's side, just remember to close any fill ports.

with those long hoses, take the radiators out and lay them down flat until the system is bleed fully.

 

[TheoneandonlyMrK]

As others have said there's too much pipe work, your defeating the head pressure of your pump.

Down size your pipe's so the runs are not extreme.

In the past when facing this issue I have at times removed half the loop leaving the second half, get it filled then add in the rest of the loop, this still isn't easy but once filled the pump might be fine pushing it around.

But to ultimately solve this I went dual inline pumps, they'll fill stuff easier.

 

[AJ847.63]

Still struggling to get the air lock out of the Corsair rad which is the one on the top in the new pic. Both have now been removed from the loop and all tubes removed from them only fittings connected. Ek filled easily but the Corsair one still has either a partial block, and it’s only partial as I can manually hack out a lung and blow past it, or a air lock still. It does not make sense though, removing the Corsair rad from the loop and disconnecting all the tubes so it’s just the rad by itself should of removed the air lock.

ive gotta get some more tubing so I can re run the loop, but there’s no point doing that and re connecting things till I break the air lock and finish filling the Corsair rad. If I can’t blow it through the pumps got no hope.

What’s funny though is I connected the ek rad to the Corsair rad connecting port 2 to port 3 just now. Blow into the left port so port 1 on the ek rad and coolant comes spraying out of the 4th port , so second port on the Corsair rad. But between the 2nd port and third port there is no coolant flow the tube is dry, so air is getting through but the air lock is stopping any coolant from flowing through the ek rad and into the Corsair rad it’s only letting air through.

 

[crispysilicon]

You need the res on the top and you need to get the air moving out now that it is. The pump alone won't do it, too much air in the system. Because of how you have the loop configured, you're going to have a hard time bleeding it out.

I would remove the radiators and res, manually manipulate the whole thing to get it as bled out as possible, then reassemble.

 

[ThrashZone]

Hi,
If you have the d5 pump pwm wiring connected to the mother board or really anything disconnect it system via 24 pin cable is not connected to the board.
24 pin cable you're not using the jumper to fake the psu out to run the molex or sata power to the pump so how is the pump powered ?

How long the piping is does not matter it's just more fluid.

 

[TheoneandonlyMrK]

Hi,
If you have the d5 pump pwm wiring connected to the mother board or really anything disconnect it system via 24 pin cable is not connected to the board.
24 pin cable you're not using the jumper to fake the psu out to run the molex or sata power to the pump so how is the pump powered ?

How long the piping is does not matter it's just more fluid.
View attachment 179401

That last point is not right, pipe length plus restrictions = head pressure (too little makes no flow).

 

[ThrashZone]

That last point is not right, pipe length plus restrictions = head pressure (too little makes no flow).

Hi,
Not in this case
Running to a different room maybe.
I've seen no issues at all I have two external mora 360mm rads too.

Dude just doesn't have his d5 powered correctly or has pwm wiring hooked to a dead board and it's not giving signal to run properly.

 

[TheoneandonlyMrK]

Hi,
Not in this case
Running to a different room maybe.
I've seen no issues at all I have two external mora 360mm rads too.

Dude just doesn't have his d5 powered correctly or has pwm wiring hooked to a dead board and it's not giving signal to run properly.

Why are the physics different where he is?
He said the pumps running so I assumed it is.

Well I'll leave you to it , good luck op.

 

[ThrashZone]

Why are the physics different where he is?
He said the pumps running so I assumed it is.

Well I'll leave you to it , good luck op.

Hi,
Wiring looks like a spaghetti factory and he's obviously powering or leak testing without the furnished ek psu plug on the 24 pin main connector which I also pointed out
I've already gone through pwm d5 and again if it's connected to a dead board it will not run very long if at all so the whole pump wiring deal is not known atm

If it's a vario 5 speed manual d5 it would be best set at it's highest setting for filling.

Hosing just looks like an octopus which I kind of like but I also use two d5 pumps and have lots more tubing

 

[thesmokingman]

Dude, is that plumbers tape on the OP's picks??