How to clear air out of water loop?

Status
Not open for further replies.

singo79

Honorable
Jan 11, 2014
28
0
10,530
Hi All,

Firstly, please accept my apologies as I know that there are many threads on here about the same topic, however none of them have proven to help my current setup.

I am a reasonably experienced PC builder, but this is the first time in venturing into a full on water cooled system. I've used factory sealed CPU coolers in the past, but this is the first time getting into the whole box of dice.

So my problem is as the title suggests, I have air trapped inside my water cooling loop that I have been unable to clear. So I should get to the nitty gritty I suppose.

I have a Corsair Carbide Air 540 Case, ASRock Z87 Extreme9/ac Motherboard, Intel i7 4770K CPU, Phobia UC-2 LT CPU Water Block, 2 x Gigabyte R9 290X GPUs with XSPC Water Blocks, 1 x XSPC EX240 Rad, 1 x XSPC EX360 Rad, XSPC D5 Photon 170 Res/Pump Combo.

Due to the case setup I have had to run my loop as follows. Please feel free to critique my setup as I am open to any idea at this stage. So the loop goes XSPC Res/Pump Combo > XSPC EX240 Rad > CPU Water Block > 1st GPU > 2nd GPU > XSPC EX360 Rad . XSPC Res/Pump Combo.

Here is a pic of the Res/Pump located in the rear of the case;
24ednid.jpg


Here is a close up of the flow rate indicator, it's totally stationary;
29f5wtf.jpg


Here is the front side of the case where the tube from the Res/Pump goes to the XSPC EX240 Rad;
fursrm.jpg


Next pic is tube from XSPC EX240 rad into the Phobia UC-2 LT CPU Water Block, it also shows the outlet of the CPU Water Block into the 1st GPU Water Block;
278cg.jpg


Next pic is the 2 x GPU Water Blocks connected by an SLi/Crossfire Water Bridge. It also shows the tube from the 2nd GPU Water Block into the XSPC EX360 Rad;
2rhtqh2.jpg


This pic shows the tube in the rear exiting the EX360 Rad and into the rear side of the case leaving the XSPC EX360 Rad;
2u94aya.jpg


This pic is the tube coming into the rear of the case from the EX360 Rad. I have connected the tube with a set of quick release fittings to enable quick and easy draining;
2dre29v.jpg


Finally, it's the two quick release fittings joined together going into the flow rate meter;
2qbycy1.jpg


I've read many of the threads on here guys and none of them have helped clear the air trapped in my loop. Shaking from side to side, front to back and upside down has not helped clear the air out. Though some bubbles have dispersed, there are still large pockets of air that I cannot get out.

Just wondering if it is because of the way I have set up the water loop?

So far the pump has been going for several hours and I've tried the tricks suggested, but unfortunately no success.

Any help would be great guys.

Muchly appreciated.
 
Solution
If your pump is running along nice and quiet, when your PC is sitting normally and you get visible flow then that to me says you should be fine.

When you tilt the pump it starts drawing air and causing cavitation I assume which is why it speeds up, whines and just plain gets noisy. This noise and cavitation is bad and can damage the pump.

Your reservoir is a decent size so I would still recommend simply topping it up and let the loop run for a day or a week and the problem should solve itself. If you were running a T line with no reservoir then bleeding would be more important (If a large bubble were to get pushed to the pump then flow could possibly stop or serious cavitation begin). It is not so much a concern however in your...

dave85uk

Honorable
Mar 3, 2012
50
0
10,630
I've not built a water cool loop myself but I have been reading lots of guides as I've been making a Loop parts list for about 4 months now and readings lots of material before I select my parts and start building in March/April, however....

I'm pretty sure I have read in a couple of places that a few drops of washing up liquid helps clear out air bubbles (actually not sure how much).

It may be more beneficial for readers for a overall/clearer pic of the entire loop, and drawing circles were the air is getting trapped... or is there just a lot of bubbles everywhere?
 

Bonecrushrr

Honorable
Jan 5, 2014
204
0
10,760
Turn the loop on and let it run. Eventually the water flow will force the air out. Water loops are not nice vertical runs with the water coming up from the bottom pushing the air up as it goes, all the twists, turns and downward flow sections try to trap air however if you let it run they will have no choice but to eventually push through.

You can not eliminate the micro bubbles that will forever be in your system... or at least its not worth it. I had the same loop on my i792 for 5 years and every time it was powered off for a day there was tiny air pockets formed at the high points in my tubing and cpu block. I would then turn my loop on, watch the bubbles get pushed out into the reservoir just to find them back in the same place next time.

If you have good flow, it is not causing noise and your not overheating because of it then just let it run and your loop will find its equilibrium on its own.

Very nice looking loop, only thing I would change was have the 240 rad between the cpu and gpu water blocks to keep the water feeding to the gpu's as cool as possible.
 

singo79

Honorable
Jan 11, 2014
28
0
10,530


Hey Dave, thanks for the reply. Whilst trying to dislodge stubborn air pockets I have noticed that the main pockets hover around the top mounted radiator, CPU Block and between the 2nd GPU and front mounted radiator. They are quite large in terms of air pockets, in fact when I tilt and shake the case about I can almost empty the tube running from the top mounted radiator to the CPU block, or the tube running from the CPU to the 1st GPU, or the tube running from the 2nd GPU to the front mounted radiator.

The line coming out of the front mounted radiator to the inlet of the res/pump is always full, there is never any air pockets trapped in this area. Bubbles/Air travel through this tube and get absorbed into the res (which is how it is supposed to work), but clearing the rest has proven more difficult.

In my limited experience I tend to think that air pockets are trapped in the radiators, but it is only a guess. I will work on some better pics and an actual diagram of my system setup to show component location and flow direction. As you indicated, this may assist others in problem solving my loop issues.

Thanks again.
 

singo79

Honorable
Jan 11, 2014
28
0
10,530


Hey BC, thank you also for the reply.

I will give your feedback a try and turn it on and leave it on. Since finishing the loop I've let it run for four or more hours, but this time also included shifting it from side to side, back and forth and on and off to try and help move the air out of stubborn areas.

I'll give it a try and let it run for a day or so and see how things pan out.

Also, I failed to mention in my original post that I am not using Distilled Water, rather I went for the EK-Ekoolant (Clear) because it was already treated and so that I didn't have to add my own treatment solution to prevent bacteria build up etc. Don't know if that is going to make a difference, but I thought I'd best be mentioning it.
 

Bonecrushrr

Honorable
Jan 5, 2014
204
0
10,760


Do your radiators have a n air bleed screws? My rad has a small bleed screw on the non barbed end (It sits so the barbs come in the bottom and the screw allows bleeding the air out the top)
 

singo79

Honorable
Jan 11, 2014
28
0
10,530


No unfortunately. They are XSPC EX240 & EX360 Radiators and they don't appear to have that ability. I initially went for cross-flow radiators but they wouldn't fit into my system, so I had to go back to standard radiators.

There aren't too many suppliers in Australia that supply different brands, most use the same wholesaler and are limited in what we can purchase in terms of brand and models.
 

Bonecrushrr

Honorable
Jan 5, 2014
204
0
10,760


If by chance you were to lay your entire case on its side would that move the 2 "inlet barbs" to the bottom and the "outlet barbs" to the top?
It should for sure help with your GPU blocks in freeing up the air. Use the whole air floats on water thing to your advantage.

Can you hear the air going through your pump?
 

singo79

Honorable
Jan 11, 2014
28
0
10,530


The only time I can really hear the pump making loud noises is when the case is tilted on it's side or on it's end, then the pump starts to whine a bit and the liquid inside starts to agitate quick fiercely.

Here is a couple of links to the products I specifically purchased.

Res/Pump Combo - http://www.xs-pc.com/water-pumps/d5-photon-170-reservoirpump-combo
Top Mounted Radiator - http://www.xs-pc.com/radiators-ex-series/ex240-dual-fan-radiator
Front Mounted Radiator - http://www.xs-pc.com/radiators-ex-series/ex360-triple-fan-radiator
GPU Water Blocks - http://www.xs-pc.com/waterblocks-gpu/razor-r9-290x-290
CPU Water Block - http://www.computerbase.de/bildstrecke/42788/22/

Hopefully this will give you a bit more of an idea as to what I am using/dealing with. I can't get pics or diagrams done at this stage as I am at work, but I should be able to organise something after that.
 

Bonecrushrr

Honorable
Jan 5, 2014
204
0
10,760
If your pump is running along nice and quiet, when your PC is sitting normally and you get visible flow then that to me says you should be fine.

When you tilt the pump it starts drawing air and causing cavitation I assume which is why it speeds up, whines and just plain gets noisy. This noise and cavitation is bad and can damage the pump.

Your reservoir is a decent size so I would still recommend simply topping it up and let the loop run for a day or a week and the problem should solve itself. If you were running a T line with no reservoir then bleeding would be more important (If a large bubble were to get pushed to the pump then flow could possibly stop or serious cavitation begin). It is not so much a concern however in your system if every once in a while a bubble or 2 will work its way out due to getting picked up from wherever it was by the turbulence at the air/water interface. Keep in mind that the 240 rad bubbles must pass by every other choke point and may even become deposited there for a while so you will see the start of the loop get better before the end of the loop.
 
Solution

singo79

Honorable
Jan 11, 2014
28
0
10,530


Thanks again BC, I'll try running it for a day or two and see how it pans out. Appreciate the assistance.
 

singo79

Honorable
Jan 11, 2014
28
0
10,530
Ok, this is some additional info that was requested earlier. I think Dave suggested this, so I have taken some more pics (still not very good images I'm affraid) as well as a loop diagram (please don't judge me on my drawing abilities lol). I've had to use a torch to provide light through the colour tubing to show whether there is liquid in the tube or not. I apologise for the pics but this is the best I could manage on my own and an iPhone camera.

First and foremost here is the loop diagram;
29wlj0x.jpg

I've highlighted the areas with red pen where the major/large air pockets are sitting and fail to move even when tilted and shaken.

This is the outlet tube of the CPU water block. Don't know if the picture actually shows it, but this whole tube is empty of liquid;
2n7in92.jpg


This pic is a continuance from the pic above. This tube is the same tube from the CPU Water Block and it goes into the 1st GPU. Hopefully the pic shows that there is only liquid just appearing in the bottom of the tube, therefore the other 99% of the tube is full of air;
2i67it.jpg


This is the outlet tube from the top mounted 240 radiator. In this pic the tube is full of liquid, however when the case is titled from side to side the tube quickly fills up with a large air pocket that consumes about 50% of the tube's volume. In this pic the tube is full of liquid when runs to the CPU water block inlet, but the outlet (as shown above) is completely empty of liquid;
20rmflw.jpg


This is the tube that runs from the bottom/2nd GPU into the front mounted 360 radiator. In this pic the tube is full of liquid, but as described above regarding the 240 radiator, as soon as you start tilting the case around a large pocket of air appears in this tube and consumes anywhere from 50 - 100% of it's volume;
5fiy45.jpg


This pic is the tube that exits out of the front 360 radiator and starts making it's way back to the inlet port of the res/pump combo. This tube has a set of quick release fittings on it for drainage purposes down the track. It also has the flow rate meter on it, which is basically stationary despite the pump being set to full speed;
zk1w6x.jpg


The PC has been running since I got home from work, which is about five hours now. I'll leave it running overnight and see what the results are in the morning and assess things more then.

In the meantime, if anyone has the time to look over the above diagram and pics and might have a bit more of an idea I would be very greatful.

Cheers.
 

dave85uk

Honorable
Mar 3, 2012
50
0
10,630
Ah ok this was a bit confusing at first but you are not having a bleeding issue. As far as I understand you have zero flow rate, bleeding can only be performed if your fluid is circulating through the whole system.

From looking at your hand-drawn picture, and the way you are explaining how it responds when you tilt it there is something restricting your flow.

Look at the tube starting from the outlet of your pump to the inlet of your 240 rad. Is this full of fluid? If so... from your drawing it looks like you are having little flow rate between your radiator and your CPU block.

The reason why you have a little fluid in each tube between your components is that there is no pressure/flow behind your cpu block and so it will trickle through due to its own fluid head (self-weight of the fluid) to the other components. I highly suspect that your 240 Radiator is clogged... try this:

1.Disconnect and create a loop with only your radiator and pump/res combo. See what happens.
2. Try then flushing out the radiator using a high pressure source.
3. Try step 1 again

It looks to me if this doesn't work the radiator is not functioning properly (and is restricting the flow rate) or your pump is damaged. Maybe try just connecting a long tube tot he inlet and outlet from the pump combo to check the pump is working correctly.

If you have ran the pump unloaded (without any fluid in the res) you have more than likely damaged your pump.
 

singo79

Honorable
Jan 11, 2014
28
0
10,530


Thanks Dave, this sounds like a sensible suggestion and though it will be a little mucking around, I am prepared to give anything a go at this stage.

I did try cleaning the components when I first got them by flushing them with distilled water and placing the caps on the port and shaking them about, but the idea about hooking the pump upto the 240 rad by itself to check flow through the rad is a very sensible suggestion. I'll be most definitely giving this a try.

I also plan on giving BC's suggestion of disconnecting the 240 rad and just plumbing the res/pump straight into the CPU block to see if the flow rate picks up.


Thanks guys for the suggestions. I will be giving it a try tonight and I'll post back on the results.
 

singo79

Honorable
Jan 11, 2014
28
0
10,530
Ok, so after almost pulling the whole loop out and testing each component, it would appear it is the Phobya CPU Water Block.

The res/pump combo is fine and pumps massive amounts of water, which is reassuring. So I hooked it up to the 240 rad and it blew straight through it, no obstruction no trickle, it was a torrent.

I then hooked up the res/pump to the 360 rad and same thing.

I then tested the CPU water block and boom! It was a trickle, it went hard to start off with and then it almost came to a standstill. The Phobya CPU water block comes with absolutely NO instructions. On their website is absolutely nothing! It says a lot about a company that can't even supply a simple set of instructions with their product, or at least supply the info on their website.

The diagram on the CPU water block itself indicated that the inlet was the arrow pointing into the water block and the outlet was the arrow pointing out of the water block.... WRONG! The inlet and outlet are completely different to what you would think. The arrow pointing in is the outlet and the arrow pointing out is the inlet.... go figure.

When I hooked the loop up in the appropriate configuration the liquid started blasting through there.... SUCCESS!

So I then put the whole loop back together and it has been going hard since. It is working those bubble and air pockets out quite nicely.

I just wanted to thank you guys for your assistance. Dave and BC you have helped me a great deal and I really appreciate it. Thanks again. I finally get to enjoy my PC at last.

So just to reconfirm, the actual post title of "How to clear air out of water loop?" is really incorrect of me. It should have been "My water flow rate is a trickle, how to fix?" That would have been more accurate, but I just wasn't aware that the flow rate and air pockets were two different problems. Something that I know very well now.
 

dave85uk

Honorable
Mar 3, 2012
50
0
10,630


Though it would likely something stopping the flow rate... good job! Have fun with your working loop!
 

Bonecrushrr

Honorable
Jan 5, 2014
204
0
10,760
Great find, did you happen to try taking the water block apart and see what was causing he restriction? That would ready suck if it had a piece of plastic or tape or something in there. Not sure if it unscrews and can come apart however considering all you have done so far it might be worth checking out. As with anything you need to be very careful when you put it back together to get an O rings back in the right spots that way they don't get pinched and leak.
 

singo79

Honorable
Jan 11, 2014
28
0
10,530


I haven't pulled the CPU block apart, I'm affraid that I'm not that confident. I did hook it up to the pump and give it a good flush once I managed to work out the issue with the inlet and outlet ports.

There does appear to be screws recessed into the bottom which would indicate that it can come apart. But as I mentioned above, I am not that confident in pulling it apart and then putting it back together correctly. Though this whole experience has been one very steep learning curve. :)
 

Bonecrushrr

Honorable
Jan 5, 2014
204
0
10,760


They are generally made to come apart for cleaning purposes once you have ran the block for a while.
 

Jonboy_Fantastic

Commendable
Sep 13, 2016
23
0
1,510
I have a very similar problem to this. I have a EK Evo Elite with an EK-XRES 140 Revo d5 pump included. the flow to the block from the res/pump combo is solid however when it leaves the evo it's a trickle and continues down until it pools right at the radiator. I've never disassembled the EVO as it came with the insert ( I1 and J2 ) pre-installed in my block. I always use a silver kill coil and plenty of deadwater in my loop and there is absolutely 0 signs of microbial growth. I also only use distilled water. I'm thinking of using a old mod top/reservoir I have ( a XSPC dual pump res combo which is horribly designed which is why I retired it ), or buying another bitspower mod top which I found on sale at a local computer store for like 25 bucks however I can't find one that's reasonably priced anywhere else.. Basically it works if the reservoir is constantly topped off however to get it started without running your pumps dry you have to hold it tiled or preferably completely sideways because the inlet for the water to the reservoir is located at the very top of the reservoir, even so the possibility for cavitation in the pump inlet assembly is very high and it destroys D5's like no other. I'd recommend not buying it. However I'm beginning to think it's simply a pressure problem. The Evo elite with it's circuitous design with the replaceable inserts and jet plates is a huge flow restriction; more so then other cpu water blocks, and I've owned a couple of them. XSPC raystorm, apogee HD and now the EK evo Elite.

To add to the problem I think it might be the design of the EK-XRES 140, as the inlet is located at the bottom of the reservoir, now I know air floats on water however pushing water through the top of the reservoir I BELIEVE bleeds out air bubbles a lot more effectively, as when the inlet and outlet are located at the same level and there is only one hole in which water is taken into the pump and air is supposed to bleed out I believe that it's more of a possibility to recycle the air bubbles back into your loop. Then again I'm not an engineer. I'm only speaking from personal experience.

By the by I've not heard of anyone bitching about the XSPC dual bay reservoir as much as I have. Either there is some secret to filling it in which your pumps will never run dry that I'm oblivious too, as even tipping it causes air bubbles to get caught in the channels in which water is supposed to circulate through both pumps. Which is dangerous for your pump, as anyone who's owned a d5 ( and I've gone t hrough maybe 5, luckily after the first I was clever enough to get protection plans, however not clever enough not to run them dry by accident ever again ) knows that running the pump dry for any period of time seriously decreases the lifespan of the pump.

I need help, any comments or advice is both appreciated and welcome.
 
Status
Not open for further replies.