• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!

Question Rad to Outlet Pump A240G AIR/Gap in Tube - Watercooling issue help!

Status
Not open for further replies.

jamieinthejar

Commendable
Mar 3, 2017
35
0
1,530
Hi Guys,

Firstly, thanks in advance for any advice/troubleshooting ideas you can offer.

Recently changed over to a new MB, RAM, CPU etc. flushed the system a few times with distilled water and then put in my new Mayhem Pastel White liquid.

The problem is as follows:

As you can see from the photos, pretty much after the CPU block, the flow stops circulating, meaning it's running really inefficiently and recycling the same liquid rather than running through the entire system/rad.

View: https://imgur.com/a/PFbFgl0

View: https://imgur.com/a/35SIfJF


After the CPU block there is liquid - although it looks like it isn't moving at all, and after the RAD there's no liquid. I knew there may be a little trapped air, but this seems more like perhaps the pump isn't strong enough or, there's a block/clogging in the CPU block/RAD.

It's also worth noting that there's a tiny little buzzing sound coming from the pump at the moment, almost like a slight vibration.

I've taken the loop pretty much mostly apart and re-assembled, tried tilting, running with the cap off/open etc. but no joy, any suggestions?

Cheers
 
Not a fan of the pastels. They tend to be far more vicious than standard colored water. Personally, it looks like there is no flow at all, either the pump isn't strong enough to push that pastel through the cpu block, or you have a blockage somewhere. The tubing before the reservoir is all air, but it's not entering the reservoir at all.

Only thing I can suggest at this point is drain it all out and flush the system backwards. See if it'll run with pure water.

You do have flow correct according to inlets/outlets right?
 
Thanks for the reply, yeah the flow is correct according to inlets/outlets, would have been worth me adding that when the system is powered off, it looks like this:

View: https://imgur.com/a/4bhjBp7


Leads me more to believe there's a blockage of some kind I guess

Yeah, the pastels look nice, but I did hear they usually need a bit more oompf from a pump to cycle.

I spent the past couple of hours doing a full flush with distilled water (2-3 flushes) and then topped it up with the last of the Mayhem White Pastel, this was the end result! When I was flushing it seemed fine and ran through the entire system, not sure why this is suddenly happening when the pastel goes in 🙁
 
I cleaned the pump specifically, I didn't clean any of the tubing/fittings fully as I was hoping that a few flushes of distilled would have been fine and didn't see any blockages, suppose I could try taking the actual fittings off and taking a look at them?
 
There might be some deposits inside of the fittings, restricting flow, that you didn't see.

It might be worth your while to operate the pump outside of the computer case (kitchen sink is a good place for me to try this), to assure yourself that the flow rate is adequate.
 
I'll give that a try tomorrow, have a look at the fittings properly and fully take apart the loop, maybe try running just the pump directly to the RAD and see if that works.

Really hoping it isn't the RAD itself, or the CPU block.
 
I'm doubting it's either of those. I'm thinking that pump is having a seriously hard time pushing the pastel through the rad. If you look at the pics, there's liquid in all the hoses except for the return to reservoir. It's like the pump gets the fluid to the res, and then the resultant air pressure that's left is too great for the pump to overcome. You'll have to try the loop using just water. If that works, you know the pastel is too thick or the pump too weak.
 
Yeah probably not a bad shout, I'll get another 5L bottle of distilled ordered this week and give it a flush, see if that goes round the system fine! If so, guess it's off to find a more malleable coolant!
 
I'm guessing an airlock is to blame. What happens when you power the pump on with the cap vented/open?

What I (think) is happening is that you have more air in your loop than you have volume in the sealed reservoir to purge. You need to allow the air to purge, first. That air in the radiator and tubing doesn't have anywhere to go since the reservoir is nearly filled and the cap is sealed.

I'm betting this is your issue.
 
Yeah I tried leaving the cap off for a while to see if that would naturally push out any air left over in the system, but doesn't seem to be the case, let it run for a good 4-6 hours and no changes, fluid in the pipes didn't increase or anything.

What is however a little interesting, is that I left it running for some hours today (sealed) and the water level had suddenly dropped a good couple of CM's and the fluid in the pipe going to the inlet was almost full, as soon as I undid the cap, there was a hissing and the water rose to the top, air lock sounds about right, but I can't really leave the cap open indefinitely can I?
 
No, but what you can do is prime the system. Empty out the fluid, you can save it.. Then with just the psu to pump (pull the mains and jumper the psu to start it) going, and about 1/4 reservoir let the pump push the fluid through as you keep putting liquid in. Uses the liquid like a plunger in a syringe and will force the air out.
 
Sounds like you have a blockage.

Silly question - when you got the EK CPU block, did you have to disassemble it and fit an internal jet plate and flow piece? It looks exactly like the block that I have and I made the mistake of incorrectly orienting the jet plate with the inlet/outlet causing flow to be completely blocked. I had to pull the block apart, rotate the piece correctly and ensure flow worked before re-installing. (blowing some air through to ensure flow worked)

fgEUQcP.png

YvfPfDt.png


This is the set of steps I'm referring to that I would check again, mostly because I ran into this exact issue.
 
Thanks Karadjgne, I'll look at doing that today/tomorrow, try and save the liquid then use it as a plunge/flush.

@rubix_1011, it's the A240G from EKWB so it's pre-built essentially, no assembling of the CPU block required and didn't have the issue in the past, sounds like the two most likely options are blockage or air lock, but I appreciate the input :)
 
Ah, OK. I tested one of these aluminum EK kits - pretty decent, honestly, just don't add copper or brass to the cooling loop.

Other than that, it would have to be an airlock like I mentioned originally. That pump should have no trouble moving coolant....sooooooo.....

I checked the installation directions and it looks like you might have the pump outlet going to the GPU block outlet, but it also mentions to use the correct inlet/outlet port on the GPU block:

It is recommended that you use the correct INLET and
OUTLET port on your GPU waterblock to ensure optimal
performance.

ZqdWpAQ.png

Rdaheok.png


Can you confirm flow direction?
 
From the pic, and yours rubix, the gpu is backwards. Looks like pump-gpu-cpu-rad-pump. Which puts the gpu inlet on top, as in the directions, so flow is into the outlet.

If the flow was the other way and pump-rad-cpu-gpu-pump, the gpu would be correct, but cpu would be backwards.

Wonder what happens if you swap the gpu plugs around and put inlet on bottom left and outlet as top right, as per written directions, not picture.
 
Yeah, really impressed with it so far, good entry option before doing a proper custom loop that's for sure, yeah as it's all aluminium, as long as I don't go adding other metals it shouldn't start ionising badly.

I'm pretty sure I've got the flow set correctly, the outlet is going into the bottom of the card, exactly as the diagram, here's a little diagram 😉

I took it apart again over the past hour/two, took all of the fittings off, all tubes etc. cleaned them mostly (there was a little white pastel gunk around, but not enough to cause a proper blockage considering the power of the pump.
View: https://imgur.com/a/Qfo20Ad


Horrible to look at, definitely going to switch over coolant in a few months.
View: https://imgur.com/a/46l6K9M


In/Out
View: https://imgur.com/a/UL7jkc6


I drew you a pretty picture, I'm happy to change it up if it's incorrect, but I'm pretty sure that's how it's supposed to be
View: https://imgur.com/a/KMAohsz


I've tried flushing it through a few times, but it just doesn't seem like it wants to go, there's some sort of dodgy air block in the RAD, even with it taken fully apart, frustrating!

Thanks for the replies so far guys, happy to switch it around if you think it's worth a shot!
 
There's not a massive amount of slack in the tubes so I may need to snip new ones, but I shall try switching over and see if that helps!

Riiiight, just went over it a few times in my head, I get it, the pump outlet is going to the outlet for the GPU, rather than the inlet so I'm essentially pumping in to a direction that would usually want to put out, makes sense! (also does gravity wise 🙄), thanks guys I'll give this a try hopefully this evening, if not tomorrow and let you know!
 
Gravity ultimately has nothing to do with it, at least once you get the loop flowing. Your pump will be pushing as much as it is pulling coolant.

But yes, the idea is to use the inlet port on the GPU block (from the pump) and the outlet port will go to the CPU inlet. You should have enough tubing left over from the kit to snip off a new length, but that's for you to confirm.
 
Only thing I can think is that the CPU block inlet/outlet is backwards, but not sure that is necessarily the issue, either. The radiator is likely just one large air pocket, but it is low enough restriction that with the reservoir vented, it should easily push through.

Here's a question for you - when you initially prime the loop, does the pump quickly push coolant through the GPU block and into the CPU block, but barely makes it out of the CPU block? Or is it a very slow process, overall?

Are you jumping the 24-pin ATX when you are doing this? Looks plugged into the motherboard, to me.

I wouldn't power the PC on and off without actual coolant flow until I had fully primed and leak tested the loop.
 
I'd have to do it again to be 100%, but if I remember correctly, the coolant instantly rushes through to the CPU block, it's after that it struggles to make it to the RAD properly.

Haven't been jumping the 24-pin ATX (got a bit lazy).
 
CPU inlet port looks correct:

6cUT60E.png


So we're back to airlock, either in your radiator or some sort of blockage.

Also, when you're priming (using the 24-pin jumper) does the reservoir get nearly empty or do you continuously add coolant while it is running? If you allow it to run all the way empty, sometimes this can cause an airlock which prevents more coolant from moving through the loop. Remember, pumps cannot push air - they can only push coolant (water) and they should never be run without liquid as this can permanently damage the impeller bearing.
 
Status
Not open for further replies.