CPU block bubble issue

[JayCobs]

I am out of ideas at this point. I have a bubble in my EK-Supremacy Copper Plexi block that won't pass through...

In the first shot the bubble is in the middle circle of the CSQ design. In the second shot it's easier to see due to it being magnified through the plexi.

For more pictures of my build you can search Tha Leviathan in the Project Build Logs.

My loop is as follows:
Dual MCP655 pumps - 480 rad - 480 rad - 480 rad - EK Supremacy Copper Plexi CPU block - Bitspower Memory block - Heat Killer 79X0 - Heat Killer 79X0 - Bitspower 150 Res - Dual MCP655

I've been running this loop for months without hiccup. The bubble has been there for the past 3 - 4 weeks. This bubble has become a thorn in my side. I have tried EVERYTHING to get the bubble to pass. I've tilted my build 90 degrees backwards which gets the bubble to exit the block. Everything is fine. However, a few hours or less later it appears again. I have recently "rinse and repeated" this 4 times, each with the bubble reappearing. I also rerouted my loop so the CPU block is first after the pumps. This didn't pop the bubble either.

I do not have air anywhere else in the loop. I don't have trickling, sloshing, or ticking noises. My loop is dead silent. This bubble grows and shrinks. As I am typing this it has become larger than the pics I have provided. Also my temps are fine. At time of typing My CPU is 29 Celsius, and GPUs are 27 degrees Celsius. I do have the CPU block and jet plate installed properly with the recommended inlet and outlet. I've read air in your loop can harm performance, but it doesn't seem to be harming mine. Unless I'm going to drop 5+ degrees with the bubble popped. I have also read people adding liquid detergent to the fluid to help with bleeding. I'm using Mayhems Pastel White, and I'm not sure how to implement that.

I am at the end of my rope. When I look at my build and see the bubble I feel the loop is not finished. Any ideas on how to fix, or pop this bubble??? :huh:

 

[SuB]

that's really strange.. the fact that it comes back almost sounds like there is a small, minuscule air-leak on the block to me..

Seems odd that it would collect there, if you leave it does it get any bigger or is that it?

It's not in a high-risk area as it's away from the centre so I don't think it will affect your performance much (don't quote me on that) It looks like there are lots of small bubbles as well ? That kinda makes me think the issue is indeed, with the block itself as you would think they'd have joined into the larger, main bubble by the time they collect there..

Very strange.

 

[JayCobs]

that's really strange.. the fact that it comes back almost sounds like there is a small, minuscule air-leak on the block to me..

Seems odd that it would collect there, if you leave it does it get any bigger or is that it?

It's not in a high-risk area as it's away from the centre so I don't think it will affect your performance much (don't quote me on that) It looks like there are lots of small bubbles as well ? That kinda makes me think the issue is indeed, with the block itself as you would think they'd have joined into the larger, main bubble by the time they collect there..

Very strange.

Odd indeed. I thought about the plexi being cracked, but wouldn't that create a leak and not bubbles? The block does not leak fluid. The bubble can grow larger, but not by much.

 

[SuB]

It depends.. but yes you would think so.. depends on the pressure difference. Does it re-appear while the system is off?

i.e. Shake/move the thing around to get rid of the bubble, does it re-appear while the fluid is still standing?

Can you give us a pick of all the fittings etc around the block instead of just a close-up? looks like you're using a long extension from the block?

 

[JayCobs]

It depends.. but yes you would think so.. depends on the pressure difference. Does it re-appear while the system is off?

i.e. Shake/move the thing around to get rid of the bubble, does it re-appear while the fluid is still standing?

Can you give us a pick of all the fittings etc around the block instead of just a close-up? looks like you're using a long extension from the block?

My rig is rarely off, so I don't know if the bubble appears in that state. I don't think the fittings are causing the bubble. And I KNOW my pump config is pushing the fluid strong. I would hate to replace this block because that is not in the budget.

 

[Remmy]

It seems to me like it may be the extension, it's O-ring isn't pinched or anything is it? And it's on nice and tight? The only other way I can see it being there all the time is if air in the system just collects on that specific bit for some reason

 

[SuB]

just to be utterly sure, the fluid is coming IN to the cpu block via that higher rised bit isn't it? Then the Ram, then GPU etc..

Are your GPU's in serial or parallel ?

I notice you've used the single o-ring crystal connections instead of the double o-ring. Are these all connected up okay? no slight off-angle tilts? or any way for air to get in?

P.S. I'm not doubting anything you've done I'm just trying to be thorough

 

[JayCobs]

It seems to me like it may be the extension, it's O-ring isn't pinched or anything is it? And it's on nice and tight? The only other way I can see it being there all the time is if air in the system just collects on that specific bit for some reason

That extension is fine. I have recently used 3 different fittings on this connection to the block and ended up with the same bubble.

just to be utterly sure, the fluid is coming IN to the cpu block via that higher rised bit isn't it? Then the Ram, then GPU etc..

Are your GPU's in serial or parallel ?

I notice you've used the single o-ring crystal connections instead of the double o-ring. Are these all connected up okay? no slight off-angle tilts? or any way for air to get in?

P.S. I'm not doubting anything you've done I'm just trying to be thorough

Yes: CPU block inlet - CPU block outlet - memory block - GPUs. All fittings are tight. C47 and C48 fittings are snug, all straight lines. Nothing off-center. GPUs are in a parallel config.

 

[SuB]

Hmm this is.. very strange...

If you lay the case on it's back...(i.e. so the block is facing up) and leave it that way does the bubble bugger off?

 

[JayCobs]

The "Off State" Test

Does it re-appear while the system is off?

Yesterday before work I disconnected all I/O, plugged pumps to external PSU, and bled the block by tilting it backwards almost 90 degrees. Then, I went to work. This is what it looked like before I went to work:

I returned home some 11hrs+ later and there are no bubbles. I turned my system on. 2hrs 50mins later, I get this:

10hrs 53mins later, I have this:

I am sooo outta ideas at this point.

What I know:
With the system off - no bubbles
With the system on - bubbles
I have no leaks
All connections are tight
Different loop configurations result in bubbles

Losing...my...mind! :cussing:

 

[PetetheSweet]

How fast is your pump running?

 

[JayCobs]

How fast is your pump running?

Normally both pumps are on 5. I have tried running the pumps individually on 5 with no dice. Do you think running them on a lower setting would help? Most of the tutorials I read a WHILE back unanimously advised running them on 5 for the bleed and then on 2 or 3 for low noise operation. Thoughts???

 

[lol]

Could it be that the coolant is somehow producing tiny gas bubbles when it heats up? Since it's forming the bubbles on the CPU block.

Or maybe it's reacting with the tubing? No idea if it's actually even a thing just throwing some stuff out there, maybe if you have the wrong coolant/tubing combination you get tiny bubbles.

That would explain why you get them when the cpu is in the first part of the loop. Maybe you need to turn down the pump to as low as you can, maybe it's sucking tiny air bubbles out of the res because they haven't had a chance to float to the top of the res yet.

If not that's another 3 things that are not the cause of the air bubble at least.

 

[JayCobs]

Could it be that the coolant is somehow producing tiny gas bubbles when it heats up? Since it's forming the bubbles on the CPU block.

I would go with that if my temps were horrible. My CPU is at 27 degrees Celsius at idle. Also this is the same equipment I used in my first build Projekt Suma Kuhl 2013 that did not produce a bubble. I'm going to turn the pumps down to one, then two, and so forth to see if that will work. I would've never thought my pumps could be pushing too hard to prevent bleeding. Thanks, and I will report back shortly.

 

[Warchild]

I think you have a blockage at the CPU.

Pump may not be able to push the water adequately due to maybe radiator flux trapped in the CPU copper plate. I think I had that problem like you.

When I took the block apart I noticed a small amount of residue built up with was enough to limit the flow from the "In" of the CPU block. It caused a bubble to form exactly like yours has. When system was off the bubble retracted inside the block so I didnt notice it.

 

[JayCobs]

I think you have a blockage at the CPU.

Pump may not be able to push the water adequately due to maybe radiator flux trapped in the CPU copper plate. I think I had that problem like you.

When I took the block apart I noticed a small amount of residue built up with was enough to limit the flow from the "In" of the CPU block. It caused a bubble to form exactly like yours has. When system was off the bubble retracted inside the block so I didnt notice it.

This was my last resort. Breaking down the build and taking a good look at the CPU block. With what you've said, I believe my early inexperience in liquid cooling has caught up with me. I can tell you that when I received my first two SR1s, I did not flush them properly. However, when I broke Projekt Suma Kuhl 2013 down I flushed the holy ghost out of all 3 radiators, and flushed the blocks. I remember how thin the space is between the fins in the CPU block, so I imagine it would take next to nothing to block one. Well, there goes Wednesday coming up. After the dentist I will go to work. Luckily the EK mount for the CPU block allows me to remove it without worrying about the back plate. This will reduce the turn around time.

Any pointers on the best way to clean a block? Very green to this part of the game, and would like to avoid as much noobness as possible.

Thanks in advance!!

 

[Warchild]

2 ways you can do it.

1. Vinegar - better result if left to soak over night but some elbow grease and a patient scrub period might work.

2. Ketchup - believe it or not but it does work. I'd sooner use vinegar myself.

And use an old toothbrush.

 

[Remmy]

2 ways you can do it.

1. Vinegar - better result if left to soak over night but some elbow grease and a patient scrub period might work.

2. Ketchup - believe it or not but it does work. I'd sooner use vinegar myself.

And use an old toothbrush.

I think the main reason ketchup works is that it has vinegar in it, so I'd just use the plain vinegar too as it's stronger. (At least I've seen people use vinegar to clean copper, I don't know about other materials).

 

[Warchild]

I think the main reason ketchup works is that it has vinegar in it, so I'd just use the plain vinegar too as it's stronger. (At least I've seen people use vinegar to clean copper, I don't know about other materials).

Wow.. what a Blonde moment I had. You are obviously correct. I should have realised its contents as I wrote my last post.

 

[JayCobs]

Heinz Distilled White Vinegar

I removed my CPU block and broke it down. This is what I found:

A while back I used some tubing from Home Depot. Lesson learned. Now I have Primochill LRT.

I let the block soak in vinegar for over 12hrs while I as at work. When I returned home I cleaned it up. Then installed it. Unfortunately I still have a bubble in the same spot.

What more can I do at this point? I'm fresh outta ideas here...