Question Buzzing noise coming from Deepcool LT720 AIO on a delay

[kybriggs]

I just built a computer and I have a worrying issue. The PC boots normally and is almost completely silent at first. At some point, usually between 2 and 60 minutes after boot, a very loud buzzing noise starts to happen. Using a stethoscope, I have narrowed it down to the CPU cooler block that sits on top of the CPU (the pump, not the RGB cover). I am reasonably sure it is not a trapped bubble (very regular, electrical buzzing sound, whereas a bubble in the loop is more irregular and distinct, especially with a stethoscope).

The onset of the noise is variable after boot, but once it starts, it does not stop on its own and goes away when rebooted. CPU and GPU load have no impact on the noise, it sounds the same when idle or when running Prime95 and Furmark at the same time. Temperatures appear to be normal, cooler is otherwise functional.

A video of the noise can be found here:

View: https://youtube.com/shorts/otthfOFTqr4?feature=share


Is this fixable or do I just return or RMA?

 

[Paperdoc]

So this noise, once it starts, does not stop. But then if you reboot it is GONE immediately? That does not sound like a mechanical or electrical problem. It DOES sound like a small air bubble get trapped and is released when the pump stops and re-starts. That makes me think in terms of how air can get to the pump? Is the pump higher up than other items like the rad?

 

[kybriggs]

So this noise, once it starts, does not stop. But then if you reboot it is GONE immediately? That does not sound like a mechanical or electrical problem. It DOES sound like a small air bubble get trapped and is released when the pump stops and re-starts. That makes me think in terms of how air can get to the pump? Is the pump higher up than other items like the rad?

Your logic is the same as mine at first, but a few things make me question this:

1. The radiator is mounted above the block, so the block is the lowest point in the loop and should be easily able to shed a bubble.
2. Tilting the computer around as no impact on the noise. I would expect at least a qualitative change in the noise with angle/gravity change
3. Changing the pump speed has no impact on the noise. I would expect the frequency of noise caused by a bubble to be related to pump frequency, but it is not in this case
4. I can listen to the block with a stethoscope, and I am familiar with liquid cooling/pump systems generally through my work. Bubbles have a distinct sound, and I do not hear it with the stethoscope.

I should also clarify: the noise goes away when the system is SHUT DOWN, left to sit, and then restarted, it does not go away if all I do is reboot. That fact at least does support the bubble theory, but the rest do not - whence my confusion.

I am currently trying to boot my computer on its side to change the orientation of the flow IO with respect to the block (while keeping the cooling block below the top of the rad) - will update after running long enough to be sure the noise would be there if it was going to be.

 

[kybriggs]

@Paperdoc Seems it may have been a bubble after all - running from boot on its side and then tilting the computer so that the IO ports to the cooling block were pointing 45 degree upwards while keeping the edge of the radiator above the block and shaking (gently) caused an audible release of air.

Once I reoriented it the noise came back quickly though. Not sure if some quick of the fluidics are causing a bubble to be able to loop through, or if there is an air pocket trapped in the cooling block itself. I will keep shaking/running on its side for now. If that fails, long term fix might be to just rotate the cooling block 90 degrees so that the pipes are oriented in the direction they are when the machine is sideways and the noise does not occur. Minor PITA since I'll have to reapply paste, but not the end of the world since the block is small enough that I can fix this without disassembling the rest of the computer.

Thanks for making me revisit a problem I thought I had ruled out!

 

[Paperdoc]

Very thorough investigation on your part. Congrats! This is NOT what I would think of as a "typical" problem easy to find.

 

[kybriggs]

@Paperdoc in the end I went the RMA route. As part of the RMA Deepcool support asked me to destroy the LT720 and send them proof of destruction instead of returning it, so I was able to drain the liquid from the cooler. In the liquid, I found a small flake of plastic. Not a bubble after all:

https://imgur.com/SY47MKU

View: https://imgur.com/SY47MKU
(upper left region). Seem like just a freak QC failure.

Deepcool support was responsive and helpful, so no harm no foul. 10/10 would RMA again.

 

[Paperdoc]

Thanks for the update. I can understand that loose debris in a loop can be a problem, but I have seen no other reports of that. In most cases of failure nobody disassembles the system to diagnose as thoroughly as you have, so this may be more common than we suspect.

 

[MeFor3]

I have a similar issue with the LT 720 but it’s more of a buzzing and it’s louder. I opened a ticket with Deeocool. Any advice?

 

[kybriggs]

@MeFor3 Not much beyond what I noted above:

To rule out a trapped bubble, dismount the cooler and hold it so that the CPU block is the lowest point and the hoses are oritented up toward the rad. Shake it to dislodge any bubbles from the block. Orient the rad so that any bubbles in it will go the end away from the hoses. Remount it, taking care at all times to make sure the CPU block stays below the rad so that no bubbles can backtrack. Mount it with the hoses pointing upward toward a top-mounted rad.

If the noise persists after doing that, RMA.