CPU overheating but AIO pump is working

Status
Not open for further replies.
Mar 14, 2018
5
0
10
Hi
My Ryzen 5 1600x is running idle at 38-42 and under 100% load up to 82 degrees celsium.
I dont think that this is ok cuz im using AIO liquid cooler with 240 radiator - its an ID-Cooling Frostflow 240l + im using good Prolimatech thermal compound.
What was wierd is that under the load i could hear pump working but the radiator was barely warm.
Fans are ok, i aplied thermal paste as required with very thin layer, tried to repaste it as well and im mounting pump properly.
I switched back to my Cooler Master Hyper 412 and got 29-30 idle and 52-53 under same 100% load as with watercooler.
I plugged AIO back to its header on a mother board and i can hear its working, bios shows its rpms around 2200 at 100%, its a 3 pin pump i tried DC pump control settings in bios, automatic, disabled. Tried other headers (fan) - no result.
Tried Arcticcooling MX-4 instead Prolimatech - no result as well.
I can even hear a noise of a running water through a pump, so why its not cooling my cpu properly? What could be a reason?
 
Solution
In the event that your AIO's impeller has been cavitating due to an air bubble, by changing the relative positions of the case and pump to accommodate your clever "iron" test may have caused the bubble to migrate from the pump to the radiator. This would result in the pump regaining flow, which in turn would transfer heat from the iron to the radiator, as one might expect.

From the information given, it appears the original problem may have been caused by an air bubble, which has increased over time due to permeation.

Depending on where the radiator is positioned relative to the pump, you can approximate the conditions of your "iron" test by once again remounting the pump, and positioning the case in such a way so that the entire...

theo.esk

Proper
Mar 14, 2018
80
0
160
i think you've misplaced the 3-pin pump plug. it's designed to connect to the PWR_FAN 3-pin header on the motherboard.
The PWR_FAN 3-pin header is: 1 = gnd 2 = +12v 3 = sense
What this do is powering the pump and sensing the rpm in the pump, you cannot control the speed of the pump with this pin header, it is designed to run at max rpm at all times. which is what you'll want the water pump to do in an AIO system.
 
Mar 14, 2018
5
0
10


Thx for the suggestion, but as i said i tried to plug it in fan header as well. Moreover, that header on my mobo is designed especially for aio pump and supports 3-pin headers as it said in manual and i can see pump rpm`s in bios. So no, its not the case :/
 

CompuTronix

Intel Master
Moderator


kurohtinp,

On behalf of Tom's Moderator Team, welcome aboard!

Your pump is failing, regardless of what you "hear" or the RPM indicated in BIOS. Decreased radiator heat output also indicates poor circulation. Just like a swimming pool pump, the impeller is magnetically coupled to the motor, so it's not directly shaft driven. This eliminates shaft seals and their potential for leaks.

If the motor hasn't failed due to excessive runtime hours, it will continue to rotate, even if the impeller has "stalled" due to the accumulation of sediment through galvanic corrosion between the copper waterblock and the aluminum radiator. It's also possible the impeller is now "cavitating" due to air accumulated from decreased water volume through gradual permeation over time. So what you "hear" may be mostly cavitation, rather than actual flow.

These problems are inherent to "maintenance free" AIO's, which custom loops don't have since they're all copper, and are routinely maintained through periodic flushing, cleaning and refilling. Since you reinstalled your Hyper 412 and your temperatures returned to expected norms, you've already proven the ID-Cooling Frostflow 240l's pump is failing. Hopefully it's still covered under warranty and can be RMA'd.

Once again, welcome aboard!

CT :sol:
 
Mar 14, 2018
5
0
10


Thx for the welcome! As the matter of fact i have exactly same thoughts as you are, well its only one logical reason which is left. Ill give it back tommorow to a seller and i will update this thread with the answer. Thx again!
 
Mar 14, 2018
5
0
10
UPDATE: My warranty period is over, it was anly 1 year. So i decided to experiment further. I took and ordinary iron, applied thermal paste to a pump, plugged it into a mobo header, heated iron rly good, and pressed a pump to its surface. And guess what? Aio radiator became hot! I have literally no idea what is going out...
 

CompuTronix

Intel Master
Moderator
In the event that your AIO's impeller has been cavitating due to an air bubble, by changing the relative positions of the case and pump to accommodate your clever "iron" test may have caused the bubble to migrate from the pump to the radiator. This would result in the pump regaining flow, which in turn would transfer heat from the iron to the radiator, as one might expect.

From the information given, it appears the original problem may have been caused by an air bubble, which has increased over time due to permeation.

Depending on where the radiator is positioned relative to the pump, you can approximate the conditions of your "iron" test by once again remounting the pump, and positioning the case in such a way so that the entire radiator is unmistakably higher than the pump.

You can then power up into Windows and perform your previous load tests using your monitoring utility to confirm the puzzling results of your iron test.

CT :sol:
 
Solution
Status
Not open for further replies.