Is my water pump dead?

[RenaKry]

I have a Deepcool Captain 360EX (the second generation red one with the longer braided hoses, not the RGB or plastic hose) cooling an overclocked i7-5820k at 1.35vcore and 1.8vin. The temps when the kit was new were good and kept my CPU well below 85C (the thermal limit that I set in the BIOS myself) under FPU loads using AIDA64 but now AIDA64 tells me the CPU is throttling between 10-20%, with CPU speeds dropping from 4.4GHz to as low as 4.0GHz according to HWMonitor. As soon as the stress test was ended, the temperature shot back down to the high-50s. This is an unrealistic load for this computer but my main concern was looking at my motherboard's POST display that was set to read CPU temps while it was idling and it was sitting steadily in the mid-50s, which is what led me to stress my computer to confirm my suspicions. For a water-cooled computer, mid-50s on idle with adaptive voltage is unacceptable, and I would like to know which part of the cooling system has failed. My money is on the pump, but something about it doesn't sit right with me. If the pump was dead, the heat would be stuck in the cold plate and the computer would crash from overheating, yet it stayed just above the hard thermal limit of 86C the whole time I was stressing the CPU with superficial throttling. I don't believe the radiator is clogged with dust nor do I think it is that hot in the room, but I could be wrong. Anyways, what do you guys think? Is it time for me to move on to custom loops or go back to good-ol' air?

[Semper]

A custom loop is an option, but be prepared for a high level of diminishing returns on cooling potential. It's not unheard of for a custom loop to carry a sticker shock of $500-600.

Was this change sudden, or has it been a thing that's changing slowly over time? see if you can feel any vibration (or hear any pump noise) coming from your block. The 360ex has a clear tube coming out of the block for water transfer, correct? see if you're able to identify any movement in that tube, things such as air bubbles. My primary concern is that you're running low on coolant, especially if you've been running the 360ex for five years.

[suchamoneypit]

Sounds like a pump issue for sure. If the CPU heat soaks, meaning once it gets hot it stays hot and starts thermal throttling, almost certainly the pump is dead or very weak. You can easily verify this by feeling the pump during operation. A dead pump will have no vibration at all at the cooling block (where the pump is on most AIOs and appears to be on yours) as the pump is working, and you will also feel the piping close to the cpu block and the block itself will be quite warm, or even hot to the touch while the tubes further away from the block will still be cool (because its not moving the hot water away). If the pump is dead do not keep heat soaking the non-moving water around the CPU block. Not only will it not get cooler, when it gets super hot, this heat doesn't move away, it sits right on the CPU and cooks it. I've had friends kill CPUs because of this.

[RenaKry]

12 minutes ago, Semper said:

A custom loop is an option, but be prepared for a high level of diminishing returns on cooling potential. It's not unheard of for a custom loop to carry a sticker shock of $500-600.

Was this change sudden, or has it been a thing that's changing slowly over time? see if you can feel any vibration (or hear any pump noise) coming from your block. The 360ex has a clear tube coming out of the block for water transfer, correct? see if you're able to identify any movement in that tube, things such as air bubbles. My primary concern is that you're running low on coolant, especially if you've been running the 360ex for five years.

 

7 minutes ago, suchamoneypit said:

Sounds like a pump issue for sure. If the CPU heat soaks, meaning once it gets hot it stays hot and starts thermal throttling, almost certainly the pump is dead or very weak. You can easily verify this by feeling the pump during operation. A dead pump will have no vibration at all at the cooling block (where the pump is on most AIOs and appears to be on yours) as the pump is working, and you will also feel the piping close to the cpu block and the block itself will be quite warm, or even hot to the touch while the tubes further away from the block will still be cool. If the pump is dead do not keep heat soaking the non-moving water around the CPU block. Not only will it not get cooler, when it gets super hot, this heat doesn't move away, it sits right on the CPU and cooks it. I've had friends kill CPUs because of this.

I do not feel any vibration when touching the block under load, but it doesn't seem to get hot either. I've had this for probably 2 or 3 years now, and I don't think coolant will just evaporate out of a sealed closed loop system. I guess my water pump really is toast. Now, where is my old Hyper 212 Evo?

[Semper]

1 hour ago, nelsonpong said:

 

I do not feel any vibration when touching the block under load, but it doesn't seem to get hot either. I've had this for probably 2 or 3 years now, and I don't think coolant will just evaporate out of a sealed closed loop system. I guess my water pump really is toast. Now, where is my old Hyper 212 Evo?

Permeation is a thing, you lose fluid over time, open or closed loop, it's physics, and there's no such thing as a free pass when it comes to physics . It's possible that it's happened in two or three years (check to see how long your warranty is, and your actual purchase date, you might still be in warranty), but i'd be surprised for it to happen that fast in a closed loop.

If your pump is running at all, under load or not, you should feel vibration. If your pump is set up for self-regulating flow based upon temperature, load would create more vibration, yeah, but you're feeling none at all, which means it's much more likely that your pump is toast.

[suchamoneypit]

18 minutes ago, nelsonpong said:

 

I do not feel any vibration when touching the block under load, but it doesn't seem to get hot either. I've had this for probably 2 or 3 years now, and I don't think coolant will just evaporate out of a sealed closed loop system. I guess my water pump really is toast. Now, where is my old Hyper 212 Evo?

liquid evaporating from closed loop systems is a real thing