The pump continues to malfunction within the tailored water cooling system.
The pump continues to malfunction within the tailored water cooling system.
Hi guys, asking here is a last resort for me because I can usually find what I need just by browsing, but now I'm completely lost. Approximately a year ago I bought components for a custom water cooling loop which, after much research, I assembled myself. I paid attention to the water level in the reservoir while filling the loop (bay reservoir) to avoid the pump (an Alphacool solo VPP655) running dry. After a couple of months, at one point the pump started making strange clinking noises. When I took the loop apart, I realized that the noises were caused by the impeller constantly making contact with the metal housing where the impeller fits. After having the pump sent back to the retailer, I learned that the cause of the failure was the plastic part, where the impeller sits on the ceramic bearing, getting damaged and thus lowering the impeller. According to the retailer, this was allegedly caused by the pump running dry which baffled me even then. Like I said, I was very careful when filling the loop and I also checked the water level in the reservoir at the time of the failure, which hadn't decreased at all. I also checked the radiator, the CPU block, everything where water should be flowing, but I found no blockages (the only thing I found is a little bit of black oxidation of copper inside the CPU block, but I assume it's very unlikely that it was the reason for the failure). The retailer offered me a new impeller for a relatively fair price, so I went with it and reinstalled the whole lot.
Fast forward a couple more months to the present. A few days ago I started hearing the same clinking noise, and I immediately knew I was in trouble yet again. I'm almost entirely sure that it's the same problem again, and I find it extremely unlikely that the pump indeed ran dry - like I said, water level hasn't decreased, the pump is directly bolted on the back of the bay reservoir and there were no leaks. Even if the pump did run dry (for example when I was filling it), why would it take 3-4 months for it to fail?
The only thing I can think of is that the water in the loop got too hot and slowly eroded the plastic, although I would assume hot water is something you anticipate while designing a water cooling pump. I'm running a fairly aggressive OC, 3570K@4,6GHz and roughly 1,35V. The CPU is delidded, new thermal paste everywhere, although it was still running fairly hot (close to 100 degrees Celsius under stress testing - I know, that's not a good number in a custom water cooling, but I guess I was unlucky with the silicon lottery).
Any ideas are greatly appreciated, although I'm primarily interested in the likely cause of the failure because I'm probably moving back to good old fashioned air cooling, this takes too much maintenance and money.
Can you share a couple of pictures of your loop routing? D5s usually work well, unless they're running dry... maybe air is stuck in the system and moves to the pump while it runs. The photos would help confirm if this is likely. Also, let me know which radiator you have, the type of tubing, etc.
The temperatures are quite high for an i5 @ 4.6, and the voltages are even higher with an AVX load—especially since it's delidded.
Sure thing,
here you go.
Rad:
Alphacool NexXxoS UT60 Full Copper 360mm Radiator
CPU block:
Alphacool NexXxos XP3
Tubing I can't recall at first, but it's definitely smaller than usual.
It's highly unlikely the radiator ran dry, I gave it a good shake for about ten minutes after filling the loop and adjusting the pump speed. If it had dried out, I have no idea what would happen, honestly—it seems like an impossible situation given what we know. I plan to take the loop apart soon and check everything carefully, but I don't expect to find anything, just as before.
Regarding the temperatures, I'm aware I probably didn't get a proper fit with the block like the first time (before the first pump failure it was stable at 60 degrees under load). I'm not sure if that matters, but the pump was set a bit lower than the second speed (raising it didn't change the temps much, just made it louder).
It wasn't the heat itself that led to failure of those parts—most pumps are built to handle temperatures above a typical watercooling loop, such as water temperatures exceeding 50-60°C or more. (Keep in mind, water temperatures differ from core readings.) What pump are you referring to? That's the key piece of information.
It seems to be a pump and reservoir unit combined, as indicated by the photo. Also, it appears to be a strong pump.
My reservoir is my setup. As you see, the pump is securely attached to the back, making it nearly impossible for it to run dry as long as the reservoir stays upright and has enough water—this explains why I'm upset about the retailer saying it runs out. I know this is the usual concern in such situations, but based on what I've seen, it doesn't seem likely. I hope you understand my frustration.
flamboush :
Sure thing,
here you go.
Rad:
Alphacool NexXxoS UT60 Full Copper 360mm Radiator
CPU block:
Alphacool NexXxos XP3
Tubing I can't remember off the top of my head, but I know that it's smaller than usual.
It's extremely unlikely that it ran dry, I rocked the case for a good 10 minutes after filling the loop while also varying the pump speed. If it ran dry, I have no idea how... frankly, it seems impossible given the facts. I'll take the loop apart in a few days and inspect everything closely, but I don't expect to find anything, just like last time.
As for the temps, I know, I probably didn't get a good fit with the block like the first time (before the first pump failure it was stable 60 degrees under load). I'm not sure if it matters, but the pump was set slightly below the 2nd speed (increasing it seemed to have made no difference in temps, it only made it louder).
From the pictures, it looks pretty unlikely that the pump would run dry. It guess it's possible to have just gotten a bad pump, but.....it's really rare. The D5s are usually VERY durable.
As rubix said, the pump isn't going to fail because of coolant temps, which...are going to be considerably lower than your core temps anyway, but.....those temps on an i5 with those settings, reaching up to 100c seem high.
And the reservoir has always been fully filled? Yes, it has. I captured a photo of it here. It's not very clear, but you can see the water level is above the 75% mark. When I was filling it, I kept it at least at 75%, never dropping below that point, even if it did, it wouldn't make sense to explain why it failed later on.