Question The absolute nightmare that has been my watercooling rig... Pump seizing edition

fureniku

Honorable
Aug 11, 2018
19
0
10,510
Hi guys!

So, this is an ongoing problem I've had coming up on 6 weeks now. I'd been posting various things on Reddit as I made or lost progress but I've come to the conclusion that an actual forum post would be much easier to maintain.

System specs:
Everything was purchased brand new August-September last year, except the 3080 + block which were mid October. System was built and first ran in early November.

Intel Core i9-10900k
NVIDIA GeForce RTX 3080 Founders Edition
MSI MAG Tomahawk Z490 Motherboard
64gb Corsair Vengeance 3000MHz RAM
Samsung EVO NVMe drive (plus a few SSDs, no mechanicals)
Corsair RM850X 850w Modular power supply
Lian Li O11 XL case
Windows 10
--- Watercooling ---
4x Corsair ML120 (Intake, sides initially now bottom over rad + I/O side), 3x Corsair LL120 fans (exhaust over rad, top)
Corsair XC5 CPU waterblock
Corsair XR5 360mm radiator
Alphacool Eisblock 3080 FE GPU waterblock
Initially Corsair XD5 pump/res, replaced with various EKWB/Lian Li G1 Distro plates with 3.1 pumps, eventually a 3.2 pump
Barrow Tech 360mm radiator added during upgrade/problem
Barrow Tech digital flow/temperature meter added during upgrade/problem


----------------------------------------
First, I'll start off with the story of how everything has happened and gone so far. There's a TL;DR below don't worry as it's pretty lengthy...

So in about mid July, I noticed that my PC was hot. very hot. CPU at almost 100c hot. I looked in my case, and realised my pump, a Corsair XD5, had died. I was getting no flow at all. The system had been running for about 8 months in total with all new parts. I was pretty annoyed, it was right as I was moving home from student accomodation so it was already a stressful time, but I contacted the store and set up an RMA.
At this point, my setup was the XD5 pump/res -> Alphacool Eisblock on a FE 3080 -> Corsair XC5 on an i9 10900k -> 360mm Corsair radiator -> back to the res. I was using EK Cryofuel cloud white fluid premix (which includes a biocide.) I had gone off how the setup looked, so figured silver lining; I'd swap over to the EK Distro Plate for my case, and sell the refurb/replacement XD5 when I got it back - that way I wouldn't have to wait a month to get up and running again either (lol). I ordered my Distro plate, some new fluid and tubing, and a few extra fittings I needed for the new setup. I also got a second 360mm rad, because the top rad always ran a little hot in my old setup anyway. It arrived and eagerly put it all together a few days later.

As I was stripping things out, I noticed there was some gunk on the inside of the tubes. foreshadowing.... I didn't really think much of it. I also noticed there was a tiny bit of algae in my CPU block and my res; which was strange, as my fluid had biocide in it. I disassembled the CPU block, cleaned it as best I could, and put everything together.

The new pump in the distro plate, a DDC 3.1, whirred away happily for my entire leak test. It worked for about a day after, before dying. During the leak test I'd noticed that the flow of fluid from the top of my case downwards was really slow, like more of a dribble than actual flow, but someone on reddit assured me it had just been due to air and it would resolve itself. Sure enough later the dribble had disappeared because the pipe had eventually filled with fluid.

Anyway, I cursed my bad luck that I'd just been sent an effectively DOA pump. Some googling had said these 3.1's were pretty weak and that it does happen. I returned it, got a refund, ordered another one the same. This time I used my brain; I bought one off eBay in slightly used condition. If the seller had it all working, then I'd know if it instantly died then the problem lied elsewhere - although it'd obviously have no warranty. A gamble for sure.

New one arrived, I fitted it, it died. I cried on reddit some more and people said once again its because these 3.1's suck, and the 3.2 and above are so much better. OK fine, replacing just the pump in this res would I guess be a little cheaper than an entire new pump/res, so I ordered a 3.2 from the only place that had one in stock, the next country over. It took a week to arrive, but during this time I treated myself to a digital flow/temperature gauge too.

I put it all together once again, turned everything on for the test and I had the same flow issue at the top. The meter was even registering 0.0L/m - this time, I took a video, so you guys can also see the problem. I'd also learned my lesson from before, and prompty turned off the system so I wouldn't kill yet another pump.

Off to reddit and people said one of my components must be very restrictive. By this point I was moving back to university, so once I'd settled in I stripped it all out again, and rebuilt with one part at a time. With just my radiators, the flow meter was saying 10-11L. With CPU and rads, it was dropping to about 6L, which is still plenty, and with just GPU... 0.24L. Ok, there's my problem! finally! I took the GPU apart and what had looked perfectly fine through the plexi, was actually a perfectly and completely sealed intake fins covered in algae. It was gross I wont lie, but I spent a good few hours cleaning out every last spec of the stuff. Rebuilt the whole system, and I was getting around 2.8-3L/m on the flow meter. I'd figured as the CPU had dropped it to 6 anyway, that was probably about right with GPU and CPU together. I finally started another leak test.
I ran it for about 8 hours and all was fine, then stopped it overnight so I could sleep. I even noted that it was still saying 2.94L as I shut it down that night. The following morning, the pump once again wouldn't start, and I felt like dropping to my knees and shouting no like they do in movies. I was completely stumped, I had no idea why the pump wouldn't even start now when it had been working when I flicked the switch off the night before.

So, I've stripped it all out again. I removed the pump, and it seems to have seized up, I couldn't even turn it with my hand. I reseated the spinny bit (technical name?), and set up a loop that was just the pump into the flow meter, back to res/pump. This got about 15L/m, but the pump started getting pretty warm with no cooling so I added a radiator, which dropped it to about 7.5L/m. That's where I am now, and it's showing 7.5L/m.
----------------------------------------


TL;DR
Corsair XD5 died, replaced with a DDC 3.1 which died, replaced with another DDC 3.1 which died, found that my GPU was caked with algae which was probably burning out pumps, rebuilt and the DDC 3.2 died, turns out it was just seized so I've freed it and it's currently running... for now

I'm currently planning to run just this pump->meter->one rad (to keep pump cool)->res setup until Friday evening, turning it off at night and making sure it restarts in the mornings. But, I wanted to get a forum thread going and gather some feedback on things;
  • What do you think could've caused it to seize up that last time, and how could I prevent that happening again?
  • Was 2-5-3L/m actually a reasonable flow rate to expect for a DDC 3.2 with two blocks and two 360mm rads?
  • Considering my replacement XD5 arrived today (brand new and sealed), would I be better off just calling it quits and going back to that, or do you think I can get this rig working? (I'd really like to just sell it as new to recuperate some costs)
  • How can I prevent such a bad algae build up from just causing all this again in another 6 months or so? My fluid (XSPC EC6) claims it needs no additives but doesn't actually say it has biocide, should I add some to be safe?
Thanks for any guidance anyone can offer, I know this has been a super long post but I'm just so stressed about it all and I'm on the verge of just giving up on water cooling, but I really don't want to if I can salvage it...

Update 1:
It's thursday morning (ish). The flow meter has died :) showing 0.0L despite visually good flow.
Reconfigured the layout so I can visually see flow into the top of the res instead. Pump still working and I can see I'm getting good flow. Who knows what L/m though :)
 
Last edited:

Ralston18

Titan
Moderator
Update your post to include full system hardware specs and OS information.

Include PSU: make, model, wattage, age, condition?

Which fans are intake, which fans exhaust?

Algae:

Out of my comfort zone somewhat but one immediate question regarding " I was using EK Cryofuel cloud white fluid premix (which includes a biocide.) "

Where are you purchasing the premix?
 

fureniku

Honorable
Aug 11, 2018
19
0
10,510
Update your post to include full system hardware specs and OS information.

Include PSU: make, model, wattage, age, condition?

Which fans are intake, which fans exhaust?

Algae:

Out of my comfort zone somewhat but one immediate question regarding " I was using EK Cryofuel cloud white fluid premix (which includes a biocide.) "

Where are you purchasing the premix?

Details are added to OP at the top.

The premix was purchased from Amazon UK: https://www.amazon.co.uk/dp/B07HYGYS6N/ref=pe_3187911_185740111_TE_item
 

Ralston18

Titan
Moderator
Do the following links match the Installation and Instruction Manual for the currently installed pump?

https://www.pumpvendor.com/media/laing/Laing_DDC_Series.pdf

https://www.ekwb.com/shop/EK-IM/EK-IM-3830046995513.pdf

Again, full disclosure, I am going out of my comfort zone here.

However, there is a consistency with respect to pump problems and that, to me indicates some common issue.

Are you able to take a couple of photographs of the cooling setup?

If so, do so, and post the photographs here via imgur (www.imgur.com).

Hopefully someone may spot something that has gone astray.
 

fureniku

Honorable
Aug 11, 2018
19
0
10,510
Do the following links match the Installation and Instruction Manual for the currently installed pump?

https://www.pumpvendor.com/media/laing/Laing_DDC_Series.pdf

https://www.ekwb.com/shop/EK-IM/EK-IM-3830046995513.pdf

Again, full disclosure, I am going out of my comfort zone here.

However, there is a consistency with respect to pump problems and that, to me indicates some common issue.

Are you able to take a couple of photographs of the cooling setup?

If so, do so, and post the photographs here via imgur (www.imgur.com).

Hopefully someone may spot something that has gone astray.

Yes, that's the right pump series. They come in 3.1, 3.2, 3.25 and 3.4 variants, with the 3.1 being a "low power", running entirely off PWM. the other three are pretty much the same with different methods of powering (molex/PWM, molex only, SATA/PWM). Currently I have the 3.2.

The rig is still stripped in the minimal layout currently, but you can see the cooling setup in the video in my OP (
View: https://www.youtube.com/watch?v=MSqsdDQ9M6Y
)

The pump seized up again yesterday, so I've started another RMA. Further searching around suggests this is still pretty common on the DDC series... I don't think I should be getting it this frequently across three different pumps though, but I think the only real option here is giving up on DDC pumps. Which is a shame because I love the back-mounted distro plate, it makes everything so much neater; I'm not a fan of the front-mounted version which has a D5 pump (plus its over double the price... costs more than the case itself).

One thing I've read someone in a similar situation did was remove the impeller from their DDC pump so fluid could freely flow past it (effectively just to plug the hole for it), then installed a D5 pump on the old pumps output. I like the sound of this idea as a way to keep my distro plate while having a D5 pump, but I'm a little concerned about the feed. The distro plate is designed so water flows down into the DDC pump, which pushes water back up out of the output towards the GPU - that output is where I'd put my new D5, but I don't know if gravity would be able to push water round and back up through there fast enough?

Plus, I'm still concerned that something must be causing this on the DDC pumps. Other people with similar setups and even the same fluid have had their rigs running for months with no problems, so why is mine dying so much? How do I know that even after this, the D5 wont just die in 2 days the same way?
 

HappyTrails

Upstanding
Oct 30, 2020
286
34
340
Sorry you have the so many issues can't imagine what may be happening. Have been running EK distroplate here with original pump. Its the pump everybody says dies very soon but I must be lucky its still ok over a year now. Have thoughts of buying a 3.2 ddc to have around as backup. On advice from others I installed an EK heatsink even thought pump ran cool it seems.

Algae had happen here too my fault. Run clear PETG and inhibitor/biocide so its easy see when algae has taken over. It wasn't effecting anything but look bad. Cleaning loop was a chore that didn't make me happy though. Hope to add more biocide each month and see if does better.

If you think loop is very clean then only suggestion here is to add something so algae doesn't come back. For me I like water with adding stuff like biocide but people say there are other premix coolants they like that do good. :)
 

fureniku

Honorable
Aug 11, 2018
19
0
10,510
A couple of the threads I'd found yesterday mentioned "black gunk" on the bearing where the impeller mounts. I took the impeller off two pumps (the current 3.2, and the 3.1 that came with my distro plate that I thought was dead), and sure enough both had this black gunk on which is why they're sticking.

My fluid is blue, and others have had all variations of fluid colour but the gunk is always black. Any speculation what it might be? Pictures here, I've cleaned the 3.1 (right) for comparison of how it should look. The 3.2 had ran for maybe 12 hours since I'd last disassembled it,and about 20 hours for its entire lifetime, so this is a pretty fast buildup of whatever this is.

MevHJnd.jpeg


6KapG1E.jpeg


So it's likely whatever this is, is what's seizing up the pumps. I can manually clean and free them but I dont exactly feel like draining my system every day, so I'm going to continue with my in-line D5 idea. I've got one on order that will arrive tomorrow, but for now I'm going to test the theory using these two (one with impeller removed as the hole plug and a second to act as the actual pump in place of where the D5 will go, just to make sure its able to flow enough)
 

fureniku

Honorable
Aug 11, 2018
19
0
10,510
So I want to post an update on this, if nothing else just for the sake of future Googlers (don't you just hate finding threads without the full info?)

So, the D5 seized up the same after about the same amount of time. I took it out, opened it, and the bearing was clean. Then the penny dropped, and D5's are slightly different - the impeller was full of residue instead.

I cleaned it out, it was about the same amount as the closer pump in my last picture. And then I stripped out everything from my PC and gave it a full deep clean. It took me like 6 hours... I just went to town on it. My flatmate thought I'd set up some weird science experiment in the kitchen as I was using one of my many pumps to send an infinite amount of tap water through the radiators.

They were the problem, by the way. I had of course flushed them both but I don't think I'd done the barrow one well enough - or they're just bad, I'm not sure. Even after 2 hours of continuous fresh flow, I was still getting tiny black particles coming out of it. These particles were mixing with the dye particles in my coolant, and coagulating on the pump's bearing, which is what gave it that black with a hint of green (which was actually UV blue).

I stripped out my blocks again and cleaned them too. Just from the tests, the CPU block (which has tiny fins, you can't even get a toothbrush between them) had gunked up with similar residue. I finished up by assembling the parts on my kitchen table, and put deionised water through it. I did a few cycles of this, flowing it for about an hour then throwing out the water and replacing it with more. The last one I left running overnight. Checked the pump this morning and it was clean! It's now built back into my PC with a fresh bottle of the UV coolant and doing yet another leak/pump test, so we'll see how its fairing this time tomorrow. I'm not going to call it "fixed" for another two weeks or so though, don't want to jynx it.

Oh, and the exact same problem had happened in my flow meter. Cleaned that out and it runs fine again now.

So, in hindsight I think the barrow radiator was causing my current problems. But all in all it was a perfect storm which is what made it so hard to debug. I believe the following happened:

  • The buildup of algae killed my original Corsair XD5 pump, because flow was restricted so badly it burnt out.
  • "I know great idea lets add a new radiator while we fix these"
  • Cue all the following pump failures being caused by the new radiator, not the same thing that killed the first.

On the plus side I've learned a lot about water pumps and fluid physics while doing all this! I've decided I like the new layout with the external D5 better than the weak little DDC in the back, so I'm going to keep it like this anyway. Y'know, unless this pump dies by the weekend and I throw the whole rig out the window.

The real takeaway from this though I think is summed up well by a JaysTwoCents video I saw. If you think your pump is dead, before you do anything else, check the impeller. Check for that buildup, because apparently its a really common problem (as far as failing pumps go) and could be anything, not just radiator gunk. After all this I'd even go as far as saying that should absolutely be a point of cleaning when doing loop maintenance, which I haven't seen anywhere before. It's so easy to disassemble them and clean them, the impeller itself is magnetic and just pulls off.

Lets hope this is my last post on this thread at least!