Help/Advice on a watercooling build

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Voodoo Rufus

Powder Junkie Moderator
 
 
 
Joined
Sep 20, 2001
Well for one, you are way over-radiator'ed. Running your radiators in a 2x2 arrangement like you switched to today would definitely increase flow, at the expense of flow velocity in the individual radiators. At ~.8GPM per rad, it won't be an issue though. Your dual D5's are fine.

The bubbles in the reservoir will leave over time.
 

spiritedandy

Premium Member
 
 
 
Joined
Apr 12, 2003
Location
Middlesbrough
Like rufus said 2 D5s fine i run 2x 480 rad cpu and gpu . Also it goes into a cooler some 2 to 3Lt and at moment ive go rig on table benching so up hill 2 ft too and all ok easily ok .

Bubbles will go just takes time looks bad clears good
 

distilld

Joined
Oct 22, 2021
Over the course of 3-4 hours my flow rate on my AquaComputer High Flow Next dropped from ~340 L/h (~1.5 gpm) to ~227 L/h (~1 gpm). Is this to be expected when first filling a loop or should I start thinking about taking the CPU block apart to check for clogs?
 

EarthDog

Gulper Nozzle Co-Owner
 
 
 
Joined
Dec 15, 2008
Location
Buckeyes!
Unless something is in the loop, flow rate shouldn't change much over such a short time.
 

distilld

Joined
Oct 22, 2021
Interesting: I turned the loop off for about an hour and then turned it back on again and the flow meter shows that the flow is back up to 1.25 gpm. I don't see any debris circulating so if that's what it is it must be really small.

I guess I'll take apart the CPU block to investigate when I drain the loop to put my replacement GPU block in. :(
 

Voodoo Rufus

Powder Junkie Moderator
 
 
 
Joined
Sep 20, 2001
Not sure. Maybe still bleeding some bubbles out. Given enough time the flow should stabilize. I've never used a flowmeter in a cooling loop before, but fluid density/viscosity will change the flow velocity.
 

distilld

Joined
Oct 22, 2021
I took the CPU/mobo monoblock apart and sure enough there was some debris built-up on the CPU fins. I cleaned it all out but now trying to get the o-rings back in place (and making them stay there!) makes me feel like I've graduated to a new level of hell. They all seem to be too big to fit in their original positions.
 

distilld

Joined
Oct 22, 2021
After many (many, many) hours, three (or five) different iterations on the tubing layout, three M2 drives, two sets of water blocks and back plates for the GPU, two motherboards, a damaged radiator, three busted sticks of RAM, and 15 meters of acrylic tubing, I’m finally putting the sides on this and calling it “good enough for now”.

IMG_3276.jpg

There were times (I'm sure you knew) when I bit off more than I could chew. But through it all - when there was doubt, I ate it up and spit it out. I faced it all, and I stood tall.

The record shows: distilld took all the blows. But...

i-did-it-my-way-my-own-way.gif
 

Attachments

  • IMG_3276.jpg
    IMG_3276.jpg
    2.3 MB · Views: 28
  • i-did-it-my-way-my-own-way.gif
    i-did-it-my-way-my-own-way.gif
    3.6 MB · Views: 25

distilld

Joined
Oct 22, 2021
I had completely forgotten that I ordered a filter from Aquacomputer several weeks ago (back when I posted about stuff clogging up my GPU block). It finally arrived today, so I installed it and took the opportunity to redo the GPU runs so that they'd be a little more level. The filter reduced my flow rate by about 20 gph but I'm still in the 1-1.5 gpm zone so I'm not going to sweat it. I'll be done again now I hope.

Updated screenshot, through the case window this time (in the last picture I still hadn't put the sides on the case):

Spoiler:


And here's a wider shot of my workspace now that I've finally cleaned up since the mess I posted earlier in the thread. Hopefully this redeems me a little bit.

Spoiler:
 

Attachments

  • IMG_3297.jpg
    IMG_3297.jpg
    2.1 MB · Views: 19
  • IMG_3302.jpg
    IMG_3302.jpg
    2 MB · Views: 18

distilld

Joined
Oct 22, 2021
Things have been (seemingly) running alright for awhile now. I still find my self having to tilt the case about twice daily to clear large bubbles out of the GPU block.

Tonight I took a closer look at the GPU block and noticed this. I'm afraid that it looks like it could be water outside of the o-ring between the plexi and metal... Is there something else it could be? The plexi doesn't look like it's actually damaged anywhere, and I can't see any clear points of failure on the o-ring itself. The screws holding the plexi to the metal are very tight - I tried to tighten a few of them to be sure. If it is coolant then what could've caused this (and what did I do to deserve this luck)? The system was on and running when I took these. It has been running nearly continuously over the past week.

How worried should I be?

Spoiler:
 

Attachments

  • IMG_3322.jpg
    IMG_3322.jpg
    775.9 KB · Views: 6

EarthDog

Gulper Nozzle Co-Owner
 
 
 
Joined
Dec 15, 2008
Location
Buckeyes!
Once bubbles are gone.... they should be gone. I don't recall ever having a loop where bubbles came back...o_O

I dont what's wrong, butthat doesn't feel right... :(
 

distilld

Joined
Oct 22, 2021
Water cooling be like

tenor.gif
 

EarthDog

Gulper Nozzle Co-Owner
 
 
 
Joined
Dec 15, 2008
Location
Buckeyes!
I've honestly not run into this many issues before in any loops I've run, or built for people over the last decade and change. I'm always nervous about taking loops apart, blocks especially. I swear they never go back the same.
 

bassnut

Premium Member
 
Joined
Aug 16, 2009
I've honestly not run into this many issues before in any loops I've run, or built for people over the last decade and change. I'm always nervous about taking loops apart, blocks especially. I swear they never go back the same.
Me neither nor can I remember any other thread with since I've been here.
 

distilld

Joined
Oct 22, 2021
Alright, I took it all apart again to clean off the o-ring in the GPU block and reseat it. I also simplified the loop quite a bit and added more fans since there clearly weren't enough in there already. That was about a week ago. Things seem good now - air bubbles are almost completely gone, flow rate is back up to 82 gallons per hour, and fan curves have been set to try to keep the water's delta under 5C. So far it's been successful at that. I ran 3dmark's stress test last night and the water stayed around 2-3C above room temp, keeping the GPU temp just over 30C.

Here's how it looks now.

Spoiler:

Post magically merged:

I swear they never go back the same.

Can confirm, this is 100% true.