Whole Room Water Cooling Part 4 - Nearing Completion...
Linus Tech Tips
·Linus Tech Tips
·2015-05-07
·
1,329 words · ~6 min read
0:00
be proud about it, man. What is this?
0:03
Wait, what is this? Okay, wait. This is
0:06
evil. This is This is Donald
0:09
from Ireland from here in Canada. Good old USA.
0:14
This is the whole room water cooling project.
0:20
This is crazy.
0:24
Day four. Day four. This was not supposed to take 4 days.
0:30
So, what's the plan? Okay, the plan is
0:34
to get these fittings to stop leaking
0:37
first. So, I have come armed with the
0:40
Uber wrench, which is kind of like the Uber mench, but wrench
0:46
here. Um, this one is almost big enough
0:52
for this, but not quite. But that's okay
0:55
because I have an even uber wrench. We'll have a plumbers's crack.
1:07
Okay, these are not going to leak this
1:10
time.
1:17
Gordon Freeman should have carried one of these. It's a lot heavier than a
1:21
crowbar.
1:24
Vice grip. Yeah.
1:28
the compression fan doing it the right way. So, we're doing a bit of a longer
1:34
tubing run this time so that we can make sure that the pump
1:39
is nice and below the reservoir.
1:47
Isn't there like a better way to see if it
1:50
uh yeah, we don't have one more than you think of
1:54
it. There we go.
1:58
In theory,
2:01
in theory, it should work. Let's fill this bad boy.
2:08
Okay, head stop.
2:18
That's not leaking, which is a good sign.
2:22
So, at this point, we're testing the loop to find out if this whole thing is
2:26
actually even working. And even though it would have been wiser to do this
2:30
before we hooked everything up to the whole room, I'm actually testing to make
2:34
sure that each of the individual systems is not leaking. So, checking fittings,
2:38
checking blocks, everything looks to be
2:41
okay. But if a leak were to be sprung at this point, it would be pretty crummy.
2:47
With the systems powered on, it looked like everything was working fine. CPU
2:52
temperatures looked normal. I could see water within the blocks, but then
2:56
temperatures started to creep up on some of the PCs and not others. And then they
3:01
would go up up up up and then they'd settle right back down to what looks
3:05
like a normal water cooled idle temperature. So, not having a whole lot
3:10
of time to figure out what the problem was, I had to do a little bit of running
3:14
around diagnosing what exactly is going on with these CPU temps.
3:20
So, the issue ended up being that some of the systems were getting some water
3:24
flow and others weren't due to air bubbles being stuck in some of the pipes
3:29
and not others. And essentially, well, our pump not having not head pressure
3:35
because we were actually getting reasonable flow out of the system once
3:39
we got the whole thing filled, but not a high enough flow rate in order to get
3:44
the water moving fast enough in order to force the bubbles down the downside of
3:50
their tubes through the systems and then back up the other side into our
3:54
reservoir.
4:11
Okay, CPU temps are 28°,
4:18
not rising. Oh, that pump just grabbed.
4:24
Can hear all the air bubbles going.
4:28
Okay.
4:31
What you mean the pump just to scratch? Uh, when it's got air bubbles stuck in
4:34
it, it spins, but not very It doesn't really pump water effectively.
4:42
Okay.
4:51
6°.
4:55
So, you're 12. Uh, the whole thing will quiet down once
4:59
the air bleeds out of the room. Here we go.
5:10
That seems to be
5:16
Okay, let's get some boot drives, shall we? All right. You know where they are.
5:20
Uh, probably in our old systems. I took them. You took them out. You took
5:26
them out and gave them to Luke. That is what you did. Are they down?
5:31
Okay. Okay.
5:35
All right. Okay. Let's do this. This is minus
5:41
because it says L.
5:50
No. What is problem?
5:56
Because the first time we bled the loop
6:01
on every one of these Mhm. we connected
6:04
them. So, right now, it's going to be
6:08
really hard to bleed the system because
6:12
these ones that we're not using
6:16
have no water flow.
6:19
disconnect, right? Yeah, we may have to drain this bad boy. God damn it.
6:24
Okay.
6:27
Okay, it's not so bad. We just have to do
6:32
what? One, two, three. Right.
6:37
Okay. Okay. Did I get Paragon working or we
6:42
get on that? Um,
6:45
might as well fire up Paragon if you want. Might as well get your system
6:49
clone so you're ready to go when the time comes.
6:54
Okay, that one.
7:11
I wish we'd planned for a more graceful way to do that.
7:17
Can I have that rag back at her? Yep.
7:20
So, what we have to do is
7:24
What do we have to do? Just put the opposite ends on them.
7:29
Yeah. Compression fittings from Yeah.
7:32
Okay. Yeah, that makes sense. Compression fittings onto here connected
7:36
to the pipe. So, what do you do now? Do you put the
7:40
compression fitting on before you attach it to the wall? Yeah. I'll try that and
7:45
use that.
7:48
Which one? This thing. Oh,
7:53
what I do to get this thing close to that, just yank
7:57
like on the tubing. Yeah, I guess that could work.
8:07
I want this pump as low as it can be.
8:11
And then I want this red tilted a little
8:16
bit. Okay, here we go.
8:21
Are you ready? All right. Hit it
8:26
for the third time.
8:33
Okay. pump them off. Um,
8:37
come up. Well, they're not enough to
8:42
Something needs to push the air
8:47
through this. And that's supposed to be that one. But
8:52
that one needs a lot more head pressure in order to make that happen.
9:03
So you can see that this is getting
9:06
even without all the other systems connected, it's getting like no flow at
9:11
all. What's different now? We're not really
9:15
getting any Oh, it's going back up. Yeah, we're not really getting any appreciable
9:19
flow in here, though. That
9:25
is kind of bizarre.
9:30
There be a blockage. No
9:39
temps now.
9:42
39. Yeah.
9:47
58 37
9:52
56
9:58
55
10:07
Oh, it's flowing now.
10:21
So, wait, the big tube's empty now. Yeah. That's terrific.
10:28
hurry. That takes me a long time. Let's have to
10:32
see. Uh yeah, but we got a prop land show, too. You don't want the uh the uh
10:39
the show where you're the first first time you're on to be terrible, right? Uh
10:44
I can come up with the opinions on the spot.
10:48
Um what?
10:52
I don't know about that deal.
10:56
That's what I do anyways. Make
11:08
sure to talk about that in there.
11:17
All that. Yeah, that's good.
11:28
Making the bubbles my [ __ ]
11:40
Running another one your way.
11:47
Oops. got too enthusiastic about my bubble
11:52
slaying. And with the bubbles out, the
11:55
pumps can help us. But with the bubbles in there, they can't do anything.
12:06
All righty then. Zero cares given.
12:19
100%.
12:23
It's running. Yeah.
12:28
No way, dog. It's running. All right.
12:34
The Cooler Master Neptune 240M features an exclusive pump design and their new
12:38
Sencio fans to provide impressive near silent performance. Click now to learn
12:43
more. again. Wow. Is there even room for that
12:47
power supply?
12:50
Should be just barely.
12:55
Fortunately, these cables are intercompatible,
12:59
so we actually only have to swap out the power supply. So, that's it for the
13:05
summer portion of this project. We actually did not finish it at all. We
13:10
got distracted with a bunch of other things. I mean, it's finished in that it
13:15
works some parts of it, but we're going to have at least one more update here
13:19
with the finishing touches as well as the final results. Stay tuned for that.