Sprocket....dripping wet (again)

[clocker]

Okay....the parts finally came.
There is no way there will be room for two radiators, so the point has become moot.
In fact, I worry about fitting in the one rad and the Eheim pump, both of which are larger than I had imagined from the pics.

Tomorrow is my day off and I'll be able to get started mounting these components and taking some pics.
Everything else is pretty much done and ready for the waterloop...the SATA cables from Ultra even came yesterday.

The DangerDen waterblock seems fairly straightforward to install although I'll probably machine the mounting studs myself...the supplied generic threaded rod is kinda cheesy looking.
I have no idea about the vibration/noise of the Eheim pump.
I can envision several isolation mounting methods, but as yet don't know how elaborate I need to be.

The radiator is going to be the biggest PITA...it has no mounting flanges ( not too difficult to fabricate) and it's sheer size really cuts down on the possible location options.

This might take a while to figure out....

[Duffman]

dont forget pics!

[clocker]

Ask and ye shall receive....
Here are the parts I'm going to try and cram into the case...

You see my dilemma.
The radiator/shroud/fan assembly is grotesquely large compared to the available space.
The pump ain't no midget, either.

My initial impression was that I might have to go with an external box to house the rad/pump, but I hate the idea and refuse to do it.
This must be completely self-contained or I won't continue.
So....

If I cut the mobo tray right where the PCB ends, the front of the case will accomodate the radiator assembly...it's close, but it works.
The pump will squeeze below the rad and the end of the video card.
Barely.
Then I just have to route the ( somewhat awkward) 1/2" tubing in as nice a manner as possible.

Before someone asks....
Yes, I have considered mounting the radiator up in the roof, but there are practical issues that make it difficult.
The side panels install first and the roof panel holds the whole skin in place, which means the roof is removed every time access to the interior is wanted.
I suppose that it is possible to get around this, but it would be a giant PITA to do well, so ( for now at least) I've dismissed the concept.
This may change, naturally.

I'm almost overwhelmed by the scope of this project...I hate the idea of hacking up the case only to run into a problem I hadn't considered, but hack away I must.
The two main construction projects are:
-properly (read: securely) mounting the radiator...there are no flanges or other attachment points as it sits now

-installing the pump...the wiring is easy, albeit time consuming, but I don't know how elaborate to make the vibration isolation mounts.

Oh boy, hours of fun.

[Duffman]

First of, wow that thing is fucking huge, but it looks like it will fit.

Heres what I'd do, bump the radiator up against the far right of the case, and put a 90degree tubing thing right off the end of the pump, and into the rad, it looks like it might just fit.

[lynx]

Have you considered mounting the radiator up in the roof.

Well, you didn't actually say not to ask.

Actually, if you've got a reservoir I'm not sure the roof would be a good place to put it. The reservoir obviously has to be at the top of the system, and would therefore have to be outside the case if the radiator is up there.

The size (and price) of the Eheim pump is exactly the reason I went for the Hydor model. It is about half the size, virtually vibrationless, and mounted on rubber suction feet. The downside is that it isn't self-priming so it is an absolute swine to get the system filled. Still, too late to worry about that now.

I came to the conclusion that the best layout was something like this:


The reservoir layout gives the best option for removal of air from the system, and the direct feed to the pump makes filling much easier. Hope that's of some help to you.

[lynx]

First of, wow that thing is fucking huge, but it looks like it will fit.

Heres what I'd do, bump the radiator up against the far right of the case, and put a 90degree tubing thing right off the end of the pump, and into the rad, it looks like it might just fit.

Unfortunately the outlet from the pump is at the top. That is a necessity with centrifugal pumps because they do not have the ability to suck.

[Storm]

that was kinda my question to........ cause i didnt hear you mention a resevoir......... are you putting one in? if not, why and is there a downside to not putting a resevoir in?

[clocker]

Thanks for the input guys.
I would have loved to put in a reservoir, but there is simply no room.
A t-line shall have to suffice.
Actually, a t-line IS sufficing...Sprocket is running on water as we speak (well, as I speak and you read).

Once I decided on a general layout the work went surprisingly fast.
As I suspected, the fit was very close, but everything slid into place as I'd hoped and seems to work well.

I leak tested the loop as I was cutting up the case (total test time was 3 hours...not the recommended 24, but I was impatient) and had no probs with water.
To begin with I am running pure distilled water...haven't decided on what (if any) additives I might use.

To early to guess the temps yet...need to let the loop get up to temp and cycle the AS5 a few times, but initial results are promising.

I also need to improve the isolation of the pump and wire it up correctly, but that can wait for a bit till I'm sure I want to keep this system in the box.

Pics soon.

[accat13]

/me patiently awaiting pics

[clocker]

Good thing you are patient...it might be a while.
Just for grins I decided to see what my Priming temps might be.
Going at it hammer and tongs right now.

100% CPU usage and 34c.
Not too shabby.