Building a Wall-Mount Metal Frame for Desktop Watercooling Components

[Windows7ge]

Thursday was a killer at work and Friday when I had planned to do some work on this my best fried decided to kidnap me so nothing got done.

 

Today however I do believe we have found a solution to the "holes too close to the corner" dilemma. 

 

 

Btw I got a new camera.  Still gotta figure out this close up macro feature but now you guys should actually be able to see what it is I'm talking about.

 

The solution I came up although not the best but still better than all the other work around I came up with was this:

 

 

That is a flexible extension with a 1/4th inch bit attached to a drill bit. This was honestly cheap to acquire (like $13 for everything) and got those holes done where the drill chuck was too wide to fit in.

 

Now I just need to finish drilling the holes on the other side and we'll start forming the openings for the 4-pin fan connectors to slot in.

[Windows7ge]

That's one done. Filed and holes tapped. Three to go. 

 

 

Oh this looks so cool.

 

[Jtalk4456]

1 hour ago, Windows7ge said:

my best fried decided to kidnap me so nothing got done.

Today however I do believe we have found a solution to the "holes too close to the corner" dilemma. 

I'd be less worried about the holes and more worried about getting kidnapped, you should report that to the police right away! XD

[Windows7ge]

3 hours ago, Jtalk4456 said:

I'd be less worried about the holes and more worried about getting kidnapped, you should report that to the police right away! XD

Is it still kidnapping if it's voluntary?

 

Anyhow THE HOLES AND SLOTS ARE COMPLETE!

 

 

Same on the other side. Not really worth including a picture about.

 

From here we need to include some holes so we can run wires out of the fame from the bottom.

 

i

 

There are still more holes that need drilling/tapping but in order to know where they go we need to move over to the 3D printer again. This wall mount rack is going to include a basic fan controller that will get screwed to the aluminum frame. We need to design a box to hold the components.

 

What I'm thinking of doing is including a series of basic switches to turn fans on and off and a 1K Ohm potentiometer. I would opt for 500 Ohms as I've never seen a fan come with a resistor that high but for the moment it's the lowest value potentiometer I've got.

 

I'm thinking of using this layout for the switches.

 

 

I also have a pair of G1/4" thread temp sensors from back in the day for monitoring fluid temps. I might just see if it's worth incorporating them here. Give some old tech a new lease on life.

 

I'm going to get designing a box but what I come up with will have to be tomorrows update.

[Windows7ge]

Small update. I think we're going to build this in the longitude direction and if possible make the package as short as possible.

 

 

• I've included the two temp sensor modules and a switch I'll use to turn them on and off if I don't want to look at their display
• The four switches to the right will control fans that I plan to gang up in pairs of 2.
• The potentionmeter will limit the current so I can control the fan speed.
• The switch below the potentiometer will short out the potentiometer. Since the potentiometers minimum value is never 0 Ohms this switch will provide unrestricted current to the fans at 12V for whatever I need maximum power for.

 

All these fans do have a PWM pin and that means I could gang them all together and have the motherboard BIOS control them linearly but for now a manual variable resistor will suit my use case. I will keep the PWM pin method in the back of my head though. Never know what I might want to do with this rig in the future.

 

The reason for building this out sideways is when pressing buttons it's going to want to flex and bend. By making it shorter it will help it to not flex as much when pressing buttons. I may also investigate designing a 90° bracket for support on the left as I'd like to mount this against the right side support.

[Windows7ge]

So it took nearly 5 hours but I have a prototype for the faceplate. I decided to modify the original design to favor the lower profile I desire.

 

 

 

Some of the tolerances are going to be pretty tight but I'll get this going on the printer and I should be able to fit everything tomorrow. If the print doesn't fail I should see the results late tonight.

 

From here I can start construction on the body.

[Windows7ge]

Tonight's progress consists of (or consisted of) redesigning the control boxes faceplate.

 

I was cocky thinking I could get it right on the first try.

 

 

The changes aren't massive but their effect is big. Two issues I ran into:

• The hole for the potentiometer wasn't large enough. I remeasured the potentiometer threads and increased the hole radius by 0.2mm
• The temp sensor holes were too loose. The goal is (or was) a friction fit but it didn't happen. These displays do have a mechanical means of securing themselves via tensioned clips. The purpose of the new slots being to hopefully take advantage of them.

This is going to take overnight to print. I will assess whether or not it worked tomorrow morning.

[Windows7ge]

The CAD for the prototype body of the controller is finished!

 

 

This will be secured to the frame using the 8 small holes while the 2 larger ones will allow wires to pass between the components in this box and the fans through the frame.

 

Zero visible wires.

 

Get this on to the printer and we'll start cleaning up the faceplate. This time our tolerances came out quite a bit better. I expect this one will be usable after I clean it up.

[Windows7ge]

This is all that's getting done tonight. I was able to fit almost all of the parts into the faceplate tonight. It looks pretty cool IMO.

 

 

Of course it wouldn't be a good hobby project if we didn't draw some blood. I'm not going to show it but I managed to drive my thumb right into an exacto-knife. Ouch.

 

We're a little over 1 hour into printing the body for this faceplate and so far it's staying stuck down to the bed so we should be on schedule to test the fit tomorrow night.

[Windows7ge]

Alright, tonight we:

1. Cleaned up the new body,
2. Reamed the 10 holes to size that secure the face-plate to the body using a 7/64th drill bit
3. Pre-tapped the 10 holes
4. Cleaned up the two 1/4th" holes
5. De-burred everything
6. Finished populating the control panel with switches.

 

Tomorrow is my late night at work so it's not likely I'll get anything done but next on the list is marking out the 8 holes that will secure this box to the aluminum frame then drilling and tapping the aluminum.

[Windows7ge]

In the past two days I:

1. Drilled, tapped, and de-burred eight 3/32" holes.
2. Drilled and deburred two 1/4" holes.

With these done all I need to do is buy 5mm M3 screws and I can attach the box to the frame.

 

 

My weekend is going to be busy so I can't say how much I'll be getting done but next in the queue is we finally get to start playing with and running electrical. This includes working with a soldering iron and heat-shrink for the wires.

[Windows7ge]

So this is where we are in the project today:

 

 

We need to:

1. 3D print four brackets in order to mount eight 4-pin fan connectors to the frame.
2. Remove all heat-shrink and sleeving from the sacrificial 4-pin Y splitters including PWM & TACH wires.
3. Mount the connectors into the 3D printed brackets.
4. Removing the male connector from the oppose end of the Y splitters and solder on our AWG26 gauge wire
5. We also need to fish wire through the frame and start mounting the various components.

How much of this will get done today I cannot say.

[Windows7ge]

This is what I got done today:

 

Finished printing the four dual holster 4-pin brackets. I additionally cleaned them up and reemed the screw holes to the proper size.

 

 

Disassembled and prepped the splitters:

 

 

I'm debating weather to just cut the male pins off or if i should modify them to suit soldering to bare wire. I mean, if they're already mechanically conjoined it kind of takes half the work out of it for me.

 

I additionally mounted the 4-pin female connectors in the dual holster brackets and verified that they all fit in the frame.

 

 

Tomorrow should be less busy than today so I hope to get more done. I realized though that there is an additional hole I need to drill for the thermal probes so I'll be doing that tomorrow before any wires actually get ran.

[Windows7ge]

Today we took a large step into completeing the eletrical.

 

First and foremost I got that hole drilled and de-burred for the temp sensors.

 

Starting the wiring I finally got to play with this silicone coated wire which I kind of bought by mistake. However this stuff is absolutely divine to work with oh my god.

So typical computer wires are coated in (I believe) some kind of PVC. This commonly has a heat resistance rating of 80°C. Meanwhile this silicone wire has a heat resistance of 200°C. The silicone did not melt what-so-ever while soldering which was fantastic.

 

The method I came up with for attaching the silicone wire to the fan cables is a combination of mechanical and solder. I cut off part of the pin normally found inside a male fan connector then utilized the locking clip as a hook for the silicone wire.

 

 

Then I soldered this together and added heat-shrink.

 

 

Because I wanted to common all the grounds together I removed the jacket from a small section in the middle of the wire for tthe plugs lower on the frame. (this wire leads to the plugs in the background)

 

 

Then I soldered over this. Made a hook and latched it to the dual 4-pins as I did with the other:

 

 

Soldered over this, added heat-shrink and we were in business. Then it was just rinse, repeat on the other side. The 12V wires had to be ran a bit differently since I want to be able to control them in zones so each bracket gets a dedicated 12V wire.

 

 

From here I was able to screw the brackets to the frame on both sides:

 

 

After this I started working on the digital displays and figuring out how I wanted to wire up the panel but I decided I had put enough work into the project for one day.

 

 

Ran a temp probe and started figuring out how to power the 5V displays alongside all the 12V equipment.

 

Tomarrow's availability is a little up in the air. We'll see what happens. Next on the agenda is wiring up the second temp sensor and starting the back side of the control panel.

[Windows7ge]

Got the second thermal probe ran. Now we can start tonight by wiring up the the temp displays. I started by removing the original wires (white/blue) and with the silicone wire soldering the modules together.

 

 

The reason for this will make sense in a bit.

 

Now things get a little complicated because we need to run some wires through the panel before we solder them.

 

 

Got all of the thermal probe wires soldered in.

 

 

Now we just clip them back together, press them in place, and prep the power wires.

 

 

I'm probably going to regret soldering the wires together this way if either display fails but a little cutting, soldering, and heat-shrink will make a display replacement relatively  pain-free.

 

 

To my delight (and surprise) I successfully didn't break either of them in this operation. This camera doesn't do these displays any justice at all they're very legible and green in person but the camera washes them out hard.

 

I was clenching the probes and the display values increased as they should.

 

 

You can see them a little bit better here:

 

 

I was just running them from a 18650 powered 5V battery bank and some alligator clips.

 

Tomorrow's agenda will be wiring up all the switches, running 5V power and possibly (if I have enough wire) running 12V power.

[Windows7ge]

Tonight starts off on a bad note. I realized I had forgotten to drill YET ANOTHER hole and I've already ran a bunch of electrical inside the tubing...

 

I decided to just go for gold and very very ginger carefuling nip a hole deeper and deeper until I just barely punched through, then widened it up.

 

 

And thank goodness I did because right below the hole was a couple of wires. If I had used my usual gung ho method of "Just push it til she goes through!" I would have chopped right through them and that would have been a real set-back.

 

Moving forward tonight I pulled a 5V wire through the tubing and I finished wiring the back of the temp displays:

 

 

So now when the switch is ON...

 

 

The displays are ON.

 

And when the switch is OFF...

 

 

The displays are OFF.

 

Now as for wiring the switches I used a multi-meter in continuity mode to find out which wire goes to which set of outlets (because why would I want to label the wires? )

 

After determining where each wire goes I soldered each according.

 

 

Then added some heatshrink:

 

 

Now the other side of these switches are a common +12V supply. Just for creativity sake I decided to connect them all together as such:

 

 

I swear this high-temp silicone wire is just an absolute treat. Try doing this with your run of the mill jacketed wire and see how big a mess it makes.

 

Finishing off the night I attached one leg of the potentiometer, pontentiometer bypass switch, and the imput of the four switches:

 

 

Unfortunately this setup and the work I've done has used up nearly all of the 20ft of wire I had and I need just a little bit more to finish it out. Good news is I have more wire on order. A lot more. This does mean through that tomorrow we will be shifting gears so we can keep moving forward while we wait on equipment.

 

I don't have a picture of it at this time but we will be disassembling, and cleaning my old watercooling apparatus. Stay tuned for that.

[Windows7ge]

I was going to start the watercooling but guess what showed up in the mail?

 

 

So i guess today we finish off the electrical.

 

For the sake of proper color/voltage I decided to use yellow for the +12V source.

 

 

Ran the line back to a molex connector and followed it up by throwing some sleeving and a little heat-shrink on the end.

 

 

 

I did a continuity test between all four lines and the frame. Each circuit is electrically isolated so the frame won't become live at 5 or 12V or worse short a line to ground.

 

I stuffed as much of the wire back into the frame as I could so I could screw the face plate on:

 

 

Although my use of heat-shrink on the connections was a nice idea I had forgotten my depth limitation and because of this the switches made it a very tight fit.

 

But man am I loving the result.

 

 

Unfortunately  the control box is only being held in by 2 screws. The 5mm ones I ordered are on backorder and won't be in for some time. It's secure enough though for us to move forward. I'll replace the two and install the eight when they do arrive.

 

Tomorrows my late night at work again. Likely won't get anything done but NOW finally next on the agenda is dismantling the old water cooling apparatus.

[Windows7ge]

@Skiiwee29OK, so NOW we finally get into the actual watercooling. I'm just poking you because I felt like teasing you.

 

This was the old wall mount cooling rig I was using to cool my system.

 

 

The back for those curious:

 

 

When I built this I was actually quite proud of my ingenuity and attention to detail.

 

Got most of it dismantled. You can see for reference why metal is going to be so much better. You can be so much more compact with it and retain the strength.

 

 

Got everything dismantled down to the fittings. God there's so much dust.

 

 

Yeah I have no excuse. This is just me being negligent.

 

 

Have to say through due to the shear amount of surface area this much dust didn't translate into any appreciable performance degradation. My 1950X and GPU's still stayed well below their maximum.

 

Unfortunately one of my GPU's has gone bunk (this was a known issue. It's not new). So it won't be going back in the system. They're both the Sapphire Tri-X R9 290X. Very old by today's standards. I just hope my still good 290X can last me until the RX 6800XT's become more affordable.

 

After an absolute ton of dusting and rinsing down the radiators with a hose I think I could get them to pass for brand new if I threw them in their factory packaging.

 

 

Few things on the agenda next. I've got a shopping list for PerformancePCs.com. I need a whole bunch of little stuff. I need to replace all of the old O-rings, I need to flush the radiators, clean the tube reservoir, and see if I can remove some of the old corrosion from the various fittings.

 

But that'll be for tomorrow. Good night!

[Windows7ge]

Today I started by flushing out the radiators with distilled water and uh...

 

 

She don't look like water coming out the other side. Looks more like milk. 

 

Got those flushed out good. then I inspected the waterblocks for clogs starting with the XSPC Raystorm Neo:

 

 

Thing's got some serious gravity to it.

 

To my surprise yes there was some gunk but really nothing of genuine concern. Explains why temps were still fine.

 

 

Got to hand it to XSPC they know how to make a complicated as hell GPU waterblock. This could have been designed so much simpler.

 

 

I hope I didn't forgot an O-ring or anything putting these back together. Went ahead and cleaned out the ress, pump, pump head, and misc components.

 

 

Gotta say, fittings are a PITA. These took many hours to clean and you probably wouldn't believe it but all of the O-rings here came from these fittings.

 

 

Next on the agenda:

1. Assemble and fit the radiators to the frame.
2. Take measurements for the mounting holes to be drilled and tapped for the ress.
3. Take measurements for the mounting holes to be drilled for the pump.

[Windows7ge]

The parts have been ordered!...but not without issue...

 

 

They are out of stock of the coolant concentrate I desire. Mayhems X1 Clear Concentrate to be exact. It seems clear coolants are very popular. I put in a request to be notified when it's restocked but in the meantime I can fill the loop with just distilled water. The X1 just acts as a good pH balancer/corrosion inhibitor.

 

We'll see what happens as parts come in and as we progress in the project. Maybe someone here on the forum knows of another place I could source it.

[Windows7ge]

Tonight starts out by sorting screws. How exciting!

 

 

Seriously though these radiators come with 30mm & 35mm screws and it's important to use the 30mm ones. As far as matching screws goes it looks like I'm shy four of the small head black ones (need 32 in total) but I guess I'll throw the copper ones in anyhow. I don't know where all my other screws have gone.

 

I'm using these 1.5mm washers to simulate the chassis you'd normally mount this to. 30mm is a bit much for a 25mm fan as far as how deep the screw goes into the radiator. You really don't want to puncture one of the tubes.

 

 

That's one radiator fully assembled. Went ahead and screwed some G1/4" black sparkle plugs in the ports we're not using and threw the compression fittings on where they need to go.

 

 

And here's the two complete.

 

 

Now comes cleaning up and installing the spacers.

 

 

The night was getting late so we only got one radiator installed but I like what I'm seeing.

 

 

I'm going to have to tinker with the spacer sizes because the bottom mounted rad is crooked. Left to right it's fine, but front to back it's tilted backwards due to my crooked brazing.

 

Tomorrow's agenda. Correct the spacers. Might have to print new ones. and mount the second radiator. If I have the time left over I'll also be measuring and drilling the holes for the pump.

[Windows7ge]

Onto...the next roadblock...

 

So the plan was to use some long stand-offs (35mm) to mount the radiator on the tubing side of the apparatus but...

 

 

The stand-offs would interfere with the compression fittings.  I thought to off-set it but due to the 71mm mounting hole gap I can't move it forwards or back without it interfering with something else...

 

Enter what all of today was spent prototyping!

 

 

 

The logic here is three 35mm stand-offs will be screwed to the frame. The two holes immediately off the center hole will secure the bracket to the frame. The center hole and two outer holes will secure the ress bracket itself to this bracket & the frame directly.

 

The tolerances on this are going to be tight. The 35mm stand-offs are only threaded 8mm deep and because of the thickness of everything that means if I'm off by any more than 1~2mm somethings not going to fit or the screws won't reach the standoffs.

 

It should be done printing by tomorrow. If it works I'll print the 2nd identical to this one.

[Windows7ge]

Only one picture for tonight. Things ran very very late but the first try prototype worked.  There was only one little error where I didn't enlarge one of the screw holes enough.

 

 

I kind of like the hidden screws design. Makes it kind of look like it's just floating there with no real support but she's bolted down solid with three brass rods inside each of those white brackets.

 

And to top things off it's LEVEL. Which I had very low hopes of actually accomplishing but somehow it worked so when it's mounted on the wall it'll look strait even though both of the aluminum supports are crooked.

 

Both tomorrow and Friday are my late days at work currently. Maybe something will get done, maybe not. I will keep everyone posted though as we continue to progress.

 

Next on the agenda is drilling the holes for the pump to mount to. I have a couple of machine screws/bolts in my stash of computer screws. I'm tempted to just use them but I'll have to test if they'd even fit. I'll also have to 3D print a special wrench/tool in order to install them.

[Windows7ge]

Guess what showed up in the mail!

 

 

I opted for plain clear tubing before for one it let's you see where the air bubbles are and two anything colored transport wouldn't match the scheme I'm going for.

 

We're picking up where we left off by assembling the remaining components for the pump along with doing all PR O-ring replacements.

 

 

I also went ahead and added sleeving on the pump cable and some heat-shrink where necessary.

 

Went ahead and assembled the ress caps. I'll use the ress itself as a means of draining the loop. Also have my little silver kill coil.

 

 

Got those installed and lined up how I want them.

 

 

 

Next up got the radiators installed.

 

 

Nothing is perfectly square but that's the limit of my current skill level and the tools I have access to. Still looks really cool IMO.

 

The fan cables flailing all over the place started to piss me off so I went ahead and took care of them next.

 

 

Starting the pluming required the changing of a few fittings. This is just sort of what happens when you water-cool. You have an idea in your head but it doesn't always translate to something that physically works. Was forced to install a couple 45's and move some things around.

 

 

 

I ran the radiators in parallel for the cleaner aesthetic. There's also the theoretical potential that if the fluid spends more time in the rads because it's moving slower that I'll see higher cooling efficiency. ¯\_(ツ)_/¯

 

Now to address an issue I didn't foresee. So I had assumed the potentiometer I picked out would be sufficient to limit the current to 8 fans. It's only 4 Watts (0.05A @ 12V). Well, as it turns out 4W is a lot for a potentiometer and when I went to perform the electrical test (that I should have done before putting everything together) the potentiometer died almost immediately. Luckily I have the bypass switch but that drives them at 12V unrestricted. Which is a little loud.

 

I have another driver which would do the job quite well but the problem is it's about the size of your average computer mouse and I have no idea where I'd mount something of that size...then I have to run it's potentiometer somewhere...I appear to have caused an electrical mess. The good news is absolutely everything else electrical works as intended and nothing is shorted to ground. So this is our only issue.

 

I'm thinking of either hanging it from the frame as it hangs on the wall or possibly rigging it to the exposed side of a 120mm fan slot then somehow running the wires...

 

Decisions, decisions. Going to have to sleep on it. I really don't want to have to dismantle and redesign the control box from scratch.

[Windows7ge]

Today's update is coming very early. I'm tired today but we are very close to finishing the build.

 

Got the quick disconnects connected to the rack.

 

 

Got the length of the CPU to GPU tubes figured out. We're running them in parralel.

 

 

It's not entirely certain where on the wall to the computer this is going to be mounted so I just wen't ahead and used up the last of my tubing.

 

 

I have a plan in mind on how I'm going to incorporate the new fan controller into this system.

 

 

I'll be using a 3D printed box & bracket, some of the remaining 35mm stand-offs I have, screws, and sleeving to mount this to the bottom of the bottom radiator. From there I can run cables into the top of the control box and replace the defective potentiometer.

 

Due to fatigue though I think this is the stopping point for today. We will investigate the box/bracket tomorrow. Aside from this little repair. We're ready to fill her up and begin leak testing. I look forward to it.