Building a Wall-Mount Metal Frame for Desktop Watercooling Components

[Windows7ge]

Alright, the goal of tonight was to get familiarized with brazing and what it taught me with the couple of hours I played with it was I need more practice and that I can expect to be doing a lot of filing.

 

Getting started this is the setup:

 

 

I checked that stone and it's flat enough for our purposes. As you can see I already brazed two 45° faces together and that went fairly well.

 

Next I wanted to see how well I could braze two flats together. This was the setup:

 

 

This was a little tougher given the clamp but it also came out pretty good.

 

 

Unfortunately brazing a 45° face to a flat proved to be the most difficult. While trying my best the torch decided to give out on me so I had to quit half way through:

 

 

To my delight though all of the joints even the unfinished one came out rock solid. I tried to break apart all three by hand and despite my best efforts none of them are budging.

 

Tomorrow I think we'll start the real deal.

[Windows7ge]

It's unfortunate to say but today is our first real set-back. What appeared to be working well was indeed not.

 

On the random test piece of brazed Aluminum 2 out of the 3 welds failed. Although they didn't break yesterday they broke today. Interesting...

 

After consulting with a neighbor, a co-worker, and factoring in my of consensus it was unanimous that the issue is not enough heat. It seems trag1c and Letgomyleghoe were both right in their suggestion that I should have gone MAPP gas.

 

Analyzing the braze it appears it did not take in various places on the failed welds. Additionally areas where we reached the melting point of the aluminum which discolored the aluminum in a way resembling wet cardboard were not apparent on the inside of the tubes. This means the surface was getting hot enough but the inside was not.

 

This weekend likely Saturday I'll be returning my Propane and buying a cylinder of MAPP but until then the project will have to be put on hold.

[IkeaGnome]

1 minute ago, Windows7ge said:

It's unfortunate to say but today is our first real set-back.

Got plenty far before you hit one. That's a pretty good run, I'd think. 

It's been fun watching the progress so far, and I'm excited to see what happens on MAPP.

[Windows7ge]

We have the MAPP.

[Windows7ge]

And I'm calling it tonight. The night is young but I'm burned out from a long work week and restless nights with no sleep.

 

Things are going quite a bit more smoothly using MAPP. Got to say I much prefer the smell of Propane though. This stuff is kind of gross.

 

 

Going by the 3D render in the original post you can see where we are headed. I have all four supports braised together (nearly had a heart attack a couple of times because each slotted support needed to be attached in a specific orientation and I though I screwed it up a couple of time...) and 3 sides of the frame but we ran into some issues with both.

 

Even with MAPP for some reason I could never get the braising rod to flow into the rear 45° joint very well. The heat just wasn't making it into those rear joints so they don't look the greatest.

 

 

I did what I could. They will be getting filed down a touch so hopefully they look a little better These braises won't have any sideways load on them as the weight will be strait down so I'm not worried about anything happening even if these happen to crack. Otherwise they came out fine and should look good after some hand filing.

 

 

We ran into a rather significant set back when I found this corner of the frame had a hair-line crack in it:

 

 

I already went through and repaired it but this proved to be difficult as for some unknown reason re-heating it opened up a ~4mm hole. This was much wider than the original gap pre-braise.

 

To fix this as I couldn't build up the braising rod material it was too liquidy while hot. I cut a small piece of Aluminum extrusion and jammed it in the hole to act as a bridge. I then filled the gaps around it and gave it a moderate filing. It still needs more work but it should be strong enough to hold. This will introduce a challenge when it comes time to feed our wires through but it was this or scrap these two aluminum sections.

 

Tomorrow I will likely begin by attaching the 4th side of the rectangular frame followed by attaching the 4 supports.

[Windows7ge]

This marks today's progress. I attached the fourth side of the rectangular frame and I successfully mounted two of the four radiator supports.

 

 

Despite how crooked all the cuts were done on the miter box I've measured the whole apparatus side to side, up & down, corner to corner. The largest discrepancy I found in overall squareness was approx 1mm.

 

In other words she's pretty damn square for being done entirely by hand.

 

I called it a night because I ran into a heating issue. Even with MAPP this one side just did not want to flow properly:

 

 

At this point in the day it was getting progressively colder and I'm thinking the cold air was making enough of an impact where it wasn't staying liquid and the Aluminum wasn't reaching the proper temp. I'll have to either warn up the garage or investigate some fire-proof insulation to help me keep the heat in long enough to braise.

 

I also ran into an issue where the post in the background has a slight backwards lean. There's any number of contributing factors to cause this but the plan is to mount the radiator with long screws and spacers so I can change the height of the spacer to compensate so the radiator is level on the wall.

 

 

I decided to give it the level check and despite what the bubble says it's actually pretty level. The stone it's clammed to measured the same discrepancy so my mounting surface is what is crooked. The bubble doesn't change place between the mounting surface and the supports so it really isn't looking too bad.

 

 

Tomorrow I'll work on warming up the garage and we will attempt to attach the other two supports in addition to the two small pieces that the pump will sit on.

[Windows7ge]

And with that the four supports are installed...but not without issue.

 

 

We're still pretty darn level which is a plus.

 

Unfortunately mistakes were made. Today in order to aid heating due to yesterday's heating issue I added a propane torch (running two torches 1 Propane, 1 MAPP). This got things hot quick...too hot in fact...

 

 

The damage is done. She melted.  Structurally it's still sound but that's a blemish I can't sand out. I'll see what I can do when we start cleaning up my shotty craftsmanship.

 

The good news is this is a great learning opportunity. Propane + MAPP + Aluminum = NO.

 

It does appear ambient temperature has a worthwhile effect on the MAPP's ability to flow the braising material into place. Today was quite a bit warmer and the braising after this mishap went smoother with just the MAPP so it's good to note the room needs to be relatively warm for things to work.

 

Another issue that occured due to overheating is I used a C clamp to hold some of the pieces in place while braising and this caused two of the supports to get squished.

 

 

If you notice the slot on the right is a lot more narrow than the slot on the left. Because of the clamping + over heating that was the weakest spot. It squished the slot and bludged the material outwards. Good news is it didn't have a significant impact on how strait the part was.

 

So another moment where something was learned. Don't use C clamps to hold a work piece while heating it. I can fix this by squeezing it with a C clamp and re-filing the hole larger again. Steps I shouldn't have to do but it's part of the process of knowing what not to do next time.

 

Unfortunately getting these on took much longer than I planned so the pump mount still isn't attached. That'll be tomorrows project.

[Windows7ge]

Today ends on a disappointing note. We've ran into the heating issue again.

 

 

This camera does a terrible job of focusing on nearby objects I know. In this blurry image the braise has taken on the stubby piece in the back but I had a significant amount of trouble getting it to take to the piece closer to us. This is the assembled frame from the last update and it seem as though it's acting as a giant heatsink.

 

To add insult to injury the air was rather cool again and this just further shows that room temp can have a worthwhile impact on braising success.

 

I risked heating the part with both MAPP & Propane again and although no damage was caused this time and it did get the area hot enough for the braise to start taking as soon as I took the torch away it would stop.

 

So I either need to grow a 3rd arm or rig something up again and pray it doesn't result in me melting the frame again.

 

That's going to have to be a project for tomorrow though. Not the best update but it's all I had time to do tonight.

[Windows7ge]

One step forward two steps back...

 

 

This is not the setup I used to try and braise these I just did this in the end to hold them in place.

 

I succeeded in using both MAPP & Propane to get the frame hot enough to start taking the rod but something went wrong and the left support barely took. The right support is rigid but there's cracks and voids so I don't trust it long term. Tomorrow night we will tackle this once again with a different approach and see if we can finish out this aggravating bastard.

[Windows7ge]

Alright. This time I freaking LAID the heat on there (MAPP+Propane) with complete and utter disregard for the fact I might melt the aluminum but this time I think we may have finally succeeded.

 

 

Giant globs of melted braising rod. If it didn't take this time I don't think it ever will.

 

 

It was still hot in this image but we successfully didn't melt the aluminum and I don't see any cracks or voids. I'm feeling confident that it took this time but I'm going to let it cool down naturally before I draw any conclusions.

 

As for a little PSA. Be careful about getting your stone too hot. See the dark spot on the stone in the middle of this image?

 

 

That is a pit or cavity that formed when the stone randomly decided to explode sending a chuck of rock up to the ceiling of the garage before landing elsewhere on the table.

 

Note to self. Don't point the torch at the rock.

 

Tomorrow I will validate that the braise took and from there we will begin filing and cleaning up the finished part.

[Windows7ge]

Damn, came back to this this morning and once again the one closer to the corner of the frame just went *POP* and came off.

 

Investigating the braise joint it's not taking very well. At the same time it looks rather dirty.

 

Tonight after work or tomorrow I'll cut a fresh piece of aluminum. File the edges down and clean up the surface of the frame better than I did before.

 

I'M GONNA GET THIS THING STUCK ON THERE IF IT KILLS ME!  IT'S THE LAST DAMN PIECE!

[Windows7ge]

Friday was a long day so nothing got done but today I had a epiphany and made this little jig to hold the piece strait in the air and allow me to braise all four sides at once. It also eliminated the stone so no more exploding rocks.

 

 

I scrapped the piece that was refusing to take to the frame. Re-cleaned the surface. Cut a fresh piece of aluminum. Hit everything with a steel brush and acetone then tried it again. Even without the stone pulling any heat away the overall frame is now too much of a heatsink so I have to use Propane + MAPP but at least it's quick and if I'm fast about it it doesn't appear to cause any damage to the frame (melt the aluminum outright).

 

It's too hot to touch right now but when it cools down we'll test if the braising took this time.

 

Lemmie tell ya, if it didn't take this time somethings getting thrown out the window.

[Skiiwee29]

<-- Moved to Build Logs -->

[Windows7ge]

2 minutes ago, Skiiwee29 said:

<-- Moved to Build Logs -->

Thought this might happen.

 

Some watercooling is planned at the end of this project when I incorporate it into my existing desktop setup. One of those "It kind of falls into both categories." situations but it doesn't really matter to me which one it's in.

[Windows7ge]

Alright it's stuck on there good this time.

 

From this point we move on to filing/grinding.

 

 

This process as expected was time consuming and tedious but I fixed up the corners, everywhere where there was excess braising material, and smoothed out every sharp edge I came across.

 

It's not going to be perfect and I don't have the best tools for this task. The after image isn't very reveling but there's not much difference to see. just a little clean-up.

 

 

I also took the liberty of fixing my mistake where the C clamp had crushed the slot for the screws on the right side top & bottom. It's still an ugly blemish but they holes line up, and are the correct size again.

 

We will see what happens tomorrow but the next step is to design and cut holes for the mounting of our 4-pin fan power connectors.

[Windows7ge]

Alright, tonight we're taking a break from the braising and the filing (may do more filing at a later time) to look into more of the 3D printing & electrical side of things.

 

The radiators I'm working with are 60mm thick and I'd like if possible to "hide" the fan outlets behind them.

 

 

I'm looking to design some 3D printed brackets that I can hopefully friction fit some 4-pin Y fan splitters into.

 

In this orientation it doesn't look like I should have any trouble designing something that fits the desired profile.

 

 

I have the de-pinning tool so I can remove the wires from the connector. I also plan to solder and properly heat-shrink the cables together so the expectations of having any issues after assembly are nil.

[Windows7ge]

The first prototype CAD model:

 

 

Huh...they look kind of like RJ-45 ports...¯\_(ツ)_/¯

 

The overall bracket is 8mm tall, 60mm wide, and 19mm long. I over sided the holes for the screws by 0.4mm as the goal is a loose fit and I oversized the holes for the 4-pin fan headers by 0.2mm as the goal is a friction fit.

 

We will see if it works as intended or if I need to make changes to the design. This will take about 45min to print so I should find out if it will do what I need tonight.

While we wait on that we need to design some spacers of varying heights due in part that nothing is perfectly square and my 10mm screws are much too long for the radiator.

[Windows7ge]

Last update for tonight but it looks like my first attempt at the 4-pin power brackets may just work. They have a very nice friction fit but are still loose enough to where the fan plug can be removed. I think this will work quite nicely. Just had to clean up the screw holes with a drill bit and I'll file down the holes for the four-pins with a knife so the connector can sit flush. but so far it's looking good. 

 

[GreatnessRD]

This looks like a fun follow. Slowly but surely you're getting it together. 

[Windows7ge]

11 hours ago, GreatnessRD said:

This looks like a fun follow. Slowly but surely you're getting it together. 

It's a process and it's slow going because of my day job and other obligations but we are making progress each day. Few set backs here & there but so far nothing that has resulted in going back to the drawing board (square 1).

[Windows7ge]

Tonight we're going to start by seeing if the screws I'm working with can self-tap into the aluminum.

 

And the answer by pre-tapping a hole using a 5/64th drill bit is...sort of...simply taking the standard M3 screw and slowly working it into the drilled hole it does form threads. Which is good. Unfortunately the aluminum sticks inside the threads on the screw and when I go to back it out ruins said threads.

 

Since I don't have a tap & die kit we're going to try creating a self-tapping screw that cuts the material away rather than pushing it by cutting a wedge out of the end of one of the M3 screws and sharpening the cutting edge.

 

A drop of WD-40 might do it some good too.

[Windows7ge]

That, worked a treat.

 

Using a small triangular file I carved a cutout in the end of one of the 10mm M3 screws (similar to what you see on the screws that come with some case fans). Following that I drilled a fresh 5/64th hole. Gave it a drop of WD-40 and proceed to tap it.

 

Not only did none of the aluminum stick to the screw but the whole process went smoother. The aluminum being 1/16th of an inch thick holds about 3 threads. With three screws per bracket that will be plenty of mounting pressure to keep things in place.

 

With that resolved we can move onto designing a tool that will show us exactly where the holes need to be drilled.

[Windows7ge]

This marks the end of tonight's progress. We have a finished prototype CAD model.

 

 

If you reference the previous prototype and look at the metal frame you can see where I'm going with this.

 

The raised portion will create a 6.4mm gap between the guide and the top of the radiator supports. This will help hide the power connectors behind the radiators. The larger holes on the guide will allow me to enter with a metal file and manually carve the rectangular holes needed for the 4-pin fan connector to partially sit inside the frame.

 

Tomorrow we'll get this going on the 3D printer and go from there.

[Windows7ge]

I think you can see where I'm going with this.

 

[Windows7ge]

Taking this picture made me realize that we have a problem. I can't drill the 3 holes closest to the left piece of metal because of the width of the drill chuck. A combination of poor planning and design has led to this dilemma.

 

Now there's a few ways I can approach this issue.

1. I could un-braise the support giving me clearance to drill those three holes.
   1. Really don't want to do that though.
2. I could design a different bracket that holds two 4-pins on-top of one another. This would enable me to use just the first 4 holes.
   1. But would require a lot of hand filing though.
3. I could go the other direction. With the first 4 holes drilled I can shift the guide to the right and drill holes under the support.
   1. The power plugs would be visible though.
   2. I'd have to really grind the braise flat in those areas so the bracket sits flush with the metal.
4. I could drill holes from the outside of the frame in.
   1. Knowing that the hole would line up on the inside would be impossible with a hand drill.
   2. I have holes on the outside I then have to fill or do something with.

Right now I'm leaning between 2 & 3. If anybody knows of a way I could drill those holes without the drill then that may be an option too. It's only aluminum that's 1/16th of an inch thick. We could possibly cob it with some vise grips and a little hand pressure...

 

Other's input would be appreciated. Tonight is cut a little short due to an obligation I have tomorrow morning.