Overclock.net - An Overclocking Community - View Single Post - Roisin Dearg, a Scratch Built Case (part 2 of Preparing for a Scratch Built Case)

View Single Post
post #4 of (permalink) Old 06-16-2017, 10:41 PM - Thread Starter
Lady Fitzgerald
Senioritis Member
Lady Fitzgerald's Avatar
Join Date: Apr 2013
Location: AZ, SSA (Squabbling States of America)
Posts: 8,231
Rep: 522 (Unique: 329)
01/08/2018 I! HATE! FILING! Today, I cut the hole for the back of the MOBO Tray into the rear panel. If I had a dollar for every file stroke it took to make that stupid, convoluted, hole, I would be the world's richest woman. By far!

I'll spare you most of the the gory details since my hands are cramping horribly from all the filing I had to do (layout was also a chore) and I'm having trouble typing but here are the pictures.

This is the rear panel with the new hole:

And here are shots of the installed MOBO tray:

I ran into a problem while installing the rivet nut inserts in the rear panel. I kinda sorta over tightened the installer cap screw on one of the inserts and stripped the inside of the hex recess in the head of the screw and rounded off the end of the driver bit. Fine, I thought, I'll just take a pair of vise grips and back out the screw. No dice, the threaded plug inside the insert would rotate. After a little (ok, a lot) of therapeutic screaming, I finally figured out I would have to use my Dremel to cut off the head of the screw. That was one hard head. I almost demolished a cutoff wheel getting it cut...and I was using a very light feed. I finally got everything to break loose and I was actually able to back out the remains of the screw by hand. Fortunately, the insert and mandrel weren't damaged and I have plenty more screws and bits.

I used a jig saw to do most of the cutting and learned (more like remembered) a couple of tricks along the way. The blade kept getting clogged with fused aluminum when I was cutting and it was a bear to clear the teeth. To prevent the clogging, I first reduced the rate of feed to reduce heating. I would also back off from the cut every two inches for a few seconds to let the blade cool a bit more. Every so often, I would rub the teeth of the blade with a piece of chalk (I figured if chalk helped to prevent aluminum from clogging file teeth, it might work for the saw teeth). It did seem to help.

Inside the notches I had to cut, it was much easier to use a cutoff wheel in my Dremel than to use the jigsaw. I found I had much better control (and had to do much less filing afterward) by using the right angle attachment on the Dremel.

I still need to make another hole in the rear panel between the back of the MOBO tray and the PSU for the bypass panel (laying out and cutting that one will also be a chore!). Then, I get to layout and cut the hole in the MOBO tray back plate behind the MOBO tray and figure out how to layout the holes for the two threaded standoffs supporting the right side of the MOBO tray (that last part has me stumped right now; I'm too pooped to think about it right now).

01/09/2018 Today went better. I cut the hole in the MOBO tray back plate under where the MOBO tray goes. This will allow for access to cables being snuck under the MOBO tray which eliminates the need for cable management holes (which are never in the right place) in the back plate.

Here the back plate's new hole (duh!):

This is the view from the right side of the case after I reinstalled the back plate:

I figured out a trick to reduce the amount of filing I need to do ( thumbsupsmiley.png ). It's still faster and more accurate to make large cutouts by using the jigsaw. However, the cuts are rather rough and not perfectly straight (it doesn't help that I can't see the line very well while cutting). Since I can control my Dremel much better when using a cutoff wheel by using the right angle attachment, I was able to use it to quickly knock off the high spots on each side instead off laboriously filing them down. All I had to do was smooth out the cuts with the files, a fairly fast process.

I also figured out how to locate the holes for mounting the 3/4" long standoffs supporting the left side of the MOBO tray. While the standoffs were screwed to the MOBO tray, I put some 3M 6 mil VHB tape on the end of each one, trimmed it to fit, then put a long piece of the release strip on to prevent the tape from gripping the back plate prematurely:

Then I reinstalled the MOBO tray, pulled out the release strips, and pushed the standoffs firmly onto the back plate. After carefully removing the screws holding the standoffs to the MOBO tray, I carefully removed the MOBO tray, leaving the standoffs stuck to the back plate. After carefully removing the back plate, I chucked a 7/64" bit into my Dremel and, using the standoffs as a drilling guide (and hanging onto the standoffs for dear life so they wouldn't break the bond of the tape and shift), I drilled holes into the back plate. I then enlarged the holes to 9/64" to just clear 6/32 screws and countersunk the holes. I found it necessary to drill a slight countersink on both ends of the standoffs since I can't get any undercut 6-32 screws in black oxide. The standoffs will stay attached to the back plate and, whenever I need to remove the MOBO tray, I'll just remove the screws from the MOBO tray side.

I also got tired of looking at the bright stainless steel screw heads on the back end of the MOBO tray so I replaced them with the black oxide screws I finally found:

I'll probably epoxy over those screws so I won't accidentally try to remove them when removing the MOBO tray from the case. I have both 4-40 and 6-32 screws in black oxide now and will use them when I'm finished with case fabrication and painting. Other than filling scratches, sanding, and painting, I'm declaring the MOBO tray back plate finished.

I still need to layout and cut the hole for the bypass panel on the rear case panel. Once I finish that, other than filling scratches, sanding and painting, the rear case panel will be finished. At that point, I'll be ready to start work on the top, front, and side panels. I already have a couple of top panels cut to size except for the fan holes. I'm going to do one with two fan holes and one with three fan holes so I can experiment with how much exhaust I will need. Once I find out which one works best, I can veneer the panel I'll be keeping.

01/10/2018 A I laid out and cut the hole for the rear I/O bypass plate this morning. This is the plate that has connectors on it to allow me to use jumpers from the MOBO rear I/O panel so I can use the two e-SATA ports on the back as internal SATA ports and relocate the two USB 3.1 Gen 2 type A ports on the rear I/O to the "front" panel. I had originally planned on putting it to the left of the MOBO rear I/O panel but there wasn't enough room there so I relocated it to above the MOBO tray rear panel. Everything was a tight fit and it was a bit tricky to layout but everything fell together slick as snot on a broom handle on the first try.

First, after laying everything out on the rear case panel, I drill holes for the rivet nut inserts and installed them. The opening itself passes so close to the inserts, I was afraid they would blow out the sides of the holes if I installed them after cutting the main hole so I installed the inserts before cutting the main hole. Once the rivet nut inserts were installed, I installed the right angle attachment to my Dremel, chucked up a cutoff wheel, and went to town. I'm still amazed by how much better control I have over the cutoff wheel when using the right angle attachment. It took very little filing to clean up the cuts (for which my hands and I are ecstatic!). When I first installed the plate, I was pleasantly surprised when it "slud" right on in (I was expecting to have to do a lot of "tune up" filing due to the tight fit). Short of filling in scratches, sanding and painting, I can't think of anything else that needs to be done to this panel so I'm declaring it done.

Ok, enough yapping. Here are the pictures. First, the hole itself:

Here, the bypass panel has been installed (I had to modify it a while back so it now needs repainting):

This is the inside view before installing the MOBO tray...

...and after the tray is installed:

This kinda sorta shows how the cables will run between the MOBO tray and the MOBO tray back plate.

I can then route the cables along the backside of the MOBO tray backplate.

I lied about the MOBO tray back plate being finished. It's a bit awkward installing the MOBO tray and it dawned on me that it's going to be a "heckuvalot" more awkward when I have the MOBO, the big NH-D15s cooler, and the graphics card installed, so I need to attach a piece of angle on the bottom of the back plate to support the weight of the tray while removing and installing the MOBO tray. I don't know when I'll get to that. I had a short night and it's just caught up with me so I need to take a nap after lunch.

01/10/2018 B I installed the angle onto the bottom of the MOBO tray back plate. It does what it's supposed to: supports the tray to make it easier to remove and install the MOBO tray.

01/13/2018 Yesterday, I finished the top panel (or as much as I can do until I paint and veneer it). First, I've been having all kinds of "fun" (polite term) using the three Morse hole saws I have. I already got rid of the Morse mandrel I had because it had almost 1/16" runout (I returned a second one because it also had excessive runout). The Milwaukee universal mandrel I replaced it with runs pretty much straight. The hole saws themselves are out of round. They have a slight bulge where the saw is welded together and they would cut between 1/16" and 1/8" oversized. It's really a pity since Morse used to make really good tools. Now, they are garbage. After the battle I had cutting the 4/12" hole for the PSU bump out cover, I decided to get a Lenox 4 1/2" hole saw. I wished I had done that a long time ago. This one was a joy to use.

I decided to use three fans on top (in addition to the one on the rear panel of the MOBO tray in back of the CPU) and located them a bit off center to ensure the best possible airflow past the PSU bump out. I added the third fan to help ensure no heat could build up in the 5.25" bay. After laying out the hole centers for the big fan holes and the fan mounting holes, I set up my drill press and its auxiliary table on top of my washing machine (after running a load of laundry this morning) and started to drill the pilot holes for the 15 holes I needed to drill or cut. I ran into a "small" snag. I could easily reach one row of six pilot holes but my drill press was 1/16" shy of having enough depth to reach the rest of the holes (mutter, mutter, mumble, mumble). I had to use my 18v portable drill to finish drilling all the holes. To cut the three big holes, I dug out my old 18v drill that has a broken speed control (it runs only at full speed for each of the two speed ranges). Doing it manually had me a bit worried but, even though I had to hang onto the drill for dear life, I had no problems cutting the holes. What's curious is a battery operated drill is more powerful than the alleged 1/3 HP motor on my drill press (the poor horse must have been sick). The cuts were far cleaner than what the Morse hole saw would have done. I only needed to clean up the edges a bit with a coarse (the forums' nannyware won't let me use the correct word) 5/8" rattail file (I would hate to run into the rat that had that tail!) and finished off the edges with a finer file. Sorry, I forgot to take pictures but you will see what it all looks like in a bit.

After all the drilling, cutting, and filing, I placed three fans over the holes and dropped in some 6-32 screws to make sure the mounting hole alignment was correct. It was; the screws literally dropped right in. In these shots, I positioned the fans so the cables would all end up in roughly the same place and plugged the cables into a fan hub to check for where to install the hub (excuse the mess; I was too lazy to clean up for the photos):

Once I determined where to put the fan hub, I took it apart, then drilled and tapped the base for 6-32 screws (the material was something like a high density polyethylene or nylon), then used the base to locate the mounting holes in the top panel. here is what the base looks like after being screwed to the top panel (I had to cut down the screws)...

...and the base with the hub itself set in place:

This is the box it came in (in case you are curious):

I'll be using the eight port version for the eight intake fans.

This is what the fans look like from above the top panel (I haven't cleaned off the layout fluid yet):

Again, I didn't take pictures of the process (sorry) but I cut and riveted some 1/2" angle to the underside of the top panel to reinforce it:

This is the topside of the top panel after cleaning off all the layout fluid (sorry about the lousy photo):

All the rivet heads and the two fan hub screw holes will eventually be covered with wood veneer. I dropped the panel onto the top of the case to make sure it still fits (it does):

Except for the rear panel, all the panels will be secured with a combination of small (but powerful!) rectangular magnets and magnetic tape. Using this method will allow me to use foam weather stripping between the magnets to ensure the panels are reasonably airtight. I had planned on using just the magnetic tape only but I was concerned that, if the panel warped a bit, it wouldn't seal up well.

Only three more panels—front, right side, and left side—to go then I should be ready to start painting. I'm saving the left panel for last because since it will have a window and will probably be a real stinker to make.

01/14/2018 A I changed my mind and worked on the large right side panel. I had just enough 1/16" x 1/2" angle to do that side so I worked on it rather than have more waste on the front panel (I'll have to get some more angle in the next day or two, if I can find it in this one horse, backward megalopolis I live in). Even though I didn't have to horse around with making fan holes, this one was more work than the top panel due ti it's size, I didn't have the aluminum sheet already cut, and there were one "heckuvalota" holes and rivets in it. I'm sorry I don't have pictures—the ones I took look "horridable" due to glare from the shiny aluminum and I'm too pooped to retake them—but the panel like pretty much like the top panel except it's way bigger and it doesn't have any fan holes.

Here are the steps to making this thing:

1. I first cut the panel a wee bit oversized using my circular saw with the carbide blade in it (I'm amazed it is still cutting so well), then filed the edges to smooth out the saw marks and bring it down to the final size (I still hate filing!).

2. I cut the angles to size, laid out the rivet holes, center punched them, drilled pilot holes, drill the pilot holes to 1/8". then deburred the holes (there were "only" 46 holes rolleyes.gif ).

3. I used fine ScotchBrite to remove the oxide from the side of each angle that had the holes drilled in it, then cleaned it with 91% isopropyl alcohol to prep the surface for the 0.006" 3M VHB tape that will be used, along with rivets to secure the angles to the panel.

4. The tape was applied to each angle.

5. Since the bottom angle is going to transfer the weight of the panel to the bottom frame cross member, I shimmed it with six layers of blue painters' tape (to provide some wiggle room when installing the panel), then used some small strips of carpet tape to fasten it to the bottom frame cross member.

6. I prepped the surface of the sheet aluminum where the bottom angle was going to go, same as I did for the angles.

7. After removing the release strip from the VHB tape, I put the sheet aluminum into the frame recess (the frame is laying on its side) and lowered it onto the tape. This was probably the quickest and most accurate way to locate the bottom angle.

8. After removing the sheet aluminum from the frame, I stuck the remaining angles to the aluminum.

9. Using the holes in the angles as drill guides, I drill the 1/8' holes through the sheet aluminum. I was able to use the drill press for this but, even with the auxiliary table installed, it was a bit awkward due to the size of the panel. I didn't have any problems; it was just mostly tedious.

10. After drilling the holes, I counter sunk both ends of each hole. Again, very tedious (longer arms would have been useful).

11. I installed 46 blind rivets that are flush on both ends when installed. That went pretty well (though, again, tedious) although I had four mandrels get stuck inside the rivet setter that required swapping out nose pieces to get them out.

Making the panel wasn't difficult; it was just a LOT of tedious work.

I don't know when I'll start on the front panel (I could use a day of rest; my tired hurts). I'll have to cut it the same way I did the other panels. I have enough aluminum angle left to make the top and bottom pieces but I'll have to find some more angle before I can finish it. I have some thicker aluminum sheet and angles on order for the left panel since that one has to be as stiff as possible since it will have a tempered glass window in it.

01/14/2018 B Today, I decided to start working on the magnet system for securing case panels to the case. I had thought that would be fairly easy (boy, was I wrong!). I started with the largest panel, the right side case panel since it will be the heaviest one. First, I secured four magnets to the case frame by installing 6-32 rivet nuts on the panel flange, then screwing the magnets down with 6-32 x 5/16" flathead screws.

Since the magnets are such powerful little things (they can give you a nasty pinch if you get skin caught between two of them coming together), I had planned on using magnetic tape on the panels to reduce the pull needed to remove the panels (I was concerned I might permanently bend a panel). I stuck four pieces of the tape to the panel but that proved to be too wimpy of a pull to remove the panel so I was back to Plan A: using magnets on the frame to secure them to magnets on the panels. I first tried temporarily taping the magnets to the panel with carpet tape. The tape held for the top magnets but pulled loose on the bottom magnets. The pull didn't feel too strong up at the top so I replaced the carpet on the bottom magnets with VHB tape. Since it takes at least an hour for the tape to reach maximum bond, I waited an hour before I tried removing the panel. That held and the pull seemed to be plenty to hold the panel in place without needing two men and a boy to remove it.

Since I didn't trust the tape alone to be strong enough over time, I drilled a hole in the panel, countersunk it, then installed the rivet. That did not end well. Those magnets are really brittle

On to plan B but, first, I had to remove the rivet. Those things are real stinkers to remove because they spin while being drilled. I found out that, if I angle the drill about 20-30°, it would chew up the rivet enough to allow it to be punched out. Plan B—reverse the rivet—didn't work on the test piece I was working with. After trying out a few other "Plans", I found that I could still use the original rivet but, instead of actually setting the rivet until the mandrel pulled through, I "squoze" the rivet setter only until I reached increased resistance (that didn't take much), then releasing the mandrel from the setter and removing it back through the head of the rivet. That held my test magnet on tight enough I couldn't pry it loose without the VHB tape. Here are the magnets installed on the panel:

This is what the rivet looks like on the outside of the panel:

Despite how little of the rivet is grabbing the magnet, it's still enough ('tis a pity I can't get them in longer lengths). To avoid depressions in the veneer after I apply it, I'll have to fill these "divots" as well as the other rivet "divots" before applying the veneer.

Because the countersunk holes are larger than the rivets, I had to get a little creative to set the rivets when they are depressed in the countersink. I used a 4-40 Nylock nut as a spacer between the rivet and the setter.

The pull it took to remove the panel seemed a little stiff while the case was still laying down on its side. Once I stood the case up, I found all I needed to do is slip my fingers behind the panel at the bottom and place my thumbs on the frame and pull. The bottom pulled loose fairly easily and, once the bottom pulled out, the top was easy to remove but, when in place, the panel is not going anywhere unless I want it to. Those four little magnets are plenty strong without being too strong.

01/14/2018 C Due to the length of the panel I just made, it flexes a bit in the middle of the sides, even with the 1/2" angle reinforcement. To prevent that flex, I cut six approximately 1" long 1/4" x 1/2 spacer blocks and used VHB tape to secure them to the panel flanges on the case frame. The tape should be plenty since there will be only very light compression forces on them.

Here is a view of the s[acer blocks on one side of the case (the other side of the case also has them)...

.and on the bottom cross member (there is also one on the top panel flange):

I will install foam weather stripping between the magnets and the spacer blocks to block as much air leakage as possible.

What nice about using the magnets to secure the panels is there are no screws or latches showing on the panels and to wrestle with.

01/20/2018 I picked up the angles I needed Thursday and finished fabricating the front panel yesterday. it was basically more of the same ol' same ol'. Here are the pictures (sorry for the quality. My camera just doesn't like to focus on shiny stuff):

I had to file down some rivet heads that were a bit proud of the surface so they wouldn't show through the veneer that's going on them. I'll have to fill in any holes and slightly recessed rivet heads with epoxy putty so they won't telegraph through the veneer.

I finally found some wood veneer yesterday. I picked up a 2' x 8' paper backed sheet of figured anigre. Even though gluing it onto the panels will be a bit of a chore, I'm looking forward to it. I have a trimmer I've used for trimming edge banding that should work fine for trimming the straight edges of the panels without knocking off the paint on the edges of the aluminum. However, I have no clue yet how I will trim the fan holes on the top panel.

Yesterday, I picked up the sheet of .090" aluminum for the front panel I ordered. It's a bit thicker than the 1/16" aluminum I've been using but I wanted the extra stiffness to help ensure panel flexing won't break the tempered glass I'm going to get for the left side panel. I also decided to reinforce the panel with 1/4" x 1/2" bar stock instead of the 1/8" x 1/2" x 1/2" angle I had originally planned on using. I was supposed to have also received that yesterday but the USPOS is having problems. It, and another package, have been showing being in limbo on their tracking page so I have no idea where they are or when they are going to arrive. If I don't see some activity by Monday, I'll have to raise a ruction with the poorly trained idiots at their "help" desk.

In the meantime, I'm going to get started cutting out the left side panel, hopefully today. That one is going to be more work because of the added thickness of the sheet aluminum and because of the window that has to be cut out.

MSI R7850 Twin Frozr 2GD5/OC Radeon HD 7850 2GB 256-bit GDDR5 PCI Express 3.0 x16 HDCP Ready CrossFireX Support Video Card
Monoprice 1x2 powered HDMI Splitter
Kingston HyperX 32GB (8 x 4GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) Desktop Memory Model KHX1600C9D3K8/32GX
Hard Drive
128GB Samsung 840 Pro SSD
Hard Drive
4TB Samsung 850 EVO SSD
Hard Drive
4TB Samsung 850 EVO SSD
Hard Drive
4TB Samsung 850 EVO SSD
Hard Drive
4TB Samsung 850 EVO SSD
Optical Drive
LG 12X BD-ROM 16X DVD-ROM 48X CD-ROM SATA Internal Blu-ray Drive CH12LS28
Power Supply
Corsair HX750W
COOLER MASTER GeminII S524 120mm Long Life Sleeve CPU Cooler
Antec Two Hundred v2
Operating System
Windows 7 Ultimate
3 x Asus VG248QE
Vizio VO320E 32" TV
Logitech G510s
Logitech M525 with Unifying Receiving
Corsair SP2500 2.1 Speakers
ASUS Xonar Essence STX Virtual 7.1 Channels 24-bit 192KHz PCI Express x1 Interface 124 dB SNR / Headphone AMP Card
LSI 9211-8i HBA card
HooToo® HT-CR001 3.5" PCI-E to USB 3.0 Multi-in-One Front Panel Internal Card Reader (6 Memory Slots, Genesys GL3220 Chipset) & SuperSpeed USB 3.0 3-Port Hub (VIA VL800 Chipset), can read & write: M2/MS Micro, High-Speed CF(UDMA), UHS-I, SD, SDHC,...
StarTech HSB220SAT25B 2 Drive 2.5in Trayless Hot Swap SATA Mobile Rack Backplane
i7-5930K Haswell-e
Asus X99-E WS/USB 3.1
Visiontek Radeon 7870 2GB with 6 miniDisplayports
G-Skill Ripjaws 64GB (8x8GB) DDR4 2133
Hard Drive
Samsung 850 EVO 4TB SSDs
Hard Drive
Samsung 950 Pro 512GB m.2 SSD boot drive
Power Supply
Corsair AX760
Noctua NH-D15S Cpu cooler
Scratch built. Currently under construction at https://www.overclock.net/t/1602023/preparing-for-a-scratch-built-case
Operating System
Win 7 Ultimate
Logitech M525
El Cheapo Allsop hard plastic
▲ hide details ▲
Lady Fitzgerald is offline