[Scratch Build] IKEA Expedit/Kallax compatible NAS server case 2.0

zrav

Limp Gawd
Joined
Sep 22, 2011
Messages
174
Hi guys,
I'm back with another scratch build project. This time it is to replace my first scratch build, the IKEA Expedit compatible server wood case. Working with laser cut aluminium sheets on my Hexplo desktop case, I gained enough new knowledge and ideas to try to improve the original server case concept. The requirements I set went further than the previous case project:
  • Must fit into an IKEA Expedit/Kallax shelf, which means a box of no more than 33.5 x 33.5 x 39cm.
  • Must hold an ATX size board, more than thirteen 3.5" and more than four 2.5" disks.
  • Decent ventilation
  • Even more living room compatible, i.e. cleaner visuals and lower noise
By using metal construction and a SFX PSU, I was able to come up with an interior layout that fits 18x 3.5" disks plus 8x 2.5" disks, perfect for a ZFS pool consisting of up to three 6-disk RAIDZ2 vdevs and a number of SSDs. The metal sheets will be joined by a mix of extruded aluminium profiles and soldering where there isn't enough space for a profile. I'm curious how I'll handle soldering with a butane torch...

qLJ4WKb.png

kiJHjYm.png

As of right now, the drawings have been sent to the laser cutting shop, and I'm in the process of ordering all the other stuff I need.
 
Last edited:

Nimisys

Fully [H]
Joined
Jun 20, 2000
Messages
18,346
Soldering? I would think you'd want the mechanical strength of Brazing for those connections.
 

zrav

Limp Gawd
Joined
Sep 22, 2011
Messages
174
I got both brazing and soldering filler material. The tensile strength difference isn't large (325N/mm2 vs 270N/mm2), so if brazing at 550C melting point doesn't work with the equipment I have, I'll do soldering at 400C.
 

zrav

Limp Gawd
Joined
Sep 22, 2011
Messages
174
After filing some rough spots, sanding the visible surfaces and drilling counter sinks I could assemble the case parts for the first time. It all fits pretty well so far.

cBIwBuw.jpg

dMBXOA5.jpg

FXtmWk0.jpg

ERTocoW.jpg

Most of the sheets of the upper part of the case will have to be joined via brazing (all except the front), as there's no room left for supporting profiles (pics show test assemble using profiles). This is going to be interesting.
 

zrav

Limp Gawd
Joined
Sep 22, 2011
Messages
174
So as it turns out it is quite difficult to heat up a large chunk of aluminium as it is a good conductor and the case offers plenty of dissipation area. With my equipment I was unable to get the proper temperatures for the 550C brazing solder. The 400C kind of worked after lots of heating, but lacked the proper capillary action to slide in the gaps. The result was quite mediocre and took a long time for just a segment of one joint, so I ordered a different solder with capillary action, which should improve things.
 

Nimisys

Fully [H]
Joined
Jun 20, 2000
Messages
18,346
So as it turns out it is quite difficult to heat up a large chunk of aluminium as it is a good conductor and the case offers plenty of dissipation area. With my equipment I was unable to get the proper temperatures for the 550C brazing solder. The 400C kind of worked after lots of heating, but lacked the proper capillary action to slide in the gaps. The result was quite mediocre and took a long time for just a segment of one joint, so I ordered a different solder with capillary action, which should improve things.
I just got done brazing flat sheets of brass and even with an oxy-acetlyne torch, it took quite a bit of time and heat to get it there. Even then it was fun trying to not hit the melting point and blowing through the work metal. Use a thermal resistive work surface if possible.
 
Top