ShorteCut

Efficiency Noob
Original poster
Mar 26, 2016
7
2
Table of Contents
======================
1. Introduction
2. Concept
3. Current state (Help!)
4. General
5. UPDATES - (Latest 2016-03-27)

======================


1. Introduction
I've always wanted a small air-cooled cube case for several reasons, but I've never found one in the mATX form-factor from any manufacturer that suits my needs; so I decided I should try and make my own. I'm currently stuck with a laptop (no specs worth mentioning), hence why I'm making this case for a new system. An issue for me is that don't have any real hardware at hand to take any measurements from. I have looked at the mATX specifications at formfactors.org and other various pages (such as protocase.com), but I can't for the life of me find accurate and specific dimensions/specs for the expansion-socket (PCIE) slots and how they should be laid out relative to the I/O cutout.

This is not really my department (this is not part of my line of study), but I still want to do it. I have worked in Autodesk Inventor before, but I've never done something like this.

TL;DR:
I want to make a custom case for my new system, but I have some conflicting specs/dimensions of the expansion-socket (PCIE) slots and how they should be laid out relative to the I/O cutout of the mobo for example. I'm uncertain which sources to trust and/or to follow (I have conflicting information). If anyone knows what dimensions I can follow, then please inform me where I can look.

UPDATE - 2016-03-27
I went with the accurate motherboard models provided here at SFF (very thankful for them!), but if you see simple yet major mistakes in Fig 5 (Update 5.1 (2016-03-27)), please inform me!


2. Concept
Below are some very early concept images. Many parts have changed since then, but the concept and aesthetics should still be the same. Lots of air-flow with 140 mm fans.

Thanks to:
Yang Ji at grabcad.com for providing the 3D-model of the motherboard.
(https://grabcad.com/library/asus-maximus-vii-gene-micro-atx-motherboard-z97-lga1150-1)

Andi at grabcad.com for providing the 3D-model of the fans.
(https://grabcad.com/library/noctua-nf-a14-fan-1)



Fig 1.



Fig 2.

The panels are going to have a matte powder-coated gunmetal-isch finish. The same for the "mesh" but in white.

All panels, except for 1, are to be made of 3 mm thick aluminium sheets (too thick to work?). I will try to make the case from as few sheets and with as few bends as possible to keep manufacturing costs down. I don't know much about material choice and/or manufacturing of it, so any help on this front would be much appreciated as well, though I'm trying to read as much as I can on my own hand. I'm going to call around to some "nearby" manufacturers, to where I live, to ask whether they can make what I want, or if it is even possible.

The motherboard tray will be one sheet/part and will be able to act as a test-bench of sorts.


3. Current state (Help!)
This is where I'm currently at (and partially stuck/need help! o_O).

As mentioned, this motherboard tray will be one sheet/part and is supposed be to act as a test-bench of sorts if needed/wanted. This is to make for an "easy" installation.


Fig 3.



Fig 4.

As mentioned in the introduction, I don't know how the expansion slots are to be laid out in relation to... well, anything really. Any help would be much appreciated! I have tried to find accurate/specific info on this, but the info seems to vary from different sources. What I've found is as seen in the images above. The standoff motherboard height I've gone with is 10 mm.

I have taken a look at the very nice model here:
https://smallformfactor.net/forum/resources/accurate-blank-motherboard-pcie-card-models.25/

Though, I'm probably missing something. The above model (link) is conflicting with other data I've found and it is hard to tell which to trust and follow.



4. General
I'm usually not active on any forum at all, but I want do this project and happened to stumble upon this nice community here! Any errors (god forbid if I posted in the wrong section) should be pointed out if any. Furthermore, as mentioned, I don't know too too much about material choice nor manufacturing (CNC-machining and bending etc.) of it. Suggestions and comments regarding this is much welcomed.

I'm working in "Autodesk Inventor 2016 Professional" as of now, but I have been looking at Solidworks aswell.


5. Updates
5.1 (2016-03-27)
Ok (very quick update this is), so I decided to go forth and trust the given accurate models from SSF-forum. It took a while for me to find this, but I'm glad that I did. It was hard to know whether the dimensions I had at hand were accurate or not pre finding this community. I added the accurate motherboard model to my motherboard tray and made changes accordingly. Results are presented below in Fig 5 with the reference point for the PCIE-cutouts being the lower right corner of the I/O cutout.

Still, if anyone reads this, and they think I've gotten it all backwards, please inform me.


Fig 5.
 
Last edited:
  • Like
Reactions: Phuncz

Phuncz

Lord of the Boards
SFFn Staff
May 9, 2015
5,802
4,894
Interesting design !

You've configured the fans in your model as all-intake, where will the air be able to exhaust ?
I would try to move one of the side fans (preferably the GPU side one) to the rear above the motherboard I/O, so you can use a large tower heatsink on the CPU, without additional fans. I'd also reverse the fan near the CPU VRMs to exhaust, so you have proper airflow with as much air coming in as going out. Fans that are all configured as intake won't perform as well because pressure is building up.
 
  • Like
Reactions: ShorteCut

ShorteCut

Efficiency Noob
Original poster
Mar 26, 2016
7
2
Interesting design !

You've configured the fans in your model as all-intake, where will the air be able to exhaust ?
I would try to move one of the side fans (preferably the GPU side one) to the rear above the motherboard I/O, so you can use a large tower heatsink on the CPU, without additional fans. I'd also reverse the fan near the CPU VRMs to exhaust, so you have proper airflow with as much air coming in as going out. Fans that are all configured as intake won't perform as well because pressure is building up.

Thanks for the feedback, much appreciated!

Yea, there is supposed to be one exhaust-fan above the motherboard I/O as you mentioned for a total of 5 fans (4 in, 1 out.) The idea was that I will be able to have them run at a low RPM, i.e low noise. If my computer can handle it, I was planning on running some air-flow simulations in Autodesk CFD (I have to learn how that program works first though). I know I just uploaded this, but there are some design bits missing (will update later). As mentioned, the render was from a very early stage :p

I'm redoing a lot at the moment, but I think I've got the motherboard I/O and expansion slots sorted out.
 
Last edited:
  • Like
Reactions: Phuncz

iFreilicht

FlexATX Authority
Feb 28, 2015
3,243
2,361
freilite.com
Welcome to the forum! Very cool idea to design the main mounting tray as an open-air case, that will indeed make installation a lot easier!

The model you've used for placement of the slots is indeed accurate, but the only source of information to really trust on in this scenario is the PCIe Card Electromechanical Specification. While you can only get your hands on the older revisions, the mechanical part didn't change since 1.0. The host you could find via google unfortunately shut down, so I just uploaded the file to my dropbox.
 

ShorteCut

Efficiency Noob
Original poster
Mar 26, 2016
7
2
Welcome to the forum! Very cool idea to design the main mounting tray as an open-air case, that will indeed make installation a lot easier!

The model you've used for placement of the slots is indeed accurate, but the only source of information to really trust on in this scenario is the PCIe Card Electromechanical Specification. While you can only get your hands on the older revisions, the mechanical part didn't change since 1.0. The host you could find via google unfortunately shut down, so I just uploaded the file to my dropbox.

Thank you!

I'll gladly take one copy for reference! For once I'm happy that standards are so hard to change. I used both SSF's model and protocases's model and followed some of protocases's "specifications". Seems to fit both models well enough. I'm sure during manufacturing that they wont be able to have an accuracy of less than 0.1 mm anyway.
 
Last edited:

ShorteCut

Efficiency Noob
Original poster
Mar 26, 2016
7
2
Depends. Pretty much all CNC manufacturing processes can achieve better accuracies than 0.1mm. Not sure about sheet metal bending.

Hm, ok, good to know. Going to make some calls to see what tolerances they can take/hold once I get a bit further. As far as I know, I'm fairly limited to what company I can turn towards. Not to keen on driving too too far.