Concept The ''Rook'' – 12.5L Vertical Mini-ITX Case | Looking for Design & Feedback

[GOBLIN85]

Hi everyone,

I'm a hobby designer and maker from Slovenia. Over the years I've designed several custom enclosures for Raspberry Pi projects and built my own open air Mini ITX PC. A few months ago my son asked me if I could build him a compact gaming PC that looked different from everything currently available. Since his desk is quite small, the goal quickly became designing a case that would have a very small footprint while still supporting standard Mini-ITX hardware and remaining practical to build and upgrade.

The original visual inspiration came from him. He liked the look of the Teenage Engineering Computer-1, so some of our earliest sketches naturally leaned in that direction. As the project evolved, however, I focused on developing our own internal layout and solving the challenges rather than trying to recreate an existing case. This project is still at the CAD stage, and before I invest time and money into building the first prototype, I'd really appreciate some honest feedback from people with much more SFF experience than me.

My goal isn't to build the smallest case possible. My goal is to build the best compact aluminum Mini-ITX case I can for my son's desk. Since this will be my first CNC aluminum enclosure, I'd much rather discover design mistakes now than after machining expensive parts.

So please don't hold back.

If you notice anything that could be improved, whether it's airflow, thermals, cable routing, structural rigidity, maintenance, hardware compatibility, or simply something that would make the case nicer to build, I'd genuinely appreciate your advice.​

Dimensions

• 180 × 215 × 325 mm
• Approximately 12.5 liters

Layout​

• Vertical orientation
• Bottom mounted SFX / SFX-L PSU
• Inverted Mini ITX motherboard
• Direct mounted GPU positioned at the top to draw fresh ambient air without using a PCIe riser. The current design is optimized for compact single fan GPUs and or a Pro card up to 210 mm in length

Hardware Support​

• Mini-ITX motherboard
• SFX / SFX-L PSU
• CPU cooler up to approximately 120 mm

Main Design Goals​

• Compact footprint for a small desk
• CNC aluminum construction
• No PCIe riser cable
• Simple airflow (different front panel designs for airflow)
• Easy assembly and maintenance
• Modular front panels with different material options

Question​

I'd really appreciate your opinions:

• Does the overall layout make sense?
• Would you change the airflow strategy?
• Is the direct mounted GPU the right approach?
• Are there any obvious manufacturing challenges I should consider before CNC machining?
• Would you optimize the dimensions differently?
• Is there anything you would change before I build the first prototype?

I'm currently preparing the first full size 3D printed prototype and will share the results and revisions as the project progresses.

Thank you for taking the time to read this. Every piece of constructive criticism is welcome.

In case not all images load correctly, I've uploaded them to my GitHub repository as well: github/GOBLIN85​

 

[samsonbl4]

Hi. Before I give you any advice, I'd like to know what's more important to you: matching the design to your render or optimizing production costs? Do you want to make the frame as a single piece, like in the render, or from separate parts?

 

[GOBLIN85]

Hi. Before I give you any advice, I'd like to know what's more important to you: matching the design to your render or optimizing production costs? Do you want to make the frame as a single piece, like in the render, or from separate parts?

Hi!

To answer directly:
● the frame is built from separate modular parts, not a single piece.
● You can't really tell from the current render, but it uses corner connectors and central frame pieces.

My main reason is flexibility and modularity. I will reuse main core parts and tweak the layout, that leaves room to optimize individual pieces for lower production costs ?

I'm currently away, but tomorrow when I'm back home, I’ll upload a photo of the first 3D printed part along with a 3D view showing exactly how the frame pieces assemble together!

 

[samsonbl4]

Okay. Here's my advices:
1. Milling internal corners on panels will be very difficult and expensive. Because the cutter is cylindrical or conical and rotates during machining, it's impossible to achieve a corner without a radius. Producing a very small radius similar to a corner will require a micro-mill and many hours of machine work. Moreover, you won't encounter this problem at the 3D printing stage. I recommend choosing a single radius for all internal corners that matches the cutter radius. I think a radius of 5 to 10 mm is appropriate.
2. The surfaces feature many sloped transitions that can be machined using a chamfer mill. To optimize the machining strategy, I recommend researching the standard angles available for such cutters and incorporating those angles into your design.
3. I think the internal ribs on the panels are unnecessary, as they won't provide significant weight reduction (and aren't needed for a desktop setup anyway) but will increase machining costs.
4. It is a good thing if the connecting elements are made of steel and the threads are located only within them. Having threads in aluminum parts—especially for panel mounting—is not the best idea, as aluminum threads can quickly fail. If I were designing such a case for myself, I would consider a structure with a steel frame (most likely made of bent sheet metal) and attach attractive aluminum panels and corner pieces to it.

 

[samsonbl4]

I think the layout you've chosen is solid, and using a tower cooler should keep airflow interference to a minimum. I can't quite tell from the render if the panel resembling the Dota logo is perforated, but it definitely needs to be. You'll need proper intake holes for the CPU cooler, as well as exhaust vents directly behind it.
Regarding the GPU, be sure to perforate the top of the side panels to allow hot air to escape. If the top is your only exhaust, the graphics card might end up recirculating its own hot air (choking on it). In SFF builds, organizing hot air exhaust is just as critical as organizing cool air intake.
You might also be able to reduce the 180mm width. Many SFF cases designed for direct GPU mounting manage to stay around ~160mm wide, though this ultimately depends on your material's wall thickness and internal hardware clearances.

 

[GOBLIN85]

Okay. Here's my advices:
1. Milling internal corners on panels will be very difficult and expensive. Because the cutter is cylindrical or conical and rotates during machining, it's impossible to achieve a corner without a radius. Producing a very small radius similar to a corner will require a micro-mill and many hours of machine work. Moreover, you won't encounter this problem at the 3D printing stage. I recommend choosing a single radius for all internal corners that matches the cutter radius. I think a radius of 5 to 10 mm is appropriate.
2. The surfaces feature many sloped transitions that can be machined using a chamfer mill. To optimize the machining strategy, I recommend researching the standard angles available for such cutters and incorporating those angles into your design.
3. I think the internal ribs on the panels are unnecessary, as they won't provide significant weight reduction (and aren't needed for a desktop setup anyway) but will increase machining costs.
4. It is a good thing if the connecting elements are made of steel and the threads are located only within them. Having threads in aluminum parts—especially for panel mounting—is not the best idea, as aluminum threads can quickly fail. If I were designing such a case for myself, I would consider a structure with a steel frame (most likely made of bent sheet metal) and attach attractive aluminum panels and corner pieces to it.

This is exactly why I wanted to share the project before building the first CNC prototype. Coming from a 3D printing background, it's easy to overlook details that have a big impact on CNC machining. Your comments have already given me several ideas. I'm now planning to redesign the internal frame and simplify the assembly to reduce CNC machining time and make the parts easier to manufacture. Regarding the threaded connections, I was also thinking about using steel threaded inserts instead of cutting threads directly into the aluminum. Do you think that would be a better approach?

One more question, if you don't mind: looking at the current design, which two or three features do you think have the biggest impact on machining time and cost? I'd like to simplify the design where it makes sense while keeping the overall concept intact. I'm also still evaluating the GPU compatibility. My initial idea was to optimize the case for compact single fan GPUs and Pro cards, but I'm wondering if it would make more sense to make the case slightly narrower and a bit longer to support a wider range of GPUs.

What would one consider a good target for 1440p gaming in a compact SFF build? By the time this project reaches the prototype stage, the GPU market will probably have changed anyway, so I'm more interested in designing around a sensible performance target than a specific model.

 

[samsonbl4]

Regarding the threaded connections, I was also thinking about using steel threaded inserts instead of cutting threads directly into the aluminum. Do you think that would be a better approach?

Yes, that's a good solution. Don't forget to use thread sealant if the inserts will be screwed into aluminum.

One more question, if you don't mind: looking at the current design, which two or three features do you think have the biggest impact on machining time and cost?

I don't want to discuss specific elements, as this is a matter of personal preference and may be important to you. If you want to reduce machining time, you can design elements and their radii so that they can be machined with a minimum number of cuts, preferably in one pass or at least two on the inner and outer contours. Also, study the standard cutter diameters and use them. For example, in my experience, a cutter with a diameter of 7 is not the popular. For example, the letter "X" can be made in two cross-cut passes with a cutter with a diameter of 8, resulting in a radius of 4 on its edges, and then finish the contours with a chamfer cutter. You may also want to consider laser cutting some of the structural elements and perforated panels.

I'm also still evaluating the GPU compatibility. My initial idea was to optimize the case for compact single fan GPUs and Pro cards, but I'm wondering if it would make more sense to make the case slightly narrower and a bit longer to support a wider range of GPUs.

What would one consider a good target for 1440p gaming in a compact SFF build? By the time this project reaches the prototype stage, the GPU market will probably have changed anyway, so I'm more interested in designing around a sensible performance target than a specific model.

MSI Inspire RTX 5060Ti, ASUS Dual RX 9060 XT, PowercolorReaper RX 9060 XT are 200-204mm in size and 2-2.5 slots thick, with dual-fan cooling systems. This is a good solution for 1440p gaming, especially if you consider options with 16GB of memory. I think these GPU should fit within your current dimensions. Similar solutions will likely be available in the future. While increasing the size will certainly improve compatibility, I'd suggest you review the dimensions of current solutions in catalogs. In my mind, dual-fan GPU tend to fall within two standard sizes: ~200mm and ~250mm, and increasing the case length by 10 or 20mm will likely add a couple of models to the GPU pool. To significantly expand compatibility, you'll need to add 50mm, which will significantly increase the volume.

 

[hrh_ginsterbusch]

So, layout-wise this mostly matches the Jonsbo T6? that one sits at 185 x 332 x 223, acc. to Geizhals. In that case (no pun intended) I'd look into what Jonsbo did right, and what they did wrong with the T6, and improve on that.

cu, w0lf.

 

[GOBLIN85]

Thanks for pointing out the Jonsbo T6.
I've been going through different builds people have posted with this case and watching a few YouTube videos about it. I think I'm starting to see where some of its compromises are, but I'm curious what you guys here think.

What do you consider the biggest weaknesses of the T6, and what would you change or improve?

It would be really useful feedback while I'm still able to make changes to my design.​

 

[hrh_ginsterbusch]

What do you consider the biggest weaknesses of the T6, and what would you change or improve?

When looking for options to replace the SG16, the T6 was a strong candidate. Availability in Europe was an issue, but the biggest weakness to me was the glass side panel.

Other potential weaknesses include the top and back panel, because of "fancy over function" looking air vents. Even with the "doesnt go fully through" design I bet they're prone to increased noise levels because of turbulence.

The carriage handle also pales in comparison to what other manufacturers, eg. Shiny Snake implemented. Ie. with reinforced case parts, like the mainboard tray in the S300, S400 and S450, or what they call "cross-bracing, Gen2 reinforced handle assembly" on the G300 product page.

Last but not least: Aluminium panels in general tend to be made more for fanciness and not utility character, starting with issues of the scratch resistance, production quality, coating, and also things like: steel allows for magnetic attachments. Aluminium obviously doesnt.

Combined with the price tag - its now sitting at approx 160 Euro; regular cost + shipping + toll fees on initial release would have been around 140ish Euro - the level of customization one would have had to invest was not cutting it for me. Like, testing a replacement side panel in 3d print, then finding a good source for single panel production in Europe with decent pricing, improving on the other shortcomings, would at least have doubled the price.

For 280 Euro you can get a used FormD T1, a fully souped-up NCase M1 v6.1, or an A4-H2O with additional custom side panels and still some pocket money to spare ...

At the end of the day, flexibility is a huge bonus.

cu, w0lf.

 

[samsonbl4]

I'd add support for narrow mATX mobo with two RAM slots and 2.5-slot GPU to the Jonsbo T6. This won't affect the footprint, and the height won't increase significantly.

 

[GOBLIN85]

I'd add support for narrow mATX mobo with two RAM slots and 2.5-slot GPU to the Jonsbo T6. This won't affect the footprint, and the height won't increase significantly.

Could you give an example of the “narrow mATX” motherboard you have in mind? What dimensions would it be?

If you have a specific motherboard in mind, could you share the model or a picture/link?
I'm curious what kind of board would fit this layout.

 

[samsonbl4]

I'm referring to mobo with dimensions up to ~200 x 225 mm. For example, the MAXSUN Challenger B760M-F, ASRock H810M-H, and ASUS ROG CROSSHAIR VIII Impact.

 

[GOBLIN85]

Design Update
"Revised ROOK Concept"

Hi everyone,

It's been a little while since my original post, but I've been working on the design and trying to incorporate a lot of the feedback you guys gave me. So here's a first look at the revised concept.

I simplified several parts to make them more practical and less time consuming to CNC machine, while trying to keep some of the character of the original design. The front has been redesigned with space for 2×120 mm fans, and the side panel is now fully perforated. This also made me wonder whether it could potentially support a 240/280 mm radiator for CPU or even GPU cooling, something I'd like to explore further.

I also moved the power button to the bottom panel, where I'm planning to have two USB ports as well. I'll go into more detail about the I/O once the design gets closer to an actual prototype.

But the main feature I'm experimenting with is the adjustable case depth.

Instead of designing the case around one fixed maximum GPU length, the rear section, together with the SFX PSU, can be moved outward and locked into several positions. The idea is that someone using a compact GPU can keep the case as small as possible, while someone upgrading to a longer GPU can simply extend the case rather than having to replace it.

I've only done some rough internal measurements at this stage, but the current positions give approximately:

*Position 1 – 204 mm
Most compact configuration
*Position 2 – 216 mm
Slightly extended
*Position 3 – 228 mm
The rear extension starts becoming visible from the side
*Position 4 – 240 mm
*Position 5 – 252 mm
Maximum extension​

The difference between each position is currently around 12 mm. These dimensions aren't final yet, but they should give you an idea of what I'm trying to achieve. I still need to finish the bottom perforated section and feet, as well as the perforated top section where the optional handles would be mounted, but I wanted to show the idea before going much further.

So now I'd really like to hear what you think.

Does the adjustable depth idea make sense to you, or am I overcomplicating something that should stay simple?
Would the ability to keep the case compact with a short GPU and extend it later for a larger one actually be useful to you?​

And perhaps most importantly: is this becoming something you'd genuinely be interested in building/testing, or would you still rather choose one of the existing SFF cases already on the market?


Don't hold back, criticism is still very welcome.