Trying (and failing) to make a Kallax PC case
I have been on quite the computer case journey. When I was younger, I wanted something intimidating. Bigger was better. Then as I got older and started moving around more, smaller became the goal.
As I got into small PCs, I realized it was not so much about the size for me, but the aesthetics. I wanted a powerful PC that would disappear into the background, and when my partner started thinking about a different case to replace their big aquarium-style one, I pitched an idea: a PC case that fits perfectly in an IKEA Kallax cube.
Initial concept
I’ve been getting more into 3D printing lately, and I was excited to take on larger project. It would be physically large but, more importantly, multi-part with specific tolerances and layouts I would have to adhere to. It felt like a more realistic design challenge than some of the things I had done where I had free reign to lay things out any way I wanted.
I mocked up a quick concept in a couple of hours and thought things would be smooth sailing. Spoiler: things were not smooth sailing.
Printing. So much printing.
I knew I would need to split the case into parts for it to fit on my printer bed, but I underestimated how much more complicated that made things.
The design called for a horizontal joining system, a vertical joining system, and a way to support the top shelf of the case. I went though many, many iterations of various ideas until I finally landed on something I was satisfied with.
I printed it out, put it all together, and it was… okay. It worked, but it felt a little too precarious for me to trust putting heavy computer parts on it. I had thoughts about fixing it, but at this point, I didn’t want to go with a completely 3D printed design. Printing the case had already taken long time and required way more filament than I anticipated, making the value proposition compared to a “real” case way less appealing. I went back to the drawing board.
Aluminum framing
After some research, it hit me: an aluminum frame with 3D printed panels. The aluminum would give me the stability I was looking for, and I could still make it aesthetically pleasing. It was perfect. Spoiler part two: It was not perfect.
I did a ton of research into aluminum framing. I wanted to use MakerBeam because I was vaguely aware of it, but it’s not widely available in the US. I ended up going with 80/20, which is basically MakerBeam in all but name. I had no issues with the product — my issues came from poor planning.
The biggest problem was the joints I chose. I went with the cheapest and most basic 90 degree joints, but these really weren’t the right tool for the job. 80/20 also offers a more robust 90 degree joint with two screws on each side instead of one, and I suspect that would have given the whole thing way more stability.
As it stood (or more like wobbled) though, I still thought it was worth salvaging. I modeled a cap to my panels that would slide into the aluminum bars, refined the dimensions, and stuck it all together.
All I could hear were sad trombone noises. It functioned, but nothing about it felt perfect. What was supposed to be a more robust take on the design was even more precarious. I had to be very careful with how I picked it up and moved it around. Still, I thought maybe it would be okay. We wouldn’t be moving the case much anyway.
Foiled by the QUBE
After all my research, iterations, and all my printing, I finished the case and set it up. My partner used it for… about a week.
I wasn’t upset; I saw all the same problems and even more. I was about to go back for yet another iteration when I happened to find a different, commercially available case and that sealed the deal: My design lost out.
In the end, did it even matter?
Something I have been tackling for the past couple years is facing failure and learning from it. As a child, lots of things I tried fell into two categories: either I was pretty good at them from the start or not good at them at all. School was something I was pretty naturally good at — I was smart. Sports were not really something I was good at. Going out of my way to make new friends was not really something I was good at.
It was convenient to just write those off as things I would never be good at. It was easy to have low expectations of myself in other areas because I was always smart; I could always fall back on that.
And sure, this is going to be true for everyone to an extent. No one is good or wants to be good at everything. As I got older though, even the things I was good at got harder. School was no longer that easy for me; I struggled not just because it was literally hard, but I felt like my identity was falling apart. If I wasn’t getting good grades, if I wasn’t smart, what was I?
Thinking back, of course I was smart. I was just in more difficult situations and, because I never really tried to overcome not being naturally good at something, I didn’t know how to face failure. I did not enjoy challenge; it scared me. So I did not try to improve. I curled up further and further into the few things I knew I could do and didn’t dare step outside them.
It took a lot of reflection and experimentation to break that cycle. I’m far from perfect at accepting failure now, but I’ve gotten a lot better. There is a version of myself that would be upset and frustrated by this project, but over the past few years I’ve worked to dive into the deep end and really embrace failure.
The harder we fail, the more we can learn because the areas for improvement are so clear. It might be painful in the moment, but refocusing on getting better for the next time is both a short and longterm fix. It helps get over the fear of failure in the short term and makes you even more capable in the long run.
This project has been a great case study. I failed at many points, overcame obstacles, and in the end, it still didn’t work. I could dwell on the wasted time, but if I look at it another way, it wasn’t wasted. I learned a number of things:
- Don’t underestimate multi-part prints.
- When you have a 3D printer, it’s easy to think everything can be solved with it. In reality, it’s a tool with pros and cons like anything else. It’s important to use it for the right jobs.
- One point of contact on each side does not create stable 90 degree joints.
- Maybe the most important one: It’s crucial to understand the parameters of the problem before starting on a project. If I had dug deeper into my partner’s thought process, I probably could have figured out the whole endeavor was an uphill battle to begin with. They weren’t that invested in the idea of a hidden case, just one that fit better on their desk.
I’m happy to have learned these lessons in kind of a painful way because they’ll stick with me that much better. I’m calling this one: task failed successfully.