The RubiX-Cube is a physical modeled cube that makes traditional cubing more interesting and difficult, but cooler; it doesn't rely on physical turns! Instead, it relies on digital turns that get translated to the cube via LEDs. That's right, LEDs. The cube will be loaded with 54 addressable neopixels that correspond to user input. It will also come preloaded with shuffles for the user to get unlimited shuffles, without actually shuffling the cube, unlike a traditional cube! Think of it as an auto-shuffler.
Basically, the way that it works is through a physical aspect, electronic aspect, and software aspect. The physical aspect includes the CAD and acrylic as the shell for the electronics to rest in. It resembles a Rubik's cube but contains holes for each 54 LEDs and divots on each row so you could tell each individual cube apart..
The components are listed above in the images!
Journal(s): Stardance Journal: : https://stardance.hackclub.com/projects/21632 Message Thread(s) (follow the links to OG message): https://hackclub.slack.com/archives/C0B04RP43TQ/p1782753392262719 PDF w/Summary + BOM: BOM+SUMMARY.pdf _____________________________________________________________________________________________
This was my original idea for the project:
I then started drafting many prompts to ask Gemini on how I'd even begin to approach the project:
I then realized that LEDs might be a bit too difficult, so I came up with two ideas (I unfortunately lost the sketches for these :C):
____
_
And then I got a response which ended up changing the projection of the project.
Additionally, the cube itself would be carried by a main code (controller code would correspond to the main code and deliver information to it) but the main code would basically use algorithims or shuffle databases from online to create shuffles and deliver that information to the cube! No informatation would come from the cube to the code, rather it would all be simulated on the computer + a reset button from the computer so that the cube could follow the simulation cube every time it updates, so no additional thinking is required on the physical cube part other than addressing the LEDs!!
However, I decided to work on this part as my projects getting reviewed, since it's free and I could do it while waiting!
I took more notes on eletrics after asking a few friends on Instagram, Discord, and Slack!
I tried to do a draft and then a formal engineering isometric + top + front + side drawing.
I needed a sliding mechanism (or so I thought) to attach the top and bottom so I could load the inside with electronics.
After MANY revisions and advise I got on doing a dove-tail joint (which was a super cool idea), I decided to just go with screws as this is a test + I'm on a time constraint :C
One more idea I had before ultimately going with the screws:
s:
I then finished up the CAD and research on materials... I switched to neopixels since they're addressable LEDs, basically meaning they could be programmed! Instead of doing the SK6812, I ultimately went for the individual ones due to spacing issues in the cube itself (a strip wouldn't work). I ultimately decided on an ESP32 as my microcontroller (MCU). However, I now had an issue with the batteries. I decided to use one 5volt battery bank for the neopixels, and a 3.3v liPo battery for the ESP32 itself to avoid building a buck booster (time constraints, but this will definitely be in my v2.))
Additionally due to time constraints, I've decided to create a custom PCB via a perfboard or a perma-protoboard so that I could map things out onto a breadboard and continue from there. A custom JLCPCB is the goal in V2 of the project!
Someone also recommended and reached out about using acrylic screens to have the LEDs really glow and stand out. Perfect! I created sliders on the CAD for them, and hopefully they turn out so that the LED color gets diffused! Super cool.
Ultimate CAD Final:
(Almost) Ultimate Schematic Final:
(Almost) Ultimate Wokwi Simulator:
And Ultimate Notes + BOM!
Thank you for reading :D