6 Axis Arm

I had a few spare NEMA 17 stepper motors sitting around, and instead of letting them collect dust, I decided to build something ambitious with them: a 6-axis robotic arm. I've always liked the aesthetic of the arms made by Standard Bots, so I set out to design something with a similar look and feel.

In true impulsive fashion, I jumped straight into SolidWorks and designed an entire robotic arm assembly purely based on what I thought looked cool. That lasted until the moment I realized I forgot one small detail: would the motors actually be strong enough to move this thing?

Spoiler — they wouldn't.

I sat down with a notepad and worked through the torque requirements for each joint based on the arm lengths (300 mm aluminum sections I also bought impulsively) and the weight of the motors. Sure enough, the required torques were way beyond what my NEMA 17s could handle.

So the next step was clear: I needed gear reduction. I researched off-the-shelf gearboxes, harmonic drives, and different reduction mechanisms, but eventually landed on cycloidal drives. They offer huge torque in a compact form factor, and they're 3D-printable — perfect for a hobby project. I decided to design my own cycloidal drives (a whole separate project on its own), and once those were in progress, I moved back to refining the arm.

And then the project hit the same wall so many of my hobby builds hit:

my 3D printer died.

I troubleshot it for months, but ultimately the project had to be paused.

Despite the stop, I have a clear plan for the next iteration:

• Make the entire arm 3D-printable, aside from hardware, motors, and the internal cycloidal drive components

• Use the electronics I already have (NEMA 17 steppers, Teensy 4.1 microcontroller, stepper drivers)

• Slim down the design so it's less bulky and more realistic

• Redesign the arms using lightweight printed parts instead of heavy aluminum

• Target performance: the arm should be able to freely move and precisely position a 1-pound payload

Even though this version never made it to physical testing, the process taught me a lot — especially about torque budgeting, realistic design constraints, and not letting looks drive an engineering project without checking the numbers first. I'm planning to return to this build once my new 3D printer arrives, and I'm excited to see how the next version turns out.