Building a 5DOF robot arm 1

It’s pretty rare that I get 6 uninterrupted hours. I’m pretty sure I spent most of the time trying not to bite my nails, which is what happens when I’m working on a really hard problem. Let’s take a tour of where I’m going here and what’s changed.

The high-level idea is a differential drive, a bit like a coreXY system. When both motors turn together the wrist (Far right) will tip up/down without twisting. If they don’t turn together the wrist tip and twist. Each motor turns a belt that runs up the arm to the wrist. One of the belts connects the beveled gear on the north side. The other connects to a shaft that runs through both north and south but only connects to the wrist on the south end. The bevelled gear spins freely on the shaft and twists the wrist. Clear as mud? Ok.

I’ve been told that I can shave a lot of weight and add rigidity by “using lots of triangles” and adding cross braces … so I have. I also added cross braces that I hope will help reduce side-to-side sway. My intuition says it’s still going to wobble a lot. Try it and see!

All the joints have been widened with slew bearings that I designed, printed, and tested already. They turn very nicely (considering my print quality) and the larger diameter will take a lot more stress than the 3mm screw and 8mm bearing I was using before.

What else, what else… Both stepper motors in this version are on the same side to make assembly easier. In my first prototype they were face to face and I had a hard time tightening the screws. I may also try putting one motor on each side but not face-to-face – it might make the belts work better.

There is a hole between the two stepper motors where I’ll put a metal spacer that manages to avoid the belts and provides an anchor for the system that will bend the elbow. Bonus: it keeps the belts from pulling the motors and bending the arm. I think I’m also going to add one on the very tail end.

This is easily the most challenging machine I’ve ever built and I’m loving every minute. I’m turning down requests left and right to make the arm do this or that because they’re all trying to pidgeon hole the machine into one specific job. I want a good general use machine and let everyone else play with ways to put it to use. Put a nozzle and build a pick-n-place. Put a Bowden extruder and 3D print. Put a claw and play Chess Clock Jenga. I hope to create an ecosystem that spawns whole new business models.

Speaking of which, I want to thank you all for reading along, using the forums, your feedback, and your help. I couldn’t do this without you. Keep it up! Together we’re going to do great things.