Approach
I have worked my way through a course at Stanford on Youtube. The course is titled “CS235: Applied Robot Design for non robot designers. How to fix, modify, design, and build robots.” The classes had excellent content as I wanted to learn more about building robots, but did not have any experiences in college doing it. The hard part of taking the course is doing the lab work with the limitations in my apartment. I have a 3D printer and some assorted hand tools. Below is a summary of the project and the changes I made to the design. I completed the lab by using the video of the lab to reverse engineer the components and build the project. I did not have access to any of the files from the actual course. I only had the video lectures.
Design&Build
Some pieces of the project that I changed from the original design were the frame, differential block, and mounting circle. These all involved 3D printed parts rather than laser cut parts. I made the frame a 3D printed base of two pieces. I printed in 2 pieces due to the size limitations of my printer. The MDF was pressed into the slots of the frame. The MDF itself was handcut with a saw and cleaned up with a dremel. The holes were then drilled for the shafts and bearings. The bearings were pressed into the MDF. The 3D printed pulleys were tapped to allow them to be mounted to the shafts with a set screw. Skate wheels were used for the friction differential and mounting them to the pulley was interesting. I needed some longer M3 machine screws and had to reprint a few pulleys for the proper hole alignment. The complete assembly is shown below along with a shot of the belt tensioner. I originally thought I could do without it, but with the belt sizes, I needed the belt tension to engage the teeth better and transfer torque.
Conclusion
Below is a video of the final project. It functions pretty well. I would have assembled the bearings better and worked on the shaft alignment a little differently if I completed the project again. I would also try to use a different material or print settings for the timing belt pulleys. They were not as accurate as I would like and the PETG was a pain to dial in on my printer.
