As a summer project, I decided to design and build a programmable camera dolly in order to take time-lapse footage. An Arduino microcontroller is the "brain" of the dolly and runs a medium servo and two stepper motors through a pair of Big Easy drivers. The steppers are both NEMA 17's and have 400 steps per revolution. The stepper that controls lateral movement moves the dolly at 0.1 mm per step while the stepper dictating rotational movement turns the camera 0.25 degrees per step under half-step mode. The whole system is held together by custom 3D printed parts I designed in Solidworks, two 3/4 in plastic pipes, and a few shafts, bearings, and belts. The LCD screen displays a simple interface that can be used to change the camera's path before it takes the photos that are stitched together to make a time-lapse film.
From this project, I learned a great deal about stepper motors, LCD interfacing, coding, and 3D printing. The entire process allowed me to discover the many caveats that become apparent as things start or stop working for no apparent reason.
There’s a minor problem with the rig. The servo doesn’t keep the camera in position 100% of the time. I’m not sure if it’s because it’s not receiving enough power through the extensive wiring or it’s just too weak. Either way, it’s a bit of a problem because when it does draw too much energy the Arduino resets.
Here are a few things I learned through the process:
3D printing
- To avoid bolts and screws. If it’s not a tank, there’s no need to design it like one.
- To use epoxy. Better yet, print everything in one piece if you can while still maintaining every parts’ integrity
- How to take into account printing tolerances along the different axes
Design
- Making sure every detail is worked out in Solidworks. It prevents things from getting messy later.
Coding
- LCD interfacing
- The AccelStepper library
For the next iteration:
-No bolts or screws
-Everything down to the wiring will be modeled in Solidworks
-Less wiring
-An infinite degree of rotational freedom around the z-axis
Update: The problem with the servo was simply that I was running it directly off the Arduino so the amperage was too low. This has been fixed with a UBEC that provides enough current and correct voltage. I finished CADding the the structural components of the camera dolly; all that's left is the wiring, which I've never done before, so it should be fun. Here are a couple pictures:
Update: The problem with the servo was simply that I was running it directly off the Arduino so the amperage was too low. This has been fixed with a UBEC that provides enough current and correct voltage. I finished CADding the the structural components of the camera dolly; all that's left is the wiring, which I've never done before, so it should be fun. Here are a couple pictures:
Many thanks to Prof. Silvan Linn at San Francisco State University for his guidance and support.
