I managed to build out about 90% of what I thought would be the moving portion of the Fern. I built custom syringe holders to control the cylinders and mounted all of it on a basic plywood base. I learned a few things from this prototype that is making me rethink the final iteration of this.
The stepper motors I’m using are small bipolar steppers mounted to a metal bracket with a threaded rod. The assembly proved to be fairly strong and the perfect size. Unfortunately, the motors are getting very hot, so I need to consider some kind of heat sink or cooling solution.
Another challenge I’m working through is that the motors have no way of being zeroed. Some of the gantry systems that use stepper motors we worked with require the user to push the main element to one corner which allows the gantry to know where it’s beginning. I can’t quite do that with this setup because everything is on threaded rods, which means that only the motor can move push the cylinders and not the other way around. One possible fix that I’m looking into is to use a slide potentiometer to give a location reading. Now that I’m using an Arduino Mega board, I’m fine sending much more data to it.
This version still has no photoresistors installed on it. I made a feeble attempt to incorporate it in the early hours this morning, but the serial communication was starting to fail on me. I’m looking into potentially scrapping serial communication and switching to something a bit more robust like running a node server. I’m not quite sure how to go about doing this, but this will make it easier for me to focus on fleshing out the neural network elements to this project without relying on shaky serial communication. At the moment, the video below shows it being controlled by a potentiometer.
I’m not yet satisfied with the look of this base. I’m hoping to clean this up and provide a more attractive enclosure that will also protect the electronics. Something waterproof would probably be smart.
I may want to retool a few aspects of the fabrication for the next iteration. The base is still lacking in my mind, and I haven’t thought through a proper enclosure strategy for the controller board. I need to consider how to attach the photoresistors to the Fern branch. Finally, I want to update the communication method to something that is more reliable and easier to debug than serial communication. Suggestions are welcome!
Some additional fabrication photos can be seen below.