CE432-1 Robotics II 2021 Fall
Balancing Car project
Name: Mychael Garcia
Email: mhgarcia@fortlewis.edu

Materials:
Introduction:

For this project, we were given the task of building a 2 wheeled car that it is all self-contained and can balance on its own. This task involved 3D printing, design skills, coding, and lots and lots of testing/calibration.

Process:

This project involved everything that we’ve been doing for the past few weeks, how to control a stepper motor, get data from the MPU 6050, and have a code that times how long it takes for it to run and use that in calculations. The first step was designing a chassis that all the components would attach to. This was done by cutting an acrylic sheet into 3 8"x3" sections and drilling holes in each corner. Using a long-threaded rod, this was cut in to 4 equal lengths roughly 10" long. Using nuts, each rod was placed into a corresponding hole and secured. The end result was a three-level frame that could be used to attach other components too.
chassis

The second process was finding a way to mount the 2 stepper motors to the chassis. This was done by 3D printing a bracket that would wrap around the motor and have two screw holes to support the motor so it would not slide around. The printer used was the Prusa i3 MK3S+ talked about in a previous project. The STL file for these brackets can be found here.
topside

The third part also involved 3D printing. We need a way to adapt the output shaft of the stepper motor to the wheels used in the project. The wheels had a hexagon shaped hole that need to be mounted to the keyed shafter of the stepper motor. To do this a small cylindrical part was made that connected both pieces. The STL file for these adapters can be found here.
tiretopside

Fourth was a bracket to mount the breadboard and Arduino to the chassis. This was also done using the 3D printer. The part had small overhang and holes that could mount directly to the chassis frame. The STL file for these mounts can be found here.
arduino mount

The fifth part was putting together and programming the whole thing. This was the hardest part and took the most time. The robot used an Arduino Uno to control the device, it would receive positioning data from the MPU 6050 and in return send a signal to the motor controllers to direct them on which way to turn to keep the car upright. The code for this project can be found here.
carwireing

After everything was connected, we started programing it. This was the most time-consuming part. After calibrating the MPU 6050, we had to calibrate the PID controller. This is a rough calculation of the error or change in angle of the car from its equilibrium state. Each error equation had its own value, P=60, I=0, D=10. These values change how quickly the motors accelerate to maximum speed. These values worked fairly well but one con was that the car is extremely twitchy and tends to fall over easily with relatively small angle changes.


After everything was running using the breadbaord the plan was to transfer everything over to a custom PCB. As this was done before the cars were running some initial guesses were made in how to connect the MPU-6050 chip and power the Arduino chip. As the PCB's take time to get made and shipped we had to make them at the time we did. Knowing everything I do now i would make changes to the PCB so it would be easier to use.
PCBscamatic

We also planned on adding a remote to control the cars direction and rotation. This was done by a partner in the class (Humberto Perez). It worked well there was just not enough time to implement it into the code as calibrating the car took longer than expected.
controlerc scamatic

Conclusion:

This was a good final project; you learn a lot about timing and more advance parts of programing. Using the breadboard as a test run to make sure everything is working then switching over to a PCB is a good idea, it does make it a bit harder to calibrate as you have to use a second Arduino to get readouts of the first one. This could be improved by adding a USB transcoder, but this would increase the wiring and another point of failure. I wasn’t able to get the PCB and remote working within time, but I was able to get the car to balance.

----------------------------------------------------------------------------------------------------

END