This week, I spent a lot of time finishing up the build for our robot. First, Hirani, Ritu, and I were able to attach the motor mounts on the robot base for our pick up mechanism. Then, Hirani and I tested out the raspberry pi hat with DC motors and stepper motors with the codes we wrote over the weekend. Here, we learned that we cannot a high threshold throttle value such as 1 for DC motors. When I ran at 1, the transistor connected to M1 and M2 terminals burned. Also, the stepper motor did not work. We looked at the pin layout, but it somehow would rotate half-way and stop moving. I was trying different combinations of the pin layouts so that the motor would rotate 360 degrees, but this also burned the M3 and M4 terminals.
I assumed that the stepper motor would not be 12 volts, but when I looked at the datasheet, it seemed the motor was able to operate at 12V. Also, I figured out from the research that I need the following pin layouts:
- coil 1 should connect to one motor port (say M1 or M3) and coil 2 should connect to the other motor port (M2 or M4).
- Black wire: first pin
- Green wire: second pin
- Red wire: fourth pin
- Blue wire: Fifth pin
Meanwhile, I finished laser-cutting every board and holes for the board to mount onto the extrusions.
Additionally, on Friday, we tested out if the machine learning object classification works as intended, where we are able to detect water bottles, soda cans, and crumbled paper. We put a rubber belt on the conveyor belt system for our final step to build a pick up mechanism. Then, Hirani and I met up on the evening to test the ultrasonic sensor with the servo motor. The problem here was the ultrasonic sensors did not work as intended as one of their sensors got damaged. We were able to test it and see it worked previously, so we decided to use a new one to test it after the demo. Finally, we tried to debug issues with sshing into Raspberry pi with a static ip address. We concluded that ssh issues may come from the school’s firewall that blocks such connection. However, we figured out we can run the codes from the monitor that connects to Raspberry Pi so that we do not necessarily need ssh.
We are currently in line with our gantt chart. Burning Raspberry pi hat was unexpected, but we anticipate to connect stepper motors on Monday after demo when the raspberry pi arrives.
Therefore, my action items will be:
- Set up a Raspberry pi monitor to control the motors
- Start looking at VSLAM and debugging its installation on Jetson
- Doing the interim demo on Monday!
0 Comments