Week 18 November 7-13

Power/Mechanical modules

Over the weekend, the new drivers were built and connected as planned to the terminal blocks for power and had the data connections from control module connected as well. Last week it was described how the power module would be wired so that any heavy current being pulled by the motors and motor drivers should only be experienced by them and not the smaller electronics in the circuit. This was done after building the motor drivers, as well as checking the voltages across all components once every potential load was introduced to the battery at the same time. This was important because it was necessary to have a fixed voltage of below 12V to supply power from the battery to control module, and a potentiometer was going to be used for this. Not having motors and drivers all connected while adjusting that resistance would not give an accurate reading of the final output voltage, but once they all were connected the right resistance was found to balance all power sources out.

Before finally connecting everything to power on, all connections were verified the proper voltages were going through, and manual control was once again tested using the direction functions from the app. This time, all four wheel motors managed to be powered and could be controlled by a user with the phone app, and the front, left, right, back, and stop commands were all performed as requested by user. Figure 1 shows the wheels in motion. PWM value was set at 50% of its full potential duty cycle as to not over-power the wheels, as this was also considered a reason the motor drivers were damaged, and it seems at the moment this would be enough speed to move the cart around an area, but verification of this will need to be done once the wheels and pillow blocks are ready and mounted. Also, verification of the voltage divider being able to perform its operation needed to be done as it was not connected right at the time the wheels were controlled due to the wires needing to be longer, but this can be done once the right wiring is done.

Figure 1: First test of wheel control with new motor drivers

Another measure was put in place between battery and the terminal blocks to help prevent an over-drawing of current or power with fuses between the battery terminals and blocks. This is done in case there is a sudden surge of current for whatever reason passing through the wires, the drivers and other hardware would be safe as the fuse would break and open the line. This was experienced when doing initial manual control tests; when first trying to move the wheels, the line between terminal and small electronics were fine and powered, but when the motors were active, they only ran for a second and power shut off. This killed the 2A fuse going to the motor terminal block, and so other fuses were tested to see if they could handle the current. While it was not confirmed the exact current being drawn with all systems working, a 10A and 25A fuse was tested to see what could handle the motors’ demand, and it seems a current between 25A and 10A is being pulled from the battery by all four motors. This is due to the 10A fuse blowing when running but then the 25A fuse controlled all motors and direction changes without being damaged.

Now that manual control is verified up to the motors, the next step is to mount the wheels and test on the ground manual control to ensure all systems can move the cart safely, meanwhile autonomous software and sensor detection will be worked on as this is done.

Control module

Monday began with implementing the code used to avoid obstacles when detected. This would be a first version to load onto the cart’s Primary controller and see if all scenarios of obstacle avoidance can be covered with this code. As long as there is no out of bounds detected by the floor line sensor, the cart’s main priority is to ensure the front sensors are cleared and move forward if faced with obstacles. If they are not, then a process of elimination will be done to see which sides have obstacles and react accordingly. If at least one on each side of the sensors has an object detected, the cart would stop and wait for clearance. If there is a line detected from an out of bounds area, the cart would act accordingly to that as well based on what direction it was moving with respect to the user.

Sensor module

Work began on finalizing the proximity sensor building such as soldering and wiring. All proximity sensors were tested that they all work after putting them together and are now ready to be mounted to the cart. Wiring will be designed throughout the week to make sure all power and data connections are made available for the sensors. The line sensor was tested to see how general operation goes. It seems it can differentiate between easily contrasting surfaces, but it will need to be tested with some other colors or surfaces to see what is a good level that can work on mostly all floor types. The sensor has an adjusting level on board that increases or decreases the sensitivity of its line detection, so it will need to be adjusted when tested on the cart.

Input module

Added code to the app that now has it require entering a pin before allowing access to controlling the cart.

Figure 2: Password screen that appears before the main screen that requires a pin for access.