Lab 4: Motors and Open Loop Control

Power Supply and Oscilloscope Setup

AnalogWrite was used to generate PWM signals for the driver, with an oscilloscope connected. The external power supply was set to 3.7V, the same voltage as the 850mAh battery, for safety

Due to limited lab time, I used the oscilloscope on the motor drivers during this step, but then wired everything up all at once as shown in my diagram and left. The single and dual motor tests below were done in code with battery power rather than a bench supply.

Figure: Bench setup showing motor driver connected to external power supply and oscilloscope probes on the output.

The following snippet is the analogWrite code used to test the motor drivers:

void loop() {
  analogWrite(4, 100);
  delay(3000);
}

Figure: Oscilloscope capture showing PWM output from the motor driver.

Single Motor — Both Directions

With the car on its side and wheels elevated, the motor was tested in both forward and reverse.

analogWrite(MOTA_FWD, 128);
    analogWrite(MOTA_REV, 0);
    delay(3000);
    analogWrite(MOTA_FWD, 0);
    analogWrite(MOTA_REV, 0);
    delay(1000);
    analogWrite(MOTA_FWD, 0);
    analogWrite(MOTA_REV, 128);

Video: Single motor spinning forward and reverse with car on its side.

Both Motors — Battery Powered

The code runs both motors forward, then reverse, then spins in each direction to confirm independent control.

Video: Both wheels spinning with 850mAh battery powering the motor drivers.

Car Assembly

All components are 'installed' inside the chassis (apologies that everything is taped in place for now, a more permanent solution is in the works I promise). Motor drivers are taped near each motor to keep wire runs short and reduce EMI, with wires twisted together for that purpose. The Artemis and QWIIC breakout board sit on the other side in a tape bundle, and the 850mAh battery next to them. As required nothing sticks out past the wheels so the car can flip without damage. I would, however, not really trust it to flip.

Figure: All components secured inside the car chassis.

Lower PWM Limit

The PWM was ramped up slowly while the car sat on the ground. Starting from rest required a slightly higher PWM than keeping the robot in motion, which makes sense since static friction is higher than kinetic friction. The lower limit was 60 for forward driving and 140 for on-axis turning.

Calibration and Straight Line

I tested for drift by running the robot along a line of tape. Hilariously, the calibration factor ended up being 1.0 anyway.

Figure: Calibration testing and results.

float CAL_FACTOR = 1.0;
int left_speed = (int)(BASE_SPEED * CAL_FACTOR);
int right_speed = BASE_SPEED;
drive(left_speed, right_speed);

Video: Robot driving straight along tape line for 2m (measured in blue tape on the floor) after calibration.

Open Loop Control

I made a sequence loaded onto the Artemis: drive forward, turn right, drive forward, turn left, drive forward, turn again, then reverse. I put him infront of a wall to demnostrate the evasion behavior I was going for. It wasn't perfect though.

    drive(50, 50); // fwd
    delay(1500);
    stop();
          
    delay(300); // turn
    drive(160, -160);
    delay(1500);
    stop();
          
    delay(300); //fwd again
    drive(50, 50);
    delay(1500);
    stop();
          
    delay(300); // turn
    drive(-160, 160);
    delay(1500);
    stop();
          
    delay(300); //fwd
    drive(50, 50);
    delay(1500);
          
    drive(170, -170); //stronger turn
    delay(2000);
    drive(-80, -80); // reverse
    delay(1500);
    stop();

Video: Open loop untethered control demo showing forward driving, turns, and a spin.