# Lesson 6: PID Loops (Feedforward for Friction Compensation)

Teachers will guide students through implementing a PID controller with feedforward to handle frictional resistance in the XRP robot’s drivetrain. Students will learn to fine-tune PID constants and use feedforward terms to improve motion smoothness.

• Completion of Lesson 5: PID Loops - Gyro Turn.
• Robot code that can already perform basic PID turning using the gyro.

This lesson builds on PID control by showing how feedforward compensation can offset predictable frictional forces that cause sluggish or inconsistent performance. This mirrors real-world FRC control strategies for mechanisms like drivetrains and arms.

Feedforward adds a predictable control term that helps the PID controller overcome known system resistance.

Formula:

Output = kP * error + kI * errorSum + kD * errorRate + kS + kV * velocity

Where:

• kS is the static friction compensation (voltage needed to overcome stiction).

Explain that while PID corrects errors, feedforward anticipates the required effort to move smoothly.

Use this concept to improve the gyro turn from Lesson 5.

public class GyroTurnCommand extends CommandBase {
    private final Drivetrain drivetrain;
    private final ADXRS450_Gyro gyro = new ADXRS450_Gyro();
    private final double targetAngle;
    private final PIDController pid = new PIDController(0.02, 0.0, 0.001);

    // Feedforward constants
    private final double kS = 0.15; // static friction voltage
    private final double kV = 0.0;  // optional if you want velocity scaling

    public GyroTurnCommand(Drivetrain drivetrain, double angleDegrees) {
        this.drivetrain = drivetrain;
        this.targetAngle = angleDegrees;
        addRequirements(drivetrain);
        pid.setTolerance(2.0);
    }

    @Override
    public void execute() {
        double error = targetAngle - gyro.getAngle();
        double pidOutput = pid.calculate(gyro.getAngle(), targetAngle);

        // Add static friction feedforward
        double output = pidOutput;
        if (Math.abs(pidOutput) > 0.01) {
            output += Math.copySign(kS, pidOutput);
        }

        drivetrain.arcadeDrive(0, output);
    }

    @Override
    public boolean isFinished() {
        return pid.atSetpoint();
    }

    @Override
    public void end(boolean interrupted) {
        drivetrain.arcadeDrive(0, 0);
    }
}

1. Start with feedforward only (kS). Increase until the robot starts moving smoothly at low speeds, then turn it down so any PID output will move it.
2. Then add PID tuning from Lesson 5 to refine accuracy.
3. Observe how feedforward reduces overshoot and startup lag.

• Have students compare turns with and without feedforward.
• Record both runs using AdvantageScope (next lesson) to visualize PID output and angle tracking.

• How does friction affect turning precision?
• Why is static feedforward especially useful on small robots like the XRP?
• What other mechanisms (like arms or elevators) might need feedforward in FRC?

60–75 minutes (including testing and tuning)

• Understand feedforward’s role in overcoming system resistance.
• Apply friction compensation to improve control smoothness.
• Tune combined PID + feedforward loops for reliable motion.