Skip to main content

Module 3 — Controls & Feedback

Objective: add sensing and closed-loop control.

Prereqs: Command-based project with subsystems.

Steps

  • Configure encoders/gyros; read positions/velocities.
  • Implement PID (position/velocity) and feedforward; tune gains.
  • Hold setpoints on a mechanism (arm/flywheel).
  • Add motion profiling or slew limits to reduce shocks.
  • Verify sensor phase/direction before tuning; protect against runaway by clamping outputs.

Deliverables

  • Mechanism that holds a setpoint within tolerance.
  • Logged tuning session (gains, observed response, final values).

Code example (simple PID)

PIDController pid = new PIDController(0.08, 0, 0);
double output = pid.calculate(arm.getPosition(), targetDegrees);
arm.setVoltage(output);

Resources

  • WPILib PID/FF examples: docs.wpilib.org/en/stable/docs/software/advanced-controls/introduction/index.html
  • CTRE: v5.docs.ctr-electronics.com
  • REV: docs.revrobotics.com

Instructions (numbered)

  1. Configure sensors (encoders/gyros); verify direction and units.
  2. Implement PID/FF; clamp outputs; add slew limits if needed.
  3. Tune gains (P first, then D/FF); record responses and final values.
  4. Test holding and moving to setpoints; measure error and stability.
  5. Log tuning session and results.

Example

  • Arm PID: kP 0.08, kD 0.001, kS/kG/kV from characterization; holds ±2° at setpoint.

Best practices

  • Verify sensor phase before tuning; safe defaults/clamps to avoid runaway.
  • Tune incrementally; log data for each change.
  • Use characterization tools for FF when possible.

Common mistakes

  • Wrong units/phase; no clamping leading to runaway.
  • Tuning without logging; guessing gains.
  • Ignoring load changes (gravity/CG shifts).

Spec notes / data to log

  • Sensors: units/phase verified; sample rates.
  • Controls: kP/kI/kD/kS/kG/kV values; clamps/slew limits.
  • Data: log setpoint vs actual, error, voltage/current during tuning.

Checklist

  • Sensors verified (units/phase)
  • PID/FF implemented with clamps/slew limits
  • Gains tuned and logged
  • Setpoint hold/move tested
  • Final gains recorded

Recommended tools

  • WPILib SysId/characterization, logging (AdvantageKit or built-in), Sim GUI for dry runs.

Sample log (template)

  • Date:
  • Mechanism:
  • Gains tried/results:
  • Final gains:
  • Notes/issues:

Photos/diagrams

  • [Placeholder: plot of setpoint vs actual with tuned gains]

Numeric Example​

Example drivetrain velocity PID loop:

  • Input: wheel speed (m/s)
  • Output: motor command (volts)
  • kP: 0.12 V / (m/s)
  • kI: 0.00 V / (m/s·s)
  • kD: 0.001 V / (m/s²)
  • kS: 0.13 V
  • kV: 1.41 V / (m/s)
  • kA: 0.09 V / (m/s²)

Expected behavior:

  • Steady-state error: < 5% of target speed
  • Overshoot: < 10% for step changes in setpoint

Data to Log​

At minimum, log the following for any closed-loop test:

  • Timestamp
  • Commanded setpoint (m/s, deg/s, etc.)
  • Measured value (same units as setpoint)
  • Controller output (volts or percent)
  • Battery voltage
  • Motor current (A)
  • Error (setpoint - measured)
  • Any fault codes or limit events

Purpose:

  • Validate tuning (check overshoot, settle time, and steady-state error)
  • Detect brownouts and current spikes
  • Compare different tuning runs over time