1. The test pilots and their checklists

Tom Wolfe’s The Right Stuff returns as the recalled work because the early astronauts were, above all, test pilots who trusted hardware only after it had been pushed through a structured matrix of conditions. They did not ask whether a machine worked in theory; they asked whether it worked at the edge of its envelope, after a failure, and in the wrong order. A LEO desktop ADCS test should be approached with the same discipline.

This entry defines the test matrix.

2. Integration checks

Before any closed-loop test, verify that the ADCS payload is a good citizen on the bus:

  • power draw at idle, peak, and safe mode;
  • data interface: telemetry rate, command acceptance, error responses;
  • mechanical interface: alignment witness marks, stiffness, thermal path, connector locking;
  • EMC: magnetometer readings with other subsystems on and off;
  • magnetic cleanliness: residual dipole of the integrated platform.

These are pass/fail. A payload that fails integration cannot be trusted in closed loop.

3. Sensor and actuator characterization

With the hardware integrated, characterize the pieces:

  • Sun sensor: field of view, accuracy versus Sun angle, response in eclipse simulation;
  • magnetometer: comparison to a calibrated reference, sensitivity to onboard currents;
  • gyro: bias, scale factor, noise, and drift over a representative interval;
  • star tracker: acquisition time, accuracy, false-star rejection;
  • reaction wheel: torque, momentum capacity, speed ripple, current draw;
  • magnetorquer: dipole moment, orthogonality, thermal rise under duty cycle.

Characterization provides the parameters that feed the estimator and controller. Wrong parameters produce wrong pointing.

4. Closed-loop pointing tests

The core of the matrix is closed-loop behavior under realistic conditions:

  • Detumbling: release from a representative tumble and measure rate decay using B-dot;
  • Sun acquisition: transition from tumbling to Sun-pointing safe mode;
  • Nominal pointing: hold Sun, nadir, and ground-station targets and measure steady-state error;
  • Payload tracking: follow a moving target with a camera or antenna surrogate and measure tracking error;
  • Slew and settle: command large-angle slews and measure peak rate, settling time, and overshoot;
  • Disturbance rejection: apply calibrated gravity-gradient, aerodynamic, and magnetic disturbance torques and observe recovery;
  • Momentum dumping: drive wheels toward saturation and verify autonomous or commanded dumping via magnetorquers;
  • Safe mode entry: inject sensor faults and verify transition to a stable safe attitude.

Each test should be run with the expected inertia and with a perturbed inertia to simulate attachment changes.

5. Autonomous attitude management tests

Finally, exercise the autonomy layer:

  • target selection when Sun, ground station, and payload requests conflict;
  • conflict resolution with different priority and tolerance inputs;
  • response to a missed ground pass or a failed payload observation;
  • re-commissioning sequence after a simulated attachment swap;
  • fault detection and isolation when a sensor or actuator degrades.

6. What this changes

  • The ADCS test has a structured matrix covering integration, characterization, closed-loop pointing, and autonomy.
  • Tests must include both nominal and perturbed configurations to represent the desktop’s modular nature.
  • The next entry will define what success and failure look like for each test.