logs
2026-08 Servo Tuning Session
Measured jitter across PWM frequencies and picked a trim strategy.
robotics
Spent the evening sweeping PWM frequencies against the neck servo. Jitter drops by
half at 333 Hz but the driver overheats above 400 Hz. Settled on 333 Hz with a
software deadband of two counts.
The procedure used is written up permanently in servo-calibration, and the
session belongs to the projects/reachy-mini journal. Next session will repeat the
sweep after thermal soak, then fold results into first-motion-milestone.
Read by