A guiding hand in engineering robotic dexterity

This diagram shows a central MPU receives power from a 24 V supply through a buck regulator and PMIC and connects to EtherCAT, USB high-speed, RS-485, and optional humanoid AI processing. The MPU communicates via CAN-FD and SPI with impedance measurement circuitry and multiple motor-control subsystems. Five single-motor boards each contain a motor-control MCU, force sensor interface, motor and gear encoder inputs, current sensing, and an HVPAK motor drive powering a three-phase motor. Five dual-motor boards provide similar control and feedback functions for dual-motor actuators. An additional dual-motor drive is controlled directly by the MPU.

Dashed lines connect all motor-control boards to a robotic hand illustration, showing that the combined system drives 17 degrees of freedom. Integrated force sensors, encoders, and current-sensing feedback enable precise motion control and dexterous manipulation.

This diagram shows power from a 12V or 24V input passes through a buck regulator and LDO to provide 5V and 3.3V supplies. A central motor controller manages two identical robotic joint control modules. Each module includes an HVPAK motor driver with gate driver, current sensor, and fault-handling circuitry. The motor controller sends three PWM control signals to the motor driver and receives three current-sampling feedback signals through ADC inputs.

Additional GPT and GPIO interfaces connect to incremental encoder feedback. Each motor driver supplies three-phase outputs, labeled U, V, and W, to a brushless motor. A Hall sensor and incremental encoder provide motor position and speed feedback.

Dashed lines connect the two motor modules to a robotic hand illustration, indicating that multiple modular joint drives can be combined to control robotic hand movements. The diagram emphasizes a scalable architecture in which a motor controller supervises individual motor-driver modules with encoder, Hall sensor, and current-sensing feedback for precise joint control.