MODULE M7 · 15.0 HOURS

Firmware Development and Control

0% complete

LEARNING OBJECTIVES

Module objectives

  1. Understand DYNAMIXEL smart actuator communication and control framework.
  2. Implement actuator and FSR sensor signal acquisition using OpenCR control board.
  3. Design real-time robot control state machine and closed-loop feedback loop.
  4. Program safe power management and torque release sequences.

Firmware Architecture and DYNAMIXEL Control

The robotic hand firmware acquires sensor data and processes actuator commands within a high-speed loop. The OpenCR 1.0 controller is based on a 216MHz ARM Cortex-M7 processor [S13] and processes DYNAMIXEL protocol 2.0 without a separate bridge [S11], minimizing latency. Each actuator supports current, velocity, and position modes, and the robotic hand uses a current-control-based torque grip strategy.

FSR Force Feedback System

The FSR 402 sensor has a resistance characteristic inversely proportional to the force applied [S12]. A voltage divider circuit is configured with a 10kΩ resistor on the 3.3V sensor rail using OpenCR’s 12-bit ADC [S13]. The divided voltage is normalized via ADC value = (V_in * R_fsr) / (R_fsr + R_ref), and this value is linked with the tendon tension of the fingers and used as grip force feedback.

Safe Control Routine

System stop is divided into two stages for safety. In the software stage, actuator torque is released (Torque Off) to immediately remove physical driving force. Before maintenance, you must physically disconnect the 3 independent power adapters and verify that all branches are less than 1V using multimeter DC mode.

WORKED EXAMPLES

Worked examples

  1. 1. Setting actuator goal position/current: An example of using DYNAMIXEL SDK to set current limit (Goal Current) for XM430 actuator and updating finger joint final position via a PID loop based on sensor values.
  2. 2. FSR voltage data filtering: Code implementation for applying a Moving Average Filter to remove noise from raw data collected from ADC and normalizing the range with upper (20N) and lower (0.2N) limits [S12].

LAB PROTOCOL

Robotic Hand Integrated Control and Precision Grip Practice

  1. 1

    Verify that the voltage of each independent branch is less than 1V using multimeter DC mode before starting assembly.

  2. 2

    Solder FSR voltage divider circuit to OpenCR's 3.3V sensor rail and connect to ADC port.

  3. 3

    Scan IDs of 11 actuators using DYNAMIXEL SDK and set initial positions.

  4. 4

    Test finger joint drive commands in no-load state and adjust tendon elongation and tension.

  5. 5

    Visualize FSR sensor data via serial monitor and tune grip force response.

Safety check
  • Never use 5V or 12V actuator power as supply for FSR sensor circuit.
  • Never approach the operating range of fingers while system is energized; use fixtures.
  • Never connect positive (+) terminals of power branch adapters to each other.
  • Before maintenance/assembly access, physically disconnect the 3 power adapters and verify that it is less than 1V in all branches by measurement.

Lab deliverables

  • Firmware source code including real-time sensor data feedback
  • Data sheet for normalization and calibration of voltage divider data
  • Actuator feedback loop normal operation logs

ASSIGNMENT

Grip State Machine Design and Implementation

Deliverables

Rubric

  • Is the current limit range (0-2.3A) according to sensor values stably controlled?
  • Does the physical tension dissipate immediately upon the torque release command?
  • Is a safe hardware disconnection procedure explicitly stated in the code?

KNOWLEDGE CHECK

Knowledge check

1What is the recommended power rail when configuring an FSR 402 sensor and a voltage divider circuit?
2What is the correct way to confirm that the system is in a 'de-energized state' when performing robot hand maintenance?
3Is it permissible to connect the positive (+) terminals of multiple independent power adapter outputs in parallel?

FIELD CHECK

Completion criteria

  • Multimeter verification completed that independent power supply and fuse protection for each branch are configured according to BOM specifications
  • Confirmed precision force signal acquisition and filtering of 5 FSR sensors via OpenCR ADC
  • Successfully performed the software torque release routine and post-physical power disconnection measurement procedure
  • The grasping state machine processes actuator and sensor data as intended, and the final report is submitted

MODULE SOURCES

Module sources

  1. OpenCR 1.0emanual.robotis.com · technical_documentation
  2. XM430-W350 e-Manualemanual.robotis.com · datasheet
  3. FSR Model 402interlinkelectronics.com · datasheet