DRV835FSV2

🐕 In-House High-Performance Quadruped Robot

📌 Project Overview

This project focuses on the design and development of a high-performance quadruped robotic dog, built entirely in-house with a strong emphasis on modularity, efficiency, and scalability.

The system is built around custom motor control electronics using the STM32G431CBU6 microcontroller, implementing advanced Field Oriented Control (FOC) and real-time CAN FD communication between joints.

Our goal is to design and develop every critical subsystem internally — including actuator electronics, firmware architecture, communication stack, and mechanical integration.

---

🎯 Objectives

• Develop high-efficiency FOC-based motor control
• Design modular smart actuators
• Implement deterministic real-time CAN FD communication
• Build scalable firmware architecture for multi-joint control
• Optimize for performance, responsiveness, and energy efficiency
• Maintain complete in-house system ownership and understanding

---

🧠 System Architecture

🔹 Control Electronics

• Microcontroller: STM32G431CBU6
• Advanced timer-based PWM generation
• Inline / low-side current sensing
• High-speed ADC sampling
• CAN FD communication interface

🔹 Motor Control

• Field Oriented Control (FOC)
• High switching frequency PWM
• Real-time current control loop
• Velocity and position control loops
• Optimized for low latency and high torque response

🔹 Communication

• CAN FD based multi-node architecture
• Deterministic joint-level communication
• Scalable bus structure for 12+ actuators
• Distributed-ready control design

🔹 Mechanical Design

• Modular actuator units
• Custom motor driver PCB
• Compact, serviceable architecture
• Strength-to-weight optimized structure

---

📂 Repository Structure

/firmware /motor_control /can_fd /drivers /middleware

/hardware /pcb_design /schematics /bom

/mechanical /cad_files /actuator_design

/docs /progress_reports /architecture_notes

---

🚀 Current Progress

• ✅ Custom motor driver PCB designed
• ✅ STM32G431 bring-up completed
• ✅ PWM generation validated
• ✅ ADC sampling verified
• ✅ Initial FOC implementation tested
• 🔄 CAN FD communication framework in progress
• 🔄 Multi-joint synchronization under development
• 🔄 Actuator mechanical refinement ongoing

---

🔜 Next Steps

• Complete closed-loop position control
• Optimize FOC loop timing and stability
• Finalize CAN FD protocol structure
• Integrate multiple actuators
• Begin quadruped gait testing
• Implement high-level motion control layer

---

🛠️ Development Tools

• STM32CubeIDE
• STM32 HAL / LL Drivers
• Custom FOC algorithms
• Oscilloscope and logic analyzer for validation
• CAN FD debugging tools

---

📈 Long-Term Vision

• Fully autonomous quadruped platform
• Advanced locomotion algorithms
• Real-time state estimation
• AI-assisted gait optimization
• Expandable research and development platform

---

🤝 Contribution

This is currently an in-house development project.
Documentation and structured releases will be improved progressively.

---

📜 License

License information will be added soon.