The Dynamic PID Tuning Platform is a user-friendly test rig built to make PID tuning easier for robotics projects. It’s specifically designed for tuning the PID settings of DJI Robomaster M3508 P19 motors, which are controlled using C620 ESCs via the CAN protocol.
With this platform, you can easily adjust PID parameters and monitor real-time feedback on setpoints, motor speed (RPM), and control outputs. The data is displayed on an onboard LCD and can also be visualized using an external Python-based software for more detailed analysis.
Watch this video to see it in action:
Dynamic.PID.Tuning.Platform.mp4
- Real-time PID Tuning: Adjust P, I, and D values using rotary encoders.
- CAN Bus Communication: Sends control data to Robomaster C620 ESCs.
- Live Monitoring: Displays motor speed, setpoint, and PID outputs on a 20x4 LCD display.
- External Visualization: Outputs CSV-formatted data to a Python-based GUI for graphing and analysis.
- At this time, the code supports controlling only one motor at a time.
- Freenove ESP32 Dev Board (or any ESP32 board)
- Robomaster C620 ESC and M3508 P19 Motor
- MCP2515 CAN module
- KY-040 Rotary Encoders (4x for P, I, D, and setpoint)
- 20x4 LCD Display with I2C Backpack (PCF8574)
For easier connections, it is recommended to use a Freenove ESP32 breakout board. You can also build the circuit on a breadboard if necessary. Using an external 5V power source for peripherals is highly recommended.
| Component | Pin on Component | Pin on ESP32 |
|---|---|---|
| MCP2515 CAN Module | CS | GPIO 5 |
| INT | GPIO 15 | |
| KY-040 Encoder (P) | CLK | GPIO 35 |
| DT | GPIO 32 | |
| KY-040 Encoder (I) | CLK | GPIO 25 |
| DT | GPIO 33 | |
| KY-040 Encoder (D) | CLK | GPIO 27 |
| DT | GPIO 26 | |
| KY-040 Encoder (Setpoint) | CLK | GPIO 39 |
| DT | GPIO 34 | |
| 20x4 LCD Display with I2C | SDA | GPIO 21 |
| SCL | GPIO 22 |
- Written in Arduino framework using PlatformIO.
- Implements PID control logic with:
- Adjustable P, I, D values using rotary encoders.
- Real-time CAN communication with the C620 ESC.
- CSV-formatted serial data output for visualization.
- Plots real-time graphs of:
- Motor RPM
- Setpoint
- PID output
- Provides GUI controls for serial port selection.
-
Hardware Setup:
- Connect all components as per the pin connections table.
- Use an external 5V power source for the MCP2515 module and the LCD to ensure stability.
-
Software Setup:
- Install PlatformIO extension on Visual Studio Code if you haven't already.
- Clone this repository:
git clone https://github.com/ShayonKhaled/Dynamic-PID-Tuning-Platform.git
- Open the project in PlatformIO and upload the firmware to your ESP32.
-
Python Visualization:
- Install required Python libraries:
pip install matplotlib pyserial tkinter
- Run the Python script for real-time graphing:
python3 pidVisualizer.py
- Select the serial port of your ESP32 from the drop down menu and hit the connect button. The ESP32 must be connected to your computer.
- Install required Python libraries:
- Power Supply: Avoid drawing power for the MCP2515 and LCD display directly from the ESP32. Use a separate 5V supply to prevent voltage drops and ensure reliable operation.
- Freenove ESP32 Breakout Board: The breakout board simplifies wiring but is not mandatory.
To house the components and provide a modular design, I used 4X 250mm DIN rails to build the test rack. The stands for the rails were designed by PlasticXYZ and can be found here: DIN Rail Test Rack Stands by PlasticXYZ.
I designed custom mounts for the following components:
- Freenove ESP32 Breakout Board
- MCP2515 CAN Transceiver Module
- Rotary Encoders
- 20x4 LCD
These designs are included in the "3D Printables" folder in both .STL and .STEP formats. You can choose to:
- Bolt the mounts directly onto the DIN rail, or
- Use DIN clips (designed by PlasticXYZ) to attach them. The DIN clips can be mounted or unmounted easily from the DIN rails. You can find the DIN clips design here: DIN Rail Clip by PlasticXYZ.
- Mounting Holes: 18mm apart. These can be modified using the provided
.STEPfiles if needed. - Fasteners Used:
- Modules: Mounted with self-tapping screws.
- DIN Clips: Mounted with M3 bolts (6mm length).
- No Heat Press Inserts: The mounts were specifically designed for simplicity, and all components are secured with self-tapping screws or bolts.
This setup provides a sturdy, modular, and customizable framework for the PID tuning platform, making it easy to assemble, modify, and expand as needed.
This project is licensed under the GNU General Public License v3.0 License. See the LICENSE file for details.