How to build myCobot 280 Pi (myCobot 280 Raspberry Pi 2023): the whole process
139 steps to build myCobot 280 Pi (myCobot 280 Raspberry Pi 2023), from buying parts, printing and assembly to software, calibration and training, with the bill of mater…
myCobot 280 Pi 2023 alone $799 (Elephant Robotics international store, 2026‑10‑31; shipping, applicable taxes and tariffs extra); with a base from $858, with a flat base and adaptive gripper $978. The hardware is closed, so there is no BOM cost
Assemble
- Place the 280PI in the Package
Refer to the picture and manually position the 280PI in a W shape.
- Use USB Audio Input and Output
The Raspberry Pi 3.5mm headphone jack does not support audio input. Use a USB interface microphone or headphones for audio input and output.
- Avoid Occupied Pins
Before wiring or adding hardware, confirm that the selected GPIO pins are not occupied. Occupied pins on the 280JN include GPIO0, GPIO1, GPIO3, etc.; consult the GPIO description in the corresponding robot arm Gitbook material for details.
- Connect the devices
Connect the MyCobot and handle to the computer.
- Connect and turn on the handle
Before running the program, connect the handle to myAGVPro and turn on the handle.
Software
- Flash the Latest atomMain Firmware
Use Atom to flash the latest version of atomMain.
- Create and Save the Test Script
Create a new Python file on the desktop, copy the provided code into it, and save it; the file must be named rasp_mycobot_test_gui.py. The code is also available from the Github download address in the document.
- Add Atom light-board test buttons
Create “Set Red” and “Set Green” buttons that call send_color("red") and send_color("green"), respectively, and place both on row 7.
- Add aging-cycle controls
Create “start” and “stop” buttons that call start_aging_test and stop_aging_test, respectively, and display them on row 9.
- Add motor-release button
Create a “Relax all motors” button that calls release_mycobot when clicked and place it on row 10.
- Add PID calibration button
Create a “Calibrate pids” button that calls rectify_mycobot when clicked and place it on row 10.
- Add scrolling log window
Create a log text widget with width 100 and height 35, connect a vertical scrollbar, and place the scrollbar on the text widget’s right side.
- Run the graphical interface
Call self.win.mainloop() to start and keep the graphical interface window running.
- Launch the myCobot test interface
Open the terminal, change to the desktop directory, and run rasp_mycobot_test_gui.py.
- Flash Atom and Transponder Firmware
Flash the `Atom` firmware into the top Atom and the `Transponder` firmware into the base Basic. Download the firmware from myCobot Software or use myStudio to flash it.
- Install and Upgrade with Pip
Install the `pymycobot` Python API with Pip and upgrade it to an available version.
- Install the API from Source
Clone the `pymycobot` develop branch and enter the `pymycobot` directory. Install it using Python 2, Python 3, or the modern Pip form.
- Uninstall pymycobot
Use pip to uninstall the pymycobot package installed from the current directory.
- Import MyCobot 280 Interfaces
Import MyCobot280 and MyCobot280Socket for the MyCobot 280.
- Select the Model Interface
Starting with v3.6.0, interfaces are differentiated by robot model; use the MyCobot 280 API documentation.
- Install required libraries
Open the terminal, switch to this folder, and run the following command.
- Connect the devices
Connect the MyCobot and the gamepad to the computer.
- Install required packages
Open a terminal, change to this folder, and run the command that installs the dependencies listed in requirements.txt.
- Set port and baud rate
Edit myCobot280_handle_control.py and change /dev/ttyAMA0 and 1000000 to the actual port and baud rate detected by the computer.
- Run the controller program
Run the myCobot280_handle_control.py program.
- Download the Matching Atom Firmware
Check the Gitbook section and use myStudio to download the matching version of Atom firmware. The document recommends downloading the latest version.
- Use GPIO on PI/JN Models
Use the source code shown in the document with 280JN as the example. For other models, change the device serial port information.
- Select the Communication Interface
Microprocessor-based robot arms support socket communication over TCP. Microcontroller-based robot arms can communicate through USB to serial port.
- Clone the pymycobot repository
Clone the pymycobot repository on the myAGVPro system.
- Enter the composite kit directory
Enter the myAGVPro_Composite_Kit directory in the pymycobot repository.
- Install project dependencies
Install the dependencies listed in requirement.txt from the composite kit directory.
- Run the myCobot test tool
Run rasp_mycobot_test_gui.py to calibrate the robot arm and detect the connection.
- Initialize 40PIN GPIO outputs
On the Raspberry Pi, import RPi.GPIO, use BCM mode, and configure GPIO 20 and GPIO 21 as outputs.
- Create the TeachingTest Class
Create TeachingTest with a mycobot instance; initialization reads the terminal size, sets the robot reference, and configures the record.txt path.
- Run the Drag Teaching Script
Open the terminal with the CTRL+ALT+t shortcut, enter the Desktop directory, and run drag_trial_teaching.py.
- Burn the latest atomMain firmware
First have the Atom burn the latest version of the atomMain firmware.
- Create and save the Python test script
Create a new .py file on the desktop, copy the document's code into it and save it.
- Install Project Dependencies
Install the dependencies listed in requirement.txt before running the control example.
- Run the Control Program
Run main.py in the project directory to start the myArm M&C control example.
- Restart Raspberry Pi Bluetooth
On Ubuntu 20.04, restart the Bluetooth service first.
- Run Bluetooth server
Run the uart_peripheral_serial.py script to start the Bluetooth server.
- Install Latest Firmware
Use the latest firmware for myBuddy.
Calibrate and test
- Select port and connect robot
Read the selected serial port and baud rate, convert the baud rate to an integer, and create the connection with MyCobot(port, baud); log a prompt when the port or baud rate is missing, and log “connection succeeded !” after connecting.
- Start cyclic aging test
Starts a daemon thread that runs _aging_test and records “Start cyclic aging test ...”.
- Stop aging test and clean up
Stops aging_test.service, removes aging_test.sh, aging_test.py, and the systemd service file, then reloads systemd. It logs success or failure.
- Rectify servo PID data
With myCobot connected, sets register 24 to 0 and registers 26 and 27 to 3 on servos 1 through 6, pausing 0.1 seconds between operations. It then reads the three register values and writes them to the log.
- Check myCobot connection
Checks whether a myCobot object exists; when disconnected it records “Mycobot is not connected yet!!!” and returns False, otherwise it returns True.
- Run servo aging motion sequence
Sets the LED to red, then moves the six joints through zero, 170°, -170°, zero, 90°, -90°, and zero positions. All movements use speed 95, with waits of 5, 3, 3, 3, 3, 3, and 3 seconds.
- Run First High-Speed Joint Calibration
Run the first five angle commands in order with speed 95. Each command uses wait(3) and sends the specified six-axis angle array.
- Continue Fifth and Sixth Axis Calibration
Continue with five angle commands at speed 95. The sequence includes returning to zero, fifth-axis angles of plus and minus 165 degrees, and a sixth-axis angle of 180 degrees.
- Switch to Green Medium-Speed Calibration
First move the sixth axis to -180 degrees and set the light to green. The following angle commands use speed 55 and begin first-axis movement at plus and minus 170 degrees.
- Calibrate First and Second Axes
At speed 55, run the commands for zero, first-axis angles of plus and minus 170 degrees, second-axis angles of plus and minus 90 degrees, and zero again.
- Calibrate Third and Fourth Axes
Continue at speed 55 with second-axis return to zero, third-axis angles of plus and minus 140 degrees, another zero, and fourth-axis angles of plus and minus 130 degrees.
- Calibrate Fourth, Fifth, and Sixth Axes
At speed 55, execute fourth-axis angles of plus and minus 130 degrees, zero, fifth-axis angles of plus and minus 165 degrees, zero, and a sixth-axis angle of 180 degrees.
- Switch to Blue Low-Speed Calibration
After returning from fifth-axis angles of plus and minus 165 degrees, move the sixth axis to 180 degrees. Then set the light to blue and begin zero and first-axis 170-degree movement at speed 15.
- Low-Speed First and Second Axis Calibration
At speed 15, execute first-axis angles of plus and minus 170 degrees, return to zero, second-axis angles of plus and minus 90 degrees, and return to zero again.
- Low-Speed Third and Fourth Axis Calibration
Continue at speed 15 with return to zero, third-axis angles of plus and minus 140 degrees, return to zero, a fourth-axis angle of 130 degrees, and a fourth-axis angle of -110 degrees.
- Connect robot and select baud rate
Select the corresponding baud rate and click Connect; use 1000000 for myCobot 280-Pi and 115200 for myCobot 320-Pi.
- Perform arm zero-position calibration
Manually move each robot-arm joint to its zero position, then click Start Calibration. Calibration is complete when Calibration End appears in the interface.
- Modify the port number
Edit the myCobot280_handle_control.py file and change the port number and rate to the values detected by your computer.
- Hold Left 1 to initialize zero
Press and hold **Left 1** for 2 seconds to initialize the robot to the joint zero state.
- Hold Right 1 for initial point
Press and hold **Right 1** for 2 seconds to initialize the robot and move it to the initial point.
- Reach the initial position first
After the program starts, click the Right 1 button first; other operations can be performed after the robot reaches the initial position.
- Initialize to joint zero
Long-press the Left 1 button for 2 seconds to initialize the robot to the joint-zero position.
- Stop torque and relax joints
Long-press the Left 2 button for 2 seconds to stop the robot's torque output and relax all joints.
- Initialize to movement start point
Long-press the Right 1 button for 2 seconds to initialize the robot to the movement initial position.
- Enable torque and lock joints
Long-press the Right 2 button for 2 seconds to enable the robot's torque output and lock all joints.
- Adjust PID to Optimize Joints
Adjust the PID using the linked document to address joint jitter, excessive joint angle deviation, or joint weakness.
- Calibrate the Arm at Zero Position
Calibrate the robot arm at its zero position using Chapter 5 of the Gitbook or the linked calibration document.
- Manually Check for Joint Gaps
For machines used for more than 3 months, follow the linked video to manually bend the joints and check for joint gaps.
- Identify the Joint Zero Marker
The joint zero point is the arched groove between the joint and the edge of the joint shell.
- Confirm the Calibrated Zero Posture
After calibration, the robot arm should have the general zero-point posture shown in the document.
- Confirm the 270 Joint Zero Posture
Pay special attention to the zero-position posture of the 270 joint shown in the document.
- Perform Zero-Position Calibration
Perform zero-position calibration using Chapter 5 of the Gitbook or the linked document.
- Configure MyCobot Pro630 network
When MyCobotPro630 is selected, control uses TCP rather than a serial port. Set COMPOSITE_KIT_HOST to the robot IP address and COMPOSITE_KIT_PORT to the port number, whose default is 5001. The robot must be on the same LAN as MyAGVPro, and the IP address and port can be viewed in the built-in roboflow software.
- Note MyArmM750 suction limitation
The myArmM750 robot arm does not support suction pumps, and the program will display a prompt when it is used.
- Run the connection check tool
Name the file myCobot_test_demo_CN.py; once run, the window lets you pick a serial port (baud list 1000000 and 115200, second selected by default), press connect to connect, press Start detection to detect the servos, test the Atom Light Board with Set Red / Set Green, and read the log output.
- Open Serial Ports in Order
First open the serial port connection of myArmM, then open the serial port connection of myArmC.
- Review Demo Test Files
Review the test case files in the demo directory and check the interfaces provided by pymycobot.
Data and training
- Start Joint Motion Recording
Calling record() clears record_list, enables recording, releases the servos, and starts a background recording thread. MyCobot320 uses release_all_servos(1), while other models use release_all_servos().
- Collect Encoder and Servo Speeds
The recording thread reads the joint encoders and servo speeds, measures the sampling interval, and adds each valid sample to record_list as angles, speeds, and interval_time.
- Stop Recording and Refocus Servos
Calling stop_record() stops recording and waits for the recording thread to finish; non-MyArm robots then call focus_all_servos(), while MyArm calls power_on().
- Replay the Recorded Joint Path
Calling play() reads the encoders, speeds, and intervals in order and sets joint positions with set_encoders_drag(). It waits 3 seconds after the first record, then replays each remaining record at its recorded interval.
Use it
- Run the control program
Run the myCobot280_handle_control.py program.
- Click Right 1 first
After running the program, click the **Right 1** button first. Other operations can be performed only after the machine reaches the initial point.
- Button 1 increases RX
Press **1** to increase the RX direction coordinate value.
- Button 2 decreases RX
Press **2** to decrease the RX direction coordinate value.
- Button 3 increases RY
Press **3** to increase the RY direction coordinate value.
- Button 4 decreases RY
Press **4** to decrease the RY direction coordinate value.
- Button 5 increases X
Press **5** to increase the X direction coordinate value.
- Button 6 decreases X
Press **6** to decrease the X direction coordinate value.
- Button 7 decreases Y
Press **7** to decrease the Y direction coordinate value.
- Button 8 increases Y
Press **8** to increase the Y direction coordinate value.
- Button 9 decreases Z
Press **9** to decrease the Z direction coordinate value.
- Button 10 increases Z
Press **10** to increase the Z direction coordinate value.
- Button 11 decreases RZ
Press **11** to decrease the RZ direction coordinate value.
- Button 12 increases RZ
Press **12** to increase the RZ direction coordinate value.
- Button 13 wakes the handle
After connection, the handle enters sleep mode when it is not used for a long time. Press **13** to wake it up.
- Button 14 checks connection
Press **14** to check the machine connection status. The atom LED flashes green three times when normal and red three times when abnormal.
- Press X to open gripper
Press **X** to open the gripper.
- Press Y to close gripper
Press **Y** to close the gripper.
- Press A to open suction pump
Press **A** to open the suction pump.
- Press B to shut down suction pump
Press **B** to shut down the suction pump.
- Hold Left 2 to relax joints
Press and hold **Left 2** for 2 seconds; the robot stops torque output and relaxes all joints.
- Hold Right 2 to lock joints
Press and hold **Right 2** for 2 seconds; the robot turns on torque output and locks all joints.
- Increase RX coordinate
Click the gamepad's 1 button to increase the RX-direction coordinate value.
- Decrease RX coordinate
Click the gamepad's 2 button to decrease the RX-direction coordinate value.
- Increase RY coordinate
Click the gamepad's 3 button to increase the RY-direction coordinate value.
- Decrease RY coordinate
Click the gamepad's 4 button to decrease the RY-direction coordinate value.
- Increase X coordinate
Click the gamepad's 5 button to increase the X-direction coordinate value.
- Decrease X coordinate
Click the gamepad's 6 button to decrease the X-direction coordinate value.
- Decrease Y coordinate
Click the gamepad's 7 button to decrease the Y-direction coordinate value.
- Increase Y coordinate
Click the gamepad's 8 button to increase the Y-direction coordinate value.
- Decrease Z coordinate
Click the gamepad's 9 button to decrease the Z-direction coordinate value.
- Increase Z coordinate
Click the gamepad's 10 button to increase the Z-direction coordinate value.
- Decrease RZ coordinate
Click the gamepad's 11 button to decrease the RZ-direction coordinate value.
- Increase RZ coordinate
Click the gamepad's 12 button to increase the RZ-direction coordinate value.
- Open the gripper
Click the X button to open the gripper.
- Close the gripper
Click the Y button to close the gripper.
- Turn on the suction pump
Click the A button to turn on the suction pump.
- Turn off the suction pump
Click the B button to turn off the suction pump.
- Wake the sleeping gamepad
After the gamepad has been connected and left unused for a long time, it enters sleep mode; press the 13 button to wake it.
- Check robot connection status
Press the 14 button to check the robot connection status; three green flashes from the atom LED indicate normal operation, and three red flashes indicate an abnormal state.
Common problems
- State-reading APIs return empty values, arrays, or results #19 #28 #39 #66
- Package installation or upgrades fail; system images contain outdated dependencies #154 #155 #157
- Robot intermittently stops responding and only a power cycle recovers it #93 #94
- Adaptive gripper sticks closed, fails to respond, or returns no status #8 #219 #231
- Motion commands fail to move, block until timeout, or run unevenly #13 #73 #79 #111 #124
- Library calls crash from missing dependencies or falsely report failed operations #3 #96 #228
- Joint limits disagree with calibration parameters, reducing positioning accuracy #10 #104 #134
- Some OS or firmware combinations fail communication or lack matching firmware files #40 #61 #224
Parts
- Ships assembled: six servo-motor joints (model not given), J1 ±168°, J6 ±180°
- Raspberry Pi 4B (BCM2711 quad-core 1.5 GHz, 2GB RAM), Ubuntu Mate 20.04, Wi‑Fi and Bluetooth
- Base: two USB 3.0, two USB 2.0, two micro‑HDMI, Ethernet, audio jack, SD slot and GPIO headers; 12V 5A DC power
- End: Atom module (5×5 RGB matrix and button), Grove, Type‑C and gripper-servo ports; the base takes LEGO holes and M4 screws, the end LEGO holes and threaded holes
- Options: flat or G-shape base, adaptive/parallel/flexible grippers, suction pump, camera flange, gamepad, AI Kit 2023
Tools
- Ships assembled; fix it with the flat base (suction) or G-shape base (clamps to a desk edge), or through the base's M4 screw holes
- Standalone use needs a monitor, keyboard and mouse
Skills
- myBlockly graphical programming (beginner)
- Python (pymycobot: joint, coordinate, gripper and IO control)
- ROS1/ROS2 with RViz and MoveIt simulation
Software
- pymycobot (MIT, v4.0.7 released 2026‑09‑05)
- mycobot_ros (BSD‑3) and packages such as mycobot_280pi in mycobot_ros2
- myStudio (firmware flashing and tutorials), myBlockly, RoboFlow; the store page also lists Mind+