Meta Terminal 3.1 - PID Tuning Dashboard

Forked from Meta Terminal III, Meta Terminal 3.1 has been entirely re-architected into a professional-grade, dual-loop PID tuning dashboard. It is designed specifically for live embedded motor control tuning, featuring synchronized real-time plotting for cascaded Angle (Outer) and Velocity (Inner) loops, automated step-response generation, and hardware-safe parameter injection.

Demo

Requirements

• Python 3.6+
• PyQt5
• pyqtgraph
• numpy
• pyserial

Integration with ut_pid_tuning (meta-embedded)

This application serves as the host-side visualizer and command center for the ut_pid_tuning.cpp unit test firmware within the meta-embedded (ChibiOS/STM32) repository.

How it Works Together

The dashboard communicates with the STM32 via a hardware Serial/UART connection using the custom Meta Shell protocol. It is specifically designed to tune Cascaded PID Loops safely and efficiently:

1. State Synchronization: Clicking "Fetch Current PIDs" sends the tune get command to the STM32, which locks the hardware UART Mutex and returns the active boot parameters for both the Angle (a2v) and Velocity (v2i) control loops.
2. Automated Step Responses: The "Run Step Sequence" button triggers a 5-stage state machine on the STM32 (IDLE -> PRE_ZERO -> STEP_POS -> STEP_NEG -> POST_ZERO). The firmware automatically handles the timing and begins streaming high-frequency telemetry.
3. Synchronized Dual-Plotting: The STM32 streams four data points per cycle (Target Angle, Actual Angle, Target Velocity, Actual Velocity). To prevent GUI lockups and visual tearing, the Python terminal buffers the serial lines and only flushes the data to the UI when a complete, synchronized 4-part frame is received. The axes of the two plots are mathematically linked for perfect visual comparison.
4. Buffer-Safe Injection: When sending new PID parameters, the GUI utilizes a queued asynchronous delay (150ms) to feed commands to the STM32 one by one. This prevents hardware buffer overflows on the serial converter and ensures the active skdThread safely absorbs the new control parameters via load_PID_params().

Recommended Tuning Workflow

When paired with the meta-embedded firmware, follow the standard robotics cascaded tuning procedure:

1. Set Tuning Mode to v2i (Velocity).
2. Tune the Inner Loop (Velocity) using P, I, and D until the step response is sharp and dampen the ringing.
3. Switch Tuning Mode to a2v (Angle).
4. Tune the Outer Loop (Angle) using P-only (Kp). If the motor begins to oscillate, either lower the Angle Kp or return to the Velocity loop to make it stiffer.

File Overview

• main.py: The main entry point of the application. Sets up the GUI and handles user interactions.
• device.py: Manages device connections and communication, utilizing pyserial for robust buffer flushing.
• chart_widget.py: Manages the layout container for the plotting widgets, enforcing edge-to-edge UI expansion.
• graph.py: Handles the underlying dual-plot architecture via pyqtgraph, enforcing dark mode styling, MATLAB color profiles, and X-axis synchronization.