Data Visualization with PIC GUI

☀️ Light

Overview

PIC GUI is a lightweight, user-friendly data visualization tool that enables real-time monitoring of variables from your embedded application. Unlike External Mode (which requires XCP protocol), PIC GUI provides a simplified approach to streaming data over UART for analysis and debugging.

PIC GUI vs External Mode

How PIC GUI Works

Data Flow

• Embedded Side: UART Tx-Matlab block packages selected variables into packets and transmits via UART
• Serial Connection: USB-Serial adapter bridges microcontroller UART to PC COM port
• MATLAB Side: PIC GUI decodes packets and makes data available as MATLAB variables
• Visualization: Custom MATLAB script plots data in real-time or saves for offline analysis

Packet Format

The UART Tx-Matlab block uses a simple, efficient packet structure:

• Start Byte: Synchronization marker (0x5A)
• Data Length: Number of variables being transmitted
• Variable Data: Raw binary values (16-bit or 32-bit integers)
• Checksum: Optional error detection

Setting Up PIC GUI

Step 1: Add UART Tx-Matlab Block to Model

• Open your Simulink model

• Add UART Config block from MCHP library → Communication

• Configure UART:

• Baud rate: 115200 (standard) or up to 2 Mbaud with FTDI cable

• Select UART peripheral (e.g., UART1)

• Assign Tx/Rx pins matching board hardware

• Add UART Tx-Matlab block from MCHP library → Communication

• Configure transmission:

• Number of variables to transmit

• Sample time (data logging rate, e.g., 1ms)

• Variable data types (int16, int32, float, etc.)

• Connect signals you want to monitor to UART Tx-Matlab block inputs

Step 2: Build and Program

• Press Build button (Ctrl+B) to generate code and program microcontroller
• Verify UART communication is working (LED blinks, no error messages)
• Connect USB-Serial adapter to UART pins on development board

Step 3: Launch PIC GUI

• In MATLAB command window, type: picgui
• Alternatively, double-click the UART Tx-Matlab block in your model
• PIC GUI interface opens

Step 4: Connect to Target

• Select COM port from dropdown (e.g., COM3, COM4)
• Set Baud rate matching UART Config block (e.g., 115200)
• Click Connect button
• Verify “Connected” status appears
• Data should start streaming immediately

Step 5: Customize Visualization

PIC GUI includes a default visualization script, but you can create custom plots:

• Click Edit Script button in PIC GUI
• Modify MATLAB code to create custom plots
• Click Start to run visualization
• Data updates in real-time as packets arrive

Example: Motor Control Data Logging

Typical use case: Monitor speed, current, and control voltage during motor operation.

Simulink Configuration

% UART Tx-Matlab block configured with 4 inputs:
% Input 1: Speed setpoint (int16, RPM)
% Input 2: Measured speed (int16, RPM)
% Input 3: Motor current (int16, mA)
% Input 4: PWM duty cycle (int16, %)
% Sample time: 1ms (1kHz data rate)

Custom Visualization Script

% Extract data from PIC GUI variables
time = data.time;
speed_ref = data.var1;
speed_meas = data.var2;
current = data.var3;
duty = data.var4;

% Create subplots
subplot(3,1,1);
plot(time, speed_ref, 'r--', time, speed_meas, 'b');
ylabel('Speed (RPM)');
legend('Reference', 'Measured');
grid on;

subplot(3,1,2);
plot(time, current, 'g');
ylabel('Current (mA)');
grid on;

subplot(3,1,3);
plot(time, duty, 'm');
ylabel('Duty Cycle (%)');
xlabel('Time (s)');
grid on;

Result: Real-time plots showing motor speed tracking, current consumption, and controller output - all captured during actual hardware operation at 2000 RPM with no load.

Advanced Features

Data Logging to File

Save streamed data for offline analysis:

% In PIC GUI custom script
% Save data structure to MAT file every 10 seconds
if mod(length(data.time), 10000) == 0
    save(['motor_log_' datestr(now, 'yyyymmdd_HHMMSS') '.mat'], 'data');
end

Real-Time Signal Processing

Apply filters, FFT, or other analysis in real-time:

% Calculate moving average of current
window_size = 50;
current_filtered = movmean(data.var3, window_size);

% Compute FFT of speed signal
fs = 1000; % Sample rate (Hz)
N = length(speed_meas);
f = (0:N-1)*(fs/N);
speed_fft = abs(fft(speed_meas));

% Plot frequency spectrum
figure(2);
plot(f(1:N/2), speed_fft(1:N/2));
xlabel('Frequency (Hz)');
ylabel('Magnitude');
title('Speed Spectrum');

Automated Testing

Use PIC GUI for automated characterization:

% Run step response test automatically
% Wait for steady-state at each speed
test_speeds = [500, 1000, 1500, 2000]; % RPM
for i = 1:length(test_speeds)
    % Capture data for 5 seconds at each speed
    wait_time = 5;
    % Analyze settling time, overshoot, steady-state error
    % Save results to report
end

Multi-Variable Analysis

Correlate multiple signals to identify system behavior:

% Plot phase portrait: current vs speed
figure(3);
plot(speed_meas, current, '.');
xlabel('Speed (RPM)');
ylabel('Current (mA)');
title('Operating Point Distribution');
grid on;

Troubleshooting

No Data Received

• Check COM port: Verify correct port selected in PIC GUI (use Device Manager on Windows)
• Verify baud rate: Must match UART Config block setting exactly
• Check wiring: Ensure Tx pin of microcontroller connects to Rx of USB-Serial adapter
• Test hardware: Try loopback test (connect Tx to Rx) to verify adapter works
• Check UART pins: Verify UART Config block pin assignments match board schematic

Garbled Data / Synchronization Errors

• Baud rate mismatch: Double-check both sides set to same rate
• Buffer overflow: Reduce data rate (increase UART Tx-Matlab sample time)
• Clock accuracy: Verify oscillator configuration in Master block for accurate baud generation
• Cable quality: Use shielded USB cable, keep cable length < 3m

Slow Update Rate

• Increase baud rate: Switch from 115200 to 460800 or 2 Mbaud
• Reduce variables: Transmit fewer signals to reduce packet size
• Use integer types: int16 transmits faster than float (2 bytes vs 4 bytes)
• Increase sample time: If 1ms rate not needed, try 5ms or 10ms
• Optimize script: Reduce MATLAB processing in visualization loop

MATLAB Script Errors

• Variable naming: Use @@PROT_9@@, @@PROT_10@@, etc. to access transmitted variables
• Array dimensions: Check that arrays have same length before plotting
• Handle clearing: Clear figure handles properly to avoid memory leaks
• Test incrementally: Start with simple plot, add complexity gradually

Best Practices

Bandwidth Calculation Example

Ensure your data rate doesn’t exceed UART capacity:

% Configuration:
% - 5 variables (int16) = 10 bytes
% - Packet overhead (start, length, checksum) = 4 bytes
% - Total packet size = 14 bytes
% - Sample rate = 1 kHz (1ms)

bytes_per_packet = 14;
sample_rate = 1000; % Hz
data_rate = bytes_per_packet * sample_rate; % bytes/s
data_rate_bits = data_rate * 10; % 10 bits per byte (1 start + 8 data + 1 stop)

% Result: 140,000 bits/s = 140 kbps

% Check against baud rate:
baud_rate = 460800;
utilization = (data_rate_bits / baud_rate) * 100;
% Result: 30% utilization - Safe! (< 80% threshold)

Integration with Other Tools

Using PIC GUI with External Mode

Both tools can coexist if using different communication channels:

• PIC GUI: UART1 for data logging (e.g., high-speed signals)
• External Mode: UART2 for parameter tuning (e.g., PI gains)

Combining with PIL Testing

Use PIC GUI during PIL tests to capture detailed execution data:

• Run PIL test to verify algorithm correctness
• Add PIC GUI logging to same model
• Capture internal states during PIL execution
• Analyze timing, overflow conditions, state transitions

See Also

External Mode | PIL Testing | Quick Start Guide | UART Tx-Matlab Block Reference | User Guide