I wanted to test our Red Pitaya Copilot on a practical development project. I chose a custom partial-discharge signal simulator: a tool for generating synthetic signals resembling the output of a PD sensor. The work brought together FPGA logic, Linux controls, and a browser interface. I then observed the generated output using the oscilloscope application on a second Red Pitaya.
The goal was to create configurable PD-like pulses with background noise and several operating scenarios. Copilot helped translate that idea into the connected parts of the application: a hardware generator, its register interface, Python control tools, and a web interface.
The simulator produces low-voltage sensor-like signals. Its healthy, early PD, moderate PD, severe PD, and manual modes provide synthetic conditions for exploring detector behavior. These labels describe test scenarios rather than diagnoses of real insulation.
Copilot helped develop the FPGA logic for pulse amplitude, event rate, ringing, pulse length, and noise. A simulated 50 Hz phase reference allows activity to concentrate within selected parts of the cycle. When enabled, the custom generator drives OUT1.
Copilot also helped connect those settings to the register map and Python controls. That made it possible to select a mode, adjust parameters, read status, and stop generation from software.
The browser interface gives the generator a practical set of controls. Copilot helped develop the interface and its connection to the hardware. The screenshot below shows the waveform preview, live readouts, and advanced settings for adjusting the simulated signal.
Figure 1. OpenPD Generator interface. The waveform preview is reconstructed from settings; live DAC, phase, and event information come from the FPGA registers.
To observe the physical output, I used another Red Pitaya as the oscilloscope. The capture below shows the extreme-failure simulation output on IN1, with irregular positive and negative spikes superimposed on a periodic baseline.
This capture provides a view of the signal through a separate acquisition path. The interface preview shows the configured waveform; the oscilloscope records the signal received by the second board. They serve different purposes when checking the generator.
Copilot also helped work on verification code and the surrounding control behavior. The project includes checks for reset, start and stop conditions, watchdog shutdown, and event-buffer overflow accounting. These checks help define what the implementation should do; a particular hardware result still needs its own measurement evidence.
My role was to define the experiment, choose the signal conditions, and assess the output. Copilot helped develop the implementation across the FPGA, software controls, and browser interface, while working through the connections between them.
The most useful part of this test was being able to carry an idea through those layers and then inspect the output with a second instrument. The simulator now gives me a platform for further experiments with synthetic pulses, noise, and detection.
Red Pitaya Copilot is an AI-assisted development tool designed to streamline custom instrumentation projects on Red Pitaya open-source FPGA platforms. It assists developers across multiple stack layers, including FPGA hardware logic development, C/Python register interfaces, and web GUI control integration.
In this test, Copilot helped write FPGA logic for generating synthetic PD pulses, configuring parameters such as pulse amplitude, event rate, ringing decay, baseline noise, and 50 Hz AC phase synchronization driving the analog output (OUT1).
Verification involves using a second Red Pitaya unit configured as an oscilloscope to measure physical analog output (OUT1 to IN1) while running programmatic verification scripts to validate register maps, start/stop conditions, watchdog behavior, and event-buffer handling.
A: Yes. Copilot helps bridge hardware registers with web applications, helping developers create browser interfaces that display real-time FPGA register data (such as live DAC voltage, phase angle, and event counts) alongside interactive waveform controls.