This simulator follows a signal across a two-layer sensor board: power enters through a connector and 3.3 V regulator, a sensor voltage is conditioned by an RC input filter, the ADC samples and holds that voltage before converting it to a 12-bit code, a serial-like transfer carries the word to a controller, and the controller's duty command drives a PWM LED output stage — all referenced to a continuous ground-plane return. Adjust supply, sensor position, injected noise, filter values, sample timing, power-path resistance and decoupling, and inject board faults to see where the signal chain breaks.
• A real-time 3D cutaway workbench of the power/sensor connectors, the 3.3 V regulator and decoupling, the RC sensor-conditioning filter, the ADC package with a sample-and-hold inset, the SPI-like serial traces and vias, the controller and clock, the PWM LED driver stage, and the ground plane/plated-through vias, with home view, focus-selected-part, toggleable full enclosure, exploded view, auto-rotate and expand controls, tappable numbered components with callouts matching the diagram reference, and a labels toggle. • Nine live controls: board input supply, sensor position (fraction of local rail), injected 50 Hz sensor noise, input filter resistance, input filter capacitance, sample request interval, power-path resistance, rail decoupling capacitance, and a board-fault selector (healthy, open analog signal trace, open serial-data connection, open power return). • Play/pause, single-step (0.1 s) and larger-step (1 s) time controls, plus a playback-speed selector from real time to 1 minute per second. • A Reset laboratory action, a Step sensor position action, a Restore board connections action, a live "what is happening" sequence narrative with component status tokens and a readings table. • Ten live metrics: local board rail, sensor source voltage, filtered ADC input, last sampled (held) voltage, last received 12-bit word, controller duty command, instantaneous output pulse state, current serial transfer bit progress, completed received words, and time since last valid word. • A Curves & measurements tab with two charts (sensor vs. filtered vs. held-sample voltage; duty command), the complete model equation set, and snapshot measurements. • An Experiments tab with four guided scenarios (sensor step, slow analog filter, data connection failure, open signal trace), a model-verification bench of independent automated checks, and a timestamped event log with a copyable trial report. • A Learn & assess tab with guided lessons (establish power and reference, condition and sample, transfer a digital word, drive and diagnose), a knowledge-check quiz with reset, and a written scope/reference statement.
The regulator produces a rail of Vreg = min(3.3, max(0, Vin − 0.3)), and the local rail sags toward Vreg minus the trace-resistance drop through decoupling. The RC input filter attenuates changing sensor input with time constant τ = Rfilter·Cfilter before each ADC request freezes (samples and holds) a voltage for a representative 5 ms conversion; that held value is quantized to a 12-bit code = round(clamp(Vsample/Vreference, 0, 1) × 4095). A representative 4 ms serial transfer then carries the word to the controller, which sets duty = code/4095 — so the analog filter response, the sample-and-hold instant, and the eventual PWM duty are three distinct, sequential stages rather than one continuous update.
If no valid word arrives at the controller for 200 ms, duty is forced to zero — this is exactly what the data-connection-failure experiment demonstrates, and it differs meaningfully from an open analog signal trace, which is pulled to zero in this fixture and therefore still produces valid (but low) 12-bit codes rather than triggering a timeout. Distinguishing "no fresh data" from "valid data reporting zero" is one of the simulator's core diagnostic points.
This is a behavioral mixed-signal teaching board with intentionally slowed ADC, serial and PWM timing (a 40 Hz LED PWM waveform, deliberately slow for observation). It is not a specific microcontroller, SPI protocol implementation, regulator transient model, or PCB electromagnetic field solver, and an open power return is treated as simply unpowered with no parasitic backfeeding. Component changes are treated as new design experiments at the retained state, not a model of physically swapping energized parts; use Reset laboratory or Restore board connections to start fresh.
No. Each conversion uses a held sample: the ADC freezes a voltage at the moment of a sample request, and that fixed value is what gets converted and transferred, even as the actual filtered analog input keeps changing afterward.
A broken serial-data connection means no words ever reach the controller, which triggers a 200 ms timeout and forces duty to zero. An open analog signal trace instead gets pulled to zero in this fixture, so the ADC still produces valid, fresh 12-bit codes — they are just low-valued, not missing.
The RC filter time constant (τ = Rfilter × Cfilter) governs how quickly the filtered analog voltage can track a changing sensor input. With a large resistor and capacitor, that time constant can reach roughly a second, so the analog signal lags well behind the true sensor voltage even while the ADC keeps sampling and converting on its own schedule.
It is a behavioral teaching model with deliberately slowed ADC, serial-transfer and PWM timing — it is not a specific microcontroller, a real SPI protocol implementation, a regulator transient model, or an electromagnetic field solver for the board traces and ground plane.