Skip to content

Operator workflow

A typical bring-up + run session, from "dashboard not running" to "5-level output on the scope."

1. Bring the dashboard up

cd firmware/stm32-f303re
dashboard\.venv\Scripts\python dashboard\run_dashboard.py

The dashboard opens in Simulator mode by default — safe for confirming the UI works before touching hardware.

2. Connect the Nucleo (live serial)

  1. Plug in the Nucleo over the ST-LINK USB.
  2. In the dashboard, choose Source → Live serial.
  3. Pick the COM port (Windows assigns one named STM32 STLink (COMx)).
  4. Click Connect.

On connect, the dashboard sends STATUS. This both confirms two-way communication and cancels the firmware's 3 s auto-start window — the operator now has full control.

The log pane shows the firmware's startup lines:

  • $A,BOOT_SELF_TEST_DONE
  • $C,mod=STAIR,fsw=500,bridge=BOTH,ffund=50,mi=0.95,cntoff=0,lock=OK
  • $P,vnom=50.00,uv=40.00,ov=58.00,oc=15.00,imbal=10.00
  • (if sensors aren't fully connected) $E,MODE_DEMOTED or $E,WARNING_OPEN_LOOP_NO_PROTECTION

3. Configure PWM (before arming)

In the PWM Config panel, set:

  • MOD → PSC (the project deliverable; bridges are thermally balanced).
  • FSW → 5000 Hz (preset dropdown).
  • BRIDGE → BOTH.
  • FFUND → 50 Hz.

Each change emits a fresh $C,... line in the log. Confirm lock=OK — if it shows lock=BAD, the PSC carrier shift didn't land; either rerun the config or fall back to STAIR_ALT.

If bench-testing below 50 V, also set VNOM to match the supply voltage (e.g. 12 for a 12 V bench supply). Confirm the new thresholds on the $P line.

4. Arm and start

  1. Check Arm live START.
  2. Click START.

The log shows:

$A,START

The FSM advances IDLE → PRECHARGE → RUN. The sensor graph starts updating at 20 Hz. The scope should show 5 distinct cascade output levels (or 5-level PWM on PSC).

5. Sustained run

Watch:

  • Telemetry sanity — vdc1 and vdc2 track the bench supplies; iout matches the load. Any NAN means the channel is unavailable (likely a sensor wiring issue).
  • Thermal balance — touch-check both bridges' MOSFETs after 5 minutes. PSC should give convergence within ~3 °C.
  • No spurious fault lines — if $F,... appears, the dashboard fault badge highlights the offending bit.

6. Stop

Click STOP. The FSM returns to IDLE; PWM outputs go off via BDTR.MOE = 0 (all MOSFETs off via the TLP250 + OSSI=1 chain).

Fault recovery

Fault What likely happened Path back
UV Bus dropped below 0.8 × VNOM. Investigate the supply. Once it recovers, CLEAR.
OV Bus exceeded 1.16 × VNOM. Reduce the supply. CLEAR.
OC Output current exceeded OC setting. Check the load. CLEAR.
IMBAL One bridge bus drifted from the other by > 0.2 × VNOM. Check the two isolated supplies. CLEAR.
SENSOR_LOST A required sensor went stuck-at-rail. Investigate wiring. RESCAN (allowed in FAULT) to re-check. If recovered, CLEAR.
MANUAL Operator pressed FORCE FAULT. CLEAR.

Scenarios (simulator only)

The scenario buttons (Undervoltage, Overvoltage, Overcurrent, DC imbalance, Sensor lost, Open loop / no protection, Mode demotion, Nominal run) play against the simulator. They never inject into the firmware. Use them to demo the fault chain to a reviewer without touching the bench.

If a simulator fault is active, use Normalize sim sensors before CLEAR — same requirement as the firmware (the active condition must be gone before the latch clears).