Skip to content

Latest commit

 

History

History
61 lines (46 loc) · 2.48 KB

File metadata and controls

61 lines (46 loc) · 2.48 KB

Bring-up checklist

Work top to bottom; each stage has a hard pass condition. Never skip to the bike with an unverified stage.

0. Firmware smoke test (no hardware but a DevKit)

  • cargo test at repo root — all host tests green.
  • cd firmware && cargo build --release.
  • Flash DevKitM-1 (cargo run --release, needs cargo install espflash).
  • Serial log shows gateway up, SDS reinit attempts cycling.
  • Wi-Fi AP vstrom-gw visible; http://192.168.4.1/ shows dashboard with link down (expected — nothing connected).

1. K-Line loopback (bench, no bike)

  • Wire L9637D per hardware/README.md, 12 V PSU on VS + pull-up.
  • Scope/logic analyser on K: verify 25 ms low pulse, then 81 12 F1 81 05 at 10,400 baud on every reinit cycle.
  • /log.txt shows the TX frames (and their echoes as RX — proves the receive path works).

2. On the bike, ignition on (engine off)

  • Battery voltage sane at J1 (≈12.5 V).
  • Console: link up, polls_ok counting.
  • /log.txt: full 61 08 responses arriving.
  • Run the M1 field-identification procedure (docs/01) and fix SdsLayout::DL650A_L7.
  • Battery voltage / gear / TPS on dashboard move plausibly.
  • Engine on: RPM matches tacho at idle and 4000 rpm.

3. OBD server (bench)

  • candump via USB-CAN dongle, or ELM327 phone app on the J2 side.
  • 01 00 answered; RPM/speed/temps read correctly against dashboard.
  • Mode 09 VIN transfers (multi-frame — watch for FlowControl handling).
  • Unplug K-Line: data PIDs go silent, bitmaps still answered.

4. WingMan (bench, then bike)

  • WingMan powered from 12 V, CAN to J2. Sniff per docs/02 M2 procedure.
  • RideLink app shows live data. Record which PIDs it used.
  • Full chain on the bike, engine running.
  • Ride test: values track; no session drops (reinits stays flat); WingMan stays happy through ignition off/on cycles.

5. DTC path

  • Unplug a sensor (IAT is the least invasive) → dash FI light, mode 03 shows the code, console lists it.
  • Clear via scan tool (mode 04) → code gone, ECU cleared (verify with dealer-mode dash check).

Field debug notes

  • Status LED (rev A board): currently unused by firmware — TODO wire to link state in main.rs once an LED exists.
  • Wi-Fi console is an OPEN AP — do not leave enabled permanently on the street build (docs/05-tunables.md).
  • espflash monitor over USB works while the board runs from 12 V.