Check and finish the imported board


Thought for 16 sec
Thought for 11 sec
Referenced guidance.
Create a file which documents ALL of the changes which we've made to the board . Every change since the initial KiCad drawing. It should be in Markdown
I’ll create a Markdown change history covering the work since the KiCad import, including reverted changes, accepted risks and the latest M4-hole/board-size changes. I’ll append the current manufacturing-rule values you requested, distinguishing verified settings from anything unspecified; no board edits.
Created Board Change History in your project files. It documents the recorded changes since the KiCad import—including connectors, routing, test pads, silkscreen, M4 holes and the 87 × 77 mm outline—and appends the manufacturing-rule values, marking unknown settings explicitly.
The file distinguishes applied changes, abandoned proposals and accepted risks, and flags the previous manufacturing ZIP as outdated. No schematic or PCB changes made.
"KiCad Board Change History" completed.
Created Board Change History in Markdown, covering recorded changes since the KiCad import, including connector consolidation, routing, test pads, silkscreen, accepted risks, and the latest M4 holes and 87 × 77 mm board expansion.
The appendix includes track width, clearance, via/drill sizes, solder-mask expansion and board-edge clearance. Paste expansion and explicit minimum via-size settings are marked unknown, rather than assumed. The history is reconstructed from saved records and live inspection—not a guaranteed exhaustive commit-by-commit diff.
No board changes were made. The previous manufacturing ZIP remains outdated following the hole and outline changes; fresh verification and exports are still needed.
Execution stopped because your ACUs ran out.

Board bring-up plan (Phase 10, rev. 2026-10-22 reconciliation)
Practical DIY procedure for the hand-soldered board. No hardware test has been performed. Every number below is a provisional check threshold derived from datasheet values or stated assumptions — not a guaranteed device spec and not a pass/fail certification. If a reading is outside a threshold, stop and investigate; a reading inside it does not prove the board is healthy. Accepted risks (Phases 1–4, 7, antenna) stay accepted; nothing here implies surge, load-dump or vehicle-transient survival.
Revision note: corrected per the qv review — removed the unsupported universal “>1 kΩ” rail criterion, clarified current-limit/inrush guidance, made firmware tests explicitly pending, defined a safe LI stimulus, rewrote isolation/grounding wording, and stated that the fitted SMD Pico is not removed.
Connector / test-pad map (verified against live netlist, 18 nets, 2026-10-22)
  • J1 (TE 282834-8, 8-pos THT): 1 /12V_IN · 2 /GND · 3–4 /K1 · 5 /LI1 · 6–7 /K2 · 8 /LI2. J1 is placed on the board bottom side (legitimate THT mounting); confirm with the user that this side is intended before soldering — not changed.
  • J2 (switch block): 1 /SW_IN · 2 /12V_SW · 3 /12V_SW (LED+) · 4 /GND (LED−).
  • Test pads: TP1 /TX1 · TP2 /RX1 · TP3 /LO1 · TP4 /GND · TP5 /TX2 · TP6 /RX2 · TP7 /LO2 · TP8 /GND (each confirmed in get_nets).
  • Other probe points: F1 pads (/12V_IN, /12V_F), D1:K (/SW_IN), U4 +OUT / D3:A (/5V), D3:K (/VSYS), C1/C2 pin 1 (/3V3).
  • Power path: J1.1 → F1 → D1 → J2.1 → external switch → J2.2/3 (/12V_SW) → U4 buck, U2/U3 VS, R1/R2, D2. U4 → /5V → D3 → /VSYS → Pico. Pico 3V3 → U2/U3 VCC, C1/C2.
Equipment & safety
  • Current-limited bench supply, DMM, optional scope.
  • Current limit is supply protection, not proof the board is healthy. A board can be faulty while drawing less than the limit.
  • Grounding / isolation:
    • Board GND, bench-supply negative, the Pico USB ground and the host PC ground all become common when USB is connected.
    • A laptop on battery alone does not guarantee isolation — its USB ground can still be tied to earth through a charger, monitor, dock, or other grounded equipment connected to it.
    • A mains-earthed scope's ground clip is earth. Connect it only to board GND (TP4/TP8 or a GND pad), and only after confirming that board GND is not already referenced to a different potential (e.g. through the bench supply, a grounded PC, or a vehicle chassis). Never clip scope ground to 12V, K, LI or any signal.
    • For vehicle connection, assess vehicle ground vs. laptop/scope ground first; if unsure, use a USB isolator or do not connect USB/scope while on the vehicle.
Step 0 — Unpowered checks
  1. Visual: solder bridges, D1/D3 band orientation, D2 band toward /12V_SW pad 1 (intended mapping, Phase 11 definition cleanup pending), U2/U3 pin 1, U4 corner polarity per silk, J1 mounting side.
  2. Short checks (DMM, power off): measure /12V_IN, /12V_SW, /5V, /VSYS, /3V3 each to GND. There is no universal resistance threshold — rails with capacitors/regulators read low and rising. Treat a steady near-0 Ω reading (≈< 10 Ω, provisional) as a suspected short and stop; otherwise record values for comparison with a known-good board later.
  3. K pull-ups: K1→/12V_SW and K2→/12V_SW should read ≈510 Ω (resistor value; the parallel transceiver path may lower the reading slightly).
  4. Switch: unpowered continuity J2.1↔J2.2 open when OFF, closed when ON (LED terminal is polarity-dependent; do not treat it as a contact).
Step 1 — First power (Pico already fitted)
  • The Pico (U1) is a surface-mounted module soldered to the board. Do not desolder it to follow a staged test. All power tests are done with it fitted.
  • Supply 12.0 V. Initial current limit ≈ 150 mA (provisional). Expected steady idle with Pico running stock/no firmware: roughly 15–60 mA at 12 V (estimate: buck efficiency assumed, Pico idle, L9637 VS ≤ 3 mA each per datasheet ON-state max, VCC ≤ 2.3 mA each).
  • Inrush vs. persistent fault: at switch-on, the buck input and output capacitors charge; the supply may briefly hit the limit (constant-current/foldback for a few ms). That alone is not a fault. A persistent fault is: current stays at the limit, rails don't rise to their expected values within ~1 s, or the supply stays in CC mode. If startup fails only because of the limit, raise it in steps (e.g. 250 mA → max ~500 mA, below F1's 1 A) and retry — do not raise further if it still won't start.
  • Stop immediately on any part heating, smoke/smell, any rail above its expected maximum (overvoltage), or a persistent current-limit condition.
  • Provisional rail checks: /12V_SW ≈ 11.5–11.8 V (12 V minus D1 drop); /5V 4.9–5.1 V (seller ±0.1 V claim); /VSYS ≈ 4.5–4.8 V (minus D3 drop; Pico VSYS range 1.8–5.5 V); /3V3 3.2–3.4 V.
  • K1/K2 idle ≈ /12V_SW (pull-ups, transceiver recessive).
Step 2 — USB and firmware (prerequisites)
  • Firmware is not provided by this project. No commands or images are specified here. All firmware-dependent tests below are PENDING actual firmware.
  • Prerequisite firmware must: configure GP0/GP1 as UART0 (TX1/RX1), GP4/GP5 as UART1 (TX2/RX2), GP2/GP6 as inputs (LO1/LO2); keep TX lines idle-high (recessive) at boot; provide a test mode that sends a short burst and prints RX and LO states over USB serial.
  • Hardware condition (Phase 11.1 ERC exception): GP1/GP2/GP5/GP6 are driven by L9637D RX/LO outputs and must be configured as inputs, never driven as outputs. Firmware has not been reviewed against this condition.
  • USB may be connected with the bench supply on (D3 isolates the buck from USB backfeed). Apply the grounding rules above. Accepted Phase 1 risk: USB-only power gives 3V3 to the transceivers while VS is unpowered.
  • Pass criteria (once firmware exists): device enumerates; TP1/TP5 idle ≈ 3.3 V; TP2/TP6 idle ≈ 3.3 V with K idle high.
Step 3 — Channel tests (one channel at a time, firmware-dependent)
K-line loopback (no external tester needed): R1/R2 already pull K up to /12V_SW. With nothing connected to J1.3/4 (or J1.6/7), transmit a short burst at 10.4 kbaud; the L9637 echoes K back on RX. Observe TP1 vs TP2 (CH1) or TP5 vs TP6 (CH2): RX should follow TX, swinging ~0 ↔ ~3.3 V; K low level should be near ground (K ON impedance 10 Ω typ / 30 Ω max per datasheet). Pass: RX matches TX bit pattern.
  • Keep K dominant (low) only for short bursts. With K held low at 12 V, each 510 Ω pull-up dissipates ≈ 0.28 W (accepted actual-part rating risk). Do not hold TX low continuously; feel R1/R2 for warmth after testing.
  • If an external K-line tester is used instead, its low level must be well within the L9637 K thresholds and it must not back-drive K above /12V_SW.
LI → LO stimulus (defined, datasheet-based): LI is a high-voltage input referenced to VS (LI low ≤ 0.45·VS, high ≥ 0.55·VS for 4.5–18 V; absolute max −24 V to VS; input current ≤ 25 µA).
  • Connect a 10 kΩ series resistor from J1.5 (LI1) to either /12V_SW (LI high) or GND (LI low). At ≤ 25 µA input current the resistor drop is ≤ 0.25 V, so levels stay valid; the resistor limits current if mis-wired.
  • Expected: TP3 (LO1) high (~3.3 V) with LI high, low with LI low. Repeat with J1.8 → TP7 (LO2).
  • Never connect 12 V (or the 10 kΩ lead) to any TP1–TP3/TP5–TP7 test pad or Pico pin — those are 3.3 V logic. Only J1.5/J1.8 receive the LI stimulus.
Step 4 — Wi-Fi/BLE and VSYS ripple observation (firmware-dependent)
  • Enable Wi-Fi TX / BLE. Input current may rise to roughly 100–200 mA at 12 V (estimate); raise the limit to ~300–500 mA if needed. Persistent CC mode = stop.
  • Observe VSYS ripple at D3:K with a short ground spring to a nearby GND pad (respect grounding rules). Informational only — no pass threshold set. Record peak-to-peak; VSYS must stay above ~4 V (provisional, well above Pico 1.8 V min). Check practical range (accepted antenna compromise).
Step 5 — Vehicle (optional, engine off, after bench pass)
  • Fuse fitted (1 A), short leads, grounding assessment done first (see Equipment & safety). No running-engine/cranking/jump-start use claimed.
Sanity review
Independent smart-friend review was not available in the original Phase 10 session; the later qv review findings have been applied in this revision. Manual checks: power-on order (unpowered → powered with Pico fitted → USB/firmware → one channel at a time → wireless) limits exposure; current limits are supply protection set below the 1 A fuse; scope grounding restricted to board GND after grounding assessment. No hardware results exist yet.
  • Connector / test-pad map (verified against live netlist, 18 nets, 2026-10-22)

  • Equipment & safety

  • Step 0 — Unpowered checks

  • Step 1 — First power (Pico already fitted)

  • Step 2 — USB and firmware (prerequisites)

  • Step 3 — Channel tests (one channel at a time, firmware-dependent)

  • Step 4 — Wi-Fi/BLE and VSYS ripple observation (firmware-dependent)

  • Step 5 — Vehicle (optional, engine off, after bench pass)

  • Sanity review