AllSpiceMirrors/incutec-OpenFC-Lite
0
Fork 0
mirror of https://github.com/incutec-hw/OpenFC-Lite.git synced 2026-08-10 06:15:48 +00:00
Files Issues Projects Releases
Files
main
incutec-OpenFC-Lite/hardware/docs/REV3_CHANGELIST.md
JustStan d80bf9baaf Fix contradictions and duplicated information 
Docs contradicted the design files and each other. Verified every claim
against hardware/*.kicad_sch and hardware/OpenFC.kicad_pcb before writing.

Contradictions:
- DESIGN.md: document that the 1.8 V gyro LDO is U15 LP5912-1.8DRVR in the
  schematic but U6 NCV8187AMT180TAG on both the board and the production
  panel, so fabricated boards ship with the NCV8187. Logged as an open Rev 3
  decision with the schematic-only pads J45/J46.
- EMC_CHECKLIST.md: LMR51420/30 -> LMR51430 (U3/U4 are LMR51430YFDDCR,
  C5219261); NCV8187 -> part-agnostic "1.8 V gyro LDO"; USB series R
  27 ohm -> 30 ohm with the real refs R10/R11 (R7/R8 are 1k ADC filters).
- rp2350_layout_notes.md: rewritten with verified refs. The old ref map was
  stale throughout (R3 is a 5.1k CC pulldown, not the VREG_AVDD filter;
  that is R12 30R + C21). Core cluster is C19, L1, C20, R12, C21, C5.
- audit_design.py: docstring said KiCad 9, the design files are 10.0.
- README.md: the project is not fully self-contained. Of 168 footprints,
  79 are project-local and 89 come from KiCad stock libraries
  (Capacitor_SMD, Resistor_SMD, LED_SMD, TestPoint, Package_SON); the
  schematics use Device/power/Connector_Generic too.

Duplications, each fact now lives in one place:
- Rev 2 rework pin mapping: REV3_CHANGELIST.md owns it, README status and
  DESIGN.md firmware link to it.
- IMU BMI270 vs LSM6DSV16XTR: DESIGN.md IMU section owns the as-built fact,
  REV3_CHANGELIST.md owns the open decision, README row shortened.
- Fabrication export procedure: README Manufacturing owns it.
- Sibling board sentence: README owns it, DESIGN.md keeps only the delta.
- DESIGN.md "Variants and revisions" split, revision content is in
  "Revisions" only.

Deleted, regenerable or dead:
- hardware/tools/design_audit.md: generated dump of a design generation that
  no longer exists (MCU U36, IMU U14, LMR51420 bucks, +12V rail, Card2
  TF-021B-H265), and it raised a false blocker claiming the CURRENT ADC has
  no bypass cap when R6 1k + C15 100n sit at U2 pad 49.
- hardware/emc.json, hardware/thermal.json: stale generated dumps, 37.5 mm
  and 217 vias against a board that is 38.9 x 38.9 mm with 634 vias.
- hardware/_filltest.kicad_pro: orphan project, no matching board.
- hardware/tools/eco_strip_sheets.py: strips sheets that no longer exist and
  would delete blackbox.kicad_sch.
- hardware/tools/rebuild_blackbox.py: one-time recovery that would regress
  the card slot to Card2 TF-021B-H265.
- hardware/tools/add_emc_note.py: never applied, duplicates EMC_CHECKLIST.md
  including the wrong 27 ohm figure.
- images/.gitkeep: directory now tracks four real files.

The three deleted dumps are added to .gitignore.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-06 10:55:07 +02:00

1.8 KiB
Raw Permalink Blame History

Rev 3 Changelist - OpenFC-Lite

Running list of changes for the next hardware revision, collected during Rev 2 bring-up and flight testing. Schematic/PCB edits are made by the maintainer. Items are appended as found; nothing here is implemented yet.

Rev 2 bring-up context: OpenDrone-Testing bring-up logs (2026-08-05 onward).

Confirmed defects to fix

  1. GND shorts on GPIO0-3, GPIO28, GPIO31 (Rev 2, KiCad-induced shorted traces on top + bottom copper). Kills stock UART0/VTX connector, GPS pads, M1/M4 motor pads; Rev 2 boards fly with the rework pin mapping baked into the Betaflight target (OPENFC_LITE_RP2350B/config.h). Root-cause the KiCad artifact in the Rev 2 layout and add DRC coverage that would have caught it. Fixing this restores UART0, GPS pads, M1/M4, UART1, I2C0 to their intended pins.
  2. Gyro LED mounted 180 degrees wrong (reversed polarity): does not light. No functional impact otherwise. While fixing: verify orientation of all power-line indicator LEDs against the footprint/courtyard markings; check whether the silk/footprint invited the wrong orientation.

Decisions needed

  1. Production IMU: Rev 2 boards are assembled with BMI270 while KiCad/BOM still says LSM6DSV16X; footprint also accepts ICM-426xx/456xx. Decide after flight testing, then align schematic, BOM, and Betaflight target defaults.
  2. 1.8 V gyro LDO, schematic and layout disagree: power.kicad_sch has U15 LP5912-1.8DRVR (WSON-6, C2876234), both OpenFC.kicad_pcb and the production panel still have U6 NCV8187AMT180TAG (WDFN-6). Fabricated boards get the NCV8187. Pick one part, then update the other file so the reference designator and package match everywhere. Same pass: the schematic has solder pads J45 and J46 that are not placed on the PCB.