Skip to content

Yaesu FTM-300DR support via microSD (#48) - #140

Merged
ww8l merged 2 commits into
mainfrom
issue-48-ftm300
Oct 12, 2026
Merged

ww8l merged 2 commits into
mainfrom
issue-48-ftm300

Conversation

@ww8l

@ww8l ww8l commented Oct 12, 2026

Copy link
Copy Markdown
Owner

Closes #48.

Programs the Yaesu FTM-300DR from its own microSD backup (FTM300D/BACKUP/CLNFTM300D.dat): the app patches the file in place (keeping CLNFTM300D.dat.orig), and the radio loads it with SD CARD → BACKUP → READ FROM SD → ALL.

What's in it

  • Memories: radios/yaesu_ftm300/. 999 records, names, and the name-length code. The layout started from CHIRP's ftm300d.py and was corrected against the radio: the offset is a u16, byte 11 carries the user CTCSS, the name-length table is a length code, and DCS is 0-based.
    • Re-encoding Tim's own 119 channels: 88 come out byte-identical, and the other 31 differ only in skip, M-GRP, clock shift and VW, which have no database column.
    • Through the dev export path: 61 of 62 channels written (the D-STAR channel excluded), and nothing outside the memory tables moved.
  • Settings: 147 fields. One spec generates both the Rust table and the form schema. They were measured with RT Systems writing the SD file: a 16-pass bit signature, an enum sweep and a text pass.
    • Decoding Tim's file reproduces what RT Systems displays for it.
    • Five fields where RT Systems and CHIRP disagreed were settled on the radio's own screen: Band Scope, Sub Band Mute, Date Format, Digital Popup "Continue", and Com Port speed/output.
  • M-GRP: a per-codeplug channel checklist on the codeplug page.
    • Stored in codeplug_memory_group (migration 0020).
    • Reaches file exporters as ExportRequest::memory_group.
    • Written as the record's M-GRP bit.
  • Wiring: card discovery, Program dialog steps, the profile-editor settings reader, the seed row and the README.

Hardware (Tim's FTM-300DR)

  • ✅ Identity read
  • ✅ One-byte name write: the radio accepted the edited file, so there is no checksum
  • ✅ Full 61-memory codeplug written through the app's Write to SD card
  • ✅ Band probe
  • ✅ Settings spot checks
  • ⬜ M-GRP after an app write: not yet checked on the radio. Tim will file an issue if it doesn't work.

Known gaps

  • Airband 1.67 / 6.67 kHz offsets display 417 Hz off; the radio's fine-step code for these isn't measured yet.
  • Com Port 57600 is withheld.
  • Settings not exposed: FM bandwidth / RX mode, Target Location, Clock Type and GPS Device. They aren't in RT Systems' dialog.

🤖 Generated with Claude Code

ww8l and others added 2 commits October 11, 2026 20:28
Programs the FTM-300DR by patching the radio's own SD backup,
FTM300D/BACKUP/CLNFTM300D.dat, which READ FROM SD -> ALL loads back.

- radios/yaesu_ftm300: 999 memories in 16-byte records with a parallel
  name table and name-length code. CHIRP's ftm300d.py record, corrected
  for this radio: u16 offset, user CTCSS bits, name-length code, 0-based
  DCS. Unused slots get the radio's own empty record; Home, PMS, VFO,
  APRS and the 0x400 block are carried through untouched.
- 147 Set Mode settings, generated with the form schema from one spec,
  measured with RT Systems writing the SD file (16-pass bit signature,
  enum sweep, text pass). Five fields where RT Systems and CHIRP
  disagreed were settled on the radio's own screen.
- M-GRP: per-codeplug channel checklist on the codeplug page
  (codeplug_memory_group, migration 0020), carried to file exporters as
  ExportRequest::memory_group and written as the record's M-GRP bit.
- Card discovery, MEDIA_WRITES steps, settings reader for the profile
  editor, seed row, README.

Hardware: identity read, one-byte name write (no checksum), full
61-memory codeplug through the app, band probe and settings spot checks
all passed on Tim's radio. M-GRP after an app write is not yet checked
on the radio.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Airband: encode the measured +8.33 kHz point as 0xD (radio displays it
  exactly) and refuse the unmeasured 8.33 kHz points with a clear error
  instead of writing a frequency the radio shows 417 Hz off.
- Text settings: a call sign the field cannot hold is refused, never
  stripped ("WW8L-9" no longer writes "WW8L9").
- Unknown enum values decode as "#N" and round-trip as themselves, so a
  numerically labelled enum (DTMF channel "1".."9" = raw 0..8) cannot
  remap one.
- M-GRP: count, checklist and export share one set -- members still in
  the codeplug; read only for formats that declare a memory group
  (CodeplugExporter::has_memory_group) and only on that radio's page.
- Shared radios/card_file.rs for the in-place + .orig-once rule (FT5D
  and FTM-300), shared codeplug channel-pool helper for both checklists.
- Real-file settings-only test: own settings = 0 bytes, one flip = 1.
- README: 147 settings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@ww8l
ww8l merged commit ebe1e32 into main Oct 12, 2026
3 checks passed
@ww8l
ww8l deleted the issue-48-ftm300 branch October 12, 2026 03:17
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Add Yaesu FTM-300 Support

1 participant