fix(bandplan): correct ARRL US plan license-class accuracy (10m Tech + 80/40/15m Tech CW + 80m CW edge + 60m channels)

[jensenpat]

Summary

Fixes #3506. Supersedes #3507 with a full-file accuracy pass verified against 47 CFR §97.301/§97.303 and the ARRL band plan (arrl.org/band-plan), plus new VHF/UHF coverage and dead-code removal.

Changes

10m (the #3506 report): Technician privileges end at 28.500 MHz.

• Added T to all segments in 28.000–28.500, including the 28.225–28.300 beacon segment that fix(bandplan): correct 10m Technician privileges in ARRL US plan (#3506). Principle IV. #3507 missed
• Split PHONE at 28.500 and removed T from everything above (28.5–29.0 PHONE, AM, SAT, repeater in/out, FM call)

80/40/15m Technician CW privileges (same class of bug, previously missing entirely):

• Added T to segments within 3.525–3.600, 7.025–7.125, and 21.025–21.200 (§97.301(e), CW only, 200 W PEP)

80m General CW lower edge:

• General CW starts at 3.525, not 3.510 — split the old 3.510–3.570 segment so 3.510–3.525 shows Extra-only

60m:

• Removed the nonexistent 5351.5 kHz channel (a stray from the WRC-15 international allocation the FCC has not implemented). The five US USB dial frequencies remain: 5330.5 / 5346.5 / 5357.0 / 5371.5 / 5403.5 kHz

2m and 70cm (new — plan previously ended at 6m, so VHF/UHF panadapters and transverter use had no band plan data):

• 144–148 MHz: EME, CW/weak-signal, SSB, 144.275–144.300 beacons, both OSCAR satellite sub-bands, linear translator in/out, FM repeater input/output pairs, simplex windows, 145.50–145.80 experimental
• 420–450 MHz: ATV, 432 MHz EME/CW/weak-signal with 432.3–432.4 beacons, 433–435 aux links, 435–438 satellite-only, repeater/simplex local-option segments (overlapping 438–444 ATV vs 442–445 repeater rendered non-overlapping by splitting at 442)
• New spots: 144.200 SSB calling, 146.52 / 446.00 national FM simplex, 432.100 SSB/CW calling, 2m FT8 / MSK144 / WSPR
• All segments E,G,T — every license class has full 2m/70cm privileges

Dead code removal (per review):

• Deleted src/models/BandPlan.h — the compiled-in kBandPlan[]/kBandSpots[] tables replaced by the JSON-driven BandPlanManager in Customizable Band Plan Overlay #425. Nothing includes the header; its stale data was now multiple corrections behind this file

Verification

• JSON validated with python3 -m json.tool
• HF segment edges cross-checked against §97.301 (license classes) and §97.303(h) (60m)
• 2m/70cm segments cross-checked against arrl.org/band-plan
• BandPlan.h deletion verified: no #include references, no build-system references, no symbol uses outside the deleted file

💻 Generated with Claude Code (Fable 5.0) with architecture by @jensenpat

[aethersdr-agent]

Reviewed the diff against 47 CFR §97.301/§97.303 and how the file is consumed (BandPlanManager::loadPlanFromJson and the #2649 license-class filter in the ATU pre-tune dialog). The data checks out — nice, careful pass. Verified specifically:

• 10m: T on everything 28.000–28.500 (including the 28.225–28.300 beacon segment #3507 missed) and removed above 28.500 — matches §97.301(e). The 28.300/28.500 PHONE split is contiguous with no gap/overlap.
• 80/40/15m Tech CW: 3.525–3.600, 7.025–7.125, 21.025–21.200 all match §97.301(e).
• 80m General edge: 3.510–3.525 Extra-only split is correct (General CW starts at 3.525 since the 2006 omnibus).
• 60m: the removed 5351.5 dial was indeed never a §97.303(h) channel; the five remaining segments match the US dial frequencies (center − 1.5 kHz).
• Parser-wise the format is fine: license is a comma-split string match, segments only need low < high and a valid color, and "version" isn't read by anything, so no bump needed.

Two non-blocking notes:

1. Stale duplicate band plan in src/models/BandPlan.h — the compiled-in kBandPlan[]/kBandSpots[] tables still carry the old, now-known-wrong data (including a // Tech privileges start at 28.300 comment). They're dead code (nothing includes the header; BandPlanManager.h notes it was replaced in #425), so this PR doesn't need to touch it, but it's now two corrections out of date. Worth a small follow-up to delete the header so it can't be reintroduced as a "reference."

2. T on DATA-labeled segments (3.570–3.600, 7.040–7.125, 21.070–21.150) — this is the right call for the file's semantics: the license field drives the ATU pre-tune class filter (#2649), where it means "this class may transmit in this range," and CW is authorized anywhere a class holds privileges. Just flagging the inherent display caveat: a Technician reading the panadapter could infer data-mode privileges on those segments, which §97.301(e) doesn't grant (CW only there). The format can't express per-mode class limits, so nothing to change here — only worth remembering if segment tooltips ever grow license detail.

Behavioral side effect worth having on the radar for QA: a Technician-filtered ATU pre-tune will now generate tune centers on the 80/40/15m CW/data ranges and 28.0–28.5, where it previously skipped them. That's correct (Techs hold transmit privileges there), but it does change where the carrier lands for that filter setting.

Thanks @jensenpat — this is a much more complete fix than #3507, and the verification notes in the description made it straightforward to check.

---

_{🤖 aethersdr-agent · cost: $15.6746 · model: claude-fable-5}