chore: BOM mass audit against vendor datasheets (2 errors found in v1.4.1 — likely more) #12

Background

v1.4.1 spot-checked hardware/bom/by_form_factor/cansat_advanced.csv and corrected two entries:

• LiPo 2S 1500 mAh: 45 g → 80 g (real pouch pack)
• HS-35HD servo: 9 g → 4 g (Hitec datasheet)

The other nine BOM files (cansat_minimal, cansat_standard, cubesat_1u through cubesat_12u) were only spot-checked on structurally-obvious entries (battery cell counts, solar cell totals). Individual component masses were not cross-referenced with datasheets.

Ask

Walk through every row in every hardware/bom/by_form_factor/*.csv and verify the Mass (g) column against the manufacturer datasheet for that part number. Correct anything off by more than ±20 %.

Candidates worth checking based on pattern-matching (not definitively wrong — just worth verifying):

• STM32F446RE Nucleo board — listed as 30 g in several files. Actual (ST datasheet): 33 g. Probably fine.
• NCR18650B cell — 50 g each (4× = 200 g in 3U). Panasonic datasheet: 48.5 g. Close.
• SG90 servo — 9 g. Datasheet: 9 g. ✓
• MLI blanket masses — quoted per-layer, worth sanity-checking against typical aerospace datasheets.
• Antenna masses (patch / whip / dipole) — varied heavily, worth a look.
• Reaction wheels in cubesat_3u / 6u / 12u — custom part, no vendor datasheet, but the quoted masses (50 g for micro, 100 g for mid, 400 g for large) should roughly match Blue Canyon / Hyperion / CubeSpace commercial catalogues.

Definition of done

• Every row in every file either matches its datasheet within ±20 % or has a Notes column explaining the deviation.
• Any row whose number changes bumps the TOTAL row accordingly.
• hardware/bom/by_form_factor/README.md bare-kit column resynced if any total moved.
• Commit message links each correction to the datasheet URL.

Good first issue for someone new — mechanical / deterministic, no design decisions, high leverage.