What happens
The sot723 footprint places its pads at the correct positions, but assigns the pin numbers 1 and 3 to the wrong pads:
fp.string("sot723").circuitJson()
| Pin |
Actual (current main) |
Expected (KiCad / Toshiba datasheet) |
| 1 |
(+0.575, 0) — alone on the right |
(-0.575, +0.4) — top-left |
| 2 |
(-0.575, -0.4) — bottom-left ✅ |
(-0.575, -0.4) — bottom-left ✅ |
| 3 |
(-0.575, +0.4) — top-left |
(+0.575, 0) — alone on the right |
So pins 1 and 3 are swapped. A real SOT-723 part placed on this footprint gets cross-wired (e.g. for a transistor, base and collector swap).
Expected
The standard 3-lead SOT arrangement, which every other 3-pin SOT in this repo already follows (sot23, sot23w, sot323): pin 1 top-left, pin 2 bottom-left, pin 3 alone on the right.
This is also exactly what KiCad's official SOT-723.kicad_mod does (pad 1 at (-0.575, -0.4) in KiCad's y-down coords = top-left, pad 3 at (0.575, 0)). The KiCad footprint references the Toshiba datasheet, and the pad size (0.45 × 0.4) and pitch (±0.575) in src/fn/sot723.ts match it exactly — only the numbering differs.
Why the KiCad parity test doesn't catch it
tests/kicad-parity/sot723_kicad_parity.test.ts compares the copper shapes via boolean difference, which is blind to pin numbers — swapped numbering produces identical copper.
(Found while comparing pin-numbering conventions across sibling footprints, same class of bug as #677.)
What happens
The
sot723footprint places its pads at the correct positions, but assigns the pin numbers 1 and 3 to the wrong pads:(+0.575, 0)— alone on the right(-0.575, +0.4)— top-left(-0.575, -0.4)— bottom-left ✅(-0.575, -0.4)— bottom-left ✅(-0.575, +0.4)— top-left(+0.575, 0)— alone on the rightSo pins 1 and 3 are swapped. A real SOT-723 part placed on this footprint gets cross-wired (e.g. for a transistor, base and collector swap).
Expected
The standard 3-lead SOT arrangement, which every other 3-pin SOT in this repo already follows (
sot23,sot23w,sot323): pin 1 top-left, pin 2 bottom-left, pin 3 alone on the right.This is also exactly what KiCad's official
SOT-723.kicad_moddoes (pad 1 at(-0.575, -0.4)in KiCad's y-down coords = top-left, pad 3 at(0.575, 0)). The KiCad footprint references the Toshiba datasheet, and the pad size (0.45 × 0.4) and pitch (±0.575) insrc/fn/sot723.tsmatch it exactly — only the numbering differs.Why the KiCad parity test doesn't catch it
tests/kicad-parity/sot723_kicad_parity.test.tscompares the copper shapes via boolean difference, which is blind to pin numbers — swapped numbering produces identical copper.(Found while comparing pin-numbering conventions across sibling footprints, same class of bug as #677.)