[volkermuehlhaus-claude]

The center-tapped fit sorted the skin sections of each half coil by corner frequency on its own. With skin elements of the two halves tied (e.g. inductances tied, resistances free, as on the bundled coil), the halves could sort differently and pair section 1 of one half with section 2 of the other, so the reported values broke the tie. Half 2 now takes half 1's order whenever skin elements are tied. The circuit itself was unaffected (the series sections of a half commute).

Reported by the snp2le port (iic-jku/snp2le#30).

Verification

• Function-level test: with inductances tied and half 2's corners in the opposite order of half 1, 1.2 breaks the tie (Lskin1_h2 != Lskin1_h1), 1.2.1 keeps it; without ties both versions sort identically.
• Not reproducible in a real fit: three synthetic center-tapped coils built with reversed corner order in half 2 were fitted with the same order in both halves (the data does not fix the order).
• All four bundled examples: .txt and _model.sp identical to 1.2 apart from the version string, so the example files are not regenerated.

🤖 Generated with Claude Code