Skip to content

EX segment resolved as absolute segment number — tag field ignored, wrong feed point on multi-wire decks #17

Description

@stevenmburns

The EX card's segment is resolved as an absolute segment number — the tag field is ignored — so any deck whose feed is not on the first wire is silently excited at the wrong point. LD's tag mapping is correct (verified separately with a two-wire discriminator deck), so this is specific to the excitation path.

Minimal reproduction

Three collinear 25-segment wires forming one 10 m dipole, fed at the true center (tag 2, its segment 13 = absolute segment 38):

CE 3-wire chain, fed middle of tag 2
GW 1 25 0 -5 10 0 -1.6666 10 1e-3
GW 2 25 0 -1.6666 10 0 1.6666 10 1e-3
GW 3 25 0 1.6666 10 0 5 10 1e-3
GE 0
FR 0 1 0 0 14 0
EX 0 2 13 0 1 0
XQ
EN
deck onec (b58c4ad) nec2c 1.3
EX 0 2 13 (center feed) 184.811 −199.602j 64.123 −55.120j
EX 0 1 13 (off-center feed) 184.811 −199.602j 184.810 −199.600j

onec returns the off-center answer for both — and its ANTENNA INPUT PARAMETERS row prints TAG 1 SEG 13 for the tag-2 request:

TAG  SEG.
  1   13  1.00000E+00 ... 1.84811E+02-1.99602E+02

i.e. the source landed on absolute segment 13 instead of tag 2 / segment 13.

Impact

Any multi-wire deck fed via a tag other than the first wire silently produces wrong results with no diagnostic. In the LD-corpus sweep for #11 this accounts for essentially all of the 258/447 diverging decks — split-element models (e.g. Cebik tutorial decks with a fed 1-segment bridge, EX 0 3 1) come out 100×+ off; driven-element beams ~2× off. Together with #16, a wild deck currently needs XQ and a feed on wire 1 to be trustworthy.

A golden deck with a tag-2 feed in the #15 CI matrix would pin this permanently.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions