Files
grabowski d05d523995
Build PCM package / build (push) Successful in 7s
Release 1.2.2
2026-07-22 16:56:26 +07:00

41 lines
1.9 KiB
Markdown

Results are unchanged from 1.2.1 for the same board and settings. This
release is about what the plugin tells you while it works, and about no
longer overstating what a frequency result means.
Progress while solving:
- The dialog used to close on OK and leave nothing on screen until the
figures appeared - minutes, on a real board, with no sign the plugin
was doing anything. A small window now stays up for that whole
stretch: the stage running, elapsed seconds, and Cancel.
- It covers the figure work as well as the solve. Laying out labels and
writing the four PNGs at full resolution is seconds on a modest board
and 10-15 on a large one, and that used to be silent too.
- Cancel stops the solve and returns you to the board with no error
figure - the run simply reports that it was cancelled.
Frequency results are described honestly:
- Nothing advertises "AC resistance" any more. At f > 0 the plugin
applies the exact 1D foil and barrel skin-effect correction and
nothing else: proximity redistribution and inductance are not
modelled, so the number is a lower bound on the resistive rise, not
an AC impedance simulation. The README headline, the PCM and plugin
descriptions, the dialog note, the CLI help and the summary line all
say so now.
- The computation itself has not changed - only its description. A
frequency result from 1.2.1 is the same number, previously labelled
in a way that invited it to be read as an impedance.
Also in this release:
- The offline runner takes --progress, so the same busy window can be
used outside KiCad.
- The frequency field keeps its specific reason for rejecting an input
("1,500" is a thousands separator, "-5" is negative) instead of a
generic "cannot parse".
The in-KiCad |J| overlay push remains experimental and opt-in, off by
default. It writes reference images to User.9-User.12 and replaces what
is on those layers.