Closes#2320. A different, Manchester-coded FSK protocol from the older
Jansite Solar TPMS in tpms_jansite_solar.c (different preamble, no CRC) --
confirmed a genuinely different wire format, not a decode bug in the
existing decoder.
Verified against 15 real messages spanning a slow 1.6-2.8 bar pressure
ramp on one real sensor: temperature matches exactly on all 15; pressure
matches within one 2.5 kPa sensor step on all 15 (display rounding, not a
decode error). No real CRC exists; a tail-byte-repeat check is used as
the integrity check instead.
Sensor ID and battery status are not decoded: doing so needs samples from
a second physical sensor, which has never been posted to the issue.