with the new ChibiOS revision we were configuring for newer revisions
of the H7, which meant that we changed the ADC config. This broke ADC
readings on revX and revY H7 MCUs.
This PR fixes it in two ways:
1) change ChibiOS config to assume XY config chips
2) use 16 bit ADC for H7, which means the chip rev doesn't matter,
and also gives us 16x the resolution for ADC readinga, so we can
read smaller voltage and current values
set_wp_destination clear yaw target
limit velocity and acceleration based on track slope
add is_active() and remove unused reached_spline_destination
init accepts desired speed
set_kinematic_limits uses current speed limits instead of defaults
add time compression to prevent target moving too fast for air
implement alternative spline
remove vel-target-length
set_wp_destination always calculates this leg
set_kinematic_limits moved to scurve
fix origin speed after spline segment
spline terrain following fix
handle s-curves with mismatching alt types
fix set_spline_destination_next
add update_track_with_speed_accel_limits
Change to next waypoint at corner apex
use scurve advance along track
remove unused definitions and out-of-date todo
set_spline_destination_next sets fast_waypoint
scurve origin speed set from spline target velocity
fixup takeoff delay
SCurve const more local variables
rename update to get_jerk_vel_pos_at_time
removed unused update
debug output in SITL
fixes to ensure finished at end of path
fixes including validity check
includes corrections from peer review
this is intended to catch problems where we have unexpectedly received
invalid arguments. For example an out-of-bounds target speed that
should have been caught somewhere else.
running out of bits so we use a single bit to capture invalid inputs to or output from a calculation
this fixes an issue with bad read of WHOAMI on a mRoPixracerPro. We
don't know how the sensor gets in a state where WHOAMI can't be read,
but toggling CS does fix it