Dev Call July 6, 2020

Issues & Pull Requests




1 Like

Attendee count (max): 26

UTC1102 -

  • Some receives send both percentage and dbi
  • Ignore one of them
  • Just use percentage
  • merged!

UTC1103 -

  • Running telem at higher frequencies floods logs at 100Hz
  • 400 messages per second
  • One timing subtraction issue
  • Can be merged after that’s fixed

UTC1106 -

  • Options for acro
  • Pushing size out?
  • Leonard devised these when redoing attitude controller
  • Sysid without the controllers getting in the way could be goodfor heli
  • Bitmask vs one-or-the-other
  • Airmode mixed in with this as an option in acro
    • Also have
  • Yaw2 / yaw3?
    • Developers can look at variations
    • “Acro locking”
  • Air-mode works well for loops and whatnot
  • Want the option to disappear for heli
  • Arming-on-a-switch and heli?
    • What happens when airmode is activated?
    • Pre-existing
    • Airmode should be independent of modes
      • Just always stabilizing
      • Helis are always stabilizing anyway
  • Leonard to look at this
  • Remove extra options that won’t be popular
  • Bill wants to check this out for Heli implications
  • Randy is worried about user support on adding extra options
  • This allows people to test their rate loops but not their stabilisation loops
  • One acro here doesn’t rely on the EKF giving an absolute attitude
    • Normal one does
      • Robust to EKF performance issues
  • Advantages to each different acro mode
    • Easy way for people to play with own attitude stabilisation techniques
    • Leonard’s not sure if the third one is overhead we don’t need ‘though
    • One and three may be useful, two may not be
  • Bill’s really not sure about that air mode on tradheli

UTC1120 -

  • Ifdefs to control bidirectional stuff
  • Couple of CI issues
  • Extracted frsky and protocol decoder about a bit
  • On correct path here?
  • Does it build with bidirectional off?
    • Yes
  • Which thread are we processing these on?
    • IO thread function
  • Need to work on CI failures
  • Needs feedback from the community!
  • Might get alignment errors
    • Want to make sure we’re using appropiate be16toh etc
    • Navio target is useful for getting the warnings

UTC1148 -

  • Video - 8 minutes long
  • PH worked out why he added the function
  • Going from q mode to forward flight is instant transition
    • So better transitions?

UTC1153 -

  • Select backend rates
  • Needs it on his 3” quad as motor RPM goes above 500Hz so can’t drive notch filter higher than that (Nyquist)
  • Adds another parameter to set the gyro sample rate
  • Need a racing-quads setup page that mentions this sort of parameter
  • Fast sampling determines raw sample rate
  • This changes the rest-of-sample rate
  • Change in behaviour for F7/H7 due to faster sample rate
  • Fast sampling vs not fast sampling?
    • Don’t have a separate parameter by sensor
  • Another parameter for each INS to specify a rate for each sensor?
  • Need more feedback to users on what the final configuration is
    • Rates that are finally coming out like we do for PWM output types
  • In the dataflash log?
  • Text to user e.g. “IMU1:8.0kHz/2.0kHz IMU2:4.0kHz/1.0kHz”
  • Param name?
    • Backend rate / frontend rate

UTC0009 -

  • Generating duplicate documents for ATC
    • invalid file in edn
  • Either agree on ranges etc
  • Or add heli file as entire parameter set
  • The two frame types have significantly different parameters
  • Parameter versioning is done on the Wiki side
  • Check with MO to see if he’s OK with adding heli as a separate parameter set
  • Ask Don too

UTC0016 -

  • Decode DSM frame markers to avoid timing related jitter
  • No Andy
  • Tridge likes the scheme
  • Lots of different variants of Spektrum receivers
    • Can’t be certain this won’t break some Spektrum receivers
    • MdB will test on some receivers
    • NeedsTesting marker added
  • Needs IO firmware to be generated and committed

UTC0017 -

  • Merged
  • Fixes compiler warnings about unaligned accesses

UTC0023 -

  • Throttle hysteresis?
    • For ICE engines may need it to keep running?
    • 5% is not a small error
  • Uses deadzone
  • MdB is happy with on-by-default
  • MdB thinks we do need the 0% guarantee
  • Sprung throttle will need to hold it down
  • Throttle suppression stop it firing up ATM
  • Merged!
  • Should close

UTC0028 -


  • Makes it easier to do loops and rolls in qacro mode
  • Manual-mix bug was found
  • This isn’t switching correctly
    • Use enumeration for arming ArmingMethod and change behaviour based on that method
  • Update_throttle_suppression is something to be very careful around
    • E.g. in auto
      • Mode check required?
  • Review added

UTC0040 -

  • Board -> firmware mapping for users is problematic
    • Fmuv2 complicates matters
  • If you have a 1MB board but its not a Pixhawk1
    • This would be useful then
  • Fmuv2 probes for lots of different sensors
    • Pixhawk1-1M does not
  • Eliminating fmuv2 might stop some people being able to update
    • Not completely redundant
  • Would have to be one of the clones that as 1M
  • Tridge can’t think of any board which needs us to build this
  • People should generally be using Pixhawk1-1M when they’re running 1MB Pixhawks
  • Fmuv4 can be removed now; just use PixRacer

UTC0047 -

  • Merged!

UTC0049 -

  • Merged

UTC0050 -

  • Allow specification of task prioritisations
  • Enumeration?
    • Priority on every line
  • Pseudo-enumeration
  • Implicit incrementing means auditing all vehicles all the time
  • Last column should not be delta
  • Tridge would like an auto-increment of line before

UTC0120 - Plane update

  • RTL autoland still needs to be fixed
  • GCS FTP stack overflow patch
    • Should already be working correctly

UTC0126 - Copter update

  • Removing compass issues

UTC0127 - close

1 Like