Dev Call July 5, 2021

Issues & Pull Requests





Attendee count (max): 20


  • Support for getting the HIGH_LATENCY2 message
  • Ground station support?
  • Several minor things to fix
  • Stephen is planning on working on the actual high-latency protocol next
    • What messages?
    • Only thing the UAV would send would be the high-latency message
    • Ground control station would rate-limit things up
    • Iridium SBD modem
    • Incorporate command-ack into high_latency2
      • Not really doing commands over the link so probably not worth it
    • Packetize response with the high_latency2 message?
      • So coalesce messages in the buffer and only send when bytes over a threshold
        • Or a timing threshold expires
      • What does the framing?
        • Depends on connection to modem
          • Modems have configuration
          • CC on other end of link can do heavy lifting
        • Connecting to the modem is what we should be aiming for

UTC2321 -

  • Merged

UTC2321 -

  • Fix DJI temp and RPM
  • Brief(?) diversion into SCurve navigation internal error
  • We should consider building with debug symbols
  • And get the dumpstacks
  • Effects of this?
    • Maybe just sit there and do nothing?
  • Dying in SITL is bad?
    • Can’t see what the vehicle would have done?

UTC2350 -

  • remove GPS status check for camera trigger
  • We shouldn’t be asking the GPS in here
  • EKF’s Absolute horizontal position instead?
  • Tridge thinks we should continue to take pictures as does Randy
  • Add a “valid” field?
    • Log the GPS status
  • We would lose the instantaneous feedback to the GCS if we continued to take photos when we don’t have a good location
  • Parameter for min gps status for taking a photo?

UTC0007 -

  • Ability to compile out deep-stall landing
  • Can be merged once the comments have been fixed

UTC0009 -

  • Merged
  • LEDs on -bdshot matekf405

UTC0010 -

  • Testing yaw control on tradheli is a problem because we’re zeroing yaw on the ground
  • Remove FF?
  • Significant yaw angle error on ground can crash vehicles, thus the current code
  • Leonard is thinking can do this when spin-armed or when disarmed
  • FF is very small on Copter’s - mostly it is P gain
  • Is the problem in pilot input or deeper?
  • Ignore yaw input after rudder-arming until it is centered?
  • Added to 4.1 list
  • Probably affects tilt-vectored

UTC0011 -

  • Low throttle stick on qloiter inhibits yaw control
  • Dead-zone of lowest throttle means you don’t get any yaw control
  • Don’t want to yaw on ground….
  • Maybe look at descent rate
  • Lift on hill can actually cause same problem on multis
    • And helis
    • Very common in quadplanes, however
  • Tailsitters yawing on ground is really bad
  • Maybe document qloiter to get yaw you need >0 throttle
  • A patch to check descent rate is coming

UTC0039 -

  • Plane beta3 is out now!
  • In-sync with Copter
  • Probably only minor fixes for now
  • Logging with forced arm fixed
  • D-shot and motor test fixed
  • D-shot buzzer was enabled by default
    • Tridge has turned it off by default
    • So no tones from motors
      • Mucked up motor test
  • D-term issues with new filters
    • Default values backed out to older more conservative values

UTC0048 -

  • Some good testing going on
  • Beta5 was a big release fixing a lot of issues
  • Some new issues
    • Guided mode update to go in
    • Heli rudder issue
    • Runcam split 4 start/stop not working
    • Some oscillations when using GPS for alt
      • EKF alt going up and down when on table
      • GPS_DELAY_MS wrong, perhaps?
    • Reports of I2c lidars not working on CubeOrange when CAN GPSs used
    • Irlock connected over i2c makes uavcan GPSs LED flash in its patterns
      • Maybe GPS PPS?
      • Maybe rapid brownouts on the flightcontroller
        • Similar to the tick-tick-tick from flight controllers?
          • Violating USB spec and drawing too much power
        • More of a “tut-tut-tut” – tridge

UTC0059 - Rover update

  • GPS-for-yaw still an issue

UTC0059 - GSoC update

  • Firmware builder is delayed by covid ATM
  • Arsh is having some waf issues
    • Will meet after call to discuss
  • Rishabh is doing well with the precision landing
    • Bunch of PRs created
      • Refactoring was merged
      • No functional change
    • Working on precland retry next
      • If it loses target it will climb back up again to try to regain target

UTC0102 -

  • Want to move Copter’s code up rather than this
  • Close