• I found that if you increase AHRS_RP_P param from 0.1 to 0.4 (means ACC more weight in attitude estimation than gyros) mitigates this issue. With AHRS_RP_P = 0.1 APM went out of control during a long 28 degree turn, now in 0.4 I can take long turns without lost of control due pitch wrong estimation. The problem still there, at least is contained...? testing at lower speed with small radio turns APM loose the control again... 

  • Hi guys, I have the same issue regarding bad AHRS pitch estimation due a long turn in FBW-A and Stab. My enviroment: FlightGear + HIL + Mission Planner 1.2.93.

    take a look to this video I made: 

  • 2 Developers

    may be inertial velocity may be better then  clean "gps or airspeed" for level corrections ?


     Vector3f                _velocity;                  // latest velocity estimate (integrated from accelerometer values) 

    or accel acceletation vector and dt

    as AP_InertialNav.cpp line 89
    // calculate velocity increase adding new acceleration from accelerometers
        const Vector3f &velocity_increase = (accel_ef + accel_correction_ef) * dt;
  • Moderator

    I have similar problem with plane descending with apparent nose up, or loosing altitude in turns because plane thinks nose is pitched up.

    Something is not quite right with AP v2.76, we've been flying two X8's, an X5 and a RVJet, all show similar behavior and cannot be trusted.

  • log file for this video


    log file 2013-12-14 08-14.log

    line 14865  mode rtl

    pitch up

    gps alt down

This reply was deleted.