Auto & Loiter Mode always Losing Altitude - Log Attached - ArduPlane User Group

Hi All ,

Hardware:

- Flight controller : PIXHACK

- GPS Ublox

- MiniOSD

Software:

- GCS : Mission Planner.

- Firmware : Arduplane 3.4

Problem:

After previous tests and posted topics here I solve the problem of pitching up and stall ,

now the plane is very stable and work well in manual and stab modes , even when I off the motor during the flight its very stable and decent with glide ,

But when try loiter mode , and Auto mode , It works good and direct to WP or start loiter ,

previous tests the throttle cut , now the throttle is off and start many time and the plane loss the altitude while trying go to waypoint , until it become near to ground I change to stab and control it to gain altitude and try again always decent ,

I am not using airspeed and my cruise throttle in set to 50% ,

What's the algorithm that control the altitude (GPS , Baro , TECS ) ?

which parameter I should change to keep altitude ?

how dose TECS control altitude ?

here is the Log file , in .Bin format Log File

Hope I can solve this problem I want to enjoy the auto modes :(

best regards..

@Abu,

altitude is controlled by sensors, software processing sensor data controlled by EKF.

As long as Powwer Control sensor is not implemented we are lost, since no direct way

to guess if altitude goes up or down if you fly at higher altitude.

Power Control Sensor is smart and inituitive.

If more energy is sent from battery to motors, there is a chance for your drone to fly up

if less, there is no chance for drone to fly up.

Easy, inituitive, fuzzy logic controller to work every time, not controlled by EKF.

GPS ionospheric error can be 5-10 metres horizontal, hard to guess vertical.

If airframe starts to resonate, IMU can generate pseudo data not useful to

control your drone.

The same for barometric sensor to generate pseudo data on airframe vibrations.

It looks like every drone should be tested on the ground on vibrating table before first take off.

Replies

Darius Jack January 25, 2016 at 3:26pm

@Abu,

altitude is controlled by sensors, software processing sensor data controlled by EKF.

As long as Powwer Control sensor is not implemented we are lost, since no direct way

to guess if altitude goes up or down if you fly at higher altitude.

Power Control Sensor is smart and inituitive.

If more energy is sent from battery to motors, there is a chance for your drone to fly up

if less, there is no chance for drone to fly up.

Easy, inituitive, fuzzy logic controller to work every time, not controlled by EKF.

GPS ionospheric error can be 5-10 metres horizontal, hard to guess vertical.

If airframe starts to resonate, IMU can generate pseudo data not useful to

control your drone.

The same for barometric sensor to generate pseudo data on airframe vibrations.

It looks like every drone should be tested on the ground on vibrating table before first take off.
- Abolo > Darius Jack January 26, 2016 at 4:28am
  
  @ Darius ,
  
  What EKF Parameter I should change ?
  
  power control sensor , I am using is
  
  CUAV IV sensor ver.3.04
  
  DC IN 2-7S Li-Po 5.4v 3.0A OUT
  
  like photo below
  
  And the battery is lipol 24 v 5300 mah
  
  but i can't read the sensor information throw the mission planner or OSD
  
  what should I change in parameter ?
- Darius Jack > Abolo January 26, 2016 at 8:00am
  
  @Abu,
  
  please read this first
  
  http://plane.ardupilot.com/wiki/tecs-total-energy-control-system-fo...
  
  http://copter.ardupilot.com/wiki/ekf-inav-failsafe/
  
  http://copter.ardupilot.com/wiki/common-apm-navigation-extended-kal...
- Abolo > Darius Jack January 26, 2016 at 10:33am
  
  @ Darius ,
  
  thank you for information , I read it and will Enable it
  
  the above link is for the copter ,
  
  AHRS_EKF_USE = 1
  
  EKF_GPS_TYPE = what option should I use as My GPS is ublox 7m ?
  
  EKF_ALT_NOISE = what is better the accelerometers or barometer,?
  
  do you think this setting will control the altitude ?
  
  what about the power module ? and how to know motor energy ?
Bruce Baxter January 25, 2016 at 9:57am

How did you solve the pitch up?
- Abolo > Bruce Baxter January 25, 2016 at 11:25am
  
  Bruce Baxter actually I did many steps until it solved and I am not sure which one solve it.
  
  - correct the C G point .
  
  - limit the pitch and roll angle ,
  
  - fly with no wind ,

of 2

This reply was deleted.