APM in a plane (real)

Hello,

after a few weeks of work - APM 2.5 with interface modules for 2 MGL Servos was installed in real plane VL-3 (ultralight). Rudder - fixed position, throttle - controled by "man" pilot.

Main purpose - as a help for pilot to fly ortophotography mission over large areas.

Last week - was a "working" test - 5 days with more than 30 hours in air.

Longest track - 62 km, shortest - 8 km. Tracks total - 88. Average single flight distance - 400 km

on board - Phase One iXA180 medium-size camera with 2-axis stabilisation - \

3691124493?profile=original

wiecie.jpg

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –

Replies

  • How did you connect MGL servos to Apm?

    • Small STM32F103 module: PWM from APM -> TTL serial (MGL protocol) -> RS-232 module -> MGL Servo.
      Also working as "false" RC receiver - delivery PWM to APM inputs ( mode selection, etc)
  • 3701746265?profile=original

    Sample foto from 850m AGL (jpg compressed 50% from RAW file)

  • Sample log file:

    - in Auto mode only roll stabilisation by APM. Elevator and throttle - hand controlled, rudder - fixed position.

    - turns - by hand. During turns auto mode active, but power (+12) to servo switched-off.

    Suggestions for tunnig - highly appreciated :)

    2014-05-23 11-13-03.zip

  • Some more photos.

    To interface module - 2 additional displays (LCD) possible connection: one is 2x16 char as a Mode and Trim display (+ 6-position rotating mode switch, + 2 trimm switches - +/- pitch and +/- roll), second 320x240 touch screen as a additional data and parameter change via touchscreen. Plus Mission Planner - via USB / DroiddPlanner via Bluetooth.

    MinimOSD + Pixim Wolf cam + 7in LCD as "bottom view" - camera installed close to PhaseOne with the same direction.

    During flight tests - most of work was PID tuning. Main difference to models - time and inertia. MGL servo is very slow - 35 deg = 1 second, so it was not easy find the propper tuning parameters. Anyway - now is flying good. X-track error dirung 60 km straight line was +/- 30m with moderate wind and turbulences, and +/- 100 m with very high turbulences (termic conditions under clouds). Roll stab - not more than +/- 10 deg. (within +/- 40 deg limit). "Gimbal" have +/- 15 deg possibility to move (due to size of bottom window for lens) so finally all photos was vertical (max difference - 2 deg).

    Plans: to transfer necessary parts of APM firmware into new board (STM32 based) with separate IMU, and that will be flexible and easy to implement in normal planes autopilot, or rather "track" pilot solution for ortophoto.

    http://youtu.be/yECnWK81F8k

    http://youtu.be/ToFUL6o9z70

    3701745042?profile=original

    3701744797?profile=original

    3701744979?profile=original

    3701745122?profile=original

  • 3D Robotics

    Impressive! Post edited to embed photo and video. 

This reply was deleted.

Activity