From PaparazziUAV
Jump to: navigation, search

Currently operating two Paparazzi powered aircraft

Parrot AR.Drone V2.0


Taking control

What framework to use to work with the ARDrone. For me this revolves around ROS and Paparazzi.

  • standard
  • standard + QGC
  • paparazzi
  • ros
    • installed ubuntu 12.04 LTS
    • installed ros-groovy-ros-full
  • node.js


  • Control
  • Access to sensors
  • Flight recorder

Video stream

  • gst stuff installed fine as per instructions from wiki


  • Flight recorder doesn't work
  • Using drotek from
  • ublox 6 works, need GPS_PORT_ID = USB_PORT_ID, already obsolete in current (<2014-06-10 Tue>) git
  • needs cdc-acm.ko kernel module
    • get linux.tar.bz2 from
    • get gcc cross compiler
    • unpack linux sources
    • cp kernel.config .config
    • set ARCH=arm, CROSS_COMPILE=arm-linux-gnueabi- in Makefile
    • make menuconfig
    • make modules
    • make modules_install INSTALL_MOD_PATH=../linux-modules


  • cp some-joystick.xml to gamepad.xml, change settings
  • exec input2ivy -ac ardrone2raw gamepad.xml


  • ardrone2_raw magnetometer problem: img_mag_raw stops reading from sensor, then TBE / toiled bowl effect, see ppz-l mails
  • fixed <2014-06-10 Tue>

Messages / Debugging / ivyprobe

  • use -c parameter for messages to select from message classes
  • (telemetry, ground, datalink, …)
  • enable debugging messages in GCS / settings app in order to
  • receive e.g. IMURAW messages

Ivyprobe needs a regexp as parameter to show you something:

  • to see all messages: ivyprobe '(.\*)'
  • to see a specific message (rotorcraft status for instance): ivyprobe '(.\* ROTORCRAFTFP .\*)'

With "messages" tool, you can also start with an option to see messages from an other class than telemetry (downlink): messages -c datalink

PID tuning

sluggish yaw response:

[GH]: I have started to fly an ardrone in manual today, and it seems that the SPMAXR which define the max yaw rate is way to high and can be reduced to 200 deg/s or less for an easier control in manual. Also the gains can be increased compared to the values in example file. I'm using PGAIN=1600 DGAIN=450 IGAIN=120 for PHI and THETA without issues (indoor flight only for now).

Btw, the mag seems to be really crappy…

Lisa/M v2 10-inch quadrotor

Setup log

DONE [2014-05-19 Mon 09:10] log / notes / ppz

  • ppz: ppz copter construction started, see below 2014-05-23
    • x ppm encoder board
    • x ppm encoder board to lisa
    • x connect GPS, telemetry modem, receiver
    • x configure Xbee for 57600 with X-CTU on windoze
    • x connect battery voltage
    • x solder LED strip red
    • x remove second satellite
    • x bind spektrum: lisa/m v2 page
    • program transmitter for ppz, set up radio receiver on ppz radio
    • config
      • roll: 11942 / -420 / -12278
      • pitch: 12236 / 42 / -12278
      • yaw: 12236 / -154 / -12278
      • thrust: -1318 / 10883
      • maybe different for spektrum?
      • RC calibration obviously not needed for spektrum
    • x ESCs OK motor order: mixer setting done
    • x configure mixer for quadx: done
    • calibrate IMU: unclear, done MAG
    • throttle adjust sensitivity
    • rate mode disable PFD?

[2014-05-20 Tue 21:41] log / notes / ppz

  • x ppz managed to upload ap.bin to lisa/m. addgroup plugdev, copy
  • udev rules for lisa/m

[2014-05-23 Fri 10:46] log / notes / ppz

  • ppz motors spinning
  • ppz now need to tune gains for RATE, ATT
    • sw/airborne/firmwares/rotorcraft/stabilization/stabilizationrate.c
    • PID formulation 1: i,d-gains independent of pgain

[2014-05-30 Fri 01:39] log / notes / ppz

  • multistar ESCs calibration trouble
    • looks like the multistar ESCs are special as they don't do
    • calibration. since i was flying well with these ESCs with the
    • naze32, it might be they're hardwired to 1000-2000us, which is
    • related to the way MultiWii handles the RC input.
  • spektrum / calibration?
  • ppz tuning
    • check GPS, separate antenna and place outwards
    • ahrs initialization
    • tuning: yaw is sluggish
    • tuning: roll/pitch is wobbly
    • disarm jumping

[2014-06-01 Sun 01:42] log / notes / ppz

  • IMU outage on the field
  • back alive on returning home
  • changed lisa power supply to separate 3A UBEC. seems better. ESC
  • BEC seemed to get pretty hot.

DONE [2014-06-04 Wed 10:51] log / notes / ppz

  • x ppz ardrone mag fix test
  • x ppz lisa disarm fix test: xbee missing
  • ppz GPS and vision fuse?
  • ppz document setup and fix minor things in wiki

[2014-06-06 Fri 09:06] log / notes

  • ppz: phi/theta: P 450, D 350, I 60, F 160
  • ppz: psi: P 480, D 100, I 10, F 250
  • x ppz: tested yaw/disarm fix from felix. works.
  • x ppz: tested mag freeze fix from daren. works.
  • x ppz: weird GPS fix electrical issue on lisa quad
  • x ppz: started rewiring with wound cables etc, slightly better, GPS
  • fix test pending

[2014-06-06 Fri 09:09] log / notes / ppz

  • ppz gps problem
    • x try other AP version: fixed
    • try other gps
    • try without ublox-ucenter autoconf
  Re: Long Beep and loses bind?
  Postby dmajchrzak » Sun Jun 01, 2014 6:10 pm
  Figured it out. The"change id" button was stuck in. I was able to pop it free and now all is well.

[2014-06-10 Tue 10:01] log / notes / ppz

  • ppz: first real flight
  • ppz: yaw does sweeping (back and forth)
  • ppz: roll / pitch still wobbly
  • ppz: baro doesn't work / no alt info

[2014-06-11 Wed 22:00] log / notes / ppz

  • ppz: test flight with IMU ok, telemetry has trouble but anyway, PID values better (policy search), Z_HZ mode semi-works, lots of oscillations

[2014-06-22 Sun 23:58] log / notes / ppz

D3 on my Lisa, looks like parts of the casing went off in addition to reading 6 Ohm both ways.

Ppz lisa D3 1.JPG
Ppz lisa D3 2.JPG