Re: [Paparazzi-devel] ardrone2_raw supported GPS?

[Christophe De Wagter]

Baro calibration problem was solved 2 weeks ago so it does not apply anymore. If the baro calibration does not download the drone will now not boot anymore as he will retry infinitely (in practice this now only happens on broken navboards)

For IMU/Mag calibration: see default calibration wiki page

For the GPS: I know specifically that the UBlox-GPS internal dynamic model, measurement noise and update rate are exceptionally good for UAV. I have no idea about the link you post. In other words: every GPS can tell you where you are but not all are equally good when used in control loops. 

-Christophe 

On Sat, Sep 28, 2013 at 5:27 PM, Ben Laurie <address@hidden> wrote:

Somehow missed this email....

On 9 September 2013 09:37, Christophe De Wagter <address@hidden> wrote:

Status:

-Given a manual check that the baro-calibration data was downloaded correctly (check raw baro temperature which should be about 20 to 40 degrees and pressure should be arround 102300 at sealevel and not 10x smaller)

How do I do this? Patch the code?

 

-Given the fact that you calibrated the magneto (incl current compensation) and accelero neutrals correctly

Likewise, how is this done?

 

-Accepting the fact that after take-off the drone moves much more laterally than with the camera-stabilized attitude trim determination of the SDK (SDK uses sonar + camera + optic flow + kalman filter to estimate which roll/pitch trim will stop the motion. Since the vision framework now works in RAW mode this can become possible in raw mode too in the future)
-Using a Ublox GPS since it works a lot better than the Sirf GPS. The accuracy of a Max6 Ublox is easily 4m while the cheap sirf is known to give larger estimation errors.

Is one of these appropriate? http://www.ebay.co.uk/itm/GPS-Smart-Antenna-GPS-03U-NEW-USB-GPS-Receiver-for-PC-laptop-compter-/190722494577?pt=UK_CE_GPS_Accessories_Software_ET&hash=item2c67f25071

 

Given the above: the ArDrone flies much much more reliably/repeatably/consistently in RAW mode than in (often unpredictable) SDK mode. 

-Christophe 

On Sun, Sep 8, 2013 at 10:34 PM, Freek van Tienen <address@hidden> wrote:

Thanks! we will test this tomorrow probably ;)

Sirf Demo is a tool to configure Sirf GPS modules. The Flight Recorder from parrot s just a simple Sirf GPS module with some extra flash memory for saving videos. Sirf Demo can change the mode(protocol) in which the Flight Recorder is sending data in to NMEA or Sirf. It can also switch the baudrate at which it is transmitting data, and some other stuff like the speed of certain messages etc.

Paparazzi currently supports both Sirf and NMEA, but I recommend you using Sirf because then the transmitter can send the position data at a higher frequency. When you configured it to Sirf and set the baudrate to 57600 or higher, and the frequency of the position messages at the highest rate you can just set the gps subsystem inside paparazzi to sirf. Then you only need to take a look inside the ardrone2_raaw makefile, and see how to configure the right baudrate in your airframe file.

2013/9/8 Ben Laurie <address@hidden>

On 8 September 2013 19:31, Ben Laurie <address@hidden> wrote:

On 8 September 2013 19:20, Freek van Tienen <address@hidden> wrote:

The ardrone2_raw does support the Flight Recorder, except for the fact that you have to manually configure it with "Sirf Demo".

Could you be a little more explicit about this? What is Sirf Demo and what would I do with it?

 

The idea is to make a module like the ublox center module, to change the configuration of the Flight Recorder.

Next to the Flight Recorder it also support u-blox GPS. This one even works with the u-blox center module to configure the GPS. You can either have a u-blox GPS with mini usb connected right away to the AR.Drone, or use a tty cable and solder that to a u-blox GPS.

The raw version is currently in testing phase, but is usable as it is right now in master. Still having some problems now and then with the baro, that it doesn't boot correctly.

You might find my fixes help with this.

 

Also having some little problems when something gets in between a propellor and all motors stop, that we sometimes can't start all motors again. But it works about 90% of the time to reset all 4 motors correctly.

And with this.

I'd love to know if they do help.

I just noticed, btw, that I meant to add a full_read() (read can read partially, particularly over serial, so this is going to cause you problems for sure). I'll do it now.

Done. Now in the same pull request.

 

 

There could alos be a "floating point error" somewhere in the code, but we aren't sure if it comes from our vision framework(which isn't in master yet) or from paparazzi somewhere.

Thank you for the cleanup! This is much better!

2013/9/8 Ben Laurie <address@hidden>

Am I right in thinking the ardrone2_raw branch does not support the Flight Recorder?

What GPS does it support?

Also, once more: is it actually usable?

Finally, see pull request #535, which fixes some bugs and does some general tidying.

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel