Originally Posted by MAPGPS

We also need to know the initial absolute direction when takeoff, so the on-drone app can calculate the Yaw needed for heading to the first GPS waypoint in the list.

Should we always put AR.Drone on the ground and facing to North of earth before takeoff?
Otherwise, how can we get the initial absolute direction? (unless a Compass used).
Or randomly fly with one direction just after takeoff, and get a rough absolute direction from GPS NMEA? (GPS can output direction only when moving).

We started with the nose pointing north to know the initial direction. this is the easy way.
we also tested a filter where we use the GPS track to calculate the heading. This only work when you fly like a "airplane". so no random flight (nose has to be in the direction of flight).
to calibrate the heading whit this filter we had to run over the field with the ardrone in our hands till the heading was correct (thats the reason we stopped whit this filter ;-) ).

microuav

Hi,

Does anyone have any flight videos (3rd person view) with the mod they can post?
I want to see the impact of the improved 'potortional' control and also the effect of mixing the rudder and aleron to smooth the flight....I have been flying well on the ipod, but the lack of rudder feedback (because of the touch screen), makes you fly back and forth in a Box! I fly my helis in smooth circles and would love to do that with the AR.D.

My AR.D did two flyaways onm wifi requiring some climbing!! - parrot repaired it twice but have RC for outdoor flight with the improved range will make all the difference...
nice job muru on the mod.looking forward to trying it.

malder

Well, you will have a heli feel once you convert.
With iDevice you should still be able to fly with out the "box" flight. I fly it with out the acc enabled. So, both thumbs on the screen and you can control both rudder and aileron at the same time. Or at least I think I can.

I think there is a video on this thread on how it looks like when it flies.
I just tried mine yesterday. Crimp 2mm connectors worked out very well.
Still using 0.03 firmware. Will update to 0.04 today before flight.

Maybe will take some videos this evening and upload them.

Originally Posted by MAPGPS

Here is a dump of the raw navdata made by Shawn Fraser:
https://projects.ardrone.org/attachments/339/Flight.zip

You can convert the TXT file into Excel format for easy to read.

Did you have a chance to look at the data and make sense out of it? TAG 2 (NAVDATA_RAW_MEASURES) has 3 acc readings, 3 gyro readings + 2 more gyro110 readings. Do you know what the latter two are about?

Hello,

i did the mod for a friend... with a AR.drone with Firmware 1.5.1, no problem, everything works fine !

We used a FUTABA F-617FS receiver... perfect... works fine !

The transmitter is a 12FH, very nice radio... we used the FM2 3 position switch mode... off course

For your information, we setted up the sensibilities of the channels in the arduino software to the maximum and we used the exponential function in the radio to calm down the Drone a bit Works fine...

Cheers.

Laurent

Did you use the 0.04 firmware? How did it handle the grass issue?

I am to try mine out in a few hours.

Originally Posted by miru

Did you have a chance to look at the data and make sense out of it? TAG 2 (NAVDATA_RAW_MEASURES) has 3 acc readings, 3 gyro readings + 2 more gyro110 readings. Do you know what the latter two are about?

I did have looked at those raw data.
The IDG-500 gyro has two separate outputs per axis for higher speed motions and lower-speed precise movements:
500°/s full scale range (higher speed)
110°/s full scale range (high precision)

They are gyro(X,Y,Z) and gyro110(X,Y).
On the ground, they have the same outputs.
On flying, they have different outputs.
The NavData_Demo Psi (Yaw) is a fusion of them (including acc).

No further analysis yet.

Originally Posted by MAPGPS

I did have looked at those raw data.
There are 2 gyroes: one gyro(X,Y,Z), another gyro110(X,Y).
On the ground, they have the same outputs.
On flying, they have different outputs.
The NavData_Demo Psi (Yaw) is a fusion of them (including acc).

No further analysis yet.

Maybe one of the gyros (110) is faster and is used for tight loop platform stabilization. I am surprised though that the substantial psi drift is not compensated for in the fusion algorithm, because the accelerometer data does not look that bad. I'll have to do some data gathering myself and figure out if it could be improved. Still leaves us with the conceptual problem of not knowing where the nose is... Received my GPS today, am not impressed with it's locking capability indoors, but will mount it on the drone so I can see how it does outside.

Originally Posted by Lorenzo29

Hello,

i'm on MAC as well...

As a serial interface, i use a FTDI Breakout 5V Sparkfun :
http://www.sparkfun.com/products/9716

On arduino IDE, it appears as :
Tools -> Serial Port =>
/dev/tty.usbserial-A700eDIY

The driver for this serial interface, as long as it is based on a FTDI is given with the Arduino IDE...
The one i use is :
FTDIUSBSerialDriver_v2_2_16.dmg
Pick it up here :
http://dl.dropbox.com/u/2463308/FTDI...er_v2_2_16.dmg

It should work better...

Cheers.

Laurent D

Hi Lorenzo29...
I ordered the FTDI Breakout 5V Sparkfun : do you have a photo or wiring diagram to connect to the Arduino? I ordered the parts and I am awaiting the arrival, I am not sure if it just plugs in and then USB's to the MAC/PC or if I need to wire it up like Miru's photo... thanks...!

Originally Posted by malder

Hi Lorenzo29...
I ordered the FTDI Breakout 5V Sparkfun : do you have a photo or wiring diagram to connect to the Arduino? I ordered the parts and I am awaiting the arrival, I am not sure if it just plugs in and then USB's to the MAC/PC or if I need to wire it up like Miru's photo... thanks...!

No wiring need. You will need to solder pins to the Arduino.
Then just plug the Ardunio to the FTDI board(female connector already on it).
Then plug it in to your Mac.

Originally Posted by actionplus

Did you use the 0.04 firmware? How did it handle the grass issue?

I am to try mine out in a few hours.

Any news on the grass situation? I'm gonna do this mod but there is no point if it doesn't fix the "not being able to fly over grass" madness

Thanx

Originally Posted by sevensixty

Any news on the grass situation? I'm gonna do this mod but there is no point if it doesn't fix the "not being able to fly over grass" madness

Thanx

Have not heard from anybody so far. On mine it does recover...

Originally Posted by MAPGPS

I did have looked at those raw data.
The IDG-500 gyro has two separate outputs per axis for higher speed motions and lower-speed precise movements:
500°/s full scale range (higher speed)
110°/s full scale range (high precision)

They are gyro(X,Y,Z) and gyro110(X,Y).
On the ground, they have the same outputs.
On flying, they have different outputs.
The NavData_Demo Psi (Yaw) is a fusion of them (including acc).

No further analysis yet.

MAPGPS, I monitored TAG 02 raw_gyros[2], the output of the X3500 gyro, over 20 minutes on the bench, it only moves +-1 tick once in a while, so it is very steady. However psi is moving quite a bit. I think they have a problem with the integration of the signal over time. When I look at the TAG 01 (time) record in navdata and compute the delta between to successive messages, it goes from 0.063001 ... 0.066599. The clock from that record is an abbreviated version of the Linux struct timeval... not very good. I think this is where the psi drift is coming from.

Originally Posted by sevensixty

Any news on the grass situation? I'm gonna do this mod but there is no point if it doesn't fix the "not being able to fly over grass" madness

Thanx

I tried mine out today. Something does not look right. Grass situation still on mine.

And when it takes off, seem not to take off straight up. Takes a few seconds before it stabilize. (non grass surface)

But again, it has been really windy this weekend.

Will keep trying.

Originally Posted by actionplus

I tried mine out today. Something does not look right. Grass situation still on mine.

And when it takes off, seem not to take off straight up. Takes a few seconds before it stabilize. (non grass surface)

But again, it has been really windy this weekend.

Will keep trying.

Thanks for the feedback, I move the cars out of the garage and test things there (no wind condition) and if everything seems to work, just fly outside. Rev 0.04 blinks the drone lights for 2 seconds when it resets the COM_WATCHDOG, so you can see when it got into trouble...