Over the weekend I did my first Balloon assisted flight.  Helium balloon carried my glider up to 13,000ft, at that point i issued a remote cutdown command to release the balloon from the plane.  

 

http://twitter.com/#!/W0OTM/status/130798913224577024/photo/1

 

We ha some issues...First know that shortly after launch the XMTR was turned off, to initiate RTL mode.  So, for the majority of the flight (20min) it was in RTL mode while suspended from the balloon.

 

After finding the wreckage, I began to filter though the logs.  (the APM survived the crash)  Here is what i found.

 

The pressure sensor began sending erroneous readings as 1137m (4,000ft) or so.  Here is a sample of the logs.

 

GPS: 71299000, 1, 11, 41.1029510, -92.4485240, 0.0000, 1137.9200, 1294.2400, 2.2300, 338.1000

GPS: 71299000, 1, 11, 41.1029510, -92.4485240, 0.0000, 950.5000, 1294.2400, 2.2300, 338.1000

GPS: 71299250, 1, 11, 41.1029590, -92.4485240, 0.0000, 821.7400, 1296.0200, 1.9200, 18.2300

GPS: 71299250, 1, 11, 41.1029590, -92.4485240, 0.0000, 650.5400, 1296.0200, 1.9200, 18.2300

GPS: 71299250, 1, 11, 41.1029590, -92.4485240, 0.0000, 452.5100, 1296.0200, 1.9200, 18.2300

GPS: 71299500, 1, 11, 41.1029590, -92.4485170, 0.0000, 238.2400, 1297.8101, 2.3700, 38.0600

GPS: 71299500, 1, 11, 41.1029590, -92.4485170, 0.0000, 15.0300, 1297.8101, 2.3700, 38.0600

GPS: 71299750, 1, 11, 41.1029660, -92.4485170, 0.0000, -137.8500, 1299.6100, 2.8800, 55.0100

GPS: 71299750, 1, 11, 41.1029660, -92.4485170, 0.0000, -316.5000, 1299.6100, 2.8800, 55.0100

GPS: 71299750, 1, 11, 41.1029660, -92.4485170, 0.0000, -439.2400, 1299.6100, 2.8800, 55.0100

GPS: 71300000, 1, 11, 41.1029700, -92.4485090, 0.0000, -523.8900, 1301.4301, 3.1900, 61.0800

GPS: 71300000, 1, 11, 41.1029700, -92.4485090, 0.0000, -582.4400, 1301.4301, 3.1900, 61.0800

GPS: 71300250, 1, 11, 41.1029700, -92.4485020, 0.0000, -623.2000, 1303.2400, 3.6500, 58.7000

 

Also, the IMU axis sensors stopped sending values at 7000ft.  Also note the HUGE difference in the MIXALT and the GPSALT.

 

GPS: 71460500, 1, 10, 41.1069680, -92.4342270, 0.0000, 140.1700, 2359.3101, 15.8400, 99.1400

ATT: -17273, -7770, 1014

GPS: 71460500, 1, 10, 41.1069680, -92.4342270, 0.0000, 140.7600, 2359.3101, 15.8400, 99.1400

ATT: -16141, -8484, 35974

GPS: 71460750, 1, 10, 41.1069640, -92.4341740, 0.0000, 141.4400, 2360.8999, 15.1800, 100.5000

ATT: -11298, -8952, 31063

GPS: 71460750, 1, 10, 41.1069640, -92.4341740, 0.0000, 141.8600, 2360.8999, 15.1800, 100.5000

ATT: 0, 0, 0

GPS: 71460750, 1, 10, 41.1069640, -92.4341740, 0.0000, 142.5400, 2360.8999, 15.1800, 100.5000

ATT: 0, 0, 0

GPS: 71461000, 1, 11, 41.1069560, -92.4341280, 0.0000, 143.3000, 2362.7300, 14.6200, 101.7700

ATT: 0, 0, 0

 

What would have caused these 2 failures? 

Views: 1204

Attachments:

Reply to This

Replies to This Discussion

Marshal I have experienced something very similar around the same altitude.
I have replicated it twice.

Around 1300m, my APM started reporting -1000m altitude (so of course started climbing).

Details are here...

http://diydrones.com/forum/topics/nasty-altitde-bug 

PS, I wonder if this is a temperature issue. At that altitude, the temperature is approx -7 celcius.

I cant imagine that cold to be an issue.  -23F (-7C) isn't THAT cold.

 

Did experience the same axis sensor issue as well?

IF it is a hardware limitation, what will it take to code the ability to disable the pressure sensor and use the GPSALT only for altitudes above lets say 1000M.

Hi Marshall,

 

Not sure if this might help you:

 

Altitude measurement and control.
//
// ALT_EST_GAIN                             OPTIONAL
//
// The gain of the altitude estimation function; a lower number results
// in slower error correction and smoother output.  The default is a
// reasonable starting point.
//
//#define ALT_EST_GAIN        0.01
//
// ALTITUDE_MIX                             OPTIONAL
//
// Configures the blend between GPS and pressure altitude.
// 0 = GPS altitude, 1 = Press alt, 0.5 = half and half, etc.
//
// The default is to use only pressure altitude.
//
//#define ALTITUDE_MIX        1

 

Regards

 

Martin.

 

www.buildyourowndrone.co.uk

Thats great, Mine was set to 1, so I changed to to '0'.

 

That still leaves the most important question of WHY I lost Attitude sensors.  Since the plane was suspended and wasn't able to respond to respond to APM commands, does the APM "giveup" after a certain time?

Hello Marshall,

I have a feeling that this is due to the change in temp, I have not got any hard and
fast proof, however I would say a more down to earth test would let us know, I
think it would be interesting to see if you get the same result when dropping
the APM into a freezer from normal outdoors temp and watching the readings that
it’s showing, as I say it’s only a feeling, but I think you will see that the
temp plays a big part in the issue you have encountered!

Please remember if you are going to do this use the same precautions as when bringing
a camera in from the cold, I would place the APM in a sealed bag to stop condensation
forming on the boards while returning to a “normal temp” I would continue to see what the APM is reporting as it warms back up again!

Regards

 

Martin.

www.buildyourowndrone.co.uk

As you say, -7C is not "that cold." But it isn't -23F. (-23F is pretty cold.)

 

-7C is 19.4F

 

 

Don't discount air pressure!
A group from here in northern Alberta went quite a bit higher with an HD video cam.
On return to earth the camera controls had jammed 'cuz the membrane style buttons had leaked at altitude and now the high pressure at the surface was causing all of them to activate at once.

Sorry Ken, ur right,  Typo.  But still I find it hard to believe that 20F is too cold for the APM.

I obtained the data sheets for the axis sensors on the IMU, and they have an operating temperature of -40C.  There is something else that caused loss of attitude sensing.

Is something that someone could test using a vacuum test rig of some kind?

Reply to Discussion

RSS

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service