I'm assembling my first quad.  It isn't stable when we try to fly it and I've been trying to figure out why.   Today I stumbled upon what probably is the answer.

When the quad is started up, the artificial horizon is steady.   It will sit there for quite some time with no problems at all.   If you pick it up and move it, however, when you set it back down the horizon won't be quite right and it will start to drift around.

Is there a procedure for testing the sensors?  Are they configurable?   Where do I go from here?

Thank you for taking a minute to read my thread, 

- David

Views: 304

Reply to This

Replies to This Discussion

Hi David,

Does the artificial horizon slowly drift back to a stable position, or does it continue to turn and change randomly after you move the quad? There are some tests that you can run in the terminal (CLI mode). Try the "gyros" test, the "adc" test, or the "imu" test, and the gyros should be very stable.

Tech Support

Thanks for the reply, it seems to continue to drift when it drifts but it doesn't always drift.   It's pretty intermittent.   I'm not sure why it sometimes seems to be drifting badly and other times it seems mostly fine, but I don't know what to expect.  

I've taken a bunch of screen shots of sensors doing various things and put them all here:


The accelerometers seem to fluctuate +- 2.   The Gyro's +- 8.   The VSI gauge alternates quite erratically +-2.   I found that running "level" under Firmware->APM Setup->ArduCopter2 sets everything level and doesn't seem to drift.

When I hold the stick down and to the right for 15 seconds, however, everything goes haywire.   It starts drifting immediately and then settles down to a slower rate and never seems right.

All these screenshots were taken on a table without moving the quad.   

Screen shot "gyros.png" is just the accelerometers and gyro's doing their thing.  gx and gy certainly aren't 0 but the display isn't drifting.   You can't see it, but the VSI gauge is fluxuating quickly between +- 2.

"Verticle Speed.png" and "climbrate and verticalspeed.png" show the vsi bouncing around. 

"drifting after 15 second down and right.pmg" shows the quad after I've held down and right on the stick for 15 seconds and performed an auto level.   This screenshot is from after the autolevel is finished.  You can see that accelerometer X seems to have picked up quite an offset.  There is also a zoomed in version of this one that better shows the other sensors.


I noticed the "yaw seems to be jumping around.png"

Is this normal or is this out of spec?  Wish there was a document some where that spelled out what to expect out of these sensors.  


Hi David,

The roll, pitch, and yaw should all be very still when the UAV is not moving. If the roll, pitch, or yaw were moving while the quad was standing still, then there may be a hardware problem, unfortunately. Please contact customer service at the store from which you purchased the boards for an RMA.

Reply to Discussion


© 2018   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service