Bitcraze Forums

Everything checks out, but it just wants to yaw in a circle.

I'm giving it just enough throttle to make it weightless (but not technically off the ground), and it just wants to yaw in circles.



Here's a shared link to a screen shot of the client-side software. Even with no yaw input, you can see the motors are clearly trying to yaw. And it doesn't stop -- it just spins weightlessly.

https://www.dropbox.com/s/6zpq3tc8slfnt99/crazyflie.png

Hi,
If you start the Crazyflie on a stable surface and connect to it do you see the yaw actual value moving in the client or does it stay fairly stable?

It thinks it's yawing while it's doing it. The gyros seem to be working just great -- I'm beginning to suspect a motor or motor driver mismatch/mis-calibration.

I assembled a second Crazieflie exactly the same way (same props, motors, soldering, etc. etc.) and it flies just great. It's a really neat little copter.

Here's another reason I think the problem copter is a motor or ESC mismatch:

1) Picture me holding the copter facing forward with thrust but zero pitch and roll.

2) As I roll the copter to the left, M4 starts speeding up and M2 slows down, so as to try to straighten the copter.

3) As I roll the copter to the right, M2 starts speeding up and M4 slows down... but it will slow all the way down to a halt. To even restart the motor, I have to turn the copter almost all the way back to level.

4) With the same test using pitch, M1 and M3 work as expected.

Is there some other diagnostic I could use? Would it be possible to return the problematic one or should I try to debug it myself?

It looks like the motors spin in the wrong direction. Could you check that the motors spin in the same direction for both copters? What you are observing with M2/M4 could be due to the fact that they spin slower in order to compensate for the yaw (the yaw will be compensated by the difference of mean rotation speed of M1/M3 VS M2/M4).

I have two identical nanos with identical wiring and identical prop placement. One flies great and one demonstrates that screwy behavior described above.



https://www.dropbox.com/s/0nlwjtop1mb2l ... 160305.jpg

Here's a brief video of what I see when I manually move the copter around. I think M4 just doesn't get as much power as the other motors. Notice that when I rotate it such that M4 should be kicking-in hard, it does indeed speed up. But when I rotate it such that M4 should back-off and M2 should be kicking-in hard, M4 shuts completely down. Sometimes I have to roll it almost all the way back to neutral just to get M4 going again.

https://www.dropbox.com/s/qzhyizczwr3n1 ... 155659.mp4

That would explain the Yaw too, because M4 just can't keep up.

I just replaced the motor mount, motor and prop on M4 and have exactly the same result.

I do a lot with Ardupilot and medium-large frames, so I freely admit that I'm new to the nano. But my best guess is either the speed controller on M4 is miscalibrated or otherwise bad. It does the correct things directionally and proportionally, but it just doesn't do it strongly enough.

And P.S. the nano that I do have working is incredibly fun to fly. So I'm a satisfied customer, but I do wish I could get this second nano up and going.