[Solved] Crazyflie Bolt - Unstable flight

[relame]

I have been developing a drone system that makes use of the Crazyflie bolt for a flight controller. I have properly tuned the drone's PID parameters within the firmware and in the past, I have been able to perform stable flights as seen in the "Stable Flight" video.

Fast forward about a month (with a few minor crashes), and the drone no longer flies in a stable manner as seen in "Unstable flight". Nothing has changed about the hardware.

To test the software, the same scripts were tested with an off the shelf Crazyflie Nano and the test was perfectly fine with stable flight.

Back to the issue, it is clear from using cfclient that the back right motor is clearly favored as seen in the "Cfclient test", and from the "Unstable flight" video, you can see the result of this favoring on takeoff.

In conclusion, something is clearly wrong with the outputs of the motors. The issue does not seem to be on the software side (commands/tuning), but is something else. The interesting part is the fact that all motors spin during flight and the drone is able to fly, albeit unstably. This fact makes me question if the issue is potentially hardware or not.

Last notes: I have checked all of the solder connections and they are all well connected. Please let me know if you have any thoughts on the cause of this issue. My best guess is that the flight dynamics have changed over time and over the use of the drone/minor crashes. This still does not explain the reason for favoring the back right prop on takeoff but may explain why it is not able to stabilize after takeoff.

Stable flight - https://youtu.be/_ivodwLl02Y

Unstable Flight - https://youtu.be/U6_M1d04bZ8

Cfclient test - https://youtu.be/9Tzpjvkp8tQ

[tobias]

My best guesses are:

• To much vibration
• ESC has been reprogrammed
• IMU is broken

I'm leaning towards that the ESC that is spinning up earlier has been reprogrammed with a new input-range. You should be able to verify this by setting manual PWM values with motorPowerSet parameter group and comparing with each other.

[relame]

Thank you for your response, I will go ahead and try and look into these possibilities and get back to you.

[relame]

I am trying to understand what you mean when you say " verify this by setting manual PWM values with motorPowerSet parameter group and comparing with each other"

My question is how do you set PWM range values manually?

and what is the PWM input range normally?

From looking at a few other forums, are the PWM values not set in https://github.com/bitcraze/crazyflie- ... ck.c#L110 ?

[relame]

I also agree that its likely the ESC values that are being sent to the motor that is spinning up early;

1. From cfclient, the IMU seems to be working properly on screen
2. Too much vibration would have been an issue during the "Stable Flight" that was performed only about a month ago

[relame]

Tobias, I finally fixed the issue and your intuitions were correct.

The external ESC on the overall build must have been burnt out because when I swapped out the ESC that was causing the motor to spin early, the problem went away.

Thank you for your help and leading me down the correct path.

[tobias]

I'm glad you found the problem! You can probably fix the ESC by reprogramming the input-range (calibrating it). Depending on what FW is on the ESC there are different ways to do this but currently requires an external programmer as we still not have a passthough interface in place.