Page 1 of 1

Finding motor Biases

Posted: Thu Jan 31, 2019 3:33 am
by dd292
Hey CrazyFlie lovers,

I have been tinkering around with crazyflie for my project. I am actually trying to find the motor biases for hover position or in simple words I am trying to finds the motor values (10000-60000) at which the pitch and roll angles would be zero without any control.

I used PID with 99% thrust to find these values, which came out to be
m1- 65535.0
m2 -65535.0
m3- 48495.0
m4- 41463.0
but somehow when I turn the controller off and use these values the system doesn't go to zero.

Am I missing something here? or is there something in the firmware that's stopping it.?
I hope I was clear in my question. Some help would be really appreciated.
I have attached a picture of my setup here.

As you can see the crazyflie can only move in 2DOF here (pitch and roll).

Re: Finding motor Biases

Posted: Thu Jan 31, 2019 9:30 am
by arnaud

If I understand well your question, you are trying to control the Crazyflie attitude by setting fix thrust value to it, without control loop.

As far as I understand, this should be impossible. The reason is that you will never find the exact bias that makes the propeller produce exactly the same thrust, you can get very close but it will not be the same. This will result in a torque being applied on the platform and the attitude will drift. A control system would continuously observe the attitude and vary the torque to reduce errors as they happen, this can be the PID controller, and it is required to control the attitude of a quadcopter.

Re: Finding motor Biases

Posted: Thu Jan 31, 2019 9:44 pm
by dd292

Thanks for your reply. So I figured out I was using the worng rpm values. Its working now :D.

Thing is I just want it to hover for 1 second or even less. Just a proof of concept for something else I am working on. Thank a lot.!