What is the standard deviation of the thrust?

Discussions and questions about the Crazyflie Nano Quadcopter
Post Reply
DarkKnight
Member
Posts: 39
Joined: Tue Jun 15, 2021 10:19 pm

What is the standard deviation of the thrust?

Post by DarkKnight »

Hello,

right now I can use the logging function to log the stabilizer.thrust, which is thrust percentage mapped to a uint16_t, where 0 is no trust and 65535 is full thrust. I have checked your documents about thrusthttps://wiki.bitcraze.io/misc:investigations:thrust, there are different ways to calculate the thrust(For example, using the relation between rpm vs thrust, or voltage vs thrust, and so on).

My first question is the percentage of stabilizer.thrust is based on what kind of relation, is it the rpm vs thrust?

My second question is how could I know the noise of the value of stabilizer.thrust? what is the standard deviation of thrust?

Thank you!
kimberly
Bitcraze
Posts: 1050
Joined: Fri Jul 06, 2018 11:13 am

Re: What is the standard deviation of the thrust?

Post by kimberly »

Hi!

Currently it's a bit quiet at the office, but in a weeks time then people come back from their holidays, who are able to help you out more with your question. So I'll remind them once they are here!
tobias
Bitcraze
Posts: 2339
Joined: Mon Jan 28, 2013 7:17 pm
Location: Sweden

Re: What is the standard deviation of the thrust?

Post by tobias »

My first question is the percentage of stabilizer.thrust is based on what kind of relation, is it the rpm vs thrust?
The thrust value is translated to a PWM signal which is driving the motors and has a linear relationship to the thrust if supply voltage would stay the same. However the battery voltage, doe to internal resistance, doesn't. You should look at the PWM vs Thust graph.

In the code there is a battery voltage compensation function that tries to compensate for this.
My second question is how could I know the noise of the value of stabilizer.thrust? what is the standard deviation of thrust?
I'm not sure why you need this? I guess it could be measured but it probably requires a test jig and a high sample rate load cell. If you want to know how much thrust you get from a specific pwm value the graphs can be used but they will differ between setups as batteries, motors and propellers are not exactly the same.
DarkKnight
Member
Posts: 39
Joined: Tue Jun 15, 2021 10:19 pm

Re: What is the standard deviation of the thrust?

Post by DarkKnight »

Hi, thank you so much for your reply.

For question1, what I am really interested is that how stabilizer.thrust is calculated? is this a desired thrust which is calculated from controller or some motor_rotor dynamics(I tried to read the code of stabilizer.c, controller.c and controller_pid.c in crazyflie-firmware code,but I didn't find it how this value is calculated)?
DarkKnight
Member
Posts: 39
Joined: Tue Jun 15, 2021 10:19 pm

Re: What is the standard deviation of the thrust?

Post by DarkKnight »

tobias wrote: Tue Aug 10, 2021 9:32 am
My first question is the percentage of stabilizer.thrust is based on what kind of relation, is it the rpm vs thrust?
The thrust value is translated to a PWM signal which is driving the motors and has a linear relationship to the thrust if supply voltage would stay the same. However the battery voltage, doe to internal resistance, doesn't. You should look at the PWM vs Thust graph.

In the code there is a battery voltage compensation function that tries to compensate for this.
My second question is how could I know the noise of the value of stabilizer.thrust? what is the standard deviation of thrust?
I'm not sure why you need this? I guess it could be measured but it probably requires a test jig and a high sample rate load cell. If you want to know how much thrust you get from a specific pwm value the graphs can be used but they will differ between setups as batteries, motors and propellers are not exactly the same.
Hi, I have another question. For the pwm vs thrust plot that your provided(see attached picture), is this relation between single motor's pwm value and single thrust? or it is the relation between mean of four pwm values and total thrust?

What's more, for above plot, the x-axis is 8-bit=256 values PWM, but the logging parameters pwm.m1_pwm to pwm.m4_pwm is 16-bit=65536 values PWM. Should I map the 16-bit values to 8-bit values if I want to use your provided thrust transfer function?
Attachments
thrust vs pwm.png
tobias
Bitcraze
Posts: 2339
Joined: Mon Jan 28, 2013 7:17 pm
Location: Sweden

Re: What is the standard deviation of the thrust?

Post by tobias »

It is the relation between same pwm value to four motors and the total thrust.

Yes, 16-bit is mapped to 8 bit.
DarkKnight
Member
Posts: 39
Joined: Tue Jun 15, 2021 10:19 pm

Re: What is the standard deviation of the thrust?

Post by DarkKnight »

For your first reply, Why do you add a question mark? do you mean for above plot, the pwm of four motors should be the same and this same pwm is the x-axis of the above plot?the y axis is the total thrust?

The true situation is that the pwm of four motors could be different, what should I do if I want to get the thrust?Could I use the mean of the pwms of four motors?
tobias
Bitcraze
Posts: 2339
Joined: Mon Jan 28, 2013 7:17 pm
Location: Sweden

Re: What is the standard deviation of the thrust?

Post by tobias »

The question mark was a mistake, removed now.
do you mean for above plot, the pwm of four motors should be the same and this same pwm is the x-axis of the above plot?the y axis is the total thrust?
Yes, the same PWM is sent to all four motors in the test. There will be individual differences for each motor so when calculating for each motor (total trust / 4) you will get the average thrust.
Post Reply