[larscc3d]

Greetings.  I'm new to RC.  I'm new to CC3D and LibrePilot.
I connected my quad to laptop appropriately, and I was able to progress through the vehicle and transmitter wizards appropriately.  Of course, once finished, I tested (without props).

At low throttle, when motors just begin to spin (say at 2% throttle), only motors 1, 2, 3 spin.  Motor 4 doesn't begin to spin until about 8-10 % throttle (significantly more).  Of course, I fiddle, adjust stick positions, adjust channel 5 and 6 switches, etc.

Then, I notice motors 1, 2, and 4 spin at low throttle.  Now, motor 3 doesn't spin until 8-10% throttle.

Yes, after more fiddling, motors 1, 3, and 4 spin at low throttle, and motor 2 takes 8-10% to start moving.

If this were one motor, I would suspect a short, an ESC, or the motor itself.  But the slow-to-spin motor moves.  I don't know how to begin debugging this problem.

Can anyone offer suggestions?  Fixes?
Thank you.

[mr_w]

You should probably redo the ESC calibration and output neutrals setup.

See https://librepilot.atlassian.net/wiki/display/LPDOC/ESC+Calibration for step-by-step instructions.

[larscc3d]

Thank you for your reply, mr_w.  I'm starting to follow your suggestion, and right away I notice that motors 1, 2, 3 are in "Bank" 1 and motor 4 is in "Bank" 2.  Is that appropriate?  If not, how do I change banks?
Thanks again.

[Mallowsthecat]

hello i suggest you redo the vehicle setup wizard then when you reach the motor output calibration please
calibrate all motors at same time instead of doing it one by one.

you will notice that when you calibrate motors one by one you will get mis-correlated neutral values
example:

motor 1 : 1041
motor 2 : 1046
motor 3 : 1039
motor 4 : 1045

the pulse values above are just enough to make motors spins but notice the values they are not correlated
please take note that when calibrating using microcontroller multiport it is not a good idea to do it one by one
because you want to sync the internal clock of the microcontroller for the multiport output of your microcontroller



you will notice that when you calibrate all motors at the same time you will get correlated neutral values
example:

at 1041 motor1 starts to spin
at 1042 motor2 starts to spin
at 1043 motor3 starts to spin
at 1044 motor4 starts to spin

notice that the neutral values are correlated with the increment of +1 the reason for this is the internal clock
tries to determine the pulse compensation with respect to the internal clock of the microcontroller, in this case
at 1044 all motors should spin since you have sync on the multiport and internal clock of the microcontroller

(the micrcontroller is your flight controller example cc3d - the microcontroller is STM32)


hope this helps

[Mallowsthecat]

and also after doing proper motor output calibraton dont forget this

on the GCS Libre pilot

goto configuration
goto stabilization
then on the bottom part find:
Zero integral when throttle is low -> enabled this then dont forget to save

[TheOtherCliff]

Thank you for your reply, mr_w.  I'm starting to follow your suggestion, and right away I notice that motors 1, 2, 3 are in "Bank" 1 and motor 4 is in "Bank" 2.  Is that appropriate?  If not, how do I change banks?
Thanks again.

This is fine.  Just make sure that Bank 1 and 2 are configured the same way, e.g. PWMSync.  That makes sure that all the outputs are configured the same way...

If the banks were configured differently, then it would be a good idea to make them the same and then redo ESC calibration and motor neutral setting.

[larscc3d]

Thank you.  Thank you.  This is starting to make sense.

So, if both are done appropriately, will the wizards produce the same results/numbers as configuring manually?

And, what does

Zero integral when throttle is low -> enabled this

do?

[Mateusz]

And, what does

Zero integral when throttle is low -> enabled this

do?

It kind of zeros I term in PID, so copter is not trying too hard to stabilize when lifting off, it makes it easier to liftoff without flipping. Once in air and ASWA is enabled, then full PID is operational.
I think it should only zero it first time throttle is low after arming. But maybe someone else knows more. It's disabled for fixed-wings I think, but for multirotors it's very hard to lift-off if this is not enabled.

[TheOtherCliff]

Wizard should set both banks the same and correctly.  "Zero the integral" is on by default and that is what the average beginner (but maybe not the intermediate pilot doing flips) wants.

Even though the best beginner values are default, these seem to be issues that beginners run into from time to time.  Either they have changed the settings and forgot, or they used someone else's setup cloud config?, UAV file? ...

[Mallowsthecat]

Zero integral when throttle is low - What is this?

Definition of terms:

PID - Proportional Integral Derivative

   Integral - in flight stabilization terms this is responsible in holding your platforms last angle, the values of Integral is mixed by a mixer
                  which is built-in inside the gyro accelerometer IC (MPU6050, iNEMO, Etc..). When values are mixed what is does is it makes
                  your platform stable (a lot happens i just made it short  )


Situation:

   Zero integral when throttle is low - DISABLED

   when you dont enable this Zero integral when throttle is low most of the time your quad will flip on take off.
   If you dont flip on take off you may flip mid flight unexpectedly and you will experience funny flying characteristics
   like very twitchy that it wants to flip. you will also notice that if Zero integral when throttle is low is not enabled
   your throttle will feel funny when taking off, your throttle may rev up on its own, or when you throttle up not all motors
   spin at the same time.

   
   Zero integral when throttle is low - ENABLED

   when enabled you will instantly feel the change, when you throttle up motors spin all at the same time, your throttle is stable
   when taking off it doesn't rev up on its own, your platform will not flip on mid air (assuming your on attitude and don't want to flip).


   Explaination:

   just a simple example imagine your normal multiplier for correction is 5, when your platform detects unwanted movement on a axis
   it applies the correction multiplier which in this example 5 this will be multiplied to your PID's imagine your I is not 0 and the value
   it started at is 1000 since Zero integral when throttle is low is DISABLED on every correction it wants to make it will multiply
   your correction multiplier to 1000 so 5 x 1000 = 5000 so just imagine what kind of correction this will do to your platform it will make
   it crazy example 10 correction factor is = 1 degree angle so 5000 is 500 degree angle, you want your platform to correct itself at a rate
   of 500 degree angle? you will flip for sure and make your platform impossible to control


   please Zero integral when throttle is low - ENABLED

[TheOtherCliff]

PID 'I' term windup...

This is worst in Attitude / Rattitude mode.  The I term winds up if the quad is sitting on a slope (any slope really) motors running, before takeoff.  It also winds up if you hold the stick to "help it" for the slope (don't do that).  With this setting disabled, it doesn't clear the windup.  With this setting enabled, it clears this windup each time you go to zero throttle.

With the setting disabled it is more likely you will get a windup that will cause a flip as soon as you get enough power to fly.

[larscc3d]

Thank you for all the info.  Very helpful.

I was able to get my quad in the air because of your comments above.  Awesome.

I'll be back with more questions as I continue to learn.

Thanks again!