strange Yaw saturation

The last Bastion to fine tune your FC matching your airframe & setup

strange Yaw saturation

Postby daniel_k » Fri May 10, 2013 1:44 pm

Hi,
after doing a lot of dry testing, I went out for the first maiden flight of my Y6 with AQ (used MK on the
same frame before).
Flying was nearly impossible: normal reaction of the copter to nick and roll command, but no reaction
to yaw commands at all (and no controller reaction while turning the craft by hand)?!?
I did this together with a college of mine (controller specialist :P ), but we had no success in controlling yaw
(either by controller setup or yaw stick commands). After trying for nearly 3 hours we gave up.

Back at home, I'm trying to analyze the problem.
What I found out so far:
After lowering the Throttle Scale factor from 0.7 to 0.3, the copter reacts to yaw stick movements
(I do the testing with all props removed).
So I set Tilt Rate PID and Tilt Angle PID all to zero to elliminate any influence and put the Yaw Rate PID
and Yaw Angle PID to the default values.
When I start the motors, the copter reacts to yaw stick movements. But with time passing by, there seems to be
a bigger reaction time between sticks change and motors changing speed. After some time, there is no reaction to yaw stick movements any more. If I start rotating the copter in yaw axis, I notice the yaw controller is still active (motors change speed as direct reaction to the rotation).
I will post the log file of my testing.
yaw_problem.jpg

From 0 - 2000 the motors react to yaw stick movement.
At about 3300 there seems to be a saturation effect (at -180, which is my Yaw Rate OM?!?) and there
is no reaction to the stick movement at about 4050. At about 5500 I rotated the copter by hand and had a controller
reaction, but still no stick reaction at about 6700.
Some informations about my copter:
Y6 with SimonK ESCs,
AHM 36-6 motors with 12x4.5 CFK props.
4s 2450 mAh battery
Mixing: Hexa coax default in GCS 1.2.1

After stopping the motors and starting again, it is the same as before: reaction to stick movements for
some time, after that: no reaction any more. There seems to be a windup effect which is resetted by stopping
the motors.

Any help would be appreciated...
Daniel

P.S. Is there some kind of diagram of the control loops (my controller specialist college is asking me this all the time ;) )?
This my be helpful to find out where the saturation is happening.
daniel_k
 
Posts: 13
Joined: Fri May 03, 2013 11:45 am
Location: Hamburg, Germany

Re: strange Yaw saturation

Postby daniel_k » Sat May 11, 2013 7:56 am

Hi,
did some further testing.
What I found so far:
My rudder scale factor seemed to be too low (don't know why, did a parameter reset in the past).
After changing it to 0.05 (same as pitch and roll scale factor), the saturation problem seems to be gone?!?
Throttle scale factor is set up by trying to reach about 700 for throttle while hovering, right?
How do I set up correct scale factor for pitch, roll and rudder?

Daniel
daniel_k
 
Posts: 13
Joined: Fri May 03, 2013 11:45 am
Location: Hamburg, Germany

Re: strange Yaw saturation

Postby chschmid » Sat May 11, 2013 8:08 am

Hi Daniel

Good to hear you found the bug.
Throttle scale factor should be as you say. The result is trottle PW 700 @ hover. If this is set, changing flight modes does not change mot RPM.

I do not know if you should change scale factor for pitch, roll and rudder.

Cheers
Christof
chschmid
 
Posts: 1800
Joined: Wed Jul 11, 2012 7:41 pm
Location: Herrliberg, Switzerland

Re: strange Yaw saturation

Postby Max » Sat May 11, 2013 8:29 am

All the scale factors are simply multiplied by actual radio input value to get a desired final value which influences amount of movement in manual control mode.

Code: Select all
      if (navData.mode != NAV_STATUS_DVH) {
          controlData.userPitchTarget = RADIO_PITCH * p[CTRL_FACT_PITC];
          controlData.userRollTarget = RADIO_ROLL * p[CTRL_FACT_ROLL];
      }


Similar for yaw and throttle. The reason it's important to get factor throttle correct, is simply because in auto mode, the AQ uses a pre-defined throttle value of 700. So if you're hovering at, say, 500, and switch to auto, the increase of 200 in the throttle value will cause the craft to rise until the AQ can stabilize the altitude. The effects of the other scale factors are simply less noticeable because there's no transition effect when going between auto/manual modes.

The control factors are used almost exclusively in control.c, except throttle factor which is also used to set maximum throttle value in control.h.

HTH,
-Max
Max
 
Posts: 2814
Joined: Mon Aug 13, 2012 9:45 pm
Location: Near Ithaca, NY, USA

Re: strange Yaw saturation

Postby daniel_k » Sat May 11, 2013 8:43 am

Hi,
thanks for the quick answers!
Which scale factor do you use for pitch, roll and rudder?
Mine were somehow messed up be resetting them to default values, I guess (or I messed them up ;) ).
My values are:
throttle scale -> 0.35
pitch scale -> 0.05
roll scale -> 0.05
rudder -> 0.05

Daniel
daniel_k
 
Posts: 13
Joined: Fri May 03, 2013 11:45 am
Location: Hamburg, Germany

Re: strange Yaw saturation

Postby Max » Sat May 11, 2013 9:39 am

Hi Daniel,

Default for yaw scaling is 0.0004.

I've never tried changing those from defaults. I get plenty of control authority in manual mode, enough to get me into trouble as it is :) I'd probably adjust that from my radio first anyway (throws/expo/rates), and only mess with the scaling settings if I can't get what I need from radio adjustments.

-Max
Max
 
Posts: 2814
Joined: Mon Aug 13, 2012 9:45 pm
Location: Near Ithaca, NY, USA


Return to PID Tuning

Who is online

Users browsing this forum: No registered users and 3 guests

cron