Oscillations found when adding a camera + gimbal

ArduCopter Copter 3.2
1

Hi Everyone

I need some help.

My Pixracer R15 flight controller has been well autotuned and flys great in every mode on a I-flight 850mm x-class frame.

Now I have added a DJI RS2 Gimbal and sony camera and now when I take off it shakes in hover. Esc and motors are not getting hot but there’s too much vibrations to safely continue flying.

I have flown with much heaver weights and they didn’t shake at all, but they were sitting on the frame directly.

Is the gimbal and camera counteracting the flight controller?

Is there a setting or values I need to add to allow for the gimbal and the camera?

2

Reduce the ACCEL values as per these calcs for minimum take off weight and maximum take off weight.
Don’t go below a 30% reduction though.

new ATC_ACCEL_P_MAX = ATC_ACCEL_P_MAX x (min_TOW / max_TOW)
new ATC_ACCEL_R_MAX = ATC_ACCEL_R_MAX x (min_TOW / max_TOW)
new ATC_ACCEL_Y_MAX = ATC_ACCEL_Y_MAX x (min_TOW / max_TOW)
3

Hey Shawn

Thanks for your reply

My values after autotune were:

ATC_ACCEL_P_MAX=122239.2
ATC_ACCEL_R_MAX=126043.1
ATC_ACCEL_Y_MAX=35389.99
The drone flys great without the gimbal and camera.

Tried reducing the weight battery by 1kg=still oscillations
Tried flying just the gimbal in a locked position and seemed to be ok, but oscillations occur when adding the camera 410g.

Is the camera changing the COG hence the oscillations?

The gimbal mount feels solid with no movement to the drone.

Tried these values with no improvements

ATC_ACCEL_P_MAX=122239.2 x (3.588/6.577)=66686.065
ATC_ACCEL_R_MAX=126043.1 X (3.588/6.577)=68761.235
ATC_ACCEL_Y_MAX=35389.99 X (3.588/6.577)=19306.5659

note:
min_TOW = weight of drone frame + small battery => 3.588kg
max_TOW = weight of drone + gimbal + camera + large battery => 6.577kg

Maybe I did something wrong should I reduce the values more?

Thanks
Razza

4

Please update to ArduCopter 4.4.1 or newer.

5

You’ve reduced that values to half their originals (almost) but I’d suggest you dont go below 2/3rds.
Like:

ATC_ACCEL_P_MAX = 122239.2 x (2/3) = 81400
ATC_ACCEL_R_MAX = 126043.1 X (2/3) = 83200
ATC_ACCEL_Y_MAX = 35389.99 X (2/3) = 23300

as you can see it’s OK to round off to nearest 100’s or 1000’s.
If stability is still an issue you might need to improve the tuning further before adding the gimbal and camera.
Or filters may need some alteration.

A before and after log might help to identify the problem

6

Please update to the latest version of arducopter and try @xfacta Shawn’s suggestions.
I recomend you to move the battery and gimbal with camera in such a way that the Centre of gravity is in the middle of the drone frame and also try to reinforce the mounting area of the gimbal and if possible try to remove extra bit of gimbal like handle and other hollow parts as it will bring all the weight close to the drone frame BUT DO THIS ONLY WHEN YOU ARE FAMILIAR WITH WORKING WITH ELECTRONIC AND KNOW HOW TO SOLDER as you might need to shift some components here and there.

7

Thanks for all the help guys!

Yesterday I followed Shawn’s advice and tried Reducing the values 10% and 20%.

10%= with gimbal and camera mounted and locked the drone responded much better and only sometime it would oscillate.

With the camera and gimbal active it still wobbled but a bit less.

20% = with gimbal and camera locked results were perfect stable flight

Unlocked the drone still oscillated but I think slightly less.

The current setup is battery mounted underneath, FC pixracer in the middle and a RS2 gimbal mounted on an isolating plate on top dead Center.

If I did a firmware update on the flight controller, wouldn’t I have to do all parameters and auto tuning again and calibrations?

Will try those suggestions next

Thanks
Razza

8

No, no need to retune or change any parameters when updating the firmware. The update process automatically updates/changes parameters if necessary.

However, when downgrading firmware you do need to manually upload an parameter file saved with that particular version

9

I was looking through parameters and noticed that I had disabled MOT_HOVER_LEARN=0 would that have an effect on the stability of the aircraft. Should it be set to 2 (learn)?

10

No, not much effect on stability, more on altitude.
Yes, Learn and Save.

11

I removed the gimbal and camera but left the gimbal Mount on, took it for small test flights.

Checked Flight data graph for DES ROLL vs ROLL, DES YAW vs YAW and DES PITCH vs PITCH and all following each other nicely.

Checked the post flight flash logs graph looking for vibrations VIBE x,y,z were within tolerance.

Checked the graph FFT GYRO 0 found only one peak at 186Hz will try and knock that down next with a notch filter.

Noticed the ACC0 FFT had 4 peaks any ideas how to to knock them down too. Or should I just let them be?

12

Ignore any and all ACC peaks. ACC data is not notch filtered.

13

Have you got that log to look at?

14

Hi everyone.

Managed to get it all sorted thanks for all the help.

Turns out along with the max values it just needed some major PID tuning as it had an aggressive tune for the new payload.

Razza