It wobbles at some time only

I am flying a large quadcopter with 15kgs of weight I flied it today but it wobbled at time of flight for some time only. i am attaching the log file for the same please tell me what is happening

https://drive.google.com/drive/folders/1fiCgJetL4kZgBUgUBkTltxaJjfiajcME?usp=drive_link

How exactly have you configured it?

Which steps did you take to tune it?

which components are you using?

How are they connected to each other?

How exactly I configured it

  • first done imu temp Cali
  • setting up the initial tune params
    -intial calibration and test flight
  • hover throttle sett and then notch filter setup
    -autotunning all the axis and testing the tune
    Problem: roll autotunes but pitch autotune too Twitchy mean it will crash any time

All the time the d value is too high and compensation to p and I it doesn’t seems good

Which component I am using

  • I used 12s 22000mah solid state battery with mn8014 100kv motor and t-motor g 30*10.5 inch propeller as propulsion and flight controller as pixhawk cube orange with adsb carrier board

How are they connected to each other

  • actually we are designing a system that detects the object from 100 ft of height and localise itself and drop the payload alongwith it should also travel 15 miles of distance in just 30 minutes

So I attached a onboard computer a gimbal camera and all the compentes are powered through a power distribution board and a pcb


the order is incorrect and you missed notch filters

Please follow: Methodically configure and tune ArduCopter using the software from GitHub - ArduPilot/MethodicConfigurator: A clear ArduPilot configuration sequence

On the software fill the component information in the component editor window. And post the resulting .zip file.

I did it after hover throttle

Sorry, I missed that. I will look at the logs tomorrow than

ARMING_CHECK,1 or at least pick everything and just deselect one thing you are working on.

Your harmonic notch filter is exactly right - good work.

That is a bad antivibration mount for a Cube and carrier. You may not even need it, but if you do then use this design:

Also you need to tidy and secure all the wiring, and remove the USB connector from the side of the Cube. Use the USB connector on the carrier board with flexible wires.
Make sure the GPS cables are secured to the frame and not just “hanging”.

If those ESCs are T-Motor Flame then try MOT_THST_EXPO,0.4 or if they are Alpha try MOT_THST_EXPO,0.6 but if you do change the expo you need to retest and set MOT_SPIN_ARM and MOT_SPIN_MIN and maybe run a fresh Autotune.
If the ESCs are out near the motors then it will be a good idea to add extra capacitors in the power wires very close to the ESCs.

Are those T-Motor Navigator “Long Time” motors? Exactly what model and which props?
I’ve seen a number of them having poor attitude control mostly because people select props that are just too big - chasing that elusive extra minute of flight time.

I would set these parameters:

AUTOTUNE_AGGR,0.08   or 0.1
ATC_RAT_RLL_D,0.0041
ATC_RAT_YAW_I,0.08
ATC_RAT_YAW_P,0.80
BATT_ARM_VOLT,44.30
BATT_CRT_VOLT,42.00
BATT_LOW_VOLT,43.20
BATT_FS_LOW_ACT,2
FENCE_ENABLE,1
INS_ACCEL_FILTER,10
PSC_ACCZ_I,0.70
PSC_ACCZ_P,0.35

and not all of those have anything to do with the wobbling, but I would say they are all important.

Let’s see the next log, and please mention more details about ESCs, motors and props.

I use t-motor flame 80av2.0 esc and t-motor anti-gravity mn 8014 100kv motor with t-motor g30*10.5" single blade propeller

I will try all of these as you suggest and will come back tomorrow

73% of the changes that @xfacta suggested are included in the ArduPilot Methodic Configurator software.

If you had used it you would not need to do the changes now.

1 Like

One thing I noticed that it heavily wobble when it’s cruising in the opposite wind direction so I also missed one thing that is wind speed estimation can it help the quadratic to be stable in the opposite wind

Yes, wind speed estimation helps to stabilize under wind conditions.
Just use the software and do all the steps that you missed. :slight_smile:

@xfacta I would ask that if I would drop the atc_rat_rll_d by 50% doesn’t it affect the Quadcopter to fly stably

The D term is more likely to induce or correct oscillations and errors with the P term. Your role D was drastically different to the pitch D and very much out of proportion with anything sensible.
That’s why I listed a new value for Roll D. It’s unlikely you use different props, motors and ESCs just for the roll axis compared to pitch, they will all be the same brand and model :slight_smile:

1 Like

@xfacta As it because of i tunned roll axis number of time the d rate comes to nearly this value but when I tune pitch axis it d value comes always nearly to roll d term but it’s p and rate p term are so that it’s very twitchy and too tight that it can’t hover stably so we manually tunned the pitch axis and generated the d term