Copter not steady (even with light wind)

ArduCopter Copter 4.3
1

Hello,
I’m having a problem with my first attempt at a 28" quadcopter.

My current build is:

  • ardupilot cubeorange
  • T-motor U8II PRO 100KV (propeller 28x9.2)
  • T-motor ALPHA 60HV 12S
  • Two Here3 GPS
    -Battery 12S Lipo 10000Mah

I have completed autotune, on all axis, and the flight was smooth on STAB, LOIT and Althold.
When wind gusts begin, copter goes almost out of control, doing sudden movements and drifting away.
Changing back to AltHold didnt’ solve the issue, so I had to turned to STAB in order to be able to land it safely.
Even in STAB though, the copter wasn’t flat and required gently moves to land it.

I’ve done a lot of test flights with different mot_thst_expo value, mot_hover, psc_accz_ but i can’t think of any more tweaks.

Here my bin log
bin log

Thanks in advance

2

Vibration = bad
Especially Z axis, which is the axis that will give you most trouble.
I advise NOT to fly again until you’ve taken some positive steps to reduce the vibrations. You do not want to risk a fly-away, although vibrations are not the worst we’ve ever seen (by a long way).
These big copters are too expensive to take chances with.

Motors are hitting minimums while trying to maintain stability
Try these:
MOT_SPIN_ARM,0.12
MOT_SPIN_MIN,0.15
MOT_THST_EXPO,0.4

Set these for additional tuning
ATC_RAT_RLL_P,0.1
ATC_RAT_RLL_I,0.1
ATC_RAT_RLL_D,0.0045
ATC_RAT_PIT_P,0.1
ATC_RAT_PIT_I,0.1
ATC_RAT_PIT_D,0.0045
INS_ACCEL_FILTER,10
INS_HNTCH_ENABLE,1 ← set this then refresh params to see the rest
INS_HNTCH_MODE,1
INS_HNTCH_REF,0.20 ← will need adjustment
INS_HNTCH_FREQ,35
INS_HNTCH_BW,17
INS_HNTCH_FM_RAT,0.7
INS_LOG_BAT_MASK,1
INS_LOG_BAT_OPT,4

Current sensing is wrong, see if you can fix that, it’s very handy to have. Or dont worry about it if you are just using the power brick for voltage. In that case set BATT_MONITOR,3

Set BRD_BOOT_DELAY,3000 to ensure those GPS units are detected properly.
Position fix and reliability might be better with
GPS_GNSS_MODE,5
GPS_GNSS_MODE2,5

Definitely set these
BATT_FS_LOW_ACT,2 or 3
FENCE_ENABLE,1
A GPS 3D Fix and setting Home can take a while, but that is better than the copter disappearing over the horizon when it tries to RTL

When you do fly next the Hover Trust will take some relearning, so just hover around in AltHold for a while and let it find the new hover value. Then try Loiter and some gentle movements.
Let’s see the .bin log file.

1 Like
3

I went through the settings and made a flight (stab & loiter) with gentle movies.
My flight was smooth, copter seemed pretty stable, however, today there was no wind at all.
Because the propellers are large and close to the main frame I’m thinking of putting the motors below the frame to reduce vibration.

Here is the new log.
Thanks

EDIT: I’m going to update this with a new log ASAP, the intensity of the wind increased.

4

Adjust these next
INS_HNTCH_FREQ,31
INS_HNTCH_HMNCS,15
INS_LOG_BAT_MASK,0
INS_LOG_BAT_OPT,0
Harmonic notch filtering will be done then :slight_smile:

I’m not impressed by the vibrations though. Z axis is still a bit worse than ideal, and Y axis is in there too, so something might be rubbing or pulling on the flight controller. Wires pulled too tight, or too loose and vibrating around?
Can you upload a pic of the flight controller and mounting?

If you can sort the vibrations a bit more, and you thought the response to RC is snappy and the copter seems stable, you should be able to try Autotune. I would do just one axis at a time if you get to that stage.

5

Ok, here it is the new log

I have to say that I’m a lot more satisfied with stability and more confident.
Ιt is worth saying that this flight was made with stronger wind and many gusts.

I’m thinking of installing longer arms in order to avoid the vibration.
Should I try Autotune now?

Again, thank you for your time, your instructions were on spot! :slight_smile:

IMG_1382
IMG_13824032×941 313 KB

IMG_1380
IMG_13801920×2560 345 KB

IMG_1379
IMG_13791920×2560 481 KB

Update: I think I have identified the cause of the vibration, it’s cause of motors and base metal to metal contact.
I will try to apply some damping.

6

Hey again,
I post this to share with you two logs after autotune completed.
Copter is steady and I’m satisfied with it’s responsiveness.
It seems that I need to study a little more the procedure for setting the filters :slight_smile:
Again, thanks a lot.

log 1
log2

7

Adding some anti-vibration or damping between motors and motor mounts has proven to make matters worse.
You count probably space those motors up on the mounts, or you might be able to find a prop mount extension so the prop sits up higher above the motor.

8

These values are exceptionally low, I believe the Z axis vibrations are affecting the Autotune result.
ATC_RAT_PIT_D,0.001
ATC_RAT_RLL_D,0.001
I think you could safely set them to 0.002

Also set these:
PSC_ACCZ_I,0.6
PSC_ACCZ_P,0.3

9

As Shawn said, don’t do that. You want everything accept the Flight Controller mount (or it’s internal isolation if it has that) to be as rigid as is practical to achieve.