I didnt see any stand-out issue in the log.
Maybe there is something loose causing that poor roll axis tune, or the flight controller is over-damped particularly that direction.
Please provide a photo of the flight controller, wiring and surrounding area.
Hi @xfacta , I have done one more auto flight after your suggestions ,is there any changes to be done please suggest. I am attaching a logs as well as FC image below.
I have attached the image of FC above and I would like to ask few suggestion from you regarding dampening other than what ardupilot document is suggested, by seeing my FC is there any possibilities that it can be dampen in better way?
Your vibrations in that latest log are not really bad, but are less than ideal, so anything you can do to improve the vibrations will definitely help your copter.
The whole time the copter is flying all the wires are vibrating and moving around.
All the wiring should be secured to the frame so it’s not putting weight on the connections and solder joints. For example the ESC solder joints should NOT be what holds all those wires in place, or we’ll soon be investigating a crash log.
That bunch of servo wires connected to the X7 outputs will be transferring vibrations - so it needs to be secured to the frame, then have just enough length and flexibilty to the connectors so the wires are not transferring vibrations or pulling the X7 one way or the other.
The GPS wire is another one that transfers vibrations - it’s usually a stiff heavily insulated cable that people leave to do whatever it likes during flight. I usually cut off some of the outer covering so there is an extra inch (2cm or 3cm) of the small flexible wires exposed. And you need some way to secure the cable while it’s plugged in and the copter is ready for flight.
What is the flight controller mounted on now? The X7 usually doesnt need anti-vibration damping unless the frame or something is particularly vibration-prone.
You can leave these set:
INS_LOG_BAT_MASK,1
INS_LOG_BAT_OPT,4
I would just run Autotune after you’ve done some more to fix vibrations.
We have 3d printed and attached silicon rubber for dampening. Earlier we hard mounted CUAV X7+ FC but observed vibration above 30. So after applying dampening vibrations are normal. But whenever we are tuning, if tuning give poor values or overshoot. We tried to increase D values but when increased diagonal arms vibrate more but even though arms vibrate in GCS the vibrations will show normal. After if we decrease D values arms vibration will go off.
Moreover even after Autotuning many time I am getting result tight in one axis than the other. For tight tuning I am doing a quick tune after autotune done. Is there any problem that may causing poor tuning?
Yes, there is something causing the issue. The incorrect sequence of steps.
Do not do quicktune after autotune, do it like this:
- perform Quicktune once only!!!
- perform autotune in roll, pitch, yaw, yawd
- perform autotune in roll and pitch
- done
Even after doing multiple autotune’s , if PID’s results in poor tuning what should be done in such case?
Change the autotune aggressiveness.
Tried with 0.075, 0.085, 0.09 and 0.1 but in one of the axis PID’s giving loose tuned values.
Then there is too much vibration noise on that axis and/or your notch filter is not doing it’s job.
You are using a very old version of the configurator, please update it.
The flight controller might be over-damped.
Pay close attention to what I said about all the wiring, and exactly how it attaches to the flight controller.
Is there a way to check the dampening manually to avoid over dampening and how to to know that the dampening has done over in test bench or in a flight.
Would you please tell me what are the settings to be changed for updating bi-directional DShot firmware for my copter.
That is explained in the Blog Post linked from the ArduPilot methodic configurator
Have you read it?
Thank you, I did read it.
@amilcarlucas Is there a way to check the dampening manually to avoid over dampening and how to to know that the dampening has done over in test bench or in a flight.
Hello, Today I have tried different dampening using 3M foam, Du bro and hard mounting CUAV X7 FC. What I observed is when tested with 3M and du bro foam and did quick and manual tune, on roll axis PIDs are getting low and D and P value is getting to low compared to pitch axis especially D on roll axis. If I increase D on roll axis am observing oscillation in arms. Vibrations are below 30.
But when hard mounted PIDs are ok in quick tune but vibrations in Y and Z are going above 30. but when using vibration dampener it is less than 20. I am attaching a logs which I have done 4 test cases today, please have a look and suggest me what best can be done.
Agree, The cube at one time is not to be isolated, Try hard mounting it with 3M tape. I also would rethink the way your arms are installed with adding more attachment to the frame with two points of clamping.
Tried hard mounting with 3M tape but observing vibration in Y and Z axis above 30. And about frame the arm is attached to carbon fiber body with support of aluminum body mount. When the top enclosure is tightened the arms will be stiff. Moreover if we add foam to FC no vibrations observed, but roll PIDs geting low compared to pitch. when given roll command it starts oscillation.
@xfacta I have done some R&D and used dampening, now the vibrations are below 15. Also as per your suggestions above did Quicktune and then moved towards autotune. Once I did autotune for all 3 axes. then again autotuned roll and pitch axes but when I fly in LOITER mode copter start oscillating. In both pitch and roll axes. Is something wrong went with frame or filters? I am attaching log below of autotuned.
After sometime I manually increased PID values and made copter tight. but this time when I maneuver at 10m/s and release the stick it is oscillating and making some jerk in roll and pitch axes.
Later i tried to decrease the PIDs but this time the copter is overshooting in loiter. Later I changed to above values which is tightly tuned. But after sometimes I just gave roll command this time copter started oscillating and crashed. I didn’t understand what went wrong. as I used above log value and did maneuver at 14m but nothing happen but only observed oscillation at jerk. But here after takeoff at 3m i gave roll command and it crashed.