HexaCopter Shaking

ArduCopter
1

Hi all,

I moved the copter from one location to another (Germany-Sweden) for field tests. Copter was flying fine in Germany (all tuned), but when I did a flight test in Sweden I observed some vibrations which I would like to ask your opinions on it.

I did all the calibrations before the test. When the copter was flying, I realized that I forgot to update the ACCEL parameters (i.e., ATC_ACCEL_P_MAX, ATC_ACCEL_R_MAX, ATC_ACCEL_Y_MAX) which were adjusted for the payload attached, however I removed the payload to just check the performance of the copter. I landed and updated the ACCEL params (however, I did not restart the copter) and then took off again, but still the copter did not perform well, the vibrations were there.

I checked the vibrations but seems not so much different from the flights in Germany, below first graph shows the vibrations from test in Sweden and second is from Germany, both seems to be below 15

image
image1364×420 12.2 KB

image
image1364×420 12 KB

Then I checked motor outputs, seems that the differences is a bit bigger in Sweden (first image) than Germany (second).

image
image1364×420 13.9 KB

image
image1364×420 16.5 KB

I am not sure why the motor differences are a bit bigger compared to Germany case, The wind was a bit higher in Sweden but I do not know if that can cause that much of a difference. Could it be because of the restart, do I need to restart the copter whenever I change any parameters? I would appreciate any help.

Here, you can find both logs
https://drive.google.com/drive/folders/1nCXdRaOO1M2_eHn6AsQ9IeandXcps-PP

Best regards

2

A lot of parameter changes are immediate and dont require a restart.
Which calibrations did you do again? Usually you would not change or recalibrate anything unless you made physical changes to the copter.
Even moving a long distance doesnt require compass calibration.

Looking at the Sweden log, you could adjust the Harmonic Notch filter a bit:

INS_HNTCH_FREQ,50
INS_HNTCH_BW,20
INS_HNTCH_REF,0.13

and it seems most of the vibration or unsteadiness in attitude control is coming from the relatively high rate PIDs. What I would do is just lower them by 10% , and D a bit more than 10% since it is quite high in proportion to P and I

ATC_RAT_PIT_D,0.01192161 -> 0.0777
ATC_RAT_PIT_I,0.215838 -> 0.1944
ATC_RAT_PIT_P,0.215838 -> 0.1944
ATC_RAT_RLL_D,0.01507968 -> 0.00824
ATC_RAT_RLL_I,0.2014846 -> 0.1813
ATC_RAT_RLL_P,0.2014846 -> 0.1813

and also reduce the overactive yaw values:

ATC_RAT_YAW_I,0.1113956 -> 0.08
ATC_RAT_YAW_P,1.113956 -> 0.80

Vibrations seem acceptable, certainly not excessive.

And the Germany log:
Vibrations are a bit less than Sweden, but it’s not a major change.
I suspect the same issues existed, like attitude control being a bit aggressive due to high PIDs, but were masked by the payload weight.
The same filter settings will work as I’ve suggested above, with the exception of needing to use
INS_HNTCH_REF,0.2
due to the increased payload. This could be the only parameter you would need to change for with/without a payload once the PIDs are a bit more universal.

In both logs the GPS update rate is poor, GPA.Delta
You could try changing to GPS_GNSS_MODE,65 and see if it’s better
or maybe do a firmware update of the GNSS unit and it might help while using GPS_GNSS_MODE,5

3

Hi Shawn,

Thanks a lot for the answer.

I did all the calibrations except compassMot, (i.e. Compass, Accelerometer, Level, Gyro, Pressure)
I actually uploaded a param file, that I was using in Germany, and right after that I got an issue “compasses inconsistent”. I did again a compass calibration and that fixed the issue.

Afterward, I did test flights and observed that it actually performs fine, still vibrations but not that much. I also did a flight test with the payload attached, which also went alright. You can find the logs in the below link, named as “payload” and “nopayload”.

https://drive.google.com/drive/u/0/folders/1nCXdRaOO1M2_eHn6AsQ9IeandXcps-PP

I am going to try to update the parameters and update here.

Best regards

4

Hi Shawn,

I think you are right, most of the vibrations come from PID, because as soon as I changed the flight mode to AltHold the vibrations disappeared and the copter seemed relaxed.

I did try the parameters you wrote down, but as soon as I took off the copter started to wobble a lot so I just brought it down really quickly (almost crashed).

Then I just replaced the parameters with the ones that I was using and it was much better I would say. I am not sure If I need to do anything now because it performs ok, I mean not perfect but not bad. What do you suggest?

Here you can find the log files, logs are named as “ShawnParams” and “OriginalParams”
https://drive.google.com/drive/u/0/folders/1nCXdRaOO1M2_eHn6AsQ9IeandXcps-PP
Thanks a lot again.

Best regards