Good day all
I finally managed to get my FFT to work.
I am trying to reduce vibrations on my quad : 450 frame , pixhackV3 , 2213 920 kV motors ,
1050 props , BLHeli32 30A ESCs
I raised the throttle to just under halfway ( props off !!! ) for a few seconds. I found to main
frequencies : at 100Hz which I assume is the motor. Then there is some frequencies at 300Hz.
I should imagine that, with props on I can expect twice … i.e. 200Hz. So the way I see it , if there are 300Hz frequencies on the AccX … this implies a 300Hz vibration in the roll-axis ??
I tried uploading the BIn file but it didn’t seem to work ??
Anyone perhaps have some comments ? Next I intend to draw the FFT graphs WITH props
and then compare the with/without plots and hopefully make something out of it.
You need to test with props, combo motor props you need to evaluate, remember to balance your props, if you still have undesirable vibrations you can test motor by motor and balance it, some videos on you tube how to do it. You can isolate your FC with some dampening method like this http://ardupilot.org/copter/docs/common-vibration-damping.html
To echo what cala2 said, you need to perform a flight before you can get meaningful data from the FFT. Generally the main excitation frequency will be caused by the motors, with another frequency being twice the main frequency caused by the blades passing over the arms.
Fly the quad around, and then post the data and we can go from there
Yes , Cala2, I am also looking at the FC. I have changed by damping there to some material
in the corners. Previously I had a lot bigger sized material.
Here is a short flight. Setup : 450 frame. 2213 920kV motors. 1050 prop. BLHeli32 30A ESCs.
Motors mounted with M3 silcon o-ring between frame and motor.
I will try and upload the BIN file but for one reason or another, it says “saved” but nowhere to
be seen.
Hello john: its not convenient to isolate the motors from the frame, its better that they are hard mounted and the structure from motor to motor need to be as stiff as possible, only dampening the FC.
Thank you for your comment. Yes , I did some tests and it seems that I don’t gain much
in damping as well.
I am now checking another prop on the same motors ( 2213 ) … this time the 9450 carbon
fiber self-tightening props.
What I also did, was to check the balance of the props ( self tightening props ) and some of them were slightly out. But what I also measure was the “wide” part of the prop close to the
shaft and compared the to sides of a props and found that that part of the measurement can
differ with as much as 0.2 mm ??? Is that enough to cause uneven lift and in doing so, add
to vibrations ? I repeated the measurement with a 9450S prop ( DJI ) and they were spot-on for both sides. Which explains why I had such low vibrations with the DJI 2312S motors with
the quick-release 9450S props !
The strange thing is, my quad’s flying seems fine with both the 1050 and the 9450 props. In
fact , they actually seem very stable. But if I look at the logs, there is something that concerns me … if is the AHRS(roll …NKF1(roll) …NKF6(roll) graphs. They dont stay together …
I have balanced the props best I can. Will see if things improve, but I am also going to look
at double dampening on the FC … i.e. …
frame … damping … aluminium-plate(for mass) …damping … FC
Whenever I try and upload a BIN-file , the percentage monitor counts up to about 35% and
then disappears. When i zip the BIN-file first , it happens at about 50% ???
I feel embarrassed that I cannot upload a BIN-file !! All I’m doing is clicking upload ICON and
selecting the file ???
OK … here we go…
Opened a dropbox account.
This file is the log-file of flight before I upgraded the damping.
Although the file above does not have clipping (clip0,1,2) there are some of the logs where
there were some clipping. Not serious, but it was clipping.
And this LOG is of a flight after I upgraded my damping to a double damping system:
frame to aluminium plate : soft foam-like strips (3mm uncompressed )
Aluminium to FC : double-sided tape 1/2 square inch in each corner (3,5mm)
FC pixhackV3 ( so it also has internal damping and temp control)
motors 2213 (LDPower)
props 9450 self-tightening carbon fiber
4S battery
ESC = BLHeli32 30A
a Dramatic improvement in the Z-axis. Also improvement in the X and Y axis but less
pronounced. Not sure why …?
One thing that is better is the AHRS vs NKF1 vs NKF6 values on the roll, pitch and yaw.
Still some work perhaps on the roll ??? But beter.
I know the damping material is not the same as the types and names that I have found in
many articles that I have read on the internet regarding damping, but those are not available
in South Africa. I thought I would start somewhere … and perhaps move on to the more
sophisticated material by ordering it online.
