Log is attached below.
Video from copter
youtu.be/Wa8I8eLglQg
I setup a simple out and back autopilot mission. The flight was successful, however the copter was not tracking the elevation. Instead it would wonder up, then down, then up, etc.
The baro and the gps altitude track nicely, so I am not thinking it’s a vibration issue. The vibration from the acc looks pretty good.
thoughts?
This is definitely vibrations. You just need better vibration isolation on the autopilot.
jschall… thanks for the input. I will review and retest the balance and the autopilot vibration isolation.
I am however having trouble understanding why this is vibrations. My vibrations are well within the ±3 on the XY and ±5 on the Z.
I have posted a vibration graph from MP showing the ACC logs. I put a shaded box around the turn around point to make it easy to find. Notice that at this point, the copter is on autopilot and flying in a stable manner. The vibrations fall to ±1 on the XY and about ± 1.5 on the Z. These vibration levels are way way within the spec of ±3 on the XY and ±5 on the Z.
You can also notice that on take off before the AUTO portion, and after the AUTO portion, the vib levels are quite nice. I may be reading this wrong, and please enlighten me if I am as that may be part of my issues.
Here is the graph of vibrations… are these not in spec?
Well, the frequency content of the vibrations matters just as much as the vibrations themselves.
Unfortunately, no way to diagnose aliasing issues by by looking at the accelerometer data.
When inertial navigation fails, it is caused by:
We can rule out #1 - the usual issue is compass calibration, but in this case yaw shouldn’t matter.
We can rule out #2 because we have 2 sensors measuring absolute height that agree with each other (GPS, baro)
What’s left is #3.