Hey Shawn, on an Auto flight Sat. 22may, while executing an Auto Land (RTL) I noticed a larger amount of oscillation on descent. Not only during the initial phase of the descent (when prop. wash is a factor) but this oscillation continued through the entire Auto land even during the last few feet when the descent was very slow. I have never seen that before. Usually any oscillation stops when the Quad enters the final phase (slow descent phase of Auto Land RTL, last few feet). Any params that I might adjust to reduce this oscillation?
Also troubling, after completing the Auto Land I attempted to stop the motors i.e. TX Thrtl. Stick lower right corner, the props continued to turn for several seconds after issuing the stop motor command. Again this is out of the ordinary, usually the props stop nearly immediately.
About the oscillation on descent, I’d suggest changing the D term just a fraction as per the “try next” section below. Definitely only do a short test in ALTHold for no more than a minute or so then check motor temperature. After autotune
ATC_RAT_RLL_P,0.0933
ATC_RAT_RLL_I,0.0933 ATC_RAT_RLL_D,0.01145
ATC_RAT_PIT_P,0.111436
ATC_RAT_PIT_I,0.111436 ATC_RAT_PIT_D,0.0135
Try next
ATC_RAT_RLL_P,0.0933
ATC_RAT_RLL_I,0.0933 ATC_RAT_RLL_D,0.01166
ATC_RAT_PIT_P,0.111436
ATC_RAT_PIT_I,0.111436 ATC_RAT_PIT_D,0.0139
I can see the props kept running but I cant see exactly why (Land is not complete). Maybe that will fix itself if we smooth out those oscillations. I’ve noticed with our set ups those sort of propblems went away with better tuning.
The alternative is set a RC channel to be the Motor Emergency Stop with RCxx_OPTION,31
The downside is it needs a dedicated switch and channel which may be a problem with some radios - and you never want to switch it accidentally!
Hey Shawn, modified the ATC_RAT’s and a very small amount of oscillation is detectable during descent. So, I am good with the way it is now. Also, conducted another go at the Vib. reduction (I was not happy the amount of Vib., although it was not to bad). The Vib is now significantly reduced! Also, no more HUD EKF alerts!?
Now finally moving on to pic. taking. Thanks again for all your great help!
2nd I perform a Dynamic balancing, I use a cell phone with a vib ap attached to the Pixhawk (4 nylon stand offs with double stick tape on the ends). Secure the Quad to the work surface with C Clamps attached to each landing strut leg. Use MP to run each motor one at a time and measure the vib for each prop mounted at four locations on the motor (0, 90, 180, 270 degrees). This identifies which location will produce the least amount of vib for the prop/motor.
What I have found is contrary to expected outcome. You would expect that once you do this dynamic bal. the prop could be replaced with another prop. using the same location on the motor. This does not seem to work, have not exactly figured out why as yet.