I had the same problem with oscillations as you have on a 30-inch engine with a 90 kV quadcopter.
The problem was solved by raising the parameter D
ATC_RAT_RLL_D, 0.01
ATC_RAT_PIT_D, 0.01
Now he flies perfectly.
1 Like
Hi Leonard,
When you talked about noise in the rate controler measured via RATE.Xout, I have seen that the Aout usualy goes up to a level than stays there for the duration of the flight. The Aout is usualy a bit higher that the P,Y,Rout values as well.
Also, how low would I dare go regarding the gyro and RATE filters to manage the noise measured via the RATE.Xout plot?
Thanks!
Aout is throttle so it will always be offset from zero.
It is very important to differentiate between noise and control output here. Control output looks a lot like noise and you should not expect to tell the difference unless high rate logging is turned on. You need to zoom into the trace to the 0.25 to 0.5 second scale to see the noise around the control signal generally.
This depends on aircraft size and how much noise you have. If I was pushed for a rule of thumb I would say don’t go below half the suggested filter values for the frame size you have.
Noise on the output is not something that is the end of the world. It tends to make an aircraft move around a little more than it should but reducing filters does similar things so it is a trade off.