@Leonardthall had mentioned his desire to start characterizing both multicopters and helicopters in the frequency domain. I am opening this topic to be the discussion thread for those participating in this characterization work.
Leonard was looking to provide a number of excitation signals to include Chirp function, sinusoide stepped in frequency and Maximal length BPSK sequence. I have incorporated a chirp function in Copter 3.5.3. I was able to run some test flights with my Ergo flybarred heli and get some frequency response data.
Aircraft and Setup
Ergo .46 converted to electric
E-flite 700 heli motor
Castle Creation 80Amp 50VDC ESC
12s 5000mAh Lipo batteries
rotorspeed: 1800 RPM
rotorhead: Miniature Aircraft X-cell 30 2 bladed flybarred head
weight: 10.5 lbs
swashplate servos: Futaba 3050 digital servos
Here is a graph of the input chirp function and the output to the mixer. It increased in frequency from 1 to 60 rad/sec over 90 seconds with an amplitude of 0.08 on Rout.
So I conducted the chirp in both the pitch and roll axes. The data was collected at 400 hz using the fast ATT logging bit. I had added the sweep as a signal that was logged with the RATE data. The data was processed using the Student version of CIFER. I had it down sample the data to 250hz for the analysis.
Here is the bode plot for the pitch axis
Here is the bode plot for the roll axis
It is interesting to see the hump in the top plot (gain plot) in both axes which is indicative of a lightly damped oscillatory mode. so if the controller was to drive the aircraft at that frequency then you could see some oscillations. It can also be noted that the roll axis has a greater bandwidth (follows your inputs up to higher frequencies) than the pitch axis which is expected. One of my questions is whether the magnitude and phase drop at higher frequencies is due to the servo or the aircraft. I don’t know what the frequency response of these servo are.
So I have a Y6 multicopter and I am interested to see how frequency response compares to these.