I am testing hexsoon td-1100 drone, with u8ııkv100 motors and 2815 t-motor props with pixhawk v6x autpilot.
In first flight, i takeoff with default gains and observed oscillations in roll axis and quickly landed. The log file is below.
Then, i lowered P and D gains, P to 0.9 and %15 decrease in D. Oscillations decreased but still have high frequency limit cycle like behaviour in PID.P and PID.D terms of roll axis. Log is below.
MOT_THST_HOVER is 0.18. I try to understand the main cause of the problem before moving into filter design.
I will appreciate if you can analyze the logs and make suggestions.
Ok you mean methodic configurator. I used most parts, apart from notch filter design since notch freq is too low around 20 and 30 hz. Not sure it is related to notch filter deficieny. Is it reasonable to design notch filter close to controller bandwith ?
@amilcarlucas if you mean setting PSC_ACCZ_I and PSC_ACCZ_P based on hover thrust, i already updated them based hover thrust which is 0.15 for zero payload and 0.19 at 2 kg payload. I used filter review tool and below are the plots. First 2 plots with zero payload and last one is 2 kg payload. At 20 and 30 hz i have problem i think but they are very close to controller bandwith so i am not sure about using notch filter. Any advise based on this would be appreciated.
So i already tuned gains such that i have stable flight but this oscillation i can not get rid off with gain tuning.
My question is, is it safe to design notch filters around 20 hz and 30 hz which are close to controller bandwitdth to reach below -50 db at these problematic frequencies. I did not add error filter, will it be also reasonable to add 8-10 hz FLTE on roll/pitch axis ?
Looking at the size of the drone oscillations at 20-30Hz are not to be expected, the vibrations should be at higher frequencies.
Have you used bigger props than recommended?
Post the .zip file from AMC so that We can see the full list of the used components and the parameter changes you did.
