That FFT notch isn’t working for you. Configure a Throttle Notch. The 1st peak ~20Hz may be too low to address with a dynamic throttle notch so set the center frequency on the ~30Hz notch with 2 harmonics. And set the Raw Log bitmask to 9 so you don’t have ridiculously large files, like you do now, that do you no good.
And it’s pointless to show an FFT graph to 1600Hz. Window it to 400Hz so you can see what’s significant.
sorry, i sincerely have no idea what i am doing. I don’t have the esc telemetry data yet,but i have order the hobbywing datalink v2 to look at the current comsumption.
No, that is not it either.
Stop doing flights until you have the datalink.
In the mean time watch videos about notch filters and read the documentation
Hi, my datalink had arrived. But it seem the esc notch filter isn’t using the esc rpm to track.
i have upload my amc,my last flight log,and the filter review result: https://drive.google.com/file/d/12H1vMK2Fp8DiO9QOuWrifBqFd1UtQB2r/view?usp=sharing
Thank you
Thank you for helping. I use hobbywing x9 plus motor that have 40 pole, shouldn’t i set the parameter to 40?
And the notch parameter is basically i need to make it more narrow and attenuate less because it already have low amplitude?
I’m not sure but I think not, hobbywing have historically some strange things with RPM reporting, if with current settings you select 2nd and 4th harmonic it attenuates reasonably well. Usually drone this size just don’t use additional RPM notches and have just LPF 20hz or lower. It’s important that it’s not very lightweight, so RPM is not low, than LPF is enough.
I recommend starting from scratch with the help of AMC, it have some predefined configs with similar configuration, there is too much things wrong in your config now: PIDs, PSC_ACCZ, MOT_THST_EXPO, compass #2, etc
It’s not safe to fly prior all this fixed.
oh okay thank i will be using only LPF then, i did start from scratch using AMC 2 days ago because i change my pixhawk 6x due to it having i2c internal error, all of the parameter i change with guidance of AMC.
Out of curiosity what Datalink do you have?
Ok, I see now. WIth AMC you need to carefully read and understand what’s written in all docs it’s referred. Besides wrong poles count your error was to keep high attenuation / bandwidth but enabling Multi-source, it gives too much delay - 460 degree at 22 hz which gives at least 60 ms delay only in filtering.
Additional RPM filtering might be helpful but just LPF gives enough attenuation for now - near -50db expected.
You also have throttle controller oscillation, need to tune
PSC_ACCZ_P = MOT_THST_HOVER (~0.15)
PSC_ACCZ_I = PSC_ACCZ_P * 2 (~0.3)
Regarding MOT_THST_EXPO, afaik it’s already linearized on hobbywing integrated motors, so 0.1 is more correct, but I did not faced problems keeping it as autoconfigurator recommends.
Compass 2 is FC internal and it gives readings affected by high current wires / batteries on copter, disable it.
Regarding PIDs, from my experience it needs much higher D to have P/D balance correct (P/I 0.15 D 0.01), after filters setup running quicktune works OK and is quick way to have some basic PIDs.
Very very true. And to prove you did that, you are supposed to change the reason changed field for each parameter you change.
@taetea use AMC do the fixes that are suggested on this thread, and post the resulting file .zip file again. I will review that one.
here is my recent AMC:https://drive.google.com/file/d/11rsFww23pNyqY5sHNQYJB5GtWBTYrWQE/view?usp=drive_link
06_telemetry.param (34 Bytes)
07_esc.param (1.2 KB)
Corrected errors in the two files, re-upload them.
And fix the clear error in the 15_motors.param file. You did not read the documentation and/or you did not follow the procedure.
I stopped looking after file 15
I seem to forgot to change the motor setting in AMC, Thank you for helping.
okay, so i turn off notch filter, set PSC_ACCZ_P/I to my hover test value, turn off compass 2 and set my current pid to the point where i dont see oscillating.
Here is my last flight, https://drive.google.com/file/d/1IqAopbJO1-Hfm-d8IDgKac81NTxYvY0c/view?usp=sharing
it seem that only LPF is not enough? when i fly i hear a high pitch sound, could this be the cause?
oh i think i might understand a little better, so right now the esc send the wrong rpm because from the prefilter it seem to be that the 1 harmonic is 34 Hz, but the esc track at less than 25 Hz(because INS_HNTCH_FREQ = 25 the notch is stuck at 25Hz). So that would mean that the esc understate the rpm right?(but why we decreasing the pole? shouldnt we be increasing it?) I will measure the rpm using tachometer tomorrow.
So, 40 counts for one revolution vs 20 counts, if you really have 20 counts but your esc thinks you have 40, you will spin twice for every one “revolution” that the esc reports. Hence an under reporting. Adjust the esc counts down.
If you really have 80 counts but have told the esc that you have 40, you will count two “revolutions” for every one true revolution. Hence an overreporting, adjust the esc counts up
