I have a T-Drones M690 with Cube, that I had to manually tune, as the Autotune does not work at all on it (I tried 3 times, and always ended badly). It is flying ok now, but there is an aggressive oscillation when I stop it fast (releasing the stick). Does someone know how to correct that?
Here is the video and log:
Please send a BIN file with ATT_FAST logging enabled
vibration in x and y axis are high with clipping, it will affect IMU and EKF measurement and copter could not brake good or keep altitude
Thank you @andyp1per and @hosein_gh !
I made two flights with ATTITUDE_FAST logging and INS_LOG_BAT_MASK = 1. First flight was shorter and smooth, with Alt_Hold, and the second of a few minutes, faster flight.
The logs files:
You have the FC in LOITER and then you give a HUGE roll input and go screaming sideways and then you let off the stick and now your complaining that the FC can’t stop the quad fast enough?
Guess what? It don’t work like that. Ever hear of inertia?
You should take out the gyro noise with the harmonic notch and then try and tighten your PIDs a little. The tune looks ok to me, but not fantastic.
Thank you, I will try that and post here the results. How do you check the quality of the PIDs on the log (what is the entries to compare)?
Well you need to switch on PID logging to truly tell
RATE will give you some indication - you want to see the different axes tracking well the desired values and you want to see as little oscillation in the OUT values (i.e. below 0.1 or 10%)
I configured the harmonic notch, and the FFT looks better, but the oscillation on the flight even looks worse. Is there anything else to do, other than starting tuning manually again?
INS_HNTCH_FREQ = 82
INS_HNTCH_REF = 0.130416
INS_HNTCH_BW = 41
INS_LOG_BAT_OPT = 2
Here is again one log for flight shorter and smooth, with Alt_Hold, and the second of a few minutes, faster flight:
You can see the overshoot and oscillation on roll for instance:
So try increasing D to see if you can eliminate that. If not reduce your PIDs by 25% say and see how that looks. Without autotune its going to be trial and error - although you may find with the gyro noise gone that autotune now works
Thank you @andyp1per,
I will only be able to fly it again next week, but I let you know the results.
I have this same copter, and I too am having trouble getting a good tune. What do you mean when you say your autotune ended badly?
My autotune succeeded after the first attempt with default settings, but I’m not happy with how it flies.
It got wild (hard to control), and I broke a pair of props.
I tried again autotune, and it gave very poor results. The roll was close the what I had, but the pitch was very different, and let the drone uncontrollable.
Here is the PIDs:
@lyon44, why are you not happy? Do it presents the same oscillation on fast stop?
I received some PIDs from T-Drones, that flies way better than what I had, but still has oscilations when I release the stick, and the Attitude is far from Desired on the logs. Autotune produced very bad behavior (uncontrolable), and I am not able to see in the logs what is the best modifications to do. Does anyone have some hints?
Here is the log of the last flight, and some small video of the oscilation:
Here is the log and video of the last flight:
Is it normal to have on the log the PIDR.P sometimes inverted from ATT.Roll and ATT.DesRoll, as in this image?
I’m working on tuning the same frame. For whatever reason it is quite tricky indeed. Would love to know what PIDs did the T-Drones sent you, would you mind sharing?
So far what I can tell is that the ESCs seem a little suspect with this frame. I’m not sure how did it work for you, but in my case it doesn’t seem to support anything but PWM and the OneShot protocol( not OneShot125).
That drone never got better. I think I got a broken one. I have others all, and all fly very well.
Would you mind sharing any PID setup still for reference? Like the one TDrones sent or if you have others that are flying well, maybe one of those?