Hi,
I’m trying to build a payload delivery quad and I followed the basic configurator so far. However I’m unsure about what I’m doing with the MOT_THST_HOVER because in one part of the configurator it requests to use the initial flight settings where I set it at 0.25 whereas my initial tune parameter suggested a value around 0.64. I have tried flying with the 0.25 value while following the other steps. I’m not seeing any visible oscillations but my CCW motors are warm to the touch and the drone does not hold the position very well, it flies in quite a bit radius whereas I need it to hold quite steady. I have completed upto the first few flights of the mandatory minimum tuning but have not run through the VTOL PID tune script yet. Till now I have followed every single step to come this far (I Hope I have not missed anything). The roll and pitch inputs seem a bit aggressive and yaw seems very unresponsive at times. I’m attaching my log files from the last few flights for reference (there are some videos in the link too):
Also on the last flight the drone exceeded the fence and completed a nice RTL. The landing is the best part so far, very smooth and reliable. In AltHld the drone drifts away quite a lot.
Vibration looks good, and learned hover thrust is 0.23, which appears about right. However, you have a massive yaw bias. I didn’t dig any deeper than that because this needs to be addressed before any further tuning. First thing to check is proper frame arm and motor mount alignment. The props should all swing in the same horizontal plane.
This is from the only log I reviewed, which appears to be your latest:
You shared .bin files, but what we need to help you are the *.param and *.json files created by the configurator.
Post a .zip file with them. I think you made some errors with the MOT_THST_HOVER and MOT_THST_EXPO
The frame that I’m using is a factory built frame with little room for me to edit. I’m using the JIS frame. Could this be a software thing I need to figure out?
No, you can’t fix a mechanical problem in software. The bias is so bad that motors 1 and 2 are saturating at MOT_SPIN_MIN at times, which will cause stability issues, if not catastrophe.
The prop tips should align when you spin adjacent ones. If they are crooked with respect to one another, you’ll get this type of bias. I don’t know how to fix your specific frame, but I do know that this misalignment is the likely cause of the issue observed.
Also would you mind telling me what you were looking for in a the graph to conclude the Motor 1 and 4 were off? I’m just trying to learn how to read these graphs better.
All 4 motors should occupy similar space on the y-axis in that graph. They won’t be perfectly matched, since they MUST vary output to produce control and stability, but they should typically average the same. In your case, motors 1 and 2 are producing significantly lower mean output than motors 3 and 4, and the flat spots at the bottom indicate saturation at the lower limit.
(edited to correct a mistake I made in numbering - I misread the colored legend when posting the first time)
Alright thankyou so much for the information. I just checked and the motors all of them are slightly inclined one way or the other. That was the only way to put the frame together. But I think I will make some adjustments (DIY) to make all the motors completely parallel to the ground with a spirit level. Maybe that can solve the YAW issue.
You MUST make some adjustments to that misalignment. The bias is significant enough that I don’t think it’s worth doing anything else with this craft until that is corrected.
I got the quadcopter working beautifully and with super stability today just by adjusting the alignment. It was an amazing suggestion. I’m just facing an issue right now that when I try to download logs (via USB through MP) the progress bar barely moves for over 5-10 mins. Any way to resolve this issue?
