Any Carbon Fiber frame…
That’s what they all say. They are like wet noodles.
3 Likes
And the main frames that are just thin PCB don’t help either.
1 Like
I guess the PCB frames are mostly DIY and for sub 250 or sub 200g builds. So may or may not be a problem
The F330/F450 have PCB plate frames. Thin FR-4.
This frame was <$25 Built many a quad with frames like these.
This all was fodder for the ever popular Hardware To Avoid Pinned Post.
1 Like
Oh those PCBs seem good enough in my case (1.8mm) unless I’m mistaken. The arms do flex though
Sir, I am considering reorganization of every component in the drone and I have these options
-
keep escs below motors and bring 2 dc wires and signal wires towards frame. Mount flight controller to frame(one of the aluminium plate),(centre of gravity) rather than using double decker platform.
-
mount escs below the flight controller, with 3 motor wires close to flight controller(2 cm) distance. Mounting flight controller to the frame same as first option.
-
keeping everything same as before, the noise peaks are above -50db even after configuration of notch filter, does that mean I have to compulsory change the mechanical configuration of the drone to reduce vibrations? because the peaks above -50 DB means that vibrations are above the acceptable range?
4)Please state a general rule while assembling a copter respect to ESC mounting, flight controller position, the routing of esc, sensor, wires. @amilcarlucas @dkemxr
Go with option 1, it provides better ESC cooling, reduces wire weight, improves your FC position, and should reduce vibrations.
Option 3 is not possible, the frequencies are not a multiple of each other, remember?
- If you use 4-in-1 ESC put it in the center.
If you use single ESCs put it in the arms, below the prop airflow, and use a capacitor in the voltage input of the ESC.
1 Like
Seeing 2 peaks instead of 1 usually means that 2 motors works harder than other 2, the reason is wrong CoG or arm twist, so motors compensate yaw.
To add some additional rigidity you can add 2 or 4 copper tubes or plates between arms, that might help.
And I think you should use any modern 4in1 esc with bdshot, this will simplify filtering alot.
1 Like
I did the changes and these are the results, there is a significant yaw bias, i thought i half burned one of the motor(motor 3(motor D)) during testing, there was a smell of copper enamel burning but that motor is still working, and the motor 3 and 4 are working hard vs motor 1 and 2.
secondly, i checked the notch filter and still there are no multiples to be operated on , the black cover(boxes/housing)
might be the reason? what do you think?
here are the logs:-
one more question :- in the filter review tool @iampete the filters only attunuate the hntch_freq in the graph i.e the frequency peaks, but as per the bandwidth the graph shoud show attnuated peaks from a range to a range rather than only showing the peaks attunated. for ex :- hntch freq is 40 and bw is 20 that means the filter will filter all frequencies from 20hz to 60hz but in the graph only 40hz peak is only attuated rest are lowred just a bit, so the graph or parameter bw is true?
I cant do anymore spending now. And modifying the frame can cause arms to break under stress, there is already a hole for the wires. Any other way?
Yes there is and it looks bad. But, you have had this problem before so you know how to fix it.
1 Like
Yes sir, but this time I will permanently fix it, it might be because of faulty motor so I am changing it and making new arms
The notch is a V shape, not a block shape, if that helps you understand what is happening
1 Like
all this time reason was a faulty motor, when i tested the motor with 17inch propeller that was beyond the motor’s limit the winding enamel got burnt and the motor started performing abnormally, the yaw bias has been fixed and this time the peaks are multiple of each other. is this configuration correct,
- I zoomed into the first peak and put the frequency same as the peak and then i adjusted the ref till the notch 1 is matched to the frequency peak i did it till 4-5 decimal places, is this correct method?
- how should be the ideal frequency resonse graphs?
-how much vibration levels are acceptable? is that vibration spike ok?
Yes, you can do that. Or you can use the formula provided in the documentation.
Remove the 2nd and 3rd harmonic
Reduce BW to 20
Activate multi-source
The documentation explains it clearly:
- Vibrations over 30 are very bad
- Vibrations over 20 are causing issues even if you don’t know it yet
- Vibrations over 15 are in a grey area - it could go either way - check clipping, it must be zero
- Vibrations below 10 are good
That is when the vehicle touched the ground during landing? Then that is normal and expected.
Just post a link to the log.
- what is multi source?
- sir the 2 and 3 harmonic peaks are present and above 50 DB so why remove the 2 nd and 3rd harmonic option in hntch_hmnch?
You mean when it landed? Come on dude.
What the docs say it is. It works for Throttle notch also. Notch Filter
I would go with this. There is a 3rd harmonic also but probably not worth addressing as it’s below -60 dB. Finally you are making progress with this thing.
1 Like
