• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

Trim Drone, roll issue?

LitterBug

Troll Spammer
#21
Take the props off and plug a battery in. Go to the output tab tab in GCS. Enable Test Ouputs. Run the slider for each motor up individually and note the value that each motor starts. Slide each one up about 50% and note the tone/speed of each motor. Do they all start at close to the same value, and do they all seem to spin at the same speed at 50%? Also keep track of which corner spins with each channel number and post the results here for that and the motor start values.

Cheers!
LitterBug
 
#22
@LitterBug, will do.

First I need to post what's happening with the settings i have right now.

drönare stabilisation tab.jpg

and this is how it flies...

The flight in this video I found it strange that it took it so long before it stopped climbing and started to descend after i lowered and the cut the throttle off.

Unfortunately i didn't get the second attempt on video since it apparently ran out of battery.
 
Last edited:

LitterBug

Troll Spammer
#24
Wow. #1 sounds really bad. Does that cause the whole frame to shake? That could be causing the gyro to go nuts. Make sure it is not loose, the arm it's on isn't cracked, and there isn't anything else loose on that arm.

Cheers,
LB
 

jhitesma

Some guy in the desert
Mentor
#26
Yeah, something sounds really wrong with #1. You could try swapping it to a different ESC to confirm it's a problem with the motor and not the ESC...but I'm betting it's the motor. Could be a bad bearing or loose magnet...but could also be a bad/iffy electrical connection or a prop or even the ESC.

Like Litterbug said - a motor that rough will play havoc with your gyros and cause all kinds of crazy stuff with any FC and/or firmware.
 
#27
Now listen to it =)
(And I do apologise for the first few seconds, that sound is awful!)


It seems to me we are getting closer to the core of the issue...

On the account of broken magnet. It seemed like the easiest to start with.
It does look ok to me (video below).
Doing this the washer under the C-washer-thingie (in lack of a better English word for it) came off. This looked coarse to me. Putting it back, I put the more normal looking side towards the motor.
That's the second time i rev it up. way better.
Third attempt ive added some silicone oil, since it seemed like less friction was a winner.
Doing this i turned it upside down lubed it, and lazy as i am, i didn't turn it back. And it was better again.
On the video its turned normal way on third rev, and upside down fourth.

Is this close enough for it to be worth doing another test flight?
And as an additional bonus questions
Why does it sound most normal when its upside down?

 
Last edited:

PsyBorg

Fly Angry
Mentor
#28
All that points directly to a bad bearing from what I can tell. I did not see any scrape marks on the stator or the magnets so a warped shaft is a less likely culprit but still possible. Its not a warped bell as again no scrape marks from binding. I would replace the bearings if you can get them or even have them. If not replace the motor depending on cost between the two choices.

You could do another test but even if it does fly That problem will still exist and surely not get better over time. At least you would confirm if a motor replacement would solve your issue.
 

PsyBorg

Fly Angry
Mentor
#31
Ok I have done a little research on this quad... I found three different builds for it. Most show opto esc's with a bec powering the FC off the Receiver. It seems on your build you are using esc's with built in bec's and have all four of them feeding power into the board.

I would start by removing the red bec lead from three of those pads and only leave one to power the Receiver and FC. That could be causing the becs to act funny and feed back into the esc's and or fight for dominance. In any case that's not good and could cause strange behaviors if it does not flat out fry things.

Make sure you cover those unused ends as they do have voltage on them when the quad is powered up.

Once you do that to arm it again and see if the behavior changes.
 
#32
With just one BEC-lead, well, the top right motor (2) there is no power to the receiver.


With another BEC-lead on (motor 1) it actually acts better, but not perfect.


I think im gonna risk another flight and see if I can get the auto tune data. And that's for tomorrow in hope that its less windy.
 
#33
Ok, im officially about to throw in the towel.

Does anyone have a spare CC3D board to donate?! Gaaah!

On a more serious note. Is there a diagnostic I can run on the cc3d board?

Unfortunatly i dont have enough control to let it do the auto tune wich seems like a really good idea. But i cant make that happen since the throttle is acting as it does. Today i also experienced that was yawing when it should have been pitching..
 
#34
more questions.
Is this article written wrong, or more likely. Is my setup wrong?

http://www.dronetrest.com/t/cc3d-flight-controller-guide/830

the flight board is mounted on correctly, but according to this the motor marked as four on the chart is what starts spinning as number one in LibrePilot and dRonin?

To get this right, i would have to remount the cc3d board with the arrow marking forward pointing towards the right. This seems not like a good solution. The other solution as I see it is to rewire the wires coming from the cc3d and onto the frame making the motors match up with the chart (and not the program)
????????????????????????????????????

Input, please!
 

jhitesma

Some guy in the desert
Mentor
#35
That motor diagram is incorrect for dRonin (and I'd assume libre pilot) that looks like *flight motor scheme. dRonin starts numbering #1 in the NW corner and goes clockwise from there.

However - you can remap those on the vehicle page so it doesn't really matter and you don't have to rotate the board which would be a totally wrong way to address that as then you'd mess up all your sensors.

The proper way to test/confirm this is to use the Outputs page to spin up just motor one and confirm it's the NW motor. If it's not then go to the Vehicle page and change the motor mapping so whichever one does spin the NW corner is listed correctly. (I'd show this on screen shots but am at work and can't do that right now.)

I doubt you've got the motors mapped wrong though as you have been flying somewhat - and if your motors were mapped wrong you wouldn't be getting off the ground at all.

I can't watch videos at work so not really sure what's going on in your last few videos.

If you haven't replaced that one horrible sounding motor yet though I would still be vibrations from that are throwing off your accels and causing problems.

I haven't seen any recent flight videos to see what exactly is wrong at this point...but if it's oscillating on stock PID's you can try turning down the P gain before running autotune. Autotune only works from an undertuned state - so if your gains are too high by default (they usually aren't but there are a lot of setup options out there) then you will have to turn them down before you can get autotune to work.
 

PsyBorg

Fly Angry
Mentor
#36
This is nagging me now...

Do you happen to have a spare servo around? If so could you try plugging that directly into your receiver on each channel and showing us how smooth it moves on each channel? That throttle issue is more and more making me think its radio / receiver related then FC / quad related.
 
#37
@jhitesma. The motor spins according to the chart in the program.
I'm starting to believe the lag/miscommunication within the drone is the problem. Since it occurs without propellers it cant have anything to do with the PID settings, it's on the floor the entire time. Still there is a huge discrepancy between sent from controller to result in motor speed that I can't explain in any other way than the cc3d-bord is retarded.
Follow this adress and youll understand what i mean. https://youtu.be/78vRdhG7O6Y
On an other note, do I need to have a serious word with your employer to be a bit more lenient on your internet usage during work hours? ;)

@Psyborg.
Here it is.
I should probably have chosen smaller servos with quicker response time.


 
Last edited:

jhitesma

Some guy in the desert
Mentor
#38
On an other note, do I need to have a serious word with your employer to be a bit more lenient on your internet usage during work hours? ;)
Heh :) It's more a politeness thing than an outright restriction. We're a small 4 man shop in a one room building. So I can watch videos...but I can't really listen to anything since it's too disruptive to the entire office.

I keep meaning to watch your videos in the evening at home but keep forgetting :(
 

PsyBorg

Fly Angry
Mentor
#39
You don't have any weird mixes or expo settings going on it the radio do you? Seems the throttle / rudder are not linear and slower then the elevator / aileron (providing you are set to mode two) do you? But then again I am functioning on cough medicine and coffee so my perception may be a touch skewed... At least its another thing in the list to check off as good.
 
#40
@PsyBorg
No, I don't have any mixes since its better to let the CC3D do that instead, otherwise it can get confused by the received signal.

Yes, I have mode two set up. That should have been pretty obvious since all motors rev up at the same time when i throttle up, no?