[charel]

Hello!

I've finished building my discovery and as soon as I take off, it wobbles like crazy and flips after a few seconds. I've also tried it with the quad attached to two strings to rule out the ground effect. Still the same. It does it in Stabilized1 and 2, and rate as well (though it takes much longer in rate mode to shake).
I'm using the following:

• CC3D (from the first OP batch) with 15.09 LP firmware from the setup wizard
• HK Blue 30A with standard FW (50Hz)
• CC3D mounted to risers on the frame, no padding/insulation at all (this has worked w/o a problem in a different quad)

I've already tried the following things:

• Calibrated ESCs
• Checked that the motors spin in the right direction
• Checked that the ESCs are connected in the right order
• Checked the orientation of the board (also responds correctly in the flight data tab)
• Redid the setup wizard (multiple times)
• Reflashed the CC3D
• Ran the motors at full speed w/o props and it had no bad effects on the sensors (the horizon stayed w/o vibration and no axis drift in the GCS)

I let it sit more than 10 seconds after powering on and after arming to make sure the sensors are calibrated. I've also tried to find a solution online, but no luck so far.

I need to add that I did shorten the props by a few mm because they were touching each other. Though I did balance them, some turned out a bit longer than others, though that shouldn't make a difference as I've flown with 12" and 11" props at the same time before (with this same CC3D) and there's even a video on YT where the CC3D manages to fly moderately okay with only 3 working motors on a quad.

Any ideas or suggestions?
I might try another CC3D I have, or a Revo mini (once I get rid of the Taulabs bootloader).

[TheOtherCliff]

This usually happens because your ESCs aren't plugged in the right places and you are off by one position.  Check how LP numbers the motors and where each motor number should be plugged in.  Different brands of firmware use different default numberings.

It can also be caused by declaring it an X when it is a + or a + when it is an X.

[charel]

• Checked that the ESCs are connected in the right order
• Checked the orientation of the board (also responds correctly in the flight data tab)

I had already done so multiple times before posting here. I went through the setup wizard a few times and also configured the quad manually in each tab again to make sure everything is correct (also spun up the motors individually to make sure direction and order are correct each time).

[TheOtherCliff]

Are you using PWMSync, PWM490 or something faster for ESC signal protocol?

Are all the necessary output banks all set to this protocol / speed (top of Output page).

Are you starting with default PID settings or did you start with someone else's configuration?

Is the wobble circular or just roll or just pitch?

[charel]

I think I've found the problem!
I had mistakenly configured the ESCs to 50Hz. I have now set the update rate to 400Hz (mode still PWM). I will try the new settings tomorrow morning.
I ordered an atmel flashing tool and will flash blheli as soon as it arrives.

Is the wobble circular or just roll or just pitch?

The wobble was pretty circular. I am still using default PID settings.

[AerialPerPlexity]

If your trying to run one shot on a cc3d it doesn't support pwm.... Ppm and something else. Just do your math

[charel]

If your trying to run one shot on a cc3d it doesn't support pwm.... Ppm and something else. Just do your math

I don't completely understand what you mean.

The receiver is setup as PPM(pin8), and the wiki doesn't say anything about PPM not supporting PWM, just that the other modes need a single-wire setup.
https://librepilot.atlassian.net/wiki/display/LPDOC/CC+Hardware+Configuration#CCHardwareConfiguration-Ports
https://librepilot.atlassian.net/wiki/display/LPDOC/PWM%2C+PWMSync%2C+Oneshot+Output#PWM,PWMSync,OneshotOutput-PWM-490Hz

[TheOtherCliff]

You will probably find that PWM 400 fixed your problem.  Personally, I would run PWMSync (1900 ESC signal max) or PWM490.
https://librepilot.atlassian.net/wiki/display/LPDOC/ESC+Calibration

[charel]

You will probably find that PWM 400 fixed your problem.  Personally, I would run PWMSync (1900 ESC signal max) or PWM490.
https://librepilot.atlassian.net/wiki/display/LPDOC/ESC+Calibration

Yes, it did fix the problem.
For the time being I'll keep it at 400Hz PWM until I have the atmel flashing tool to flash blheli.

I also learned not to maiden under a tree when I have no idea what the throttle response will be. The quad shot right from the ground into a branch and I broke an arm again. A lot of hot glue later, I tried again and it flies pretty stable; no wobble, almost no drifting.
It's really slow to respond (about a second), but that'll hopefully be due to the standard PID settings.

Thanks Cliff for the help!
I'll see if I can change the thread title to mark that this problem is solved.

[OptimaZe]

I think I've found the problem!
I had mistakenly configured the ESCs to 50Hz. I have now set the update rate to 400Hz (mode still PWM). I will try the new settings tomorrow morning.

I know this is an old topic, but I want to thank you for this post. I have been flying quads for several years now, all of which have run either Cleanflight, or Betaflight. I recently picked up an old Eachine Racer 250 for dirt cheap with a CC3D FC in it. I got it to use as an LOS quad to teach my nephews how to fly. I have zero experience with this FC.

I updated it from Openpilot to the latest LibrePilot and got it all set up. However, no matter what I did, I could not get a stable hover. I went through the setup wizard numerous times, soft mounted the FC, balanced props and messed with PID's for a couple days without success. After a few hours of reading through Google search results I came across this post, and low and behold the quad is flying perfectly. Hats off to you for sharing your fix. You made my day. I owe you a beer! 

[robertr]

 I owe lots of people beers haha.