What could be the reason of different servo output for neutral position?

Hi,

I have set up my quad apm 2.8. Calibrated accels and reciever. When I arm and put every stick into middle position. I get the servo outputs as given in the image. As you can see everyting is in neutral position. What could be the reason for this different outputs as we expect all be the same?


What type of transmitter are you using?
Could your trims be out of center?

The trims are all centered. I am using Turnigy 9x transmitter with dsss and Turnigy 9X 2.4GHz 8Ch Receiver (V2). As I check the radio pwms on mission planner they are all centered, too. So seems no problem with the rx. Everything on rx side is normal. All calibrations are normal also.

What could be the problem other then trims?

I really new to MissionPlanner and ArduPilot myself but here’s another guess.
The list you see is the servo “Output”. My guess is it’s the PWM signal to each ESC. Since you’re quad is armed, it’s trying to drive the ESCs to level the quad. I’d image these values would fluctuate as you move the flight controller.

I hope someone corrects me if I’m wrong.

Yes you are right. FC should always be trying to drive ESCs to level the quad. In a case like mine in which the quad is already levelled, pwm outputs to ESCs should be single value. In a propeller mounted case, my quad leans on a direction when I throttle up even it is leveled. I am trying to learn what could be the problem of getting this outputs not same for all ESCs when quad levelled.

It works quite meaningfull when I move FC, the servo outputs fluctuates as it should be. Such, when I roll right it drives the right motors more etc. But, in a levelled case there should not be any difference on the speed of the motors. Thats my problem.

thanks.

Sorry, it took me a while to understand the problem.
Thanks for your patience.

Are you sure this is a problem?
Does the 'copter fly level? Is the output you posted displayed while the 'copter is in the air? I’d think the outputs would be unpredictable if the 'copter isn’t allowed to move itself.

You are wellcome.

Yes it is a problem for me since quad leans to a direction on takeoff and will probably lean same on fly. As you might predict the pwm outputs all be the same for levelled position. But thats not in my case.

Yes I pick up the quad and hold it in fly level by my hand. Changing the roll and pitch by my hand fluctuates the outputs in a correct manner. But the output pwms are not correct for the level in the air forced by my hand.

Again this is a problem for me since take off leans on a side and it will probably lean again on the air.

As you can probably tell by now, I’m just guessing at possible solutions. As long as I’m guessing, I’ll add one more.

Is your 'copter balanced? Is the weight distributed evenly? A photo of the 'copter might help someone diagnose the problem. Plus, I just like to see photos of people’s quadcopters.

I saw in an earlier post you mentioned something about not being sure your ESCs were properly calibrated. Did you get this resolved? Is this the same quadcopter?

The earlier post was about another build months ago. I dont remember how did I solve this problem.

For this build ;

----Is your 'copter balanced?----
Yes it is

----Is the weight distributed evenly?
Yes. All the things are centered.

----A photo of the 'copter might help someone diagnose the problem.
I use the card itself without copter only card but still same problem.

The copter will never be 100% level. As soon as you raise the throttle, all stabilization calculations start to run. It is never a good idea to keep the throttle up without actually lifting off. The PID algorithms try to compensate for small errors in attitude without being able to really correct the copters orientation, because it sits on the ground. Especially the I-term of the PID gets bigger and bigger and if you finally decide to lift off, the copter flips over.
If you are sure the motor order and prop rotations are correct (check them again!), the proper procedure is to arm the copter and raise the throttle to about hover throttle, then raise it a bit more until the copter takes off. Go to a height >1m to get out of groundeffect, which otherwise might make the copter unstable.