First off, I confess that I am not as technically savvy as any of you.

I am trying to get my 2 axis gimbal to operate correctly.  I am controlling it with a DJI Naza.  I am using the F1 and F2 outputs to control the pitch and roll.  The problem is the servos do not stop once they get going in a particular direction.  If I tilt the copter sideways, the gimbal will start rotating to correct for the motion, but rather than just correct, it continues to rotate until I tilt the copter in the opposite direction. same is true for the pitch.  Also, I have the manual pitch adjustment controlled by my VR dial on my Futaba 8J radio.  Rather than just move the gimbal pitch incrementally based on the dial motion, it initiates motion in the direction of the dial.  If I move the dial slightly, the gimbal goes in that direction, but rather than move a bit and stop, it just keeps going until i turn the dial in the opposite direction.

I have tried different end point and gain values in the setup software with no success.  I am wondering if my problem lies in too much voltage going to the gimbal servos.  I have an 11.1v lipo feeding the Naza, and I pulled up the spec sheet on the servos and their operating range is 4.8-6v.  Is it possible that I am sending too much voltage to the servos, or is it fair to assume that the NAZA controller automatically drops the voltage down on the F1 and F2 outputs?

Any help or links to help would be greatly appreciated.

Views: 1607

Reply to This

Replies to This Discussion


It sounds like the flight controller is expecting standard fixed rotation servos and is outputting a constant value for the set point. Most gimbals are configured to operate this way. In other words, if you rotate the copter to the right, the flight controller is telling the gimbal to rotate left at, lets say, and arbitrary value of 95. Now as long as you are holding the copter to the right, that 95 value is getting sent to the servo. On standard servos this is not a problem because 95 corresponds with an angle. On continuous rotation servos, 95 is the speed at which the servo rotates. So you can see, on your setup that 95 value will just cause the servo to rotate and never stop until the copter is level again at which point the flight controller outputs 0 again.

I hope this helps you understand the problem. I do not have a solution for you because I do not have much hands on time with the Naza. They have a configuration to account for the different kind of servo you have. What you would want is the flight controller to output a value that is related to the rate of rotation instead of the angle. 



Adam, thanks.  You have given me something to research and helped me to understand the process.  Can you tell from these specs if I have standard fixed rotation servos?  They were the ones recommended by the gimbal supplier.

Model: GS-D9257 360°
Size: 35.5mm x 15.0 x 28.6mm
Weight: 25g
Torque (4.8v): 3.8 kg-cm
Torque 6v): 4.2 kg-cm
Speed (4.8v): 0.08sec/60 °
Speed (6v): 0.07sec/60 °
Operating voltage: 4.8v-6v

D9257 features
* 1250μs broadband

* 760μs narrowband
* Coreless Motor
* D, dual ball bearing
* Torque, high-speed, high-resolution
* Material with international brands.

It looks like I am trying to use an open loop servo in a closed loop situation.  Is there any way to reprogram a servo from open loop to closed loop? 


I think you are better off just getting new servos, its an easy fix to convert closed loop servos to open loop. But not as easy the other way around.

The way you usually hack a servo from a normal closed loop to a open loop servo is to replace the potmeter in the servo with two resistors of equal value. To go the other way around you would have to find  a potmeter and replace the two resistors with it. And then you would need to find a reference point to fix the potmeter to. This is usually the servo outputshaft, or on larger gimbals en external reference for instance to the roll or pitch axis on your gimbal.

Thanks Anders.  I think I am starting to get on the right track. I opened up one of my 360 degree open loop servos and found that the potentiometer is actually still in there.  They just cut off the plastic piece that attaches to the gears.  That will make it easier for me to just solder in the leads for an external pot attached to the pitch and roll axis.  I may have to do some calculating when attaching and calibrating a 10 turn pot to the roll axis, since it is one of those designs that mounts the camera in the center of a wheel.  Thanks for the input.

Reply to Discussion


© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service