I have short question about the alexmos gimbal controller. The Controller receives the gyroscope and accelerometer sensor data from the IMU, that is mounted on the camera. In optimal case, the camera should stay in perfect position, which means, that there would be no gyroscope data, since there is no movement. So the only data for positioning would be the accelerometer. Is there a second IMU onboard that receives the gyrometer data?

All the gimbal controllers from aliexpress seems not to have an IMU onboard, but it that case the controller can only use the accelerometer, right?

  • $\begingroup$ What makes you think that being in some "perfect" position means that there is no data from the sensor? For how long do you think the system maintains this state? I'm not sure I understand this question: you ask about controlling the state of a system and try to point out that reaching the desired state is apparently problematic. I guess the opposite is true: you use the controller to reach exactly the desired state. $\endgroup$ Nov 5, 2016 at 21:57
  • 1
    $\begingroup$ I mean in stationary position, the velocity remains near to zero in the camera frame. So we have only small, noisy gyro data, which seems to be useless and we can only use the accelerometer, which is only for positioning in the long run and not for fast angular deviations. $\endgroup$ Nov 6, 2016 at 7:03

2 Answers 2


The alexmos and other similar gimbal controllers, are using the camera mounted accelerometer and gyroscope in order to identify the angle and angular rate (and possibly higher order dynamics) of the camera platform, in an attempt to keep the angle the same as the set point. You are right that if the platform was doing it's job perfectly that the sensors would "see" no change, but this is not the case. The platform moves constantly but the sensors are able to sense very small movements very quickly much smaller than could be picked up by the cameras themselves and correct before they become errors that could be see in the video. Some controllers take input from the flight controller or gyros onboard the craft in order to predict or lead movements, in order to make the response to stimuli faster, and enable proactive correction as opposed to just reactive movement.


I am not familiar with the gimbal controller you are using, but you cannot stabilize a gimbal with only an accelerometer. You must be able to measure the states or change in states of the gimbal for stabilization. Can you please provide a link to the controller you have the question about.

  • $\begingroup$ If we look for example at this one here: de.aliexpress.com/item/… There seems to be only one sensor attached to the camera. So the gyrometer can only measure the deviation from the stationary position, and that should be really small for stability and also the data is pretty much noisy for small gyro values. I think it would make more sence to attach the sensor to the base, so the gyrometer can measure the disturbance. $\endgroup$ Nov 6, 2016 at 6:55
  • $\begingroup$ I think there is also an accelerometer onboard, right? Its that smaller TSSOP chip invensense? $\endgroup$ Nov 6, 2016 at 8:02
  • $\begingroup$ The link is not in English. Also, I don't understand the point you are trying to make about the gyro measurements being small. Even if they are small, they still contribute to stabilizing the gimbal. I think having the sensor attached to the camera makes the most sense because you want to stabilize the camera. $\endgroup$
    – Ralff
    Nov 9, 2016 at 1:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.