I am refining an externally-mounted camera system for light aircraft that is attached to the wing strut, and would like to stabilise the camera in one axis so that the roll of the plane doesn't affect the angle of view of the oblique photograph.

There's a Nikon D5600 DSLR in the housing, currently fixed to the metal framework - gets good results, but as the aircraft rolls in flight the angle of view changes (it's nominally 45°). I would like to stabilise in one axis only (i.e. perpendicular to the direction of flight) - researching gimbals shows a lot of options for 3-axis stabilisation but not single-axis, and wondering what the options might be?

Can 3-axis gimbals be set to only work in one axis?

  • Would you be willing to DIY or are you looking for an off the shelf part? – Jack Soldano Sep 3 '20 at 8:39
  • @JackSoldano as you may notice DIY is a theme of this project! Would be interested to build something. – Simbamangu Sep 3 '20 at 8:52
  • Excellent, thought that might be the case, I've added a rough answer to the question although if you want to go this route this might be a better question for the arduino.stackexchange. – Jack Soldano Sep 3 '20 at 9:18

I would recommend looking at single axis solar tracker designs for the mechanical design.

These will be some form of motor connected to a plate (Where the camera could mount) that get close to 180° rotation of that single axis.

You will then want an Arduino with a gyroscope sensor like the MPU-6050 to detect the current orientation of the plane. The Arduino would then adjust the single axis motor to counteract this rotation, you may need a way of identifying the current location of the rotating axis either using an encoder or something in the mechanical design.

Some References:



  • The second link shows a good example - though the product there isn't very smooth, one can probably optimise for a better movement? – Simbamangu Sep 3 '20 at 11:31
  • You could probably use a hobby servo such as those found here. These are a good match for an Arduino and can have a range of ~180 degrees or more. – BobT Sep 3 '20 at 15:25
  • @Simbamangu yes that example is very basic and uses fairly low-quality components. If this were built with machined parts & higher torque servos the stability could be much better. BobT linked parts look pretty good and would likely do the job. The 'Smoothness' would also depend on how well the code is written, you could apply lots of averaging for example to the axis sensor data to 'smooth' the movement. This sort of system would require a fair amount of tuning. – Jack Soldano Sep 3 '20 at 15:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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