Having got to the point where I have engineered a multi-rotor flying camera mount that works to a very high level of stabilization (sub .05 deg.) I need to complete the puzzle with some help in designing some boards to compliment the APM system. Then the community at large can buy off the shelf, here at DIYDrones, the right components to do the job properly.

I am not an EE so I probably don't know what I am talking about but I do know the aerial camera mount business inside out. As I see it, the task is this. Camera mounts need to have a very different approach to that of stabilizing a multi-rotor copter. It is a lot simpler for a start. 300/500 deg/sec gyros typically output 2mv./ deg/sec but Invensense for example make gyros specifically for the job of stabilizing camera IS lenses and camera platforms the IDG 1150 for example is a 20 deg/sec. gyro that can output 50 mv/deg/sec. other gyros are available that have dual output that amplify the signal by a factor of 4, but as I see it this only amplifies the critical area which is at the very initial point of movement - noise. Sub .00? is normally only achievable with mil. spec. fiber optic gyros which are way beyond affordability in our case, but as Mems systems have come along in leaps and bounds recently. I believe we can do this to create top professional results at very little cost. Such high sensitivity modules would only work in a vibration free environment and this problem has been addressed and a complete solution is now available. Along the way a multi-rotor craft has been developed that not only is aerodynamically better than conventional designs but it uses materials that float on water and give better crash resistance and weight. Such a craft can be made using cheap materials that are freely available and can be fabricated with simple tools.

The 5/6 axis system explained. We all know that it is only possible to move a camera thro 3 axis of movement - pan, tilt and roll. the typical high end gyroball device uses the inner axis principle to fine tune the high accuracy movements.

Our model can create that outer axis and with a just little more work can be made to integrate perfectly with an inner axis system. The inner axis can be a simple 2 axis module device or a more complex 3 axis. The 2 axis module would be attached to the camera base plate and would have pwm in /out and utilize 2 axis acc. to reference gravity. The 3rd axis again a stand alone module would be in the Z axis (Pan) and integrate with the FC's mag.

Once this has been achieved we can step into the really exiting stage which is to create a total system where the operator simply fly's the camera and the copter follows the camera's desired track. it will use optic flow technology to look at the cameras LCD and augment stabilization and create an auto object tracker. WOW.


My prototypes are made using components such as heading lock gyros that have been modified to accept high sensitivity gyros but these are created as simple test bed components and need a lot of sorting out  before they could be used commercially.

I have only just started with APM so do excuse my ignorance if I made false assumptions at any time.

Most of my model development was done with a simple KK FC so as to create a reliable benchmark. I later moved onto the MK system which did not cut the mustard. It became known as the MK fireworks kit as ESC's and power distribution boards would burst into flames for no apparent reason and then there was the death roll gyro problem where hundreds of duff FC boards were sold to the unsuspecting masses. It would try to transit cumula granite resulting in the spectacular drop kick bang smash. They were the days! Onwards and upwards as they say. Interested anybody?

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones


  • It uses two or three 4100 20c lipos. max duration is 12 mins. I did not check the weight recently as I keep changing camera set-ups.and stripping out anything on the camera this is not needed. being made from foam and carbon it is a very light structure. Balsa wood is also used throghout.
  • T3


    What kind of duration are you getting?



  • Developer

    Best videos I have seen! Super smooth. Most excellent engineering work and editing!

    I have  MultiPilot v2 w ST32 is on its way from Italy.

    I hope to get familiar with vision systems and multi core development.

    All up weight with large camera?

  • Really nice stable footage Denny which makes a pleasant change. What size battery and how long

     can you stay up?

  • http://vimeo.com/21241710


    Yes it floats

  • Steve, there are a lot of beautifully made copters and camera mounts out there. I don't want to upset anyone with even a hint of criticism. However what I have done is a total breakaway from conventional thinking. My system just would not work outside of the custom airframe. However I do expect to have some electronics that will help your systems work to a very high standard. I can't say at this stage when that will be, but things are starting to move along at last.
  • Part two is not up yet as I have to clear this with my client as they now own it.
  • Part one clips were shot with many different platforms. Close in with a heavy lift Okto plus a Hughes 500 and a Hughes 300 plus a Cessna 150 strut mounted and a Ramphos amphib. microlight. The opening shot was done with it's time lapse feature from a boat. It is the technology demonstrator that is used to create the new flying camera mount. You can see now why I made it to float. almost everything I shoot is over water.

    I will be selling the DR-7's after I get the FCM-1 project finalised and into production as a kit.. 

  • Moderator
    Thats very impressive, all of that was from your Octo? (part 2 is not up yet)
  • The video looks really good.

    Do you plan to offer this as a standalone mount? Or the mount will only be offered integrated with the frame? I have a really nice all carbon quad that I am looking for a good mount for. I probably wont lift a 5D but I want to get really stable video.
This reply was deleted.