This is main board normally put on Gimbal frame near the Quad . 


This is the two options imu available normally is on camera module: 

2 axis with 3 accelerometer and 3 gyroscope 3 magnetometer in one version we use mpu6050 and in other we use mpu 9150 1 chip 9 DOF sensor.




Some friends start to design different kind of gimbal for our new board. The main difference respect of standard gimbal is that the mechanical frame need to be develop for each video sensor that you would like to use for your video. This frame is for Gopro Hero 2-3 in the design is important to move CG of camera for obtain better result during gimbal stabilization :)


This is the caratheristics of our : VR Gimbal :

Micro Controller is STM32F1 at 72 mhz.

3 Axis Brushless Direct Drive gimbal :

  • Roll    (MOT1)
  • Pitch  (MOT2)
  • Yaw    (MOT3)

For each channel we have 3 high power pwm output .


I2C IMU that support 3 axis . It  is on camera module , so we can know exactly the position of camera and control in realtime the 3 Brushless motors.

We are developing two option :

  • more traditional approach : MPU6050 (6 DOF) + HMC5883 (3 DOF)
  • new sensor from inversense MPU 9150 (9 DOF)


1 USB port.

1 Serial port that support mavlink protocol.

1 Power module that support until 3S battery.

4 Radio RC Input for :

  • Control ROLL Setup.
  • Control Pitch Setup.
  • Control YAW Setup.
  • IR control repeater.

4 Analog Input 0-3.3v

1 IR Out , so is pssible to control a remote camera by its ir receiver.


The first batch of production will be available in 3 week it's better to reserve your board because there is a big request of this gimbal.

for more info send a mail to

original blog post :


Roberto Navoni


E-mail me when people leave their comments –

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

Join diydrones


  • Neat the 9150 more stable?

    Would love to see how well it stabilizes video.

  • Moderator


    could be possible use a serial port with mavlink protocol for share 2 wire connection without add another radio module.

    The bigger use sensor that we yet use in a lot of design so it's a safe and affordable design. The MPU9150 instead is a new component and we don't have enough experience on it to choose to use its magnetometer instead of external hmc5883 that work fine in our platform.

  • Moderator

    @Oliver ,

    i agree that there are a lot of development on this kind of platform i prefer to don't clone an arduino boar but i would start a new design based on a powerfull STM32 bit micro controller . That kind of micro normally used for motor control in industry market , so is possible to implement a lot of tips for a good control of brushless gimbal :)

    We are working on a first release of firmware that will be available with the first batch of eletronics :)

    I know that the 3 axis AP library control developed for servo , but could be upgrade to brushless gimbal , too :)


  • Hi Roberto, This is great!! I was looking for a 3 axis gimbal. 

    I have some suggestions. The whole idea of Ardupilot is to make everything as much simple is possible, so why don't share the I2c bus with ardupilot to command the gimbal? Instead of 4 cables with PWM signal, we are going to connect just one 3 wire cable, like the PPM sum option. Is for this reason you have a com port shared with the mavlink?

    Also what is the difference between the 2 different IMU's, other than having all in one package on the MPU9150?

  • Oooh I was just about to click go on a iflight one.

    Robert I will instead wait for yours. Could you put me down for one please?

    Many thanks

  • AlexMos is planning a 3rd axis add-on board ->

    Anyone know how will the GUI/config utility of the VR Gimbal board will compare to the latest from AlexMos..

    Software seems to be the biggest difference at the moment between the AlexMos' controller and the open source Martinez one. There seems to be very limited documentation or development happening with the Martinez variety, even though RCTimer, iFlight and many others now have cloned versions of their own.

    Side note: looks like iFlight just released a series of brushless gimbals too ->

    Steve: Arducopter's current gimbal stabilization routine also does not let you configure PIDs or enable rate based control. It only has most basic of options to set angle and output range. Flight controllers like Wookong, OpenPilot, and Hoverfly would be a better better comparison for servo controllers  And dedicated gimbal control boards like Radian, Tarot ZYX, Skyline, Hoverfly's and Smoooth should outperform most flight controllers. APM never really made it to the party.

  • Distributor

    subscribed! :) (PM me for distributor details... just like the rest!) 

  • I just bump into this thread.. been dying to ask if somebody would add source dode for brushless gimbal. can't we just add codes without any new added gadjet? not possible isn't it?

  • Moderator

    @Steve , the technology that we are developing is similar to DJI Zenmuse , but will be opensource :) And less expensive 

    The main difference is that on APM or on VRBRAIN , too normally we use servo for move gimbal ... in Brushless gimbal we use special brushless motors for more the gimbal , the result is incredible smooth.



  • Moderator


    yes it's very nice simple but work very  nice ... so the paradigma with brushless gimbal change a lot need to have a perfect mecanich that is around your camera .... and then move it with brushless motors and use imu as virtual potentiometer. I had doing a version of code that use analog output and standard servo but it don't work so good as a brushless gimbal.

This reply was deleted.