Pixhawk Hardware Limitations


I wanted to ask about the limits of the accelerometer and gyroscope in the pixhawk. I wish to use the pixhawk as a flight controller for a turbine powered plane capable of speeds upwards of 100m/secs.

I know this is a bit much for the poor pixhawk but I am curious about the hardware limitations of the device.

If there are other more robust sensors which can be attached to the pixhawk to make it capable of supporting a vehicle at that speed, that would be great as well.

Thank you,


The accelerometer measures accelerations, and the gyro measures angular velocities. If your plane does not exceed the acceleration range or angular velocity range of the sensors, then it should work just fine. Of course the dynamics of such a plane are probably vastly different so the fixedwing controller might need some changes or at least re-tuning. Also the state estimation (e.g. EKF2) and the GPS model chosen needs to be looked at.

Most use a UbloxM8N, but i would go for one with a higher update rate.

Max navigation update rate NEO-M8M/Q in GPS & GLONASS mode = 10 Hz
Dynamics  4 g
Velocity 500 m/s

Think about update rate! At a 100m/s and 5hz gps (that’s common) each update is ging to be 20m away…

acc&gyro should be no problem, they don’t know the speed and can handle 16g normally. Turbines don’t vibrate, so they will be happy.

Thank you for the reply.

So I have to use a GPS with a higher frequency, somewhere around 50hz. But is there a module available which is compatible with pixhawk? Also won’t using a GPS with a higher frequency make the results less accurate?


You don’t need GPS updates this fast because you can predict the motion of the aircraft using the IMU (mag, gyro, accel) quite well for short time horizon. Also, I don’t know if it’s possible to have such fast update rates given the signals that are received from the satellites.

Assuming the G’s a model-jet can do are similar, and might be higher, than a real aerobatics 1seater aircraft. I think that px4 with the right settings won’t make a mess of the situation, but i don’t know how precise you want to be. (nor how precise it will be)

A paper explaning some of the troubles that come with gps, speed and G’s: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5461065/
This compares their specialised (mems)ins/gnss system (hardware similar but might be better than a pixhawk+ublox ) to “gps + a fiber optical gyro” and “ring laser gyro”.