Increase hovering stability - What can I do?

[janis_blank]

Hello,

I just built my first DIY quadcopter using an F500 frame. It uses a (vibration dampened) Pixhawk 2.4.8 controller and a normal m8n GPS. Maybe I had false assumptions, but I'm shocked at the high instability when hovering at a stationary point. It sways around like it's drunk, sometimes more than 2 metres in every direction.
So far I heard of 2 ways to remedy that:

• Using an RTK system
• Using an optical flow module

Does anybody have any experience with that.

Best regards

 

[Bannerman]

I am an electronic technician, but don 't have specific knowledge of drone GPS systems. However, hardware & sofware issues are possible causes. Here are some quick guesses:
1) there may be a delay in the system reporting the GPS position to the control system,
2) the sofware may be slow in processing the data.
Both of these might result in slow system response, causing the observed oscillation.
3) GPS signal may be weak in that location, try a place away from trees & buildings.
4) There may be some some configuration setting that will provide the correct damping function

 

[janis_blank]

I am an electronic technician, but don 't have specific knowledge of drone GPS systems. However, hardware & sofware issues are possible causes. Here are some quick guesses:
1) there may be a delay in the system reporting the GPS position to the control system,
2) the sofware may be slow in processing the data.
Both of these might result in slow system response, causing the observed oscillation.
3) GPS signal may be weak in that location, try a place away from trees & buildings.
4) There may be some some configuration setting that will provide the correct damping function

Thanks for answering,
but I don't think that much can be gained from tweaking the current GPS, as the mentioned 2 metres are well within the expected error range of these devices. Also, the test took place on a wide open field and the software (Ardupilot) doesn't allow for much other configuration that plugging the GPS in (at least to my knowledge).

 

[Dugdog47]

I don't know much about ardupilot but I thought that firmware had a self tuning mode you could put on a switch?

 

[Bannerman]

Thanks for answering,
but I don't think that much can be gained from tweaking the current GPS, as the mentioned 2 metres are well within the expected error range of these devices. Also, the test took place on a wide open field and the software (Ardupilot) doesn't allow for much other configuration that plugging the GPS in (at least to my knowledge).

My little Eachine E58 (no GPS), when properly gyro calibrated and trimmed, will hover almost perfectly still. I suspect there is a function within your system (possibly a basic accelerometer sensor device) that will enable the normal basic stability you want, when correctly set. Does the manufacturer of your kit have a support forum ?

Another thought: how does the drone behave if you switch off the GPS ?

 

[janis_blank]

Sorry, I was on holiday.
I just wrote a script that autonomously starts the drone, makes it hover for a few seconds in 3 metres height in guided mode and then lands it (the drone has a Raspberry Pi microcomputer onboard). This completely eliminated human interaction via radio. The hovering was very stable this time, so the error might be with the remote not being configured correctly or with me not operating the remote correctly and involuntarily causing the movement.
Since the drone is supposed to fly autonomously most of the time anyway, I'm fine with these results for now. I need to test it further to see if the stability persists in different situations. However, It may take some time until I can start testing again as the project is currently on hold.