Jump to content


Multirotor tuning


  • Please log in to reply
7 replies to this topic

#1 momentsky

momentsky

    Key Member

  • Members
  • PipPipPip
  • 50 posts
  • LocationSaint-Petersburg
  • Country: flag of Russian Federation Russian Federation

Posted 02 February 2012 - 11:40 AM

Now I'am tuning the 4 rotor X-config aircraft.
This frame flied well powered with korean board. So I know she can :)

Does it need to be trimed on the transmitter, or it can be done in gcs? Or it can be combined?

As I see the acceleration graph a notice the huge vibration influence to the sensor. Though the main display show the level close to normal.

Where i can find the issues concerned vibration influence to sensors and what values of accel data (vibration oscilation ammount of axises) are ok and wich are not?

Just for now on I am studing the stabilisation tuning doc. Since it does not want to fly well with the default attitude stabilisation settings.

Here is GFX mesured on the ground with props close before takeoff.

Posted Image

#2 D-OCHO

D-OCHO

    Advanced Member

  • Members
  • PipPipPip
  • 286 posts
  • LocationEurope
  • Country: flag of Netherlands Netherlands


Posted 02 February 2012 - 12:07 PM

Wrong info

Edited by D-OCHO, 02 February 2012 - 12:08 PM.

Open Pilot CC Quad (35 Mhz), Open Pilot CC Minsoo Mini Quad (2,4 Ghz), Blade MCP X,Trex-250, KDS450S + Co-pilot II, Trex-600E + Flymentor,  Spektrum DX7, Futaba T9CP Super, Phoenix RC Simulator, Traxxas 1/16 E-Revo VXL (TQ 2,4 Ghz).

#3 D-Lite

D-Lite

    Core Team

  • Members
  • PipPipPip
  • 968 posts
  • Country: flag of Germany Germany


Posted 02 February 2012 - 01:24 PM

View Postmomentsky, on 02 February 2012 - 11:40 AM, said:

As I see the acceleration graph a notice the huge vibration influence to the sensor. Though the main display show the level close to normal.

This looks okay for me. Full throttle vibrations will be much higher and there is a limit where it starts to influence the attitude estimation. The only thing you can do about this is reduce the vibrations or mount the board with vibration isolation mounts. But for the moment, it should be okay as it is.


View Postmomentsky, on 02 February 2012 - 11:40 AM, said:

Just for now on I am studing the stabilisation tuning doc. Since it does not want to fly well with the default attitude stabilisation settings.

That's exactly the right thing to do. A little tuning and it should fly nicely. It's all done in the GCS stabilization settings. If it behaves too sluggish, carefully increasing the Kp factor for rate and attitude may already help. But best is to follow the tuning doc.

#4 momentsky

momentsky

    Key Member

  • Members
  • PipPipPip
  • 50 posts
  • LocationSaint-Petersburg
  • Country: flag of Russian Federation Russian Federation

Posted 02 February 2012 - 01:47 PM

View PostD-Lite, on 02 February 2012 - 01:24 PM, said:

This looks okay for me. Full throttle vibrations will be much higher and there is a limit where it starts to influence the attitude estimation. The only thing you can do about this is reduce the vibrations or mount the board with vibration isolation mounts. But for the moment, it should be okay as it is.
Ok, nice news! :)
Shure it should be some job done for the vibration reduction as you said. That's right. But as for now i want it to go on first.

View PostD-Lite, on 02 February 2012 - 01:24 PM, said:

That's exactly the right thing to do. A little tuning and it should fly nicely. It's all done in the GCS stabilization settings. If it behaves too sluggish, carefully increasing the Kp factor for rate and attitude may already help. But best is to follow the tuning doc.
Yeah. On the way.

#5 momentsky

momentsky

    Key Member

  • Members
  • PipPipPip
  • 50 posts
  • LocationSaint-Petersburg
  • Country: flag of Russian Federation Russian Federation

Posted 02 February 2012 - 01:49 PM

And what is about trimming?
Should it be trimmed in the TX or it should be trimmed in the GCS?
Or it should be self-corrected via stabilization?

#6 D-Lite

D-Lite

    Core Team

  • Members
  • PipPipPip
  • 968 posts
  • Country: flag of Germany Germany


Posted 02 February 2012 - 03:30 PM

View Postmomentsky, on 02 February 2012 - 01:49 PM, said:

And what is about trimming?
Should it be trimmed in the TX or it should be trimmed in the GCS?
Or it should be self-corrected via stabilization?

If you are in attitude mode and the board is perfectly leveled, it should self correct. But if the board isn't mounted 100% flat or something else with the frame's mechanic isn't perfect, the board thinks is leveled while the frame is slowly drifting away to one side. You can correct this via the TX trimming or by making small adjustments in the GCS to Settings/AttitudeSettings/AccelBias. It's a bit a matter of taste but I think most people would say the second one is the cleaner solution.

#7 A*Morale

A*Morale

    Core Developer

  • Members
  • PipPipPip
  • 194 posts
  • LocationRome/Italy
  • Country: flag of Italy Italy


Posted 02 February 2012 - 04:22 PM

In the rare case that you have too much vibrations that influences bad the attitude estimation, you can try to raise a little the GyroTau parameter, as shown here:
http://wiki.openpilo...ilization Panel
I did this on one quad that actually has four motor with wear out bearings.
This is expecially useful for a quad that is going to be used mostly for hovering.
Alessio my Blog
Posted Image

#8 momentsky

momentsky

    Key Member

  • Members
  • PipPipPip
  • 50 posts
  • LocationSaint-Petersburg
  • Country: flag of Russian Federation Russian Federation

Posted 04 February 2012 - 08:00 PM

View PostA*Morale, on 02 February 2012 - 04:22 PM, said:

In the rare case that you have too much vibrations that influences bad the attitude estimation, you can try to raise a little the GyroTau parameter
And what is your exact value?