Compass interferences

I have an APM2.5 with internal compass, my compass motor interference is a little high 30-40% I calibrate compass, etc. but always the copter turns 20-30 degrees rigth when I take off and stays there, can I resolve changing COMPASS- DEC? my value is -0.05788675 any suggestion?

Another suggestion for compass motor interference? (not possibility to change  gps in  short time)

 

SUMMARY DISCUSSION: (for newbies)

I'm  trying to summarize the most important things from this post for newbies here, if someone notice any mistake or something usefull to write here too, please let me known (Orthographic mistakes are welcome too ;)  if someone wants to re-write better, I change it, english isn't my native language)

WARNING!!! : BE CAREFULL WITH YOUR COMPASS INTERFERENCES!!!, you can crash your copter if you don't paid attention to them.

 

REMEMBER: If your compass read is inconsistent with your GPS your copter LAND by default, you can switch to stab or Alt Hold if you like to abort LAND http://copter.ardupilot.com/wiki/configuration/ekf-inav-failsafe/

 

BE CAREFULL: Not use magnetized screwdrivers or bolt in your ship. (better plastic or aluminium)

                              Prefer to calibrate compass outside far from metal things

                              Avoid fly compass dependent modes near buildings or other metal things.

                              Check your compass before calibrating that have correct orientation on MPlanner map

                              Do the compass mot 

 

READ INSTRUCTIONS: http://copter.ardupilot.com/wiki/ac_compasssetupupadvanced/

 

SOLVE YOUR INTERFERENCES: you can buy an external compass and set on a mast far from interferences (like this http://store.3drobotics.com/products/hmc5883l-triple-axis-magnetometer )

 

OR FOLLOW THIS SUGGESTIONS:

Forrest:

Until you can get an external compass, try different orientation/location of the:

- battery 

- creating more distance between the APM and wires between the battery-ESC

- creating more distance between the APM and power distribution board or power module

- creating more distance between the APM and ESC (2 cm can make a big difference)

You can test the distance theory by putting the APM up high away from the electronics.  In your experiments, the ship does not have to take off so the changes can be temporary and not fly worthy.  You just need to know if changing the orientation or location of something with an electrical field stops the initial rotation as you power up. Take the props off and just watch MP (does the heading change as you power up).

ColinC suggestions:

 buy a cheap hand-held compass.  I got mine at the local Scout shop for under $5.  You can learn a lot just by passing it around your rig.  You can also learn a lot about your room just by walking around.

The compass and the level are the two basic tools.  Oldies but goodies. Cheap and very low current draw.

And look's ColinC interesting test here:  http://diydrones.com/forum/topics/compass-interferences?groupUrl=arducopterusergroup&groupId=705844%3AGroup%3A394475&id=705844%3ATopic%3A1907171&page=5#comments

 

Enjoy your fligth :)                                         

 

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

Join diydrones

Email me when people reply –

Replies

  • "in fact Einstein was right : everything is relative. Our problems come from basing our crafts attitudes and corrections on absolute measurements that are drifting and unreliable."

    So true. Also from a pervasive assumption that our problems would all go away if we could just get a little more accuracy or precision. We really need to develop a Science of Close Enough.

    Cheers,
    Colin
  • Thank's Colin great explanation,  I'm not going to calibrate anymore in my office, near PC, metal roof, etc., I read many crashes here flying near buildings with compass problems, it's a high risk to fly automatic modes in those places, bad GPS and bad compass, explosive combination :o

    • MR60

      RTK GPS systems would be a good solution as it uses a reference GPS on the ground to provide centimeters relative positionning accuracy to the "flying GPS" on the craft.

      I would add that the same should apply for baro : one reference baro on the ground too to correct the baro drift of the flying baro.

      So my hardware wishes for the next releases are:

      -A telemetry module on the ground integrating a reference baro and a reference GPS.

    • MR60

      in fact Einstein was right : everything is relative. Our problems come from basing our crafts attitudes and corrections on absolute measurements that are drifting and unreliable.

  • :O, The roof house  is metal, I flyed aprox 45 degrees different from one place to the other.

    Thank's for the laywoman explanation (I learn another term too ;) ) I think I found my frame problem as you diagnosed  some weeks ago :O; I have to do some other test to confirm.

    In the mag log are three values, This one, other is offsets that give an idea of metal interferences and corrections inside the ship, correct? and the last one what it's means in layman terms ?

    • HI. For reference here is a magnetometer rotated in normal space:3702554840?profile=originalHere, next to a radiator:

      3702555053?profile=original

      And here, next to a speaker:

      3702555066?profile=original

       

      Radiator:

      3702554939?profile=original

      Speaker:

      3702555110?profile=original

    • I should just mention that the mag is at the edge of the turntable, so the distance changes as it rotates.

       

      3702689944?profile=original

    • Developer

      yeah, great analysis!

    • MR60

      so you got those levels of change by simply rotating the stool.  Wow.  no more calibrating around my desk.

      thanks for the lesson Colin.

  • Look's what happens today, I was testing another change on my cuad, I fly in front of the house then I move to one side about 30 mts to an open field and looks mag log

    3701959574?profile=original 

This reply was deleted.