What are effects of bad compass data, really ?

[lesd]

Engineering question: What might be the effects of compass interference ? For example, flying near a power line ? I am guessing that since the compass data is used to hold YAW angle, the YAW might be wrong, but would this be a cause of total fly away ? Wouldn't the flight controller see that the GPS data is showing position way off, and revert to something more safe, like ignoring the compass data and 'limp' home with inertial and GPS based navigation only ?

 

[RoyVa]

My understanding is the compass calibration sets up North position for the GPS and
after its calibrated your good to go unless you take your quad miles away.. re-calibrate is recommended. Also want to be sure then you do the initial compass calibration that their are no metal interferences which would give you a bad calibration (setting North position) which could lead to quad traveling incorrect GPS coordinates. Any others have any insight on the compass calibration.
Interference front cell towers and other electronic interferences could cause you to lose communication between the RC and quad.

 

[yorlik]

very interesting info. I have never had any idea how compass was used with the gps.

how did you come to this understanding please?

 

[lesd]

yes, please elaborate.

 

[ianwood]

http://www.phantompilots.com/threads/compass-calibration-a-complete-primer.32829/

 

[RoyVa]

very interesting info. I have never had any idea how compass was used with the gps.

how did you come to this understanding please?

It comes from an understanding of copter control functions, I can point you to a site that explains it also. Do a google search on Quadcopter control functions. Interesting stuff. Http://Diydrones.com/forum/topics/Quadcopter-control-functions-layers then. read under orientation. Hope this helps
I.E.
Quadcopter starts to set orientation to face certain direction. It needs a sensor that gives different readings when it rotates horizontally. This feature is achieved by using magnetometer. Magnetometer is simply an electronic compass, it gives readings in X-Y-Z directions, but what we focus here is the Z-Axis assuming quadcopter is horizontal; some math is required to get the absolution magnetic field direction if quadcopter is not horizontal in case we look for high accuracy.
So in essence it tells the copter direction... North mainly.

 

[landmannnn]

The processor combines inputs from the compass, GPS, accelerometer and barometer to calculate position, orientation stability and direction.
Combining the sensors improves accuracy.
If one of these sensors gives a false reading then the phantom will miscalculate either position, orientation, stability or direction.

So no, unlike your car, it won't go into limp mode and get home safely if a sensor is wrong, it is likely to crash instead.

 

[lesd]

Crash is no good for craft owner.
Doesn't have to be this way.
It should be able to do crude but effective navigation with no compass, based solely on Gyro/accelerator readings and the GPS for location.
In effect, doing semi inertial based nav, but because the gyros used are cheap, the drift would be causing it to have a very 'drunken' return to home.
How? You start off in a direction ( based on managing yaw by using accelerometer/gyro ) and see from the GPS if you are going in a relatively correct direction. Corrections made base on the error.
This is similar to the same way you navigate with a GPS when walking, where initially you don't know what direction to walk, but as you start you can see you are in the correct direction or not and turn based on what you see your direction is on the GPS screen.

Doable, but it's easier to crash the toy and make the user buy another one, since it's not a $100 million satellite or something .

 

[ianwood]

Crash is no good for craft owner.
Doesn't have to be this way.
It should be able to do crude but effective navigation with no compass, based solely on Gyro/accelerator readings and the GPS for location.
In effect, doing semi inertial based nav, but because the gyros used are cheap, the drift would be causing it to have a very 'drunken' return to home.

It's an interesting idea but not without challenges. Gyro and accel may help but not much. It would be mostly GPS. Effectively you're using Course Over Ground (COG) as a substitute for heading when the compass produces bad data. The first problem is the FC can only detect really bad compass data. Moderately bad compass data is indistinguishable from good data.

You would need 10+ sats so you can quickly and accurately determine COG. Anything less and you'll have to cover a lot of ground before the correction. Variable or shifting winds will cause problems with using COG to adjust direction of travel.

If the compass got knocked out in light and consistent winds with really solid GPS, it could work.

 

[RoyVa]

While all GPS receivers can tell where you are, entry-level models have no way of ascertaining which direction you are facing. So you have to start moving and walk a few steps at a good pace before it can accurately point you towards your destination. This is why mid-range and high-end units have an electronic compass — to show the direction you should travel even while standing still.
This can be found in any GPS operation information. Compass Needed !!!And so back to the Magnetic Compass.
GPS manufacturers recognise that whilst a differential compass is fine most of the time, users still like the security of a magnetic compass. The traditional magnetic compass has been fitted with electronic sensors and bundled into some of the GPS devices. This is what is meant when a GPS is advertised with an electronic compass. It costs a little more but allows you to take bearings when stationary or out of sight of the satellites.


Hope this helps


Roy