DJI Mavic, Air and Mini Drones
Friendly, Helpful & Knowledgeable Community
Join Us Now

GPS and compass

Ralph thompson

Well-Known Member
Joined
Feb 26, 2019
Messages
793
Reactions
577
Location
Montreal, Quebec, Canada
A question to the technical experts please, for the Mavic 2 series, are the GPS and compass totally independent of each other? I've read that the compass sensors are mounted on the same board as the GPS (not that makes them connected electronically in any way). Flying the M2P in an area where there is no GPS signal, the compass does seem to read correctly.
So in a situation where there is no GPS but with the compass is pointing in the right direction and no command to "calibrate compass", plus IMU normal, then it should be totally safe to fly out of a GPS shaded area into a region where the M2P picks up sufficient GPS to fly normally in GPS mode? And then to fly back into the GPS shaded area. Lets say the visual and ultrasonic sensors are turned off so it starts and ends in ATTI mode (obviously being very attentive to wind, drafts and drift). I'm aware also that the GPS signals can be compromised causing the drone to get confused about its location, how bigger risk is this?
Sorry, bit of a mixed up question but just trying to better understand the inner workings.
 
are the GPS and compass totally independent of each other?
Yes .. they are independent, one provides directional data, the other provides location data.
They are two different devices.
then it should be totally safe to fly out of a GPS shaded area into a region where the M2P picks up sufficient GPS to fly normally in GPS mode? And then to fly back into the GPS shaded area.
Yes, you can fly in/out of a GPS denied area, just be cautious if close to obstacles.

 
Yes .. they are independent, one provides directional data, the other provides location data.
They are two different devices.

Yes, you can fly in/out of a GPS denied area, just be cautious if close to obstacles.
Speaking of the compass. Why is it's calibration so important or maybe not important. When asked, I never calibrate it anymore, unless the drone simply won't allow me to take off unless I do. What else does the compass do besides show the orientation of the drone on the map window?
 
Tks META4. I can partly answer this cgmaxed. The GPS, compass and IMU outputs are all combined in the control processor & depending on deviations from commanded position & the position the drone thinks it at, the controller speeds up or slows down one or all of the propellors. If one of the inputs is in error or inconsistant then controller can get confused. Let’s say yr compass indicates an incorrect heading but the GPS & IMU are indicating movement in a different direction. The consequence can leave the drone beyond your control.
I have had compass errors in flight, along with poor GPS & strong winds. Probably because I calibrated the compass near a heavy iron cover. It’s not fun trying to control a drone without these crucial control inputs. I take compass, IMU calibration seriously and check under Advanced<Sensors before every flight to confirm the state.
 
Speaking of the compass. Why is it's calibration so important or maybe not important. When asked, I never calibrate it anymore, unless the drone simply won't allow me to take off unless I do. What else does the compass do besides show the orientation of the drone on the map window?
The calibration process subtracts the magnetic field of the aircraft so that the compass accurately measures just the external (presumed to be the earth's) magnetic field. If it is out of calibration then the compass heading will be wrong. For small errors the most obvious indication is that it will tend to crab rather than fly straight forwards. Larger errors may result in uncontrolled flight.

Note that calibration is absolutely not correcting for declination or for external magnetic fields that locally distort the earth's magnetic field.

For more details:

 
Thank you sar104, that is really interesting. It fits with my experiences and observations and explains several uncertainties in my mind. I'm now much more comfortable knowing about these simple preflight checks to confirm compass status.
 
  • Like
Reactions: sar104
It's also important to consider the magnetic conditions at the take-off point. Taking off in a magnetic field then flying out of that field leaves the drone controller unable to maintain stable flight. Happened to me once, and with no control inputs the drone started skating sideways towards a tree. The only way to save the drone was to visually watch the drone while experimenting with the controls to see which combination brought the craft back to safety.
 
It's also important to consider the magnetic conditions at the take-off point. Taking off in a magnetic field then flying out of that field leaves the drone controller unable to maintain stable flight. Happened to me once, and with no control inputs the drone started skating sideways towards a tree. The only way to save the drone was to visually watch the drone while experimenting with the controls to see which combination brought the craft back to safety.
You were very lucky and had a very mild yaw error.
Most end within seconds in a high speed crash.
 
... and well prepared. Extensive practice in ATTI mode for just such an occurrence gave me the ability to react quickly to a hazardous situation and to save the drone. If you complete the training for a Part 107 license and take that stuff seriously, you WILL be a better, safer pilot.
Most yaw error situations are uncontrollable and over before the most experienced pilot could react.
 
It's also important to consider the magnetic conditions at the take-off point. Taking off in a magnetic field then flying out of that field leaves the drone controller unable to maintain stable flight. Happened to me once, and with no control inputs the drone started skating sideways towards a tree. The only way to save the drone was to visually watch the drone while experimenting with the controls to see which combination brought the craft back to safety.
Just to be clear on this, the control failure doesn't occur when leaving the area of magnetic disturbance, even though that may be when a compass error is reported. Control failure will occur whenever the IMU heading is incorrect, which will start inside the interfering magnetic field.
 
Not sure I understand this sar104. I assume when the drone boots up, the IMU has no info on heading. Presumably the compass provides that data. The IMU just integrates angular acceleration and it’s added to the compass stated heading. But if now you climb away from the interference field, the compass swings around and you have an anomaly in the control computer. No angular acceleration but a change in heading. Correct?
 
Not sure I understand this sar104. I assume when the drone boots up, the IMU has no info on heading. Presumably the compass provides that data. The IMU just integrates angular acceleration and it’s added to the compass stated heading. But if now you climb away from the interference field, the compass swings around and you have an anomaly in the control computer. No angular acceleration but a change in heading. Correct?
Not exactly. The compass does initialize the IMU heading, but it's not a subsequent discrepancy between the compass and the IMU headings that causes loss of control - it's simply the IMU heading being wrong. If the IMU heading is wrong then when the FC attempts to correct for positional drift then it corrects in the wrong direction - that's the source of uncontrolled flight.
 
  • Like
Reactions: Rich QR
The calibration process subtracts the magnetic field of the aircraft so that the compass accurately measures just the external (presumed to be the earth's) magnetic field. If it is out of calibration then the compass heading will be wrong. For small errors the most obvious indication is that it will tend to crab rather than fly straight forwards. Larger errors may result in uncontrolled flight.

Note that calibration is absolutely not correcting for declination or for external magnetic fields that locally distort the earth's magnetic field.

For more details:

SAR104, another great conversation, and thanks for sharing your considerable knowledge. Now I can understand the recommendation to recalibrate "3. When the mechanical or physical structure of the Phantom 3 Professional has been changed." (DJI P3P manual page 45)

But it also recommends: "2. When flying in a new location or in a location that is different from the most recent flight." I'm guessing that's because the earth's magnetic field has both vertical and horizontal components and the compass needs information about that in order to tease direction with respect to the earth's local surface from the observed field. Is that so?
 
But it also recommends: "2. When flying in a new location or in a location that is different from the most recent flight." I'm guessing that's because the earth's magnetic field has both vertical and horizontal components and the compass needs information about that in order to tease direction with respect to the earth's local surface from the observed field. Is that so?
No it's not.
There's no physical reason that your drone's compass would need to be recalibrated because you've moved from where you last flew.
The calibration does nothing except identifying and measuring the magnetic fields that are part of the drone.
It has nothing to do with the earth's magnetic field or anything outside of the drone itself.

DJI has a very poor record an documenting compass calibration in their manuals, even including misinformation and have been the cause of lots of myth and confusion.
 
Last edited:
SAR104, another great conversation, and thanks for sharing your considerable knowledge. Now I can understand the recommendation to recalibrate "3. When the mechanical or physical structure of the Phantom 3 Professional has been changed." (DJI P3P manual page 45)

But it also recommends: "2. When flying in a new location or in a location that is different from the most recent flight." I'm guessing that's because the earth's magnetic field has both vertical and horizontal components and the compass needs information about that in order to tease direction with respect to the earth's local surface from the observed field. Is that so?
As @Meta4 stated, and as explained in more detail here, the calibration is just to subtract the internal magnetic field of the aircraft. The compass is primarily used simply to determine magnetic north from the horizontal field component to initialize the IMU heading and then as a low-gain correction for IMU heading drift. The vertical component isn't really relevant except as a check that the total field vector is nominally as expected in magnitude and direction. Nor is the calibration used to determine the local magnetic declination - it cannot do that since it has no way to measure the direction of true north. Declination is calculated from location by the FC using a fairly simple global model for the earth's field.
 
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Forum statistics

Threads
130,586
Messages
1,554,118
Members
159,586
Latest member
DoubleBarS