The compass, what Mavic uses it for?

[AlanTheBeast]

You simply have this wrong w.r.t. rotary-wing aircraft, my friend.

"Heading" is defined as the compass direction the longitudinal axis of the aircraft is pointing, which invariably is the direction the "nose" is pointing.

"Course" is the compass direction of the aircraft's velocity vector, or put more simply the compass direction the aircraft is moving over the ground.
[COMMENT: NO. That is the track]

In fixed-wing aircraft, course and heading are the same, so are sometimes lazily used interchangeably. However, the definitions above have always been in place.
[COMMENT: No. Aircraft are not immune to crosswind, and so heading, track and desired course are different things]

With rotary-wing aircraft like helicopters and multirotors, this relationship is broken, and aircraft are free to point in a different direction than they are flying -- i.e. have a different heading than their course.

In FW aircraft heading, track and course are not the same (see my post above). ie: an aircraft in a crosswind needs to adjust heading to maintain a track (desired course).

A helicopter is of course free (like a drone) to not fly longitudinal heading to get to a point (you could elect to fly DC to NYC sideways), but in practical "navigation" terms, helicopters are no different than FW aircraft.

 

[Aerial West]

Is this entire argument to discover that they work together to establish position and course?

Because I don't think it will end up being that DJI put a whole system in there that they didn't need.

 

[Logger]

In FW aircraft heading, track and course are not the same (see my post above). ie: an aircraft in a crosswind needs to adjust heading to maintain a track (desired course)......

Exactly and it would look like this.



Further more Fixed Wing aircraft can also diverge course and track from heading by flying in a crabbed state either by side-slipping or thrust Asymmetry regardless of wind. So somewhat like a drone, not travelling through the air along its longitudinal axis.

 

[Aerial West]

I crabbed cessna 107/g from Ohio to Southern cali last year. Im not sure I ever had a straight heading, ha.

 

[Will.i.am]

I said I was done with this conversation, but it has popped back up and I still see much confusion. The original question - why does the Mavic need a compass - is a good one, and I thought the answer was simple. I think the confusion comes partly from the use of the term 'compass' and common association of it with the aircraft heading. The compass in the Mavic (otherwise known as the magnetometer) is an essential component of the IMU. The other two components are the gyroscope and the accelerometer. Here is a detailed discussion of the purpose of each component, and the real answer to the OPs question. Some in this thread have wrongly suggested that the compass is not necessary, and it would be possible to fly the Mavic with just information from the GPS or derived from measurements done with it. As the linked discussion concludes, and what I have been trying to explain, is that the IMU in the Mavic could not function (or at least function well enough for flight) without all three components. I will quote the relevant statement if you don't feel like reading the whole thing -
"A magnometer senses magnetic flux density across the pitch-, roll- and yaw-axis. A system could use this data, combined with data from the accelerometer and gyroscope, to estimate the aircraft's orientation relative to the Magnetic North Pole and use this feedback to manipulate angular velocity such that a desired angular position is maintained in the yaw-axis. Without a magnometer, only the angular velocity can be maintained in the yaw-axis (not the angular position, which exists only relative to some known direction)."

Conclusion - the compass is needed by the IMU in the Mavic, and the Mavic could not fly without one.

 

[Nomand Eloc]

You guys sound like a bunch of smart nerds arguing over smart stuff. That said I finally tried to take off while my Mavic was warning me to calibrate the compass, mind you I was in a safe location in the middle of the Pismo beach sand dunes with nothing to hit but sand and my friends 50,000$ truck. Taking off from the hood, or bonnet for me English friends proved interesting. The Mavic went up a few and just started to slowly spin. It would not stop rotating even with opposite correction. I had to hand catch the bird due to the sand only landing option, which was quite funny and challenging to say the least. That said I think the compass is quite useful on this flying pieces of plastic and magnesium.

 

[sar104]

I said I was done with this conversation, but it has popped back up and I still see much confusion. The original question - why does the Mavic need a compass - is a good one, and I thought the answer was simple. I think the confusion comes partly from the use of the term 'compass' and common association of it with the aircraft heading. The compass in the Mavic (otherwise known as the magnetometer) is an essential component of the IMU. The other two components are the gyroscope and the accelerometer. Here is a detailed discussion of the purpose of each component, and the real answer to the OPs question. Some in this thread have wrongly suggested that the compass is not necessary, and it would be possible to fly the Mavic with just information from the GPS or derived from measurements done with it. As the linked discussion concludes, and what I have been trying to explain, is that the IMU in the Mavic could not function (or at least function well enough for flight) without all three components. I will quote the relevant statement if you don't feel like reading the whole thing -
"A magnometer senses magnetic flux density across the pitch-, roll- and yaw-axis. A system could use this data, combined with data from the accelerometer and gyroscope, to estimate the aircraft's orientation relative to the Magnetic North Pole and use this feedback to manipulate angular velocity such that a desired angular position is maintained in the yaw-axis. Without a magnometer, only the angular velocity can be maintained in the yaw-axis (not the angular position, which exists only relative to some known direction)."

Conclusion - the compass is needed by the IMU in the Mavic, and the Mavic could not fly without one.

The confusion is rather inevitable in these kinds of discussions. The link that you provided does provide some background but nowhere does it mention exactly how the data from the sensor suite is combined. It's worth revisiting precisely what type of data are provided by the various sensors in terms of flight stability:

GPS: (1) Instantaneous 3-D absolute position from data TOA with relatively poor resolution compared to the required stability of the aircraft and (2) instantaneous 3-D velocity from carrier signal Doppler shift, again with limited resolution for flight control.

Magnetometers: Instantaneous aircraft orientation relative to the local component of the earth's magnetic field. Since the magnetic field is a simple vector, no orientation information about the axis parallel to the magnetic field.

Rate gyros: Accurate instantaneous rate of change of orientation in °/s around the three principle axes of the aircraft.

Accelerometers: Accurate instantaneous acceleration components along the three principle axes of the aircraft. But since acceleration is indistinguishable from local gravity, these data do not uniquely define orientation relative to the gravitational field unless the aircraft is stationary or moving at constant velocity. These are therefore both relative and absolute data.​

In principle, given a starting location and orientation (GPS and magnetometers), the rate gyro and accelerometer data can be integrated with respect to time to give the resulting location and orientation (pure inertial guidance), but these kinds of sensors have both noise and bias/drift errors, and those sum under integration such that after some (relatively short) time both location and orientation are significantly incorrect. Integrating the rate gyro and accelerometer data from the DAT files is quite instructive for seeing how fast that drift occurs.

To overcome these limitations, a Kalman filtering scheme (or equivalent) is to correct the time-integrated rate gyro data with the absolute accelerometer and magnetometer data, resulting in quite accurate instantaneous pitch, roll and yaw values. This could not be achieved without the magnetometers. This doesn't require any GPS data, and is how the aircraft holds attitude just fine in ATTI mode.

Accurate position is similarly achieved by using absolute GPS data to correct for cumulative errors in the integrated accelerometer data that would otherwise result in drift in position. The FC could use just accelerometer data to attempt to hold position, but it would start to drift with time as the errors went uncorrected with any absolute position data so in ATTI mode it does not even try.

 

[AlanTheBeast]

Exactly and it would look like this.

View attachment 21692

Isn't the selected course 325° there? (VOR)?

 

[Big Fil]

A mavic, or any other drone/rc heli is perfectly capable of flying without a compass or GPS. You may not be able to fly it, but there are plenty of people that can fly it with out GPS/compass aids. Hint, ATTI mode does not rely on compass or GPS. The only sensors at play in ATTI mode are the IMU (gyros/accelerometers) and the pilot's eyes. Where the compass becomes necessary is for P-mode flying and all the other intelligent flight modes which are reliant on P-mode/GPS based flying.

 

[Logger]

Isn't the selected course 325° there? (VOR)?

Perhaps you meant 335, but no. Tail of the VOR needles simply indicates AC is currently on the 335 radial or course from an auto tuned VOR station behind the AC that happened to be nearby. Ground track is the small green drift diamond at 328 and the flight planned course over the ground is the green line. Sorry to digress though. Was simply backing up the concept that FW heading does not have to match track or course due drift in this case or crab in other situations.

 

[Logger]

A mavic, or any other drone/rc heli is perfectly capable of flying without a compass or GPS... Hint, ATTI mode does not rely on compass or GPS.....

Whilst capable of doing so. Hint, the number of compass issues that have led to Mavic prangs would suggest it is not perfect at doing so

 

[sar104]

A mavic, or any other drone/rc heli is perfectly capable of flying without a compass or GPS. You may not be able to fly it, but there are plenty of people that can fly it with out GPS/compass aids. Hint, ATTI mode does not rely on compass or GPS. The only sensors at play in ATTI mode are the IMU (gyros/accelerometers) and the pilot's eyes. Where the compass becomes necessary is for P-mode flying and all the other intelligent flight modes which are reliant on P-mode/GPS based flying.

While we don't know the full details of the FC control schemes, the compass is almost certainly playing a role in ATTI mode as the magnetometer data are required in the Kalman filter to correct for the drift in the integrated rate gyro data. In other words the FC cannot even maintain attitude without including the magnetometer data.

 

[AlanTheBeast]

While we don't know the full details of the FC control schemes, the compass is almost certainly playing a role in ATTI mode as the magnetometer data are required in the Kalman filter to correct for the drift in the integrated rate gyro data. In other words the FC cannot even maintain attitude without including the magnetometer data.

I doubt that very much. The Magnetometer is a low noise heading reference and little else. Plenty of drones do not have a magnetometer.

 

[sar104]

I doubt that very much. The Magnetometer is a low noise heading reference and little else. Plenty of drones do not have a magnetometer.

It may seem unlikely or counter-intuitive, but the strapdown, 9-DOF MEMS IMU sensor packages such as those used in DJI UAVs almost always employ a sensor fusion scheme that combines integrated rate gyro data with instantaneous magnetometer and accelerometer data to determine orientation. The magnetometers are not used as a simple compass to determine heading - pitch, roll and yaw are instead all extracted from the orientation quaternions (or Euler angles).

It's not that an IMU cannot function at all without magnetometer data, it's just that time-integrated gyro bias errors, in particular, lead to accumulating yaw errors even when stationary, and under significant rotation and acceleration that also leads to pitch and roll errors since the accelerometers cannot reference an unambiguous z-axis. The addition of the magnetometer data improves the solution considerably.

For a simple explanation of the background to this I liked this report, but if you do a search for articles on IMU sensor fusion there are many out there. It is possible, as I mentioned, that DJI ignored the prevalent technology and went with a different solution, but that seems unlikely to me. However, if anyone has definitive information on that subject I'd be very interested.

 

[LivinLarge]

and

 

[sar104]

and

As I said, the point is not that the IMU cannot function without the magnetometers, it is that it works much better with them. The question that started this thread was about how the IMU uses the compass, not how well it copes without it.

 

[LivinLarge]

As I said, the point is not that the IMU cannot function without the magnetometers, it is that it works much better with them. The question that started this thread was about how the IMU uses the compass, not how well it copes without it.

"it is that it works much better with them"
Yes, if the compass is working correctly and yes if there is no magnetic interference. If the flight controller is able to detect that the compass is not working correctly or that there is interference and switches to ATTI mode, then the bird can still be flown home assuming you are within sight.
Early Phantoms did not automatically switch to ATTI mode when there was a compass failure or interference and this caused many flyaways. If the pilot was able see there was a problem and manually switch to ATTI mode, then the Phantom could be flown home. I had this happen several times with a P1.
How accurate does the compass need to be?
Here is an example of a compass being intentionally offset from the axis of the aircraft by 10 or 15 degrees. The quad still flew fine but had a slight J hook at the end of its travel.