I will admit to not having fully read all the comments in this topic. But DJI does have flyaway insurance....if your drone is bound to the remote. Did I miss something where that did not happen? Or does the OP simply not have the Care Refresh insurance?

hunterws Posted at 10-29 11:04
buuuuut, dji care refresh is sold on the premise it's to protect against pilot error, warranty would take care of manufacture defects.

buuuuut, dji care refresh is sold on the premise it's to protect against pilot error, warranty would take care of manufacture defects.
Care refresh and warranty claims require the drone or wreckage be returned to DJI.
Unfortunately for the OP, he chose to fly in a location where recovery isn't possible.
There's no indication of any malfunction or defect.

GaryDoug Posted at 10-29 12:16
I will admit to not having fully read all the comments in this topic. But DJI does have flyaway insurance....if your drone is bound to the remote. Did I miss something where that did not happen? Or does the OP simply not have the Care Refresh insurance?

Or does the OP simply not have the Care Refresh insurance?
He hasn't mentioned Care Refresh or "fyaway" insurance, which suggests he has neither.

Labroides Posted at 10-29 15:34
Or does the OP simply not have the Care Refresh insurance?
He hasn't mentioned Care Refresh or "fyaway" insurance, which suggests he has neither.

Suggests? Why not if we ask? I have never seen you assume anything ;-) By the way, the flyaway insurance is part of the Care Refresh package and does not require return of the drone. It only requires you to bind it to a remote. All of mine (except the M2P) are bound to their remotes.

GaryDoug Posted at 10-29 19:15
Suggests? Why not if we ask? I have never seen you assume anything ;-) By the way, the flyaway insurance is part of the Care Refresh package and does not require return of the drone. It only requires you to bind it to a remote. All of mine (except the M2P) are bound to their remotes.

There's a difference between assuming and noting that something suggests something.

You're correct about Care Refresh.
I was confusing it with DJI Care which requires recovery and returning the wreck.

Labroides Posted at 10-28 04:55
You should check your facts before going off with your simplistic assessment.
Have a look at the flight data here:
https://www.phantomhelp.com/LogViewer/1BVLY437WMOBXTNYIJCP/

And exactly what are you getting out of the flight data that I am not.  Show me where there was a SAT lock?

Labroides Posted at 10-29 15:31
buuuuut, dji care refresh is sold on the premise it's to protect against pilot error, warranty would take care of manufacture defects.
Care refresh and warranty claims require the drone or wreckage be returned to DJI.
Unfortunately for the OP, he chose to fly in a location where recovery isn't possible.

There's no indication of any malfunction or defect.
Nope, just pilot error, which is what this was, and djicare is suppose to cover.
I get it, you gotta get the body or no go.  Maybe he can still get to it if it's important enough ...?

Mhinc Posted at 10-29 23:28
And exactly what are you getting out of the flight data that I am not.  Show me where there was a SAT lock?

I pointed you to the flight data in which it is obvious.
If you can't read that, I can't help you.

Mhinc Posted at 10-29 23:28
And exactly what are you getting out of the flight data that I am not.  Show me where there was a SAT lock?

at 1.9 seconds in Flight > message >  "Homepoint updated. RTH altitude adjusted to 30 m"

This message in only there if the GPS icon is white coloured = enough satellites to set HP
(before that amber coloured = GPS lock)

cheers
JJB

hunterws Posted at 10-30 03:52
There's no indication of any malfunction or defect.
Nope, just pilot error, which is what this was, and djicare is suppose to cover.
I get it, you gotta get the body or no go.  Maybe he can still get to it if it's important enough ...?

Hi,

Yes there are indications of other malfunctions in the flightlog.

During autotake off the vertical speed shows negative numbers while drone is climbing, see VPS height indication.
Baro height from zero at takeoff to minus 267 meters at end of the flight.
Vertical speed to max -25 mtr/sec down!
If these values are correct than the splash in the water had to be real close to the start point, actually it splashed down 124 meter away after 15 seconds.

With a 'normal' yaw compass error these baro height and vert speed values are never seen like this, as this part of the IMU datalines are not affected by a yaw channel error.
But is not sexy to mention this, becsause this means that it is not a 'normal' yaw/compass error fly away....

AND it is normal to get a max power warning during a fly away as the drone does correct to the max. Max power warning at 3.8 seconds in flight, with normal pitch and roll drone angles. (less < 10 degrees both)
BUT in a max power correction yaw/compass error these angles 'normally' high values!

So many out of the normal indications for just a simple yaw/compass error.

cheers
JJB

Looking at both the flight log and flight playback video, at 11 seconds, for instance, the horizontal speed is 47 mph and vertical speed is 78 m/s(140 mph) also IMU altitude is -600 ft. Seems to me that either the nav processor or the gps unit fried. The numbers are way too far off to be a simple compass/gps lock error.

Mobilehomer Posted at 10-30 07:00
Looking at both the flight log and flight playback video, at 11 seconds, for instance, the horizontal speed is 47 mph and vertical speed is 78 m/s(140 mph) also IMU altitude is -600 ft. Seems to me that either the nav processor or the gps unit fried. The numbers are way too far off to be a simple compass/gps lock error.

Agreed.......

Mobilehomer Posted at 10-30 07:00
Looking at both the flight log and flight playback video, at 11 seconds, for instance, the horizontal speed is 47 mph and vertical speed is 78 m/s(140 mph) also IMU altitude is -600 ft. Seems to me that either the nav processor or the gps unit fried. The numbers are way too far off to be a simple compass/gps lock error.

In post #101 the actual flying speed is shown in meter/sec

cheers
JJB

JJB* Posted at 10-31 01:20
In post #101 the actual flying speed is shown in meter/sec

cheers

Yes, I converted it. Still it indicates more than just a simple compass error.

Mobilehomer Posted at 10-31 05:13
Yes, I converted it. Still it indicates more than just a simple compass error.

I know  ;-)  as you have read in all my posts in this thread.

Just trying to explain that the indicated speed in the app is only correct if there are no yaw/compass/speed/imu errors.

If the GPS is killed to use by the software, than the GPS coordinates are still recorded.
And use of the coordinates over time gives the correct speed.

cheers
JJB

You don't need to test as many flyers have done it for you already and most have ended badly.

Appart of the fact that it would be my decision, lets assume this is how it happens. If thats wrong please correct me:
Drone starts up -> gets apparent orientation (where is it facing) from magnetic compass -> lifts off -> change in magnetic field is significant -> apparent orientation changes -> drone has trouble

There are at least 4 things that make this unlikely:
1. The drone doesnt need to know its orientation relative to earth to hover. All it needs to know is its angle relative to ground (gyroscope) and acceleration (to stay at one position). All of these are constructively fix, no user input. It also wants to hover facing the same direction, which it doesnt need to know relative to earth, just maintain whatever it is currently (eg. other axis of accelerometer provides this information as rate of change). In any case, this would only result in a rotation around its axis, not flying anywhere, no matter how much it would have to correct itself.
2. If the magentic compass would be THAT important, we would see such crashes all the time, since the magnetic fields on this planet are everything but constant. Not only are there very local deviations, if you were to travel 10'000 km and then take off (without calibration) it would guarantee that the done crashes. Also see 4.
3. Flying next to anything that affects the magnetic field should also make the drone crash, why would it be limited to the startup? If its apparent orientation changes mid flight it would have the same issue. If not: Why not?
4. Flying inside builings (wood/steel/concrete/combination) is no problem, despite them beeing a massive issue for a compass since the magnetic field is all over the place.

Just search for yaw error in google and you should find many incidents.

Yes, searching ''yaw error drone'' results in incidents of crashes and someone saying something about yaw error, just like in this thread here. Thats neither an explanation nor proof of it beeing related to the compass. (side not: Some of them, eg. the first hit, say this is not user error). Here is an example on youtube with massive changes of magnetic field, the drone only starts to drift a little bit more (also due to bad GPS):

Eheran Posted at 11-1 00:01
You don't need to test as many flyers have done it for you already and most have ended badly.

Appart of the fact that it would be my decision, lets assume this is how it happens. If thats wrong please correct me:

There are at least 4 things that make this unlikely:
1. The drone doesnt need to know its orientation relative to earth to hover.
This is only partly correct.

The drone needs to know its orientation to make any corrective movements.

In a yaw error situation, the problem becomes obvious after the drone changes its orientation when away from the magnetic interference..
That's when there is conflicting data from the gyro sensor and compass starts.
When the drone attempts to correct, but moves in the wrong direction because of the incorrect initialisation of the gyro, it senses that it is further off course and corrects again & again etc.
So it wants to hover but cannot correct the uncommanded movement that has started.


2. If the magentic compass would be THAT important, we would see such crashes all the time, since the magnetic fields on this planet are everything but constant. Not only are there very local deviations, if you were to travel 10'000 km and then take off (without calibration) it would guarantee that the done crashes. Also see 4.
This is not true at all.
The local magnetic variation is irrelevant in drone flying.
There is no need to recalibrate the compass after travelling any distance.
I'm not sure why you suggest that variations in the earth's magnetic field would crash drones.

The issue in this thread is a yaw error - the incorrect initialisation of the gyro sensors due to local magnetic interference.
The most common cause is powering on on reinforced concrete.

3. Flying next to anything that affects the magnetic field should also make the drone crash, why would it be limited to the startup? If its apparent orientation changes mid flight it would have the same issue. If not: Why not?
Flying close to a large steel object will upset the compass and can cause erratic flight.
Try it yourself sometime by slowly lowering your drone toward a steel roof.
As you get closer (probably <2 metres), you'll see the effect of disrupting the compass start.
Climb away from the steel to recover.

But doing that does not give the gyro sensor a false heading like it gets on powering up in a magnetically contaminated environment.
Climb away from the steel and the compass and gyro data agrees again.

4. Flying inside builings (wood/steel/concrete/combination) is no problem, despite them beeing a massive issue for a compass since the magnetic field is all over the place.
If the effect was massive you'd see problems as I outlined above.
If the drone is a metre or so away from the concrete, the magnetic interference isn't enough to cause issues.

JJB* Posted at 10-31 05:25
I know  ;-)  as you have read in all my posts in this thread.

Just trying to explain that the indicated speed in the app is only correct if there are no yaw/compass/speed/imu errors.

What if the GPS module goes bad? A very rare occurrence, but it could happen. That might explain the loss of GPS lock with 19 satellites.

Mobilehomer Posted at 11-1 05:17
What if the GPS module goes bad? A very rare occurrence, but it could happen. That might explain the loss of GPS lock with 19 satellites.

There was nothing wrong with his GPS hardware.
The problem was that his IMU's gyro sensor was wrongly initialised.
The flight controller couldn't deal with the data conflict, so stopped using the GPS data, although it continued to record GPS data.

Mobilehomer Posted at 11-1 05:17
What if the GPS module goes bad? A very rare occurrence, but it could happen. That might explain the loss of GPS lock with 19 satellites.

Hi,

I think there was that nothing wrong with his GPS.
DJI software kills the use of GPS data if there is a yaw/compass conflict.

Not sure about that yaw gyro initialisation though.....see the following chart

#5 ChartInChart (top = 1) ;
red cross = max power load warning. (just after takeoff)
black line = not enough power

#4 ChartInChart
Blue line = Vert speed blue curved line ; going into high negative values! (while drone raised height)(during AutoTakeOff, just 1.7 secs in flight, the verticcal speed goes from zero to minus 1.4 m/s down!, IMU calculates this....)
Purple line = compass line, see it changes at takeoff, and at record 60 the yaw reference line remains steady.
(purple line and Blue line stayed along each other till record 60, meaning that there is no conflict)

#1 ChartInChart
Blue line = baro height, going down!!

BUT  all this before compass line goes weird !! This is seen in the $4 chart above record # 60, at that point the compass heading was changing without RC input.

Conclusion :
if you look only at the compass data then simple cause : yaw/compass error > flyaway
if you look to all data than my conclusion : mayby another error in the hardware too, cannot say if this error caused the compass change thus fly-away OR only the yaw/compass conflict did cause the fly away.

It is always better to look to all data in in full context than just an easy pick for yaw/compass error.

If only DJI would have a look at this!
cheers
JJB

Labroides Posted at 10-26 05:16
I doubt it, but do some testing with a magnetic compass and see if teh watch can deflect the compass at a similar distance.

Would it be a good practice to bring a magnetic compass along, so we can check any interference with the compass before powering up our drone.  

This is a really good thread to follow, with a lot of info.  As for me, i am a newbie and know nothing about how gyro and IMU systems work.  But from my own experience, yes, I encountered some issues as I launched mine on a boat before.  The drone was unstable up in the air, fortunately i didn't fly it off too far and was able to bring it back.  Since then I've been researching a lot online and found out that I should have launched it up high up in the air when on a boat, which I tried doing a couple times and they were all successful.

I was never aware of the "initialization" being done at power-up.  So, if I power it up indoor with heavy interference, and then launch it outdoor, it could potentially have "fly away" issue?  Is there any way to reset or re-initialize the drone other than powering it off and on again outdoor?

EnjoyFlying Posted at 11-1 08:36
Would it be a good practice to bring a magnetic compass along, so we can check any interference with the compass before powering up our drone.  

This is a really good thread to follow, with a lot of info.  As for me, i am a newbie and know nothing about how gyro and IMU systems work.  But from my own experience, yes, I encountered some issues as I launched mine on a boat before.  The drone was unstable up in the air, fortunately i didn't fly it off too far and was able to bring it back.  Since then I've been researching a lot online and found out that I should have launched it up high up in the air when on a boat, which I tried doing a couple times and they were all successful.

Would it be a good practice to bring a magnetic compass along, so we can check any interference with the compass before powering up our drone.
No need to, your drone's magnetic sensor is a close equivalent.

So, if I power it up indoor with heavy interference, and then launch it outdoor, it could potentially have "fly away" issue?  
Yes

Is there any way to reset or re-initialize the drone other than powering it off and on again outdoor?
No

Labroides Posted at 11-1 12:39
Would it be a good practice to bring a magnetic compass along, so we can check any interference with the compass before powering up our drone.
No need to, your drone's magnetic sensor is a close equivalent.

But you said that if the interference is not stronger than the earth magnetic field, then it won't trigger any alert, and in that case, would those be enough to interfere a compass?  If not, then like you said, it will be a no need.

So, the only way (most reliable way) to check is thru checking the drone icon in the map, making sure that it matches with the drone.  Fly away would be a nightmare experience to me, and I hope we all can learn a solid way to avoid it.  Thanks to everyone for all the good tips and info.

OP - thanks for sharing your experience as this is very useful, especially to me as a noob.  Hope DJI can help you more on this, and do keep us updated.

EnjoyFlying Posted at 11-1 13:14
But you said that if the interference is not stronger than the earth magnetic field, then it won't trigger any alert, and in that case, would those be enough to interfere a compass?  If not, then like you said, it will be a no need.

So, the only way (most reliable way) to check is thru checking the drone icon in the map, making sure that it matches with the drone.  Fly away would be a nightmare experience to me, and I hope we all can learn a solid way to avoid it.  Thanks to everyone for all the good tips and info.

But you said that if the interference is not stronger than the earth magnetic field, then it won't trigger any alert, and in that case, would those be enough to interfere a compass?  If not, then like you said, it will be a no need.

We are talking about magnetic interference with a field strength approximately the same as the earth's. but which deflects the compass (which is usually very close to the problem)  so that it reads incorrectly.
The compass programming can't tell if that is the earth's normal field or not so it raises no alarm.

The way to avoid this problem is to confirm that the drone icon on your map view points the same compass direction as the actual drone.  eg.  drone facess east & icon points east.
It helps to avoid likely problem spots like any reinforced concrete surface.

Johnny_J Posted at 10-26 08:09
Wrong again! A gyro is trying to keep it's position by means of a fast spinning disc with substantial weight.
Whether it is spun up manually by human force or by a motor does not change the behavior of the gyro.
Without saying it, you assume that the gyro will spin forever. That is not true. Either the disc eventually stop due to friction or it keeps on spinning powered by whatever.

Hi Johnny,
Not all gyros deal with mass spinning about a rotational axis.  Long ago before solid state devices were used in everything that may have been the case, but they have come a long way since then.  Cheaper/Compact alternatives to the mechanical spinning gyros is the Ring Laser Gyro (RLG).  No moving parts, they operate on the Sagnac Effect using an inferometer to measure the interference created between 2 laser beams when the sensor is moved in a direction.  

Another type of gyro is the MEMS type, chip based sensors that use coriolis effect as opposed to spinning wheels.  These chip based MEMS gyros are most likely the type used by DJI in their drones, not the older spinning wheel type.  

Cheers,
Hazr

JJB* Posted at 2021-10-24 08:48
Hi,

Best to get help is to uplload your flightlog for this flight.

Hello my Mini 2 flew away by itself aswell, i have the flight logs, but cant post links, how can i reach support ?

The Drone flew perfectly for some time but suddenly started to fly backwards on max speed without my input, the return home signal could stop the reversing for a short time but then aborted and the drone flew away until i crash landed it in a field.

ExoTriXo Posted at 4-24 11:06
Hello my Mini 2 flew away by itself aswell, i have the flight logs, but cant post links, how can i reach support ?

The Drone flew perfectly for some time but suddenly started to fly backwards on max speed without my input, the return home signal could stop the reversing for a short time but then aborted and the drone flew away until i crash landed it in a field.

Hi,

Share your log, use this link > https://www.phantomhelp.com/LogViewer/Upload/
Post the uploaded link on here.
Or give a link to a cloud with the log there.

cheers
JJB