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

Mavic 3 Cine hit a tree

mikeol999

Member
Premium Pilot
Joined
Jan 11, 2022
Messages
10
Reactions
2
Age
61
Location
Ballarat, Victoria, Australia
This is why a pilot should check drone arrow orientation on the map, to the actual drone orientation on the ground before take off (when GPS mode is active), and why the lady always says "Home point recorded, please check it on the map."

Having the low magnetic influence causes this interference (and not give any warning).
It doesn't put the drone into atti mode, it just causes the drone to fly off whiele attempting to adjust corrections, totally the wrong way, usually a very fast path, no control, generally until signal is lost, or a crash happens.
A true flyoff but totally unavoidable.

I have read that it can correct itself in a short time in the air, one or two cases I think in the crash section threads over the years . . . but usually this doesn't happen fast enough.

If the map arrow and ground direction are different, all you need to do is shut down, move the drone a little to a new take off point, and try start up again.
This usually fixes the issue.
I think it did give the message that the homeport had been updated, but maybe not. Can't remember now, but I usually make sure it has. however, I was only flying a few feet away to take a photo so may not have worried about it. I haven't found any advice on the DJI website related to the Mavic 3 where it is suggested that you should check if the map arrow and ground direction are the same
 

Ralph thompson

Well-Known Member
Joined
Feb 26, 2019
Messages
558
Reactions
380
You’ll know your in ATTI mode because the drone will drift with the wind. you have to be quick on the controls and fight the drift. A colleague of mine taught drone flying with an old Phantom which had an ATTI mode switch on the controller. You could try to find an old phantom to practice. You can also fly indoors, in a barn or A/C hanger, maybe under a large overpass where the GPS signal gets blocked. ATTI mode isn’t difficult you just have to be super attentive otherwise the drone will quickly drift away.
 

mikeol999

Member
Premium Pilot
Joined
Jan 11, 2022
Messages
10
Reactions
2
Age
61
Location
Ballarat, Victoria, Australia
You’ll know your in ATTI mode because the drone will drift with the wind. you have to be quick on the controls and fight the drift. A colleague of mine taught drone flying with an old Phantom which had an ATTI mode switch on the controller. You could try to find an old phantom to practice. You can also fly indoors, in a barn or A/C hanger, maybe under a large overpass where the GPS signal gets blocked. ATTI mode isn’t difficult you just have to be super attentive otherwise the drone will quickly drift away.
Yeah, it was drifting. Couldn't get it to hold still to get the photo I wanted. That was when I tried to bring it back home. But, it got faster and faster as it came towards me and then didn't seem to respond when I tried to get it to stop. Flew straight past and into a tree behind me. Had never heard of ATTI before, nor what it does. If it was confused with GPS etc, it should either not take off or tell you to do an emergency landing. No point giving what seemed a casual warning and then be uncontrollable
 

Meta4

Well-Known Member
Joined
Oct 30, 2016
Messages
12,433
Reactions
14,647
Age
66
Yeah, it was drifting. Couldn't get it to hold still to get the photo I wanted. That was when I tried to bring it back home. But, it got faster and faster as it came towards me and then didn't seem to respond when I tried to get it to stop. Flew straight past and into a tree behind me. Had never heard of ATTI before, nor what it does.
In this incident atti mode wasn't the real problem.
It was a symptom of the yaw error issue which caused the runaway crash at the end of the flight.

If it was confused with GPS etc, it should either not take off or tell you to do an emergency landing. No point giving what seemed a casual warning and then be uncontrollable
GPS wasn't a problem.
A conflict between the compass and the gyro sensor data was the issue and everything else followed on from this.
There was no way for the flight controller to tell there was a problem to be able to prevent takeoff.
In the terminal stage of a yaw error, there is no way to control the drone to make an emergency landing.
Preventing a yaw error from happening is the only way to deal with them.
 

MAvic_South_Oz

Well-Known Member
Premium Pilot
Joined
May 25, 2017
Messages
7,937
Reactions
7,225
Age
60
Location
South Australia, the great southern land
I think it did give the message that the homeport had been updated, but maybe not.

It does still give the standard GPS mode message, even with the mild magnetic interference, and conflicting compass data between the compass itself and the flight controller.

The " . . . please check it on the map" is basically wanting you to check the map, general location, and correct drone arrow and ground alignment ARE correct.

What's important is doing that check . . .

This is why a pilot should check drone arrow orientation on the map, to the actual drone orientation on the ground before take off (when GPS mode is active), and why the lady always says "Home point recorded, please check it on the map."

I haven't found any advice on the DJI website related to the Mavic 3 where it is suggested that you should check if the map arrow and ground direction are the same

You possibly won't see this info in general under advice on the DJI forum, but this forum section (Crash analysis) is littered with posts outlining this problem and it's easy prevention.

It's certainly nothing to do with atti mode, that's an entirely different scenario, and one rarely encountered . . . IF a pilot waits for GPS mode to appear, and yours probably did if the HP recorded message did come up, as this is what is announced when the GPS function kicks in.
 

MAvic_South_Oz

Well-Known Member
Premium Pilot
Joined
May 25, 2017
Messages
7,937
Reactions
7,225
Age
60
Location
South Australia, the great southern land
It was a symptom of the yaw error issue which caused the runaway crash at the end of the flight.

Ok, as no one else has seem the flight record . . .

Was it a short flight ? (I imagine it must have been.)
Did it have the yaw error the whole flight time ?

Must admit, I have only seen the yaw error happen right from the flight take off, not in the middle or end of a flight.
 

Meta4

Well-Known Member
Joined
Oct 30, 2016
Messages
12,433
Reactions
14,647
Age
66
Was it a short flight ? (I imagine it must have been.)
Most, but not all yaw error incidents result in short flights.
The difference between the incorrect gyro heading and the true compass heading varies and this has a lot of effect on how the flight is affected.
This flight was longer than most yaw error incidents, which accounts for the (longer than most) period of atti mode flight, before things became irretrievably bad.
Did it have the yaw error the whole flight time ?
Yes .. because the yaw error is created at startup,
For this flight it is first seen in the flight data at 3.6 seconds when the drone climbs to 2 ft away from the magnetic interference and the compass starts to give normal readings.
As the drone gained height, the difference between compass and gyro sensor data increased.
 

MAvic_South_Oz

Well-Known Member
Premium Pilot
Joined
May 25, 2017
Messages
7,937
Reactions
7,225
Age
60
Location
South Australia, the great southern land
Most, but not all yaw error incidents result in short flights.
The difference between the incorrect gyro heading and the true compass heading varies and this has a lot of effect on how the flight is affected.
This flight was longer than most yaw error incidents, which accounts for the (longer than most) period of atti mode flight, before things became irretrievably bad.

Yes .. because the yaw error is created at startup,
For this flight it is first seen in the flight data at 3.6 seconds when the drone climbs to 2 ft away from the magnetic interference and the compass starts to give normal readings.
As the drone gained height, the difference between compass and gyro sensor data increased.

Thanks.
How long was this flight ?

It does seem very different to most yaw error flights, as the OP's M3 didn't fly off very much sideways, or in a big arc, at very quick speeds, like most of these incidents.

Just "drifting" . . .

"It would not hover in one place and it would keep drifting at the same altitude without me touching the controllers. So, I tried to bring it home. However, it took off with increasing speed and fly over my head and into a tree. "

It might be as simple as the yaw error being VERY slight, rather than 90 or 180 degrees often seen.
 

Meta4

Well-Known Member
Joined
Oct 30, 2016
Messages
12,433
Reactions
14,647
Age
66
How long was this flight ?
About 18 minutes
It does seem very different to most yaw error flights, as the OP's M3 didn't fly off very much sideways, or in a big arc, at very quick speeds, like most of these incidents.
Unfortunately there is no position or speed data available for the end of the flight, but the OP described a typical high speed runaway when things went bad at the end.
 
  • Like
Reactions: MAvic_South_Oz

Cookedinlh

Well-Known Member
Joined
Oct 17, 2016
Messages
454
Reactions
213
Location
Sarnia ON Canada
Often you can never say for sure the compass was the culprit BUT . . . I learned the hard way never take off from concrete (due to rebar), some sandy beaches are really bad too (high iron content) or near vehicles <10-15 ft (too much metal) or even near a fence especially chain link<10-15ft. Look for open grass dirt tarmac or . . what I prefer always is HAND LAUNCH. . .and (careful doing that in windy conditions). Then no matter what do a "hover check" at 10 ft and Wait for the "Home point has be set please check it on the map" before doing any maneuvers. Then just before departing check SATS >10 and heading matches in the way you're pointing . . just like real pilots do . . . saves a lot of agony trying to figure out what happened AFTER the fact. Also if you're new t the game try this "ZEN and the ART of Done Happiness" . . . and have more fun! . . I've even gone to using a Megnetometer APP available for Android and iPhone . . it works . . . very sensitive
 

mikeol999

Member
Premium Pilot
Joined
Jan 11, 2022
Messages
10
Reactions
2
Age
61
Location
Ballarat, Victoria, Australia
PS: Back to my original question - should you submit a service repair request or a DJI Express Care request, my advice is to go down the service repair request route first. Notwithstanding, DJI's push for you to use DJI Care Refresh, the repair may not be very expensive (just over $100 in my case) and you can always swap over to Care Refresh if it ends up being cheaper. So, provided you have not lost the drone, I would get the quote first. Also, I have found that neither option guarantees you a new drone. It is either repair or replacement, with the replacement sometimes being new and sometimes being a previously repaired drone.
 

Meta4

Well-Known Member
Joined
Oct 30, 2016
Messages
12,433
Reactions
14,647
Age
66
PS: Back to my original question - should you submit a service repair request or a DJI Express Care request, my advice is to go down the service repair request route first.
Great result Mike.
Not many people would have had a Mavic 3 for >12 months.
Wouldn't a repair, be a warranty case (no charge)?
 

DJI Drone Deals

New Threads

Forum statistics

Threads
123,288
Messages
1,450,967
Members
150,360
Latest member
Allen13331