My Phantom 4 PRO flew away - Please help

Joined
Jul 21, 2018
Messages
10
Reaction score
2
Age
31
Hello everyone.

I am geologist in Brazil.

Last year I saved a lot of money to be able to buy a Phantom 4 Pro.

I used it a lot in the first half of 2018 on my work. I used it to take aerial photography of mineral areas.

All these areas are located in rural places, with no interferences like power, wifi, etc.
So today i went to take pictures of a granite area in the middle of nothing. At least 5 miles away from the nearest road.

Anyway, it all started good, like it always had. However, suddenly the drone started flying away.
Since it was such a rural place, I could not follow it through the mountains and everything, so I just started trying to make it come back and pressing the "come home" button.

It didnt work. It flew away. I didnt see it again and probably wont see.

Here is the flight log of this last flight: DJI Drone Help

My question with this topic is: Does DJI warranty cover this type of case? I am just desolated.
 
It didnt work. It flew away. I didnt see it again and probably wont see.
Sorry for the loss, however this is obviously pilot error, due to inexperience flying in ATTI mode. The RTH button has no effect in ATTI mode. You apparently had a compass issue right from the start, and the best case of action would have been to land immediately. That said, this is not something that would be warranty covered, IMHO. I have not looked at the full .csv log, to verify, but I don't think that would be necessary in this case.
 
Anyway, it all started good, like it always had. However, suddenly the drone started flying away.
Since it was such a rural place, I could not follow it through the mountains and everything, so I just started trying to make it come back and pressing the "come home" button.

It didnt work. It flew away. I didnt see it again and probably wont see.
The usual cause of the yaw error/compass error as shown in the flight data is launching from a steel or reinforced concrete surface.
It would appear that your compass was very close to a magnetic field when the Phantom was initialising and this has screwed upthe compass.
Having the compass error makes the Phantom ignore GPS data and it loses the horizontal position holding of GPS.
Sometimes a Phantom with a yaw error flies off at speed but yours was moving at 0-2 metres/sec for most of the time, possibly just drifting on the wind.
When the error message showed up, the Phantom was 46 metres up directly above the launch spot.
That would have been a good time to bring it down.
But it was sent higher (up to 120 metres) and drifted away.
The log doesn't show any movement of the right joystick during the flight.

At the end of the data, the Phantom was 440 metres away and still drifting with the battery at 87%.
Without GPS, it could not return home and would have continued to drift until the battery reached critical low voltage level and then landed.
It might have drifted for something like 10-15 minutes at 0-2 metres/sec
So it may be anywhere from 600 metres to 1800 metres away (or perhaps more?) off to the northwest.
 
  • Like
Reactions: BJ114 and Loz
From looking at the flight log u had issues from the start, it looks like it warned u a few times that things were not right. Sorry to heart lost your drone but from what I see it was definitely your fault, the aircraft should have been landed before it flew away on you. You tried to push it hoping the software would save the day but gps wasn’t working to fly back home.
 
Thanks everyone for the answeres.

So, the things are:

1) As you guys can see on the sattelite images, it was a rural area with absolutely none eletromagnetic interferences whatsover.

2) There were no strong winds or anything like that in the area (as we can see in the flight log, there were no "high wind, be cautions" message).

3) I lauched it from the soil. No concrete, no steel, nothing like that.

4) When I lauched it, I checked and everything was NORMAL. Compass, GPS, battery, you name it...

5) The first (and only) error message that appeared was: "compass error (or sth like that)".

6) When the error showed up, I saw it turned to ATTI mode and tried to bring it down but it was going away very fast so I flied it high for it not to crash into the moutains nearby.

7) I tested and flied ATTI mode before. The thing is that on yesterday's flight, the drone would not answer my commands to come back. It would only answer my commands to go higher and lower and as it was fast, I thought it would be wiser to send it high while I figured out what to do. Since it would not obey my commands to come back (hand flying or through RTH button).


I am not saying it is not my fault, I just feel it weird that I have to take responsability for a drone that suddenly changes the flying mode and STOP ANSWERING MY MANUAL COMMANDS.
 
2) There were no strong winds or anything like that in the area (as we can see in the flight log, there were no "high wind, be cautions" message).
First, there was no suggestion that there were strong winds. Even a mild breeze ( Especially at altitude will cause drifting in ATTI Flight )

4) When I lauched it, I checked and everything was NORMAL. Compass, GPS, battery, you name it...
This can very well be true. Compass Errors more often than not occur not too long after takeoff.

As @Meta4 pointed out above, there were no pitch and roll commands ( Right Stick ) issued by you according to the log file for the entire flight. Also as was mentioned, the RTH button is useless once switched to ATTI.

So in taking your word at the fact that you "issued" right stick commands, this would be one of two issues ( Of which I have never personally seen ) Either two receiver channels on the aircraft failed, or the same applies to the RC. Below is is a full graph of the RC inputs to the aircraft for the duration of the log. Notice the last two remain at zero movement. This is really here nor there in relation to the incident, but to throw out one thing, you took off well before the home point was set. You were already at ~12 meters when it updated.
Stick Inputs.png


Home Point.png
 
  • Like
Reactions: David Kraft
Hello everyone.

I am geologist in Brazil.

Last year I saved a lot of money to be able to buy a Phantom 4 Pro.

I used it a lot in the first half of 2018 on my work. I used it to take aerial photography of mineral areas.

All these areas are located in rural places, with no interferences like power, wifi, etc.
So today i went to take pictures of a granite area in the middle of nothing. At least 5 miles away from the nearest road.

Anyway, it all started good, like it always had. However, suddenly the drone started flying away.
Since it was such a rural place, I could not follow it through the mountains and everything, so I just started trying to make it come back and pressing the "come home" button.

It didnt work. It flew away. I didnt see it again and probably wont see.

Here is the flight log of this last flight: DJI Drone Help

My question with this topic is: Does DJI warranty cover this type of case? I am just desolated.


I believe the question was if DJI warranty covers this type of case ?

We had two incidents that occurred with an older firmware when in Tripod mode where the drone entered into Atti Mode and than stoped responding to any commands.
The Data report showed some Yaw Errors that occurred at that time of the fly aways..

On both occasions DJI gave us a coupon for 100% the cost of the new drone.

Its a very time consuming process to work with DJI but if you are patient , they will determine if it fits within the warranty.
Dont be surprised that every email written will require a day for them to respond, as we never got more than 1 email a day and had a total of about 30 back and fourth as
they asked different questions in regards to the flight, not everything is a cut and dry as the data report.

Note: We called every few days to check on the case once it was submitted to DJI to keep us in the light.
Approved Vendor
Phantomrain.org
 
I believe the question was if DJI warranty covers this type of case ?
Indeed it was. Each case will have it's own variables, and such that a warranty case is not up to "us". We wish to assist in any analysis that could be done from this end to allow the owner to make a sound decision as to how to approach the incident with DJI, for any warranty related specifics. DJI also uses the data ( and most instances will request such) to verify any claims made by the claimant for warranty purposes. This is another reason we try to provide as much information as possible, with as much relevant data as possible. It makes for a very strong case when you have actual data to back up any claims. Cut and Dry, no. But very close to that.
 
What could he have done to save the bird? He says he tried to land it in ATTI but it wouldn’t obey his inputs.
 
What could he have done to save the bird? He says he tried to land it in ATTI but it wouldn’t obey his inputs.
That is the question. If indeed he commanded right stick inputs. The aircraft never received them, according to the data files. The best option would have been at that point, to full down throttle. At the very least the aircraft would have been in a recoverable position, but to the OP's point there was no way that he would have known that the commands weren't received, other than visually.
 
  • Like
Reactions: Timinator and RPP
What could he have done to save the bird? He says he tried to land it in ATTI but it wouldn’t obey his inputs.
The problem was made worse by ascending instead of descending. It just flew higher and in a more unrecoverable area. The aircraft responded to altitude inputs so if it were me, as soon as I saw that it was not responding to other control commands I would have gone to stick down and have it land or crash where it was. At least that way I would have a better chance to recover the aircraft.
 
  • Like
Reactions: RPP
to throw out one thing, you took off well before the home point was set. You were already at ~12 meters when it updated.
launching with anything other than a fix from a dozen satellites on the GPS is risky.
This is not necessarily a problem and in some locations it is the only way you can fly.
If the launch point skyview is blocked by trees, terrain or buildings, it might be necessary to launch in atti mode and climb to clear the obstructions and get a good GPS signal.
The Phantom will record a homepoint when it gets GPS.
That appears to have been the situation in this case.
 
This is not necessarily a problem and in some locations it is the only way you can fly.
Completely agreed, that is quite possible.
 
Reading with interest, out of curiosity those two ‘flat line’ stick input graphs appear remarkably ‘flat’. I’ve no experience in reading these to be clear, I’m just asking, is that normal? Looking at the two graphs above these there appears to be what I’m guessing is ‘normal’ human movement of the sticks. I realise they may just not have been moved, but is that typical?
 
Unfortunately the observations above are correct. The aircraft switched to ATTI mode due to a yaw error. The winds were not excessive (10 - 15 mph above tree level) but will have caused the aircraft to drift.

2018-07-21_[14-04-03]_03.png


But your recollection appears to be incorrect - you made no inputs to the elevator or aileron sticks after takeoff:

2018-07-21_[14-04-03]_01.png


The suggestion that the RC may have been faulty is not the explanation for that, since you used the CSC to start the motors, as is visible at the start of the record:

2018-07-21_[14-04-03]_02.png


Both elevator and aileron were registering full scale at startup, indicating that they were working properly.
 
  • Like
Reactions: Harleydude
Both elevator and aileron were registering full scale at startup, indicating that they were working properly.
Good point. My question is @sar104 is why you do not see this in the .csv from Phantom Pilots? Quite honestly, I didn't see this until I used the offline converter. All of the right stick inputs in the Phantom Pilots conversion are at Zero from start to finish. Is this a sample rate thing in the conversion?
 

Members online

Forum statistics

Threads
143,066
Messages
1,467,354
Members
104,934
Latest member
jody.paugh@fullerandsons.