Please select Into the mobile phone version | Continue to access the computer ver.
My Spark flew away! Lost my drone
12Next >
6356 72 2017-12-31
Uploading and Loding Picture ...(0/1)
o(^-^)o
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Hello Guys,First of all, a happy new year!

I wanted to do a quick flight with my Spark tonight.
My plan was to go to a height of around 100-150 meters.
I waited for the satelites to come online, until the Spark said 'Ready to GO (GPS)'.
My home point was set correctly.

She took off, and climbed to around 100-150 meters.
I let the Spark hold her position for around 3-4 minutes.
Everything worked perfectly.

But, all of the sudden i got a lot of error messages.
The GPS did not have any signal (altough it showed me 14 satelites at full strenght), and the aircraft switched to ATTI mode.
She flew away, i tried to correct her but she did not respond.
Any commands i made with my RC didn't do anything.
She just flew away in the same direction.

I could see from the ground that all 4 lights were indicating red.

She flew away, I lost signal and i could not find her back.
I lost my drone.
I tried to search in the last location, with my RC, hoping that the signal would come back, but it didn't.

I have no idea what went wrong here.
Does DJI cover my problem with their warranty?
Because i have no idea what i could have done wrong.

My flightlog: http://www.phantomhelp.com/LogViewer/W6Y4SZ6PRISTRPRGCEF8/

Kind regards,

2017-12-31
Use props
S.J
Captain
Flight distance : 322454 ft
Kuwait
Offline

happy New year 2018
2017-12-31
Use props
S.J
Captain
Flight distance : 322454 ft
Kuwait
Offline

you may have a valid case for a warranty claim
2017-12-31
Use props
S.J
Captain
Flight distance : 322454 ft
Kuwait
Offline

did you ever do a imu or compass calibration
2017-12-31
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Hi,

Yes, i did.
But I always fly my drone in the same area, but I didn't do a IMU or compass calibration before my flight today.
2017-12-31
Use props
Paul Lee
Second Officer
Flight distance : 120623 ft
United Kingdom
Offline

Possibly had to fight high winds?? I see today in Leeuwarden you are experiencing sw winds of 20mph which is consistent with the direction you Spark headed.
How was the wind yesterday and at 427ft it is going to be much much stronger ........

Current conditions in your area Wind Speed


Again current conditions.... But this is gust speed

But hopefully someone else can come along and analyze your data... This is just my thoughts....

2018-1-1
Use props
massipos
lvl.1

Italy
Offline

Sorry for what happened. You can see that the auto-landing phase has begun, and the wind pushes in the direction of wide open spaces, probably AC still intact and that is waiting for you somewhere. Keep looking in the open area.
2018-1-1
Use props
rimza
Captain
Flight distance : 3307 ft
Brunei
Offline

Which version of DJI GO 4 app do you have on the device (android or IOS)?
it is also interesting in one of the remark, "In Flight, working IMU encounters heading exception,please switch to atti mode if craft behave abnormally; Yaw Error"


when there is no option to manually switch to ATTi on spark.
2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Hi guys,

@Paul Lee , the wind yesterday was not that strong as it it now.
But there was quite some wind.
The spark did not have any trouble in the wind, until ofcourse the ATTI mode activated, and it immediately flew away with the wind. (GPS not holding it's position anymore).

@massipos We have looked yesterday in the fields for around 2/3 hours.
Also, my RC did not pick up any signal from the Spark, so i'm afraid that it's lost, or crashed in the water.

@Rimza I'm using version 4.1.20 of DJI GO 4 on my iPhone 6S Plus.
It is indeed strange that it said to manually switch to ATTI mode,  which i could not.

But in my flight log, i do see a lot of Yaw errors, what does this error mean?

Thanks for the help guys!

2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Also weird,

I just noticed that there is no date between 4m 7.3s and 4m 23.3s in my flightlog.
What could have happened in those seconds?
2018-1-1
Use props
Wachtberger
Captain
Flight distance : 261509 ft
Germany
Offline

If the wind speed information posted above is correct, this would explain most of the incident I believe. The problems started at 3m 47.4s in about 130 (!) meters altitude. I cannot say what might have caused the errors but they made the Spark switch into ATTI mode. In that high altitude the wind might have been even stronger. More experienced pilots will be able to tell you if in that situation you still had any chance to keep control. Have you made any attempt to land immediately when the error messages started? The flightlog then shows that Spark is rapidly gaining speed (probably pushed by the wind). At 4m 23.3s Spark switched into auto landing and at 4m 37.6s the flightrecord stops in about 63 meters altitude and about 470 meters away from you (if at that point the flightdata is still correct). It might have crashed into a high building. If there is a high building in the direction your Spark had taken I would start searching there.
2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Wachtberger Posted at 2018-1-1 02:51
If the wind speed information posted above is correct, this would explain most of the incident I believe. The problems started at 3m 47.4s in about 130 (!) meters altitude. I cannot say what might have caused the errors but they made the Spark switch into ATTI mode. In that high altitude the wind might have been even stronger. More experienced pilots will be able to tell you if in that situation you still had any chance to keep control. Have you made any attempt to land immediately when the error messages started? The flightlog then shows that Spark is rapidly gaining speed (probably pushed by the wind). At 4m 23.3s Spark switched into auto landing and at 4m 37.6s the flightrecord stops in about 63 meters altitude and about 470 meters away from you (if at that point the flightdata is still correct). It might have crashed into a high building. If there is a high building in the direction your Spark had taken I would start searching there.

Hi,

Immediately after the first error, i tried to get it down.
But the wind took it as soon as it entered ATTI mode.
Ofcourse the wind played a huge factor in this accident, but this accident shouldn't have happened.

I don't know why the Spark had GPS issues, while having full strength and around 17 satelites.

There are no high building, only wide open spaces, and water.
2018-1-1
Use props
JJBspark
Captain
Flight distance : 12165479 ft
  • >>>
Netherlands
Offline

But why would this Spark switch to autolanding having enough batt power left?
Mayby a mis-pushing-button on the app screen, instead of pushing the RTH  button?

Yaw errors and compass error are the worst thing having a Spark, imo its a bad luck monday morning made Spark!
Or the spark did encouter a magnetic cloud passing by.....

cheers
JJB

2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

JJBspark Posted at 2018-1-1 03:01
But why would this Spark switch to autolanding having enough batt power left?
Mayby a mis-pushing-button on the app screen, instead of pushing the RTH  button?

Hi,

I know for sure that i haven't pressed the auto landing button.
The spark entered auto landing by it self, i could not stop it.

I tried pressing the RTH button on the RC, but nothing happened and it just continued to descent, until my signal was lost.
2018-1-1
Use props
Wachtberger
Captain
Flight distance : 261509 ft
Germany
Offline

PilotWoevie Posted at 2018-1-1 03:00
Hi,

Immediately after the first error, i tried to get it down.

I really don't want to add to your pain (and I really feel with you!) but the accident would not have happened if you had not decided to fly in strong wind and in particular not at that high altitude.
2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Wachtberger Posted at 2018-1-1 03:11
I really don't want to add to your pain (and I really feel with you!) but the accident would not have happened if you had not decided to fly in strong wind and in particular not at that high altitude.

Yes, that's true.
But i also find it strange that the spark just lost all GPS connections, entered atti mode and just flew away from me without entering RTH or anything.

It didn't have any problems with the wind at altitude, the problems began when it lost it's gps and couldn't hold it's position anymore.
2018-1-1
Use props
Wachtberger
Captain
Flight distance : 261509 ft
Germany
Offline

PilotWoevie Posted at 2018-1-1 03:17
Yes, that's true.
But i also find it strange that the spark just lost all GPS connections, entered atti mode and just flew away from me without entering RTH or anything.

It did not lose the GPS connection, you can see this by the fact that there is still updated speed and distance information recorded. But it has decided to switch into ATTI mode. Which potential external interference might have triggered this to happen is the key question I believe.
2018-1-1
Use props
Paul Lee
Second Officer
Flight distance : 120623 ft
United Kingdom
Offline

PilotWoevie Posted at 2018-1-1 03:09
Hi,

I know for sure that i haven't pressed the auto landing button.

Lets say wind strength yesterday was 35mph at the height you were flying at even at top speed going forward the Spark would be blown backwards.... So my thoughts are it did indeed try to return home as your log suggests but due to it not being able to travel in the required direction due to the wind it timed out.
My knowledge of Yaw and Pitch leads me to believing that the wee Spark was almost being blown around in circles and having difficulty correcting itself again due to the high winds... This is what I think all your Yaw errors may have stemmed from.Again just my opinion......And I am no expert....Just interpreting what I am reading.
2018-1-1
Use props
Wachtberger
Captain
Flight distance : 261509 ft
Germany
Offline

In any case you should contact DJI support and make the flightrecord available to them. I believe they can see much more from it than we can.
2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Hi guys,

The drone did initiate the landing, altough it was not a return to home landing.
Just like when you hit around 10-15% of battery, it will just land, where ever it is at that moment.
That's exactly what happened to my drone, it wanted to land.

I e-mailed DJI today, but its the first day of January so i'm not sure when they will reply.
2018-1-1
Use props
Wachtberger
Captain
Flight distance : 261509 ft
Germany
Offline

PilotWoevie Posted at 2018-1-1 05:46
Hi guys,

The drone did initiate the landing, altough it was not a return to home landing.

I am optimistic that you will get a response tomorrow and cross my fingers for a good outcome!
Concerning the auto landing I believe that Spark has initiated it because RTH was not possible in ATTI mode and connection to your RC was lost. In this regard I have to maybe correct my initial analysis of your flight record. It can well be that only because of connection loss you did not receive any data anymore and the flightrecord stopped while Spark has continued its flight in landing mode. Then the question is where it might have landed.
2018-1-1
Use props
asourceunknown
lvl.3
Flight distance : 7021 ft
United States
Offline

I have another theory as to what might have happened. The yaw error might've been caused by a large gust of wind bending or damaging one of the props in some way to make the flight unstable and the drone start spinning. This would cause the compass and GPS issues and likely force the drone into ATTI mode. It would also explain why your attempts to bring the drone back home were unsuccessful. I say all this because of how the drone reacted in this YouTube video:

The uploader cut one of the sparks props to test what would happen and you can see the drone rapidly developing a spinning yaw movement and attempting to auto land. So my best guess is that wind or something else messed up one of your props.
2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

@Wachtberger Yes, but it's so hard to find it, the chance is 1 out of 1000

@asourceunknown I saw on the live feed that everything was steady, it was just drifting away.
Maybe your right, but i won't know how to find out without having my aircraft ;)

2018-1-1
Use props
jcgoobee
First Officer
Flight distance : 476119 ft
United States
Offline

asourceunknown Posted at 2018-1-1 06:03
I have another theory as to what might have happened. The yaw error might've been caused by a large gust of wind bending or damaging one of the props in some way to make the flight unstable and the drone start spinning. This would cause the compass and GPS issues and likely force the drone into ATTI mode. It would also explain why your attempts to bring the drone back home were unsuccessful. I say all this because of how the drone reacted in this YouTube video: https://www.youtube.com/watch?v=BgzjIOUpO0o

The uploader cut one of the sparks props to test what would happen and you can see the drone rapidly developing a spinning yaw movement and attempting to auto land. So my best guess is that wind or something else messed up one of your props.

This is quite interesting. I've never heard of this before. Thanks for educating me (us)
2018-1-1
Use props
jcgoobee
First Officer
Flight distance : 476119 ft
United States
Offline

PilotWoevie Posted at 2018-1-1 06:35
@Wachtberger Yes, but it's so hard to find it, the chance is 1 out of 1000

@asourceunknown I saw on the live feed that everything was steady, it was just drifting away.

According to the flight log, the AC started showing error at 3.47.4 with YAW error. Perhaps one of the motor has malfunctioned and causing the AC to yaw left and/or right uncontrollably? Yet, it does not explain why it went from GPS to ATTI mode though but that could be the first sign the machine malfunctioned started from the motor malfunction and not caused by high wind speed.

I also noticed that your take off status only showed 8 SATs, I would usually wait until I get at least 13-14 SATs before I feel comfortable of taking off. Like what Captain has said earlier, I'm also confident that DJI will get back to you after the New Year holiday. Stay tuned. Too much speculations may not help you much, so lets wait for the official feedback from them then go from there.
2018-1-1
Use props
S.J
Captain
Flight distance : 322454 ft
Kuwait
Offline

PilotWoevie Posted at 2018-1-1 03:09
Hi,

I know for sure that i haven't pressed the auto landing button.

Is there a possibility that there are other SPARK users in your area and having found a lonely SPARK tried to connect with yours !!  This is a very difficult scenario but may be it is possible that somebody stole your SPARK in midflight !
The reason for my above conculson is based on your flight log data where it says that  home point not updated  or something of that phrase .
2018-1-1
Use props
JJBspark
Captain
Flight distance : 12165479 ft
  • >>>
Netherlands
Offline

What happend ( imo) is that this Spark got a YAW error, in the flight log the heading is stable all the time so why a YAW error?? (happens more to sparks, read this forum!)
After some time a Yaw error is mostly followed by a compass error, wich kill the GPS mode and will switch to ATTI.
In the flight log you can see that the pitch and roll angles are zero, wich means that the wind is taking the drone , gone by the wind!

See my logs attached. Cheers JJB
Spark 2018-01-01a.png
Spark 2018-01-01b.png
2018-1-1
Use props
_pk
Second Officer
Flight distance : 696 ft
Poland
Offline

1) To the people who claim that Spark was not able to withstand the wind. It was able, since before the yaw error it was keeping the position.

2) This is yet another case of a yaw error followed by a fly-away. Since it happened 427ft altitude it is EXTREMELY unlikely to be caused by any magnetic interference.

3) The OP should definitely start a case with DJI support.

Just my 3 cents.
2018-1-1
Use props
Wachtberger
Captain
Flight distance : 261509 ft
Germany
Offline

JJBspark Posted at 2018-1-1 08:01
What happend ( imo) is that this Spark got a YAW error, in the flight log the heading is stable all the time so why a YAW error?? (happens more to sparks, read this forum!)
After some time a Yaw error is mostly followed by a compass error, wich kill the GPS mode and will switch to ATTI.
In the flight log you can see that the pitch and roll angles are zero, wich means that the wind is taking the drone , gone by the wind!

Your software is looking better and better! When will you release it? And do you have any theory what could be the cause of yaw errors?
2018-1-1
Use props
Tviscomi
Captain
Flight distance : 106886 ft
  • >>>
United States
Offline

S.J Posted at 2018-1-1 07:22
Is there a possibility that there are other SPARK users in your area and having found a lonely SPARK tried to connect with yours !!  This is a very difficult scenario but may be it is possible that somebody stole your SPARK in midflight !
The reason for my above conculson is based on your flight log data where it says that  home point not updated  or something of that phrase .

Yeah...that would be impossible.   We "busted" this myth over on another forum.
2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Thanks for all the answers guys.
I just came back home from searching the spark again, but no luck.
I think it’s lost forever.

What does a YAW error actually mean?
Does this have anything to do with the RC input?

About the wind, I don’t think the wind played a issue here. (other then that the spark got taken away by the wind after entering ATTI mode, but this is logical.

I’m now waiting for DJI to respond on my mail.
2018-1-1
Use props
Gunship9
Captain
United States
Offline

_pk Posted at 2018-1-1 08:33
1) To the people who claim that Spark was not able to withstand the wind. It was able, since before the yaw error it was keeping the position.

2) This is yet another case of a yaw error followed by a fly-away. Since it happened 427ft altitude it is EXTREMELY unlikely to be caused by any magnetic interference.

I am not sure it is a yaw error followed by a flyaway so much as a yaw error, ATTI mode putting the pilot in charge of flying, and the pilot failing to be able to fly a drone.

It is in ATTI mode!  Fly it back to yourself like you do all your RC airplanes and helicopters that don't have GPS station keeping.

I weep for the future since full size helicopters now have GPS locked hover ability.  If the GPS is lost, now full size helicopters will fly away out of control because professional pilots can't fly without GPS autopilot help.
2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

Gunship9 Posted at 2018-1-1 09:54
I am not sure it is a yaw error followed by a flyaway so much as a yaw error, ATTI mode putting the pilot in charge of flying, and the pilot failing to be able to fly a drone.

It is in ATTI mode!  Fly it back to yourself like you do all your RC airplanes and helicopters that don't have GPS station keeping.

I actually tried to control it, but it didn’t respond to my controls.
Remember that the aircraft entered a immediate auto landing, I couldn’t cancel it.

It’s much easier to actually tell someone to fly their drone, then actually doing it in the exact same conditions.
I have a Phantom 4 for about a year now, so I know what I was doing.
The drone did not respond to any of my commands.
2018-1-1
Use props
S.J
Captain
Flight distance : 322454 ft
Kuwait
Offline

PilotWoevie Posted at 2018-1-1 09:44
Thanks for all the answers guys.
I just came back home from searching the spark again, but no luck.
I think it’s lost forever.

Yaw error is something to do with  left and right turn of the AC housing . But this means that either you have a motor problem or a very strong  wind or it can be somebody else controlling your drone !!
Again the doubt is over the flight log data which says "Home point nöt updated  Yaw error "
2018-1-1
Use props
nixuspix
Captain
Flight distance : 2864974 ft
  • >>>
Latvia
Offline

I personally am getting more and more convinced, that among all possible differencies with Spark behaviour during flyaways, there is a big group of cases, when AC is totally out of  uncontrol in ATTI mode ( the sticks movements and/or buttons do not respond), thus making any rescue attempt unavailable
2018-1-1
Use props
Bright Spark
First Officer
Flight distance : 22129 ft
United Kingdom
Offline

A yaw error is when the cpu thinks the ac has turned , but the gyros say it hasn't.
So the heading now conflicts with the compass
input, and no heading  can be derived on where to fly. Similarly vice versa if the compass suffers interference.
BTW were there any high wind messages?
2018-1-1
Use props
PilotWoevie
lvl.4
Flight distance : 16437 ft
Netherlands
Offline

I do not recall seeing any high wind velocity messages.
2018-1-1
Use props
Gunship9
Captain
United States
Offline

PilotWoevie Posted at 2018-1-1 10:02
I actually tried to control it, but it didn’t respond to my controls.
Remember that the aircraft entered a immediate auto landing, I couldn’t cancel it.

It is easy to tell someone to fly their drone because it is common for drone pilots to say it is not following their control inputs but the flight log to show it did follow their stick movements.  It is common for pilots to not be able to handle an unstabilized nose in hover, due to reversed controls.  It is also common for pilots to have lost orientation of their drone because they flew it far enough away or into the sun so it is just a dot or black profile.  

A profile that gives no indication of which way the drone is facing.  Hard to tell the front from the back or the sides if all four motor lights are red.  So, the pilot holds the wrong stick command like grandpa standing on the car's gas when he thinks it is the brake pedal.

What did your flight log record as the stick movements?  How do they compare to the video feed from the camera?
2018-1-1
Use props
Kloo Gee
Captain
Flight distance : 16641309 ft
  • >>>
United States
Offline

Would it be possible for you to also upload the .DAT file from your phone to some sort of a file sharing service (like DropBox, GoogleDrive, or similar)?  It seems different bits of  info may be stored in different log files.  I'm trying to learn what exists where and to see if the extra information would be useful in trying to diagnose why the YAW Errors occur.  The file would be named something along the lines of:   YYYY-MM-DD_HH-MM-SS_FLYXXX.DAT and would correlate with the date/time FlightRecord you posted to PhantomHelp.  
2018-1-1
Use props
asourceunknown
lvl.3
Flight distance : 7021 ft
United States
Offline

Honestly the more I think about it the more I think these increasingly complex drones must be a nightmare to design firmware for. I am a currently a computer science student and with the number of sensors on this little drone it seems nearly impossible to account for all scenarios of failure and write code that will given the situation make the best decisions. I think I have been on the forums long enough to easily see that the firmware is not sufficiently robust to handle many edge-case scenarios and as a result we see fly-aways and crashes.

My guess is that the number of situations that the drone will automatically correct and safely land are minimal and many errors are going to be entirely up to the pilot to correct. If this is the case then the drone likely has some default modes that it enters when it detects errors in the compass, gps, VPS, etc. Such as going into ATTI mode, or something else.

My thoughts here are that with higher drone sophistication comes with the need for significantly smarter software/firmware to correct and likely learn from environmental factors and overall drone health, but at that point we're talking about running a full blow AI on the AC (maybe not practical yet).

Overall I myself plan to stay well within line of sight and fully aware of my drones orientation, you don't have to go super high to get good shots.
2018-1-1
Use props
12Next >
Advanced
You need to log in before you can reply Login | Register now

Credit Rules