Yet another "Spark Falling Out of Sky" thread
1251 9 2017-8-6
Uploading and Loding Picture ...(0/1)
o(^-^)o
mmatos
lvl.2
Flight distance : 179770 ft
United States
Offline

I posted in a prior thread however I'm finally home from my vacation so I can provide more details around my situation which seems to be somewhat different. Unfortunately DJI Support told me I could keep flying, and I didn't realize that my flight logs would get overwritten so quickly, so I do not have the data directly from the aircraft which would have included ESC temperatures, etc.

I'm attempting to see if I can find a pattern. This was all in upstate NY on a sunny day, light wind, about 90 degrees outside. (Note: My primary use is in TX, over water, with weather around 100 degrees.)

Details (click each timestamp to view flight record):

  • 11:30A: Battery #2 / intense 12 minute flight with lots of Sports Mode. Battery was at 6% at landing. Max battery temp 125 degrees.
  • 11:43A: Battery #1 / Not so intense flight where I started in Sports mode to position the aircraft, then started an active-track session around a house.
    - Start with 100% battery
    - 1m22s into flight, "Subject Lost" followed by "Mode changed to GPS_Atti". Battery: 90%, 95.2F.

    - Spark motors stop spinning and it begins tumbling to the ground. No further data is logged from this point.
    - During the fall, the motors start spinning up again, the aircraft rights itself but lands hard in some brush.
    - I still have video signal, GPS location, and attempt to tell it to take off. The aircraft reports that the angle is too great.
    - After a few minutes of searching the brush, I finally locate the aircraft. The interesting thing to note is that despite having video signal and using the GPS to locate the aircraft, none of this data is in the flight log.
  • 11:55A: Battery #1 / Test flight immediately after finding aircraft
    - Battery at 64%, 109F. While attempting to take off I noticed it's a bit bouncy. Land 30 seconds later, Battery: 61%, 110F. Restart aircraft to see if I can fix it.
  • 11:57A: Battery #1 / Easy-going flight back to the house while I walk behind it
    - Battery at 59%, 111F, Land at 39%, 116F
  • At this point I contact DJI support who says I can continue using the aircraft. I ask my daughter to help me record some .
  • After an hour of trying to replicate the issue at low altitude (running in direct sunlight, trying ActiveTrak mode, etc.) I give up. I've flown a dozen times since without issues, however I am being a bit more careful since my primary use case is and/or concrete...

Unfortunately, I didn't know at the time where the various pieces of data/types of logs lived, and have lost the most important data by now. When possible, I am attempting to have a secondary video recorded so that if it happens again, we'll have more evidence.

The summary:

  • Aircraft Firmware: V 01.00.0500
  • Remote Controller Firmware: V 01.00.0300
  • Battery Firmware: (whatever shipped with the A/C firmware above)
  • DJI GO 4 App Android:V 4.1.4
  • Aircraft Mode: ActiveTrack + Video, followed by loss of subject then crash.
  • Battery: 90%, 95F, 17 charges to date.


I lot of folks have speculated it may have something to do with the battery, however given that my aircraft remained powered on, linked to my RC with video and flight control, I don't believe this is a battery issue, at least in my case.

Personal Speculation: I think there's a bug/issue with the DJI Go app itself (bear with me) because:

  • I haven't yet seen any reports in the forum where this happened while not using the app. (In all cases, the app was being used.)
  • It doesn't make sense that a bug in the aircraft code would cause the DJI Go app to stop logging locally.


As someone who has suffered through the DJI Go app crashing frequently on a Nexus 7 device, my theory is that the DJI Go app is perhaps sending an invalid/incomplete command to the aircraft causing it to suddenly shut down. I will try to test this theory by forcing random link disconnects between the app and RC, however I'm not optimistic that I can get it to crash again (ha) since I've had a dozen successful flights since then.



2017-8-6
Use props
eylneb
lvl.4
Flight distance : 122477 ft
Malaysia
Offline

please raise a support case... since you have ur drone & its working.. better to retrive the info asap
2017-8-6
Use props
mmatos
lvl.2
Flight distance : 179770 ft
United States
Offline

eylneb Posted at 2017-8-6 18:06
please raise a support case...

Did you read the post?
2017-8-6
Use props
eylneb
lvl.4
Flight distance : 122477 ft
Malaysia
Offline

ops sorry i've missed the earlier few sentences.. cheers.. was "excited" that your drone survived.. some people that have it falling like a brick lost their drones completely hence theres no data for analysis... but yours survived but too bad it was written over..
2017-8-6
Use props
mmatos
lvl.2
Flight distance : 179770 ft
United States
Offline

After thinking about this a bit more, another theory is that there's something up with the IMU. In fact, I think this might be more likely:
  • I've noticed that ActiveTrack and Gestures are a bit jerky sometimes when losing track/recognizing the subject.
  • The crash happened right after (somehow) losing track of the house.
  • After the crash, the aircraft was a bit jerky when ascending/descending, although that seems to have cleared up.


So...now I'm wondering if temperature + the sudden jerk after losing track of the house caused the IMU to misread things, thereby shutting down the motors for a few seconds? It still doesn't explain why the app would have stopped logging anything but that could be a separate issue...
2017-8-6
Use props
Kyokushin
lvl.4
Flight distance : 296381 ft
Poland
Offline

Good analysis.
Could you also write when you bought a spark, what is the product code and production date (http://djiserialdecoder.x10host.com here you can decode s/n)?

DJI GO is buggy as hell, but for me it look as some kind of short circuit, hardware issue in ac or battery. When a short circuit is apearing then some safety fuse is turning at all.

I saw few times in a case of short circuits when devices just turned off, or made quick restart as in Your case.
2017-8-6
Use props
Enri
lvl.4
Flight distance : 683501 ft
Italy
Offline

Well I think batteries at 125 F it's over normal operation. I also live in a very hot area we are a month around 38C.
I think it's not safe to connect the second battery and fire it immediately.
Craft will cool down burning the second cold battery.

2017-8-6
Use props
fans340b9dbb
lvl.1
Flight distance : 15640 ft
South Korea
Offline

you can take log use dji assistant2 from DJI SPARK (Not DJI APP)
2017-8-6
Use props
DJI Thor
Administrator
Flight distance : 13602 ft
Offline

We are so sorry about this case, if the drone had crashed before, please check the battery slot to see if it is broken. And if a Spark crashed, please export the flight data on the drone via DJI Assistant 2 first if it is possible in case it is overwritten. Besides, we recommend sending the drone in for repair if it is necessary.
Here's the tutorial for exporting the data:
2017-8-7
Use props
hallmark007
Captain
Flight distance : 9827923 ft
  • >>>
Ireland
Offline

Not to sure why you wouldn't have your logs, dji go stores up to 50 flights, you could have also synced your flights to dji cloud.
Anyways there is always dat file logs in your black box of aircraft. Videos don't tell a lot, in the second one you say jumpy but it sounds like you push left stick full forward and then let go then full forward again if you play it back you can hear double revving, if you raise AC in one single motion of just pushing left stick forward you won't get this jerky move, unless you have a problem with IMU.
It maybe wise to send craft back for analysis, your theory that there is a glitch in the app may or may not be correct but might need some checking out.

You mention AC going into gps Atti mode this is just normal gps mode.
2017-8-7
Use props
Advanced
You need to log in before you can reply Login | Register now

Credit Rules