Spark tried to force landing. And other strange behavior.

[Giancarlo Solarino]

Last Saturday I experienced some stranges behivors on my spark.
Everything was up to date with last week's updates.(Controller FW, Drone FW, Battery and App.)



This is what I did:
1.Turned on Controller.
2. Turned on Spark.
3. Connected using wifi to controller.
4. Opened the app(Iphone6+) and changed frequency to 2.4.
5. Connected OTG cable to Phone.
6. Connected OTG cable to Controller.
7. Turn on Airplane Mode.
8. I took of from a 2th floor balcony using palm take off.(with cammera pointing at me, press power button twice and it took off.). and the activated SPORT Mode.
9. I turnned it around and flew about 340 m streight. So Far everything good.
10. I dediced to bring it back and pressed and hod RTH button on RC (Altitud was 32. and RTH altitud was configured at 30m) .
11. Drone started to came back but gainning altitude at the same time (FIRST STRANGE BEHIVOR)
12. It came back until 44m distant and 75m ALTITUDE!.
13. I cancelled RTH pressing RC button and started to bring it down on my own, full stick down.
14. Then drone enter AUTOLANDING MODE. and I couldnt cancell it. When pressed cancell the app was not doing anything.
15. Suddndly landing stopped and I was able to get it closer to me.
16. AUTOLANDING STARTED AGAIN.
17. I was able to canceled after pressing the button so many times until it did.
18. I bring it 1 m close to me and tried to land it in the front yard of the house. Full stick down and the drone descended SO SLOW. If you see the logs you'll see that descending speed was about 1km/h. It was very dificult to bring it down.
19. When finally get to the ground it did it fast and bounced a couple of time hitting the floor until it finally landed.
This was really scary and strange.
When I saw the logs I noticed that after palm launched there is a message saying 'Camera Settings Adjusted to ActiveTrack'. Which makes sense for the palm launch. And noticed that there was a FORCED LANDING message on the logs.
I didnt noticed if gesture mode got stucked for not changing it. I think the stranges things made by the drone may be related to this.
Greatfully the drone is OK. But I wanted to share this expirience in case it works for someone else or if someone knows what was the reason.
logs:
Airdata UAV - Flight Data Analysis for Drones
DJI Flight Log Viewer - PhantomHelp.com

 

[brzemia]

You started from a palm. Did the spark set homepoint at the begining ?

 

[Giancarlo Solarino]

Yes, as soon as it took of the home point was recorded.

 

[Spark 317]

Step #3 conflicting with steps #5 & 6?
If you connect with one, you shouldn't have to connect to the other? I'm thinking cable or WiFi, you choose.

 

[KNAVE]

When I take off by hand and plan to fly with controller I don't use the double power button tap method for that. I simply press and then slide take off on my phone screen with one hand and hold the drone with my other. It's not as awkward as it sounds and the drone thinks it is taking off from the ground even though I am holding it. I do the same thing to land. If you just catch the drone while it is landing it will think it hit the ground and shut down the motors.

 

[minidrohn]

Hello,
i fly with Remote+WIFI+IPAD in the evening to film the sunset with the new update and face a similair topic. There was also the problem that it came down very slowly and it sounds that is stucks on that in intervalls. It takes a long time to get the drown down to flor with different states of stucking inbetween.
Never seen that before.

Back home I will also take a look to the logs. Gestique mode was enabled but not used in this situation. So normal start from ground.
regards,
minidrohn

 

[gnuPilot]

I need to try this:

 

[KNAVE]

I wonder if dji will ever release a drone with a built in parachute or at least the option for one. Too bad 3dr changed tunes or the solo could have had one.

 

[Dronason]

Coming down very slow is probably because the vertical distance sensor reported a short or wrong distance for whatever reason. At some distance reported by it, it take priority over the barometer altitude sensor.
Maybe you can try to retrieve the .DAT file (flight log) from the Spark using DJI assistant 2. There is more info that from the app flight record.

 

[Blade Strike]

I would really recommend not flying with OTG. It's not supported and app doesn't function fully with it. As long as you turn off BT the connection to the craft is solid!

Without BT


With BT On

 

[gnuPilot]

I would really recommend not flying with OTG. It's not supported and app doesn't function fully with it. As long as you turn off BT the connection to the craft is solid!

Without BT
View attachment 776

With BT On
View attachment 777

What do you mean, what doesn’t work if you use OTG?

 

[Blade Strike]

What do you mean, what doesn’t work if you use OTG?

Apps not fully functional... update is very unreliable as well. This is unsupported method and really should be used.

 

[gnuPilot]

Apps not fully functional... update is very unreliable as well. This is unsupported method and really should be used.

What do you mean that the app isn’t fully functional? I’ve not seen anything different besides custom button assignment getting reset.

 

[Blade Strike]

What do you mean that the app isn’t fully functional? I’ve not seen anything different besides custom button assignment getting reset.

Basic things like the downlink signal strength...

 

[Dronason]

Also you cannot select which frequency band you want to use.
But all these seems to be fixable issues.
The question is if DJI will fix that and support OTG, not that there are currently some issue with existing workaround.

 

[Blade Strike]

Also you cannot select which frequency band you want to use.
But all these seems to be fixable issues.
The question is if DJI will fix that and support OTG, not that there are currently some issue with existing workaround.

I don't believe so.. the spark is not a MFI device.

 

[Dronason]

I don't believe so.. the spark is not a MFI device.

Sorry, that's what measurement shows. The spark is not involved directly it just decide which channel it will use, it is the RC which can use the 2 bands at same time, not the Spark.
At power-up, when the Spark decides for a band (based on available channels in each band I assume) one of the band is used:

1. It uses 2.4 GHz, then when the RC connect it generate Spar-RC wifi at 5.8 GHz, Spark don't show any SSID as it is in RC mode
2. It uses 5.8 GHz, then when the RC connect it generate Spar-RC wifi at 2.4 GHz, Spark don't show any SSID as it is in RC mode

From that:
1a) You are in Europe and you use android, then you are stuck as the wifi channels used by the Spark / Spark-RC are illegal in Europe (149-165) and not detected by the tablet. You need to
1b) Not in Europe or (in Europe and you use an IOS device), then the tablet can connect to these wifi channels.
2) Android and IOS tablet can connect to the RC.

What I noticed is that the RC select always a channel that is already used and at high level (by me and for 2.4G, cannot check 5.8 GHz). There would be really better channel to choose.

Using illegal channels in Europe is a big flaw of the Spark as use of OTG is almost a must for android users.
I personally purchased my first iDevice because of this and the fact that development of android version of GO4 app was fare behind IOS version.

 

[Blade Strike]

Sorry, that's what measurement shows. The spark is not involved directly it just decide which channel it will use, it is the RC which can use the 2 bands at same time, not the Spark.
At power-up, when the Spark decides for a band (based on available channels in each band I assume) one of the band is used:

1. It uses 2.4 GHz, then when the RC connect it generate Spar-RC wifi at 5.8 GHz, Spark don't show any SSID as it is in RC mode
2. It uses 5.8 GHz, then when the RC connect it generate Spar-RC wifi at 2.4 GHz, Spark don't show any SSID as it is in RC mode

From that:
1a) You are in Europe and you use android, then you are stuck as the wifi channels used by the Spark / Spark-RC are illegal in Europe (149-165) and not detected by the tablet. You need to
1b) Not in Europe or (in Europe and you use an IOS device), then the tablet can connect to these wifi channels.
2) Android and IOS tablet can connect to the RC.

What I noticed is that the RC select always a channel that is already used and at high level (by me and for 2.4G, cannot check 5.8 GHz). There would be really better channel to choose.

Using illegal channels in Europe is a big flaw of the Spark as use of OTG is almost a must for android users.
I personally purchased my first iDevice because of this and the fact that development of android version of GO4 app was fare behind IOS version.

MFI = Made for iOS..

This won't happen... it was sold as a wifi device.

 

[Dronason]

MFI = Made for iOS..

This won't happen... it was sold as a wifi device.

Sorry, I cannot follow you.
What will not happen ?
Support of OTG or they will fix issues?

And sorry, I thought that MFI = "Multi Frequency Interface".

 

[Blade Strike]

Sorry, I cannot follow you.
What will not happen ?
Support of OTG or they will fix issues?

And sorry, I thought that MFI = "Multi Frequency Interface".

No that's MiMo type technology that the spark does not have ..

Since the RC is not MFI certified by Apple it's not going to be supported over OTG for both apple or android.

By the way most 5.8 over in CE land are now only available for indoor use to my knowledge.. I could be wrong on that I'm in FCC land.