Welcome DJI Spark Pilot!
Jump in and join our free Spark community today!
Sign up

Spark goes haywire, again!

dropro

Member
Join
Jun 12, 2017
Messages
18
Age
53
I recently posted that my spark went out of control.
I decided to try it again, in a controlled environment.
First, I downloaded and installed the latest firmware. Then I made sure I had a full battery.
Took the spark to an open place, on top of a mountain.
Turned it on, waited for it to pair with my phone. Waited until it had home point, multiple satellites and ready to fly gps. No calibration of compass or IMU required. All ok.
This time, I decided to start it from the ground. Aka no gestures. Used slide to launch. Lifted it in air about 8ft from ground. Hovering in place and waited, checked that all was still ok.
Then I slowly started to yaw. It yawed and then it took off at full speed, just like last time!!!!

Tried to land it, but it was not responding to any controls.
It behaved EXACTLY like in the first time: did a semi circle while losing height.
I still ran after it but waited to see what it would do. It flew through trees, scraping them, hitting branches. When it finished its semi circle it came to a stop.. I did not have to catch it and injure my fingers. Put my hands underneath it, caught it and turned it off.

Conclusions:
1. I must have a defective Spark
2. Defective Sparks are really dangerous.
3. When spark goes haywire, it seems obstacle avoidance is not working
4. When spark goes haywire, phone is useless in trying to control it.
5. I now officialy HATE my Spark

These two incidents have completely blown my confidence in these. It's not fun anymore.
Based on these experiences, as soon as I get back home, this drone is going back to DJI.
It will take a lot to convince me they are safe. So sad.
 
I recently posted that my spark went out of control.
I decided to try it again, in a controlled environment.
First, I downloaded and installed the latest firmware. Then I made sure I had a full battery.
Took the spark to an open place, on top of a mountain.
Turned it on, waited for it to pair with my phone. Waited until it had home point, multiple satellites and ready to fly gps. No calibration of compass or IMU required. All ok.
This time, I decided to start it from the ground. Aka no gestures. Used slide to launch. Lifted it in air about 8ft from ground. Hovering in place and waited, checked that all was still ok.
Then I slowly started to yaw. It yawed and then it took off at full speed, just like last time!!!!

Tried to land it, but it was not responding to any controls.
It behaved EXACTLY like in the first time: did a semi circle while losing height.
I still ran after it but waited to see what it would do. It flew through trees, scraping them, hitting branches. When it finished its semi circle it came to a stop.. I did not have to catch it and injure my fingers. Put my hands underneath it, caught it and turned it off.

Conclusions:
1. I must have a defective Spark
2. Defective Sparks are really dangerous.
3. When spark goes haywire, it seems obstacle avoidance is not working
4. When spark goes haywire, phone is useless in trying to control it.
5. I now officialy HATE my Spark

These two incidents have completely blown my confidence in these. It's not fun anymore.
Based on these experiences, as soon as I get back home, this drone is going back to DJI.
It will take a lot to convince me they are safe. So sad.

Please report how DJI responds.
 
Check out your TXT flight log to see if it explains what happened. You can upload and view it online here. If you'd like other people to review and comment on your flight log, then please post a link back here after you upload it.
 
  • Like
Reactions: Jagerbomb52
Maybe check your sensors before flying..
Where did you last do your compass dance??
Did you ever crash before?? Usually after a crash, an IMU calibration may be needed.
 
Check out your TXT flight log to see if it explains what happened. You can upload and view it online here. If you'd like other people to review and comment on your flight log, then please post a link back here after you upload it.
Would be glad to get your comments on the flight log.

I looked at it and saw that during a certain portion of the flight it has a compass error, during another it has a speed error, but it does not have any of these at the beginning of the flight and there were no metals to interfere in the area. The strange thing is the flight path it follows. I don't know if this viewer can report user input. I know I tried to get it to land as soon as it took off- but I cant see this in the logs.
 
Your flight log is showing compass errors. That most likely means one of the following:

1) The compass was last calibrated near a magnetic metal source. It needs to be recalibrated.
2) The Spark was near a source of magnetic metal when you took off.
3) The compass was near a strong magnet (e.g. a subwoofer in a car) and needs to be degaussed.
4) The compass is faulty and needs to be replaced.
 
  • Like
Reactions: Mike and TavisB
Your flight log is showing compass errors. That most likely means one of the following:

1) The compass was last calibrated near a magnetic metal source. It needs to be recalibrated.
2) The Spark was near a source of magnetic metal when you took off.
3) The compass was near a strong magnet (e.g. a subwoofer in a car) and needs to be degaussed.
4) The compass is faulty and needs to be replaced.
Thanks for that subwoofer comment, that got my attention instantly! Does this mean our compasses are subject to corruption when the aircraft is not powered up? I was unaware of that.
 
Does this mean our compasses are subject to corruption when the aircraft is not powered up?
It means the compass could be magnetized by a strong magnetic source. It's not a common problem, but it does happen sometimes.
 
  • Like
Reactions: Nikon1
It means the compass could be magnetized by a strong magnetic source. It's not a common problem, but it does happen sometimes.

Wow - That's Great information for me. I was figuring on the Spark living in my hatch - in which also resides a 12" AlumaPro SubWoofer. I may have to modify the GPC case with shielding when it comes.

Thanks for that insight, Mike!
 
Your flight log is showing compass errors. That most likely means one of the following:

1) The compass was last calibrated near a magnetic metal source. It needs to be recalibrated.
2) The Spark was near a source of magnetic metal when you took off.
3) The compass was near a strong magnet (e.g. a subwoofer in a car) and needs to be degaussed.
4) The compass is faulty and needs to be replaced.
I am very familiar with compass issues, however the logs do not explain the following:
1. Why did it take off?
2. Why was it moving when it did not have compass errors? See all the logs...
3. Without user input and even with compass errors it must hover based on the VPS, especially since it was at low altitude. But it didn't...

And other question: is there any other data that we can glean from this log?
User input?
Why was obstacle avoidance not used? See how it cut through the trees?

To me- this drone is just not reliable this way. But we have to figure what's wrong so dji can fix it.
 
Those are great questions -- most of which would be difficult to answer. Let's start with the things that are easily fixed first:
  • Since this incident, did you try calibrating your compass again in a location you know is free of all magnetic metal sources? If not, try that and see if you're able to reproduce that erratic flight.

  • Has your Spark even been near any strong magnetic sources (e.g. a subwoofer in a car)?
 
Those are great questions -- most of which would be difficult to answer. Let's start with the things that are easily fixed first:
  • Since this incident, did you try calibrating your compass again in a location you know is free of all magnetic metal sources? If not, try that and see if you're able to reproduce that erratic flight.

  • Has your Spark even been near any strong magnetic sources (e.g. a subwoofer in a car)?
First, the flight was not erratic. It was unexpected and uninitiated but took a nice curve path.

I had calibrated the spark before the first incident in Berlin. So it was calibrated just before the first flyaway. In both cases it was open space without any metal or magnets in the area and it was not next to a subwoofer.

compass calibration is not (should not) be voodoo magic. If the compass has a bad calibration, then you will get notice of it as you turn it on. This is being checked before every flight. But that was not my case.

On the other hand, compass calibration can be affected by magnetic fields while in flight. Again, if you look at the path of my flight - that was not the case- flying in open place with just some trees. As a matter of fact, very similar trajectory in both flights.

Finally, in both cases it also lost altitude while curving. That can not be caused by compass issues - I think...

There is much more going on here... unfortunately....
 
If you are sure you have a good compass calibration, then you should return your Spark and get a replacement. If it was my Spark, I'd trying calibrating the compass again before going through that trouble.
 
  • Like
Reactions: turbodronepilot
If you are sure you have a good compass calibration, then you should return your Spark and get a replacement. If it was my Spark, I'd trying calibrating the compass again before going through that trouble.

Absolutely - will redo the calibration and retest.
Do you know how to extract user input from the logs?
Also, how do I submit the log to DJI for evaluation?
 
Do you know how to extract user input from the logs?
You can find the recorded stick inputs in the CSV file that's linked on my log viewer after uploading a TXT flight log.

how do I submit the log to DJI for evaluation?
I'm not sure if the Spark logs flight data internally. If it does, there's seems to be no way of accessing it from the DJI Assistant 2 app.
 
You can find the recorded stick inputs in the CSV file that's linked on my log viewer after uploading a TXT flight log.
I did not see any- are you sure it logs these for the spark?

I'm not sure if the Spark logs flight data internally. If it does, there's seems to be no way of accessing it from the DJI Assistant 2 app.
I was able to export data but it is complete gibberish to me.
 
You can find the recorded stick inputs in the CSV file that's linked on my log viewer after uploading a TXT flight log.


I'm not sure if the Spark logs flight data internally. If it does, there's seems to be no way of accessing it from the DJI Assistant 2 app.
I saw something on the DJI forum about the "BLACK BOX" option. I think it's the flight recorder that DJI now request if users have an error.
 
I did not see any- are you sure it logs these for the spark?
You're right -- the CSV has no data for the remote controller inputs. The data is either being stored elsewhere in the TXT flight log or DJI GO is not logging it. I did not attempt to dig through the DJI GO source code to verify.

I was able to export data but it is complete gibberish to me.
You were able to export the DAT flight logs?
 
You're right -- the CSV has no data for the remote controller inputs. The data is either being stored elsewhere in the TXT flight log or DJI GO is not logging it. I did not attempt to dig through the DJI GO source code to verify.


You were able to export the DAT flight logs?
I was able to export files - but they are not the DAT files. There is a folder called panic but it is empty. There are other numbered files but they are rather small- don't know what their use is.

In any case, I used the DJI Assistant to send a copy of the TXT files to DJI. Did this 3 days ago along with an incident report. So far, they did not even acknowledge receiving my files...
 
I recently posted that my spark went out of control.
I decided to try it again, in a controlled environment.
First, I downloaded and installed the latest firmware. Then I made sure I had a full battery.
Took the spark to an open place, on top of a mountain.
Turned it on, waited for it to pair with my phone. Waited until it had home point, multiple satellites and ready to fly gps. No calibration of compass or IMU required. All ok.
This time, I decided to start it from the ground. Aka no gestures. Used slide to launch. Lifted it in air about 8ft from ground. Hovering in place and waited, checked that all was still ok.
Then I slowly started to yaw. It yawed and then it took off at full speed, just like last time!!!!

Tried to land it, but it was not responding to any controls.
It behaved EXACTLY like in the first time: did a semi circle while losing height.
I still ran after it but waited to see what it would do. It flew through trees, scraping them, hitting branches. When it finished its semi circle it came to a stop.. I did not have to catch it and injure my fingers. Put my hands underneath it, caught it and turned it off.

Conclusions:
1. I must have a defective Spark
2. Defective Sparks are really dangerous.
3. When spark goes haywire, it seems obstacle avoidance is not working
4. When spark goes haywire, phone is useless in trying to control it.
5. I now officialy HATE my Spark

These two incidents have completely blown my confidence in these. It's not fun anymore.
Based on these experiences, as soon as I get back home, this drone is going back to DJI.
It will take a lot to convince me they are safe. So sad.
Interesting about the curved flight path. When I had my flyaway, it flew in a large arc, but in my case, it ascended, not descended.
 

Members online

No members online now.

Forum statistics

Threads
14,593
Messages
118,799
Members
17,987
Latest member
csdisme