2 x 360 and crash landing - how did it happen ? (Flight Log Attached)

[alexsutherby]

HI ,

i am relatively new to flying drones , went out today on a nice open area and for some reason when i was attempting to bring the drone back to me it flipped over twice and crash landing breaking my camera off

Has anyone any ideas how this would happen ?

Flight log attached

 

[Oso]

HI ,

i am relatively new to flying drones , went out today on a nice open area and for some reason when i was attempting to bring the drone back to me it flipped over twice and crash landing breaking my camera off

Has anyone any ideas how this would happen ?

Flight log attached

Here is the log uploaded to the flight log viewer:
DJI Flight Log Viewer - PhantomHelp.com

Check out the dramatic jumps in speed in one tenth of a second. For example, at 15 seconds it goes from 4mph to 30mph in .1 seconds. At 17 seconds it goes from 14.1mph to 87.3mph in .1 seconds.

There are many examples of this throughout the flight, and they seem to be happening in exactly 1 second increments along with multiple speed errors.

It looks like the flight controller was seeing discrepant data so it ignored GPS data and put you in ATTI mode over and over. There were also magnetic interference errors along the way.

My guess is that you were trying to land in ATTI mode and the P3 was moving horizontally when it touched the ground so it flipped over.

 

[alexsutherby]

Wow 250mph is impressive

I'm not sure what I'm looking at tbh. Do you know why the speed error is there ? Or is there any indication of why it crashed ?

 

[bsartist]

Were you switching it in and out of ATTI mode? Or was it going into ATTI mode by itself.

 

[alexsutherby]

Were you switching it in and out of ATTI mode? Or was it going into ATTI mode by itself.

Must have been by itself. It was still around 20m up when it flipped over twice vertically then crashed straight down and broke the gimbal off and two propellers

 

[quaddamage]

Can you share the DAT log?

 

[alexsutherby]

Can you share the DAT log?

Hi I've attached it the post. Oso has also put a link in their comment

 

[sar104]

That is one sick aircraft, as the selected plots below demonstrate without further commentary. The telemetry suggests that there were multiple problems culminating in the FC completely locking up.

In the last 6 seconds the telemetry data all flatlines, but at strange values in some cases:

I am puzzled though - with all those warnings, why did you keep it in the air for so long?

 

[alexsutherby]

That is one sick aircraft, as the selected plots below demonstrate without further commentary. The telemetry suggests that there were multiple problems culminating in the FC completely locking up.

View attachment 98273

View attachment 98274

In the last 6 seconds the telemetry data all flatlines, but at strange values in some cases:

View attachment 98275

I am puzzled though - with all those warnings, why did you keep it in the air for so long?

i was trying to bring it home to land. Hopefully no major damage done and it's the gimbal support clip to replace. There was a lot of interference in the area too

 

[quaddamage]

Hi I've attached it the post. Oso has also put a link in their comment

I can only see the TXT log from the app, not DAT log from the Flight Controller.

 

[BudWalker]

Here is the log uploaded to the flight log viewer:
DJI Flight Log Viewer - PhantomHelp.com

Check out the dramatic jumps in speed in one tenth of a second. For example, at 15 seconds it goes from 4mph to 30mph in .1 seconds. At 17 seconds it goes from 14.1mph to 87.3mph in .1 seconds.

There are many examples of this throughout the flight, and they seem to be happening in exactly 1 second increments along with multiple speed errors.

It looks like the flight controller was seeing discrepant data so it ignored GPS data and put you in ATTI mode over and over. There were also magnetic interference errors along the way.

My guess is that you were trying to land in ATTI mode and the P3 was moving horizontally when it touched the ground so it flipped over.

Those jumps in the speed are an artifact of the converter you're using. To get more accurate results try using the TXTlogToCSVtool converter. It shows a top speed was around 39 mph.

This plot also shows that gpsHealth was only 3 (it needs to be 4 or 5) and numSats = 8. It also shows that the FC was rapidly switching between using, and not using, GPS data to compute coords. The green background is where GPS is being used and the purple background is where GPS is not being used. This rapid switching is the reason for all the weird data.

But, then why is gpsHealth < 4? For that we'll need the .DAT. @alexsutherby there is some info on this here. It'll be large so you'll need to Dropbox or GoogleDrive it.

 

[BudWalker]

....

In the last 6 seconds the telemetry data all flatlines, but at strange values in some cases:
.....

Maybe this occurred after the AC hit the ground? The .DAT would tell us more.

 

[sar104]

Maybe this occurred after the AC hit the ground? The .DAT would tell us more.

Maybe - but it did not appear to have descended, and it was still transmitting telemetry - but unphysical telemetry. It kind of looks like the FC crashed first, digitally speaking.

 

[BudWalker]

Maybe - but it did not appear to have descended, and it was still transmitting telemetry - but unphysical telemetry. It kind of looks like the FC crashed first, digitally speaking.

Could be... The FC had to be hurting with all that switching. I'm hoping @alexsutherby can provide the .DAT.

 

[Oso]

There are many examples of this throughout the flight, and they seem to be happening in exactly 1 second increments along with multiple speed errors.

Those jumps in the speed are an artifact of the converter you're using .

I assumed the erroneous speed data was due to the FC trying to reconcile the discrepant data. It never occurred to me that it may be from the flight viewer itself. Thanks for that insight.

 

[BudWalker]

I assumed the erroneous speed data was due to the FC trying to reconcile the discrepant data. It never occurred to me that it may be from the flight viewer itself. Thanks for that insight.

To be sure the speed data may not be accurate because of the GPS problems. The 39 mph values come directly from the .txt file which come directly from the AC. The 80+ mph values come from calculations done using data in the .txt file. These calcs are restricted to the 10 HZ sampling rate of the .txt file.

The SPEED_ERROR_LARGE errors that you see are the FC saying that coords data it sees don't match the coords data it expects

 

[msinger]

Those jumps in the speed are an artifact of the converter you're using. To get more accurate results try using the TXTlogToCSVtool converter.

My log viewer was showing a speed calculated by the GPS location (which is why it did not match some other converters). I just updated it to show the "Custom.HSpeed" in the "Speed" column of the main table view. The generated CSV now contains an "HSpeed" column and a "GpsSpeed" column (the old "Speed" column).

Here's the flight log re-uploaded with the above changes applied:
DJI Flight Log Viewer - PhantomHelp.com

 

[sar104]

My log viewer was showing a speed calculated by the GPS location (which is why it did not match some other converters). I just updated it to show the "Custom.HSpeed" in the "Speed" column of the main table view. The generated CSV now contains an "HSpeed" column and a "GpsSpeed" column (the old "Speed" column).

Here's the flight log re-uploaded with the above changes applied:
DJI Flight Log Viewer - PhantomHelp.com

That would certainly explain the speeds if they were calculated from the lat/long data that I posted in #8 rather than the direct IMU values.

 

[bsartist]

That would certainly explain the speeds if they were calculated from the lat/long data that I posted in #8 rather than the direct IMU values.

What about those negative altitude values?

 

[sar104]

What about those negative altitude values?

Those are clearly meaningless - the unphysical telemetry that I mentioned earlier.