Hi Dji,


Yesterday I had a my first random flyaway on my p3a which lead to a bad crash. have not had a  single issue with it previously. always maintained and pre flights done,  compass cal, i had 12 sats for gps where i was flying minimum wind, clear skys and nothing out of the ordainairy. abit about the flight, i took off on an oval had for about 15 seconds i had control. flying at aprox 30m it became speratic and i lost all control of the aircraft. it flew in line of sight still incressing speed very quickly banking hard to one side, to my eyes it look to lift up high and invert over itself, and include a few rotations before it decended crashing into trees and into a off road. upon getting to the crash site i noted extreme damage to the top shell, bottom shell, camera gimble have all broken. camera and battery was completely removed from phantom. flying where i was on an oval i did not record the flight as my flight area would only have included the oval with direct line of sight, I have retrieved the flight data from the p3a, i dont believe that i was incontrol of the dji flight software/hardware at all after the initial 15seconds and look forward to how to proceed from here.

Rob

also to add. Controller was in gps mode and all firmware upto date

http://www.phantomhelp.com/logviewer/upload/           Let everyone take a looksee .

http://www.phantomhelp.com/logviewer/1KHLXDP6WVAAW39XHSYP/

The first thing I noticed was you took off with a battery at 47%, which is dangerous. You should always start a flight with a freshly charged, fully charged battery. Almost immediately the battery started to show low voltage on some of the cells. I also notice you were actually flying in ATTI mode, or fully manual. You may have thought youy were flying in P-GPS mode, but rthe record says not.

Wow 101 mph. Doesn't seem it had control.

i had previously used a percentage in the small flight minutes before this flight, it was approx 60% i reiterate the controller was in gps mode when i took off. i had not switched the controller to any other mode.  ive flown over 50 hours with this drone and this is the first time i have encountered this behavior. my plot for this flight was to fly center of the oval and figure 8 around the posts and land. the drones speed and distance from my designated flight area are clearly a miss.

There are a number of things in the recorded data that are unusual and I'm not sure which data is accurate and which is false.
I can't tell what you did with the joysticks from the data - that's just garbage numbers and I doubt the speed and altitude data.
What can you say about the battery?
How long since it was fully charged?
Was it showing 47% at the start of the flight - is that number believable

Labroides Posted at 2018-6-3 18:59
There are a number of things in the recorded data that are unusual and I'm not sure which data is accurate and which is false.
I can't tell what you did with the joysticks from the data - that's just garbage numbers and I doubt the speed and altitude data.
What can you say about the battery?

the battery was charged to 100% the day before flight. ive not noticed anything different from this battery to any of my other battery's. all dji official. never dropped etc. as soon as i realized i was not in control of the drone mid oval i release my hands from the joysticks as i am under the impression that it would come to a complete stop and not continue to move about. minus slight stopping distance. on app i do recall the battery being accurate to data outputted.

Mr_Rob1322 Posted at 2018-6-3 19:05
the battery was charged to 100% the day before flight. ive not noticed anything different from this battery to any of my other battery's. all dji official. never dropped etc. as soon as i realized i was not in control of the drone mid oval i release my hands from the joysticks as i am under the impression that it would come to a complete stop and not continue to move about. minus slight stopping distance. on app i do recall the battery being accurate to data outputted.

If the battery was freshly charged, that indicates the battery data is also unreliable.
It's lots of crazy numbers.
There is almost nothing in that flight data that I would trust.
I can only guess that if the flight controller was being fed crazy numbers then it's going to do crazy things.
I'd want to have that drone checked over properly before flying again.

Labroides Posted at 2018-6-3 19:15
If the battery was freshly charged, that indicates the battery data is also unreliable.
It's lots of crazy numbers.
There is almost nothing in that flight data that I would trust.

it was fully charged. it had also seen a smaller flight prior to this one. hence not being at capacity, as stated, the drone is completely un fly able. where do i go from here ?

Mr_Rob1322 Posted at 2018-6-3 19:32
it was fully charged. it had also seen a smaller flight prior to this one. hence not being at capacity, as stated, the drone is completely un fly able. where do i go from here ?

So it wasn't fully charged but had a previous flight that used part of the battery charge?
I still don't believe some of the battery data.
The cell voltage falling to 2.7V in 16 seconds is crazy if true.
The Phantom would have stopped flying before the cell voltage reached 3.0V.
The joystick data shows the joysticks going to positions that are off the scale, the altitude goes to -400 ft on generally flattish terrain, the speed is showing up to 90mph.
Given that all of these fields are full of crazy numbers, I wouldn't even be sure that the atti/P-GPS data is correct.

There were some stray electrons running around inside that Phantom and I can't make anything out of the data.

Was it under warranty?

Labroides Posted at 2018-6-3 19:54
So it wasn't fully charged but had a previous flight that used part of the battery charge?
I still don't believe some of the battery data.
The cell voltage falling to 2.7V in 16 seconds is crazy if true.

yes thats correct. i believe about 8ish mins of standard hover flight.
im also unsure of why the data shows some of the data it does, but it hasnt been altered.
i believe that the gps in the drone could have been faulty and caused the flyaway and extreme damage rendering it un fly able.

Bashy Posted at 2018-6-3 19:59
Was it under warranty?

no i dont believe so.

Oh dear, as to how to proceed then would be to buy another one

Mr_Rob1322 Posted at 2018-6-3 20:54
yes thats correct. i believe about 8ish mins of standard hover flight.
im also unsure of why the data shows some of the data it does, but it hasnt been altered.
i believe that the gps in the drone could have been faulty and caused the flyaway and extreme damage rendering it un fly able.

A GPS unit doesn't have any influence on altitude data or cell voltage data or joystick data but all of these had unbelievable numbers.
I would suspect the problem comes from an electronic fault somewhere around the flight controller and it's scrambling all the data such that the flight controller couldn't function normally.

Hi Rob, we're sorry to read your post. The data analysis is only available when the drone is still under warranty period. If you need to repair the drone, please kindly start a case to send the unit in: https://www.dji.com/support

Did the flight path follow the track reported on the map on the PhantomHelp page?  If so, it would seem the GPS was functioning and that data should be reliable.   For the aircraft to invert and tumble something severe had to happen, e.g. birdstrike, tree branches, thrown or broken prop.  It would have to have been blowing a gale to have overcome the aircraft's natural tendency to remain upright and tip it over, something I dare say you would have noticed.  When you recovered the wreckage, were any of the props missing or snapped?  My guess based on the clear terrain was that something happened to a prop which would cause the rise and roll-over you described.  Even though you weren't recording to your micro SD card, it's possible the low resolution FPV video might still be present on your device and could be downloaded for further clues.  Any photos of the damage you can share?

The flight data shows you were using Litchi for flight control.  Any chance you also had DJI Go open in the background?

The stick movements recorded (after transposing the data in the log to the correct stick function positions) roughly relate to your movements initially, but there seems to be something else happening on top of that.  It's like there are additional forces from a high wind causing it to rotate and descend uncontrollably (which might be feasible in ATTI mode), or it is receiving additional "phantom" signals from a second app running.  I can't explain the reported 100+mph speeds.

endotherm Posted at 2018-6-4 02:49
Did the flight path follow the track reported on the map on the PhantomHelp page?  If so, it would seem the GPS was functioning and that data should be reliable.   For the aircraft to invert and tumble something severe had to happen, e.g. birdstrike, tree branches, thrown or broken prop.  It would have to have been blowing a gale to have overcome the aircraft's natural tendency to remain upright and tip it over, something I dare say you would have noticed.  When you recovered the wreckage, were any of the props missing or snapped?  My guess based on the clear terrain was that something happened to a prop which would cause the rise and roll-over you described.  Even though you weren't recording to your micro SD card, it's possible the low resolution FPV video might still be present on your device and could be downloaded for further clues.  Any photos of the damage you can share?

The flight data shows you were using Litchi for flight control.  Any chance you also had DJI Go open in the background?

flight path was fairly accurate to what i saw. i did not see a bird strike nor did it encounter the trees until there was no control. props were intact and only slight contact damage, not sure how to retrieve any fpv footage. im certain i only record to drone onboard. photos incoming when im home. litchi has only been installed on this device so i can be certain that it dji software was not to be mixed up with the litchi software. wind speeds were averaging on 12-16kmph on the day and timeframe flown. looks like ill open a repair ticket and go from here such a shame and costly

Sorry to hear of this incident.   I think we all dread this possibility.

A long shot but in the absence of other explanations.....
Were you in an isolated area?   Is it possible there could have been law enforcement or military personnel anywhere in the vicinity?
A large number of police departments and military bases have acquired anti-drone devices and it stands to reason that there would be some wanting to test them on a live subject.
However, I was under the impression that they would cause a drone to land, not lose control....but it's probably experimental.
A long shot...but more and more of these anti-drone devices are filtering into local hands.

Another thing that can cause this is to be in the line of a microwave transmission tower.   A sudden burst of high energy microwaves while your Phantom is within a mile or so of a microwave transmission tower can cause similar results.   I always carefully check for the location of "cell towers" and ANY transmission towers before flight.

All said, I would say your flight with a non fully charged battery was the most likely culprit as others already mentioned.   Even a slight drop in current flow from the battery which could even be be caused by poor battery contact can cause this to happen.   More probable on a non fully charged battery.  I prevent those types of current drop outs by using Noalox on the battery contacts.  Never had a problem so far (knock on wood).


Looks like you'll be buying a replacement unfortunately.

Damage after being disassembled includes: Props, Legs, Yaw arm, Flex Cable, Gimble Vibration Plate, Dampening Rods, Stickers, Top and Bottom Shell, battery compartment thing, and weirdly enough the rubber usb protector. the above ordered inc screws for piece of mind. including postage damage has so far cost shy of $300

Cessna172 Posted at 2018-6-5 04:26
Sorry to hear of this incident.   I think we all dread this possibility.

A long shot but in the absence of other explanations.....

'A large number of police departments and military bases have acquired anti-drone devices and it stands to reason that there would be some wanting to test them on a live subject.'

He is in Australia, in a purely urban area. Neither the police or military here have or use those devices, they would not be permitted in our country.

Cessna172 Posted at 2018-6-5 04:26
Sorry to hear of this incident.   I think we all dread this possibility.

A long shot but in the absence of other explanations.....

None of the above
It's always better to look for believable and likely causes before randomly suggesting the unlikely or impossible.
A large number of police departments and military bases have acquired anti-drone devices and it stands to reason that there would be some wanting to test them on a live subject.
The whole flight was over in 16 seconds and the recorded data was looking dubious from 3 seconds.
There isn't that level of drone paranoia in Australia, drones aren't assumed to be a threat.

I'm pretty sure that the police in a quiet Australian country town aren't equipped for that and wouldn't be that trigger happy anyway.
I think we can discount that particular idea.

Another thing that can cause this is to be in the line of a microwave transmission tower.   A sudden burst of high energy microwaves while your Phantom is within a mile or so of a microwave transmission tower can cause similar results.
Nothing like that in the vicinity but they wouldn't have the effect that was observed anyway.

I would say your flight with a non fully charged battery was the most likely culprit as others already mentioned.
And a partially charged battery wouldn't do that either but the OP has confirmed that the battery was properly charged.

Mr_Rob1322 Posted at 2018-6-4 18:38
flight path was fairly accurate to what i saw. i did not see a bird strike nor did it encounter the trees until there was no control. props were intact and only slight contact damage, not sure how to retrieve any fpv footage. im certain i only record to drone onboard. photos incoming when im home. litchi has only been installed on this device so i can be certain that it dji software was not to be mixed up with the litchi software. wind speeds were averaging on 12-16kmph on the day and timeframe flown. looks like ill open a repair ticket and go from here  such a shame and costly

The FPV footage is usually cached in a folder in your app directory on your device.  Instructions will vary.  I'm not entirely sure how Litchi handles it or if it caches at all, but we can try.  What device are you using?  iOS or Android?  PC or Mac?

You mentioned "stickers" before.  Are you referring to the small ones that come on a standard body or have you applied aftermarket skins to the top of the shell, potentially upsetting GPS reception?

None of the above
It's always better to look for believable and likely causes before randomly suggesting the unlikely or impossible.
The whole flight was over in 16 seconds and the recorded data was looking dubious from 3 seconds.
There isn't that level of drone paranoia in Australia, drones aren't assumed to be a threat.

I'm pretty sure that the police in a quiet Australian country town aren't equipped for that and wouldn't be that trigger happy anyway.
I think we can discount that particular idea.

In all fairness, I missed his location......my bad.  I mistakenly thought he was in the USA where I have read many stories of anti-drone devices being developed and sold.

Nothing like that in the vicinity but they wouldn't have the effect that was observed anyway.

Unless you have an advanced degree in Electromagnetic Theory, I'm not sure you're qualified to make that conclusion.
Not only have many people reported incidences similar due to that type of interference, I have had it happen to me personally.
That you would say that indicates you don't have much actual flying experience with drones.
I am not saying "THIS is what happened"....just offering possibilities.
Why not let others decide for themselves which information is useful and which is not rather than spending our time attempting to destroy what others post?



And a partially charged battery wouldn't do that either but the OP has confirmed that the battery was properly charged.

You are wrong.  Go back and read Post #7

Cessna172 Posted at 2018-6-6 08:07
None of the above
It's always better to look for believable and likely causes before randomly suggesting the unlikely or impossible.
The whole flight was over in 16 seconds and the recorded data was looking dubious from 3 seconds.

Based on his profile under his avatar, it seems he has flown a very large distance and with multiple DJI products.  EM disturbances may upset telemetry transmission or control signals, but wouldn't cause the mess of data recorded, including altitude and speed mis-readings.  It's highly unlikely to be caused by a microwave burst.

With regards to the battery, read post #14:
So it wasn't fully charged but had a previous flight that used part of the battery charge?
yes thats correct. i believe about 8ish mins of standard hover flight.

Battery levels haven't been reported for recording bad data nor taking over flight controls.  Flying on partially depleted batteries that have been stored for some time are responsible for calculating bad capacity values and unpredictable shutdowns.  This does not apply to a short hop taken immediately before the current flight.  It only seems to happen if the used battery has enough time to completely cool down and stabilize, perhaps at least overnight.

Those are some crazy extreme numbers in the "flight log". Minus -439 feet altitude and high 101.3 mph velocity.

Cessna172 Posted at 2018-6-6 08:07
None of the above
It's always better to look for believable and likely causes before randomly suggesting the unlikely or impossible.
The whole flight was over in 16 seconds and the recorded data was looking dubious from 3 seconds.

Not only have many people reported incidences similar due to that type of interference, I have had it happen to me personally.
Many people report flying saucers and even being abducted by space aliens too.
Sometimes the real explanation behind what people think they observe is quite different.
Just because someone thinks that a cell tower did things to their Phantom, doesn't mean a thing without confirming data.
There are people living in my area that claim all kinds of crazy things about cell towers.  

That you would say that indicates you don't have much actual flying experience with drones.
Perhaps you believe that but it doesn't mean it's true.

I am not saying "THIS is what happened"....just offering possibilities.
Why not let others decide for themselves which information is useful and which is not rather than spending our time attempting to destroy what others post?
What you were throwing out there as "possibilities" just weren't possible as they don't match the data and don't match what happens in a Phantom.
Forum members are welcome to decide which information is useful and I hope that I've provided something which really is as it was based on actual knowledge of how Phantoms work and the actual data from  this flight rather than just suggesting maybe it was an asteroid, maybe  space aliens used a tractor beam etc.

And a partially charged battery wouldn't do that either but the OP has confirmed that the battery was properly charged.
You are wrong.  Go back and read Post #7
You mean where he says .. i had previously used a percentage in the small flight minutes before this flight
He had a fully charged battery, flew a bit, landed and shortly after flew a bit more.
That's not an issue and is just like flying one flight on one battery.
But even if he had a battery that had been sitting around for a few weeks discharging, it still wouldn't do what is showing in the actual data.
It wouldn't provide crazy speed, altitude and battery data.
It would simply show a rapidly falling voltage with the other data behaving normally and the Phantom autolanding because of a critically low battery level.