Welcome.
I had a terrible crash today with Dji Phantom 3 Professional.

I was flying over my village making video with about 10m/s speed in P mode.
Then suddenly my Phantom tuned upside down and made a free fall from about 120m.
The hit was that hard that gimball with camera broke off and flew 10meters from the crash place.
Battery was charged for 91% before crash.
GPS signal was very good, i have done compass and imu callibration before flight.
All firmwares and DJI app was up to date.

I was so scared after this event, but luckily no one was injured.
Certainly there was an issue with the drone, because it it impossible to flip the drone manually.
I am very disappointed with DJI product.
I am sure that there was an issue with hardware or there are some underdevelopements with DJI firmware.
For 1200$ product i was expecting high quality product with high level of refinement in terms of hardware and firmware.

Here is a video from the crash, i can provide flight log as well in txt and dat format.Please give me an advice what should i do now?

I'm sorry for your crash, what is your email and I will look into the flight.

Thank You for the reply. My email is razor[at]razor.pro

About an hour ago, I had crash with my DJI Phantom 3 Advanced, too.
I was flying in P mode, drone was about 10 m away from me. Then I switched mode to F(follow me). First moment everything was ok, then suddenly Phantom turned away from me and started to fly in different direction with high speed. I tried to stop it, but it was not responding, altough DJI app was showing full video and transmitter range. Drone hit the building and fell on the ground. Gimball and camera were broken and fell off somewhere (I think they are on the roof of the building). Also, one leg and propeller are broken.
DJI app showed no warnings. GPS connection was great(about 20 satellites). Battery was showing 50% before crash. Firmeware and DJI app were up to date.
I had no movie from the crash, because my sd card was full. but I can provide flight log from DJI app.
What should I do now?
I think it was bad day for DJI drones

romanoitalino Posted at 2016-8-5 08:16
About an hour ago, I had crash with my DJI Phantom 3 Advanced, too.
I was flying in P mode, drone wa ...

First off, it is not polite to hijack someone elses' thread, you should start your own. In any event, the Follow Me function relies upon the phone/tablet you have connected to your RC  unit providing a strong and accurate GPS location for the aircraft to follow. The fact that it was showing 20 satellites has nothing to do with the issue, because those are the satellites the aircraft is seeing, not your phone.


It sounds very much like your phone/tablet gave the aircraft an erroneous reading to lock onto and the aircraft went after it. This is quite common, as the GPS receiver in most phones is pretty crappy and often not accurate.

Were you looking at the drone? from the video id say a bird (Eagle or Hawk) got it. Happened to a mate of mine and looked a lot like that!

@Geebax i was using Galaxy S5, as recommended on official DJI website. While takeoff the drone saw about 15 satellites. @brucey791 of course, there were 2 persons with me and they were watching the drone all the time.
update:
Sorry for incomprehension, i thought that Your answers were directed to me.
@romanoitaliano im sorry for your loss, but please make a separate thread.

romanoitalino Posted at 2016-8-5 06:16
About an hour ago, I had crash with my DJI Phantom 3 Advanced, too.
I was flying in P mode, drone wa ...

Sorry for your loss, romanoitalino. Please get in touch with support.eu@dji.com and then send it in for diagnosis and data analysis. You can sync all your flight records before contact our tech support.

Either a bird snap or a propeller got loose, I hope it didn't cause any further damage upon hitting the ground since you were flying over (your) neighborhood..

razor@razor.pro Posted at 2016-8-5 01:11
@Geebax i was using Galaxy S5, as recommended on official DJI website. While takeoff the drone saw a ...

sorry for stealing your thread

DJI-Paladin Posted at 2016-8-5 03:15
Sorry for your loss, romanoitalino. Please get in touch with  and then send it in for diagnosis an ...

thank you for your response. I will send an email with all the data

Hi razor - my theory ...

According to your altitude and speed graph, there was a sudden increase of altitude by 100-feet at approx 12:03:30 followed by an increase in horizontal speed at around 12:03:50 ... I see no evidence of either during your controlled flight segment in the video, therefore this must of happened after the drone inverted.

With the drone inverted, whether with all props in place or missing one, there is no way it's going to climb and accelerate horizontally, therefore an outside force must have been in play.

My best guess would be a raptor strike from below where it grabbed a landing skid and began lifting it away from the strike point.  The subsequent horizontal speed increase would be consistent with the raptor then starting to fly off with it in near horizontal flight.

At roughly 12:04:25 where both speed and altitude plummet (no pun intended) it would appear to be the descent after the raptor dropped the drone having determined it was non-edible and not a threat.

It's a shame the video terminated before ground impact as it may have borne out my theory above, and may have had a glimpse of the raptor flying away from the falling drone.

@Old Geezer this altitude raise was controlled by me while starting from my yard i have starting to make video while the drone was above my yard at certain height. The low altitude was keeped low long before Flying high because i was Flying over my yard near me for a long Time.

DJI-Ken Posted at 2016-8-4 22:56
I'm sorry for your crash, what is your email and I will look into the flight.

Dear Ken, a im still waiting for Your contact information, so i can send You all necessary data. I had already gave You my email address.

razor@razor.pro Posted at 2016-8-6 00:32
Dear Ken, a im still waiting for Your contact information, so i can send You all necessary data. I ...

That email does not have any flight logs associated with it, can you sync your flight records.
Also, I am only tech support and I am in the US. You need to contact support and start an ticket.
I was only offering assistance to look at the flight to see if I saw anything unusual/

razor@razor.pro Posted at 2016-8-5 14:18
@Old Geezer this altitude raise was controlled by me while starting from my yard i have starting to  ...

OK - thanks for that clarification - it makes a lot more sense of the graph now.

Finally!  Someone else who had a Phantom flip over and crash!  And I have been led to believe from DJI that it was all my fault and no one else ever heard of such.  I have been fighting DJI for over a month now trying to get someone to provide me with the flight data logs but they refuse, claiming no logs exist!!!!!  My Phantom 4 crashed from some 200 feet upside down.  I have all the proof that I am correct but DJI will not acknowledge.  Now I have someone else with the same issue.  
Thanks,
Jim
WA5TEF

I wonder if the gimbal broke and flew in to the props. Suddenly I see what looks like one of the arms is in view then crashing. It could have been a bird strike from below that pushed the gimbal over and in to the prop. My guess is the gimbal broke and caused the accident.

WA5TEF Posted at 2017-3-5 18:00
Finally!  Someone else who had a Phantom flip over and crash!  And I have been led to believe from DJI that it was all my fault and no one else ever heard of such.  I have been fighting DJI for over a month now trying to get someone to provide me with the flight data logs but they refuse, claiming no logs exist!!!!!  My Phantom 4 crashed from some 200 feet upside down.  I have all the proof that I am correct but DJI will not acknowledge.  Now I have someone else with the same issue.  
Thanks,
Jim

Sorry for your crash, but just because this guy had a P4 crash doesn't mean the two crashes (yours and his) were caused by the same thing.  Personal I think his was caused by a bird strike, happens all the time. That's wouldn't be DJI's fault is that was the case.

definitely could have been a bird.. or a motor that failed.. the p3 tried to upright itself a few times that's for sure, so something was pushing it over...

WilliamM Posted at 2017-3-7 09:46
Sorry for your crash, but just because this guy had a P4 crash doesn't mean the two crashes (yours and his) were caused by the same thing.  Personal I think his was caused by a bird strike, happens all the time. That's wouldn't be DJI's fault is that was the case.

Sorry William H, if it had been a bird stike that knocked it over, it would have righted itself within 4 or 5 feet.  That fact could be proved by the fact it was still powered own and flying when it crashed!  The video I have after the crash shows full power to lights, battery, etc.  
Thanks anyway,
Jim
WA5TEF

definitely more to the story!

WA5TEF Posted at 2017-3-17 11:32
Sorry William H, if it had been a bird stike that knocked it over, it would have righted itself within 4 or 5 feet.  That fact could be proved by the fact it was still powered own and flying when it crashed!  The video I have after the crash shows full power to lights, battery, etc.  
Thanks anyway,
Jim

There's many crashes that are known to be cause by bird strikes, they don't always right themselves. Maybe a bump that causes a little wobble, but a good hit that causes a flip over? There not that good to figure out the correct propulsion per prop to stop a tumbling craft. Not sure about your crash, but in the OP's case don't remember if he said all four props were still on when it crashed. A lost prop is also another possibility, rare but it does happen.

I had a vary similar crash when my P3 inverted and fell from the sky. Here is the video and logs

After approx 35 successful flights with my P3 advanced, today it inverted mid-flight and fell from the sky. I was flying over water and the P3 fell into the water and was never recovered :-( .

There was a bit of a breeze on the day of the flight, not too strong and I was flying into the wind (backwards, see video) when the crash happened. I was high than all obstacles (accept maybe birds?!).



Any idea what could have happened? I double check all props were tight before take off. However, for the past few flights I've been getting "Motor Idling (Loose or Missing Propeller)" errors but each time I landed and checked they were super tight. Maybe this was a sign of early motor failure? Looking at the flight log it seemed to automatically switch to ATTI mode after "speed error" then there was an MFI error then a "Compass Error" then crash. All happened in 0.2s!

http://www.phantomhelp.com/LogViewer/TULEQNT0P5F4PCEQA22I/

http://app.airdata.com/main?share=ykxfBp

What can I learn from this? Thanks.

fansb2281168 Posted at 2017-6-9 16:54
I had a vary similar crash when my P3 inverted and fell from the sky. Here is the video and logs

After approx 35 successful flights with my P3 advanced, today it inverted mid-flight and fell from the sky. I was flying over water and the P3 fell into the water and was never recovered :-( .

One question: You are saying your Phantom 3A crashed, however the video you referenced is titled "RIP DJI Phantom 4"... so which one is it?

SkyerUAS Posted at 2017-6-9 18:38
One question: You are saying your Phantom 3A crashed, however the video you referenced is titled "RIP DJI Phantom 4"... so which one is it?

Sorry, it was a  P3A. Video has now been corrected. Drone has now been recovered and all props were intact:

More discussion here with full flights logs and HD video (recovered from drone): www.phantompilots.com/forums/pro-adv-discussion.81/page-3

This thread is worse than a spaghetti dinner -- it has been hijacked by multiple fliers and crash descriptions and I've got no idea who is answering who any more.  If you've had an incident and want the community to help, start your own thread and discuss one incident at a time.  In pretty much every case we will want access to a flight log that allows us to download csv and kml data (tick to share them in AirData if you are using that), or use Phantom Help and check that the link you are posting actually points to your flight.  I'd love to help but the (lack of) quality of the questions is a discouraging start.

endotherm Posted at 2017-6-26 22:43
This thread is worse than a spaghetti dinner -- it has been hijacked by multiple fliers and crash descriptions and I've got no idea who is answering who any more.  If you've had an incident and want the community to help, start your own thread and discuss one incident at a time.  In pretty much every case we will want access to a flight log that allows us to download csv and kml data (tick to share them in AirData if you are using that), or use Phantom Help and check that the link you are posting actually points to your flight.  I'd love to help but the (lack of) quality of the questions is a discouraging start.

Sorry, I really wanted to start a new thread. However, new users are not allowed to start a new thread :-(

I started a thread on Phantom pilots and had lots of insight but still not cause has been identified: https://phantompilots.com/thread ... page-4#post-1209238

fansb2281168 Posted at 2017-7-1 16:18
Sorry, I really wanted to start a new thread. However, new users are not allowed to start a new thread :-(

I started a thread on Phantom pilots and had lots of insight but still not cause has been identified: https://phantompilots.com/threads/p3-inverted-and-fell-from-the-sky-into-a-lake.116755/page-4#post-1209238

O.K., I'll forgive you this time.

I read the thread on Phantom Pilots (which also suffers from hijackers...). My first observation of the video is that the aircraft seems to invert and slide into the water directly -- there is not the typical tumble in all x-y-z axis movements one would expect.  It's almost like it was grabbed by a large raptor and lowered to the water.  I can't say that I have examined a fall from the sky that appears to be so smooth and direct.

I tried to look at the flight logs, the link to Phantom Help doesn't point to your flight, and the AirData link doesn't share the kml and csv options, which is what I need for a proper analysis.  If you could fix that up and repost it, it would be helpful.

Flying backwards is not a problem and aerodynamically it is no different than flying in any lateral direction in a symmetrical "X" configuration quadcopter.

You complained of getting "Motor Idling (Loose or Missing Propeller)" errors.  This message is a result of examining the loads on the motors against the speed of rotation.  Usually you get this on the ground spinning the motors with no load (props) fitted.  The calculation would be based on using genuine DJI props, which are typically well balanced with little to no difference in weight.  Using third-party props could well cause this relationship to change from what is expected, resulting in an error message. DJI seem to have tightened tolerances in later firmware releases so they seem to be less forgiving of an imbalance and pop up a warning more often.  That doesn't mean it is immediately likely to crash, just that it is out of tolerance.  The same effect could be caused with genuine props, but ones that have sustained dings and damage from crashes, or having been sand-blasted by flying in "hostile" environments and worn down.  The first time I saw this, the person had mixed up P2 and P3 props and was getting the message all the time, even though they were tight and it appeared to fly well.  The P3 uses 9450 props and the earlier props were 9443. The props weren't as wide and didn't droop as much as a P3 prop.  As such the weights would have been  way out of tolerance, explaining why he was constantly getting error messages.  In your case, who knows what weights and drags your third-party CF props have.  DJI like to control what you use with their equipment (a bit like Apple) -- they already prevent you using non-genuine batteries so that the flight characteristics are a known quantity.  Perhaps the constant warnings are to discourage you from using non-genuine products?

I see you have recovered the aircraft and found the props were still attached (and presumably undamaged?).  Were the motors able to be rotated freely, or had any of them seized up?   If you are able to power up the aircraft, you can extract the aircraft internal flight log.  It is a DAT file that will show up if you enable data mode from the Go app, and read it off the aircraft via the front USB port next to the front name plate.  This will contain additional data beyond that which we normally examine via Phantom Help etc.  It would contain motor speed and loads which may correspond with motor failure/prop loss/abnormal battery discharge etc.  This might provide more information to determine a cause if the standard flight log is unsuccessful.  This is the same log that DJI examine when you send the aircraft in for repair and warranty claims.  It is stored on an internal sealed SD card and is unlikely to have been tampered with, so they trust it more than you sending in any old flight log.

If you can start by fixing the links to a normal flight log with access to kml and csv data, we can start there.

endotherm Posted at 2017-7-1 21:01
O.K., I'll forgive you this time.

I read the thread on Phantom Pilots (which also suffers from hijackers...). My first observation of the video is that the aircraft seems to invert and slide into the water directly -- there is not the typical tumble in all x-y-z axis movements one would expect.  It's almost like it was grabbed by a large raptor and lowered to the water.  I can't say that I have examined a fall from the sky that appears to be so smooth and direct.

Thanks a lot for your insight. That could explain for loose prop error I was getting

>I see you have recovered the aircraft and found the props were still attached (and presumably undamaged?). Were the motors able to be rotated freely, or had any of them seized up?

Yes, the props were still tightly attached ann undamaged when the aircraft was found. After disassembly and long drying time I've managed to power up the electronics and everything seems work. I've managed to spin the motors on the bench. The motors  looked a bit rusty from a week spend underwater, however they all powered up and spun!

>If you are able to power up the aircraft, you can extract the aircraft internal flight log.  It is a DAT file that will show up if you enable data mode from the Go app, and read it off the aircraft via the front USB port next to the front name plate

Yes, I managed to get the DAT file from the aircraft. Here is the raw .DAT file : https://drive.google.com/file/d/ ... EE/view?usp=sharing

Here is a link the the DJI Go log files uploaded to:

http://www.phantomhelp.com/LogViewer/TULEQNT0P5F4PCEQA22I/

http://app.airdata.com/main?share=ykxfBp

I don't wish to hijack the thread either, but I will say I had a very similar incident with my P3A in May.  Total mystery for what had been a very reliable bird for almost two years.  It was only about 20 feet from me and about 20 feet high when it suddenly swung over on its side and powered down into the concrete right near me.  Of course, the shell is destroyed, but motors seem fine, as does the gimbal.  More testing to be done as time permits.  This all happened with me looking directly at it.  No prop breakage or motor failure that I could detect by ear.  Of course, when it hit the concrete (on one leg edgewise)  it killed that prop, but motors all seem smooth and run 'fine'.  I hope you get your P3P fixed.  I'm going to try rebuilding mine with a new shell and see what happens.

fansb2281168 Posted at 2017-7-7 09:06
Thanks a lot for your insight. That could explain for loose prop error I was getting

>I see you have recovered the aircraft and found the props were still attached (and presumably undamaged?). Were the motors able to be rotated freely, or had any of them seized up?

I'll examine your files as time permits.

Neither of your flight log links work.  The PhantomHelp link takes you to the phantomhelp site with a generic page of links, nothing specific to your flight.  If you upload your flight log and get a page in your web browser with a map on the top third of the page and rows and columns of numbers beneath, copy the page address and paste that back here to share it.   With AirData, after you upload your flight you have to tick "share kml and share csv data"  when you enable sharing the link, otherwise we can't download all the telemetry data recorded in the flight.  Air Data provides a lot of conclusions by default, but does not show the data unless you tick the share options.

I've gone through and looked at the DAT file and nothing looks particularly obvious to me indicating a fault.  I would suggest you request BudWalker from PhantomPilots to have a look at the DAT file for you and see if he can see anything obvious.  He is the author of DatCon which is the .dat file viewer/converter I use.  If anyone can make sense of it, he is your man.

As I said previously, the crash looks strange and the aircraft seems to have slipped into the water without tumbling.  I can see  from the positions of the sticks in the log you didn't fly it into the water.  Your aircraft is pitched upwards, which is correct for flying backwards and ascending.  You were hitting almost +30° which is about the maximum tilt or roll in controlled flight.  This indicates about full stick deflection on both sticks, which is what we see in the logs.  The change in flight and the start of the loss is obvious. We see during the crash that the only movement is a body roll to its starboard to -65° which shows in the data.  It seems to have righted itself momentarily before hitting the water, but that isn't reflected in the video.

In the pitch/roll/yaw columns, I highlight the backgrounds green for positive and red for negative;  the higher the nose gets, the greener it is etc.  We can see the red backgrounds develop suddenly in the roll column, which is the only departure from normal flight.  Yaw maintained a steady heading and the pitch was normal, more or less.  A "normal" 3-axis tumble looks more obvious:
with lots of deep red to deep green orientation changes beyond ±30° "controlled flight", and wildly changing yaw heading.  This is typical of a single motor/prop failure.

The rest of the data sheet above shows the motor being commanded on and off to regain control and shows all the motors responding.  There is no obvious failure, and the motors report positive and negative loads on the motors.  This indicates the props are working.  This was confirmed when the wreck was recovered.

I thought it was strange looking at the wreckage that the base of the gimbal was broken and torn from the aircraft.  The rest of the airframe looks unscathed.  I didn't think it would have hit the water that hard to break the gimbal like that.  I would have expected the soft suspension rubbers to have been torn or disconnected, but not snapping the metal and disconnecting the base from the shell.

I'd like to see how you go after BudWalkers analysis.  Good luck.

endotherm Posted at 2017-7-7 16:03
I've gone through and looked at the DAT file and nothing looks particularly obvious to me indicating a fault.  I would suggest you request BudWalker from PhantomPilots to have a look at the DAT file for you and see if he can see anything obvious.  He is the author of DatCon which is the .dat file viewer/converter I use.  If anyone can make sense of it, he is your man.

As I said previously, the crash looks strange and the aircraft seems to have slipped into the water without tumbling.  I can see  from the positions of the sticks in the log you didn't fly it into the water.  Your aircraft is pitched upwards, which is correct for flying backwards and ascending.  You were hitting almost +30° which is about the maximum tilt or roll in controlled flight.  This indicates about full stick deflection on both sticks, which is what we see in the logs.  The change in flight and the start of the loss is obvious. We see during the crash that the only movement is a body roll to its starboard to -65° which shows in the data.  It seems to have righted itself momentarily before hitting the water, but that isn't reflected in the video.

Wow, thanks so much that's amazing analysis. Very impressive software to visualise the .dat file.

Your analysis from the .dat file correlates with what I remember. At the time of the crash I was flying backwards (about full stick) when I saw the video feed suddenly invert.

I'm currently in the process of rebuilding the drone.  I will let you know my experience when I try and fly it again. I would be very interested if BudWalker could provide any further insight.
Very sorry for the dead links, it looks like phamtom help viewer link has expired. Is there any further data to be gained from the DJI Go logfile that you can't already obtain from the .dat?

fansb2281168 Posted at 2017-7-10 06:48
Wow, thanks so much that's amazing analysis. Very impressive software to visualise the .dat file.

Your analysis from the .dat file correlates with what I remember. At the time of the crash I was flying backwards (about full stick) when I saw the video feed suddenly invert.

The charts above are just visualising the data in Excel, with a bunch of conditional formatting and data bar built-in functions that I have set up for analysis.  It makes it a LOT easier to relate to and analyse if everything is colour coded and shapes, rather than a block of numbers that doesn't jump out at you.  DatCon decodes the whole dat file in more detail than the txt file on your device, and in higher time resolution.  There is nothing to be gained from the txt file now that we have the dat.  I was more pointing out that the other links didn't work and we had NO data to look at before.  The dat file is the holy grail of flight records, as you can see the file is pretty big.  We often don't see them especially if the aircraft isn't recovered, however most of the time the relevant data is contained in the txt file.

Do you think I should try and make a claim with DJI? I bought the aircraft 2nd hand and it's currently in pieces. I fully disassembled to properly dry it. Do you think DJI would accept my claim?

fansb2281168 Posted at 2017-7-19 16:05
Do you think I should try and make a claim with DJI? I bought the aircraft 2nd hand and it's currently in pieces. I fully disassembled to properly dry it. Do you think DJI would accept my claim?

You would probably have trouble proving purchase of the aircraft if you bought it in a used condition. You would have to have the original owners receipt, that receipt would have to be from an authorised DJI dealer and it would have to be within the warranty period. Then ort course you have to have proof, in the form of a flight record, that the product failed.

Geebax Posted at 2017-7-19 16:57
You would probably have trouble proving purchase of the aircraft if you bought it in a used condition. You would have to have the original owners receipt, that receipt would have to be from an authorised DJI dealer and it would have to be within the warranty period. Then ort course you have to have proof, in the form of a flight record, that the product failed.

yeah, don't think I would get very far. Probably more productive to try and repair the bird myself :-/

fansb2281168 Posted at 2017-7-20 05:21
yeah, don't think I would get very far. Probably more productive to try and repair the bird myself :-/

Or buy a used one.

well today my dji spark did the same thing am hearing it just flip over and freefall to the ground breaking 2 motor as far as i can tell but dont know why this hapend it was  recording at about 2 meter of the ground stable with no warning drone crashes i have the video and and the logs sad its was my 4th time out batery wss fine just had update the app but before the crash i did 2 ten minuts  fly with 2 bateries the on my 3rd batery is when this happend ihope dji can honestly tell me why this happen i dont whant to scare away from dji just stared to enjoy flying and i do want to upgrade to the new zoom 2 which is my next step up well my  email is caraudiosystem3g@gmail.com