Hello,


I post a part of this message on an other thread (http://forum.dji.com/thread-91866-1-1.html) but people ther advised me to post a new thread.

As for what happened to me today, I was flying my drone at around 1.5 meters high, I had just taken off and was stabilized, the Mavic was maybe 2 or 3 meters away from me, so there was no connection problems, the battery was around 60% and I had just flewn it for a few minutes (2 flights of 3/4 minutes each) just before, when... Suddenly, I saw all the propellers stop at once, for no apparent reason, and the Mavic fall down in front of me. Fortunately, nothing broke. After that, I flew again for another few minutes and nothing wrong happenned. Since then, I keep thinking of what would have happened if I had been flying it at 30m above the ground or worse, above water like 'Sad OwnerOfMavic' (post#40 on the aforementioned thread).

Now I'm really preoccupied because I don't know if this will happen again or not and I'm a bit scare to use it again, which is a shame since I just bought it and I definitely want to enjoy my purchase as much as DJI will enjoy my money... ;-)

I saw on this forum that my problem is not isolated and that it seems to have happened to a lot of people recently.


I must also say that my case is not about someone who would try to blame a manufacture default for a flying error or a unwanted collision resulting in a dead drone since my drone is still OK, this is really what happened to me today. I say that because I read a few replies suggesting it is sometimes the case.




The only indication I can give you now is that when it happened I was getting a warning from my phone that its memory was almost saturated (because of the datas DJI GO 4 was wrinting on it), but I don't see how this could make the engines stop mid-air.


Finally (I know, I write long posts, I'm old school...), I keep reading that I should post my DAT.File... Well, as I'm very knew in the world of DJI and drones, I just don't know how to do that. If someone can help me on this that would be very appreciated since I'm really willing to post whatever is required for anyone to help me understand and solve my problem.


Thank you.

The instructions for obtaining the file are here, and you should use DropBox to provide us with access to the file.

THank you for your answer and your link.

In the meantime I registered to app.airdata.com and uploaded the flight. Is this enough with this link ?
http://app.airdata.com/main?share=fYkWiR

Thanks again.

odradek Posted at 2017-4-27 14:09
THank you for your answer and your link.

In the meantime I registered to app.airdata.com and uploaded the flight. Is this enough with this link ?

Is this enough with this link ?

No, not if you want someone on this forum to diagnose what happened. That's just the flight log from your mobile device; what's needed here is the DAT file off the Mavic itself, which includes a lot more information about the state of the aircraft during the flight.

DroneFlying Posted at 2017-4-27 14:12
Is this enough with this link ?

No, not if you want someone on this forum to diagnose what happened. That's just the flight log from your mobile device; what's needed here is the DAT file off the Mavic itself, which includes a lot more information about the state of the aircraft during the flight.

Ok, I'll do that. In the meantime, the video of the flight is there.



Thank you for your time.

Contact DJI support and send your log file to them, their engineer might find the reason and fix it. This will help Mavic's community to avoid worse problem. Thanks

Although I don't have reason at this point to think it had anything to do with your crash, you should avoid flying under the canopy of trees where the sky will be blocked out. That weakens the GPS signal as it did in this case, and there's also a possibility that you can wind up colliding with one of the nearby trees. Instead you should look for an open area to fly where the Mavic will have a clear view of the sky.

DroneFlying Posted at 2017-4-27 14:24
Although I don't have reason at this point to think it had anything to do with your crash, you should avoid flying under the canopy of trees where the sky will be blocked out. That weakens the GPS signal as it did in this case, and there's also a possibility that you can wind up colliding with one of the nearby trees. Instead you should look for an open area to fly where the Mavic will have a clear view of the sky.

Thank you for your comment.

I know I should fly on an open area but this was like a test and I was very very cautious to avoid any collision with trees or branches. As you can see in the flight log, My maximum speed was 9.40 mph and my maximum altitude was 10.2 ft.
But, as you say, I'll fly open areas in the future :-)

But we still shouldn't be limited to not flying under tall trees especially if we have multiple GPS Signals.

FrequentFlyer Posted at 2017-4-27 14:34
But we still shouldn't be limited to not flying under tall trees especially if we have multiple GPS Signals.

Perhaps I'm overly cautious, but I did notice that during the OP's flight that his GPS was "poor" part of the time and never went above "good". It's just a personal preference / suggestion and should be taken as such. But flying under and near trees isn't something I'd recommend to a new drone pilot.

FrequentFlyer Posted at 2017-4-27 14:34
But we still shouldn't be limited to not flying under tall trees especially if we have multiple GPS Signals.

Op had approx 10 sats,you need minimum 6 to fly so let's say he has gps 6 satellites and glonass 4 he is very much on the limit for stability, that's why they say you need a minimum of 11 satellites to fly with stability because at least then you will know you have at least 6 gps or glonass satellites . So always be careful when flying under cover you really should have 11 satellites shown.

hallmark007 Posted at 2017-4-27 14:59
Op had approx 10 sats,you need minimum 6 to fly so let's say he has gps 6 satellites and glonass 4 he is very much on the limit for stability, that's why they say you need a minimum of 11 satellites to fly with stability because at least then you will know you have at least 6 gps or glonass satellites . So always be careful when flying under cover you really should have 11 satellites shown.

Hello Hallmark007,

I understand the GPS is important to fly, but can it be the reason the propellers stopped all at once ?

odradek Posted at 2017-4-27 15:17
Hello Hallmark007,

I understand the GPS is important to fly, but can it be the reason the propellers stopped all at once ?

No it can't but it important to know that flying undercover like you were you risk going into Atti mode further more you run the risk of not having any VPS to keep your AC stable, I seen nothing in your log or video that shows motors cutting out, however this might show in your dat file.

For testing your AC this area would be considered unsuitable.

OK, thank you for your comment.

Here is the DAT file, or at least, I hope it is the right one. I tried to open it with notepad but it's all in Chinese...

https://www.dropbox.com/s/9jgsgw ... 8_00-39-10.DAT?dl=0

Let me know if you find something.

Thank you.

Flying indoor or undercover need a good skill to fly so it is no no to beginner, I am totally agreed with that but stop motors in the air is nothing to do with that.
Better for you and for the Mavic's community, you should contact DJI support and send them the log file so that their engineer can investigate the problem.
I just worry that there is the bug in firmware or DJI app which shut down the power because it thought that the Mavic was in critical error state.

Whether it had 1 GPS or 100 GPS signals...motors should not just randomly stop.  

Also, the video I posted on #5 is from the cache of my phone. The actual video from the SD card inside the Mavic stops 1 second earlier. Apparently, The Mavic kept sending video to my phone for 1 second more than the video it wrote on the SD card. Then, my phone recorded the noise of the propellers stopping in the air, the ugly sound of the crash and then the expression of my surprise (in French : "hé bé !). But both videos stopped before the fall, indicating that the power was off when it started falling down.

It's so strange that I keep getting mentioned here on that forum.
Your case is similar, so I think that something must be wrong with this european batch. I had exactly the same video feed in cache - frozen for 1 second.

So do I, especially since my Mavic was manifactured 1 day before yours, on the 23rd of march 2017... Very strange indeed... I hope that if there are a lot of us who report the problem to DJI, they will acknowledge that it is THEIR fault and that your MAVIC being part of this bad patch will be replaced with a new one... ;-)

odradek Posted at 2017-4-27 14:32
Thank you for your comment.

I know I should fly on an open area but this was like a test and I was very very cautious to avoid any collision with trees or branches. As you can see in the flight log, My maximum speed was 9.40 mph and my maximum altitude was 10.2 ft.

Your motors were still running when it hit the ground because they can be heard in the video. Also were you reving the motors? The motors sounded like they were being given a hard time.

Hello Jenee,

No, my motors were not running when it hit the ground. I saw them stop in front of me. As I wrote in my 1st post, I was only 2 or 3 meters away from it. The thing is that the Mavic being only 1.5 meters high, it took less than a second between the stop of the motors and the crash itself. It went down instently. But I'm sure of it : the propellers all stopped at once before the fall. Besides, they were not given a hard time since I was nearly stationnary at the moment it happened.

odradek Posted at 2017-4-28 01:59
Hello Jenee,

No, my motors were not running when it hit the ground. I saw them stop in front of me. As I wrote in my 1st post, I was only 2 or 3 meters away from it. The thing is that the Mavic being only 1.5 meters high, it took less than a second between the stop of the motors and the crash itself. It went down instently. But I'm sure of it : the propellers all stopped at once before the fall. Besides, they were not given a hard time since I was nearly stationnary at the moment it happened.

It's difficult to tell if your motors were running when it hit the ground, you were pushing elevator almost right up to when your AC went down, you can see this in your log and your motors are responding you can also see this in the log, I'm not to sure , this could be a problem of stability and possibly VPS not working.
It's probably only one that will be sorted from dat file, looking to see what others looking a dat file make of it.

hallmark007 Posted at 2017-4-27 14:59
Op had approx 10 sats,you need minimum 6 to fly so let's say he has gps 6 satellites and glonass 4 he is very much on the limit for stability, that's why they say you need a minimum of 11 satellites to fly with stability because at least then you will know you have at least 6 gps or glonass satellites . So always be careful when flying under cover you really should have 11 satellites shown.

Number of satellites is meaningless and a security blanket that everybody wrongly holds on to.

It is the Horizontal and vertical dilution of precision (HDOP/VDOP) that is important NOT the number of satellites the aircraft can see in either the Navstar of Glonass constellations.

The bar graph display is far more useful in ascertaining whether you have a good positional accuracy.

hallmark007 Posted at 2017-4-28 02:20
It's difficult to tell if your motors were running when it hit the ground, you were pushing elevator almost right up to when your AC went down, you can see this in your log and your motors are responding you can also see this in the log, I'm not to sure , this could be a problem of stability and possibly VPS not working.
It's probably only one that will be sorted from dat file, looking to see what others looking a dat file make of it.

Hello Hallmark007,

It's very nice from you to get an interest in my post and to try to help me find an answer to my problem, and for that I thank you. But, with the risk of repeating myself, I can assure you that the propellers all stopped turning, all at once, for no reason, in front of me, provoking the fall. I'm not saying that I'm 100% sure that it is DJI's fault with a manufactured default or a frimware problem. I'm new to the world of drones, I might have done something wrong, I don't know and that's what I'm here to find out, but, please, believe me with my description of the event because it is as accurate and sincere as it can be.
Thanks again for your interest and answers.

I looked at this flight. In summary, there was an unexplained complete power failure. It's just like several incidents that have been analyzed over in mavicpilots.com. In those incidents there was always the question - Did the battery become disconnected? Your incident provides an example where the battery didn't come loose.
Here is one of those threads
http://mavicpilots.com/threads/mavic-pro-dropped-out-of-the-sky.14482/
Also take a look at
http://mavicpilots.com/threads/will-my-mavic-fall-out-of-the-sky.14969/


With these types of incidents the battery data won't help since it is sampled at 10 Hz and has a 1.5 sec delay. Instead, the motor voltage data which is sampled at 50 Hz is used.
At the very end of the flight recording each of the motor voltages drops from about 11 volts to about 7 volts.

Hello Bubwalker,

I've been here since only yesterday but your name got mentionned several times by people telling me that it would be very helpful to get your analysis on this matter. It seems like you're a kind of a legend around here... Well better than a legend since you really exist ! :-)

Anyway, thank you for your analysis that confirms what I saw.

I imagine that the people at DJI will have the same conclusion as yours.

Since I got your attention, I have another question for you : Do you think this could be Firmware related ?

I checked the SN with the tool online and it appears that my drone was built the day before two other drones which were on the wrong end of similar problems (one of them ending down in a lake in Poland...), the 23rd of march of this year. I hope it's the firmware (I downgraded to 400 this night as a precaution). Because if it's not the firmware, I guess I have an administrative fight in front of me to get a fine working drone from DJI to replace mine... :-(

odradek Posted at 2017-4-28 02:48
Hello Hallmark007,

It's very nice from you to get an interest in my post and to try to help me find an answer to my problem, and for that I thank you. But, with the risk of repeating myself, I can assure you that the propellers all stopped turning, all at once, for no reason, in front of me, provoking the fall. I'm not saying that I'm 100% sure that it is DJI's fault with a manufactured default or a frimware problem. I'm new to the world of drones, I might have done something wrong, I don't know and that's what I'm here to find out, but, please, believe me with my description of the event because it is as accurate and sincere as it can be.

Well odradek you were closest to the situation, I don't dispute what you say happened, it was that you were so close to the ground, I suppose for you that was lucky, but for the exercise of power cutting out if you had have been 3 meters, maybe then it would have been much clearer that your motors cut out..

odradek Posted at 2017-4-28 03:06
Hello Bubwalker,

I've been here since only yesterday but your name got mentionned several times by people telling me that it would be very helpful to get your analysis on this matter. It seems like you're a kind of a legend around here... Well better than a legend since you really exist ! :-)

Sorry, IMHO, this isn't a firmware issue. I know you probably waited a long time to get your Mavic. And, now it's got to make an unplanned trip to DJI repair.  It isn't fair.

hallmark007 Posted at 2017-4-28 03:08
Well odradek you were closest to the situation, I don't dispute what you say happened, it was that you were so close to the ground, I suppose for you that was lucky, but for the exercise of power cutting out if you had have been 3 meters, maybe then it would have been much clearer that your motors cut out..

Yes, definitely, but if it had been 3 meters high it could have broken something on it... So... Anyway, with the DAT file analysis from BubWalker I guess we have a clearer view...
Thanks for helping me.

Tharg (from the Posted at 2017-4-28 02:41
Number of satellites is meaningless and a security blanket that everybody wrongly holds on to.

It is the Horizontal and vertical dilution of precision (HDOP/VDOP) that is important NOT the number of satellites the aircraft can see in either the Navstar of Glonass constellations.

You are correct, but you will find most around seem to only understand or at least put more emphasis on the number of satellites they are reading, rather than the graph bars.
And I do understand what your saying about HDOP/VDOP, but try explaining to those who just read the manual taking the advice and looking at their monitor and it seems a lot easier to read what the number of satellites are, as to trying as they should to work out there connection rate according to the graph .

Here's one simplified (I hope) explanation of the results of different satellite counts.  Let's assume you have 12 sats in view of the antenna.  They are constantly moving in relation to your location because each is in a separate orbit.  Now let's say that all twelve are pretty much overhead - not exactly, but spaced fairly close to one area of the sky from you.  Because of that, your GPS is going to calculate location based on the signals from just one area of the sky.  The horizontal resolution is going to be poor, because they are so close together.  There is little "depth perception".  Now, let's imagine that you only have only 8 sats in view, but that those 8 are spread all over the sky at the time of the reading.  The calculated resolution will be much better than when you had 12 all in one area.  Having lots of sats in view isn't nearly as important as the quality of the calculated position that you would get if that same lot was spaced out all over the sky.  Think triangulation!

All that said, as has been stated before, none of this would have anything to do with the motors stopping.  I just wanted to try to help with some explanation of the HDOP terminology, for those who haven't been around since the days of Garmin I, II, III, etc.     
Gary

BudWalker Posted at 2017-4-28 02:55
I looked at this flight. In summary, there was an unexplained complete power failure. It's just like several incidents that have been analyzed over in mavicpilots.com. In those incidents there was always the question - Did the battery become disconnected? Your incident provides an example where the battery didn't come loose.
Here is one of those threads
http://mavicpilots.com/threads/mavic-pro-dropped-out-of-the-sky.14482/

A drop of voltage from 11V to 7V is a serious problem.

Hopefully DJI will pinpoint the problem and sort it out.
So no one gets seriously injured by this.

Mcflying Posted at 2017-4-28 04:58
A drop of voltage from 11V to 7V is a serious problem.

Hopefully DJI will pinpoint the problem and sort it out.

You're right, because, at the end of the day, that's the major concern.

I contacted the DJI support and am waiting for instructions.
I'll let the MAVIC community know about what's done by DJI to solve those kind of problems.

Are we sure these arent battery issues and not AC issues? OP said he has no problems on another flight leading me to believe that maybe something is wrong with his battery.
I dont think I would send the AC back

Glad your AC wasn't damaged! Seems like the simplest explanation is that the battery was not completely seated and shook loose during flight, cutting power to the AC. I make a habit of ensuring the battery snaps in on both sides and isn't loose as part of my preflight checks.

Another forum member posted video of batteries that demonstrated a surprising amount of play when connected, which is not good -- does yours exhibit any play?

Ex Machina Posted at 2017-4-28 05:16
Glad your AC wasn't damaged! Seems like the simplest explanation is that the battery was not completely seated and shook loose during flight, cutting power to the AC. I make a habit of ensuring the battery snaps in on both sides and isn't loose as part of my preflight checks.

Another forum member posted video of batteries that demonstrated a surprising amount of play when connected, which is not good -- does yours exhibit any play?

I don't know if something is wrong with the battery, but two things I'm sure of is that when it crashed on the ground, the battery remained in its seat. I guess if it had not been properly inserted it would have been ejected during the crash. Second thing, the battery shows no play at all once inserted.
My conclusion is that if something went wrong with the battery, it was internal but not mecanical.

odradek Posted at 2017-4-28 06:10
I don't know if something is wrong with the battery, but two things I'm sure of is that when it crashed on the ground, the battery remained in its seat. I guess if it had not been properly inserted it would have been ejected during the crash. Second thing, the battery shows no play at all once inserted.
My conclusion is that if something went wrong with the battery, it was internal but not mecanical.

Here's another question for you, in case you happened to notice and remember... Was the battery itself still powered on after the crash?

Yours is one of the few cases where the crash impact was minimal, causing no apparent damage to the aircraft or battery. So, we could probably assume that if the battery was still on when the Mavic dropped, it would probably still have been on when you went over to retrieve it. This would tend to indicate the problem was within the aircraft.

On the other hand, if the battery itself was powered off, then it's possible the battery did in fact shut down, causing the loss of power to the aircraft.

Jason Lane Posted at 2017-4-28 06:27
Here's another question for you, in case you happened to notice and remember... Was the battery itself still powered on after the crash?

Yours is one of the few cases where the crash impact was minimal, causing no apparent damage to the aircraft or battery. So, we could probably assume that if the battery was still on when the Mavic dropped, it would probably still have been on when you went over to retrieve it. This would tend to indicate the problem was within the aircraft.

Well I don't remember. I didn't notice if the battery was still powered on. I would say it was, but I'm definitely not sure and I wouldn't use this information because I'm far from 100% sure.

AG0N-Gary Posted at 2017-4-28 04:45
Here's one simplified (I hope) explanation of the results of different satellite counts.  Let's assume you have 12 sats in view of the antenna.  They are constantly moving in relation to your location because each is in a separate orbit.  Now let's say that all twelve are pretty much overhead - not exactly, but spaced fairly close to one area of the sky from you.  Because of that, your GPS is going to calculate location based on the signals from just one area of the sky.  The horizontal resolution is going to be poor, because they are so close together.  There is little "depth perception".  Now, let's imagine that you only have only 8 sats in view, but that those 8 are spread all over the sky at the time of the reading.  The calculated resolution will be much better than when you had 12 all in one area.  Having lots of sats in view isn't nearly as important as the quality of the calculated position that you would get if that same lot was spaced out all over the sky.  Think triangulation!

All that said, as has been stated before, none of this would have anything to do with the motors stopping.  I just wanted to try to help with some explanation of the HDOP terminology, for those who haven't been around since the days of Garmin I, II, III, etc.     

Yes but I think the message is, your graph will give you a better indication of what signal you have rather than the amount of satellites shown, but the problem is most seem to rely on sat count, and some even ignore graph altogether.

odradek Posted at 2017-4-28 06:31
Well I don't remember. I didn't notice if the battery was still powered on. I would say it was, but I'm definitely not sure and I wouldn't use this information because I'm far from 100% sure.

No worries. Thanks for trying and for giving your best answer :-)