Mavic down

[BudWalker]

Hi I cut the excel file I've been looking at and you can see the takeoff GPS height is 681 and at the last entry is 682 yet the relative height is still at 47.8. The motors say that they have power as well. This is what I have been going off as I said I'm not an expert so if I'm reading it wrong please let me know. As for the drone its with DJ now so I don't think I can have another go at the data from that. Those two extracted files were the ones I thought came from the drone.

Dropbox - Cut 2018-03-11_16-33-27_FLY056.csv

You probably spent a lot of time cutting down the .csv in an attempt to make it easier for others to look at. But, a lot of the info I would want to look at is missing. I suggest that you provide the original .DAT so that we can have a look.

 

[Ronsta@Mav]

You probably spent a lot of time cutting down the .csv in an attempt to make it easier for others to look at. But, a lot of the info I would want to look at is missing. I suggest that you provide the original .DAT so that we can have a look.

Hi BudWalker here is the link to the DAT file

Dropbox - 2018-03-11_16-33-27_FLY056.DAT
Thanks

 

[Ronsta@Mav]

They sent me an email last night saying that the data was in conclusive and can't really find what happened, they offered me 30% off the parts. What do you think I should do? I don't really want that drone back if it's a lemon, they are admitting some level of fault I think.

 

[BudWalker]

Hi I cut the excel file I've been looking at and you can see the takeoff GPS height is 681 and at the last entry is 682 yet the relative height is still at 47.8. The motors say that they have power as well. This is what I have been going off as I said I'm not an expert so if I'm reading it wrong please let me know. As for the drone its with DJ now so I don't think I can have another go at the data from that. Those two extracted files were the ones I thought came from the drone.

Dropbox - Cut 2018-03-11_16-33-27_FLY056.csv

I get it now. Maybe I'm following too many incidents. This is the cut down version of the FLY056 that we've been looking at. I thought you had managed to retrieve FLY056 from the Mavic and this .csv came from that. If I had looked at it a bit closer I would have seen that. Sorry about that.

You're right about relativeHeight. The relativeHeight signal exists in the Mavic .DAT, but in a tablet .DAT it doesn't exist and is computed by DatCon. Essentially relativeHeight = Barometer:Smooth - Barometer:Smooth(at launch). There is a delay incurred in computing the smooth value - that's why the relativeHeight makes it look like the Mavic stopped recording at 47 meters. I'll change DatCon so that it uses Barometer:Raw to compute relativeHeight.


Knowing now the problem with relativeHeight the battery issue/disconnect supposition doesn't make any sense. I took another look and have to agree with DJI - there isn't enough data in the tablet .DAT. The fact that you've retrieved FLY060.DAT from the Mavic makes it likely that FLY056 exists on the Mavic as well. If you're FTP savvy you could try using it instead (I've been using it lately and like it much better than using DJI Assistant 2).
1) Connect USB port to the PC and then start a FTP client (I use FileZilla).
2) connect to 192.168.42.2 with UserName = nouser, Password = nopass, port = 21
3) On the remote site (i.e., the Mavic) navigate to the /blackbox/flyctrl directory
You should see FLY056.DAT

Notwithstanding your comment it looks to me like you've become quite the expert with .DAT logs.

 

[BudWalker]

Extracted with 1.2.0 and attempted to open with 3.4.0. I'll send you a link.

For those still following this issue the root cause of this problem lies with ExtractDJI. In sar104's case there already existed a FLY060.DAT causing ExtractDJI to create a FLY060(2).DAT instead of overwriting the existing FLY060.DAT. ExtractDJI actually reports this but it's very easy to miss. I'll change ExtractDJI so that a popup occurs that the user has to explicitly acknowledge.

 

[sar104]

For those still following this issue the root cause of this problem lies with ExtractDJI. In sar104's case there already existed a FLY060.DAT causing ExtractDJI to create a FLY060(2).DAT instead of overwriting the existing FLY060.DAT. ExtractDJI actually reports this but it's very easy to miss. I'll change ExtractDJI so that a popup occurs that the user has to explicitly acknowledge.

You are too kind. I would not say that it was an ExtractDJI problem - more just user incompetence.

 

[BudWalker]

You are too kind. I would not say that it was an ExtractDJI problem - more just user incompetence.

Oh no, with all due respect I think you are certainly mistaken. Clearly the problem lies with incompetence of the ExtractDJI developer.

 

[Simmo]

Back to your corners..... wait for the bell...

 

[sar104]

Oh no, with all due respect I think you are certainly mistaken. Clearly the problem lies with incompetence of the ExtractDJI developer.

Don't fight me on this - you are doomed to lose. I have the all the best incompetence. Tremendous incompetence. #biglyincompetent.

 

[Ronsta@Mav]

I get it now. Maybe I'm following too many incidents. This is the cut down version of the FLY056 that we've been looking at. I thought you had managed to retrieve FLY056 from the Mavic and this .csv came from that. If I had looked at it a bit closer I would have seen that. Sorry about that.

You're right about relativeHeight. The relativeHeight signal exists in the Mavic .DAT, but in a tablet .DAT it doesn't exist and is computed by DatCon. Essentially relativeHeight = Barometer:Smooth - Barometer:Smooth(at launch). There is a delay incurred in computing the smooth value - that's why the relativeHeight makes it look like the Mavic stopped recording at 47 meters. I'll change DatCon so that it uses Barometer:Raw to compute relativeHeight.
View attachment 35625

Knowing now the problem with relativeHeight the battery issue/disconnect supposition doesn't make any sense. I took another look and have to agree with DJI - there isn't enough data in the tablet .DAT. The fact that you've retrieved FLY060.DAT from the Mavic makes it likely that FLY056 exists on the Mavic as well. If you're FTP savvy you could try using it instead (I've been using it lately and like it much better than using DJI Assistant 2).
1) Connect USB port to the PC and then start a FTP client (I use FileZilla).
2) connect to 192.168.42.2 with UserName = nouser, Password = nopass, port = 21
3) On the remote site (i.e., the Mavic) navigate to the /blackbox/flyctrl directory
You should see FLY056.DAT

Notwithstanding your comment it looks to me like you've become quite the expert with .DAT logs.[/QUO

Thanks for that I wish I still had the AC because the 60 and 61 was all I got off the drone using assistant 2. My bad I'll know for next time.

Cheers

 

[Ronsta@Mav]

Don't fight me on this - you are doomed to lose. I have the all the best incompetence. Tremendous incompetence. #biglyincompetent.

Hi sar104 is there any chance of more flight data to be in 60 or 61 because when I use datcon it tells me there is more than one flight record in there. When I try to extract the files using DJIextract it tell tells me that the file wasn't created by Assistant 2 and just stops there. Any chance someone can extract and place a link back here if there is something useful in there?

Thanks

 

[BudWalker]

Hi sar104 is there any chance of more flight data to be in 60 or 61 because when I use datcon it tells me there is more than one flight record in there. When I try to extract the files using DJIextract it tell tells me that the file wasn't created by Assistant 2 and just stops there. Any chance someone can extract and place a link back here if there is something useful in there?

Thanks

DJI_ASSISTANT_EXPORT_FILE_2018-03-12_08-23-21.DAT is the file that DJI Assistant 2 created. It's a compressed file that contains FLY061.DAT and FLY060.DAT. ExtractDJI will decompresses files created by DJI Assisitant 2. In this case DJI_ASSISTANT_EXPORT_FILE_2018-03-12_08-23-21.DAT is decompressed to create FLY061.DAT and FLY060.DAT.

When DJI_ASSISTANT_EXPORT_FILE_2018-03-12_08-23-21.DAT is submitted to DatCon it invokes ExtractDJI which sees that there are two .DATs. DatCon then offers offers to process the first .DAT (FLY061.DAT) and tells that you'll need to use ExtractDJI explicitly to obtain the other .DAT.

FLY061.DAT and FLY060.DAT weren't created by DJI Assistant 2 - that's why ExtractDJI gives that error message

 

[Ronsta@Mav]

DJI_ASSISTANT_EXPORT_FILE_2018-03-12_08-23-21.DAT is the file that DJI Assistant 2 created. It's a compressed file that contains FLY061.DAT and FLY060.DAT. ExtractDJI will decompresses files created by DJI Assisitant 2. In this case DJI_ASSISTANT_EXPORT_FILE_2018-03-12_08-23-21.DAT is decompressed to create FLY061.DAT and FLY060.DAT.

When DJI_ASSISTANT_EXPORT_FILE_2018-03-12_08-23-21.DAT is submitted to DatCon it invokes ExtractDJI which sees that there are two .DATs. DatCon then offers offers to process the first .DAT (FLY061.DAT) and tells that you'll need to use ExtractDJI explicitly to obtain the other .DAT.

FLY061.DAT and FLY060.DAT weren't created by DJI Assistant 2 - that's why ExtractDJI gives that error message

Ok thanks so the one we are looking for is probably still the AC correct?

 

[BudWalker]

Ok thanks so the one we are looking for is probably still the AC correct?

It might be on the Mavic. It's not in any of the files that you've put in your Dropbox link.

 

[Fullerj27]

Hi first time user long time reader, I'm hoping someone here can help me out. This is how the flight went I unpacked the drone from the carry bag, unfolded the arms, checked the propellers, checked the battery was in firmly, did all pre flight checks, it was later in the afternoon and the wind was minimal, proceeded to takeoff. I did about 15 minutes of flying up and down the creek getting different shots of the creek, I decided to do a point of interest shot. So I climbed up a little and activated the control for (POI). It was going along fine and then I got the message to say the cache was full so I decided to land the drone and attend to this . So I hit the (RTH) on my device all seemed fine the drone did it usual functions came over head and began to descend. As you will see on the flight record it only came down a few meters and then it just fell without warning (it was like it landed at 47.3 mtrs above the ground).
The DJI go app stopped recording at this time as well witch is strange, the Dat files from the flight keep recording all the way to the ground showing a rapid decent in the last 4secs of the flight and seems to have voltage to the motors all the way down as well.
I'm not an expert on this so I could be reading the data incorrectly. any help would be greatly appreciated.

Ill place the dropbox link below.
The files in question are dated 11/03/2018 at 16:34 for the txt file
and the Dat file is 11/03/2018 FLY056 16:33.27

drone data

Thanks

Was it a DJI battery, or 3rd party aftermarket? if it was the latter, these batteries are known to cause problems. Your scenario sounds just like a 3rd party battery incident.

 

[motopokep]

No offense to you, personally, but I get a kick out of the "control the drone yourself" crowd when they are not really in control at all. I'd like to see them control a quad if all they had were four sliders that controlled each motor's speed individually.

Even a simple landing of a Mavic by pulling down the left stick is only telling the FC that you want to descend. The FC does the actual work, including deciding that it is close enough to the ground to auto-land (as it announces, "Landing").

To me it is silly to not trust a computer to fly, when it is flying it anyway. How many people use Litchi, AutoPilot, et al? Lots. And they are usually not beginners. Yet those apps are often used for preplanned missions where, yep, the FC is doing the "flying."

Sure, if you don't understand RTH or configure it correctly it can lead to a surprise. But that surprise can come at any moment anyway on a disconnect. Why not fully trust it, and test it, so that there is no surprise? You don't wait for a fire to find out fire drills would have helped.



Mike

The reason I bought the Mavic is because my previous GPS drone, Upair One, fell out of the sky after I activated RTH. It fell down like 400 feet and I couldn't do anything. There have been many many reports of flyaways and crashes with all brands of GPS drones when RTH was activated. Given that DJI products do tend to have the least issues with RTH, they still happen. Maybe it's pilot error due to strong winds, maybe the compass needed calibration, maybe there was magnetic interference at take off. Too many factors can affect RTH. That is the reason I am part of the control it manually crowd, I'm not taking any chances. I've only used RTH about 5 times, couple times tested it out of the box, and couple times was too far away and lost signal and used RTH to regain signal and fly back manually. If you want to use RTH in every flight, it's your prerogative, just like it's ours to fly back manually. If your sportbike had a button to do a wheelie, would you use it?

 

[Cheech Wizard]

Sorry for your first crash experience.
Luckily there are a few heros here that can help you determining what caused your RTH crash.

I have said it many times and I keep repeating (only as a friendly advise from a pilot who started flying manual drones well before any autonomy was available);

RTH is a fail safe that hopefully will work when it's absolutely needed, not to be used as a normal flight mode, although most people seem to think it is and claim that it is absolutely fool proof because they never had a problem with it. But by far most problems occur during RTH. Wether by wrong settings by the pilot, a GPS or compass interference, obstacles, a low Sun, and maybe even because of a barometer issue while landing.

It is always better and safer to control the drone yourself, after you learned to master it. Manual take off and landing is safer than any algorithm written by a programmer somewhere in China. The amount of variables in a real life environment are countless and no algorithm in a consumer electronics world is capable of dealing with all of them. 99% RTH works ok but you never know when that 1% is up. Same for auto take off and landing.

I would still use the "Precise Takeoff" feature in case I did need RTH. You don't have to use it every time you land, but better safe than sorry.

 

[Tinmania]

If you want to use RTH in every flight, it's your prerogative, just like it's ours to fly back manually. If your sportbike had a button to do a wheelie, would you use it?

As far as RTH I don’t use it for every flight, but I test it enough to have confidence enough to do just that, should I desire.

What I don’t think, however, is that I am in direct control of my MP when it is flying. I understand that the FC is doing it, and accept it. So when someone says something like, “Tried RTH once or twice a year ago but prefer to be in control so never touched it since,” I do have to roll my eyes a bit.

Back to the motorcycle analogy: I see where we might differ. If my toy RC sport bike had a button to do a wheelie then of course I’d use it.



Mike

 

[motopokep]

As far as RTH I don’t use it for every flight, but I test it enough to have confidence enough to do just that, should I desire.

What I don’t think, however, is that I am in direct control of my MP when it is flying. I understand that the FC is doing it, and accept it. So when someone says something like, “Tried RTH once or twice a year ago but prefer to be in control so never touched it since,” I do have to roll my eyes a bit.

Back to the motorcycle analogy: I see where we might differ. If my toy RC sport bike had a button to do a wheelie then of course I’d use it.

Mike

Yes, the toy RC sportbike is a better analogy when discussing an unmanned drone, but my point was about trusting advanced technology. If one wouldn't trust an automatic wheelie button that could give an uncontrolled lift on your bike and flip you backward, one shouldn't trust a single button to auto bring back your drone, that could have erroneously recorded the home point to be on the other side of the globe. I'm still baffled why my Upair One fell out of the sky when I activated the RTH switch, after having come back via RTH so many times before.

I agree that you are never in complete control of the mavic when flying it manually, but you're much more in control of it than when using RTH. I would compare RTH to intelligent modes like tapfly, terrain follow, etc. You basically stop using stick input and let the drone fly on programmed code, which is much more advanced than the code used for stick input. I've seen a few crashes reported on here due to intelligent flight modes failing or RTH problems. I don't recall reading about the Mavic all of a sudden having a mind of its own and flying away or crashing down when you are just flying it with stick input (I'm disregarding prop and battery failure and strong wind occurrences).

RTH has been discussed here over and over. I'd love to see that feature work flawlessly, but for now with all the crashes and flyaways reported, not so much for the Mavic, but more so for other drones, including my own experience, I'll use it as a fail safe feature only. Besides, I enjoy navigating the drone back manually, changing altitude on the way back. For example, you get a nice descending view on the return home when you're up high, say 300-400 ft, and do a full (left) throttle down and full forward on the right stick, looks kinda like an airplane landing, versus the RTH view of coming in at a certain altitude and just lowering down.

 

[Danilo831]

Yesterday for the first time I was try to flight with my first brand new Mavic Pro. The previous days I did the software upgrade of the Mavic as requested from the application, I m using an iPhone X. As I m traveling around the world, the weather was really good to make a nice footage in St. Maarten (Caribbean Island), so I prepared my self to flight, first the remote controller took a bit time to connect with the Mavic, I restart 2 or 3 times Controllers and Mavic together once I get finally connected I attempt to take off, the Mavic rich an height of about 15 meters steady for 5 second or less, start spinning around to 360 degrees without touching anything on the controller than he made a crash landing, likely me and my mate we where a bit far, was really scaring. While crashed the battery went off, gimbal detached and broken, few scratches around and few damage in the plastic here found some pictures.

I would like to know there is something wrong with Gyroscope or Magnetic Compass?

Is DJi will replaced it