Hit something in the air, drone never fell

[Echo5delta16]

YES!!! thats how mine acted but the gimbal bounce was a lot more intense.
However the first time it happened a few days ago there was guts from a bug or a small bird on the prop when i got it home. Watching the flight record it shows my gimbal go fast up and down. i dont recall if that was me trying to find the ground incase i crashed or if it was a glitch making it move..
I would like to record my phone while im flying so i can have a record(completely separate from Dji app) of it until i find out whats causing it does anyone know how to record what happens on your phone?

I am wondering if what visually appears to the operator as collision could be the Gimbal misbehaving along with the camera resetting. Had it happen twice with .600 where the sudden jolt in video makes it appear like I struck something in clear air. Both occasion the video recording to SD trips off however the fpv view recovers promptly. @BudWalker seemingly you are not able to pull the Gimbal pitch angle or camera state at present? As it would be good to see what was going on with the camera in these events.

This is the live video feed in one such event. Doesn't show a lot other than the instant it starts you can see the gimbal recovering from a jolt. The moment prior which was not recorded was where it looked like a collsion. Then in the last 4 second the camera flashes bright, refocuses then a climb starts.

View attachment 11311

As best I can tell video starts at ~551 and goes for 24 seconds until 574 as mavic starts to climb. Just prior to the start of the video where the black vertical line is in the log you cans see four of five spikes in the gyro degrees. But they are simply me rotating the Mavic to take some stills JPG+Raw images. Maybe recording the raw images cause the camera to misbehave although it has never caused an issue in the past.
Either way it visually looked to me like a collision when in reality it was not.

The second event was more visually exciting although I have no cache record. Likewise the DAT for it revealed nothing untoward as far as I can see. Would be interesting to see what the gimbal pitch was doing in these events.

 

[BudWalker]

That low battery part was after it hit something. I had walked around the pond to get to it and i was flying it back to my vehicle while walking behind it to see what would happen if i pushed it past low battery.. the spot where i sat for a long time before that low battery part is where it stopped and i saw the camera go sideways on the screen but it the video was apparently lost.

Did the gimbal incident occur here?

 

[Logger]

....BUT in both of my events the signal strength was reduced to the point that the TXT file has no gimbal telemetry recorded for a period before and after the events. even though the Mavic was relatively close by and flight telemetry was still OK. ie when I play back the flight in GO the gimbal position greyed out about 20 seconds before each of the events.

Actually I was incorrect here. Greyed out gimbal in Go playback just meant I was not moving gimbal at time it. The telemetry was in the TXT file after all. Doesnt really show much other than the simultaneous spike of all three gimbal axis toward the end of the climb at t578 corresponds to the end of the cache video when the cache video quit and restarted a new recording and weak signal message. I think the Yaw spike at t548 & t576 is misleading as it it simply passing through 180 degrees.

 

[Echo5delta16]

Did the gimbal incident occur here?
View attachment 11315

Yes right there!

 

[Echo5delta16]

Actually I was incorrect here. Greyed out gimbal in Go playback just meant I was not moving gimbal at time it. The telemetry was in the TXT file after all. Doesnt really show much other than the simultaneous spike of all three gimbal axis toward the end of the climb at t578 corresponds to the end of the cache video when the cache video quit and restarted a new recording and weak signal message. I think the Yaw spike at t548 & t576 is misleading as it it simply passing through 180 degrees.

View attachment 11314

It's a bit confusing, are you talking about my data or yours?

 

[Logger]

No I am sorry - didnt mean to hijack your thread, Was talking about my data. Seemingly my telemetry did not really capture the gimbal misbehaving. Although I am no expert.

Be interesting to see what @BudWalker see's in your data at that spot and if it is a similar issue.

 

[VE79]

Hi.

Had to sign up to tell my story...or struggle with Mavic Pro...
And at first, you did not hit anything mid air. It is most probably a software issue, I think.

I received my MP bit over a month ago, from the very first flight I faced problems, and one of those is just like yours.
At first, when it happened, I was sure the Mavic is falling when I was only looking at the screen, but luckily the drone was so close to me I could see it was not.
Other problems were losing the GPS, going from ~15 satellites to lost GPS all at once. I managed to pull it down to the field with no damage.
And once I did auto take off, Mavic went unresponsive and I had to take it down with my hands, it just hovered at 1,2 meters.

I took my Mavic back to the local dealer who sent it to the service to Netherlands. It took one month to get it back, and they said it is fine, no problems found.
They didn't bother to look at the video clip at the memory card where the gimbal went crazy.
Original comment from the service: "Warranty, tested OK. after changing the controls from custom to mode 2 the unit flew perfectly. there was no drift or GPS loss, even indoors. The gimbal did not twitch or unstable. The customer did not send in any proof or evidence to support the case with."

Here is the first flight, straight after the service from this morning:

password for the video is DjiMavic.
Problems @ 1:40, 2:40 and 4:10.
Before the flight I calibrated everything.

I took it back to the shop and I have to say, I'm a little bit pissed off right now.


Br,
Veli

 

[Echo5delta16]

Hi.

Had to sign up to tell my story...or struggle with Mavic Pro...
And at first, you did not hit anything mid air. It is most probably a software issue, I think.

I received my MP bit over a month ago, from the very first flight I faced problems, and one of those is just like yours.
At first, when it happened, I was sure the Mavic is falling when I was only looking at the screen, but luckily the drone was so close to me I could see it was not.
Other problems were losing the GPS, going from ~15 satellites to lost GPS all at once. I managed to pull it down to the field with no damage.
And once I did auto take off, Mavic went unresponsive and I had to take it down with my hands, it just hovered at 1,2 meters.

I took my Mavic back to the local dealer who sent it to the service to Netherlands. It took one month to get it back, and they said it is fine, no problems found.
They didn't bother to look at the video clip at the memory card where the gimbal went crazy.
Original comment from the service: "Warranty, tested OK. after changing the controls from custom to mode 2 the unit flew perfectly. there was no drift or GPS loss, even indoors. The gimbal did not twitch or unstable. The customer did not send in any proof or evidence to support the case with."

Here is the first flight, straight after the service from this morning:

password for the video is DjiMavic.
Problems @ 1:40, 2:40 and 4:10.
Before the flight I calibrated everything.

I took it back to the shop and I have to say, I'm a little bit pissed off right now.


Br,
Veli

Yea that was the same thing that mine did. So DJI doesn't view this as a problem??

 

[VE79]

Yea that was the same thing that mine did. So DJI doesn't view this as a problem??

I really don't know.
It seems they didn't even bother to take a look. Just kept it there a few weeks and sent it back. There was a test flight report included, stating all is ok, but for me it does not look like it's ok...

Let's see, the local "service" said they will replace it with a new one if something is found. I consider that video itself as finding "something", but maybe they are struggling finding 5 minutes to watch it, as they did't watch the previous one.
Service in quotation marks because in contrast to advertised they can't do anything to DJI's here in Finland, they're just forwarding those to Netherlands. One of the main reasons choosing the local dealer instead ordering it just from somewhere was the local service...

 

[BudWalker]

Yes right there!

The reason that I asked is because there isn't anything in the .DAT to indicate a problem here. But, the .txt shows the gimbal making abrupt pitch and roll changes for no reason.

The minor changes at 1150 are associated with AC pitch and roll and are normal. The changes at 1161 occur while the AC is stable in pitch and roll. That's 90 degrees down followed by 90 degrees CW roll. This kinda looks like the gimbal pitch and roll that my Mavic does after the battery is turned on.

I wasn't able to find anything the .DAT that coincides with this gimbal event in the .txt. In particular, the accelerometer and gyro data is relatively stable.The timescales are offset where time 1290 in the .DAT coincides with time 1161 in the .txt.


Seems clear that there was no mid air collision and what you were seeing was the result of the gimbal movements. Other than that there isn't much I can add. The data that would tell us what caused the gimbal changes may be in the .DAT. But, CsvView/DatCon isn't smart enough to extract it.

Since this is the second time this has happened maybe it's time for a trip to DJI repair?

 

[Echo5delta16]

The reason that I asked is because there isn't anything in the .DAT to indicate a problem here. But, the .txt shows the gimbal making abrupt pitch and roll changes for no reason.
View attachment 11345
The minor changes at 1150 are associated with AC pitch and roll and are normal. The changes at 1161 occur while the AC is stable in pitch and roll. That's 90 degrees down followed by 90 degrees CW roll. This kinda looks like the gimbal pitch and roll that my Mavic does after the battery is turned on.

I wasn't able to find anything the .DAT that coincides with this gimbal event in the .txt. In particular, the accelerometer and gyro data is relatively stable.The timescales are offset where time 1290 in the .DAT coincides with time 1161 in the .txt.
View attachment 11346

Seems clear that there was no mid air collision and what you were seeing was the result of the gimbal movements. Other than that there isn't much I can add. The data that would tell us what caused the gimbal changes may be in the .DAT. But, CsvView/DatCon isn't smart enough to extract it.

Since this is the second time this has happened maybe it's time for a trip to DJI repair?

 

[Echo5delta16]

Sigh.........I guess so.

 

[BudWalker]

Sigh.........I guess so.

I know you didn't wanna hear that. How long have you been waiting for your Mavic?

 

[dareich]

Hey guys. I have had the same issue, but it happens several times on every flight. I've tried lots of stuff to try to resolve the problem and the closest so far is clearing the video cache. I got 9 minutes glitch free instead of the usual 3 minutes but still glitches. I made a video from cache compiling just a few of the many resets I've seen.

I'm going to see if turning off the caching helps. That would suck though since I like to watch the videos on the phone later.

BudWalker if you want more flight data I would gladly share mine. Would a share from AirData be enough or is there more to it. I'm not sure where to get the DAT and other files quite yet.

 

[BudWalker]

Hey guys. I have had the same issue, but it happens several times on every flight. I've tried lots of stuff to try to resolve the problem and the closest so far is clearing the video cache. I got 9 minutes glitch free instead of the usual 3 minutes but still glitches. I made a video from cache compiling just a few of the many resets I've seen.

I'm going to see if turning off the caching helps. That would suck though since I like to watch the videos on the phone later.

BudWalker if you want more flight data I would gladly share mine. Would a share from AirData be enough or is there more to it. I'm not sure where to get the DAT and other files quite yet.

The AirData share won't help much. The .txt file from the tablet is what is needed. Go to https://www.phantomhelp.com/LogViewer/upload/ to see instructions on how to retrieve the .txt. You can upload the .txt to that site and provide the link that it produces. Or, since it's a .txt, it can be uploaded here; just push the "Upload a File" button.

I'll be on travel for the next couple of days and will have to look at it after that. Maybe @Robbyg could have a look before then.

 

[Robbyg]

I would be happy to take a look at it.
I have come across a few Youtube videos over the last 6 months with people having the same issue. Unfortunately the end result was the person sending it back to DJI for repair. Looking at the problem it almost seems like the controller for the Gimbal is being reset in Midair.

My personal advice is do not waste time, send it in under warranty now. Everyday you wait is another day that you can scratch off of your one year warranty

Rob.

 

[dareich]

Great advice. I'll call them next week. I'll upload a small flight log from today. Turning off video caching did not help. Still glitched. I tried sport mode for the first time and it flaked out at max speed. Over a road. Little freaked out by that.

Glitched a D and also while landing.

 

[highcrow]

I am wondering if what visually appears to the operator as collision could be the Gimbal misbehaving along with the camera resetting. Had it happen twice with .600 where the sudden jolt in video makes it appear like I struck something in clear air. Both occasion the video recording to SD trips off however the fpv view recovers promptly. @BudWalker seemingly you are not able to pull the Gimbal pitch angle or camera state at present? As it would be good to see what was going on with the camera in these events.

This is the live video feed in one such event. Doesn't show a lot other than the instant it starts you can see the gimbal recovering from a jolt. The moment prior which was not recorded was where it looked like a collsion. Then in the last 4 second the camera flashes bright, refocuses then a climb starts.

View attachment 11311

As best I can tell video starts at ~551 and goes for 24 seconds until 574 as mavic starts to climb. Just prior to the start of the video where the black vertical line is in the log you cans see four of five spikes in the gyro degrees. But they are simply me rotating the Mavic to take some stills JPG+Raw images. Maybe recording the raw images cause the camera to misbehave although it has never caused an issue in the past.
Either way it visually looked to me like a collision when in reality it was not.

The second event was more visually exciting although I have no cache record. Likewise the DAT for it revealed nothing untoward as far as I can see. Would be interesting to see what the gimbal pitch was doing in these events.

I made a new thread about it, and your issue is very similar to mine. My mavic stopped recording video when this thing happened and altitude was showing -8ft when it was flying over 30ft.

I guess I need to contact DJI to solve this issue..,

 

[rvaidyan]

Could you please post an update? I have been having the same problem with several of my videos. It seems to be doing this reset in midair and it is terrifying if the MP is a little ways away. I usually quickly bring it back home without trouble, but the video stops recording whenever it happens.

 

[McDerp]

So I have a question for you all.
My first outside of my sight flight went great except for at one point I hit what I think was a bird. On the screen the gimbal or drone went sideways, spun twice and was slow to respond for about 45 seconds then everything went back to normal. Problem is that the video was recording when I hit what ever and the impact made the camera stop recording. I hit the record button and it started recording again with no issues. I had taken a few POI shots and was still recording when I hit something and the whole clip seems to have disappeared.
Today the same exact thing happened, I was flying and hit something and camera went crazy then stopped recording even though it never fell from the sky. I was at a lower altitude on the other side of a pond so I left it there and walked to see and NOTHING was within 100 feet of it. But again, the whole clip stopped. Has anyone had this issue before or know what could have happened to the video?

I have had the same thing happen to me in this exact scenario earlier today. The file is too big for me to post the video of it happening but it scared me to death. I thought I was going to lose the drone. What could be happening?