1st crash in over a year, what happened??

[badbrain]

Hey,

Drone “drifted”/flew into a side of a building at speed. Sideways.

Have been having incessant remote beeping noises for the last month. Since last update. It goes after remote calibration, but always comes back. Also did an IMU calib, with same result, solved for a few flights.

The drone also drifts back. When I do remote calibrations, I could see the remote going back on its own (24%). Then restart drone and no response at all for 20 secs or so. Just hovering.

Finally, took it to official DJI store, they had a quick look and said it seemed ok. Of course,
No beeping that time.

Tried it that evening, beeping. Which I videoed. When outside to see if it drifts, about 15m up, BAM, into the side of a building.

Who can check my logs please? No problems at all for 1st year, no crashes, nada.

Many thanks

 

[Moose1967]

My guess will be interference or compass was messed up a bit - after having some drifting issues myself I decided to do a compass calibration and imu calibration out at my remote camp - never had an issue since - i think even though the app says they are successful it's always better to do them away from wifi, metal etc yo get a "real good calinration"

Look forward to seeing what the data you added says.

 

[Fly Dawg]

Drone “drifted”/flew into a side of a building at speed. Sideways.

According to this data set : DJIFlightRecord_2019-07-16_[21-32-57].txt You applied full aileron and kept it applied until it hit the building sideways. I don't see any "drifting".

 

[Meta4]

Drone “drifted”/flew into a side of a building at speed. Sideways.

No mystery.
The data looks like this: DJI Flight Log Viewer - PhantomHelp.com
From 0:41.5 you had the right stick pushed hard to the right.
At 0:43.4 the drone collides with something, you let go of the sticks and the drone came tumbling down.

 

[badbrain]

Thanks guys. Exactly what I thought happened.

BUT, I wasn’t pushing full on on the sticks. Just tapping to keep straight.

As mentioned, the drone has been drifting back for the last month, every other flight. Remote calibration every other flight because of beeping. When doing the calibration, can always see the right one going down on its own. Plus neither stick works for at least a few goes when doing calibration.

Could it be connected?

I’m a very careful pilot and never go “full on” on the sticks. At least it still flies, though.

Should I try DJI in your opinions? Or some glue? Or nothing?

 

[badbrain]

No mystery.
The data looks like this: DJI Flight Log Viewer - PhantomHelp.com
From 0:41.5 you had the right stick pushed hard to the right.
At 0:43.4 the drone collides with something, you let go of the sticks and the drone came tumbling down.

That’s a great read out, thank you ?. So strange though, I was fighting drift with taps (as the speed shows) then speed suddenly just increases soo much, that’s when it went crazy. Could there be something sticking inside the remote to cause the constant backwards drifting and this accident? Plus the beeping sounds?

 

[badbrain]

Screw it, I’ll try DJI. Only live 25 mins cycle away, would be silly not too.

Plus I already went to see them the same day as this incident, 4 hours before, to report drifting and remote beeping.

 

[Fly Dawg]

I’m a very careful pilot and never go “full on” on the sticks. At least it still flies, though.

You did in this case. You were issuing aileron commands, but the last one shown in red below was the full on that crashed the aircraft. The aircraft was responding as it should and was NOT drifting.

 

[Kilrah]

Seems like you carry something magnetic on you close to the remote.

 

[Meta4]

Thanks guys. Exactly what I thought happened.
BUT, I wasn’t pushing full on on the sticks. Just tapping to keep straight.
...
I’m a very careful pilot and never go “full on” on the sticks.

The data doesn't agree with your memory.

The 364s and 1684s indicate a joystick pulled or pushed to the limit.

I’m a very careful pilot and never go “full on” on the sticks.

There's no reason to avoid pushing or pulling the joysticks to the limit (unless you are close to an abstacle).

Seems like you carry something magnetic on you close to the remote.

Where does that idea come from?

 

[Astom22]

The data doesn't agree with your memory.

The 364s and 1684s indicate a joystick pulled or pushed to the limit.

There's no reason to avoid pushing or pulling the joysticks to the limit (unless you are close to an abstacle).

Where does that idea come from?

I'm not sure this is possible on the Air. Could he have somehow adjusted his rates to where a slight input could have a full effect, like I'm able to do with my racing drone and my Taranis remote and flight controller?

Just a thought. I haven't messed with my air much, other than basic camera and remote settings.

 

[Lon Denard]

Hey, you're not one of these guys that put one of those magnetic micro USB charge ports on your remote, are you?

Smart Watch? The Samsung Galaxy S3 Frontier is capable of utilizing mag strip readers.

Old/ damaged or aftermarket props?

 

[Kilrah]

Where does that idea come from?

On the controllers that use magnetic sticks having a magnet close to them in the wrong place can cause unwanted control inputs even though the sticks aren't moved. There's a guy who used a magnetic usb connector on the side and he recently posted it gave him 30% input. It would also affect scaling so even a small stick movement could end as 100%.
AFAK the MA controller isn't one of them, but there may be multiple versions.
At least that's the only thing that would explain "not touching the sticks" and "getting stick input logged".

 

[Lon Denard]

On the controllers that use magnetic sticks having a magnet close to them in the wrong place can cause unwanted control inputs even though the sticks aren't moved. There's a guy who used a magnetic usb connector on the side and he recently posted it gave him 30% input. It would also affect scaling so even a small stick movement could end as 100%.
AFAK the MA controller isn't one of them, but there may be multiple versions.
At least that's the only thing that would explain "not touching the sticks" and "getting stick input logged".

Stop being so smart! Yea I've seen that reported on other remotes so I'd guess it's possible with this one as well.

 

[Olav van Gerven]

This funny, or actually it is NOT funny but:
I observed several times this week that after about 45 to 60 seconds after take off, the drone would suddenly -without any control action- make a left or right yaw of about 80°. I posted the log and from what was read there, it showed a strong ya move. Now I am old and I might forget things and yes, I am willing to admit that I might have done that once, but from 10 flights I observed it more then once.
Then yesterday, I started a flight on a path between trees, ontedning to move slowly up so I could get a shot of several big big piles of wood ready to be transported. Again, I oberserved the move and then the drone started drifting. In a reflex, I pulled it up maximum speed, and avoided the trees (but only just), then regained control and brought it back in.
My thought reading about this crash was "hey, it seems not only to happen to me" and second, I wonder if it might be caused by having the drone when I'm hiking in a backpack. I never got the message to callibrate anything, but could it be, that thsi constant transport confueses the compass or IMU? (and now, don't shoot me if it is not, please)

 

[BudWalker]

This funny, or actually it is NOT funny but:
I observed several times this week that after about 45 to 60 seconds after take off, the drone would suddenly -without any control action- make a left or right yaw of about 80°. I posted the log and from what was read there, it showed a strong ya move. Now I am old and I might forget things and yes, I am willing to admit that I might have done that once, but from 10 flights I observed it more then once.
Then yesterday, I started a flight on a path between trees, ontedning to move slowly up so I could get a shot of several big big piles of wood ready to be transported. Again, I oberserved the move and then the drone started drifting. In a reflex, I pulled it up maximum speed, and avoided the trees (but only just), then regained control and brought it back in.
My thought reading about this crash was "hey, it seems not only to happen to me" and second, I wonder if it might be caused by having the drone when I'm hiking in a backpack. I never got the message to callibrate anything, but could it be, that thsi constant transport confueses the compass or IMU? (and now, don't shoot me if it is not, please)

This incident probably was caused by launching in OPTI mode in a visually challenging environment. Until 67 secs there 6 or less sats and the MA was using the vision system, not GPS, to navigate. Looking at the environment it's not surprising that there was some drifting. At 67 secs the FC had seen enough GPS data to 1) raise the navHealth (aka gpsHealth) from 0 to 4, and 2) start using the GPS to navigate.


After 67 secs the log doesn't show any drifting.

Did the other incidents happen in flights where the MA had to rely on the vision system at launch?

 

[Olav van Gerven]

This incident probably was caused by launching in OPTI mode in a visually challenging environment. Until 67 secs there 6 or less sats and the MA was using the vision system, not GPS, to navigate. Looking at the environment it's not surprising that there was some drifting. At 67 secs the FC had seen enough GPS data to 1) raise the navHealth (aka gpsHealth) from 0 to 4, and 2) start using the GPS to navigate.
View attachment 77901

After 67 secs the log doesn't show any drifting.

Did the other incidents happen in flights where the MA had to rely on the vision system at launch?

Thanks for looking into this.
The other incidents happened in open field. Normally I wait for the message "homepoint has been recorded" and the green signal on the screen about the GPS. I'm pretty new to flying the MA, so this was my first flight in a relatively narrow area. Realizing that between the high trees it could be difficult to get sufficient GPS, I took of hoping to get better GPS getting a little bit higher. Then indeed I got the GPS and the message "homepoint" and was able to pull up the drone further. Afterall seemed to have been not the best idea, though I'm a little bit proud that I pulled up in a refelex and avoided the trees. This was one of the things that I learned from reading this forum and watching other people crash.

 

[BudWalker]

Thanks for looking into this.
The other incidents happened in open field. Normally I wait for the message "homepoint has been recorded" and the green signal on the screen about the GPS. I'm pretty new to flying the MA, so this was my first flight in a relatively narrow area. Realizing that between the high trees it could be difficult to get sufficient GPS, I took of hoping to get better GPS getting a little bit higher. Then indeed I got the GPS and the message "homepoint" and was able to pull up the drone further. Afterall seemed to have been not the best idea, though I'm a little bit proud that I pulled up in a refelex and avoided the trees. This was one of the things that I learned from reading this forum and watching other people crash.

Drifting is the subject of both your post here, including logs, and this thread. The other incidents you described are a different phenomena (uncommanded yawing) that are being addressed in a different thread
Strange move
If you would like to continue pursuing the uncommanded yawing phenomena you should continue posting in that thread. I suggest that you heed @Fly Dawg 's request for a more concise statement of the problem and log files that illustrate the problem.

 

[badbrain]

Thanks people, for “translating” the data and the smart arse remarks

No magnets. Memory is 100%. No pilot error. Sometimes things go wrong and not always pilot’s fault at all.

Went to DJI shop (only one in France) and had their tech team look it over, data and machines.

Remote problem. “Something” was sticking in the remote, causing it to block and over exaggerate input, and move on it’s own.
Wouldn’t tell me exactly. Oh well, knew it wasn’t me, especially when I have to remote calibrate every power on/battery change.

Actually have a video of remote “moving” 35%, on it’s own.

Get it looked at “professionally” Olav.

Add that to the list of very rare, but possible, incidents.

New MA and remote next week. Yippee!!

 

[badbrain]

Finally received it. Proof of zero pilot error and a wonky remote.

Anyone with drifting or remote problems, film it (I did) and send it to DJI, if under warranty. Luckily EU has a 2 year one.