Need help. DJI won't cover flyaway crash. Ignores flight data.
- Thread starter miltondave
- Start date
The compass seems to have been reporting the heading incorrectly. Compare the last section if the footage with the DJI GO app replay and its clear that the aircraft heading depicted on the map is significantly different to the actual AC orientation revealed by the camera. To the extent your stick inputs appear ignored it might be they were applied in the direction of flight given the compass error.
The influence of the bridge on the compass is not unexpected. Your only option is to switch to ATTI mode and fly well clear of the structure. I have been in a situation where my heading was almost 180 deg out from actual, the pitch controls were reversed and I had to land like that. It wasn't fun. The heading depicted in the map did show the error though so I could comoensate.
The influence of the bridge on the compass is not unexpected. Your only option is to switch to ATTI mode and fly well clear of the structure. I have been in a situation where my heading was almost 180 deg out from actual, the pitch controls were reversed and I had to land like that. It wasn't fun. The heading depicted in the map did show the error though so I could comoensate.
I'm not sure the heading direction on the app is accurate though. When I turned the map on my app The Heading also turned with my map orientation. See these screenshots Imgur: The most awesome images on the Internet
But if the compass was acting incorrectly, why would the aircraft accelerate in the way that it did? Why didn't it stop moving when I pressed the RTH function? Why didn't it ascend to the preset height? And why did it descend into the water when it was a 165 feet from the home point? The AC clearly knew it wasn't over the home point. I'm not trying to call you out here, I'm just genuinely curious. I really appreciate your feedback.
And what really makes me suspicious, is that DJI continually refuses to acknowledge the events before the RTH function. Even when I call them out on it. Over the phone or through the complaint process they just act like I didn't even mention the events prior to the RTH. Makes me think they know it was a malfunction... but maybe I'm reading too much into it
- Joined
- Jan 2, 2017
- Messages
- 610
- Reaction score
- 303
- Age
- 62
The final piece of the puzzle no one seems to have gotten: VPS doesn't work over water. ESPECIALLY very smooth, mirror-like water.
What happens is the cameras see a reflection of the Phantom that moves perfectly with it so it looks stationary. The flight controller thinks it's fixed stationary over a point on the ground, when it is in fact moving and drifting all over the place.
If it encounters this condition when moving fast already (like under stick command, transitioning from shore to mirror water surface) it will just continue on in whatever direction its momentum sends it, like a hockey puck on ice.
This, of course, creates all sorts of inconsistencies with what's coming from GPS, confusing things even more.
I've directly experienced this issue with VPS, thankfully without any consequence. I've watched the video several times, and the behavior once over the water is EXACTLY how mine behaved. Thankfully I switched to ATTI within a few seconds, and got control back.
There's a setting to turn VPS off for exactly this reason (low-level flight over water), and I believe water is even noted in th comments in the setting for a reason to turn it off.
Unfortunately, this one looks like Pilot Error.
What happens is the cameras see a reflection of the Phantom that moves perfectly with it so it looks stationary. The flight controller thinks it's fixed stationary over a point on the ground, when it is in fact moving and drifting all over the place.
If it encounters this condition when moving fast already (like under stick command, transitioning from shore to mirror water surface) it will just continue on in whatever direction its momentum sends it, like a hockey puck on ice.
This, of course, creates all sorts of inconsistencies with what's coming from GPS, confusing things even more.
I've directly experienced this issue with VPS, thankfully without any consequence. I've watched the video several times, and the behavior once over the water is EXACTLY how mine behaved. Thankfully I switched to ATTI within a few seconds, and got control back.
There's a setting to turn VPS off for exactly this reason (low-level flight over water), and I believe water is even noted in th comments in the setting for a reason to turn it off.
Unfortunately, this one looks like Pilot Error.
I turned VPS off for this very reason as per the owners manual and Tom's Tech Time recommendations. I was thinking I might fly out I've the river if I got confident and turned ff the VPS in case. Also there was snow on the ground and the temp was above freezing. I was worried that shinny snow might screw with the VPS.
- Joined
- Jan 2, 2017
- Messages
- 610
- Reaction score
- 303
- Age
- 62
Well, then that wasn't it. The behavior over the water was exactly like a fooled VPS, though.
Wouldn't it be great to nail a bug in DJI GO, that you flipped the switch, but VPS stayed on? Board intelligensia, any way for us to determine that?
- Joined
- Jan 11, 2017
- Messages
- 159
- Reaction score
- 83
- Age
- 41
It definitely hit the sticks/grass just before heading over the river, you can see them move after it hits them. I am totally out of my element here but based on that, plus flying near/under the bridge, low over water, and minimal flight time under your belt, I could see Dji sticking with pilot error. Shitty.
Sent from my iPhone using PhantomPilots
I can only my give you my anecdotal assurance but the drone definitely did not hit the grass. I watched it the whole time it flew away. The grass is moving due to the downward force of the rotors. Even if it did hit, it was well after the beginning of the flyaway and it would've only grazed the landing gear. I would hope that wouldn't cause the drone to go crazy.
And yea I can see them holding onto the pilot error story. But the fact is that many people fly these drones around, under, and over Large bridges and building without issue. Also, DJI hasn't mentioned the bridge and ignores the beginning of the flyaway, choosing to focus only on the RTH and nothing else.
- Joined
- Dec 29, 2016
- Messages
- 97
- Reaction score
- 16
- Age
- 61
Sorry that happen to you I had a major crash myself but mine is pilot errand I saw all the information you have look like something mess up your compass and drone became confused and was trying to remember the flight path wish you all the best mike
Sent from my iPad using PhantomPilots
Sent from my iPad using PhantomPilots
Callings me out? lol.... I'm happy to be corrected and learnt something here. Why the depicted heading might change when moving the device is beyond me (you have seemingly demonstrated it does). I had compared the heading at start and towards the end of the clip and was recounting my own experience for comparison. In my case the heading was reversed from actual due to the environment I was flying in. I must confess I don't replay my flights in the Go app. I will be now though for curiosity value.
I have flown near and under things since but only in ATTI mode and close enough to see the orientation of the AC. I set RTH behaviour to hover also, last thing I want is an auto ascent into an object above.
Are you able to upload the flight to healthydrones.com. Would be interesting to see if there are any different clues there.
- Joined
- Sep 2, 2016
- Messages
- 262
- Reaction score
- 52
- Age
- 45
Lol yea... I just wanted to make sure I didn't sound dickish or over defensive
Here the HD link HealthyDrones.com - Innovative flight data analysis that matters
TBH I don't remember what the mode I had it on was called. I did change it because I didn't like the default mode. It felt backwards.
Yes, stay away from rebar and concrete. Not good at all for the compass.
Sent from my SM-G900V using PhantomPilots mobile app
- Joined
- Sep 2, 2016
- Messages
- 262
- Reaction score
- 52
- Age
- 45
If the brigde was interfering with compass it would tell u to recalabrate,,,or u get in flight warning,,
- Joined
- Sep 2, 2016
- Messages
- 262
- Reaction score
- 52
- Age
- 45
DJI Phantom 3 Flyaway crash:
Once the phantom has compass malfunction on the air, the firmware logic calculates the wrong result. Even just for hovering (with no input command), it will drive it to wrong dirrection. That's what causing "toilet bowl effect".
The dangerous thing is, the phantom didn't even realize that its compass is malfunctioning, so it won't show any error message and keep calculating wrong values of trajectory over and over.
How to avoid this is make sure your actual heading is match with the heading on the map. Maybe only your self who can confirm the situation on that location, between the onboard video vs the flightlog video on your 1'st post. Make sure it is perfectly match the orientation (N,E,S,W) not only the rotation movement on the map.
And this youtube video i post here is one example, that phantom didnot realize its compass malfunction, so it won't show any error message. The worst thing is the pilot also didnot aware about it.
I believe this is exactly match your case, i can see it from the same "toilet bowl" trajectory.
Sent from my D5503 using PhantomPilots mobile app
Last edited:
- Joined
- Sep 2, 2016
- Messages
- 262
- Reaction score
- 52
- Age
- 45
Lucky me, everything went as expected. No compass error, no heading mismatch (actual vs map), RTH right on the homepoint next to where i stand.
There also electrical powerline beside the bridge.
A lot of good memories with my old P3s...
Sent from my D5503 using PhantomPilots mobile app
Last edited:
Similar threads
