I recently lost my DJI Mini 3 Pro. I sent it flying at the bottom of a famous scenic spot in the Tiger Leaping Gorge in China. I always had direct line of sight, and the flight recorder from the DJI remote shows I always had a 5-bar signal strength (heck, I could even still start downloading the video to the remote when it was down... unfortunately, it was only 400k/s, and the drone shut down before I could download the full-quality video)

Here's how it went:


I sent it around 450 meters away into the gorge before stopping. I hovered for a few seconds, then turned 180° to prepare for flying back. A few seconds more went by, and I pressed record before pushing forward to come back. I immediately noticed something seemed strange. The drone seemingly didn't move forward, and I realized it was drifting to the right and was about to hit the gorge wall. I tried at the last second to steer it to the left away from the wall, but it seems like there was no response, and it crashed into the cliff wall and fell into an unrecoverable place.


From the flight recording and the preview video that I retrieved from the remote, the drone responded to yaw commands but was clearly unresponsive to forward and lateral inputs. Also, there was little wind in the gorge, and I can't believe a gust of wind pushed it so consistently that way. I also never encountered a situation where the wind was strong enough so that a stick input would not be able to counter it (it's impossible the wind was over 40 km/h; you can look at the plants in the video if you don't believe me)

I understand that the flight recording shows a weak GPS signal (it went stable around 13 satellites, below 16), and that hovering might be unstable. What I don't understand is that the drone did not respond to my stick input. It didn't go into ATTI mode and was always in P-GPS.

My drone is not under the DJI Care plan anymore, so I am not expecting any compensation here. I just want to understand what happened because that doesn't give me confidence in buying a new DJI drone. I am also an FPV pilot, so maybe I'll just stick to FPV...


For anyone who wants to have a look, the video below contains the flight recording analysis with the two videos I captured during the flight (including the last seconds).
https://youtu.be/9M5u8ltYY0E?si=yqlN1wTWOdiwrkAW&t=105

The full flight record from the DJI remote is also available here.
https://drive.google.com/file/d/1bCI5ePqldLHtm04P5I105UNT6cOklxWb/view?usp=sharing

It would be awesome to understand what happened.

Sorry, the links were not properly formatted:

Video here:
https://youtu.be/9M5u8ltYY0E?si=yqlN1wTWOdiwrkAW&t=105

Flight recording here:
https://drive.google.com/file/d/1bCI5ePqldLHtm04P5I105UNT6cOklxWb/view?usp=sharing

It would be awesome to understand what happened.

It seems to me that the main reason is the loss of GPS signal, despite the fact that the data showed 10-13 satellites, the signal from them was unstable and the data was only from satellites directly above the drone, because it was blocked by mountains. This is a bad situation. Stabilization by lower sensors did not work because the drone was above the water flow.

Serg SSA Posted at 9-20 07:09
It would be awesome to understand what happened.

It seems to me that the main reason is the loss of GPS signal, despite the fact that the data showed 10-13 satellites, the signal from them was unstable and the data was only from satellites directly above the drone, because it was blocked by mountains. This is a bad situation. Stabilization by lower sensors did not work because the drone was above the water flow.

I agree the loss of GPS signal is the likely cause of the issue. However, from what I understand, the drone should try to hover in that case  (with potential drift because of the wind) and still respond to stick inputs, not go full speed to the side and back when I'm trying to move forward.

djules75 Posted at 9-20 07:30
I agree the loss of GPS signal is the likely cause of the issue. However, from what I understand, the drone should try to hover in that case  (with potential drift because of the wind) and still respond to stick inputs, not go full speed to the side and back when I'm trying to move forward.

This is unfortunate, I believe if you fill out a flyaway report, they will give you a new drone, or at least 40% off compensation if you buy another one. And I'm also assuming you did not retrieve the drone?

djules75 Posted at 9-20 07:30
I agree the loss of GPS signal is the likely cause of the issue. However, from what I understand, the drone should try to hover in that case  (with potential drift because of the wind) and still respond to stick inputs, not go full speed to the side and back when I'm trying to move forward.

However, from what I understand, the drone should try to hover in that case  (with potential drift because of the wind) and still respond to stick inputs, not go full speed to the side and back when I'm trying to move forward.

If the GPS signal is not stable, the drone cannot hover and hold the GPS position, the drone can use the downward vision system to hold the position, but only if the surface below allows it. You had a water flow below, it confused the downward vision system and the drone began to move with the wind or tried to hold the position with the water flow.
Here's an example, the drone, even with a good GPS signal over water, began to move to the side due to ripples on the water. I barely managed to avoid a collision with trees.

djules75 Posted at 9-20 07:30
I agree the loss of GPS signal is the likely cause of the issue. However, from what I understand, the drone should try to hover in that case  (with potential drift because of the wind) and still respond to stick inputs, not go full speed to the side and back when I'm trying to move forward.

If the drone decides that the GPS quality is poor and VPS is unavailable then the drone will switch to ATTI mode.

In ATTI mode the drone will, in the absence of any stick inputs, fly level but will NOT fight wind, thus is can remain level yet be blown about by any wind.

Your google page wants me to log in or something.
You would do better to ulpoad that log to
https://www.phantomhelp.com/LogViewer/Upload/
and post the resulting URL here.

The problem with GPS in canyons, deep gorges or anywhere that a large part of the sky is blocked by terrain, is that you can't get a good spread of sats which is needed to provide good location data.
If you are only able to see sats that are bunched in a small part of the sky, your GPS reliability falls.
Your number of sats was never

From 2:17.4 your GPS reliability fell to 2/5 and stayed there except for two 3-second intervals.
The drone only has full GPS reception when the GPS reliability is 4 or 5.
The VPS data shows that the downward sensors were able to detect the eight of the drone above the ground below it.
But unless the lighting was good AND the ground texture was suitable, VPS won't be able to provide horizontal position holding.

Despite the display indicating the drone was in P-GPS mode, with GPS reliability of only 2/5, GPS could not help with horizontal position holding.
If VPS also couldn't help, your drone would have been in Atti Mode, subject to drifting and without any "brakes".
Despite this you were using full stick moves and flying at 10 m/s with the drone 473 metres away and out of sight when the drone collided and tumbled down.

Flying at speed, close to obstacles and without GPS is always risky.
Obstacle avoidance is unable to help without GPS.
In this case, the outcome was not surprising.

djules75 Posted at 9-20 07:30
I agree the loss of GPS signal is the likely cause of the issue. However, from what I understand, the drone should try to hover in that case  (with potential drift because of the wind) and still respond to stick inputs, not go full speed to the side and back when I'm trying to move forward.

However, from what I understand, the drone should try to hover in that case  (with potential drift because of the wind) and still respond to stick inputs, not go full speed to the side and back when I'm trying to move forward.

I also noticed a peculiarity: when the drone loses GPS stabilization, it stops responding to the right stick for a few seconds, but continues to respond to the left stick.

Sean-bumble-bee Posted at 9-20 11:11
If the drone decides that the GPS quality is poor and VPS is unavailable then the drone will switch to ATTI mode.

In ATTI mode the drone will, in the absence of any stick inputs, fly level but will NOT fight wind, thus is can remain level yet be blown about by any wind.

Log here:
https://www.phantomhelp.com/logviewer/20240920SFF6F0BRNQ8V95LB9NWAIJG5UKJORYA0#

The youtuve video is also a replay of the log with the camera recordings from the drone synced and layered on top.

Labroides Posted at 9-20 12:16
The problem with GPS in canyons, deep gorges or anywhere that a large part of the sky is blocked by terrain, is that you can't get a good spread of sats which is needed to provide good location data.
If you are only able to see sats that are bunched in a small part of the sky, your GPS reliability falls.
Your number of sats was never

I was not flying at speed. The drone was hovering just before it started drifting by itself and I tried to regain control. If you look at the youtube video I posted starting minute 1:45 where you can see the camera recording from the drone along with the flight recording. I am trying to go forward full stick and last second to the left to avoid the wall. You can clearly see there is no forward speed registered.

Also, it was not windy. Even if it was a gust of wind, then it would need to be over 40km/h constantly to counter my forward command and have a zero or negative absolute forward speed.

I have always been careful with windy conditions, but even when the wind was greatest, I could always move the drone against the wind with manual stick commands.

Labroides Posted at 9-20 12:16
The problem with GPS in canyons, deep gorges or anywhere that a large part of the sky is blocked by terrain, is that you can't get a good spread of sats which is needed to provide good location data.
If you are only able to see sats that are bunched in a small part of the sky, your GPS reliability falls.
Your number of sats was never

That's interesting. In my case, that would have been the opposite. I managed to apply yaw left and right with the left stick, but the forward and left movements from the right stick didn't seem to have any effect.

djules75 Posted at 9-21 04:36
I was not flying at speed. The drone was hovering just before it started drifting by itself and I tried to regain control. If you look at the youtube video I posted starting minute 1:45 where you can see the camera recording from the drone along with the flight recording. I am trying to go forward full stick and last second to the left to avoid the wall. You can clearly see there is no forward speed registered.

Also, it was not windy. Even if it was a gust of wind, then it would need to be over 40km/h constantly to counter my forward command and have a zero or negative absolute forward speed.

I was not flying at speed. The drone was hovering just before it started drifting by itself and I tried to regain control.
The recorded flight data shows a speed of 10 m/s which would be consistent with your joystick inputs.
Your drone didn't "start drifting by itself".
You started it moving forward and expected it to respond as it does with full GPS.
But without any horizontal position holding, your drone will continue to drift when you go hands-off.
It has no way to brake or rapidly change direction.
It's like driving on ice with no brakes.

I am trying to go forward full stick and last second to the left to avoid the wall.
But you also said:
I was not flying at speed. The drone was hovering just before it started drifting by itself and I tried to regain control.
You are confused about what you were doing.
According to the recorded flight data ...
At 3:28.4 you pushed the right stick fully forward and still held it there until two seconds after the collision.
You  also had the left pushed stick fully to the left from 1.2 seconds  before colliding, not releasing until two seconds after the collision.
This accelerated to a speed of 10 m/s at the time of collision.

I agree the loss of GPS signal is the likely cause of the issue.
Losing GPS doesn't cause crashing.
Flying inappropriately close to obstacles, without an understanding of how the drone performs without GPS causes crashes.

However, from what I understand, the drone should try to hover in that case  (with potential drift because of the wind)
Your understanding is flawed.
WIthout any horizontal position holding, your drone will continue to drift when you go hands-off.
It has no way to brake.

... and still respond to stick inputs, not go full speed to the side and back when I'm trying to move forward.
Your drone did respond to your joystick inputs, but not to your expectation.
When flying without GPS, your drone doesn't have any brakes.

Also, it was not windy.
I never suggested wind played any part in the incident.
It was all down to poor piloting and lack of how the drone flies without GPS.

Labroides Posted at 9-21 15:17
I was not flying at speed. The drone was hovering just before it started drifting by itself and I tried to regain control.
The recorded flight data shows a speed of 10 m/s which would be consistent with your joystick inputs.
Your drone didn't "start drifting by itself".

"You are confused about what you were doing.
According to the recorded flight data ...
At 3:28.4 you pushed the right stick fully forward and still held it there until two seconds after the collision.
You  also had the left pushed stick fully to the left from 1.2 seconds  before colliding, not releasing until two seconds after the collision.
This accelerated to a speed of 10 m/s at the time of collision."


Have you looked at the youtube video I posted?


You appear to only be looking at the flight recording data without looking at the camera recording from the drone.
You are telling me that I flew straight into the wall full speed. This is not what the video will tell you. I was trying to move away from it.
The cliff wall was to the RIGHT of the drone, so, of course, I was trying to move left.
Are you telling me that moving left accelerated the drone to the right?...

djules75 Posted at 9-21 17:26
"You are confused about what you were doing.
According to the recorded flight data ...
At 3:28.4 you pushed the right stick fully forward and still held it there until two seconds after the collision.

By the way, this is the exact reason why I made this youtube video with the footage from the drone (which is the preview video from the remote since I lost the drone). Without the footage, your conclusion would be difficult to contradict.

https://youtu.be/9M5u8ltYY0E?si=vzm853SP82rKsXJe&t=105

djules75 Posted at 9-21 04:27
Log here:
https://www.phantomhelp.com/logviewer/20240920SFF6F0BRNQ8V95LB9NWAIJG5UKJORYA0#

Thanks......

djules75 Posted at 9-21 17:26
"You are confused about what you were doing.
According to the recorded flight data ...
At 3:28.4 you pushed the right stick fully forward and still held it there until two seconds after the collision.

You are telling me that I flew straight into the wall full speed. This is not what the video will tell you. I was trying to move away from it.
The cliff wall was to the RIGHT of the drone, so, of course, I was trying to move left.
Twice I've pointed out that you had the right stick full forward and the left stick full left until well after the crash.
The right stick moves the drone forward.
The left stick rotates the drone anti-clockwise .. it does not move the drone to the left.
This is very basic stuff that any drone pilot should inderstand..

Are you telling me that moving left accelerated the drone to the right?
I've tried to explain to you how the drone behaves without horizontal position holding but you aren't understanding.
You propelled the drone forward and also commanded it to rotate anti-clockwise.
1.  That doesn't move the drone left .. and
2.  That doesn't eliminate the drone's forward momentum.
Without horizontal position holding, the drone continues in the direction it is travelling.

You asked for help understanding the circumstances of your crash.
I analysed the data to find out what actually happened and gave you a detailed explanation, which I have had to repeat in even more detail because to don't understand drone basics and you don't want to accept blame for the crash you caused.

You may be in denial, but the recorded data is clear.
You crashed your drone into an obstacle.
You were flying close to an obstacle at speed.
Contributing factors were that you were too far away to see the drone properly.
You didn't have the benefit of horizontal position holding that GPS gives
And you didn't understand that your drone would behave differently without GPS.

Without the footage, your conclusion would be difficult to contradict.
The recorded data is very difficult to contradict, but you want to ignore the story it tells.

djules75 Posted at 9-21 04:38
That's interesting. In my case, that would have been the opposite. I managed to apply yaw left and right with the left stick, but the forward and left movements from the right stick didn't seem to have any effect.

This is the effect I was talking about, I observed this behavior of the drone. Now the labroid will tell me that I also do not know how to control the drone) But I have a lot of practice in controlling FPV drones and I know how the drone reacts to sticks without GPS

djules75 Posted at 9-21 17:38
By the way, this is the exact reason why I made this youtube video with the footage from the drone (which is the preview video from the remote since I lost the drone). Without the footage, your conclusion would be difficult to contradict.

https://youtu.be/9M5u8ltYY0E?si=vzm853SP82rKsXJe&t=105

It might be useful if you can retrieve the DAT flight log from the screen device. The log's name will have the same date stamp as the .txt flight log and a similar time stamp, though the time stamp may be in a 12hr clock rather than a 24hr clock.
The DAT's name will also contain "FLY234".

If you have the DAT then upload it to a file hosting website and post the URL here. There is a slim chance that the DAT may be readable but I offer no guarantees.

It might also be useful to you to post your storry on
https://mavicpilots.com/forums/crash-flyaway-assistance.85/
along with the Phantomhelp ( PH ) URL and the  URL of the DAT's page.
Alternatively I recollect MavicPilots permits the upload of the .txt and DAT flightlogs directly to the website though it wiould save PH upload traffic if you just post teh PH URL

With regards to video, don't bother with including a replay of the log, just show the entirety of the last video shot by the drone, which is about 11 seconds long

Serg SSA Posted at 9-22 02:44
This is the effect I was talking about, I observed this behavior of the drone. Now the labroid will tell me that I also do not know how to control the drone) But I have a lot of practice in controlling FPV drones and I know how the drone reacts to sticks without GPS

I didn't see you commenting on the recorded flight data.
Did you bother to analyse it yourself?

Labroides Posted at 9-21 22:49
You are telling me that I flew straight into the wall full speed. This is not what the video will tell you. I was trying to move away from it.
The cliff wall was to the RIGHT of the drone, so, of course, I was trying to move left.
Twice I've pointed out that you had the right stick full forward and the left stick full left until well after the crash.

Thanks for assuming I don’t understand drone basics. It’s not because I’m level one here that I am a beginner. My drone is registered in China, where I currently live, which uses a different forum. I’m here on the English forum to be able to discuss in a language I’m more comfortable with.

I also fly FPV by the way, and I have a pretty good understanding and experience of momentum, including in windy conditions.

Now, you told me you carefully analyzed the data I am really sorry to ask you to go over it again with me step by step.

After flying forward into the gorge:
- Around 3:01 flight recorder data, I stop moving forward and stop recording the video. Just before stoping the video, you can see the drone coming to a standstill and hover.
- The flight data also confirms a speed close to zero from 3:02 to 3:16 (so long for the momentum, which if I understand correctly, is what carried my drone towards the gorge wall…). I switch to picture mode and take a picture of the gorge.
- At 3:16, I go full yaw left to turn the drone around 180 degrees and I leave it hovering while I take a picture another picture.
- From 3:16 to 3:28, there is a 1.5-2 mph drift that I didn’t notice when looking at the remote. Likely because of the wind and poor GPS signal.
- At 3:27, I press video record and move the right stick forward to come back. The drone starts picking up speed… but not forward. Instead, it starts moving to the right and slightly backwards (it was hovering just a couple of seconds before that).
- At 3:31, I yaw slightly to the left so that by going forward and back it would fly away from the cliff wall (yes, not the best move here, but the drone should still have picked up some speed forward into the canyon).
- At 3:32, I’m trying a last ditch effort to steer the drone to the left with a full left on the right stick.
- Just after the crash, I keep the right stick forward and left for two seconds because… well… I WAS REALLY UPSET and couldn’t believe what just happened.


Again, clearly, you haven’t look at the YouTube video. I will post the relevant segments directly below.

At the last second, I am applying left lateral movement with the right stick full left. And you’re telling me that has accelerated the drone into the wall.
(By the way, even though I had been applying full forward movement from 3:27, the video also doesn’t show the drone moving forward in any way).


Oh, and this:
“Twice I've pointed out that you had the right stick full forward and the left stick full left until well after the crash.
The right stick moves the drone forward.
The left stick rotates the drone anti-clockwise .. it does not move the drone to the left.”

Thanks for being condescending, but it’s simply wrong. I didn’t do that and the flight data shows it. look again. Plus, if that was so, the drone would have rotated on itself going into the crash, which it didn’t.

djules75 Posted at 9-22 04:51
Thanks for assuming I don’t understand drone basics. It’s not because I’m level one here that I am a beginner. My drone is registered in China, where I currently live, which uses a different forum. I’m here on the English forum to be able to discuss in a language I’m more comfortable with.

I also fly FPV by the way, and I have a pretty good understanding and experience of momentum, including in windy conditions.

Sean-bumble-bee Posted at 9-22 04:18
It might be useful if you can retrieve the DAT flight log from the screen device. The log's name will have the same date stamp as the .txt flight log and a similar time stamp, though the time stamp may be in a 12hr clock rather than a 24hr clock.
The DAT's name will also contain "FLY234".

Thanks. I will try to do that tomorrow, I don't have the remote handy.
Also uploading the relevant parts of the video below, because without seeing the footage from the drone along with the flight data, I agree that it's difficult to understand, which is why I posted this story in the first place.

Labroides Posted at 9-22 04:27
I didn't see you commenting on the recorded flight data.
Did you bother to analyse it yourself?

If I haven't published my analysis here, it doesn't mean that I haven't analyzed the flight )))
Besides, unlike you, I also watched the video. And most importantly, I also got into similar situations and I know from practice that the drone doesn't always work out the movement of the sticks, there is a certain delay, especially for the right stick.

Serg SSA Posted at 9-22 05:07
If I haven't published my analysis here, it doesn't mean that I haven't analyzed the flight )))
Besides, unlike you, I also watched the video. And most importantly, I also got into similar situations and I know from practice that the drone doesn't always work out the movement of the sticks, there is a certain delay, especially for the right stick.

So you didn't analyse the data and don't believe what the data clearly shows?

djules75 Posted at 9-22 04:51
Thanks for assuming I don’t understand drone basics. It’s not because I’m level one here that I am a beginner. My drone is registered in China, where I currently live, which uses a different forum. I’m here on the English forum to be able to discuss in a language I’m more comfortable with.

I also fly FPV by the way, and I have a pretty good understanding and experience of momentum, including in windy conditions.

I didn’t do that and the flight data shows it.
OK .. I have to apologise.
I did my initial assessment of the data in a hurry and made a mistake.
You didn't push the left stick, you pushed the right stick forward and left.
But that doesn't change the assessment.
You gave the drone forward momentum and tried to alter course towards the left but still kept up the forward momentum.
The drone picked up some leftward momentum, but without position holding, the drone also kept its forward momentum.
That's what happens when you have lost horizontal position holding ability.
Your drone kept skidding forwards and only went a little left.

My assessment, which is confirmed by the data remains.
The recorded data is clear.
You crashed your drone into an obstacle.
You were flying close to an obstacle at speed.
Contributing factors were that you were too far away to see the drone properly.
You didn't have the benefit of horizontal position holding that GPS gives
And you didn't understand that your drone would behave differently without GPS.

Labroides Posted at 9-22 07:17
I didn’t do that and the flight data shows it.
OK .. I have to apologise.
I did my initial assessment of the data in a hurry and made a mistake.

And you didn't understand that your drone would behave differently without GPS.

The author told you that he flies FPV drones, so he understands how a drone behaves without GPS, I also have such experience and it is clear to me how to react to deviations. I don’t think that the author made stupid mistakes, and the video confirms that he controlled the sticks correctly. I wonder if you yourself fly FPV and have you ever found yourself in similar situations?
I’ll repeat once again, I noticed a situation where in such moments for 1-2 seconds the drone reacts only to the left stick.
Has anyone noticed something similar?
Here's an example: I tried to use the stick to compensate for the drone's leftward movement, but the drone didn't respond to the right stick, only the left stick allowed me to turn it and not crash into a tree.

djules75 Posted at 9-22 04:51
Thanks for assuming I don’t understand drone basics. It’s not because I’m level one here that I am a beginner. My drone is registered in China, where I currently live, which uses a different forum. I’m here on the English forum to be able to discuss in a language I’m more comfortable with.

I also fly FPV by the way, and I have a pretty good understanding and experience of momentum, including in windy conditions.

Hi

Had a look at your log and think wind and slow reactions on your part were the main reason for your loss.

Have attached a few items to show you my reasoning.

djules75 Posted at 9-22 04:53
Thanks. I will try to do that tomorrow, I don't have the remote handy.
Also uploading the relevant parts of the video below, because without seeing the footage from the drone along with the flight data, I agree that it's difficult to understand, which is why I posted this story in the first place.

"I agree that it's difficult to understand, which is why I posted this story in the first place."  I do not think the replay on the log incorporated with the flight bvideo helps anything especially if the person looking has a relatively small screen, e.g. a laptop.

The attached show the results from processing the PH CSV with CsvView and it seems that the left right yaw tewitch seen in the video matches to the aileron inputs between 209 seconds and 212 seconds. If that is correct then the other stick commands fall into place.

The second attachment shows the GPS count and quality for the whole flight.

Pleomax Posted at 9-22 08:41
Hi

Had a look at your log and think wind and slow reactions on your part were the main reason for your loss.

Hi and many thanks for the detailed log visualization.

I agree with you that it is the most probable explanation. The gorge was getting narrower there, and it's possible that a gust of wind might have arisen even though it felt calm from where I stood.

From the wind data, at 9m/s, moving forward full stick would have barely been able to compensate for the wind speed in N mode (10m/s). This would explain why I felt the drone was not moving forward at all. This also would match the trajectory curve in the last seconds (absolute forward speed = zero, lateral speed too high to fight with a last second reaction from my part with the right stick to the left command).

In that case, I guess that had I promptly switched to sports mode, I could have saved the day.

Could you help me confirm one last thing though?
1. Having wind data means the drone had still GPS positioning (though weak), correct?
2. If GPS signal was weak, I guess this would affect the wind data and the wind might have been stronger than shown on your visualization. Is that also correct?

I have to admit that although it seems likely from your data that I fell victim to a gust of wind, I had a hard time going with this hypothesis because it wasn't windy where I stood and also, the drone kept recording for a minute with a view on the gorge's wall vegetation that stood still, with no sign of any wind. But yes, I know, it's a narrow gorge and I was 400m away, so what do I know...

Posting that last minute of the recording as a tribute to my lost drone, if you can bear looking at some still vegetation...

djules75 Posted at 9-22 09:46
Hi and many thanks for the detailed log visualization.

I agree with you that it is the most probable explanation. The gorge was getting narrower there, and it's possible that a gust of wind might have arisen even though it felt calm from where I stood.

Hi

Glad to have helped you.

Sorry for your loss.

Wind data is not derived from GPS AFAIK.

The wind data comes from the decrypted Flightlog csv by Phantom Help ( thanks to msinger) so I can't answer you.

djules75 Posted at 9-22 09:46
Hi and many thanks for the detailed log visualization.

I agree with you that it is the most probable explanation. The gorge was getting narrower there, and it's possible that a gust of wind might have arisen even though it felt calm from where I stood.

Wind 10 m/s is a very very strong wind. I do not see any signs of such wind in the video. I provided my video when the drone moved independently to the left and I tried to stop it with the right stick, but it did not work. The wind was 0 m/s and I was standing 8 meters from the drone. Please note, for example, if you are flying and press the stop button, then after that the drone will not respond to the right stick for 1-2 seconds. Perhaps something similar happens when GPS is lost.

Serg SSA Posted at 9-22 12:23
Wind 10 m/s is a very very strong wind. I do not see any signs of such wind in the video. I provided my video when the drone moved independently to the left and I tried to stop it with the right stick, but it did not work. The wind was 0 m/s and I was standing 8 meters from the drone. Please note, for example, if you are flying and press the stop button, then after that the drone will not respond to the right stick for 1-2 seconds. Perhaps something similar happens when GPS is lost.

Hi Serg SSA

If you watch the flight animation at approx 202 secs the wind from the SW direction starts to increase.

The AC is facing in the direction of the wind and starts to move back with the wind because there is no input from the RC.

Low quality GPS can not hold AC in position.

Wind speed tops out at 9 m/s and even with RC inputs can not overcome wind speed ending in the loss of the AC.

Had the operator switched to Sport Mode as he suggested the outcome may have been different.

Regarding your own incident, upload the flight log for analysis.

Today my drone took off after take off was in the sky 3 or 4 minutes and I just heard it speed up and sounded louder than usual and disappeared from sight lost connection with it couldn’t get it to return to home point found it few hrs ago crash landed in a man’s garden luckily I drew my phone number and ring if found on the top on a sticker guy rang me I was ecstatic to get it back but it’s absolutely ruined humble no longer exists in the drone 1 arm is snapped but actually still powers up and calibrates itself…. Absolutely devastated I suffer real bad social anxiety and anxiety and depression the drone got me back out side loving my life now is destroyed and spent everything I had too buy it  now sitting Looking at ways to glue it all back together and make if fly without a gimble but I know it’s not gonna fly again MEHHH

Pleomax Posted at 9-22 10:12
Hi

Glad to have helped you.

May I ask how you got this visualization? You wrote "thanks to msinger". I'm assuming this is @msinger user from Phantom Help. Correct?

Is it his own tool? Do you think I could ask him about how the wind data is derived?

Thanks in advance.

djiuser_IKJfGCdTh9Dg Posted at 9-22 15:47
Today my drone took off after take off was in the sky 3 or 4 minutes and I just heard it speed up and sounded louder than usual and disappeared from sight lost connection with it couldn’t get it to return to home point found it few hrs ago crash landed in a man’s garden luckily I drew my phone number and ring if found on the top on a sticker guy rang me I was ecstatic to get it back but it’s absolutely ruined humble no longer exists in the drone 1 arm is snapped but actually still powers up and calibrates itself…. Absolutely devastated I suffer real bad social anxiety and anxiety and depression the drone got me back out side loving my life now is destroyed and spent everything I had too buy it  now sitting Looking at ways to glue it all back together and make if fly without a gimble but I know it’s not gonna fly again MEHHH

If you want to find out the cause of your incident and how to prevent it happening again, start a new thread and post your flight data.

Go to: https://www.phantomhelp.com/LogViewer/Upload/
... where you'll find instructions.
That will give a report you can post a link for the report here.

Or if you use Airdata, post a link to the Airdata report for the flight.

Or just post the .txt file.

If you used the RC controller, you'll find your files here:
Android\data\dji.go.v5\files\FlightRecord

djules75 Posted at 9-22 17:35
May I ask how you got this visualization? You wrote "thanks to msinger". I'm assuming this is @msinger user from Phantom Help. Correct?

Is it his own tool? Do you think I could ask him about how the wind data is derived?

Hi

The visualization is my own. It is an Excel addin "EXCEL FLIGHT DISPLAY" that I have developed and has many more features including many charts.

It uses the decrypted csv of the flight log, the decryption is by Phantom Help Log Viewer.

I'm sure you could ask in the forum on PH as to how the wind is calculated but not all AC models provide the wind speed data.

Pleomax Posted at 9-22 13:51
Hi Serg SSA

If you watch the flight animation at approx 202 secs the wind from the SW direction starts to increase.

Regarding your own incident, upload the flight log for analysis.

Unfortunately, the log file was not saved, the flight was a long time ago. I got into situations with loss of GPS several times and each time the drone behaved in the first 1-2 seconds as if it did not respond to the right stick, then control returned and with my experience in FPV I could control the drone even without GPS.
I conducted experiments with a small GPS signal jammer for testing, but it has too small a range to get a reliable result.