Mini 3 Pro crash on takeoff - Help Required please

[BrAinZ]

I have owned various DJI drones for years now, and have made hundreds of successful flights without any previous problems.

A couple of weeks ago I was on holiday in Lapland and went to take a quick couple of low level shots across the hotel location with my Mini 3 Pro.
The Previous day, I had made a successful flight but had encountered a slight issue where the drone was not flying straight, but was drifting left or right, even when holding position, it was turning by itself. Because of that I recalibrated the controller sticks.

I did the same as I always do, and started the auto take off procedure, but instead of the props spinning up as normal, they only spun very slowly. As I was about to abort the take off to investigate the slow spinning props, the drone then suddenly spun up and shot forward and down, straight into a fresh pile of snow.
As the snow was very powdery, it went straight into the vents and all around the motors. I immediately dried it off as much as possible, and took it back to our room to dry off.

Although I did fly the drone the next day, I was concerned as to why it crashed, and also any longer term water damage. So I decided to return it to DJI under warranty.

Although they are saying the “Aircraft core board module malfunctioned”, they still want to charge me for the repair. They are saying that the crash was “likely” caused by an unstable platform at launch and/or an angle of -11°

Having done hundreds of launches I really feel that the drone did something very unusual, causing the crash. I had 32 satellites and home point set.

I will try to attach the flight data, and would really appreciate a second opinion.

 

[BrAinZ]

Here’s the Airdata link to the flight…

 

[Yorkshire_Pud]

The log on Phantomhelp = 3ADQAHV280G787FTUW4L

It might be that the DAT is readable, do you have it? If so, can you post it?
The time stamp should be similar to the time stamp of the .txt log but 'predate' the .txt by a few seconds and if probably contains FLY036.

That I remember there are several posts or threads that mention the Mini 3 is quirky when hand launching. You'd need to do a search for them but it might be better to hand launch using a CSC to start the motors and the throttle up to actually launch.

I do not have a Mini 3 with which to experiment but I have noticed that, launched via a CSC and throttle up from a sloping surface, other DJI drones first level themselves and then take off.
Your log starts with -13.7deg of pitch, I can never remember if that is nose up or nose down.

 

[slup]

Pretty sure that this was the culprit...

AIMU attitude restricted. Ensure aircraft is level (Code: 16100015).; Aircraft pitch angle too large. Ensure aircraft is level before taking off. (Code: 1610001E).

 

[BrAinZ]

Thanks, But surely this wouldn’t cause the drone to take off and flip over. If the software KNOWS the angle is too high, surely it should not try to take off. Simple enough programming to check that.

It’s also weird that the drone props were spinning slowly before going to high speed and flipping. Again, just doesn’t seem like it acted logically?

 

[slup]

Thanks, But surely this wouldn’t cause the drone to take off and flip over. If the software KNOWS the angle is too high, surely it should not try to take off. Simple enough programming to check that.

It’s also weird that the drone props were spinning slowly before going to high speed and flipping. Again, just doesn’t seem like it acted logically?

The slow turning props could have been a sign that the drone were at the margin tilt wise for a take off... with the vibrations it just went over the right side of the threshold & tried to take of. Depending on the surface, the drone may have catched slightly with the front legs making it tilt forward further & the flight controller tried to correct it with commanding the motors more... & it ended with a nose dive in the snow.

 

[Etbrown4]

My experience with the dji warranty was different. I had a gps mismatch error upon takeoff, and resulting minor crash, not due to any nearby metal present. Sent it back to dji, they said the imu was bad and they sent me a spanking new drone. I did end up communicating with supervisors several layers up, but it was all collegial.

 

[Meta4]

My experience with the dji warranty was different. I had a gps mismatch error upon takeoff, and resulting minor crash, not due to any nearby metal present. Sent it back to dji, they said the imu was bad and they sent me a spanking new drone. I did end up communicating with supervisors several layers up, but it was all collegial.

GPS mismatch messages are almost always a meaningless false alarm.
They did not cause your crash.
You never posted your flight data to find out the actual cause of the crash, but it would have shown something quite different.

 

[Etbrown4]

The idea behind my post was not to discuss technology or causes, but to give some encouragement to other owners still in warranty with a bona fide failure, that if they persist, they can likely request and receive a free replacement.

 

[BrAinZ]

The log on Phantomhelp = 3ADQAHV280G787FTUW4L

It might be that the DAT is readable, do you have it? If so, can you post it?
The time stamp should be similar to the time stamp of the .txt log but 'predate' the .txt by a few seconds and if probably contains FLY036.

I backed up all the internal storage and memory card before sending the drone off to DJI, but there' doesn't appear to be a DAT file for this date, only the day after FLY037 on-

These were the two data graphs that DJI sent back. I still don't think that this a good enough reason to say the drone is not at fault though. Not even reason for it to flip itself and crash. there was definitely nothing in it's way or holding it back so the fact it crashed was entirely due to it's own software/hardware.

 

[Yorkshire_Pud]

That was rather strange, using a manual start, CSC. and launch I was able to start the motor as some quite large tilts, as I am hand launching I tend to roll the drone. BUT when I used the automated procedure it required the tilt to be quite small otherwise it refused to start the motors, issuing a 'tilt to large' warning.
I am going to do a pitch version now.

 

[Yorkshire_Pud]

Plots for nose down Mavic Mini.

 

[BrAinZ]

That was rather strange, using a manual start, CSC. and launch I was able to start the motor as some quite large tilts, as I am hand launching I tend to roll the drone. BUT when I used the automated procedure it required the tilt to be quite small otherwise it refused to start the motors, issuing a 'tilt to large' warning.
I am going to do a pitch version now.

I guess that’s my point to DJI. If the drone knows the tilt is too large, then it shouldn’t have taken off. That should be simple enough to set in their software. In my case, it took off and immediately went forward, flipping and down. That’s just illogical, and no reason for it to do that.