uncontrolled vertikal ascent

[AliAli]

Hey everyone. I am new to drones and just got the mavic mini, my first post here.

I can't explain the drone's behavior on this flight - it stopped responding to the controls on the ascent and ascended so high until it lost signal. He also exceeded the set maximum height. Then, fortunately, the automatics worked and he came back. Proper stress dose… Altitude above ground from video 430 m :-(

Thank you in advance for every comment.

 

[Meta4]

Hey everyone. I am new to drones and just got the mavic mini, my first post here.

I can't explain the drone's behavior on this flight - it stopped responding to the controls on the ascent and ascended so high until it lost signal. He also exceeded the set maximum height. Then, fortunately, the automatics worked and he came back. Proper stress dose… Altitude above ground from video 430 m :-(

You were lucky not to lose the drone.
The drone showed problems holding altitude and the flight data shows uncommanded ascent and descent.
There are many Not Enough Force/ESC Error messages, which indicates that at full power, the propellers are not giving full thrust.
They have flattened and the change in shape is causing the issue.
Change to new propellers and it should fix the problem.

 

[Gagey52]

Welcome to the forum from the land of Oz. Not a great way to start with almost loosing your drone, anyway, look forward to your input to the forum.
Regards

 

[Deleted member 103366]

This is your flight log on Phantom help log viewer

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

and the CsvVview of motor speed, height and in the shaded areas "not enough force"

 

[slup]

...the propellers are not giving full thrust.
They have flattened and the change in shape is causing the issue.
Change to new propellers and it should fix the problem.

That's not correct... this have nothing to do with props at all.

This is a flight controller that gets a major brain fart for some unknown reason ... if this craft still is within warranty I would contact DJI about this.

Below is from the second of in total 2 major FC errors making the craft ascend uncommanded ...

The red graph is the height relative the HP
The green graph is the vertical velocity ... negative value is ascend & positive descend
The blue graph is a calculated velocity difference between the IMU & GPS ... normally this is very close to zero
The black graph is the throttle input ... negative value for descend & positive for ascend
Pink background color is tripod mode
Green background color is auto landing ... this time by means of an app command done by the pilot



Note how well everything lines up to the very left in the chart, a negative throttle input (black) means a increased positive vertical velocity (green) ... & the height (red) decreases. All good so far.

At 555,9sec I've placed a chart marker ... it's there the FC get it's second major brain fart during this flight. The red increases, the black throttle is neutral & the green vertical velocity starts to go more & more positive step wise which should mean a descent but here the barometric sensor instead reads a ascent (red).

All goes haywire at about 590sec ... the craft ascends up rapidly even though the craft is in auto landing mode with a full negative throttle from the pilot ... all this goes on until about 625sec with the craft at approx 177m height ... and all goes back to normal.

Looking at how the FC commands the motors during this period ... the command percentage goes erratic up near 100%... it's clear that the flight controller want's the motors to spin faster... and all back to normal from 625sec.



... As the FC commands nearly 100% to the motors ... it also gets nearly full RPM's seen below



If we then check the DAT log event stream we see this when it all starts ...



And when all goes back to normal ... "fault off" as nothing have happened!

 

[Gagey52]

That's not correct... this have nothing to do with props at all.

This is a flight controller that gets a major brain fart for some unknown reason ... if this craft still is within warranty I would contact DJI about this.

Below is from the second of in total 2 major FC errors making the craft ascend uncommanded ...

The red graph is the height relative the HP
The green graph is the vertical velocity ... negative value is ascend & positive descend
The blue graph is a calculated velocity difference between the IMU & GPS ... normally this is very close to zero
The black graph is the throttle input ... negative value for descend & positive for ascend
Pink background color is tripod mode
Green background color is auto landing ... this time by means of an app command done by the pilot

View attachment 141966

Note how well everything lines up to the very left in the chart, a negative throttle input (black) means a increased positive vertical velocity (green) ... & the height (red) decreases. All good so far.

At 555,9sec I've placed a chart marker ... it's there the FC get it's second major brain fart during this flight. The red increases, the black throttle is neutral & the green vertical velocity starts to go more & more positive step wise which should mean a descent but here the barometric sensor instead reads a ascent (red).

All goes haywire at about 590sec ... the craft ascends up rapidly even though the craft is in auto landing mode with a full negative throttle from the pilot ... all this goes on until about 625sec with the craft at approx 177m height ... and all goes back to normal.

Looking at how the FC commands the motors during this period ... the command percentage goes erratic up near 100%... it's clear that the flight controller want's the motors to spin faster... and all back to normal from 625sec.

View attachment 141967

... As the FC commands nearly 100% to the motors ... it also gets nearly full RPM's seen below

View attachment 141968

If we then check the DAT log event stream we see this when it all starts ...

View attachment 141969

And when all goes back to normal ... "fault off" as nothing have happened!

View attachment 141970

That’s amazing stuff there slup, when I get more time going to have to learn how to do that. Well done.
Regards

 

[AliAli]

Slup: Exact description, thank you. I still have a lot to learn about reading flight protocols.
Wasn't there a problem flying in the twilight after sunset? I'm afraid to try again.

The magic of the unwanted - Brno from height 356 m :-(

 

[slup]

...Wasn't there a problem flying in the twilight after sunset?

In general, flying in darker conditions, will only disable possible obstacle avoidance & visual positioning (downward) sensors.

In your case with a Mini 1 you only have VPS sensors ... flying in the dark will at most make them useless, meaning that height measurements above ground directly under the Mini will be deactivated, that the landing protection will not be working (you need to be careful when landing to avoid hard landings) ... & that hover at lower heights can be somewhat drifty compared to when the VPS work & can lock on the ground pattern.

That it was dark had nothing to do with this incident ... contact DJI for this, you never know when this can happen again.

 

[BOSSMAN1]

I've just lost my drone. All normal until about 8 minutes in to the flight the **** thing just kept going up.. and up.... and up. I lost contact with it going past 990m height at 2mps vertical speed, with the battery down to less than 10% by then and everything bleeping and flashing. No direction control possible, no vertical control after those first few minutes. "Find my drone" shows me where it's supposed to be but four of us scouring the area can't find it. I suspect that was the last position the phone got signal and it could have travelled some distance in the wind at that height, even though on the ground there's hardly a breath. BUMMER! BTW the "cannot return automatically" message did come up.

 

[Meta4]

Find my drone" shows me where it's supposed to be but four of us scouring the area can't find it. I suspect that was the last position the phone got signal

Find my Drone isn't magic.
It can't ever do anything more than show the last location data that the app received.
If you post your flight data, it might give some more clues to explain what happened and what direction to search.

Go to DJI Flight Log Viewer | Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
That will give you a detailed report of the flight.
Come back and post a link to the report it gives you.
Or .. just post the txt file here.

 

[Andris]

@slup Where can I view the dat log event list in a log file? (I know csvview, datcon, extractdji, etc tools but I think I am missing a tool )

 

[slup]

...Where can I view the dat log event list in a log file?

By pushing the marked button in CsvView when you have loaded a DAT log...

 

[Andris]

Thank you @slup!

I found some interesting part in the graphs also: I can see neutral position of the throttle stick but relative height goes up and down. Also the height_ctrl error shows up in the log too.

What was your conclusion about this issue? Processor on main board or IMU module? (I already swapped the GPS/barometer module to rule out a barometer issue.)

Thanks!

 

[slup]

...What was your conclusion about this issue?

You have my analyze of this threads incident in --> Post #5

 

[Andris]

I further analyzed the issue. I attach a new screenshot.

In the picture you can see that the baro and the gps height readings are going together. So if I am correct the problem is not with the GPS or the baro. However I see that the acceleration on the Z axis is going up and down rapidly without touching the throttle stick. So I assume that there is something wrong with the IMU? Maybe some inside dampening is defective? What do you think?