Crash in the mountains

[GrafMalt]

Hi, a while ago my dji mini crashed while traveling through Tasmania. Since them I am slightly reluctant to use it (I had 2 times that it crashed afterwards). Maybe someone can help me what the root-cause was?

The mobile phone txt log is here. Let me know if the DAT file is required and I will follow the instructions to upload it here.

Mobile Phone TXT Log: DJI Flight Log Viewer - PhantomHelp.com

Thanks in Advance

Malt

 

[Meta4]

Hi, a while ago my dji mini crashed while traveling through Tasmania. Since them I am slightly reluctant to use it (I had 2 times that it crashed afterwards). Maybe someone can help me what the root-cause was?

The flight data is a little complicated.
It would help if you could give some description of what happened beyond just saying that it crashed.

 

[iamcostas]

Hi.
First of all , reading the flight log gave me the opportunity to study and learn more about the drones.
I noticed something through the reading. How many times through this flight you pressed the return home button?
Also I notice a few up side down drops at the battery levels which I have no idea if it is normal or not. How many spared batteries do you have?
I hope I gave some tiny help with my observations.

 

[boblui]

Judging from the reluctance of the drone to tilt forward when forwad control was applied and when countering the wind, the crash was likely due to the widely reported uncommanded descent problem of the Mini and the root cause is deformed props, very often those on the rear motors. It seems that your Mini was having the same problem.

Suggest to update the firmware and FLY APP to the latest version and make a test flight. The motors with deformed props will beep and twitch after landing. Replacing those props will usually solve the problem.

 

[slup]

Yeah ... a plain & simple "Uncommanded descent" case unfortunately. Those have been quite rare lately as from DJI introduced the warning system regarding over revving motors in the firmware.

Looking into your flight a bit & depicting the crucial flight telemetry in graph form below shows ...

...that it in total was 3 major occasions where the Mini descended rapidly without stick commands, have marked those 3 periods with a red ring on the time line.

The red graph shows the height relative the HP, read from the barometric sensor ... the blue graph is the vertical speed reading, positive values mean descent & negative ascents. Lastly the green one tells us how you applied the throttle to regulate the height (throttle = left stick forward or backwards ... forward stick for ascending means positive values).

It's easy to see that in those marked 3 periods you had full left stick forward for ascend command ... but the vertical speed anyway went to big positive values --> the Mini lost height rapidly without you commanding it.

(click on all charts below to make them larger)


So why this behavior ...

Representative for "Uncommanded descent" cases have always been motors that can't spin faster as they are already on max RPM. This usually results in a bad ability for the Mini to maintain a proper tilt angle ... which means a very bad heading speed, not at all on par with the specifications for the Mini.

If we check off how the Mini tilts, how the heading speed performance is & how it was commanded by the sticks.

Have here below chosen a period just before you commanded RTH (the light blue background field) ... there you had some full elevator stick commands (full forward on the right stick = blue graph) ... according to the specification for a Mini it should be able to maintain a max tilt angle of 20 degrees in P-mode. Have placed the markers in the chart in a typical place where the elevator command was full ... but the Mini could just produce in average 10 degrees tilt. This is also reflected in the green heading speed graph with a really low speed.



This behavior with over revving motors (typically in the rear) leading to low tilt angles & bad heading speeds sooner or later develops to a more critical state where the props no more can generate enough thrust to generate lift either ... which leads to these uncommanded descents.

Usually in events like this a lot of "Not Enough Force/ESC" errors is generated ... & this was the case in your flight also.

Here from the flight log event stream ... at 143,5sec into the flight the first error appeared.

Not Enough Force/ESC Error. Max. Stromleistung des Fluggeräts erreicht. Flughöhe verringern und vorsichtig fliegen. Sollte das Problem weiter bestehen, sofort landen (Code: 30168)


All the wind warnings was in reality not true ... those are a result of wrong calculations coming from the bad tilt together with max stick commands ... the wind was in reality manageable, only a couple occasions is noted with gusts up to approx 7m/s.



So ... as said earlier, the root cause to all this is flattened or warped props. So see to that you change your props & update the Mini's firmware to the latest together with the DJI Fly app, then you should get warnings if this with deformed props are about to force the motors to go for max RPM again.

If you attach the mobile device DAT log we can out from there pinpoint which prop that was under performing this time ... the correct DAT ends with FLY017.DAT.

 

[Karlewski]

Very impressive analysis!

 

[GrafMalt]

I have to admit I am amazed how fast and thorough the analysis went - thanks to all and especially thanks to slup for the very detailed analysis. I remember the first time it fast descent down and I kinda panicked pulling the controller up - it was saved then for a while but when it happens the second time it shortly stopped and then went down - thankfully not in the lake but beside of it and thankfully not too far away from the way.

I just exchanged propellers and will try it soon in a safe environment to see if that fixed the problem.

Out of curiosity: Can snow / mountains be a problem for the dji mini? (e.g. because of reflection of light or because when you fly in a valley below my own position?) Might be a stupid question but want to prevent (i) loosing my mini and (ii) littering the alps with my plastic toy

 

[slup]

... I just exchanged propellers and will try it soon in a safe environment to see if that fixed the problem.

Also be aware about how you're storing your Mini ... it's been a lot of speculations around why the props gets deformed in the first place, much points to the Fly More Case & how the props gets/needs to be folded (according to DJI instruction) to fit in there. Otherwise this will happen again ... I propose that you only use cases as transportation devices only, when home keep the Mini free standing in a shelf with spread out props away from all kind of heat sources, then you can be sure that nothing is pressing on & deforming them.

Can snow / mountains be a problem for the dji mini?

...because of reflection of light
...because when you fly in a valley below my own position?)...

Generally no ... but in very high mountains the max take off altitude needs to be taken into consideration. For the Mini it is 3000m ASL ... this as the air is thinner & the props no more can make enough thrust to fly stable ... (these altitudes can also generate uncommanded descent with max motor RPM's).

Flying down into valleys (below your HP) shouldn't mean any problems in general, just negative heights in your app ... but be aware of not loosing a free line of sight between the RC & Mini. Also a very narrow valley can shield off large portions of the sky ... meaning that you might loose GPS coverage. Then the wind conditions in a valley can be a lot different than up on height ... the wind speeds can increase in valleys as they can serve as funnels.

Snow & reflections ... well from the manual you can read out that the visual sensors on the Mini's belly can't lock on to homogeneous colored/patterns or reflective surfaces ... this means that the horizontal & vertical stability gets affected. According to the spec, below 10m from ground the VPS sensors come into play ... without a proper ground lock you must expect more drifts than usual when landing or hovering below 10m.

 

[Crouching Leopard]

Yeah ... a plain & simple "Uncommanded descent" case unfortunately. Those have been quite rare lately as from DJI introduced the warning system regarding over revving motors in the firmware.

Looking into your flight a bit & depicting the crucial flight telemetry in graph form below shows ...

...that it in total was 3 major occasions where the Mini descended rapidly without stick commands, have marked those 3 periods with a red ring on the time line.

The red graph shows the height relative the HP, read from the barometric sensor ... the blue graph is the vertical speed reading, positive values mean descent & negative ascents. Lastly the green one tells us how you applied the throttle to regulate the height (throttle = left stick forward or backwards ... forward stick for ascending means positive values).

It's easy to see that in those marked 3 periods you had full left stick forward for ascend command ... but the vertical speed anyway went to big positive values --> the Mini lost height rapidly without you commanding it.

(click on all charts below to make them larger)
View attachment 120794

So why this behavior ...

Representative for "Uncommanded descent" cases have always been motors that can't spin faster as they are already on max RPM. This usually results in a bad ability for the Mini to maintain a proper tilt angle ... which means a very bad heading speed, not at all on par with the specifications for the Mini.

If we check off how the Mini tilts, how the heading speed performance is & how it was commanded by the sticks.

Have here below chosen a period just before you commanded RTH (the light blue background field) ... there you had some full elevator stick commands (full forward on the right stick = blue graph) ... according to the specification for a Mini it should be able to maintain a max tilt angle of 20 degrees in P-mode. Have placed the markers in the chart in a typical place where the elevator command was full ... but the Mini could just produce in average 10 degrees tilt. This is also reflected in the green heading speed graph with a really low speed.

View attachment 120795

This behavior with over revving motors (typically in the rear) leading to low tilt angles & bad heading speeds sooner or later develops to a more critical state where the props no more can generate enough thrust to generate lift either ... which leads to these uncommanded descents.

Usually in events like this a lot of "Not Enough Force/ESC" errors is generated ... & this was the case in your flight also.

Here from the flight log event stream ... at 143,5sec into the flight the first error appeared.

Not Enough Force/ESC Error. Max. Stromleistung des Fluggeräts erreicht. Flughöhe verringern und vorsichtig fliegen. Sollte das Problem weiter bestehen, sofort landen (Code: 30168)

View attachment 120797
All the wind warnings was in reality not true ... those are a result of wrong calculations coming from the bad tilt together with max stick commands ... the wind was in reality manageable, only a couple occasions is noted with gusts up to approx 7m/s.

View attachment 120798

So ... as said earlier, the root cause to all this is flattened or warped props. So see to that you change your props & update the Mini's firmware to the latest together with the DJI Fly app, then you should get warnings if this with deformed props are about to force the motors to go for max RPM again.

If you attach the mobile device DAT log we can out from there pinpoint which prop that was under performing this time ... the correct DAT ends with FLY017.DAT.

Wow.

 

[LakeWorthB]

Yeah ... a plain & simple "Uncommanded descent" case unfortunately. Those have been quite rare lately as from DJI introduced the warning system regarding over revving motors in the firmware.

Looking into your flight a bit & depicting the crucial flight telemetry in graph form below shows ...

...that it in total was 3 major occasions where the Mini descended rapidly without stick commands, have marked those 3 periods with a red ring on the time line.

The red graph shows the height relative the HP, read from the barometric sensor ... the blue graph is the vertical speed reading, positive values mean descent & negative ascents. Lastly the green one tells us how you applied the throttle to regulate the height (throttle = left stick forward or backwards ... forward stick for ascending means positive values).

It's easy to see that in those marked 3 periods you had full left stick forward for ascend command ... but the vertical speed anyway went to big positive values --> the Mini lost height rapidly without you commanding it.

(click on all charts below to make them larger)
View attachment 120794

So why this behavior ...

Representative for "Uncommanded descent" cases have always been motors that can't spin faster as they are already on max RPM. This usually results in a bad ability for the Mini to maintain a proper tilt angle ... which means a very bad heading speed, not at all on par with the specifications for the Mini.

If we check off how the Mini tilts, how the heading speed performance is & how it was commanded by the sticks.

Have here below chosen a period just before you commanded RTH (the light blue background field) ... there you had some full elevator stick commands (full forward on the right stick = blue graph) ... according to the specification for a Mini it should be able to maintain a max tilt angle of 20 degrees in P-mode. Have placed the markers in the chart in a typical place where the elevator command was full ... but the Mini could just produce in average 10 degrees tilt. This is also reflected in the green heading speed graph with a really low speed.

View attachment 120795

This behavior with over revving motors (typically in the rear) leading to low tilt angles & bad heading speeds sooner or later develops to a more critical state where the props no more can generate enough thrust to generate lift either ... which leads to these uncommanded descents.

Usually in events like this a lot of "Not Enough Force/ESC" errors is generated ... & this was the case in your flight also.

Here from the flight log event stream ... at 143,5sec into the flight the first error appeared.

Not Enough Force/ESC Error. Max. Stromleistung des Fluggeräts erreicht. Flughöhe verringern und vorsichtig fliegen. Sollte das Problem weiter bestehen, sofort landen (Code: 30168)

View attachment 120797
All the wind warnings was in reality not true ... those are a result of wrong calculations coming from the bad tilt together with max stick commands ... the wind was in reality manageable, only a couple occasions is noted with gusts up to approx 7m/s.

View attachment 120798

So ... as said earlier, the root cause to all this is flattened or warped props. So see to that you change your props & update the Mini's firmware to the latest together with the DJI Fly app, then you should get warnings if this with deformed props are about to force the motors to go for max RPM again.

If you attach the mobile device DAT log we can out from there pinpoint which prop that was under performing this time ... the correct DAT ends with FLY017.DAT.

Quick question, these plots were made with CsvView? Wind one also?

 

[slup]

Quick question, these plots were made with CsvView? Wind one also?

All from CsvView besides the wind bar chart which is averages from a payed Airdata subscription. The mobile device TXT log doesn't have any wind calculations in CsvView ... the DAT log have though.