Critical low battery, and unable to maneuver!

[scubaddictions]

While being stupid today, I was flying the M2P around draining a battery in preparation for deep cycling, so deliberately ignoring the low battery alerts but keeping an eye on voltage/percentage. Around 10% I was on my way home and only a few hundred feet away, maybe 100 feet up when it tried to force auto-land. This is where things get weird: it would NOT let me control the drone in any lateral direction. I was locked in a descent, though with left stick up I could arrest the descent and force it to hover but it would not allow me to fly forward in the direction of home. As far as I can see in the manual and various videos online this should not be possible, everyone states that you can still maneuver the drone to avoid obstacles during the auto-descent. In fact, on my first drone (Spark) that definitely saved my bacon one day, out over the ocean at zero percent (!) but it still let me proceed forward to a safe landing.

During today's flight the drone was in P mode, so my first thought was to over-ride with Sport mode, no change. Things were getting dicey. Then I remembered that I had hacked the drone to replace Tripod mode with Atti mode, which worked like magic. Once in Atti mode I immediately regained forward control and flew home to a safe landing.

Later in the evening I was running down another battery (while being much smarter, low altitude, hovering over the home point) and I could not duplicate the lateral lock-out condition. Even at critically low, fighting the forced descent, I could still maneuver the drone in any direction needed.

Any thoughts on why it locked me out on the first flight? All I can think of is that I was beyond some threshold, either too high or too far from the home point so I wouldn't allow me to keep fighting my way to get home. Doesn't make sense to lock out completely though, auto-landing or not the pilot should still be able to control the drone to avoid water or obstacles.

 

[Meta4]

Any thoughts on why it locked me out on the first flight? All I can think of is that I was beyond some threshold, either too high or too far from the home point so I wouldn't allow me to keep fighting my way to get home. Doesn't make sense to lock out completely though, auto-landing or not the pilot should still be able to control the drone to avoid water or obstacles.

It's not possible to tell anything without seeing data.
But the idea of running your batteries down to somehow manage your batteries better is a myth.
Don't do it, don't put drone in this situation because there's nothing to gain from it.

 

[new2mavic]

why it locked me out on the first flight? All I can think of is that I was beyond some threshold, either too high or too far from the home point so I wouldn't allow me to keep fighting my way to get home.

This is a mystery but what you have mentioned could not be the reasons. In the following flight of my M2P, autolanding was triggered at 11% battery. The craft was much further away and higher than your craft was when it happened but I had no problem in controlling the craft :

 

[Phantomrain.org]

While being stupid today, I was flying the M2P around draining a battery in preparation for deep cycling, so deliberately ignoring the low battery alerts but keeping an eye on voltage/percentage. Around 10% I was on my way home and only a few hundred feet away, maybe 100 feet up when it tried to force auto-land. This is where things get weird: it would NOT let me control the drone in any lateral direction. I was locked in a descent, though with left stick up I could arrest the descent and force it to hover but it would not allow me to fly forward in the direction of home. As far as I can see in the manual and various videos online this should not be possible, everyone states that you can still maneuver the drone to avoid obstacles during the auto-descent. In fact, on my first drone (Spark) that definitely saved my bacon one day, out over the ocean at zero percent (!) but it still let me proceed forward to a safe landing.

During today's flight the drone was in P mode, so my first thought was to over-ride with Sport mode, no change. Things were getting dicey. Then I remembered that I had hacked the drone to replace Tripod mode with Atti mode, which worked like magic. Once in Atti mode I immediately regained forward control and flew home to a safe landing.

Later in the evening I was running down another battery (while being much smarter, low altitude, hovering over the home point) and I could not duplicate the lateral lock-out condition. Even at critically low, fighting the forced descent, I could still maneuver the drone in any direction needed.

Any thoughts on why it locked me out on the first flight? All I can think of is that I was beyond some threshold, either too high or too far from the home point so I wouldn't allow me to keep fighting my way to get home. Doesn't make sense to lock out completely though, auto-landing or not the pilot should still be able to control the drone to avoid water or obstacles.

It is it possible when changing the paramaters of the assistant 2 that you also changed how the forced landing works by enabling the Atti mode.

I assume you changed to 3 correct > So you would need to change it back and test it out.

Phantomrain.org
Gear to fly your Mavic 2 in the Rain.

 

[slup]

Let's first take a look at the flight log before starting up any try & error exercise ... indications of what happened is most probably in there.

Start up by sharing the .TXT log from this flight which is stored in the mobile device you flew with, read up here on how to retrieve --> Mavic Flight Log Retrieval and Analysis Guide (read under section 3.). Once retrieved attach it here in a new post.

 

[scubaddictions]

Flight record attached.

 

[scubaddictions]

It is it possible when changing the paramaters of the assistant 2 that you also changed how the forced landing works by enabling the Atti mode.

Possible, but on the follow-up flight I was able to freely maneuver at critically low battery without changing parameters.

 

[scubaddictions]

But the idea of running your batteries down to somehow manage your batteries better is a myth.

The recommendation is still included in the manual for the Intelligent Flight Battery:

 

[slup]

Flight record attached.

Well, unfortunately the mobile device .TXT log doesn't give away any reason for your described behavior ... it just confirms it.

The chart below covers the yellow marked flight path in the sat picture ... the incident happens where I've drawn a red circle.



The chart below deserves some explanation ...

The white background color is P-mode ... from where the colored backgrounds starts the low battery autolanding have been initiated. There the light blue field is P-mode, the green is Sport mode & the darker blue is where you initiated the ATTI-mode.

The black graph shows the heading speed in m/s
The red the height in meters
The blue the elevator (your right stick, positive = forward, negative = backwards)
& the purple graph the throttle (your left stick, positive = ascending, negative = descending)



It's clear that something is blocking your elevator command ... you try & try but nothing happens with the heading speed. At the same time we see that the throttle stick have effect & ascends the AC step wise as the autolanding will make it decend every time you release it.

As soon as you switch over to ATTI-mode the AC starts to listen to the elevator stick & the heading speed picks up.

We need to dig deeper ... try to attach the mobile device .DAT log ending with FLY061.DAT ... perhaps that can give us more to go on.

 

[scubaddictions]

We need to dig deeper ... try to attach the mobile device .DAT log ending with FLY061.DAT ... perhaps that can give us more to go on.

Can you confirm the file number please? I have similar numbers, but a search of the entire DJI folder doesn't yield a FLY061.

In fact I only have one .DAT file in the MCDatFlightRecords folder dated from yesterday, ends in FLY120.DAT. I'm curious, I did two flights yesterday with the same mobile device and drone, plus a third flight with the same mobile but different drone. I didn't exit out of DJI Go 4 in between flights, I'm wondering if it created one monolithic DAT file for the whole day?

Thanks!

 

[slup]

Can you confirm the file number please? I have similar numbers, but a search of the entire DJI folder doesn't yield a FLY061.

In fact I only have one .DAT file in the MCDatFlightRecords folder dated from yesterday, ends in FLY120.DAT. I'm curious, I did two flights yesterday with the same mobile device and drone, plus a third flight with the same mobile but different drone. I didn't exit out of DJI Go 4 in between flights, I'm wondering if it created one monolithic DAT file for the whole day?

Thanks!

That is the DAT index specified in the .TXT log ...



Covering this flight ...



The mobile device .DAT log runs from AC power on to off ... have nothing to do with closing DJI GO4 app in between flights.

 

[scubaddictions]

That is the DAT index specified in the .TXT log ...

View attachment 115590

Hmmmm. That said, is there a good reason why the specified DAT file wouldn't exist on my mobile?

 

[scubaddictions]

Hmmmm. That said, is there a good reason why the specified DAT file wouldn't exist on my mobile?

Additionally, any idea why the mobile only created a single DAT on a day that included three different flights?

 

[DanMan32]

We have found that the Fly app seems to delete logs, particularly DAT logs when synced with DJI. Perhaps Go is doing the same thing.

 

[scubaddictions]

We have found that the Fly app seems to delete logs, particularly DAT logs when synced with DJI. Perhaps Go is doing the same thing.

Once deleted, any way to force retrieveal or re-create it?

 

[slup]

Hmmmm. That said, is there a good reason why the specified DAT file wouldn't exist on my mobile?

Well ... it's not an unknown problem, deleted or not created DAT logs have been happening from time to time. No root cause have been firmly established, it's so far been combinations of phone OS, different versions of app & which app. It's also known that versions of remotes for M2's can totally skip writing DAT logs.

This copied from the link I provided earlier in the thread ...

Once deleted, any way to force retrieveal or re-create it?

No unfortunately ... further more the DAT log in the aircraft itself isn't readable for anyone outside DJI either.

 

[scubaddictions]

Well ... it's not an unknown problem, deleted or not created DAT logs have been happening from time to time. No root cause have been firmly established, it's so far been combinations of phone OS, different versions of app & which app. It's also known that versions of remotes for M2's can totally skip writing DAT logs.

This copied from the link I provided earlier in the thread ...

View attachment 115592

No unfortunately ... further more the DAT log in the aircraft itself isn't readable for anyone outside DJI either.

So, end of the road? Do you think it's worth looking at the single DAT created yesterday? **EDIT** Just crunched it through Airdata, looks like the single mobile DAT from yesterday was from my first flight of the day on a P4P. So, I don't have a DAT from either flight made yesterday with the M2P. Bummer.

 

[slup]

So, end of the road? Do you think it's worth looking at the single DAT created yesterday?

Put it up ... nothing to lose, will look it over sometime during tomorrow if somebody else haven't tried already, well over bed time in Sweden now ?

 

[scubaddictions]

So this is a mystery. I wanted to see how Airdata would interpret the available DAT files on my mobile, so for the single recorded DAT file saved yesterday I did a manual upload. It processed, and it was for my first flight on a P4P. Rather than uploading the rest of the DAT files manually/individually I downloaded the Airdata UAV app and turned on auto-sync. After processing, Airdata is showing data for all flights logged on my mobile, (including the glitched flight yesterday) not just those with DAT files.

I'm guessing that Airdata is more than happy to use the TXT flight logs when DAT files aren't available.

In the end though it appears DJI Go 4 on this device is only saving DAT files for about half my flights, that is, I have a bit more than twice as may TXT files as I have DAT files. Those numbers appear accurately reflected within Airdata. I can't see a pattern either, it saves a DAT file at random for both of my primary drones but randomly skips flights, even skips whole days of flights. That's very poor reliability! DJI Go 4 is running a fairly recent (about two weeks old) clean install on this mobile.

Just verified that the missing DAT files weren't synced to the cloud and deleted locally. Synced a different device and verified that my missing flights aren't in them.

 

[Meta4]

Any thoughts on why it locked me out on the first flight? All I can think of is that I was beyond some threshold, either too high or too far from the home point so I wouldn't allow me to keep fighting my way to get home.
Doesn't make sense to lock out completely though, auto-landing or not the pilot should still be able to control the drone to avoid water or obstacles.

I can't find a reason for being locked out, but you were only locked out from 22:06.9 - 22:52.9.
After that you had full directional control.
The data shows you changing the Flight Mode switch to Atti Mode at 22:52.9, which unlocked whatever glitch was preventing full control.