Mavic lost in sea

[demme]

The compass/yaw errors are equipment problems - nothing that you did. The stick inputs are real and the aircraft responded appropriately during both segments of the flight. It appears that it went into autoland after losing uplink in ATTI mode, which was why it descended to the water, but then uplink was restored at 903 s and a few seconds later at 912 s you gave it throttle input that caused it to ascend again. It also dropped back into ATTI mode and started drifting at that point. It's not clear if it ever touched down, but it was close.

Yeah the uplink was restored at 903 s because I was running to the direction where I saw it went down, I then gave throttle input to ascend...after that it basicly just flew away and I could not stop it.

 

[sar104]

Yeah the uplink was restored at 903 s because I was running to the direction where I saw it went down, I then gave throttle input to ascend...after that it basicly just flew away and I could not stop it.

You still had control and could have flown it back, but it was in ATTI mode and I'm guessing that you are not familiar with flying in that mode. It's certainly tricky, especially if you don't have a clear view of the aircraft and it's orientation.

 

[demme]

You still had control and could have flown it back, but it was in ATTI mode and I'm guessing that you are not familiar with flying in that mode. It's certainly tricky, especially if you don't have a clear view of the aircraft and it's orientation.

Well, I've flown in ATTI mode indoors, but as you say it's a bit tricky, especially when the aircraft is a far away.

So I guess DJI will find that it was pilot error rather than aircraft...

 

[sar104]

Well, I've flown in ATTI mode indoors, but as you say it's a bit tricky, especially when the aircraft is a far away.

So I guess DJI will find that it was pilot error rather than aircraft...

Except that all the magnetic errors led to the ATTI mode in the first place, so they may be willing to accept some of the fault because of that. Certainly worth asking.

 

[Montypaks]

What do you think caused those errors?

 

[demme]

Except that all the magnetic errors led to the ATTI mode in the first place, so they may be willing to accept some of the fault because of that. Certainly worth asking.

Yes I will. I've got a case number with them know. I've just synced all flight records to the DJI server, extracted the RC logs and all the stuff they wanted. Just sent it all to them. Pretty sure they are going to find a reason/fault outside warranty which makes the case invalid.

One other thing, regarding all your help, much obliged btw, so very, very kind of you to help out like you did.

Still many thoughts in my head and many hours in the forest searching for it.

It makes me wonder, u wrote:

"it was last recorded heading NE at over 50 km/hr, at 59.74614994° N, 14.52274262° E, at an altitude of 115 m with 28% battery remaining when downlink was lost permanently."

And then the wind blowing East making it drift like u showed on the map.

How ever, how can you know that it did not land or crashed down right away above mentioned position? Or that it came so fast that the breaking took many meters b4 it stopped?

Also, yeah, I saw wind blowing East on airdata.com, but I'm thinking, would not a few m/s blowing from another direction put it drifting on a total different path than the one you painted?

 

[demme]

What do you think caused those errors?

well I have no idea, but out of 3200 error codes there are over 1000 saying "compass error", also the log is clearly showing that I'm pressing RTH without any response from the mavic

 

[sar104]

Yes I will. I've got a case number with them know. I've just synced all flight records to the DJI server, extracted the RC logs and all the stuff they wanted. Just sent it all to them. Pretty sure they are going to find a reason/fault outside warranty which makes the case invalid.

One other thing, regarding all your help, much obliged btw, so very, very kind of you to help out like you did.

Still many thoughts in my head and many hours in the forest searching for it.

It makes me wonder, u wrote:

"it was last recorded heading NE at over 50 km/hr, at 59.74614994° N, 14.52274262° E, at an altitude of 115 m with 28% battery remaining when downlink was lost permanently."

And then the wind blowing East making it drift like u showed on the map.

How ever, how can you know that it did not land or crashed down right away above mentioned position? Or that it came so fast that the breaking took many meters b4 it stopped?

Also, yeah, I saw wind blowing East on airdata.com, but I'm thinking, would not a few m/s blowing from another direction put it drifting on a total different path than the one you painted?

You cannot know that it did not land or crash at that point, but there is no reason to assume that it did since it was not at critical battery level. As for the wind direction, the aircraft was in ATTI mode and so AirData is not giving much useful data. Far more useful are the periods of no stick input that show a very consistent speed and direction of drift, and that's what I used.

But you are correct - if that were to have varied, or if the aircraft were to have continued to receive any stick inputs after the downlink disconnected, then that would change the picture. And even if that were not the case there is uncertainty in the battery depletion rate on drifting, the battery critical level for landing, and the amount of drift during descent. I'd put at least a kilometer of uncertainty on the guess.

 

[sar104]

well I have no idea, but out of 3200 error codes there are over 1000 saying "compass error", also the log is clearly showing that I'm pressing RTH without any response from the mavic

It could not RTH because it was in ATTI mode.

 

[demme]

It could not RTH because it was in ATTI mode.

Yeah.. But what really caused it to go in ATTI mode, it was in open spaces all the time and it had strong GPS signal....hmm

 

[sar104]

I explained that previously - it has nothing to do with the quality of the GPS reception, serious compass errors always result in the FC switching to ATTI.

 

[demme]

I explained that previously - it has nothing to do with the quality of the GPS reception, serious compass errors always result in the FC switching to ATTI.

Oh okay sorry. Are we talking about the compass in the drone or my mobile device? Either way all check points were OK before take off. I even used a landing/takeoff pad.

I've never had problems with the compass before, this said, I recently purchased a crystalsky which I noticed had a much more sensitive compass than a mobile phone.

 

[Montypaks]

He's referring to compass on the quad.

So why is the compass on the Mavic so **** sensitive. It's like both compasses can error out and the AC goes into some suicidal state.

Is this a product design issue? Why is such a fragile part of the AC so critical?

 

[sar104]

Oh okay sorry. Are we talking about the compass in the drone or my mobile device? Either way all check points were OK before take off. I even used a landing/takeoff pad.

I've never had problems with the compass before, this said, I recently purchased a crystalsky which I noticed had a much more sensitive compass than a mobile phone.

The compasses on the aircraft. There have been several reports of this recently and it is not clear whether there is a hardware problem, a firmware problem, or if it is just coincidence.

 

[sar104]

He's referring to compass on the quad.

So why is the compass on the Mavic so **** sensitive. It's like both compasses can error out and the AC goes into some suicidal state.

Is this a product design issue? Why is such a fragile part of the AC so critical?

There have been a few cases of that with no clear cause. It doesn't appear to be a widespread problem though. And just to be clear, the aircraft goes into ATTI mode which is perfectly flyable. A problem for Mavic pilots, however, is that since the aircraft has no selectable ATTI mode (unlike the Phantoms), many have no experience flying in ATTI.

 

[demme]

There have been a few cases of that with no clear cause. It doesn't appear to be a widespread problem though. And just to be clear, the aircraft goes into ATTI mode which is perfectly flyable. A problem for Mavic pilots, however, is that since the aircraft has no selectable ATTI mode (unlike the Phantoms), many have no experience flying in ATTI.

And since they brag about their 7km range, if it goes into atti mode 4km away and the compass is loco... Kinda makes it hard to fly it home in atti mode then.

Yes yes, someone will say, always keep a line of sight... Hey let's come up with this new technology called occusync with 7km range and expect all pilots to keep a LOS, because the compass might **** you up at a random point.

 

[rfc]

I did not know that there's no switch selectable way to put the Mavic into ATTI. Is there a way to do it from DJIGO4?
If not, is there no way to practice flying in ATTI mode?
And finally (forgive all the questions, but my only experience is in a P3), If you have video from the screen, why is ATTI so difficult to fly (I mean that if you roughly know where the bird is, and you know where you are...what's the problem?)

 

[sar104]

And since they brag about their 7km range, if it goes into atti mode 4km away and the compass is loco... Kinda makes it hard to fly it home in atti mode then.

Yes yes, someone will say, always keep a line of sight... Hey let's come up with this new technology called occusync with 7km range and expect all pilots to keep a LOS, because the compass might **** you up at a random point.

Or to paraphrase more accurately, if your aircraft suffers a relatively rare compass issue, and you are outside LOS, and you don't know how to read position and orientation information from your screen (which is still there in ATTI mode), and you have no experience flying in ATTI mode, then you may lose the aircraft. Correct.

 

[sar104]

I did not know that there's no switch selectable way to put the Mavic into ATTI. Is there a way to do it from DJIGO4?
If not, is there no way to practice flying in ATTI mode?
And finally (forgive all the questions, but my only experience is in a P3), If you have video from the screen, why is ATTI so difficult to fly (I mean that if you roughly know where the bird is, and you know where you are...what's the problem?)

No - not on a Mavic. And agreed - ATTI is trickier than P-GPS but not impossible. You can still use telemetry and still have video as you pointed out. But if you panic and just start randomly applying stick inputs, then it's unlikely to find its way back.

 

[demme]

I did not know that there's no switch selectable way to put the Mavic into ATTI. Is there a way to do it from DJIGO4?
If not, is there no way to practice flying in ATTI mode?
And finally (forgive all the questions, but my only experience is in a P3), If you have video from the screen, why is ATTI so difficult to fly (I mean that if you roughly know where the bird is, and you know where you are...what's the problem?)

Hi, There's no way to put it in Atti mode manual, it's in goes into atti mode if it doesn't have enough GPS satellites. So mine went into atti mode at the same time the compass stopped working. If I went north(home at that point), it went maybe East or something... And so on.