Lost my Phantom 4 Advanced +

To add to my previous post - the appropriate corrective action would have been to switch immediately to ATTI, which would have stopped the incorrect attempts by the FC to correct for drift, and then fly it back. This was almost certainly due to severe magnetic interference at the launch site.
So that would have worked? ATTI needs nothing from the compass, I guess that does make sense, I have never had any remote aircraft before I got my first Phantom.

Rod
 
I say non-responsive, because they were basically at the "Zero" input point. If the op were applying both throttle and rudder, which were responding, I can't fathom to see how it is not possible for the OP, under duress at some point apply Aileron and Elevator. That just does not make sense to me. Would you agree with that statement?

I cannot know for certain whether the OP touched the right stick or not - all I can say is that the log says that he did not. And that appears to me to be a more likely scenario than that inputs were applied, but ignored by the aircraft and not recorded, especially since nothing else about this record suggests any elevator or aileron inputs - it appears to me that the aircraft behavior is fully explainable by the large compass error. Are you seeing anything in the record that does not seem consistent with that?
 
So that would have worked? ATTI needs nothing from the compass, I guess that does make sense, I have never had any remote aircraft before I got my first Phantom.

Rod

Yes - the flight behavior appears to be entirely due to the FC attempting to correct for some initial drift to the NE after takeoff but, since it was facing the opposite direction than the yaw indicated, actually driving it hard in that direction. A positive, rather than negative feedback loop. Switching to ATTI would have stopped that, and it would have simply held attitude and altitude. It would have continued to drift NE due to the wind, but that could have been corrected by appropriate stick inputs, in contrast to the complete lack of control that results from a large compass error and an FC still trying to control the aircraft's position.
 
  • Like
Reactions: RodPad
Sorry Rod, I may have misunderstood your post, ASAP in the UK is a common acronym for "as soon as possible", I am not sure if it is universal. I may have not understood your comment completely, so apologies.
My statement was in the OP's deference, once the errors started, he couldn't or (as I'm understanding a little more) didn't know how to bring it down. He didn't have a ASAP, after the first error he was done, all the errors that you commented on, he was not in control. ;)

Rod
 
  • Like
Reactions: Neon Euc
I cannot know for certain whether the OP touched the right stick or not - all I can say is that the log says that he did not
I was asking for your opinion on that, because no I did not see anything inconsistent with the large compass error, however, as I said I cannot see how there could be no inputs from the right stick. Which would lead me to believe that there are two scenarios involved. One the major compass errors and primariy from all indications the actual cause of the incident Two, would be possibly a controller problem with the right stick. That was really where I was going with that.
 
Yes - the flight behavior appears to be entirely due to the FC attempting to correct for some initial drift to the NE after takeoff but, since it was facing the opposite direction than the yaw indicated, actually driving it hard in that direction. A positive, rather than negative feedback loop. Switching to ATTI would have stopped that, and it would have simply held attitude and altitude. It would have continued to drift NE due to the wind, but that could have been corrected by appropriate stick inputs, in contrast to the complete lack of control that results from a large compass error and an FC still trying to control the aircraft's position.
OK, Thanks, that statement will help many and we can urge others on learning ATTI.

ATTI is your personal Fail Safe. Learn ATTI and practice if you can see it, you can save it... (Well something like that)

Rod
 
I was asking for your opinion on that, because no I did not see anything inconsistent with the large compass error, however, as I said I cannot see how there could be no inputs from the right stick. Which would lead me to believe that there are two scenarios involved. One the major compass errors and primariy from all indications the actual cause of the incident Two, would be possibly a controller problem with the right stick. That was really where I was going with that.

Agreed - the later data do not rule out a stick problem, but note that the right stick worked just fine to start the motors with a CSC. While it may seem odd that the OP did not touch that stick (and he does not mention applying any elevator or aileron - just climb and steering), I see no reason to invoke an entirely separate failure mode that isn't needed to explain the event. Even if he had applied elevator and aileron I'm not sure he could have recovered the situation - these incorrect positional holding attempts are one of the few events that really render these aircraft uncontrollable - toilet bowl in the case of errors around 90° and uncontrolled linear flight for 180° errors.
 
If you look at the yaw data, the aircraft heading was initialized at around 350°. After takeoff that heading changed to around 45° (NE) and the FC, detecting drift, applied positive (rearward) pitch. That should have resulted in the aircraft moving to the SW but, instead, it moves NE - indicating that it was actually facing SW. Still based on a yaw of NE, the FC applies even more positive pitch to attempt to correct the motion, but that just makes it worse. RTH doesn't work for the same reason and in the end it cannot resolve the compass errors, drops into ATTI and continues to drift away to the NE.

I'm sitting here reading through the analysis offered up by the forensics experts on this thread and I'm, well, envious that you can look at squiggles on a graph and perform a post mortem like you do, time after time, with all the ease and finesse of ordering lunch at McDonalds. I don't consider myself an idiot (although there are others who would) but, what you do with that data is impressive, to say the least. If you ever decide to host a series of classes on how to interpret DJI flight data, let me know because I'd sign up. It would be helpful if you accepted credit card payment for the tuition as well. Thanks.
 
I see no reason to invoke an entirely separate failure mode that isn't needed to explain the event.
I completely agree, but I did want to point out that part of the data was just odd to me. It is just so very rare to see two separate unrelated issues simultaneously. I just wanted to verify that I was looking at this correctly.
 
(and he does not mention applying any elevator or aileron - just climb and steering)
Just an added note on that. Would you not consider the right stick steering? Otherwise, you only have yaw and throttle. That is neither here nor there at this point, I think we know what the primary issue was. ;)
 
Before the OP goes looking, somebody check my thoughts. Drifting at 15 mph for 15 minutes, would be 3.75 miles, this is what I came up with.
3.7miles a.jpg

Rod
 
I'm sitting here reading through the analysis offered up by the forensics experts on this thread and I'm, well, envious that you can look at squiggles on a graph and perform a post mortem like you do, time after time, with all the ease and finesse of ordering lunch at McDonalds. I don't consider myself an idiot (although there are others who would) but, what you do with that data is impressive, to say the least. If you ever decide to host a series of classes on how to interpret DJI flight data, let me know because I'd sign up. It would be helpful if you accepted credit card payment for the tuition as well. Thanks.

Thanks for your comments. It's not as hard as it looks, especially if you have a background in a scientific field (I'm a practicing physicist) and if you have spent some time analyzing this type of data.There's always a tendency to be over-impressed by the work of others in areas in which you have less experience, but once you get into it then it makes a lot more sense.

As a result of widespread interest I am putting together a guide to DJI log analysis, but I do this stuff for fun, so it's not going to cost anything.
 
Just an added note on that. Would you not consider the right stick steering? Otherwise, you only have yaw and throttle. That is neither here nor there at this point, I think we know what the primary issue was. ;)

Right stick is not really steering. You can apply aileron to add sideways motion, but when you remove the input the heading hasn't changed.
 
Thanks for your comments. It's not as hard as it looks, especially if you have a background in a scientific field (I'm a practicing physicist) and if you have spent some time analyzing this type of data.There's always a tendency to be over-impressed by the work of others in areas in which you have less experience, but once you get into it then it makes a lot more sense.

As a result of widespread interest I am putting together a guide to DJI log analysis, but I do this stuff for fun, so it's not going to cost anything.
I’m interested, never attempted to use. Over 4 years flying drones have crashed a lot, 95% pilot error, 5% possible Russian meddling, but one was an attempt to take off at a water side pier that evidently had lots of rebar, compass error, so moved to grassy area until gps ready, but due to trees had to move back to pier. Should have hand launched(never done) but set on pier, got no compass alarm and quickly took off but then it went crazy going wrong direction right into building and barely climbed to miss it and went out of sight behind it. Hit rth and she came back erratically but managed to hand catch. Just an example of compass problems around buried metal.
 
I'm sitting here reading through the analysis offered up by the forensics experts on this thread and I'm, well, envious that you can look at squiggles on a graph and perform a post mortem like you do, time after time, with all the ease and finesse of ordering lunch at McDonalds. I don't consider myself an idiot (although there are others who would) but, what you do with that data is impressive, to say the least. If you ever decide to host a series of classes on how to interpret DJI flight data, let me know because I'd sign up. It would be helpful if you accepted credit card payment for the tuition as well. Thanks.
^^What he said. I'm not stupid. I have been a Ham, FF/EMT, collected all sorts of certifications and taken various classes. I play around with computers and tech all the time but this sort of data analysis is out of my league. Color me impressed.
 
HUM?
Did @Fosterlandsvakt get lost looking for his Phantom?
I hope I didn't lead him to a bad part of town...

Rod
 
  • Like
Reactions: Neon Euc
[QUOTE = "RodPad, post: 1344107, medlem: 26059"] HUM?
Föll @ Fosterlandsvakt bort och letade efter sin Phantom?
Jag hoppas att jag inte ledde honom till en dålig del av staden ...

Tappen [/ QUOTE]
No im still here and looking for my phantom 4 advance. Thanx 4 all tips. I have look where your maps says i coud be but not found it yet
 
  • Like
Reactions: pa2x

Members online

Forum statistics

Threads
143,055
Messages
1,467,298
Members
104,919
Latest member
BobDan