Pursuing a refund - Flyaway not found...

I had the same type of flywaway the OP described about 2 weeks ago, l was able to bring her in for a somewhat hard landing after a few minutes of wrestling with the controls. Here's a copy of my log from that particular flight, maybe someone could chime in as to what could have caused this to happen...

EDIT: I had 2 previous flights back to back from the same location with no problems.
 

Attachments

  • DJIFlightRecord_2015-07-26_[20-38-55].txt
    256.5 KB · Views: 245
I had a similar experience,
However mine also some GPS compass errors logged on the controller.
With no Phantom, all DJI did was offer 30% off retail cost of a replacement.
I bitched and the offered me 50% off the cost instead.
I had to pay for the shipping back of my controller and they eventually sent me a new unit.
I'm glad there was at least some resolution. How long did the process take from sending the log file to hearing back from DJI? And then how long from hearing back from them to receiving your replacement?
 
OP I'm sorry for your loss, really. It is stomach wrenching when it takes off on it's own mission and you just stand there helpless.

However -

Good luck with getting anything from DJI - and even better luck with your credit card company. Unless you can prove that the goods were faulty, they won't entertain you. Don't ask me how I know this (previous experience!). I've had phantoms for over two years, had two fly away and ditch in water, and both times the evidence pointed to (1) pilot error in the case of my P1; and (2) external magnetic interference in the case of my V+. DJI didn't respond to either incident, and CC company (Nationwide) just about told me to feck off.

If I was you, I'd take the earlier advice and go and search around the area ringed and see if you can salvage what's left of your aircraft. From the clips posted, esp the second video clip, where the aircraft looks to be flying away sideways, I would conclude that something affected the compass. Maybe not your fault, but I'd wager that DJI would say that it wasn't their fault either, if you can get them to reply to you.

Sorry if this all sounds harsh, but after four phantoms, three flyaways (one recovery!) three rescues from trees and countless heart-in-the-mouth incidents you just become numb to reports of flyaways.

Doesn't the fact that I have the logs which show a successful flight several minutes earlier from the same location, no errors, calibration performed, and the Pilot app clearly showing a home point AND RTH being activated and ignored, constitute proof?

It can't be right that a Phantom can fly away with no notice at any given time, the user having done everything required for a safe flight, and be completely on his own. If I would have had a compass error showing then I would never have taken off. I've had compass errors in the past, and I've recalibrated before take off. Surely, this is DJIs role... to ensure that any time there is a compass error the app flags this up.

Would a compass error cause the drone to move without any input from the RC? I understand that this would only cause the drone to move in the wrong direction when it's told to move (this is if the compass error is static -ie. mistaking North for South. If the compass error was dynamic in nature, then the drone wouldn't have headed off in the smooth trajectory that it did)

If it is true that there was a compass error, then any claim that this is my fault, or at least not DJIs fault, would imply there were two simultaneous malfunctions: the one that made it move without a command to do so; and the one that made it move in the wrong direction when RTH was activated.

If north is in-fact south, but I don't walk in either direction, then I still don't move. In the same way, if north was south, and I took off but didn't tell my drone to move, then it should just hover in one spot.
 
Actually, a compass error could make it drift off. The GPS lock tells it where to be. The compass tells it what direction its facing... movement shows variations in location, so it references the compass to see what movement is needed to return. If the compass data is erroneous then it would correct to the wrong direction... leading to a cascade of false movement.
 
I've experienced this 'type of behaviour" when flying near metal. Apparently it totally screws up the compass and your ability to input directions and engages in a random flightpath. If you took off on concrete it might have been the rebar used to stabilize the mortar.
I hope you will finally find your P3. I use Trackimo but hope I never have to go looking'
 
Ok. So you're saying that if point X is where the drone should be, the drone is programmed to move toward this point. So say A is the actual drone position and this is north of X, then the drone will move south to reach it. But if a compass error tells the drone that north is south, and south is north, then it will actually continue to move further away while trying to reach it's location, ie. further and further north.

Is there a failsafe for this kind of scenario? The only solution would be to correct to compass, switch off the compass, or quickly realize what is going on and do the opposite, ie. Fly north when you want to go south. True?

Regardless, isn't there any kind of indication that the app gives that the compass is wrong? And why would the app show the compass as correctly callibrated when it is not? Also, how would this compass calibration change from correct to incorrect in a matter of minutes without moving location or changing any settings? Any locational interference can't have been present because I have flown there several times with no problem. What could be the cause?
 
The app has no way to validate the data from the compass. To do that you whould have to see in the app that the drone is being presented as facing south, but in reality is facing north. The only way to recover from this kind of failure would be to switch to atti mode and fly manual.. Which in many cases will lead to a crash since few people can actually accomplish this. ESPECIALLY when you cant even see it!
 
  • Like
Reactions: atki
Ok. So you're saying that if point X is where the drone should be, the drone is programmed to move toward this point. So say A is the actual drone position and this is north of X, then the drone will move south to reach it. But if a compass error tells the drone that north is south, and south is north, then it will actually continue to move further away while trying to reach it's location, ie. further and further north.

Is there a failsafe for this kind of scenario? The only solution would be to correct to compass, switch off the compass, or quickly realize what is going on and do the opposite, ie. Fly north when you want to go south. True?

Regardless, isn't there any kind of indication that the app gives that the compass is wrong? And why would the app show the compass as correctly callibrated when it is not? Also, how would this compass calibration change from correct to incorrect in a matter of minutes without moving location or changing any settings? Any locational interference can't have been present because I have flown there several times with no problem. What could be the cause?

Additionally, you write like, right from the start of the flight, the drone was trying to get back somewhere. But when it initially took of it was already in that place, so didn't have to move. Can this "cascade" be caused by a drone movement of just a couple inches?

Imagine a room with no breeze, and a perfectly flat surface from which to take off, and the most perfect take off. Now, if the compass was wrong, but the drone flew directly up and hovered in the spot, the compass error would cause no movement at all, false or otherwise.

Now imagine a field with some wind, an uneven take off, or something similar. Would this very very slight deviation from the take off point coupled with the faulty compass, cause a fly-away?
 
  • Like
Reactions: atki
Ok. So you're saying that if point X is where the drone should be, the drone is programmed to move toward this point. So say A is the actual drone position and this is north of X, then the drone will move south to reach it. But if a compass error tells the drone that north is south, and south is north, then it will actually continue to move further away while trying to reach it's location, ie. further and further north.

Is there a failsafe for this kind of scenario? The only solution would be to correct to compass, switch off the compass, or quickly realize what is going on and do the opposite, ie. Fly north when you want to go south. True?

Regardless, isn't there any kind of indication that the app gives that the compass is wrong? And why would the app show the compass as correctly callibrated when it is not? Also, how would this compass calibration change from correct to incorrect in a matter of minutes without moving location or changing any settings? Any locational interference can't have been present because I have flown there several times with no problem. What could be the cause?

Check my posts

You can see the overlay of stick input vs actual actions.
There were two back to back flights where it happened.
First time I recovered. Second I did not.
The phantom appears to actually tilt on a 45 degree angle (not just the gimbal) as it take a angled dive.
Atti mode mad no difference as the phantom would not level.
 
The app has no way to validate the data from the compass. To do that you whould have to see in the app that the drone is being presented as facing south, but in reality is facing north. The only way to recover from this kind of failure would be to switch to atti mode and fly manual.. Which in many cases will lead to a crash since few people can actually accomplish this. ESPECIALLY when you cant even see it!

Then what is happening when the app tells you to calibrate?

I assume it's noticed the error and asked you to fix it. IE. It's seen that the data for the compass is invalid, hasn't it?
 
Then what is happening when the app tells you to calibrate?

I assume it's noticed the error and asked you to fix it. IE. It's seen that the data for the compass is invalid, hasn't it?
Its probably seeing a movement in the compass but no movement in the gyro
 
i had the same problem (immediately flying left after take off), but i didn't use auto-takeoff. my bird did not respond to my request to return to home (other than turn to face home as it continued on its crazy flight away from me). i lost connection to it about a mile away. i jumped in my car and drove to the last known spot, and i was able to reconnect. it had drifted a hundred feet or so from that spot and was slowly drifting further away. in my panic, i didn't think of switching to ATTI mode (though i have read that it allows the pilot to regain control), but will if it happens in the future. i learned a big lesson in that it pays to not turn off your controller and to take it with you when searching for a lost Phantom. i also purchased a retractable dog leash to attach to the bottom of my Phantom to prevent this type of fly-away in the future. once i get it up and confirm that it is stable, i will detach the leash and continue flying.
 
Actually, a compass error could make it drift off. The GPS lock tells it where to be. The compass tells it what direction its facing... movement shows variations in location, so it references the compass to see what movement is needed to return. If the compass data is erroneous then it would correct to the wrong direction... leading to a cascade of false movement.

Compass problems do not cause the Phantom to drift off in one direction. It will cause it to circle as in TBE. This does not look like a compass related issue.
 
Compass problems do not cause the Phantom to drift off in one direction. It will cause it to circle as in TBE. This does not look like a compass related issue.

Oh come on guys! Now I don't know what to think :(

I appreciate all your responses. I think I'll refrain for jumping to conclusions about causes, seen as their seems to be disagreements between people who are much more experienced than I am.

All I know is:

1. First flight was successful
2. Home point and RTH were responsive on first flight
3. I did not change settings
4. I did not change location
5. I did not receive any error messages
6. From the outset of my second flight I had no control
7. RTH did not work
 
  • Like
Reactions: dalebb
Compass problems do not cause the Phantom to drift off in one direction. It will cause it to circle as in TBE. This does not look like a compass related issue.
I disagree.. it would totally depend on what is actually happening. There are far too many variable to simply say that a compass, that has the ability to input a near infinitely vast spread of input to the NAZA to only cause one end result. A false, yet steady compass output not only could lead to this flight path, but would also attribute to the steady increase in speed as the controller keeps fighting to correct. At the end of the day its all nothing but speculation. All we can do is look at the data we have and apply some logic to it. I dont think anyone has said THIS is what happened, or THIS isnt what happened. Only what might have.
 
I disagree.. it would totally depend on what is actually happening. There are far too many variable to simply say that
a compass, that has the ability to input a near infinitely vast spread of input to the NAZA to only cause one end result. A false, yet steady compass output not only could lead to this flight path, but would also attribute to the steady increase in speed as the controller keeps fighting to correct. At the end of the day its all nothing but speculation. All we can do is look at the data we have and apply some logic to it. I dont think anyone has said THIS is what happened, or THIS isnt what happened. Only what might have.

Noted. Thanks Streve.
 
Your prior flight really has no bearing on anything. A light bulb works fine the first 10000 times its turned on. That doesn't mean it should work the 10001 time. Eventually electronics fail. Some prematurely. I'm not discounting the fact that it sucks. Just that it will all fail.. at some point in its life. You just got screwed with something failing very early!
 
I disagree.. it would totally depend on what is actually happening. There are far too many variable to simply say that a compass, that has the ability to input a near infinitely vast spread of input to the NAZA to only cause one end result. A false, yet steady compass output not only could lead to this flight path, but would also attribute to the steady increase in speed as the controller keeps fighting to correct. At the end of the day its all nothing but speculation. All we can do is look at the data we have and apply some logic to it. I dont think anyone has said THIS is what happened, or THIS isnt what happened. Only what might have.

You can disagree but you'd be wrong. You don't understand how the Phantom aligns itself to a point.
 
Your prior flight really has no bearing on anything. A light bulb works fine the first 10000 times its turned on. That doesn't mean it should work the 10001 time. Eventually electronics fail. Some prematurely. I'm not discounting the fact that it sucks. Just that it will all fail.. at some point in its life. You just got screwed with something failing very early!

I mention the first flight to show that something other than the location or set-up of the craft was responsible.
 
  • Like
Reactions: III% Streve

Members online

Forum statistics

Threads
143,066
Messages
1,467,358
Members
104,935
Latest member
Pauos31