Flight log doesn't show real course during flyaway

[F-visuals]

Hi everyone !
I come to you as I wish I could find some answers to what happened with my P3P lately.
Please find flight log over here :
Flight Log
And please look at the video that will 'prove' flight log is just totally wrong (and will show you how "f..." close I was from a "disaster").
Footage

So basically what happened here: I was flying waypoint without troubles. Suddently I got this error message saying that the drone was switching to "atti mode" and that's when the drone went all nuts. First it speened and then became really really hard to control with actions not in adequation with control sticks movements.
I barelly managed to bring it back and roughly landed it.

What's bothering me here is that the flight log is just REALLY FAR to tell what indeed happened.
If you have a look at the "large map" in the general tab it shows that the serie of events happened quite the same point one the map when all these events actually happened while I was far away trying to get control back.
Looking at the flight log, well it says the gps signal stayed strong and good all time. If so, why the flight log doesn't show the real course of the bird ??? (looks like the last location of the drone was well registered before flyaway but once it switch the first time to atti, it disappear from earth, you know, like a drone getting zero gps signal after popping out the gps module cable for example)
The list of notification is also really weird (especially at event N while I was actually landing and it says I reached the maximul altitude flight O__o)


I actually had the same flyaway next day. I made a IMU calibration before in the morning (but not a compass one - might be a mistake ?)
Flight Log
This time I was lucky too but less. Managed to "crash" it from not too high in a field as there was no way to bring it back (stick responses were just too wrong). But it took me quite some time (and luck) to find it as I lost visual with it during flyaway ;-)


To give a little bit of background here, I recently crashed in shallow waters (total pilot stupid mistake). I made it repaired and they changed the motherboard. I only made a clean compass calibration, assuming they did on their side the IMU one (of course now they won't admit if they forgot...)
So I'm somehow wondering if it could be the result of a problem with gps module ? I mean, drone went in waters and if they just said that motherboard needed to be changed, it might also be possible that gps module could have been damaged...



Drone is now in service and they say everything's just fine with it...
They say that as I was flying close to a no fly zone, the compass was affected. Hummm ok but that flight close no fly zone was the second flyaway. How can you explain the first one ?

Anyway I find hard to believe there's nothing wrong with the drone. Does anyone see any valuable information in flight logs that could help spot the problem ?

Thanks very much for any input.
Best

 

[Meta4]

So basically what happened here: I was flying waypoint without troubles. Suddently I got this error message saying that the drone was switching to "atti mode" and that's when the drone went all nuts.
Anyway I find hard to believe there's nothing wrong with the drone. Does anyone see any valuable information in flight logs that could help spot the problem ?

Unfortunately HealthyDrones only gives a basic summary with no details.
If you want to know what happened go to https://www.phantomhelp.com/LogViewer/Upload/
Follow the instructions to upload your flight record.
Come back and post a link to the report it provide.

BTW .. what HD does show is that your Phantom was in atti mode which might explain the difficulty you had with the controls.
Have you had experience with atti mode to know what to expect and how to handle it?

 

[BobMcKinlay]

I've had similar situations. Although it's just a suspicion, I think it's from poor compass calibration. I know many advise against it (more chances for something to go wrong, not needed, etc), but I tend to calibrate the compass before every session. Prior to doing that I had 2-3 unexplained flyaways. Afterwards, none. I've also made sure to calibrate in an open field with no RC, phone/tablet, keys anywhere on my person. No hard evidence, just an idea.

 

[F-visuals]

Thanks guys for input.
@Meta4
conditions of flight were not exactly good, pretty windy. Can't give a number but it was strong, that's for sure.
It was probably one of the first time I really had to fly atti. I tried once to show a friend how crazy a drone would drift in big wind. And that's it. Never really realized that one day it might save my bird. Never realized how incomplete a pilot can be if he doesn't know to pilot without assistance. I was just like probably 90% of drone operators, trusting the gps mode with all my heart.
I can't say how much I do think this is a big mistake and I can only invite people to learn to fly, for real, in atti mode. Just hope that my experience will help other avoid this huge amount of stress that squeezed my heart at the time....
However, regarding this experience, on the next day, with quite the same windy conditions, I decided that it was definitely time to give a try to atti. Didn't do a whole crazy flight in this mode but switched time to time to let the drone drift and try to take control back. It wasn't easy but nothing to compare with difficulties I had previous day.
But when the second flyaway happened, it was just a total different story. It was really impossible to manage anything. The whole time remote switch stayed on P mode. Now I do realize that when I saw **** was happening, I should have manually switch to "real" atti mode A. I regret not doing that as I somehow have the feeling things would have been better and more easy to manage.
So yes, inexperience with atti wasn't a good point but honnestly the GPS atti was acting really really funky. Difference during the same flight and same conditions were to extreme between this mode and the real atti "A". On one side a flight with something quite not responsive to any input on sticks and on the other, a more difficult time to fly but really and honestly manageable.
I'm gonna send the log to the adress your gave me and will be back there.

@BobMcKinlay
thanks for your point of view. I'm more of the kind of not doing my compass all the time, especially as living in urban area. So that's why each time I really move of location, I just find a "clear" spot and do a clean compass. And that's what I did when taking the drone back (it actually crashed in waters very far from home when I was traveling). So a poor compass calibration, I don't know. It's hard to tell. I have the feeling the spot I used was good but you never know what's underground

Anyway, thanks both of you !
Best

 

[F-visuals]

Here we go for the flight records
Log first flyaway
Log second flyaway

 

[Meta4]

Here we go for the flight records
Log first flyaway
Log second flyaway

Both of those flights exhibit the same problem.
The first starts at 13:50 with Speed Error and GPS positioning off message.
These go on for about a minute and for this period, the Phantom is getting bad data from the GPS which explains the crazy scattered position plot and some crazy speeds.
The Phantom switched to atti itself for some of this time because of a conflict between the compass and GPS data.

The second flight starts with the same error messages at 6:59 and again switches to atti because of a data conflict.

My guess is a problem with the GPS system in your Phantom.
Having this on two separate flights is very suspicious and definitely not normal or caused by anything you could do.
Perhaps a loose connection or a faulty GPS component??
But definitely needing to be fixed before you fly again.

 

[Luap]

You could post the dat log of the flight you posted the footage from.
I or others could then check further.
Here is an example from a recent incident Standard - Insight on my Fly Away? - HealthyDrone & YT Link
And some data you get out of it in below video.
The no rotation is the yaw values from the log. You will not that P3 orientation doesn't match the footage.
The 130 degree rotation I adjusted manually to match the footage. When it returns home I had to adjust to 40 degrees.

 

[mdowney58]

Holy COW! There were several darned close calls in that flight video!

Sent from my SM-N910V using Tapatalk

 

[F-visuals]

Thanks guys
Much appreciate input and concern.
@Luap
I'd be more than happy to post the dat log. However as briefly mentioned the drone is at the repair shop at the moment. As soon I'll get it back I will post it !

Best !

 

[Mimoid]

A compass error in flight will immeditely incapacitate most of the automatic guidance functions of the Phantom (incl RTH).

If you encounter a compass error, don't panic, but treat your Phantom as a fully manual drone, without any automation whatsoever. Abandon all GPS functionality and immediately switch to Atti mode. Fly home manually in FPV mode and land asap. Whatever you do, do not engage RTH, as the Phantom will then fly to where it THINKS is the home point - which in case of a compass error scenario is most likely completely wrong!

As long as you have FPV, you are in control, and you should make sure to remain in control! Don't lose the RC signal or the FPV signal, or you will be in a world of hurt.

I would think that the majority of the reported flyaways is caused by wind drift effects in Atti mode, manual flying in distress and with loss of orientation or an initiated auto-navigation function with a scrambled compass.

That is why I strongly advocate against recalibrating the compass more often than absolutely necessary. Every compass calibration should be followed by a strictly controlled test flight, at least 100 meters away from the home point, and with a section of the flight in each cardinal direction. If you have managed such a test flight at one place with green status all the way and without strange gimbal tilt aberrations, your compass is good and perfectly calibrated. From there on, you really don' t have to recalibrate again. In fact, avoid it like the plague. Compass calibration is kind of like a one-time activity, unless you travel 100 km or more.

//Tom

 

[F-visuals]

Thanks Tom
Your insight is very valuable and confirm me what I was thinking. From the moment I saw the bird getting weird I should have switched to A mode straightforward !
These two flights have definitely teach me much more than any others.
And I somehow hope a lot of pilots will see this threads and stop to be like the dumb...ss I was.
GUYS, please learn A mode before you have to face a **** like I did. For the ones that had a look at the video, you'll understand that of course I could have crashed the drone but that's not what matters. This drone and I mean by that the unaware pilot in case of technical failure I was, could have cause an accident, possibly injure people, possibly kill them without wanting to sound too dramatic...

 

[Mimoid]

Thanks Tom
Your insight is very valuable and confirm me what I was thinking. From the moment I saw the bird getting weird I should have switched to A mode straightforward !
These two flights have definitely teach me much more than any others.
And I somehow hope a lot of pilots will see this threads and stop to be like the dumb...ss I was.
GUYS, please learn A mode before you have to face a **** like I did. For the ones that had a look at the video, you'll understand that of course I could have crashed the drone but that's not what matters. This drone and I mean by that the unaware pilot in case of technical failure I was, could have cause an accident, possibly injure people, possibly kill them without wanting to sound too dramatic...

True.

Flying in Atti mode should be fully mastered once you begin to venture beyond visual reach. That way you can rely on your own ability to get the drone back to you without relying on the built-in automatic navigation functions.

It's actually quite good to see that it is not so difficult to fly in that mode, as long as you are sufficiently high up so you don't accidentally drift into a tree or something. You will feel safer and much more relaxed once you have been able to fly a predefined circuit a thousand feet out in Atti mode only.

As long as you have visual feed from the drone to your tablet/smartphone you are good to go, really!

//Tom

 

[F-visuals]

Hi everyone,
just took back my bird today. They say they haven't found anything suspicious, which sounds pretty strange.
Anyway, @Luap, I extracted the 2 dat logs. However, not sure it's the good ones as I have no clue on how to "read" that
DAT
Much appreciate any opinion !

Best

 

[Jesse James]

Dropped my in water total lose. Buying new one today. Found them cheap $100 dollars Cheaper then eveywhere else


Sent from my iPhone using PhantomPilots mobile app

 

[Jesse James]

Good luck. Water and droins don't mix.


Sent from my iPhone using PhantomPilots mobile app

 

[Luap]

Hi everyone,
just took back my bird today. They say they haven't found anything suspicious, which sounds pretty strange.
Anyway, @Luap, I extracted the 2 dat logs. However, not sure it's the good ones as I have no clue on how to "read" that
DAT
Much appreciate any opinion !

Best

Had a look at the 2 Dat log but does't appear to match the video footage. If you have Dats dated from day you posted footage could you post them?

 

[F-visuals]

Hi, @Luap
sorry about the mistake, it was indeed wrong flight dat log. I corrected the zip file. Link remains the same

Thanks !

 

[Luap]

This looks like it might be correct Dat but as you say flight log looks total wrong. As said above seems like some hardware related GPS issue.
I synched the footage to the telemetry when P3 tips over at landing.

 

[Luap]

Hi, @Luap
sorry about the mistake, it was indeed wrong flight dat log. I corrected the zip file. Link remains the same

Thanks !

Any chance you have other dats? I'm not sure this is the correct one. Do you have the video footage from start to land?