DJI Mavic, Air and Mini Drones
Friendly, Helpful & Knowledgeable Community
Join Us Now

Odd Flight Problem

Romonaga

Well-Known Member
Joined
May 6, 2020
Messages
51
Reactions
65
Age
60
Location
Montgomery Il
Yesterday, my daughter and I went flying at a area close to our house that has some interesting creeks and areas to fly. Flight conditions were good, no real wind, good visibility (5mph). I was flying the Mavic 2 Pro, she was flying the P4.

After flying my daughter and I were talking, and we both experienced an issue with our drones. No drones were harmed or lost, but I am a loss to explain or understand what we both experienced, in the same spot.

At different times, and each alone and not sharing the airspace, we went down to this section of the creek to film over the water, My altitude over the water was 10-15 feet, my daughter was feeling a bit braver and was flying a bit lower, but not low enough to account for what she saw and I saw.

First what I experienced. I flew into this area using activetrack, once I got to the bank I took control of the drone and proceed with manual flight. About 20 feet out and 12 feet up, the drone started to descend with no inputs on control. I bring her back to height, and she starts to descend and yaw to the right. At this point, I am not sure whats going on, I assume user error, I raise her to about 15 feet and proceed to fly her back to me. Mind you the complete time this is happening, she is only about 20 feet from me. As I am flying her back, she starts to descend rapidly, she did respond to controls and I was able to bring her back.

Now, this is a new bird, so I assumed that I had done something wrong, after all this is the first time using activetrack. But when I went back and talked to my daughter and told her what happened. She asked about where I was, and told me she had the same issues in that same spot, and when reviewing the film from both drones, you can clearly see this happening.

In my film, you can see me controling the bird as it is descending, and you can see the very puzzled look on my face as it was trying to take a bath, as well as clearly see the controller in my hands. I assumed that I did something wrong when I turned off active track.

I do know that flying over water has its risks and that drones can get confused, and I am willing to accept error on my part. However, I think my distance from water was correct, and outside this small section of the creek, I have never seen this happen. I do not like flying over water as it is, but some fantastic shots can be had.

One detail that may be important. The take off point was higher then where the creek bed was, by about 10 feet.

So anyone have any idea? I do plan on downloading the logs in an attempt to analyses.

Edit to fix facts, as memory is often wrong.
 
Last edited:
So anyone have any idea? I do plan on downloading the logs in an attempt to analyses.
Key word: Logs. Please do upload them, there are members here who can help with you out with answers. I suppose you do know how to upload Txt log file, if not, take a look here (section "3. Log file retrieval"):


Quote from the guide:
"If you want help analyzing the data then you can either retrieve the logs and post them here directly, or upload them to PhantomHelp or AirData and post the link back here."



Hope this is helpful.
 
  • Like
Reactions: Romonaga
Key word: Logs. Please do upload them, there are members here who can help with you out with answers. I suppose you do know how to upload Txt log file, if not, take a look here (section "3. Log file retrieval"):


Quote from the guide:
"If you want help analyzing the data then you can either retrieve the logs and post them here directly, or upload them to PhantomHelp or AirData and post the link back here."



Hope this is helpful.
Thanks for the refresher course, I do know the logs are valuable, but I wanted to get details recorded least I forget. Once I have time today, I will pull the logs.
 
  • Like
Reactions: Rockycat and Dedal
The interesting points would be between H and J. While I was lower then I thought in my flight, it still does not account for what I experience. Or, it could be that I did nose dive the bird while driving her forward in a rush to get away from what I perceived as abnormal flight.

Again, 2 birds a P4 and the Mavic 2 Pro both experienced very abnormal flight in this same location, no other location that day were problematic.

Here is my uploaded flight. If someone can tell me what I did wrong, that would be fantastic.

H02m 07s7.9 ft296 ftModeMode changed to P-GPS
I03m 09s10.8 ft317 ftTipObstacle Avoided. Revise Flight Route
J03m 46s12.5 ft289 ftModeMode changed to ActiveTrack
 
To narrow it more. I do not remember applying down thrust at 3m9.6s until 3m17.5s This is after the warning about obstacle, just before the bird was rising and falling, so I started to bring her home. Speed was not excessive enough to think speed related. So did I, or did I not apply down thrust at this time, or what exactly did I do wrong.
Again, starting home point was higher then this spot.

3m 9.4sP-GPS19satellites11.2ft5.9ft7.2mph317.4ft85%16.102V4.028V4.027V4.03V4.021V0.007V
3m 9.5sP-GPS19satellites10.8ft5.6ft6.4mph316.9ft85%16.264V4.072V4.075V4.07V4.046V0.029VObstacle Avoided. Revise Flight Route
3m 9.6sP-GPS19satellites10.5ft5.2ft6.4mph316.5ft85%16.264V4.072V4.075V4.07V4.046V0.029V
3m 9.7sP-GPS19satellites9.8ft4.9ft3.8mph316.1ft85%16.264V4.072V4.075V4.07V4.046V0.029V
3m 9.8sP-GPS19satellites9.5ft4.6ft2.3mph315.9ft85%16.264V4.072V4.075V4.07V4.046V0.029V
3m 9.9sP-GPS19satellites8.9ft4.6ft1.4mph315.9ft85%16.264V4.072V4.075V4.07V4.046V0.029V
3m 10sP-GPS19satellites8.5ft4.6ft0.9mph315.8ft85%16.264V4.072V4.075V4.07V4.046V0.029V
3m 10.4sP-GPS19satellites6.6ft4.3ft0.2mph315.6ft85%16.264V4.072V4.075V4.07V4.046V0.029V
3m 14.3sP-GPS19satellites5.9ft4.9ft1.6mph313.5ft85%16.24V4.062V4.062V4.06V4.055V0.007V
3m 15sP-GPS19satellites5.9ft4.9ft0.2mph312.9ft85%16.222V4.058V4.057V4.06V4.051V0.007V
3m 17.5sP-GPS19satellites7.9ft6.9ft0mph312.6ft84%16.169V4.06V4.058V4.04V4.013V0.047V
3m 19.5sP-GPS18satellites8.9ft8.5ft0mph312.7ft84%16.225V4.059V4.057V4.06V4.052V0.007V
3m 20sP-GPS18satellites9.2ft8.5ft0mph312.7ft84%16.215V4.058V4.055V4.05V4.048V0.01V
3m 20.1sP-GPS19satellites9.2ft8.5ft0mph312.7ft84%16.215V4.058V4.055V4.05V4.048V0.01V
3m 21.9sP-GPS18satellites9.2ft8.9ft0mph312.4ft84%16.219V4.059V4.057V4.06V4.048V0.011V
3m 22.7sP-GPS19satellites9.2ft8.9ft0mph312.3ft84%16.223V4.06V4.058V4.06V4.05V0.01V
 
Last edited:
Well posted logs and lots of detail, post was locked, removed posts in hope of getting unlocked, but still locked. Please remove this thread as I have no desire to retype and at this point not sure what I did wrong.
 
While my post has been restored, I still do not know why when I posted my logs my post was flagged for moderation. While it is nice to see my post restored, I would still like to understand why posting my log file resulted in moderation.

I truly did want some feedback on what we experience.
 
While my post has been restored, I still do not know why when I posted my logs my post was flagged for moderation. While it is nice to see my post restored, I would still like to understand why posting my log file resulted in moderation.

I truly did want some feedback on what we experience.
Hi Romonaga! I have seen this happen a few times. The forum moderators seem to be employing a system of sorts whereby file attachments to posts made by new/recently joined members are automatically flagged for review by a mod before they appear in the post (for security reasons I presume). But once they are reviewed and approved, they do appear. I am basing this solely on what I have observed in my time here and don't really have the details of how this is implemented. However, I can tell you that you are not being personally targeted here and you shouldn't take this personally. Perhaps a moderator could drop by and enlighten us all; @dirkclod, @Thomas B
 
at this point not sure what I did wrong.
It's unlikely that you did anything wrong.
Forums use software to pick out items that might be moderation-worthy, but it's up to human eyes to check them.
Often the software will pick up scams and spam but occasionally completely innocuous posts get picked out because of some combination of words or something else triggers the software.
It's not something you need to worry about.
 
Hi Romonaga! I have seen this happen a few times. The forum moderators seem to be employing a system of sorts whereby file attachments to posts made by new/recently joined members are automatically flagged for review by a mod before they appear in the post (for security reasons I presume). But once they are reviewed and approved, they do appear. I am basing this solely on what I have observed in my time here and don't really have the details of how this is implemented. However, I can tell you that you are not being personally targeted here and you shouldn't take this personally. Perhaps a moderator could drop by and enlighten us all; @dirkclod, @Thomas B
Mods do not hold anything back the system holds Threads/post in queue that have key words that might relate
to spam and I don’t know what all else. Mods do look at them and if it looks ok we approve it. If it’s questionable it’s left to a Admin to clear it.Nobody is being singled out .
 
While my post has been restored, I still do not know why when I posted my logs my post was flagged for moderation. While it is nice to see my post restored, I would still like to understand why posting my log file resulted in moderation.

I truly did want some feedback on what we experience.
Then you should have waited and not deleted your own post
which I just restored.
 
I imagine that this is the spot where you felt the oddities ...?

1589376252659.png

Looking into the data during the upper red flight path over the water during length of the black arrow doesn't reveal anything abnormal regarding Yaw (Dark yellow) vs. Rudder inputs (Yellow) ... they corresponds perfectly to each other. Equal with the Elevator inputs (Grey) vs. Heading speed (Black).

But looking into the height changes reported from the VPS sensor (Magenta) vs. Throttle inputs (Green) vs. vertical Z-speed (Blue, positive=descend) we see uncommanded movements. As seen where the markers are in the chart the VPS height (Magenta) have been increased uncommanded since approx 172sec (flat water surface I assume). After 184,5sec the VPS height instead drops together with a positive vertical Z-speed, meaning descend (Blue). The total VPS height descend is from it's highest 6,4m down to 1,3m, this with two instances of positive Throttle inputs included (Green) to counter the descent. At 191,5sec the vertical Z-speed is again zero & the VPS height is stable. To the note the height change is also recorded by the barometric height (Red) with a tad of delay.

The only explanation I can think of here is that the water surface below the AC tricked the VPS sensor, perhaps it was streaming water just there which created extensive reflections & the VPS tried to first adjust up ... then just later decided to adjust back but then went to far perhaps still misjudging against the water.

1589376418865.png
 
  • Like
Reactions: Romonaga
Mods do not hold anything back the system holds Threads/post in queue that have key words that might relate
to spam and I don’t know what all else. Mods do look at them and if it looks ok we approve it. If it’s questionable it’s left to a Admin to clear it.Nobody is being singled out .
Please understand, I was worried I did something that was breaking the rules.
 
Please understand, I was worried I did something that was breaking the rules.
No you just got caught up in queue. If you had did something wrong you would have known it ??
 
I imagine that this is the spot where you felt the oddities ...?

View attachment 101192

Looking into the data during the upper red flight path over the water during length of the black arrow doesn't reveal anything abnormal regarding Yaw (Dark yellow) vs. Rudder inputs (Yellow) ... they corresponds perfectly to each other. Equal with the Elevator inputs (Grey) vs. Heading speed (Black).

But looking into the height changes reported from the VPS sensor (Magenta) vs. Throttle inputs (Green) vs. vertical Z-speed (Blue, positive=descend) we see uncommanded movements. As seen where the markers are in the chart the VPS height (Magenta) have been increased uncommanded since approx 172sec (flat water surface I assume). After 184,5sec the VPS height instead drops together with a positive vertical Z-speed, meaning descend (Blue). The total VPS height descend is from it's highest 6,4m down to 1,3m, this with two instances of positive Throttle inputs included (Green) to counter the descent. At 191,5sec the vertical Z-speed is again zero & the VPS height is stable. To the note the height change is also recorded by the barometric height (Red) with a tad of delay.

The only explanation I can think of here is that the water surface below the AC tricked the VPS sensor, perhaps it was streaming water just there which created extensive reflections & the VPS tried to first adjust up ... then just later decided to adjust back but then went to far perhaps still misjudging against the water.

View attachment 101193

Yep, that area was the spot, more so over the water then land. With the most concerning being when I was bringing it back and it decided it wanted a bath. I did not think I requested it to descend, your work confirmed it.

Thank you so much, for confirming what I thought I experienced. Often, our memory is tainted during these events and I wanted to make sure of no error on my part.

I know flight over water is something that requires caution, its also why I fear flying over water.

As to your suspicion, the water in the center was calm, however closer to the bank, as you can see, there are things in the water that did cause rippling. I can see how this ripple could cause reflections that could fool the sensors.

In the end, the bird was safe, did not even consider this a close call. Will just have to be on guard over water, as i normally am. Did I mention I hate flying over water.
 
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Forum statistics

Threads
130,586
Messages
1,554,112
Members
159,586
Latest member
DoubleBarS