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

Mavic Mini "Flyaway" - forensic help requested

rwilliam99

Well-Known Member
Joined
Apr 12, 2020
Messages
291
Reactions
290
Age
60
Location
Oregon, USA
I took my Mavic Mini hiking to a waterfall last weekend. Getting a good GPS lock is difficult in a waterfall amphitheater, so I got it flying and hovered it as high as it would let me (hoping that would be more likely to get a signal up higher). I let it sit there for several minutes while it attempted to find satellites. When I thought it had a good lock, I started the video and started moving up the waterfall. When it got to the top of the waterfall, it did a corkscrew turn and flew into a tree. I was able to get it out of the tree, however it fell and broke the gimbal.

I've spent hours trying to figure out what happened so that I can learn from this crash, but I'm still a little perplexed. As best I can tell, my guess is that due to the poor GPS signal it had, it encountered some kind of wind vortex at the top of the waterfall (which made it do the corkscrew) that blew it into the tree.

I've viewed the flight log using airdata.com and I've attempted to process the .DAT file from the SDCard using both CSVView and DatCon, but due to the extremely large size of that file I've been unable to glean any insights from it. I was trying to just cut out the last portion of the .csv file it produced but have been unable to figure out exactly where to cut it. I've followed the steps in the sticky post and continue to try and glean useful information but feel like I'm not making much progress at this point.

Any tips or hints or insights from the flight log would be appreciated. Thank you!
 
Have a look at the webpage
Lower down are instructions concerning how to retrieve the flight logs, upload the relevant flight log to tat page and poet the reulting URL here.
You are interested in the logs on the screen device. From memory the fc.log.log DAT on the mSD card contains the DATs from several flight and as you said it is enormous, it takes a very long time to process.

CsvView will process both the .txt and ,DAT flight logs from the screen device.

Sorry about the drone.
 
Last edited:
Have a look at the webpage
Lower down are instructions concerning how to retrieve the flight logs, upload the relevant flight log to tat page and poet the reulting URL here.
You are interested in the logs on the screen device. From memory the fc.log.log DAT on the mSD card contains the DATs from several flight and as you said it is enormous, it takes a very long time to process.

CsvView will process both the .txt and ,DAT flight logs from the screen device.

Sorry about the drone.
I had only done the airdata.com one before, but here is the phantomhelp one:

 
  • Like
Reactions: dug611
Looking at that phantomhelp result it looks like things went sideways at 5m 2.2s.

I can't find a description for what "GPS Position Mismatch" means - it shows it had 8 satellites at that point, but it was also generating the mismatch errors as well.
 
At a quick glance the satellite count is 'poor' the whole way through the flight, is that normal for this Mavic Mini and or the area?
The drone seems to be above VPS height for the majority of the flight so I would have thought you were dicing with it going into ATTI mode for a lot of the flight.
In ATTI mode it is at the mercy of any wind unless you counteract that wind. Towards the end of the flight the GPS count drops to 5, which is bad.
From what I have read what we are really interested in is the GPS health, column 28 of the PH csv "OSD.gpsLevel" and that's 1 and less from 5:16 onwards. 4 might be the minimum necessary to set the home point, I think the max is 5.

I think this will come down to bad GPS, no VPS and basically the drone being blown into the tree.

I wouldn't have flown anything other than overhead with those GPS levels, out of doors my Mavic Mini generally has 10 or more satellites, often in the higher teens and sometimes breaks into the 20's.

But you really need @slup. @sar104, or @Meta4 to weigh in here.
 
At a quick glance the satellite count is 'poor' the whole way through the flight, is that normal for this Mavic Mini and or the area?
The drone seems to be above VPS height for the majority of the flight so I would have thought you were dicing with it going into ATTI mode for a lot of the flight.
In ATTI mode it is at the mercy of any wind unless you counteract that wind. Towards the end of the flight the GPS count drops to 5, which is bad.
From what I have read what we are really interested in is the GPS health, column 28 of the PH csv "OSD.gpsLevel" and that's 1 and less from 5:16 onwards. 4 might be the minimum necessary to set the home point, I think the max is 5.

I think this will come down to bad GPS, no VPS and basically the drone being blown into the tree.

I wouldn't have flown anything other than overhead with those GPS levels, out of doors my Mavic Mini generally has 10 or more satellites, often in the higher teens and sometimes breaks into the 20's.

But you really need @slup. @sar104, or @Meta4 to weigh in here.

What is VPS?

When I fly it in my backyard, with a clear view of the sky, I get 11-12 satellites with this drone. Waterfall amphitheaters are difficult for GPS due to the geography (steep, high walls on 3 sides), so I knew that going in, but I thought I had gotten it to lock on enough satellites - I did this a couple weeks prior at another waterfall and all was good - that waterfall was a bit more open than this one was, however which may have contributed to its success.

I'm still very novice, but am trying to learn from my failures (without destroying my mini in the process).
 
  • Like
Reactions: WAXMAN
VPS refers to the downward looking visible light and IR sensors on the bottom of the drone.
They have limited range but when the drone is low enough for them to work they help the drone maintain postition when the drone is over suitable surfaces.
IMO VPS position holding is, when available, far more accurate than GPS.
Note VPS is also responsible for the descent speed control whan the drone is near the ground AND the landing protection

Not being nasty but it appears you should read the drone's manual.
https://dl.djicdn.com/downloads/Mavic_Mini/Mavic_Mini_User_Manual_v1.2_enI.pdf

other useful and not so useful downloads can be found on
 
  • Like
Reactions: WAXMAN
VPS refers to the downward looking visible light and IR sensors on the bottom of the drone.
They have limited range but when the drone is low enough for them to work they help the drone maintain postition when the drone is over suitable surfaces.
IMO VPS position holding is, when available, far more accurate than GPS.
Note VPS is also responsible for the descent speed control whan the drone is near the ground AND the landing protection

Not being nasty but it appears you should read the drone's manual.
https://dl.djicdn.com/downloads/Mavic_Mini/Mavic_Mini_User_Manual_v1.2_enI.pdf

other useful and not so useful downloads can be found on

No worries - I've read the manual - more than once, but there is a lot to consume and learn. Maybe a refresher is in order...

Thanks for the feedback.
 
  • Like
Reactions: oztx
I took my Mavic Mini hiking to a waterfall last weekend. Getting a good GPS lock is difficult in a waterfall amphitheater, so I got it flying and hovered it as high as it would let me (hoping that would be more likely to get a signal up higher). I let it sit there for several minutes while it attempted to find satellites. When I thought it had a good lock, I started the video and started moving up the waterfall.
Your drone didn't have reliable enough GPS data to record a home point until 5:05.7.
The crash happened just 6 seconds after that at 5:11.4.
I can't find a description for what "GPS Position Mismatch" means - it shows it had 8 satellites at that point, but it was also generating the mismatch errors as well.
It's a meaningless false alarm.
Most of those didn't appear until after the crash when the drone was tilted at crazy angles.
When I fly it in my backyard, with a clear view of the sky, I get 11-12 satellites with this drone. Waterfall amphitheaters are difficult for GPS due to the geography (steep, high walls on 3 sides), so I knew that going in, but I thought I had gotten it to lock on enough satellites
Without good GPS data, your drone has no ability to hold position horizontally.
It cannot stop and will continue drifting when you take your hands off the sticks.
It will also be blown on any wind.
Flying with obstacles close by without full GPS reception is just asking for trouble.
 
Last edited:
Your drone didn't have reliable enough GPS data to record a home point until 5:05.7.
The crash happened just 6 seconds after that at 5:11.4.

It's a meaningless false alarm.
Most of those didn't appear until after the crash when the drone was tilted at crazy angles.

Without good GPS data, your drone has no ability to hold position horizontally.
It cannot stop and will continue drifting when you take your hands off the sticks.
It will also be blown on any wind.
Flying with obstacles close by without full GPS reception is just asking for trouble.
I see many people getting 20+ satellites - is it possible the GPS receiver in my drone is faulty? Or do newer models have better GPS receivers? My backyard has little to no obstacles - seems like I should be able to do better than 11-12 sats. I'm considering upgrading to a mini 3 pro which also has 3 direction obstacle avoidance and I'm hoping a better GPS receiver.

I'm learning the hard way just how important a good GPS lock is for the drone. I thought I was being safe, but obviously not. I've even been seeing people reporting that the "The Home point has been updated" prompt may not even be good enough for a positive lock.

Thank you for the feedback and info.
 
I see many people getting 20+ satellites - is it possible the GPS receiver in my drone is faulty?
No ... your original Mini can receive signals from 2 satellite systems (GPS + Glonass).
It's normal for you to have sat numbers in the low teens with a clear sky.
Later models with access to three different sat systems are receiving more sats.
But the number of sats doesn't mean much.
Yours never had enough sats and sats with enough spread to provide good data until five minutes into your flight.
This was entirely due to the poor environment you chose to fly in.

GPS receivers need a clear view of a large part of the sky.
Obstacles like terrain , buildings and forest canopy block satellite signals and reduce or prevent your receiver doing what it should.
Or do newer models have better GPS receivers? My backyard has little to no obstacles - seems like I should be able to do better than 11-12 sats. I'm considering upgrading to a mini 3 pro which also has 3 direction obstacle avoidance and I'm hoping a better GPS receiver.
The GPS receiver in the Mini 3 receives sats from three sat systems.
But this doesn't mean that either is better or worse.
The number of sats is not what matters and after you have enough and are getting good location data, more sats won't make anything any better.

I've even been seeing people reporting that the "The Home point has been updated" prompt may not even be good enough for a positive lock.
If someone suggested that, they were wrong.
Your drone cannot and won't record a home point until it has good location data.
Read this post where I explained some more about this:
 
  • Like
Reactions: Ty Pilot
No ... your original Mini can receive signals from 2 satellite systems (GPS + Glonass).
It's normal for you to have sat numbers in the low teens with a clear sky.
Later models with access to three different sat systems are receiving more sats.
But the number of sats doesn't mean much.
Yours never had enough sats and sats with enough spread to provide good data until five minutes into your flight.
This was entirely due to the poor environment you chose to fly in.

GPS receivers need a clear view of a large part of the sky.
Obstacles like terrain , buildings and forest canopy block satellite signals and reduce or prevent your receiver doing what it should.

The GPS receiver in the Mini 3 receives sats from three sat systems.
But this doesn't mean that either is better or worse.
The number of sats is not what matters and after you have enough and are getting good location data, more sats won't make anything any better.


If someone suggested that, they were wrong.
Your drone cannot and won't record a home point until it has good location data.
Read this post where I explained some more about this:

Thank you for the additional info. I was just looking at the specs for the original mini and the mini 3 and saw the difference in the sat coverage. I use GPS frequently when hiking so I'm pretty familiar with how it works and where it has issues. I've occasionally had issues with GPS reception even when I had a good view of the sky

One last question - I had taken off and was just hovering, waiting for the GPS to lock in - I flew as high as I could without a lock (about 15'), with the hope that a bit higher would get better reception. Is that the best course of action? Or would it be better to leave it on the ground and wait until the sats lock in (the icon turns white and the home point gets updated) before taking off?
 
I've occasionally had issues with GPS reception even when I had a good view of the sky
If your drone has a clear, unobstructed view of half the sky or more, it shouldn't have any problem with GPS reception.
One last question - I had taken off and was just hovering, waiting for the GPS to lock in - I flew as high as I could without a lock (about 15'), with the hope that a bit higher would get better reception. Is that the best course of action? Or would it be better to leave it on the ground and wait until the sats lock in (the icon turns white and the home point gets updated) before taking off?
Your drone rated GPS data reliability as 3/5 until 5:05.5, 100+ feet up from the launch point.
It's inability to get good GPS down low was because obstructions were blocking the skyview and the sats in the blocked part of the sky.
Leaving the drone down low, it's unlikely that it would have ever achieved good GPS reception.

You need to understand that the drone won't get GPS reception in locations where obstructions block the sky view and to fly in those areas will be quite risky.
 
You need to understand that the drone won't get GPS reception in locations where obstructions block the sky view and to fly in those areas will be quite risky.

I completely understand that - I've lost one drone and crashed one for this very same reason. I'm trying to mitigate some of the risk by understanding it more fully and ways I can improve my odds/reduce risk. If I want to take videos of waterfalls, it is an inherent risk due to the geography - there isn't anything I can do about that other than decide not to fly, which kind of negates one of the biggest reasons I bought a drone.

Thanks again for the insightful responses. They have been helpful.
 
  • Like
Reactions: Ty Pilot
Don’t forget that the satellites are always moving, and the raw satellite count doesn’t 100 % of the time mean that you have a good fix. That is the goal of having more than 3 satellites for a 3 dimensional fix. Trees, canyon walls, etc further degrade the validity of a satisfactory fix.
 
Just for reference, across all my Mavic Mini flight I have 297 outdoor flights, all (UK).
The average maximum number of satellites is 17.33, the average "average" number of satellites is 14.36.
The maximum number of satellites recorded is 22.
3 flights had maximums less than 7 and 12 flights had maximums less than 10.
 
Just for reference, across all my Mavic Mini flight I have 297 outdoor flights, all (UK).
The average maximum number of satellites is 17.33, the average "average" number of satellites is 14.36.
The maximum number of satellites recorded is 22.
3 flights had maximums less than 7 and 12 flights had maximums less than 10.

Is that the original mini? or 2 or 3? Just curious how the different satellites might affect these counts.
 
Or learn to fly in full manual mode without gps.

Yeah, a year or more ago I did some practice trying to simulate this. I forgot exactly what I did - I think I taped some aluminum foil over the GPS receiver or something. I do remember flights being VERY different in that scenario - you couldn't take your hands off the sticks at all, but it was really good practice.

That is what threw me about this flight - it seemed like all was well - it was stable the whole way up the waterfall and then when it got to the top, it just did that weird corkscrew turn and went all wacky. It was good until it wasn't. :)

I hadn't flown in a while either which probably contributed to this. I need to make sure I fly more regularly so I don't get rusty. Just like life, owning a drone is a continual learning experience it seems....
 

DJI Drone Deals

New Threads

Forum statistics

Threads
130,412
Messages
1,552,352
Members
159,420
Latest member
shandilya