UPDATE:


I contacted DJI support, provided flight data, they examined it, and are providing a replacement Mini 2 as a warranty case.

I am grateful, I am relieved, and overall I view this as a positive experience (well, I will once I get my replacement Mini 2, but I assume that will be on the way shortly - the weather is too snowy and wet here to fly right now anyway).


So apparently, this was not "pilot eror" (which allows me to stop kicking myself), and the root cause was some issue that caused the drone to fail to climb from negative altitude to an altitude above (approximately) "0" even when I tried and tried to climb.  It was apparently not a battery issue, not due to cold weather, etc., and not pilot error.

Thank you to all who contributed to this thread and helped me work towards a resolution.  I look forward to flying again soon (once the replacement drone arrives and the weather clears).

Blue sky and tailwinds.

---

I lost my Mini 2 a couple days ago. I'm trying to understand what went wrong.

I've had my Mini 2 for about 4 months, have been flying it regularly. Other than flying it into a small tree on one of my very first flights (months ago), I've had good luck with it (that first collision was with a small tree, at low speed, it only fell a few feet onto tall grass, so a relatively benign crash landing, with only some minor prop damage which was easily fixed with new props -- so I learned to stay out of the trees...).  I have flown it successfully many times, in many different environments since (including on some cold days recently, and on some warmer but quite windy days -- times when I did get the High Wind warnings).  Never had any problems.  Until a few days ago, when the flight ended badly.

It was a cold winter day (around 32°F/0°C). We had received several inches of snow, but wind was fairly calm (I'd estimate surface winds below 5 mph).

Side note: I am a licensed airplane pilot, I own a small Cessna, and have been flying (airplanes) for many, many years -- I can estimate wind speeds fairly accurately, I think.

This time (as I always do), I started with a 100% full battery, which I had been keeping warm in my coat pocket before the flight.  I went up to a local park late in the day, before sunset but shortly before sunset.  I launched the drone from a flat, public park space that is on a high hill/small cliff, where the land slopes away, steeply down to an ocean bay.  So the launch point was at "0" feet elevation (AFAICT, the Mini uses altitudes that are relative to the elevation of the launch point).  I launched, went up to a fairly low altitude (maybe 50 feet above the launch point), then flew forward a ways, out over the "cliff" and then out over the bay.  The Mini was always within sight though just barely.  Because my launch point was up on the cliff, while over the bay, the altitude shown on the controller may have been roughly 50 feet, but altitude over the ground (and water) was more like 150-200 feet, which is what I'd expect.

Question: I assume that the Mini uses relative altitudes, calculated from, and relative to its launch point/home point, NOT "true" above-ground altitude, derived from GPS data.   Is that correct?  That has always been my assumption, but this is the first time (at least the first time recently) where I started from a "high" elevation, then flew the drone down to a lower altitude that was over much lower terrain.  I probably should have paid closer attention to the aititude readouts throughout the flight, but I was concentrating on other things, and I've always assumed the altitude displayed was always a number based on its start point, not a measurement of it's actual hight above the ground.


I flew around, dropped down to fairly low over the water (maybe 20-30 feet above the surface of the bay, where some sailboats were docked), shot a little video of the boats. Then when the battery was starting to get near 50%, I started bringing the Mini back to me.  I'm generally quite conservative with remaining battery time, and try to land the drone with at least 20-30% left (a habit I have developed from managing fuel reserves in actual airplanes with myself in them -- having skin in the game makes one REALLY aware of how much time you have left before the engine gets quiet...).  So to bring the drone home, I started climbing up towards the cliff top, and flying back towards the little park on the cliff-face where I waited.

Note: The whole time, I was in Cine Mode.  In fact, I put the controller in Cine Mode on Day One whehn I bought the drone, and have never switched to anything else (I am now reading and learning that Sport Mode may provide more "power" when needed - lesson learned).

The drone started gaining altitude (it needed to get up 100 feet or more to reach the cliff where I was standing, and clear some trees between the water and the cliff). And it was getting closer to me.  It was heading in the right direction and climbing up, all seemed good.  But as I watched it get closer and higher, it stopped climbing.  It was a few feet lower than where I was standing, and refused to go any higher.  It was still a ways out, so I flew it towards me, while it was maybe 10 or 15 feet too low.  It wasn't quite high enough to clear the tops of some trees between its present position and where I was, but it was close -- it probably need to go up about 15-20 more feet -- so I manoeuvered it around/through the trees, getting closer to me, but even with full "up" stick, it would not climb any higher.  I kept trying that, then I saw the battery levels were getting pretty low.  I hit RTH and it continued to come towards me, but at the same, just-a-bit-too-low altitude.  I gave up on RTH and started looking around for a place to set it down.  Unfortunately, there was no place reasonable to land it where it was or nearby: it was above a steep, wooded slope that I could not reach, I could try going back out and down to the area by the sailboat docks, but that was some distance away.  I wasn't sure I could get there.  Started getting low battery alarms.  I probably wasted some energy/battery flying back and forth, looking for a flat, accessible place to put it down, but there just wasn't anyplace that looked like it would work.  I gave up on that as the alarms continued and escalated, deciding to try one more time to see if I could get it back to where I was.  It refused to gain more altitude. I flew it towards me, between the taller trees, getting it closer and closer.  It was still about 10 or 15 feet below me, and in between there were some bushes that looked to be right at its altitude, and finally a chest-high chain-link fence along the cliff edge, but it was getting close. My hope was to get it close enough that it might just barely clear the bushes and fence, or failing that, at least it might get to a point where I could land it and work my way down to it. I flew it closer and closer to me.


Finally, with the battery alarm blaring, the Mini hit the bushes: about 25 or 30 feet horizontally from me, and about 20 feet down the slope - and tumbled.  I could see it -- just barely -- sticking out of the snow, at the base of some bushes.  While not exactly a pretty landing, it looked like a completely survivable crash.  It probably fell no more than 3 or 4 feet vertically, ending up in small brush sticking out of a snow drift -- all pretty soft stuff.  I could see the red light blinking.  Barely.


My first instinct, of course, was to hop the chain-link fence, and scramble down the snow-covered slope to get it.  But...that is a steep cliff.  The Mini may have just been 20 feet down the slope, but beyond that, the terrain drops VERY steeply, almost vertically, for another hundred feet or so.  All of it was covered in snow and ice and small brush.  And it was starting to get dark.  I took a hard look at the slope and the conditions.  It looked like a perfect place to slip and fall to one's death, or at least make the evening news.  It was getting colder, too, and more snow was forecast for the night.

If I had had rock climbing gear handy (a harness, good ropes, and someone to call 911 if I slipped), I could imagine attaching the rope to the chain-link fence I was standing behind, rapelling down the cliff just a bit, and bush-whacking down the snow-covered slope to fetch the thing.  It really wasn't THAT far away.  But I didn't have any of those things handy.  I watched the red light blinking as darkness fell.  It was oh-so-close.  But I concluded that scrambing down to try to recover the Mini could easily be a fatal mistake, and I didn't want to win the Darwin award for 2021.  I sighed and went home (trying to think of anyone I knew who had good rock-climbing equipment...).  That night it snowed again, the temperatures dropped.  It has been icy since then.  It has been 3 days now.  I have not gone back to recover it, deciding that it's still too dangerous to try and scramble down that cliff. The weather forecast says the termeratures will rise above freezing tomorrow, we may get some rain, all the snow should melt in the next day after that.  Once the snow and ice are gone, I will go take another look and see if it's recoverable without risking a deadly fall.  I'm guessing that 3 or four days left outside in freezing temperatures and wet weather is not going to do it any good.  This Mini may have to be written off as an expensive lesson.

I keep teling myself: it's better to lose this drone than to have the same thing happen to my airplane while I'm flying it, and that's certainly true.  Still, it irks me, and I'd like to know what exactly caused this.

I've flown this Mini in much colder termperatures.  I've flown it in much greater wind.  I've flown (a few days before) when it was both a lot colder and a lot windier, and had no trouble then.

I can clearly take away the lesson that I need to be more conservative, when it comes to a combination of cold + wind + distance.  But I'm still stumped about exactly why I could not get the drone to continue to gain altitude.  On the way back from flying down low, I needed just another 20 or 25 feet of altitude, and it just would not go any higher.  It had flown much higher than that, earlier on this flight.  But once I sent it down low to poke around the sailboats, when I tried to bring it back up to me, it would just not go that high.  I am at a loss as to why it wouldn't (I had just updated firmware an hour before the flight, which makes me wonder, but I'm trying not to go there).  It had enough battery to make it back up that high (at least for a while), I'm pretty confident it wasn't fighting much wind.  There was no visible moisture (so I assume not much risk of airframe icing).  It seems like it just refused to continue climbing when I kept pushing full "up stick".  


Looking for input.  Any thoughts on why it wouldn't go up those last 20 feet and clear the treetops/fence to reach safety?  Thanks for your input.

That's way too long for an account of your incident.
Too much to read.

But if you want help, post your flight data and I'll read that and see what story it tells.
Go to http://www.phantomhelp.com/LogViewer/Upload/
It has instructions to find the .txt file from your phone or tablet which you can upload to that site
Upload it and it gives you a summary of the flight data.
Post a link to that summary.

Or just post the .txt file

One other thing you may not be aware. REGARDLESS of where the switch is on the controller, it ALWAYS STARTS in normal mode, you then switch it to one of the other two while airborne.

Mobilehomer Posted at 1-1 14:21
One other thing you may not be aware. REGARDLESS of where the switch is on the controller, it ALWAYS STARTS in normal mode, you then switch it to one of the other two while airborne.

Well, that would be news to me. So, settings from the Mode switch on the controller are simply ignored, and the only inputs from the Mode switch that are honored are those made while the drone is in flight?

If so, then it's been in Normal mode since the first flight back in July....

(Looking for the "Gobsmacked" Emoticon...).

It HAS been in Normal mode since its first flight.

Labroides Posted at 1-1 14:10
That's way too long for an account of your incident.
Too much to read.

Here the link to the flight data on phantomhelp.com (I think):

https://www.phantomhelp.com/LogViewer/BDWM87486PQC6WPYSMDQ/#

This is the first time I've seen this kind of data, so I'm probably not interpreting everything correctly, but at first glance, there are a few items that jump out at me...

Looking at the values beginning at 17m 4.6s, it shows Flight Mode "Go Home", Message: "Low battery. Return to home promptly.; Critical low  battery. Aircraft returning to home  (Code: 30184).; Low battery.  Returning to home."  

I assume that this was when I hit Return to Home, because 1) I had been unable to get it to climb any higher for several minutes time, and 2) there were treetops between the drone and the safe landing area that were higher than the drone's altitude -- it needed to go a litle higher.  I figured using RTH it would climb higher (I don't recall the exact RTH altitude setting but am pretty sure it was 100 feet or higher; the drone altitude at that point shows as roughly 6 feet).  Not only did it not climb to a higher altitude, it also didn't move horizontally, showing 0 mph for nearly a minute.  

After 30 seconds had passed, and it was pretty clear to me that it was NOT "returning to home" (Flight Mode switches to "Go Home" from 17m 4.6s to 17m 38.7s, during this time it was not moving at all, not climbing), when I switched back to trying to hand-fly it back (note Flight Mode changes back from "Go Home" to P-GPS), it still did not climb, it didn't even move at all (0 mph until about 17m 54s), then it slowly moved but never got any higher than a few inches more -- all this time I was pushing "full up" on the stick. This is when it just felt "stuck" and was not responding.  Battery here was down to about 17%, but all it needed was to climb up about 15 feet higher and it was just a short hop to the landing point.  But it just seemed "unwilling" or unable to climb.

Insights welcomed.

Hello dear @Burstmode ... that is really a very exciting story from you, I read everything carefully.

I would go back and try to rescue the drone with a long pole with a net attached. It's a shame about
the poor little drone that may freeze and cry in the snow :-)

Yes, why it didn't rise in height is also a mystery to me.
I was always very afraid of such events, on my 8 or so flights since September 2021, but mostly on sunny days
with little wind. Nevertheless, it could happen every now and then as it does with you, that the drone just doesn't
do what you do on the RC controller.

I don't have that much money to just write off a Mini 2 Combo for around 580 euros and lose it.
I couldn't buy a new one right away.

I would recommend you, if you still know the place where the poor little Mini2 is, go back and
rescue her with some auxiliary equipment such as a long pole or abseil down and climb
down to the Mini2 to recover her.

I wish you good luck here from Germany in USA.

Pinarek Posted at 1-1 16:48
Hello dear @Burstmode ... that is really a very exciting story from you, I read everything carefully.

I would go back and try to rescue the drone with a long pole with a net attached. It's a shame about

Thank you @Pinarek.

The location of the crash is very frustrating. It is soooooo close, and I know exactly where it is. But it is on a very high, very steep slope -- a cliff really. To safely retrieve it, I would need a climbing harness, strong rope, and know how to attach things. I have never done rock climbing and have no experience rappelling.  The slope is currently covered with snow and ice.  Although the drone is only about 20 feet from the fence (I went back and measured it), it's not in a safe place unless you have good equipment and know how to use it.

I also thought about a long pole with a net attached.  The pole would need to be over 20 feet long, and that is not easy to find (I checked local swimming pool supply stores and amazon - lots of 12-foot poles but nothing quite long enough).  I've also tried to imagine other creative ways to reach it (throw a large net off the cliff and try to pull it up; maybe find someone with a BIG drone that could hook it and fly it back! Hire someone who is more desperate and less careful than me...).  I am open to suggestions.

The snow and ice will probably be gone in another day or two.  I will take another hard look at the location when the snow clears. But if I slip and fall off that cliff, if the fall doesn't kill me right away, my wife has told me she will.  So, I'm still thinking about how much my drone and my life is worth...

Thanks for the encouragement and Frohes neues Jahr!

You are correct heights are relative to the take off point, take off point is the zero of height.
Note you can replay the flight in the app WITH TELEMETRY, see the profile or me section on the 'front page' of the app.
I haven't looked at the csv yet but the information displayed on the Phantom help page seems to tell an entirely different store from the one you wrote.

You might like to replay the flight on the Phantom help page by SLOWLY dragging the green button back and forth across the the screen and watch both the map and the data that scrolls in the area beneath the green button. Then compare that to the replay of the flight in the app.
  
Have you posted the correct flight?

The phantom help page indicates that the drone was above the take off point from 14:48 onwards until 18:24  by about 7.5ft in places, I do not know if that was sufficient to clear the treesand was once again above the take off point for a short while slightly later.

Assuming the above is correct I am sorry to say I think you simply misinterpreted what you were seeing and flew the drone until its gas tank was empty.

EDIT     having produced some graphs from csvview I have to own up and say the struck out was wrong, apologises, it seems there is something odd going on with the drone.

Thanks for the wishes for the new year 2022, I wish you and your good drone that you and the
Mini2 get home safely and you and your wife stay healthy, Omicron Covid lurks everywhere.

I think there will be a possibility, how old are you? You must have young, strong friends,
I would still be able to do abseiling at the age of 68, I was once a tank driver in the
GDR People's Army in 1973 and learned a lot there.

Sure I'm older now, but "where there's a will there's a way" !!
and you just have to go through it and want to.

Burstmode Posted at 1-1 16:03
Here the link to the flight data on phantomhelp.com (I think):

https://www.phantomhelp.com/LogViewer/BDWM87486PQC6WPYSMDQ/#

Here's the sad tale that the data shows:

You started the flight with a battery that showed 89%.
You flew and hovered and a Low Battery warning came up at 16:27.8 (20%).
At that time the drone was at about launch height and 600 ft out.

You brought the drone back a little closer and then went further out to 720 ft.
A Low Battery RTH commenced at 17:04.6 (15%)

I can't see a reason, but the RTH stalled and left the drone hovering
At 17:38.7 (14%) you cancelled the RTH, left it hovering for 15 seconds and started to fly towards home.

At 18:06.2 you stopped flying and hovered for 4.5 sec before resuming the flight back.

At 18:17.4 you stopped flying forward and tried to climb, but the battery was down to 10% and the drone only climbed slowly to 5 ft higher than home.

At 18:27.5 (10%) a critical low battery warning shows and the drone started descending as it is programmed to.
The drone is still 366 ft from home, but you leave the right stick untouched and use precious time and battery fighting against the autodescent.

It's not until 18:58.8 (7%) that you try flying forwards again after hovering for almost a minute.
At 19:10.3 (6%) the drone crashes into something at approximately home height, but 259 ft from home and the show is over.

Just as in a real plane, it's not a good idea to run the fuel gauge down to almost empty before landing.
It's good practice to come in with enough remaining battery to allow for unexpected problems.
Leaving the drone hovering for almost a minute at a critical phase of the flight wasn't very helpful either.

Sean-bumble-bee Posted at 1-1 17:54
You are correct heights are relative to the take off point, take off point is the zero of height.
Note you can replay the flight in the app WITH TELEMETRY, see the profile or me section on the 'front page' of the app.
I haven't looked at the csv yet but the information displayed on the Phantom help page seems to tell an entirely different store from the one you wrote.

Yes, I am 100% sure that is the correct flight (it was the last flight -- pretty easy to know which one it is, and I've never flown in this location before; the map confirms it).  Yes, I have dragged the green button to review the position and compare the altitudes throughout the flight. Still is a mystery as to why it would not gain altitude.

I also carefully examined the altitude values listed.  The numbers shown would not be enough to clear the treetops between the drone's position and the safe landing area (which is what I could see) -- it would have needed to be a bit higher -- maybe another 20 feet, I was not far away and could see it easily (there's also a chain-link fence that's about 4 feet high right along the cliff edge which it would not have cleared, even if there were no trees). That's why I kept trying to gain altitude -- I could see it would hit the treetops and probably not even clear the fence.

The biggest mystery to me is why the drone would not climb any higher.  I was constantly on the "up" stick most of the time since I started returning the drone.  When I hit RTH it never got any higher, either, so it was not gaining altitude at all, either when flown manually or following the RTH command (it also did not move at all, forward, backward, sideways, when I hit RTH....it just hovered; you can see it sitting at 0 mph).  I kept trying to get it to climb but it would not, and it did not climb when in Go Home mode.

It went into "Go Home" flight mode at 17m 4.6s.  It never moved (sat at 0 mph and did not change altitude or position) until 17m 38.7s, when it went back to P-GPS flight mode.  It just sat hovering for over 30 seconds (in Go Home mode) and did not move until I took back the control and tried to get up higher -- but it did not gain any altitude, either when it was (supposedly) in RTH mode, or after I gave up on that and tried manually climbing up.  By the time it ended up in the bushes on the cliffside, I had been trying to get the thing to climb up higher for several minutes, but it could/could/did not.  The only thing explanation can think of is that the temperature was too cold (around 32°F/0°C) but I see reports from others saying they fly their drones at much, much colder temperatures and have no problems.

Beats me.

Hello there Burstmode. I am sorry to read and to know what happened to your DJI Mini 2 and thank you for reaching out. Since this unfortunate incident happened and you have lost the said DJI drone. I would recommend for you to contact our DJI Support Team at https://www.dji.com/support?site=brandsite&from=nav for further assistance. We have a team that will do there best to help you and give out the best resolution for the said issue. Again, I am sorry for your loss and thank you.

@Burstmode had you updated the Fly Safe database shortly before the flight, by any chance?

-N.O.T.

The battery temperature vary from 66F to 99F,  "BATTERY.temperature [F]" column 91? in the csv.

I hate to be that guy with hindsight. But maybe at some point I would have just backed off and landed it in the marina parking lot. Always have a plan B.

Another thing. Seattle is very damp with the usual onshore winds. Maybe the props were iced. Some can fly at lower temperatures with a much lower humidity.

No Original Thought Posted at 1-1 19:46
@Burstmode had you updated the Fly Safe database shortly before the flight, by any chance?

-N.O.T.

Yes, I did.  About an hour before the flight (as I was topping off the batteries in the smart charger).  Also did a firmware update, so everything was up-to-date.

Labroides Posted at 1-1 18:06
Here's the sad tale that the data shows:

You started the flight with a battery that showed 89%.

Thanks for the input, @Labroides.  

Battery was topped off and full, removed from the smart charger just a few minutes before I headed out the door (grabbing the freshly-topped-off batteries was literally the last thing I did before I headed out).  Drove less than 5 minutes to the launch point.  Launched the drone no more than 10 minutes from when I walked out the door.  So while the numbers show 89% at launch, I don't know how that battery could have had more juice in it.  Knowing it was cold outside, when I headed out from my home, I stuck two batteries (one of which I used for the last flight) in my coat pocket to keep them from getting chilled.

All of the hovering you see late in the flight data, was me trying to just climb that last 20 feet to clear the treetops. The RTH program appears to have been having the same trouble -- the drone just sat hovering, it didn't climb higher, it didn't attempt to move towards the RTH point, either (which was good, because it would have run into the trees).

(And yes, the "don't run out of fuel" analogy for airplanes is perfectly valid, but what happens when you get down to less than 30% of your fuel capacity is not...an airplane will climb perfectly well until the moment it runs out of gas...but it seems my Mini 2 just will not climb once the batteries reach 28% capacity....look at the numbers.. . ).

And yes, I understand that burning gas versus drawing down a battery, you get different behaviors towards the end, different technologies, I get that.  All I needed was another 20 feet of altitude and would have been fine.  But it seems in this case, once the battery got down to 29 or 28%, that drone was not going to gain any more altitude no matter how hard I asked for it or how long I tried (RTH produced the same non-results).  That's definitely not what I had expected.  Maybe that's considered A Feature, Not A Bug -- it's reasonable to ignore commands to climb higher at some point -- when the juice is running out.  It had never occurred to me though that it would stop accepting commands to climb higher as soon as the battery level hit....28%.  I would expect that behavior certainly at 5% or maybe 10%. That was a surprise.  Maybe it's all about how LIPO batteries discharge and this is how it's got to be, because of the battery technology, I don't know. This may be an intentional design decision. If it is, I wish I had known about it.

What I do know is that I was trying to climb, ever since the drone was down by the sailboats.  The RTH program appears to also have been trying to climb.  But if you look at the numbers from my flight data, it shows that once the battery was down to 28%, after that, it never climbed more than a few inches, despite spending a lot of its dwindling battery reserve trying.  THAT was a surprise.  If I had known that, I would have been more conservative.

The major lesson here, for me at least, is that I can not expect this drone to climb at all, once the battery level is below 30%.  It may fly fine in a level attitude for some while, and of course it can descend well, below 28%.  But if you need to get over a line of trees to safely land, those trees better not be more than a foot or two above your drone's altitude if you expect to keep it out of the trees.

I hope this helps others avoid a similar fate.  Below 28% battery, expect that your only options are moving horizontal or downward. Below 30%, there really is no option to go up, even a little.

GaryDoug Posted at 1-1 20:16
I hate to be that guy with hindsight. But maybe at some point I would have just backed off and landed it in the marina parking lot. Always have a plan B.

Another thing. Seattle is very damp with the usual onshore winds. Maybe the props were iced. Some can fly at lower temperatures with a much lower humidity.

I tried that -- briefly.  Problem was that in order to reach a flat area near the marina I had to get the drone quite a big further west (in between the cliff I was on, and the marina parking lot, the terrain is wooded and near vertical).  Getting to me, up on the cliff, was way closer. But that required getting over the trees in between, and although I thought "it's just 15 or 20 feet higher" that was 15 or 20 feet higher than it would go.  Unfortunately, the inability to gain altitude was not evident until the drone had left the marina far behind (it climbed well as it was leaving the marina and heading up the hillside ahead....it just stopped climbing 10 seconds too soon).  Of course, at the time, I did not understand what was happening.  I still don't understand completely (is this "Expected Behavior" or caused by some condition, eg it's cold outside), but it seems evident from the battery and altitude numbers that once the battery got down to 28%, it was a dead duck, and would never clear the trees).

Hindsight, as you say.  Now I know, after your battery gets under 30%, the only vertical option is going to be...downward.

It was -6C and a humidity of 66%, at time of flight, i don't see icing as a factor to be honest, not with such low humidity i would be surprised and the warning data doesn't show props struggling,

But there are 2 things you've exaggerated though, the temp and the battery, but thats that's by the by,

I have no input other than where the drone may be but you already know that, i cannot think of a reason as to why it wouldnt ascend other than your max height setting, but that can't be because you took off and went higher than the point where it struggled, i cannot see any flight restrictions in that area either, RTH alt was 157ft so thats not it either, but something did stop you going higher, we can see you trying to climb from 18:10.7 to 18:17.5 and there was hardly any ascent even at full throttle. I cannot see what that could be. True that prop icing may cause this but the warnings do not support that, we should see some motor errors or battery draw errors, not to mention the low humidity, needs high to have enough moisture in the air so i think we can rule out icing.  Home point was not updated during the flight at a lower level as this would cause a problem for the height but this too was not the case. Im lost, perhaps JJB* may have more insight as to why it wouldnt ascend, anyhoo, heres a nice image for you

Burstmode Posted at 1-1 21:11
Thanks for the input, @Labroides.  

Battery was topped off and full, removed from the smart charger just a few minutes before I headed out the door (grabbing the freshly-topped-off batteries was literally the last thing I did before I headed out).  Drove less than 5 minutes to the launch point.  Launched the drone no more than 10 minutes from when I walked out the door.  So while the numbers show 89% at launch, I don't know how that battery could have had more juice in it.  Knowing it was cold outside, when I headed out from my home, I stuck two batteries (one of which I used for the last flight) in my coat pocket to keep them from getting chilled.

All of the hovering you see late in the flight data, was me trying to just climb that last 20 feet to clear the treetops.
All that hovering was time and battery that wasn't being used to bring the drone closer.
When RTH was stalling, you were already in a lot of trouble, having ignored low battery warnings and flown further out.
When things became even worse, you should have just flown toward home.
The time and battery wasted out there would have brought the drone back.

what happens when you get down to less than 30% of your fuel capacity is not...an airplane will climb perfectly well until the moment it runs out of gas...but it seems my Mini 2 just will not climb once the batteries reach 28% capacity....look at the numbers.
The analogy is still good.

This may be an intentional design decision. If it is, I wish I had known about it.
Drones normally can climb right down to critical low voltage (and even after that if you push it).
Something abnormal happened to yours and the most likely explanation would be a loss of thrust due to prop icing.
We already know the temp was low.
Was the air moist too?
The tiny props of the Mini are particularly prone to losing lift with icing in cold moist conditions.
Last winter there were quite a few that went down because of icing.

The loss of your drone was due to a combination of factors working together to create a hopeless scenario.

The major lesson here, for me at least, is that I can not expect this drone to climb at all, once the battery level is below 30%.
The major lessons should be:
Be cautious and get your drone back before the voltage gets to 20%
Avoid flying the Mini in cold, moist conditions.

Burstmode Posted at 1-1 21:22
I tried that -- briefly.  Problem was that in order to reach a flat area near the marina I had to get the drone quite a big further west (in between the cliff I was on, and the marina parking lot, the terrain is wooded and near vertical).  Getting to me, up on the cliff, was way closer. But that required getting over the trees in between, and although I thought "it's just 15 or 20 feet higher" that was 15 or 20 feet higher than it would go.  Unfortunately, the inability to gain altitude was not evident until the drone had left the marina far behind (it climbed well as it was leaving the marina and heading up the hillside ahead....it just stopped climbing 10 seconds too soon).  Of course, at the time, I did not understand what was happening.  I still don't understand completely (is this "Expected Behavior" or caused by some condition, eg it's cold outside), but it seems evident from the battery and altitude numbers that once the battery got down to 28%, it was a dead duck, and would never clear the trees).

Hindsight, as you say.  Now I know, after your battery gets under 30%, the only vertical option is going to be...downward.

Now I know, after your battery gets under 30%, the only vertical option is going to be...downward.
That's nonsense.
If it was true, we'd all have experienced it and many would have lost their drones.
And yours did respond to throttle input after 30%.
At 18:18.4 (11%) you pushed the left stick up.
The drone climbed from 3 ft to 5.2 ft
It had probably shaken some ice off.
Unfortunately for you, it hit Critical Low Voltage shortly afterwards and it was all downhil from there.

Thanks for the nice image.  It just needs to be covered in snow and ice.  Forecast is for rain all day tomorrow, the snow should be gone soon.  Maybe I'll try cliff-diving on Monday (must check my life insurance first...).

I did not check the temperature locally that afternoon.  It was cold.  It may well have been below freezing, though didn't feel like -6C to me.  Humidity was fairly low (way below 66% I would guess, though probably a bit higher down by the marina than up on the cliff).  Airframe icing seems like a stretch, but I'm new to cold weather flying this thing.  Not a lot of wind.  Air temps down close to the water were probably a few degrees higher than up on the cliff.

I had left 3 batteries in the DJI "smart charger" all afternoon.  Charger lights showed all were full before I gathered up gear to go, and it was a short drive to the launch location.  I recognize that the technology (indicator lights on the charger, digital readings from the internals) are not precise and exact, so between both systems, maybe the 89% shown was the best it could approximate.  In any case I'm confident that battery was about as "full" as one could get it when I jumped in the car for the short drive to launch it.  FWIW I always head out with fully-charged batteries (I'm obsessive about that). Battery was purchased along with other gear end of July.  The Mini has made 77 flights since new (based on the number of Flight Record text files).  I have six batteries, and rotate between batteries 1-3 and 4-6.  Not a large number of battery cycles through its life. Everything was purchased new (AFAICT) from major retailers July 2021, nothing damaged AFAIK.

Something felt like it was preventing the drone from climbing after it almost got high enough.  It seemed to just stop climbing completely at around 14m 43s, at 28% battery.  If it had climbed just another 20 feet before it gave up climbing, as it was coming up from the marina (as I was trying to get it to do), I would have had plenty of time to bring it in safely.  A good portion of the battery (the last 28%, actually)  was spent trying (in vain) to climb a bit more, hovering, and looking for a low point to sneak in between the tallest trees and get it in closer (pretty much everything after 14m 43s).

I flew the drone a few times in the preceding days (it was equally cold then, probably colder), and had no problem other than freezing hands.

There might be an opportunity here to improve the warning language.  Instead of "Low battery. Return to home promptly.; Critical low battery. Aircraft returning to home (Code: 30184).; Low battery. Returning to home." (I was already painfully aware of all that, except it was NOT actually returning to home, it was just hovering), maybe something more like "It will NOT go any higher, quit trying to climb and just land it NOW wherever you can" might give someone a fighting chance.  [shrug]

Labroides Posted at 1-1 22:21
Now I know, after your battery gets under 30%, the only vertical option is going to be...downward.
That's nonsense.
If it was true, we'd all have experienced it and many would have lost their drones.

If it was true, we'd all have experienced it and many would have lost their drones.

I have no idea what anyone else's experiences have been.  I only know what happened to me.  It's pretty clear to me that the drone would not climb adequately after some point -- and that behavior was quite different from all my previous (though admittedly limited) experiences, where the drone was always responsive.  I'm just attempting to understand why.

It seems you are confident that iced-up props prevented it from climbing adequately.  OK.  I appreciate your perspective.

I used the 66% and temps from the local airport, there are a few local personal weather stations close by and they are fairly close to the airport give or take, allowing for the quality of both stations they are
somewhere close to the airport data, It's my thing, weather, i own a pro station (Davis VP2+)
One station at North Beach, whilst a cheaper quality station, temps are normally close for this make, humidity you have to allow for +/-10% and it looks like its closer to the minus  for this one
Low - High - Ave
Temperature-4.7 °C        -8.3 °C        -5.8 °C
Dew Point        -5.3 °C        -10.6 °C        -7.9 °C
Humidity        97 %        70 %        86 %

Another at hilltop, its actually a Davis VP2 so you can take those figures to the bank
Temperature  -3.8 °C        -7.6 °C        -4.8 °C
Dew Point         -5.6 °C        -10.6 °C        -8.1 °C
Humidity          89 %        66 %        78 %

You were flying between those 2 stations as seen on Weather Underground, hence i used the official data from the Airport, but you can see, the Davis station pretty much agreed.

Battery, what can i say, if the lights are still flashing whilst charging, they are not full, yours were far from full, you deffo had one battery that wasn't full. There are no discrepancies in the battery percentage, certainly not like you're thinking, once full, these batteries say 100% in the Fly app, there is no margin, now then, when only at 89%, there we have a margin because you're starting with a less than full battery, this reading may not be accurate and could be less than what its stating, hence why you must start a new session with a freshly, fully charged battery, they've been know to dropped in percentage quickly and not manage to return home, your situation isn't that mind, there's something that we are missing and it might well be that DJI will need to investigate the data for you, and i will recommend opening i a ticket and sending it off for repair, for 1 the issue at hand and 2, the drones been out in the wet, so you will also need a new battery under warranty.

Any cached video from that flight on your device? would like to see the conditions please...

Labroides, If it was an icing issue, wouldnt the motors record an error, be it in speed or battery draw? I have only known my props to ice when there is a higher humidity, high enough to cause at least fog patches and that's around 97%+  There has to be a lot of water vapour in the air for fog and lest say the humidity was 75% (middle of the road between stations and official) that's still not high enough, i could conceive that moisture can come off the water but the humidity that day doesn't support that and same with the warnings, I would like to see the video because ice build up is not uniform and would trow out the balance, it would be interesting to see if this is shown on the video, I take it you think Ice was a factor? If so, then i will not rule it out 100%, there could be extenuating circumstances, sun melt for example may cause moisture in the air at that locale, i ruled it out as it was -5/-6C but, as there is no weather station at that exact spot (see map below) then we don't exactly know what was going on, but i do know it was at least partly sunny at the station on the point (see map) so i would say it was sunny or partly sunny at the flight spot too. Just one more point re ice, if icing prevented lift then wouldnt it be safe to say that with no up stick it would slowly drop on its own? I don't see evidence of that, but then i spose the IMU would make the motors work harder to keep alt, mm, stupid thought lol, ok another thought, i posgted a video the other day re ice and that knocked the drone form the sky, totally different to this scenario, in the posters scenario, the battery dying and drone on decent that clipped a bush that made it fall, not the drone  just dropping from the sky. Icing would be a good reason, i just dont see data either weather or flightlog to agree atm, unless I'm missing something?

.

What a weird story….
Hope you can still recover the drone

But the explenations of data that I read (dont know how myself) cant be true  either?

16:27.8 (20%).
17:04.6 (15%)
17:38.7 (14%)
18:17.4 10%
18:27.5 (10%)
18:58.8 (7%)
19:10.3 (6%)

That is 2.5+ Hours on a thirth of the battery?

What was the max flight altitude set at in the app?
You can see it in the flight log with the Airdata UAV app (dji flight log viewer does not show that)

The max height limit was set at 400ft, it is in the Phantomhelp csv, column DP / 120 I think.

Burstmode Posted at 1-1 20:39
Yes, I did.  About an hour before the flight (as I was topping off the batteries in the smart charger).  Also did a firmware update, so everything was up-to-date.

Hmm, this is the third or fourth report of the Mini 2 suddenly failing to climb during flight when a Fly Safe DB update was completed shortly before the flight.

IIRC in all cases the aircraft was only able to fly up to 0m - in some cases the aircraft was erroneously reporting negative height when it was still above the take of height, but would then not climb to above 0m.

I'm afraid it doesn't help you in the situation you have, but the other users reported that after landing the aircraft performed perfectly on subsequent flights.

The difference that I can see here is that you updated about an hour before the flight. The others, I believe, had only updated a few minutes before.

Have a look on this forum for posts about the aircraft failing to climb... It'll be little comfort, I suspect, but it may point to this not being pilot error.

I assume the app was connected to the aircraft while you updated the database? (I don't think it's possible to update without an aircraft connection?

Sean-bumble-bee Posted at 1-2 03:02
The max height limit was set at 400ft, it is in the Phantomhelp csv, column DP / 120 I think.

Ok, did not know that

Just a gentle reminder that airframe (meaning props here) leads to a loss of lift - NOT a loss of ability to climb.

Depending on the severity of the icing this loss of lift means anything from:

- loss of ability to climb (mild icing)
- reduced ability to climb
- loss of ability to keep the aircraft airborne (severe icing)

It would be very unusual if the icing was exactly the right amount to mean that the aircraft could maintain its current height and yet not be able to climb very slowly (due to reduced lift generation).

I think prop icing is a red herring here.

And this is not a battery level thing, either. As others have https://youtu.be/IsJlooyTr2Q the aircraft will climb with the battery at almost zero. Obviously not advisable unless you are doing it to save your drone by avoiding imminent danger.

I strongly suspect the Fly Safe DB issue that others have reported. Have a read of the other reports and judge the similarities...

Bashy Posted at 1-2 01:03
I used the 66% and temps from the local airport, there are a few local personal weather stations close by and they are fairly close to the airport give or take, allowing for the quality of both stations they are
somewhere close to the airport data, It's my thing, weather, i own a pro station (Davis VP2+)
One station at North Beach, whilst a cheaper quality station, temps are normally close for this make, humidity you have to allow for +/-10% and it looks like its closer to the minus  for this one

Labroides, If it was an icing issue, wouldnt the motors record an error, be it in speed or battery draw?
When the icing is so bad that the drone can't hold altitude and just keeps sinking, it will give a motor speed error.
I take it you think Ice was a factor?
I can't think of any other explanation for the incident.
In many, many incidents, I've not seen anything that can cause the drone to fail to climb like that except icing props on a Mini.

Just one more point re ice, if icing prevented lift then wouldnt it be safe to say that with no up stick it would slowly drop on its own?
That would depend on how much the props had iced.
Too much and they won't give enough lift.
But with less, the props could have enough to hover, but not enough to climb.

And the data shows that the props had recovered some climbing ability just before the autoland initiated.

Here's two examples of the Mini 2 suddenly failing to climb mid flight after a Fly Safe DB update.

https://forum.dji.com/thread-253146-1-1.html
https://forum.dji.com/thread-253288-1-1.html

I'm searching for others...

Bashy Posted at 1-1 21:32
It was -6C and a humidity of 66%, at time of flight, i don't see icing as a factor to be honest, not with such low humidity i would be surprised and the warning data doesn't show props struggling,

But there are 2 things you've exaggerated though, the temp and the battery, but thats that's by the by,

Hi Bashy,

Thanks for mention my name  ;-)

Interesting case this is! Why did it not climb in SmartPower RTH ?? even better, why did it not climb with 26% battery level?

I have no clue, but the countdown from 10 to 0 after wich the SmartPower RTH starts...this countdown stopped at the number 3. Why?

In the log the GoHome mode is written, but was it really in a RTH mode?
Mini2 steady hovering....
And, message "Ascending and returning to home" written in the log at17m38 ; the moment RTH was cancelled....uh ??

From 15m13s with 26% battery and further UP steering but Mini2 did not climb.See the upper baro height line in my chart, pink coloured where UP input was applied but no joy.
So this not going up already was there before the SmartPower RTH was activated.

Icing on the blades? nah, hover performance was so normal, don`t think so.Mini2 is not known of it bad batt power performance, i can fly UP with 1% battery left....

About this flight, if you are away from HP only 220 meter with 26% battery level than juice enough to fly and land at HP, if it only did climb with RC up input....
4.3 % battery usage for 1 minute flying is normal value, for sure in cold temps.

cheers
JJB

I pulled out these pictures from the flight log @Burstmode uploaded to the phantom help site.
That doesn't look too bad place to retrieve it.

Hannibal_ Posted at 1-2 01:52
What a weird story….
Hope you can still recover the drone

its minutes not hours, 16 minutes 27 seconds 8 milliseconds
=================================================

I didn't realise there were other similar issues referencing the update else that would have been a direct hit, good catch...

That climb looks doable BUT NOT from the top, go up from below....

You flew near the warmer more humid air the propellers iced the Icing on the propellers would not allow it to gain the proper altitude, poofkaboom crash landing? Air is always warmer more dense over open water in colder ambient temperatures.


Do not attempt recovery without the proper PPE .

Labroides Posted at 1-2 03:43
Labroides, If it was an icing issue, wouldnt the motors record an error, be it in speed or battery draw?
When the icing is so bad that the drone can't hold altitude and just keeps sinking, it will give a motor speed error.
I take it you think Ice was a factor?

It is a tough call for a definitive answer, but as there's now talk of similar issues after the latest update could lean us in that direction perhaps?

Having produced the following from CsvView I have to own up and say the tail end of the conclusion that I drew in post #10  was wrong. It seems there is something odd happening with the drone, apologies.
I also see I have been gazumped with graphical stuff lol.
If possible I'd be inclined to add the DAT log for the flight to any correspondance sent to DJI.



,

Bashy Posted at 1-2 04:40
its minutes not hours, 16 minutes 27 seconds 8 milliseconds
=================================================

Had to laugh at anyone thinking those times were hours and minutes!
Imagine the mini 2 flying for hours on end. lol.