Please select Into the mobile phone version | Continue to access the computer ver.
Sudden 170m (560ft) "free fall" without an apparent reason
12Next >
2885 42 2020-1-25
Uploading and Loding Picture ...(0/1)
o(^-^)o
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

So for the weekend I went a bit off the grid and thought, that I could try out a bit higher altitude of 300m (~900ft) from the start point and eventually to ~1000m / 3300ft final altitude, but mini made it only halfway through the original plan...It suddenly stopped ascending and started to "free fall" at a constat 6m/s (~13mph).
During the fall I tried some inputs to figure out the problem, but it was obvious - vertical control was completely gone - I could still control the position, so I moved it a bit further from the cottage to make the fall a bit safer and then I heard it buzzing and eventually saw the mini falling down to the snow... I promptly took out the battery, brushed the snow off and put it all to dry off next to the fireplace...After uploading the flights logs I expected something obvious to put the blame on, but alas, there was nothing...
After about an half hour I tried to take off again - I did a brief trip around the cottage without any problem... (and put the drone to the dry warm environment again, just to be sure)

So I am just uploading logs here as a curiosity...
My guess for the cause would either be a Sport mode (almost every problem I have seen on the forum was related to the sport mode in one way or another - I usually don't use it though),
or a humid , windless slightly freezing -2C (~28F) weather... (condensing humidity maybe?)

App, FW - everything was up to date, calibrated all sensors before the take-off.
Baro and Sonar sensors were working just fine, position reading was rock solid, battery levels were stable, wifi signal was full whole flight.


Guess I was lucky I have decided to try something stupid during a calm weather and in an open field covered in a 1ft layer of snow.

https://app.airdata.com/flight/1dbc34c400f27c7b5cf2d5ea8a0f051d/GENERALNotifications&val=detailed

(I put the "free fall" in quotation marks, because it was not a real free fall, since the gyro and motors were still running and tilt for a position movement was working as well.Mini seems to have taken no damage at all.)

2020-1-25
Use props
Francoisd
Captain
Flight distance : 66352 ft
France
Offline

humid , windless slightly freezing -2C (~28F) weather... (condensing humidity maybe?)

I would think about a propeller freezing regarding to the ground temperature....
2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

Francoisd Posted at 1-25 07:51
humid , windless slightly freezing -2C (~28F) weather... (condensing humidity maybe?)

I would think about a propeller freezing regarding to the ground temperature....

I thought so as well, but the constant ascend stopped so suddenly (from 4 m/s ascend to 0m/s in 13s) and I cant really recall any noticable frost on the props... But I might be wrong.I did a 10 min flight 1-2 hours prior to this one and the conditions were pretty much the same (except altitude was only up to 40m-60m [130-200ft]) - no problems at all...

I guess my main issue is, that Mini seemed to completely calmly accept its state... It knew it was falling, but it didn't try to put more power to the motors, to compsensate for that - even if that would not be enough, then a warning about max power reached would be expected right? But it was not there...
It just seemed like "I am just going to fall now. Bye..." - The fall was so calm and stable, so I exptected minimal damage (there was none in the end)
2020-1-25
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Online

There have been a few cases of dropping off in cold temperatures, rated at 0 to 40c, it seems that IMU getting to cold beyond working temperature gets confused or stops working, nothing confirmed but it’s the most likely scenario, I have done some testing and have seen a rise in altitude as well as drop in altitude, mu guess is when summer comes along we will see a lot less of this.
It’s great your drone survived hopefully you get to fly again soon .
2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

hallmark007 Posted at 1-25 08:21
There have been a few cases of dropping off in cold temperatures, rated at 0 to 40c, it seems that IMU getting to cold beyond working temperature gets confused or stops working, nothing confirmed but it’s the most likely scenario, I have done some testing and have seen a rise in altitude as well as drop in altitude, mu guess is when summer comes along we will see a lot less of this.
It’s great your drone survived hopefully you get to fly again soon .
If that would be the case, then i wonder which one of those sensors gave out...
As I said, baro, sonar, gyro, acceleration values were all spot on... The flight lasted only 1.5minutes in whole :/

It descended as if i was the one intentionally telling it to deacend at full speed.
2020-1-25
Use props
virtual
Captain
Flight distance : 4897142 ft
Czechia
Offline

As similar issues have been reported it would be better to not push the limits, especially in S-mode. I hope there will be fix soon...
I believe that even if some overload protection is aplied the bird can not fall down suddenly.
2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

virtual Posted at 1-25 08:44
As similar issues have been reported it would be better to not push the limits, especially in S-mode. I hope there will be fix soon...
I believe that even if some overload protection is aplied the bird can not fall down suddenly.

Yeah, no way I am going to go over 60m height and use S mode anytime soon
2020-1-25
Use props
virtual
Captain
Flight distance : 4897142 ft
Czechia
Offline

hallmark007 Posted at 1-25 08:21
There have been a few cases of dropping off in cold temperatures, rated at 0 to 40c, it seems that IMU getting to cold beyond working temperature gets confused or stops working, nothing confirmed but it’s the most likely scenario, I have done some testing and have seen a rise in altitude as well as drop in altitude, mu guess is when summer comes along we will see a lot less of this.
It’s great your drone survived hopefully you get to fly again soon .

But there are many videos with Mini in snow or temperatues bellow 0°C and S-mode should warm the electronic a bit, on the other hand the airflow during fast ascent can push more the cold air through the cooling holes on the belly. Anyway it happens and it is good to be aware.
2020-1-25
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Online

virtual Posted at 1-25 09:22
But there are many videos with Mini in snow or temperatues bellow 0°C and S-mode should warm the electronic a bit, on the other hand the airflow during fast ascent can push more the cold air through the cooling holes on the belly. Anyway it happens and it is good to be aware.

This will give you an idea how temperamental drone can get in cold temperatures.

2020-1-25
Use props
Miixxa
lvl.3
Flight distance : 280682 ft
Finland
Offline

Welp, judging by the battery voltage changes and the conditions it probably just froze up. -2C at ground level can be several degrees lower at any 50m jump upwards so Mini might have faced -10C temps up there depending on the weather.
So I would say that either your props froze up and/or battery/esc came incapable of providing sufficient power. Descent rate seems to be similar as in other cases of Mini losing it's lift.
I've suffered a similar uncotrollable descend without any warnings or errors, so I'm pretty sure your props collected ice. Nothing to do with modes or sensors. Just look it up, there's several topics about it on the forum. Even the slightest humidity at zero C or lower can accumulate enough ice to bring MM down.

Why people keep pushing this starter level drone in a way off spec conditions is beyond me...(granted that I've done it also, but stopped after couple of close calls )
2020-1-25
Use props
DJI Stephen
DJI team
Offline

Hello and good day Gu5s. I am sorry for the trouble this has caused and thank you for sharing these information with us. The cold temperature might have affected the flight performance of your DJI Mavic Mini. The DJI Mavic Mini has an Operating Temperature Range of 0° to 40°C ( 32° to 104°F ). In addition, I will post an official DJI tutorial video on Flying in Hot and Cold Weather. Thank you and fly safe always.

2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

As I said, I know I was pushing it, but it's best to learn it the hard way yourself, luckily all ended well, so crossed fingers for any other silly daredevils and keep flying safe ;)
2020-1-25
Use props
Hamiltom
Second Officer
Flight distance : 82927 ft
Canada
Offline

Looking at the voltage graphs there is a dramatic drop at around 1m11s in both cells. There are other drops, but this is the only one that both cells are affected. This corresponds to the beginning of the rapid descent. I would take from this that the battery simply could not deliver enough current to gain further altitude. I wonder if it triggered a software response to protect the motors? This would make sense, but then why no warnings? Myself, I would keep an eye on that battery and ascend in short, controlled bursts with pauses in the future to give the battery recovery time. I also would accept the limitations that come with the design of the Mavic Mini and govern my flying accordingly.
2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

Hamiltom Posted at 1-25 10:03
Looking at the voltage graphs there is a dramatic drop at around 1m11s in both cells. There are other drops, but this is the only one that both cells are affected. This corresponds to the beginning of the rapid descent. I would take from this that the battery simply could not deliver enough current to gain further altitude. I wonder if it triggered a software response to protect the motors? This would make sense, but then why no warnings? Myself, I would keep an eye on that battery and ascend in short, controlled bursts with pauses in the future to give the battery recovery time. I also would accept the limitations that come with the design of the Mavic Mini and govern my flying accordingly.

Hah, I thought "how smart" when I put labels on the batteries, shame i don't remember which one was it...
2020-1-25
Use props
virtual
Captain
Flight distance : 4897142 ft
Czechia
Offline

hallmark007 Posted at 1-25 09:30
This will give you an idea how temperamental drone can get in cold temperatures.

https://youtu.be/KMTJVyBQLag

Mini did pretty well at 0°F, stayed under controll with all the warnings, I would expect that it'll be completely frozen for the 2nd take-off (if not at the very beginning).
2020-1-25
Use props
virtual
Captain
Flight distance : 4897142 ft
Czechia
Offline

Miixxa Posted at 1-25 09:33
Welp, judging by the battery voltage changes and the conditions it probably just froze up. -2C at ground level can be several degrees lower at any 50m jump upwards so Mini might have faced -10C temps up there depending on the weather.
So I would say that either your props froze up and/or battery/esc came incapable of providing sufficient power. Descent rate seems to be similar as in other cases of Mini losing it's lift.
I've suffered a similar uncotrollable descend without any warnings or errors, so I'm pretty sure your props collected ice. Nothing to do with modes or sensors. Just look it up, there's several topics about it on the forum. Even the slightest humidity at zero C or lower can accumulate enough ice to bring MM down.

"Why people keep pushing this starter level drone in a way off spec conditions"
I blame youtube with all those volcanos and freeriders shots!
2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

Hamiltom Posted at 1-25 10:03
Looking at the voltage graphs there is a dramatic drop at around 1m11s in both cells. There are other drops, but this is the only one that both cells are affected. This corresponds to the beginning of the rapid descent. I would take from this that the battery simply could not deliver enough current to gain further altitude. I wonder if it triggered a software response to protect the motors? This would make sense, but then why no warnings? Myself, I would keep an eye on that battery and ascend in short, controlled bursts with pauses in the future to give the battery recovery time. I also would accept the limitations that come with the design of the Mavic Mini and govern my flying accordingly.

Well looking a the log again, i probably wouldn't make that association, since that (arguably) "considerable" voltage drop happened ~14s after the start of the fall.
But sure, I labeled the batteries to figure out such malfuctions or irregularities in advanced, before it can cause any bigger losses.

(BTW all of the 3 batteries had around 3 full charges in total - since I got my Mini fly more combo, so I don't expect them to be worn out a bit)
The more I think about it, the more plausible the ice build up on the blades seem to be. I have flown in colder conditions without issues, but the air was definitely much drier.
2020-1-25
Use props
BobWinNV
Captain
Flight distance : 284209 ft
  • >>>
United States
Offline

Gu5s Posted at 1-25 10:10
Hah, I thought "how smart" when I put labels on the batteries, shame i don't remember which one was it...


You were using battery 1U5X9AUEXP00L5



2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

BobWinNV Posted at 1-25 12:47
You were using battery 1U5X9AUEXP00L5

hah true, it was in the flight logs, thanks Bob
2020-1-25
Use props
BobWinNV
Captain
Flight distance : 284209 ft
  • >>>
United States
Offline

Here is a plot of your logs for Altitude, Speed (XY), Pitch, Throttle commanded, Pitch commanded, Drone heading, Vertical Speed, and Cell voltage for both cells.



2020-1-25
Use props
Hamiltom
Second Officer
Flight distance : 82927 ft
Canada
Offline

It is interesting that the Sonar Altitude is way off from 44" in height to 107' (11.9 to 17.1s). That seems to correspond to a voltage peak. I should have wrote 1.011s as the beginning of the rapid descent.That was when the cells had the downwards spike. It appears the descent was controlled to a degree. Is there a way to read the pitch/roll, etc. from flight logs?   I think that video would not be reliable due to the 3 axis gimbal.
2020-1-25
Use props
BobWinNV
Captain
Flight distance : 284209 ft
  • >>>
United States
Offline

Flight Summary of your logs from FRAP

2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

BobWinNV Posted at 1-25 13:53
Flight Summary of your logs from FRAP

[view_image]

Interesting, that seems a bit like a battery overload or something? Can that happen? As I said, I usually don't use sport mode, so this was new for me... I should definitely be wary of this battery.

Btw yes Hamilton - the fall seemed pretty controlled and stable, i still had a position control over the drone, it just kept descending at a very constant pace.
2020-1-25
Use props
BobWinNV
Captain
Flight distance : 284209 ft
  • >>>
United States
Offline

Gu5s Posted at 1-25 16:31
Interesting, that seems a bit like a battery overload or something? Can that happen? As I said, I usually don't use sport mode, so this was new for me... I should definitely be wary of this battery.

Btw yes Hamilton - the fall seemed pretty controlled and stable, i still had a position control over the drone, it just kept descending at a very constant pace.

The AirData battery information implied that this was the first time that you used this battery if I remember correctly.
2020-1-25
Use props
Miixxa
lvl.3
Flight distance : 280682 ft
Finland
Offline

Gu5s Posted at 1-25 16:31
Interesting, that seems a bit like a battery overload or something? Can that happen? As I said, I usually don't use sport mode, so this was new for me... I should definitely be wary of this battery.

Btw yes Hamilton - the fall seemed pretty controlled and stable, i still had a position control over the drone, it just kept descending at a very constant pace.

Still ice buildup as far as I'm concerned.

Bob's graphs back this up with the force errors and self oscillating pitch during rapid descend...
2020-1-25
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

BobWinNV Posted at 1-25 17:36
The AirData battery information implied that this was the first time that you used this battery if I remember correctly.
Mmm, afaik I have flown all 3 of my batteries, whether this one has not yet made a full-charge cycle (25-30min flight time) yet - is certainly possible...

Btw to answer some other comments - this was a short flight in the session, mavic was just taken outside and still had a room temperature so in case the culprit was an ice built up on the props, then it probably only 1 took minute of rapid ascend to render the props useless.
2020-1-25
Use props
Sigmo
Second Officer
United States
Offline

This is interesting.

When having some odd problems of my own, indoors at normal room temperature, what seems to have happened was that I had a number of factors that made the mini struggle.

First, the elevation here is around 5150 feet above sea level.

Second, I had the prop protectors installed, which adds weight.

Third, I have had damaged props installed.

And fourth, the mysterious "drops" all occurred after the battery level went below 50% as indicated on the display.

Replacing my prop blades has helped a lot.  And damaged blades may be similar to iced-up blades in their lack of efficiency.  It also seems that at my elevation and with the prop protectors in place, the poor mini is working very hard, indeed at all times.  So it's flying near the limits of thrust just hovering in my office here.  But the one environmental factor that's favorable is that the temperature in my office is always around 72 degrees F.

What seems to happen to me under these conditions is that the battery percent indication is not very accurate.  It seems that the reported percent remaining is overly optimistic.  I would imagine that when DJI has written the code that interprets battery voltage and then estimates remaining life, they've done this with the drone flying in warm temperatures, near sea level, and without the prop guards in place.  And perhaps when we put the batteries under a much greater load (as I do at my elevation with the prop guards in place), that throws the battery level estimate off.

When I see 50% battery remaining displayed, it seems like I really have about 20% left (indoors at room temp, but with the high elevation and heavy prop guards installed).

Perhaps when people stress the mini with heavy loads, and particularly in cold conditions, the batteries can suddenly lose enough voltage that the ESCs can no longer supply full power to the motors, and we get drops in altitude.  Lots of things could contribute.  Flying in S mode.  Flying high in cold conditions such that frost does form on the prop blades.  Cold reducing battery performance, etc.

I think, when we're stressing these little drones heavily, we need to take the battery level indication with a huge grain of salt.  And not only won't you get as much more flight time, but it does seem that as the battery voltage drops down, we can see reduced available power even though the battery indication would make you think all was well.

I think we tend to believe that the drone will have full power available until the batteries indicate, perhaps, 10% or less.  But I'm beginning to think that's not the case.

If the whole system used a DC-DC upconverter to power everything, then it could theoretically provide full power to the motors right up until the battery was almost dead.  But is that how it works?  It's just as likely that they use the "raw" battery power to feed the ESCs.  And if that's the case, then the maximum power available to the motors would be dependent on the battery voltage.  And even with a DC-DC upconverter ahead of the prop motors, there may be a battery voltage at which the converter can no longer produce the intended motor drive voltage.

Beats me! But there does seem to be some evidence suggesting that as the battery drains, you can no longer expect full power.

The suddenness of the drops that people experience could be a combination of things.  Icing could happen very quickly as one reaches an elevation where the conditions are suddenly just perfect for the formation of the ice on the props.  So that could easily be all it was.  But it could also be a case of icing slowly building on the props such that the RPM demanded of them to maintain or keep increasing elevation went higher and higher, and then, as the batteries dropped lower in charge, the available power to maintain the high RPM was being lost at the same time as the icing was requiring more RPM.

I'm just tossing it out there that batteries discharging at an unexpected rate might be throwing off the percent life indication because that indication is made based on assumptions of sea level, warm air, perfect props, etc.

It would be neat to have a better understanding of how the code for all of this is actually written.
2020-1-26
Use props
virtual
Captain
Flight distance : 4897142 ft
Czechia
Offline

IMPOV there was gradual loss of vertical speed (begins at 43s time mark) on full up stick input, ending with Not enough power warning at 1m 03s time (the highest altitude reached). It looks like some kind of overload protection was aplied (for speed controllers or to prevent motor overheat?). It doesn't look like battery issue as the voltage level under heavy load droped slowly from 3,8 to 3,7 or 3,65 Volts, that shouldn't be a major problem keeping in mind the winter temperatures (BTW have You updated all 3 batteries?). It doesn't look like propeller problem either as it lost all the power in 15s, too fast for icing IMPOV as there are videos where Mini is flying in -17°C...
I think that Mini was simply pushed too hard in these conditions - too long at full power in S-mode.
2020-1-26
Use props
Sigmo
Second Officer
United States
Offline

There's certainly a lot going on here.  It makes it so hard to troubleshoot after the fact.

One thing to consider about icing though, is that the conditions to get icing are not just about temperature.  There's also the state of water droplets, humidity, etc., in the air at that time.  I'm no expert on this, but I do know that what often causes icing to surprise a pilot and even cause a crash is the presence of water droplets that are still liquid, but already below the freezing point.  When they encounter an object or surface, or shock, that can trigger sudden crystallization all at once.  So called "icing conditions".

So even though the temperature might not be as cold as when you've had no problems, if you do happen to fly into a zone where these supercooled water droplets exist, you could get sudden icing and be surprised.

So that's something to watch out for and consider.
2020-1-26
Use props
Miixxa
lvl.3
Flight distance : 280682 ft
Finland
Offline

virtual Posted at 1-26 03:19
IMPOV there was gradual loss of vertical speed (begins at 43s time mark) on full up stick input, ending with Not enough power warning at 1m 03s time (the highest altitude reached). It looks like some kind of overload protection was aplied (for speed controllers or to prevent motor overheat?). It doesn't look like battery issue as the voltage level under heavy load droped slowly from 3,8 to 3,7 or 3,65 Volts, that shouldn't be a major problem keeping in mind the winter temperatures (BTW have You updated all 3 batteries?). It doesn't look like propeller problem either as it lost all the power in 15s, too fast for icing IMPOV as there are videos where Mini is flying in -17°C...
I think that Mini was simply pushed too hard in these conditions - too long at full power in S-mode.
You actually described here exactly what icing does:
Gradually accumulates on the props until it loses lift, sufficient power or both. It was collecting ice all the way up there, the ice didn't just snap suddenly on the props.
That's why it gradually slows down until saying "that's all folks" and descents uncontrollably when the motors are incapable(software limited) of spinning the props fast enough to provide lift, probably cause of disturbed air flow and/or increased rotational mass.

Also temperature isn't the key here, it's literally the moisture in the air which tends to be higher in higher temps and vice versa. I'd say that a drone flying in -17C or lower temps will not collect ice, as the air is relatively dry at those temps...
2020-1-26
Use props
WrongWay Feldman
Second Officer
Flight distance : 19885 ft
United States
Offline

Wouldn't the apparent reason be flying something called Mini 33 hundred feet in freezing weather?
2020-1-26
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

WrongWay Feldman Posted at 1-26 15:46
Wouldn't the apparent reason be flying something called Mini 33 hundred feet in freezing weather?

It was meant to be a hyperbole... I had a pretty good Idea what went wrong and I knew I was doing something risky, the title was implying that the drone acted as if nothing was wrong during the fall... (Some members of forum, had to dig for a brief power warning and fluctuation of the Voltage) You would expect warnings, errors flashing left and right during that (right?) since it knew it knew it could not keep up the altitude from the sensors, but from the user perspective (looking at the dji fly app) it just seemed as if the mini "just gave up".
(BTW yesterday I saw a low temperature warning in one youtube video - I have never seen such warning in my flights)

Next day I did a brief flight at lower altitude (same weather conditioons) and that was a definite proof of what have happened - Upon returning, the edges of blades had accumulated quite a noticeable layer of ice (it didn't affect the control that much yet) and the blades had water droplets all over.
2020-1-26
Use props
Sigmo
Second Officer
United States
Offline

Gu5s Posted at 1-26 22:44
It was meant to be a hyperbole... I had a pretty good Idea what went wrong and I knew I was doing something risky, the title was implying that the drone acted as if nothing was wrong during the fall... (Some members of forum, had to dig for a brief power warning and fluctuation of the Voltage) You would expect warnings, errors flashing left and right during that (right?) since it knew it knew it could not keep up the altitude from the sensors, but from the user perspective (looking at the dji fly app) it just seemed as if the mini "just gave up".
(BTW yesterday I saw a low temperature warning in one youtube video - I have never seen such warning in my flights)

I had my several "drop" incidents, flying indoors, at 5150' elevation, with the prop guards in place, and with  damaged blades.  This combination of things may have put the poor mini into a situation that might be similar to what you got with the icing on the blades.  Basically, very high power required from the motors to be able to hold vertical position.

And what I noted was that I got NO warnings whatsoever.  No extra load detected, no low battery, nothing like that.  Instead, the Mini simply descended from where I was flying at about 1.5 meters above the floor down to within, perhaps, 15 CM (about a foot) above the floor.  And from then on, for the remainder of these flights, I could not raise the mini no matter how much upward command I gave with the left stick.  The Mini did not crash, and I could fly it back and forth, left and right, spin it CCW or CW, etc.  And I was able to fly like this for another six or seven minutes.  This sudden descent happened at about 50% indicated battery level.

Putting in a fresh battery, the Mini would do the same thing.  I could fly normally until about 50% power at which point, the Mini would descend to just above the floor for the remainder of that battery's charge.  About 6 or 7 minutes of "good flying" followed by the drop and about 6 minutes of "floor skimming".

I initially thought that this might have been caused by my having NOT properly updated the firmware in two of my three batteries when I had updated from 1.00.0300 to 1.00.0400.  So I thought it might be some sort of miscommunication between the batteries and the Mini that caused it to either "protect" the batteries too soon, or something like that.  But when I rolled things back, and tested with different combinations of firmware in batteries and the mini and remote control, etc., that hypothesis didn't pan out at all.

Instead, the conclusion I came to was that I had a nick in one of the blades which had suddenly propagated from that stress riser, turning into a split that then made that blade lose a lot of its former lift, and that was why this behavior appeared coincidentally with my update from 1.00.0300 to 1.00.0400.  Coincidences are something to be very skeptical of when troubleshooting.  But in this case, that's my best guess, because my tests of various battery firmwares in combination with various Mini firmwares didn't make any difference.  What DID make a difference was replacing all of my props.  With the old bad props, I got the problem.  With the new, presumably good props, I never saw the issue no matter the firmware combinations.

But what is interesting to me about your incident when compared to mine is that in both cases, we got no dead battery indication.  Yet the drones seemed to descend, but in a somewhat controlled way.  I got the feeling with mine that the only thing that kept it from touching the floor was "ground effect" holding it up.  It was as if the drone was protecting itself or the batteries OR the batteries were simply providing everything they had to give, but that was now not enough (between the loss of thrust and the battery level).  And as I say, that happened to me at about 50% indicated battery level.

So I do think that when we have something happen where the poor Mini is being forced to work extra hard, AND the battery level is insufficient to meet that power demand, we're of course going to see the Mini drop down.

I don't know if there was any protection happening in my case, or things just reached the point where that's simply all the poor Mini (or its battery) had to give.

But regardless, it's interesting, and for me, it's a lesson that I need to be extremely careful to have all of the conditions correct when flying the Mini (or any drone).  And with the Mini, at higher elevations, and certainly when flying with the prop guards mounted, things are very critical because I'm already flying at or beyond what the Mini was designed to do.  I've seen a warning from DJI that says you should not fly with the prop guards in place at elevations over 1500 Meters.  I'm right at or slightly above that level.  So I can't complain about any of this.

All I'm offering here is that in my situation, I'm able to test the Mini right at its design limits.  So perhaps I can do some tests for us all that might act as simulations of what people at lower elevations experience if they encounter something like this apparent icing you had.  And maybe some of my flight logs could be helpful to the community.  I have a situation where I could perform experiments in an indoor and controlled situation that might prove useful.

Plus, I'm bored because it's always so windy here that I can't fly outdoors much, so there's that.  ;)

I think others could do similar testing by, for example, installing their prop guards and perhaps attaching small weights to it to get their Mini to be right at the limits, too.

Flying indoors here, with the prop guards on, I get flights of only 12 to 13 minutes before the battery is dead.  So that gives an idea of how hard the Mini is working.

I'll bet with "iced" props, if you could maintain that in a constant way, you'd see something similar.  I guess someone could stick bits of tape strategically to their props to simulate icing in a controlled and constant way.

2020-1-27
Use props
Sigmo
Second Officer
United States
Offline

Miixxa Posted at 1-26 15:03
You actually described here exactly what icing does:
Gradually accumulates on the props until it loses lift, sufficient power or both. It was collecting ice all the way up there, the ice didn't just snap suddenly on the props.
That's why it gradually slows down until saying "that's all folks" and descents uncontrollably when the motors are incapable(software limited) of spinning the props fast enough to provide lift, probably cause of disturbed air flow and/or increased rotational mass.

That sounds right.  And that's all interesting!

Where I work, we have a water intake into a river.  When the conditions are just right, we get ice formation on the Johnson screens over the intakes.  This freezes off and plugs the water inlets, so we can no longer pump water from the river.

This phenomenon occurs only when the river temperature is "just right".  If the weather is very cold, it doesn't happen.  And if it's warmer, of course it doesn't.  But there's a sweet spot for this crystallization to happen, and it's when the river water is apparently right at the freezing point, but not actually frozen (it never happens when this area is frozen over, either).  The working hypothesis here is that the turbulence created as this supercooled water flows through the Johnson screens triggers or "nucleates" the crystal formation, and then things progress rapidly, freezing the intakes off.

For our Minis, I wonder if we can get ahold of some of that de-icing stuff they spray on aircraft wings at the airport when they are afraid of icing conditions?  Hose down our props with it.  ;)

Actually, I'm afraid to fly into any clouds of mist, ice crystals, etc., because of the potential of wetting the electronics in the Mini.  But I've seen some very beautiful videos people have made by flying into or at least close to clouds or fog banks.  I like the results, but it seems dangerous.
2020-1-27
Use props
Miixxa
lvl.3
Flight distance : 280682 ft
Finland
Offline

Gu5s Posted at 1-26 22:44
It was meant to be a hyperbole... I had a pretty good Idea what went wrong and I knew I was doing something risky, the title was implying that the drone acted as if nothing was wrong during the fall... (Some members of forum, had to dig for a brief power warning and fluctuation of the Voltage) You would expect warnings, errors flashing left and right during that (right?) since it knew it knew it could not keep up the altitude from the sensors, but from the user perspective (looking at the dji fly app) it just seemed as if the mini "just gave up".
(BTW yesterday I saw a low temperature warning in one youtube video - I have never seen such warning in my flights)


So your main issue is the fact that it didn't warn you in advance? I'm not trying to be be rude, just asking.
If that's the case you have to understand the technical limitations of the Mini. It has no way to know that itcs collecting ice, other than it needs to amp up the motors to provide the commanded lift. Yes it could have warned you(and maybe it did briefly, have you checked the recording from the app?) but I'm pretty sure the outcome would have been the same, as it couldn't have descended fast enough to prevent ice accumulation growing critical. Failing to notice the user of max power/motor errors might be due to the recent firmware update where they obviously tweaked the system to not show power warnings so hastily and limited the power output of the ESC. (Previous firmware was showing warnings even in light breezes).
Also those temp warnings can be triggered but at least in my case they only trigger in really cold, like in -10.
2020-1-27
Use props
Miixxa
lvl.3
Flight distance : 280682 ft
Finland
Offline

Miixxa Posted at 1-27 01:25
So your main issue is the fact that it didn't warn you in advance? I'm not trying to be be rude, just asking.
If that's the case you have to understand the technical limitations of the Mini. It has no way to know that itcs collecting ice, other than it needs to amp up the motors to provide the commanded lift. Yes it could have warned you(and maybe it did briefly, have you checked the recording from the app?) but I'm pretty sure the outcome would have been the same, as it couldn't have descended fast enough to prevent ice accumulation growing critical. Failing to notice the user of max power/motor errors might be due to the recent firmware update where they obviously tweaked the system to not show power warnings so hastily and limited the power output of the ESC. (Previous firmware was showing warnings even in light breezes).
Also those temp warnings can be triggered but at least in my case they only trigger in really cold, like in -10.
...damn character limit.

As I was saying, the only temperature sensor onboard the Mini is for battery temps, so if you slam a pocket warm battery in and start flying -> no temp warning. Also the batt temp probably continued to grow throughout the flight as the motors were continuosly drawing more power.

There are clear built in compromises in MM's construction by default cause of the size and there's no way around it. Some might think of them as design flaws but I consider them as they are, forced technical solutions to manufacture a 249g "as full blooded as possible" cameradrone...
2020-1-27
Use props
Miixxa
lvl.3
Flight distance : 280682 ft
Finland
Offline

Miixxa Posted at 1-27 01:25
So your main issue is the fact that it didn't warn you in advance? I'm not trying to be be rude, just asking.
If that's the case you have to understand the technical limitations of the Mini. It has no way to know that itcs collecting ice, other than it needs to amp up the motors to provide the commanded lift. Yes it could have warned you(and maybe it did briefly, have you checked the recording from the app?) but I'm pretty sure the outcome would have been the same, as it couldn't have descended fast enough to prevent ice accumulation growing critical. Failing to notice the user of max power/motor errors might be due to the recent firmware update where they obviously tweaked the system to not show power warnings so hastily and limited the power output of the ESC. (Previous firmware was showing warnings even in light breezes).
Also those temp warnings can be triggered but at least in my case they only trigger in really cold, like in -10.
...damn doubles...
2020-1-27
Use props
Gu5s
lvl.4
Flight distance : 284728 ft
Czechia
Offline

Miixxa Posted at 1-27 01:34
...damn doubles...

Hah, funny that the character limit is present only on mobile version...

Well since the mini knew it was descending, although the inputs were completely oposite, you would expect something like "Can't keep sufficient propulsion for unknown reasons, try to land safely immediately"

Or something like that. I know the outcome would be the same, since at that point was already too late, but you would at least have a better idea of what was happening since in the couple of first seconds i had no clue, whether the motors had stopped.
2020-1-27
Use props
virtual
Captain
Flight distance : 4897142 ft
Czechia
Offline

Miixxa Posted at 1-26 15:03
You actually described here exactly what icing does:
Gradually accumulates on the props until it loses lift, sufficient power or both. It was collecting ice all the way up there, the ice didn't just snap suddenly on the props.
That's why it gradually slows down until saying "that's all folks" and descents uncontrollably when the motors are incapable(software limited) of spinning the props fast enough to provide lift, probably cause of disturbed air flow and/or increased rotational mass.

But there was another flight in the same place and very similar altitude drops are reported in higher temperatures as well, so I still don't think that iceing was the main problem (although is clear that Mini is not overpowered at all and paper thin blades would be very sensitive to demage, dirt, ice or whatever).
My suspicion is that onboard electronic can not handle long ascend in S-mode (the highest power demand) or overeacts to rising temperatre in this situation.
BTW You say that ice didn't snap suddenly, but that's how it looks. If there was lack of lift I would expect "power warning" 10-15s earlier, as soon as Mini is not able to contine ascend at the same speed...
2020-1-27
Use props
WrongWay Feldman
Second Officer
Flight distance : 19885 ft
United States
Offline

What I find amusing is whenever there is a report of someone flying something called a MIni in a way it was not intended to fly these posts become an episode of Air Disasters.
2020-1-27
Use props
12Next >
Advanced
You need to log in before you can reply Login | Register now

Credit Rules