Thanks Pat that was what I was guessing, that they would take the drone and say nothing was wrong. I am curious when you crashed on driveway did you look at the flight logs to see if that was represented?
uncommanded descent
- Thread starter tino.v
- Start date
I used original props but I got different performance from 2 separate installations. I just got my drone about 3 weeks ago. I was flying with the original on the drone props when I ordered another set from DJI. I decided to change them from the installed props to the new ones. All of a sudden I was getting a listing in the drone and it sounded much louder. I switched them back and the listing went away and they are much quieter. No matter what props I use when I full throttle across a field the drone climbs in altitude as much as 8 to 10 feet and then will drop about half the altitude to lets say 5 feet.
I still think the "listing" is a red herring, unrelated to your issue. If the drone wasn't drifting in the direction of the tilt (which should be very noticeable), then the tilt was simply the drone holding position in a light breeze. That's how it holds position, and in still air it isn't physically possible to tilt without moving in that direction.
Hi Robert,
Yes, I looked at the flight logs in AirData initially before sending back to DJI. And then after I got the drone back, I learned how to download the files from my phone and plot the data using CsvView as described in the post I linked to. The post includes a plot that shows the aircraft descending into the driveway without against the wishes of the throttle command.
Pat
And DJI thought that was not an issue? I am starting to think I made a bad decision buy buying this drone. I have not heard the same problems with other DJI drones, just the mini, some speculate that they hurried production to save their company once the new laws and regulations went into affect.
If one or more props were, shall we say, malformed, might that cause the drone to list just to keep it in the right position???
That would just cause the drone to spin those props faster until they were generating as much thrust as the other props. The thrust is still aligned with the motor shafts, so if the drone tilts, it will put some thrust in that direction.
Like a lot of forum speculation, that doesn't really add up.
There's been no regulatory changes that would have made any difference in the 6 months since DJI launched the Mini.
No-one knows how long it was in development prior to launching in time for Christmas buying.
Not directly on topic but I had an interesting experience today when a prop blade departed at altitude. The video footage shows the view cruising high above the trees, then a rapid serious upset & recover to a hover, then the dreaded sinking began. It did not respond much to stick inputs on the way down though it did seem to be trying; moving a little in all the commanded directions except for 'up', unfortunately. Landed successfully sideways 
50 feet up in a shag bark hickory.
Thinking about it, I'm surprised it didn't just fall out of the air with one blade gone but no, it automatically righted itself. That's pretty clever, and the third time I've seen it do that, the other two were caused by clipping invisible branches. There's a lot of deciduous trees on my place and no leaves at present.
I located it using Find My Drone, got it down using a 12' stick and a 40' ladder and took it back to base. The retaining screw was still in place, no trace of the blade. I replaced all the props with new Dji ones that coincidentally arrived today, recalibrated everything and flew a few tests. I'm very happy to say the test flights were perfect except for one gimbal jump which I hope isn't too chronic.
50 feet up in a shag bark hickory. Thinking about it, I'm surprised it didn't just fall out of the air with one blade gone but no, it automatically righted itself. That's pretty clever, and the third time I've seen it do that, the other two were caused by clipping invisible branches. There's a lot of deciduous trees on my place and no leaves at present.
I located it using Find My Drone, got it down using a 12' stick and a 40' ladder and took it back to base. The retaining screw was still in place, no trace of the blade. I replaced all the props with new Dji ones that coincidentally arrived today, recalibrated everything and flew a few tests. I'm very happy to say the test flights were perfect except for one gimbal jump which I hope isn't too chronic.
You got lucky I guess, but the question is why did the blade come off in mid flight? I know about those invisible branches, I'm in the northeast and there are no leaves on the trees now so those branches are hard to see. I always try to fly above the trees, unless I am in a big open field.Not directly on topic but I had an interesting experience today when a prop blade departed at altitude. The video footage shows the view cruising high above the trees, then a rapid serious upset & recover to a hover, then the dreaded sinking began. It did not respond much to stick inputs on the way down though it did seem to be trying; moving a little in all the commanded directions except for 'up', unfortunately. Landed successfully sideways
Thinking about it, I'm surprised it didn't just fall out of the air with one blade gone but no, it automatically righted itself. That's pretty clever, and the third time I've seen it do that, the other two were caused by clipping invisible branches. There's a lot of deciduous trees on my place and no leaves at present.
I located it using Find My Drone, got it down using a 12' stick and a 40' ladder and took it back to base. The retaining screw was still in place, no trace of the blade. I replaced all the props with new Dji ones that coincidentally arrived today, recalibrated everything and flew a few tests. I'm very happy to say the test flights were perfect except for one gimbal jump which I hope isn't too chronic.
Did the blade come ripped off clean or damaged ?
I've seen some DJI craft able to fly with half a blade but not too good, looks like the Mini wont fly at all.
I've seen some DJI craft able to fly with half a blade but not too good, looks like the Mini wont fly at all.
Blade was completely gone, no trace, just the screw remained. To be fair, after the blade separated the Mini was able to resume flying, slowly in an almost vertical downward direction. It had already self-recovered after the blade separated and didn't just plummet, instead going into a straight and level slow sink. I had almost no control after the upset, it did all that by itself. It was just unlucky that it was over a tall tree when the blade separated. So I give the MM full points for fault tolerance, if it had happened 10 seconds before or after it would have landed gently in a field. I'm not inclined to give the design of the prop attachment too many points though.
Full disclosure, they were not Dji props, they were third party. The MM's fault tolerance is the real story here. Dji props for me from now on.
Full disclosure, they were not Dji props, they were third party. The MM's fault tolerance is the real story here. Dji props for me from now on.
Yeah aftermarket props probably wouldnt help. Mind you DJI props can fail too but seemingly much less likely given it puts their reputation into question if one of their drones is downed. Aftermarket dont care so much.
If you do go aftermarket, try stick to the better known brands like Master Airscrew (MAS) as less likely to fail as they also have a reputation to uphold.
GerdS
Well-Known Member
I continued with testing yesterday after updating my MM to firmware .0500.
Behaviour is unchanged, it is still ascending by itself if flying full speed and later after stopping it will descend on its own for a part of the additional height accumulated before.
My next test was finding out if there is a specific speed limit and yes, I found it to be approx. 20km/h or about 5.5m/s.
As long as the MM flies slower than that (no matter of P or S mode), it will keep altitude in normal tolerance of maybe 1m.
But as soon as it flies faster it will do uncommanded ascending.
My theory is, that the altitude control loop within the flight controller does no longer rely on barometric data when flying very fast and then only relies on the accelerometers, maybe to avoid false barometric input caused by the air blowing onto the sensor.
When the MM will come to a stop the flight controller might recognise the false altitude and after a while it might try to correct it. If the pilot did same before already by stick input, the flight controller might not recognise that and then overcompensate, which looks like uncommanded descending to the pilot.
And if the pilot tries to correct by pushing the throttle stick forward, his command may add onto the internal descending command of the flight controller which will result to zero, so the impression is that the MM will not or only accept slowly and with delay an ascending command.
Behaviour is unchanged, it is still ascending by itself if flying full speed and later after stopping it will descend on its own for a part of the additional height accumulated before.
My next test was finding out if there is a specific speed limit and yes, I found it to be approx. 20km/h or about 5.5m/s.
As long as the MM flies slower than that (no matter of P or S mode), it will keep altitude in normal tolerance of maybe 1m.
But as soon as it flies faster it will do uncommanded ascending.
My theory is, that the altitude control loop within the flight controller does no longer rely on barometric data when flying very fast and then only relies on the accelerometers, maybe to avoid false barometric input caused by the air blowing onto the sensor.
When the MM will come to a stop the flight controller might recognise the false altitude and after a while it might try to correct it. If the pilot did same before already by stick input, the flight controller might not recognise that and then overcompensate, which looks like uncommanded descending to the pilot.
And if the pilot tries to correct by pushing the throttle stick forward, his command may add onto the internal descending command of the flight controller which will result to zero, so the impression is that the MM will not or only accept slowly and with delay an ascending command.
This was certainly not in my case. If you relook at my log, you will see that I was cruising for minutes in Cinematic/Tripod mode (+/- 4 m/sec) at a height of 7 m above the ground, not windy at all, not asking much of the power of the AC, when it suddenly loses altitude and almost hits the ground.
GerdS
Well-Known Member
- Joined
- May 27, 2017
- Messages
- 14,458
- Reactions
- 19,469
- Location
- Los Alamos, NM
- Site
- www.mavicpilots.com
GerdS
Well-Known Member
Up to now I always did my test flights at quite low altitude, few meters above ground, means within the range of the VPS system.
Today I repeated my test flight, but with a starting altitude of 20m which is outside of VPS range.
And there was no more uncommanded ascending, even when flying up to 46km/h in S mode! Altitude always stayed witrhin +/-1m.
So I guess this firmware bug has something to do with the VPS system...
Today I repeated my test flight, but with a starting altitude of 20m which is outside of VPS range.
And there was no more uncommanded ascending, even when flying up to 46km/h in S mode! Altitude always stayed witrhin +/-1m.
So I guess this firmware bug has something to do with the VPS system...
Great info. Was wondering why my mini would not keep altitude and ascend while flying. When going full throttle on low altitude (<5m) it climes without any input. Pretty scary if you need to keep altitude. Hope the bug is resolved. Did not noticed this in the previous fw. |Now on 05
Mine is still climbing 10 feet per leg after updating firmware and replacing the props with Dji. All my flight start heights for these tests have been 20' and higher. So no difference here.
Question, how high is the limit of the VPS?
Question, how high is the limit of the VPS?
Last edited:
Similar threads
DJI Drone Deals
1. Mini 2
2. Mini 3 Pro
3. Mini 4 Pro
4. Air 2s
5. Air 3
6. Avata 2
7. Mavic 3 Pro
8. Mavic 3 Classic
2. Mini 3 Pro
3. Mini 4 Pro
4. Air 2s
5. Air 3
6. Avata 2
7. Mavic 3 Pro
8. Mavic 3 Classic
New Threads
-
-
-
Drones play critical role in NYPD action at Hamilton building at Columbia U- Started by Droning on and on...
- Replies: 1
-
-
Forum statistics
