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

Lost control of my Mini....I think I know why, but am hoping for some confirmation.

douglas_in_philly

Well-Known Member
Joined
Dec 29, 2019
Messages
167
Reactions
217
Age
53
Location
Philadelphia
Hi,

I was playing tennis on an indoor court today, in a building that is much like a corrugated metal airplane hangar. I had no GPS signal, but assumed I'd still be able to control the Mini. I took off, flew up about twenty feet, then forward maybe 60, and then the controls just didn't respond right at all. I'm assuming that the lack of GPS, and the metal nature of the building (and all kinds of metal stuff inside), and also possibly wifi interference within the building (it's a building on the campus of the college where I work), led to this crazy behavior. I was very lucky that #1) none of my friends playing tennis got hit (they didn't even realize I was having a problem), and #2) the Mini "crashed" into a big bulky net (used for softball/baseball batting and pitching practice), and wasn't damaged in the least bit.

If anyone wants to take a look at my logs, and let me know if they see anything conclusive, or if simply based on my description, can come to a conclusion, I'd appreciate it.

I'm not sure if there's a better way to share the flight data, but I uploaded it to the two sites below, and here are the links to view it:



Here's a link to the brief video shot by the drone.


Thanks!

Doug
 
I was very lucky that #1) none of my friends playing tennis got hit (they didn't even realize I was having a problem), and #2) the Mini "crashed" into a big bulky net (used for softball/baseball batting and pitching practice), and wasn't damaged in the least bit.

Lucky also the guy at 1:10 didn't get hit as it went past !!

Did your stick inputs have an effect on the mini behaviour ?
It looked like it went into atti mode a lot in between brief GPS contact, but it looks like no satellites were recorded.

In atti mode the drone will go forward and not brake, you have to slow / stop with your left stick.
A bit of a balancing act and not easy for a new drone pilot.

I don't think you ever had any real control over it from take off, just failed to see this and probably had too much trust in the usual flight behaviour when full GPS is available outside.
 
  • Like
Reactions: Saigondroner
Hi,

I was playing tennis on an indoor court today, in a building that is much like a corrugated metal airplane hangar. I had no GPS signal, but assumed I'd still be able to control the Mini. I took off, flew up about twenty feet, then forward maybe 60, and then the controls just didn't respond right at all. I'm assuming that the lack of GPS, and the metal nature of the building (and all kinds of metal stuff inside), and also possibly wifi interference within the building (it's a building on the campus of the college where I work), led to this crazy behavior. I was very lucky that #1) none of my friends playing tennis got hit (they didn't even realize I was having a problem), and #2) the Mini "crashed" into a big bulky net (used for softball/baseball batting and pitching practice), and wasn't damaged in the least bit.

If anyone wants to take a look at my logs, and let me know if they see anything conclusive, or if simply based on my description, can come to a conclusion, I'd appreciate it.

I'm not sure if there's a better way to share the flight data, but I uploaded it to the two sites below, and here are the links to view it:



Here's a link to the brief video shot by the drone.


Thanks!

Doug
Hi,

It is my morning here, and have run off......but I took a quick look at your logs, and agree with your conclusion.

You never had any GPS lock, that is normal indoors. But it also did not ever report RC signal strength. You seemed to have taken off with no GPS, nor HP recorded.....and then lost all control of the MM. It is a very common thing to occur when indoors, especially in metallic structures.

There are members that are much more qualified to analyze further.

Happy Lunar New Year! :cool:
 
  • Like
Reactions: Gindra
@douglas_in_philly Sure looks like it was responding to me....Even with the ATTI switches

1580000201322.png
 
  • Like
Reactions: Saigondroner
Lucky also the guy at 1:10 didn't get hit as it went past !!

Did your stick inputs have an effect on the mini behaviour ?
It looked like it went into atti mode a lot in between brief GPS contact, but it looks like no satellites were recorded.

In atti mode the drone will go forward and not brake, you have to slow / stop with your left stick.
A bit of a balancing act and not easy for a new drone pilot.

I don't think you ever had any real control over it from take off, just failed to see this and probably had too much trust in the usual flight behaviour when full GPS is available outside.
That may very well be what was going on. I did NOT realize that in ATTI mode, I had to "brake" in a way other than how I do it in GPS mode. Whoops! I got lucky!
 
  • Like
Reactions: Capt KO and Gindra
That may very well be what was going on. I did NOT realize that in ATTI mode, I had to "brake" in a way other than how I do it in GPS mode. Whoops! I got lucky!

Yep, lucky no damage to the little mini for sure, that net worked well.
Do check props well for fine damage though, easy to miss fine chips etc.
 
  • Like
Reactions: Gindra
The entire flight was done with vision positioning on for 95% of the time ( pink time interval below ) so the craft should be able to hold position like when GPS is available. If there are insufficient high-contrast features like markings on the ground, vision positioning may not be able to work effectively but even so, the resulting drifting of the drone should still be manageable.

1580034935153.png

As far as I can see, the response of the drone to elevator input was normal. It has even applied brake automatically when the elevator stick is released.

1580044525232.png

The logged messages have not indicated any loss of control signals throughout the flight :

0.0 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
0.0 NO GPS. FLY WITH CAUTION
0.0 0.0 0.0 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
0.2 1.1 1.1 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
3.2 3.2 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
3.5 3.5 JAIRCRAFT IN ATTITUDE MODE. UNABLE TO HOVER. FLY WITH CAUTION (CODE: 30004)
3.5 ATTITUDE MODE. FLY WITH CAUTION
4.1 4.1 IN FLIGHT0
22.6 BGPS SIGNAL WEAK. ALTITUDE LIMIT SET. WAIT FOR GPS SIGNAL TO RETURN TO NORMAL TO CONTINUE ASCENDING!
27.5 53.5 BGPS SIGNAL WEAK. ALTITUDE LIMIT SET. WAIT FOR GPS SIGNAL TO RETURN TO NORMAL TO CONTINUE ASCENDING!E
54.6 56.6 BGPS SIGNAL WEAK. ALTITUDE LIMIT SET. WAIT FOR GPS SIGNAL TO RETURN TO NORMAL TO CONTINUE ASCENDING!
58.6 60.3 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
60.3 NO GPS. FLY WITH CAUTION
60.4 60.4 JAIRCRAFT IN ATTITUDE MODE. UNABLE TO HOVER. FLY WITH CAUTION (CODE: 30004)
60.4 ATTITUDE MODE. FLY WITH CAUTION
61.6 BGPS SIGNAL WEAK. ALTITUDE LIMIT SET. WAIT FOR GPS SIGNAL TO RETURN TO NORMAL TO CONTINUE ASCENDING!
61.6 61.6 IN FLIGHT0
64.6 72.3 JAIRCRAFT IN ATTITUDE MODE. UNABLE TO HOVER. FLY WITH CAUTION (CODE: 30004)
72.3 ATTITUDE MODE. FLY WITH CAUTION

As far as I can see there has not been any loss of control but the vision positioning system might have become ineffective at some points due to the lack of ground features. The resulting drift might have made you feel that the drone was behaving abnormally. You can try again in indoor enviroment with the downward camera covered by opaque tape and practice controlling the drone when neither GPS nor visual positioning is available.
 
Last edited:
  • Like
Reactions: Gindra
The entire flight was done with vision positioning on for 95% of the time ( pink time interval below ) so the craft should be able to hold position like when GPS is available. If there are insufficient high-contrast features like markings on the ground, vision positioning may not be able to work effectively but even so, the resulting drifting of the drone should still be manageable.

View attachment 92101

As far as I can see, the response of the drone to elevator input was normal. It has even applied brake automatically when the elevator stick is released.

View attachment 92105

The logged messages have not indicated any loss of control signals throughout the flight :

0.0 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
0.0 NO GPS. FLY WITH CAUTION
0.0 0.0 0.0 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
0.2 1.1 1.1 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
3.2 3.2 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
3.5 3.5 JAIRCRAFT IN ATTITUDE MODE. UNABLE TO HOVER. FLY WITH CAUTION (CODE: 30004)
3.5 ATTITUDE MODE. FLY WITH CAUTION
4.1 4.1 IN FLIGHT0
22.6 BGPS SIGNAL WEAK. ALTITUDE LIMIT SET. WAIT FOR GPS SIGNAL TO RETURN TO NORMAL TO CONTINUE ASCENDING!
27.5 53.5 BGPS SIGNAL WEAK. ALTITUDE LIMIT SET. WAIT FOR GPS SIGNAL TO RETURN TO NORMAL TO CONTINUE ASCENDING!E
54.6 56.6 BGPS SIGNAL WEAK. ALTITUDE LIMIT SET. WAIT FOR GPS SIGNAL TO RETURN TO NORMAL TO CONTINUE ASCENDING!
58.6 60.3 BGPS SIGNAL WEAK. HOVERING UNSTABLE. FLY WITH CAUTION (CODE: 30002)
60.3 NO GPS. FLY WITH CAUTION
60.4 60.4 JAIRCRAFT IN ATTITUDE MODE. UNABLE TO HOVER. FLY WITH CAUTION (CODE: 30004)
60.4 ATTITUDE MODE. FLY WITH CAUTION
61.6 BGPS SIGNAL WEAK. ALTITUDE LIMIT SET. WAIT FOR GPS SIGNAL TO RETURN TO NORMAL TO CONTINUE ASCENDING!
61.6 61.6 IN FLIGHT0
64.6 72.3 JAIRCRAFT IN ATTITUDE MODE. UNABLE TO HOVER. FLY WITH CAUTION (CODE: 30004)
72.3 ATTITUDE MODE. FLY WITH CAUTION

As far as I can see there has not been any loss of control but the vision positioning system might have become ineffective at some points due to the lack of ground features. The resulting drift might have made you feel that the drone was behaving abnormally. You can try again in indoor enviroment with the downward camera covered by opaque tape and practice controlling the drone when neither GPS nor visual positioning is available.
Thank you, very much, for the analysis! It was just a really weird flight. It had me thinking I was using the controls incorrectly--and maybe I was, but from the point at which the drone started going backwards, I felt that I had literally no control over it. Nothing I did would stop it from coming back toward me. I could slow it down from going backwards, but it would immediately begin again as soon as I released the joystick.

Either that or I'm crazy, which is a distinct possibility. I may give it another try soon, to see if I get a repeat performance.

Thank you to everyone for their assistance with this!
 
  • Like
Reactions: Gindra
..... from the point at which the drone started going backwards, I felt that I had literally no control over it. Nothing I did would stop it from coming back toward me. I could slow it down from going backwards, but it would immediately begin again as soon as I released the joystick.....

OK, I can see what you mean now.

A reason I can think of is that the IMU is not calibrated correctly making the drone believe that it is level when it actually is not.

With GPS / vision positioning unavailable, the drone can only rely on the IMU data to tell whether it is moving. An off-calibrated IMU will cause the drone to persistently drift in a particular direction. You can check it by covering up the downward camera ( NOT the infrared sensors ) with opaque tape and flying the drone indoor. Make sure that the prop guard is installed. The drone should drift right after take off and you will have to fight with it in the manner you have described.

An IMU calibration should fix the problem.
 
Last edited:
  • Like
Reactions: BudWalker
OK, I can see what you mean now.

A reason I can think of is that the IMU is not calibrated correctly making the drone believe that it is level when it actually is not.

With GPS / vision positioning unavailable, the drone can only rely on the IMU data to tell whether it is moving. An off-calibrated IMU will cause the drone to persistently drift in a particular direction. You can check it by covering up the downward camera ( NOT the infrared sensors ) with opaque tape and flying the drone indoor. Make sure that the prop guard is installed. The drone should drift right after take off and you will have to fight with it in the manner you have described.

An IMU calibration should fix the problem.
This behavior is probably caused by the FC using the Vision System to navigate. It's an open question, at least to me, how the FC fuses the VS and IMU data. @douglas_in_philly if you could retrieve the .DAT from the device running the Fly App maybe we could learn more. Look here

Retrieving a V3.DAT from the tablet

to see how to retrieve the .DAT. The .DAT will have a label ending in FLY038.DAT

The .DAT will have the VS data in the VIO record. Comparing the VIO data against the IMU data is likely to be a bit tedious. I'm hoping for a team approach to looking at this data.
 
  • Like
Reactions: Gindra
Thank you, very much, for the analysis! It was just a really weird flight. It had me thinking I was using the controls incorrectly--and maybe I was, but from the point at which the drone started going backwards, I felt that I had literally no control over it. Nothing I did would stop it from coming back toward me. I could slow it down from going backwards, but it would immediately begin again as soon as I released the joystick.

Either that or I'm crazy, which is a distinct possibility. I may give it another try soon, to see if I get a repeat performance.

Thank you to everyone for their assistance with this!

This event obviously developed quickly, and you may have misremembered what happened and what you did. Firstly, as noted, the aircraft never had any GPS position data, and it was relying entirely on VPS:

Status.png

As you can see, it was in P-GPS mode for most of the flight except for a couple of seconds at 60 seconds and then after 72 seconds.

Looking at the records for elevator, pitch and forward velocity for those parts of the flight shows the following. Note that the velocity data (green) vanish to zero during ATTI mode since the aircraft has no position data in ATTI:

Vel_forwards.png

The backward motion started around 56 seconds, and was commanded by full back elevator. It stopped at 65 seconds and again at 70 seconds when you centered the elevator, as expected in P-GPS mode.

When it switched to ATTI at 72 seconds it was reporting a backwards velocity of around 2.5 m/s, with full back elevator. The velocity data stop at that point but you did not release the elevator until after 75 seconds and only applied a very brief (1 second) forward elevator at 76 seconds to try to arrest the motion, then you returned to backward elevator. The pitch data follow the elevator commands appropriately, and so it's not at all surprising that it kept moving.
 
This event obviously developed quickly, and you may have misremembered what happened and what you did. Firstly, as noted, the aircraft never had any GPS position data, and it was relying entirely on VPS:

View attachment 92111

As you can see, it was in P-GPS mode for most of the flight except for a couple of seconds at 60 seconds and then after 72 seconds.

Looking at the records for elevator, pitch and forward velocity for those parts of the flight shows the following. Note that the velocity data (green) vanish to zero during ATTI mode since the aircraft has no position data in ATTI:

View attachment 92112

The backward motion started around 56 seconds, and was commanded by full back elevator. It stopped at 65 seconds and again at 70 seconds when you centered the elevator, as expected in P-GPS mode.

When it switched to ATTI at 72 seconds it was reporting a backwards velocity of around 2.5 m/s, with full back elevator. The velocity data stop at that point but you did not release the elevator until after 75 seconds and only applied a very brief (1 second) forward elevator at 76 seconds to try to arrest the motion, then you returned to backward elevator. The pitch data follow the elevator commands appropriately, and so it's not at all surprising that it kept moving.
I agree. Probably not caused by a confused Vision System. That being the case no need to get the .DAT.
 
  • Like
Reactions: Prismatic

DJI Drone Deals

New Threads

Members online

Forum statistics

Threads
130,599
Messages
1,554,238
Members
159,603
Latest member
refrigasketscanada