Freefall of a Mavic Pro

[Simmo]

What is a parameter mod ?

The parameters in the firmware can be altered, allowing different functions that DJI didnt intend you to use.
Such as running the motors when upside down.

 

[Porky]

Well, as someone with real experience, I can tell you that my Mavic fell from over 100 m altitude (where it was when the battery died) and hit the desert surface flat. That surface was volcanic cinder and dust from the Colorado River delta so it sort of cushioned the landing. I replaced the battery and it flew just fine but the camera had been wrecked.

As I said in my post, “I’m no expert”

 

[Mako79]

Test at your own risk..

Quick motor start and shutdown

g_config_rc_cfg_start_motor_verify_tim_ms: 10
g_config_rc_cfg_stop_motor_verify_tim_ms: 10
g_config_rc_cfg_emergency_stop_verify_time_ms: 10​

Motor restart while inverted

g_cfg_debug_motor_no_start_motor_check: 1
g_config_fdi_open_close_auto_stop_motor_check: 1​

 

[netfolks2000]

Test at your own risk..

Quick motor start and shutdown

g_config_rc_cfg_start_motor_verify_tim_ms: 10​

g_config_rc_cfg_stop_motor_verify_tim_ms: 10​

g_config_rc_cfg_emergency_stop_verify_time_ms: 10​

Motor restart while inverted

g_cfg_debug_motor_no_start_motor_check: 1​

g_config_fdi_open_close_auto_stop_motor_check: 1​

If it requires all that, then the mystery continues. How did the guys in the two "thrilling" videos manage to land their drones? They never mentioned anything like that, just flew their drones supposedly "as is", stopped motors in flight, fell hundreds of feet, restarted and managed to land safely. The whole point here was asking how that was possible, and it appears, without their own input, or someone else that has done it, we won't know. From what you folks have said here it should not be possible without that software change.

 

[Mako79]

If it requires all that, then the mystery continues. How did the guys in the two "thrilling" videos manage to land their drones? They never mentioned anything like that, just flew their drones supposedly "as is", stopped motors in flight, fell hundreds of feet, restarted and managed to land safely. The whole point here was asking how that was possible, and it appears, without their own input, or someone else that has done it, we won't know. From what you folks have said here it should not be possible without that software change.

Luck with both..
1st video, s/he was praying it would recover and luckily it did.
2nd video, the mavic was well balanced enough that it didn't invert immediately (as you can see from the first video that a small free-fall will will NOT flip the mavic over).

 

[Brunovic]

This is not lucky at all...this forum is not very friendly to discuss about it....just google for it....you will find many things you can and can't do with your Mavic Pro 1...super easy to do...

 

[MrMavic]

Why would anyone want to purposley put their expensive drone in a situation like that for excitement. Kind of like sticking your head in a alligator's mouth to get a cool pic.

 

[old man mavic]

Why would anyone want to purposley put their expensive drone in a situation like that for excitement. Kind like sticking your head in a alligator's mouth to get a cool pic.

exactly

 

[msinger]

Why would anyone want to purposley put their expensive drone in a situation like that for excitement

It's fantastic content for viral YouTube videos

 

[netfolks2000]

It does help us understand the complexity of the programming in our drones. It makes me angry when anyone tries to trick us with some stunt, however, showing something that is impossible or inadvisable, even harmful if tried. Those videos make me more conscious about how I move my sticks and avoiding EVER placing them in either of those two "shut off" positions when in flight. Since both of the producers of the two videos above were given invitations to come here and elaborate and didn't show up it gives me some doubts about their processes and intentions. Thank you all for your input.

 

[tom351]

If the Mavic is in free-fall- I don't see how it could even know that it is inverted (ignoring any aero-drag) . There may be something that I do not know or understand about it's sensors, but an accelerometer would show no force in any direction during a free fall- and assuming the Mavic uses accelerometer to sense the direction of gravity to determine AC orientation then the Mavic would not be able to tell that it is inverted (because of the free fall)- which might explain why they were able to restart while inverted without changing parameters (if the video is real)

Why an accelerometer shows zero force while in free-fall

 

[lolo780]

Gyro.

 

[sar104]

If the Mavic is in free-fall- I don't see how it could even know that it is inverted (ignoring any aero-drag) . There may be something that I do not know or understand about it's sensors, but an accelerometer would show no force in any direction during a free fall- and assuming the Mavic uses accelerometer to sense the direction of gravity to determine AC orientation then the Mavic would not be able to tell that it is inverted (because of the free fall)- which might explain why they were able to restart while inverted without changing parameters (if the video is real)

Why an accelerometer shows zero force while in free-fall

It's not in free-fall in the physics sense. Free fall is a state of zero measurable acceleration in the frame of reference of the object. In a gravitational field that means uniform acceleration at the rate of the local gravitational field - in the case of the earth, 9.81 m/s². However, as an object falls through the atmosphere its initial acceleration is decreased by the force of aerodynamic drag opposing its weight, and fairly rapidly asymptotes to zero at the terminal velocity. At that point the 3-axis accelerometer will measure an acceleration of one g upwards, which will define its orientation. Even in pure free fall, the 3-axis magnetometer (compass) is measuring an inclined magnetic field, so that also gives absolute vertical orientation data.

It's actually primarily the rate gyros that track orientation, provided that there is accelerometer and/or compass data to permit the sensor fusion scheme to correct for bias and drift in the gyros, which will exist in this case for reasons described above.

 

[netfolks2000]

If the Mavic is in free-fall- I don't see how it could even know that it is inverted (ignoring any aero-drag) . There may be something that I do not know or understand about it's sensors, but an accelerometer would show no force in any direction during a free fall- and assuming the Mavic uses accelerometer to sense the direction of gravity to determine AC orientation then the Mavic would not be able to tell that it is inverted (because of the free fall)- which might explain why they were able to restart while inverted without changing parameters (if the video is real)

Why an accelerometer shows zero force while in free-fall

Well, that is an eye-opening comment from someone who obviously knows a bit about the technology inside. Very good point. Thank you for commenting. That could explain the whole thing. My turning it on its side and upside down is nullified as evidence it won't start because it would have to be at zero-gravity in freefall for the accelerometer to not feel gravity, not prevent a start of the motors.

 

[sar104]

Well, that is an eye-opening comment from someone who obviously knows a bit about the technology inside. Very good point. Thank you for commenting. That could explain the whole thing. My turning it on its side and upside down is nullified as evidence it won't start because it would have to be at zero-gravity in freefall for the accelerometer to not feel gravity, not prevent a start of the motors.

No. You need to read post #33.

 

[tom351]

Thanks Netfolks2000, but SAR104 is correct, I was not considering the gyrometer- which would be able to determine orientation. I guess it might be possible that the aircraft programming over-rides the orientation requirement for motors to spin if the accelerometer senses that the aircraft is falling rapidly. But I am just speculating there.

 

[JanB]

Have you ever seen the semen of a dandelion falling? It tumbles all the way but in the end it is never upside down for just minimal time. I don't think that are gyros of the drone which will stabilize the free fall since the rotors are not blocked by cutting the motors - they are still rotating by the blowing air and that is what will stabilize.
I have done those free fall experiments a lot of times, themost important to do is -as it is said before- to pay attention that they are really started, that means you may have to try it several times till you see the rpm mounting.

 

[netfolks2000]

Thanks Netfolks2000, but SAR104 is correct, I was not considering the gyrometer- which would be able to determine orientation. I guess it might be possible that the aircraft programming over-rides the orientation requirement for motors to spin if the accelerometer senses that the aircraft is falling rapidly. But I am just speculating there.

I trust what SAR104 says, although I barely understand it. I've seen many of his analyses of drone crash data. My conclusion is that no one who values their drone should ever try a mid-air shutdown of motors. I assume those who did and didn't land safely after an attempted re-start were too embarrassed to share an unsuccessful story.

 

[JanB]

What tom351 is assuming that in free fall the gyrometer is able to start the motors for orientation? Nonsense.

 

[netfolks2000]

Have you ever seen the semen of a dandelion falling? It tumbles all the way but in the end it is never upside down for just minimal time. I don't think that are gyros of the drone which will stabilize the free fall since the rotors are not blocked by cutting the motors - they are still rotating by the blowing air and that is what will stabilize.
I have done those free fall experiments a lot of times, themost important to do is -as it is said before- to pay attention that they are really started, that means you may have to try it several times till you see the rpm mounting.

Excellent point. The blades will be turning, and pretty fast in the airflow as it goes down. Just their spinning would give stability to the aircraft. I just took my drone out and blew on a blade to confirm to myself that the blades, however, are going backward. So when the start command is issued that must be quite a strain on the motors to suddenly have to change direction while spinning. I really think you have a good point that illustrates how the drone could be stable. I wonder if it turned upside down a moment if the center of balance would get it back over before the blades started spinning. It may be that it never does turn sideways or upside down because the blades would start spinning backward as soon as the motors stopped.