Labroides Posted at 1-29 19:10
P4s and M2s are just as susceptible to Yaw Errors.

I never had a yaw error but I joined DJI well after everything was matured.  They probably corrected it with secret revision updates.

I’m kind of amazed here, for 3 posts it was being said that somehow dji was culpable because it allowed drone to take off and it should not allow drone to take off unless it had enough satellites.
Well actually what we seen was the new added feature in last FW working correctly exactly as it should and completely safe.

How it works is simple , you cannot take off unless you have enough gps or it’s sufficiently bright enough to be in opti mode, which is exactly what the OP did , in opti mode you are completely safe and your craft will not climb above 16ft unless it has gps lock, again this was exactly what happened here, and so neither dji nor gps were responsible for this crash.

It seems that this drone is a beginners drone, but in reality it’s just a drone that beginners decide for whatever reason they want to own and fly, all dji drones from phantom to mini are beginner drones, they all have beginner modes on them, in fact the drone that attracted by far the biggest interest for beginners was the Mavic pro, this was also said to be a beginners drone , flying the mini or a Mavic 2 is basically the same thing except the mavic2 is probably safer and equipped with more redundancy and the simple reason why is it costs 3 times as much, Mavic mini is a beginners drone simply because of its price nothing else and anyone who can fly it can fly any Mavic .

hallmark007 Posted at 1-29 19:18
I’m kind of amazed here, for 3 posts it was being said that somehow dji was culpable because it allowed drone to take off and it should not allow drone to take off unless it had enough satellites.
Well actually what we seen was the new added feature in last FW working correctly exactly as it should and completely safe.

Someone put a log where the attitude indicator went 90° left and right 4 times in less than 1 second.  If that isn't enough to ground the mini then idk what is.

They are probably the same supplier that supplied for the Boeing MCAS.  That one that supplied to DJI for the Mavic Mini if not the same flawed methods were used.  Unless you are an electrical engineer that knows what is wrong, we will never ever know.

Labroides Posted at 1-29 18:57
Your error is that you are confusing a simple warning of magnetic interference with a Yaw Error.

Bothersome would be two other things:

"The trouble is that you can put your drone down in a small magnetic  field, one that isn't strong enough for the flight controller to  identify as being different from the earth's magnetic field, but is  enough to cause the drone's IMU to initialise incorrectly."

Disagree, based on my own repeated testing with Mavic Pro Platinum involving different amounts of steel, and drone placed at different distances and angles to steel.  With some of testing posted here (DJI Forum).



"A simple check of the drone's orientation in the directional indicator would identify this problem before launching."

That would be Pre-liftoff check.  Not a a post-liftoff - pre-flying of mission check.


Since OP said drone takeoff location was dirt path upstream of bridge, what do you think was source of magnetic field?

AntDX316 Posted at 1-29 19:33
Someone put a log where the attitude indicator went 90° left and right 4 times in less than 1 second.  If that isn't enough to ground the mini then idk what is.

They are probably the same supplier that supplied for the Boeing MCAS.  That one that supplied to DJI for the Mavic Mini if not the same flawed methods were used.  Unless you are an electrical engineer that knows what is wrong, we will never ever know.

Thought from log, that was after Mavic Mini did it's fly-away, and had crashed into bridge's support cables.  

Not really amazed a countering possibility as to cause of flyaway and crash, brought out a person who has compensation interests involving drones.

HedgeTrimmer Posted at 1-29 19:45
"The trouble is that you can put your drone down in a small magnetic  field, one that isn't strong enough for the flight controller to  identify as being different from the earth's magnetic field, but is  enough to cause the drone's IMU to initialise incorrectly."

Disagree, based on my own repeated testing with Mavic Pro Platinum involving different amounts of steel, and drone placed at different distances and angles to steel.  With some of testing posted here (DJI Forum).

Disagree all you like but you're far out of your depth., as you showed in post #12 already.
Correction ... that's post #13

Dirty Bird Posted at 1-29 20:10
A piece of ferrous metal may, or may not, be magnetized.  If it's not magnetized, then you could lift off from it at will with no effect.  If it is magnetized, unless it happens to coincidently be aligned with Earth's magnetic field, then you have problems.  Any ferrous material or structure beneath the surface can potentially could lead to issues.  

Want to test your pilot skills?  Sit you drone on a large speaker for a few minutes then take it up for a spin.  I did this by mistake once with my Vision+.  I had to degauss the compass to get it to calibrate properly.

I don't think a magnetic field would affect the compass.  The Skydio 2s battery is attached by super strong magnets that are built into the aircraft.  There is no clip at all, not even a slot for the battery.  It's held on by a magnet and nothing else.

AntDX316 Posted at 1-29 22:42
I don't think a magnetic field would affect the compass.  The Skydio 2s battery is attached by super strong magnets that are built into the aircraft.  There is no clip at all, not even a slot for the battery.  It's held on by a magnet and nothing else.

I don't think a magnetic field would affect the compass.  
Of course you don't.
That's because you don't know the first thing about a compass and you're an idiot.

The Skydio 2s battery is attached by super strong magnets that are built into the aircraft.  
And ??
One of the features of the that drone is that it has no compass.

Hi Ewa,

Just a little question, at wich heading was your drone at the ground before take off?
In the direction of the red line?
Am i right to think that during the climb to 30 meters you drone was aligned with the bridge?
As your said  "I started a slow pitch to the right, to look along the bridge"

During the climb in the flightlog the heading of your MM was a steady value of 142, that is not looking to the bridge at all.
So if you state that you watched the bridge during that climb, than it is for sure a compass error.
So wrong compass heading to the actual flying heading, thus conflicting values to the IMU, thus fly-away.

See the chart, if this heading was the actual heading and the MM moves away that way  than the correcting action was oke, fwd and right roll to maintain stabilized hover.
But the craft heading was not 161 but aligned to the bridge, so fwd and roll will move the MM away and more away from its position.

cheers
JJB

Labroides Posted at 1-29 23:14
I don't think a magnetic field would affect the compass.  
Of course you don't.
That's because you don't know the first thing about a compass and you're an idiot.

You are talking about a "normal" compass.  We should assume it uses a different compass type on the Skydio 2 at least.  A lot of people stick to what they were taught but new methods supersede the old methods completely.  At least be open to change.  It's only a matter of time before new methods super obsolete old methods to the point of throwing people into bankruptcy or worse for not being able to keep up.  People think the problem is at the very end but it was actually a build up of unseen issues that were never addressed.  Do you want a demo of the battery?

AntDX316 Posted at 1-29 23:37
You are talking about a "normal" compass.  We should assume it uses a different compass type on the Skydio 2 at least.  A lot of people stick to what they were taught but new methods supersede the old methods completely.  At least be open to change.  It's only a matter of time before new methods super obsolete old methods to the point of throwing people into bankruptcy or worse for not being able to keep up.  People think the problem is at the very end but it was actually a build up of unseen issues that were never addressed.  Do you want a demo of the battery?

You are talking about a "normal" compass.  We should assume it uses a different compass type on the Skydio 2 at least.  
You must mean one of those new-fangled compasses that don't respond to the earth's magnetic field?

A lot of people stick to what they were taught but new methods supersede the old methods completely.  
New methods, like idiots know more than people that know stuff?
Like knowledge and experience are over-rated and unnecessary?

It's only a matter of time before new methods super obsolete old methods to the point of throwing people into bankruptcy or worse for not being able to keep up.  People think the problem is at the very end but it was actually a build up of unseen issues that were never addressed.  Do you want a demo of the battery?
Did I already say that you are an idiot?

Labroides Posted at 1-29 23:46
You are talking about a "normal" compass.  We should assume it uses a different compass type on the Skydio 2 at least.  
You must mean one of those new-fangled compasses that don't respond to the earth's magnetic field?

I've updated the post to have a video.

The Saint Posted at 1-28 15:18
^if the bridge beams caused this drone to flyaway and crash....I'm speechless.

speechless too... How can this toy be called a drone? This thing can never be trusted, it is obstructed by everything ..... temperature, wind, water, buildings, bridges anything ... I wonder how people ignore such events and lie to themselves saying that everything is ok with this bird

Donsir_LTU Posted at 1-30 00:08
speechless too... How can this toy be called a drone? This thing can never be trusted, it is obstructed by everything ..... temperature, wind, water, buildings, bridges anything ... I wonder how people ignore such events and lie to themselves saying that everything is ok with this bird

I think you’ve just included all drones there ;+):::

AntDX316 Posted at 1-29 19:33
Someone put a log where the attitude indicator went 90° left and right 4 times in less than 1 second.  If that isn't enough to ground the mini then idk what is.

They are probably the same supplier that supplied for the Boeing MCAS.  That one that supplied to DJI for the Mavic Mini if not the same flawed methods were used.  Unless you are an electrical engineer that knows what is wrong, we will never ever know.

I can see from most of the thread, that you are just being awkward or it’s all tongue and cheek.

However I was discussing the simple fact that this was in no way due to a problem with gps as one poster was trying without any success to convince OP it was something dji was culpable of allowing to happen to his drone.

Djis new failsafe came into play perfectly in this flight and as it turned out is a great help and lesson to all who fly mini.
Craft will not take off unless it has enough sats Or light is sufficient to operate in opti mode, in opti mode craft is limited to 16ft unless it gains gps lock, this worked like a dream on this flight and for others flying mini they can now see how it works.

To many putting out completely wrong information about craft going into atti mode etc, not helping anyone .

hallmark007 Posted at 1-30 03:39
I can see from most of the thread, that you are just being awkward or it’s all tongue and cheek.

However I was discussing the simple fact that this was in no way due to a problem with gps as one poster was trying without any success to convince OP it was something dji was culpable of allowing to happen to his drone.

You can go about how it "should" perform but if it's defective inside there is nothing anyone can do.  It's like the MCAS failure.  Not all of them fail.

AntDX316 Posted at 1-30 03:58
You can go about how it "should" perform but if it's defective inside there is nothing anyone can do.  It's like the MCAS failure.  Not all of them fail.

You see you weren’t around when they launched spark , just as many problems with new users, or Mavic pro twice as many problems again with new users, it will always be the same with drones that new users flock to, I think if you look around the forum, very few problems for users that fly other drones .

hallmark007 Posted at 1-30 04:35
You see you weren’t around when they launched spark , just as many problems with new users, or Mavic pro twice as many problems again with new users, it will always be the same with drones that new users flock to, I think if you look around the forum, very few problems for users that fly other drones .

Yeah, I had a late 2018 model Spark.  Secret hardware revisions help a lot.

Labroides Posted at 1-29 21:26
Disagree all you like but you're far out of your depth., as you showed in post #12 already.

"Disagree all you like but you're far out of your depth., as you showed in post #12 already."


hm7 much?

LordyLordy#12

Dirty Bird Posted at 1-29 20:10
A piece of ferrous metal may, or may not, be magnetized.  If it's not magnetized, then you could lift off from it at will with no effect.  If it is magnetized, unless it happens to coincidently be aligned with Earth's magnetic field, then you have problems.  Any ferrous material or structure beneath the surface can potentially could lead to issues.  

Want to test your pilot skills?  Sit you drone on a large speaker for a few minutes then take it up for a spin.  I did this by mistake once with my Vision+.  I had to degauss the compass to get it to calibrate properly.

Thanks, but I know better than to place my MPp on top of my speakers.

Something to watch out for.  Even if a piece of metal is not magnetized, it can act as a shield to Earth's magnetic field.  Seen this when I was testing magnetic field Apps and MPp near large metal barn.  Gauss level would dip from open area.

JJB* Posted at 1-29 23:35
Hi Ewa,

Just a little question, at wich heading was your drone at the ground before take off?

Interesting analysis JJB, leading to good question!

Donsir_LTU Posted at 1-30 00:08
speechless too... How can this toy be called a drone? This thing can never be trusted, it is obstructed by everything ..... temperature, wind, water, buildings, bridges anything ... I wonder how people ignore such events and lie to themselves saying that everything is ok with this bird

Mavic Mini is a drone.  But like all things it has its limitations.  

Petty to hold extreme temperatures against Mavic Mini.
Off base to claim water is an issue.  MM is meant to fly, not submarine.
As for buildings and bridges, virtually all drones are going to be obstructed...

Only one is arguably fair, wind.  Mavic Mini has barely enough flight-speed to fight against average wind speeds through out world.  Solution is obvious, don't fly MM in windy conditions.  Light breeze or still air only.

Mavic Mini's real problem is DJI marketed it to general public, which have unrealistic expectations.  

Donsir_LTU Posted at 1-30 00:08
speechless too... How can this toy be called a drone? This thing can never be trusted, it is obstructed by everything ..... temperature, wind, water, buildings, bridges anything ... I wonder how people ignore such events and lie to themselves saying that everything is ok with this bird

I think wind for instance, Mavic mini is rated to be flown in highest wind 18mph, something like a Mavic pro is rated to be flown in winds no stronger than 22mph, so pretty clear that a drone for its size weight and a third of the price its not to far behind the Mavic pro for handling the wind . And this has been tested .

Just wondering: what to do in such compass error? Is there something to do or the drone is doomed to crash?
If so, DJI should work on an emergency feature that disables GPS and compass only to leave it to the pilot to save the drone. That seems elementary if there is currently no way to save the drone

JJB* Posted at 1-29 23:35
Hi Ewa,

Just a little question, at wich heading was your drone at the ground before take off?

Hi, AFAIR before start drone was positioned to the bridge - so to the nord-west, not south-west, as on image. After start I made some maneuvre to make sure all is right - including rotating and short flight in all direction. Then I rotated drone again to the bridge (nord-west) and lifted up. I had the view across the bridge. Then I started pitch right, to take o photo along the bridge. And then it flew away.

Samoth Posted at 1-30 11:05
Just wondering: what to do in such compass error? Is there something to do or the drone is doomed to crash?
If so, DJI should work on an emergency feature that disables GPS and compass only to leave it to the pilot to save the drone. That seems elementary if there is currently no way to save the drone

That’s what the pilot is for, there are checks you can do before flying including what you intend to do and what measures you intend to take if something might go wrong, it’s probably not just a coincidence that the vast amount of users don’t lose their drones.
This crash was caused by simple interference picked up on the ground, it could have been avoided and I posted in an earlier post how it happens and how to avoid it .

Ewa i Piotr Posted at 1-30 11:11
Hi, AFAIR before start drone was positioned to the bridge - so to the nord-west, not south-west, as on image. After start I made some maneuvre to make sure all is right - including rotating and short flight in all direction. Then I rotated drone again to the bridge (nord-west) and lifted up. I had the view across the bridge. Then I started pitch right, to take o photo along the bridge. And then it flew away.

Hi Ewa,

Thanks for your answer and the answer to what happend.

Compass heading in your log was not the actual heading of your drone.
So that is the reason for your fly-away, as described earlier.

Best is to always check before takeoff if drone heading in the app (map view) is the same as the actual situation.

cheers
JJB

HedgeTrimmer Posted at 1-30 09:58
Interesting analysis JJB, leading to good question!

Thanks, and his answer is what i expected. So now we know for sure why it flew away.

cheers
JJB

JJB* Posted at 1-30 11:31
Hi Ewa,

Thanks for your answer and the answer to what happend.

Thanks for advice! Will do - next fly.

HedgeTrimmer Posted at 1-30 09:48
"Disagree all you like but you're far out of your depth., as you showed in post #12 already."

My mistake ... that should have been #13

My analysis and diagnosis turned out to be correct.  #69 is confirmation.

Dirty Bird Posted at 1-30 11:40
Immediately take manual control & get the bird on the ground.  Power off, move to a new starting location, restart the aircraft.  Often times you don't have to move far.  If the interference is caused by metal in concrete, a bridge, or pier, you can usually just launch from your hand & that's enough distance to alleviate the issue.

If you wanted to be extra cautious you could add a compass to your kit & test the launch spot by placing it on your intended launch spot & comparing the pointer to magnetic north.

Pretty funny answer What to do when the drone spirals out of control? "Take manual control" Well, duh But how? Thankfully i never found myself in such a situation but I imagine when the drone goes diving like it did here all of a sudden, it won't respond to commands either. At least not in the way you're expecting it to, since you'll be asking to go right and the drone will do something different because of the messed up compass, right?

A compass error is one thing but a yaw error can be much, much more serious.
It all depends on how far out the initialised IMU values are.
If the difference is small, you get a chance to bring the drone back but if the difference is large, the drone accelerates quickly and usually crashes before you can even try to do something about it.

"After start I made some maneuvre to make sure all is right - including rotating and short flight in all direction. Then I rotated drone again to the bridge (nord-west) and lifted up."

Leading to several things...

1) What was source of Magnetic Inteference on dirt trail that was strong enough to mess with Mavic Mini's compass, not result in compass interference warning, and source was not obvious to OP?

2) Brings back reason to use a SmartDevice with it built-in (or downloaded) compass app or magnetic field detector app to check takeoff area for source of magnetic interference.

3) Flyaway as result of magnetic interference shows it is pointless to do a post power-up / mission pre-flight check involving: Lift-off, turn-right, turn-left, move-forward, move-backward, slide-left, slide-right, etc. to verify drone's Compass and IMU are in sync.

Dirty Bird Posted at 1-30 06:38
I suggest testing your belief by placing your drone on a large speaker for a few minutes then take her out for a spin.

What would happen?

HedgeTrimmer Posted at 1-30 18:28
"After start I made some maneuvre to make sure all is right - including rotating and short flight in all direction. Then I rotated drone again to the bridge (nord-west) and lifted up."

Leading to several things...

2) Brings back reason to use a SmartDevice with it built-in (or downloaded) compass app or magnetic field detector app to check takeoff area for source of magnetic interference.

Just checking the orientation of the drone and comparing to what the directional indicator in the app shows would be sufficient.

I am not sure if it applies here but I have experienced substantial compass problems in relation to bridges. I am not talking drone but boat. My boat has autopilot which is compass driven. I also have a large compass mounted in the boat.

Below is a photo of a bridge I regularly pass under while in this 400ft wide canal. If running dead center in the canal all is well. If I move to one side when I pass under - even though I still maybe 75 ft from the concrete piling - my compass will swing about 90 to 100 degrees.  As a result, if on autopilot it will steer the boat and make a hard turn right at the shore.  (Don't ask how I know that!) I have watched the compass swing going through here dozens of times.  I wouldn't dare fly my Mini anywhere near it.

Labroides Posted at 1-30 19:49
2) Brings back reason to use a SmartDevice with it built-in (or downloaded) compass app or magnetic field detector app to check takeoff area for source of magnetic interference.

Just checking the orientation of teh drone and comparing to what the directional indicator in the app shows would be sufficient.

Believe it has been discussed before here.

Something like 15 or more degrees of compass deviation from actual is enough to cause flight problems when drone clears a magnetic field and compass jumped to actual, resulting in conflict with IMU.

Issue being, it would be difficult to look at drone's orientation to within 15-degrees vs. GO-4 App's compass pointer.  Adding to difficulty is compass is a pointer, without degrees (digital display of).

Agree you could see 45-degrees or more, and guess down to half of 45-degrees (22-1/2 degrees).

Some have suggested using SmartDevice's compass to aid in checking drone's orientation.  Which will work for most SmartDevices.  But forget using DJI's CrystalSky's built-in compass via GO-4.  The bloody thing, update after update, which was supposed to fix it, still jitters like toy Chattering-teeth on Crack.

Raymacke Posted at 1-30 20:02
I am not sure if it applies here but I have experienced substantial compass problems in relation to bridges. I am not talking drone but boat. My boat has autopilot which is compass driven. I also have a large compass mounted in the boat.

Below is a photo of a bridge I regularly pass under while in this 400ft wide canal. If running dead center in the canal all is well. If I move to one side when I pass under - even though I still maybe 75 ft from the concrete piling - my compass will swing about 90 to 100 degrees.  As a result, if on autopilot it will steer the boat and make a hard turn right at the shore.  (Don't ask how I know that!) I have watched the compass swing going through here dozens of times.  I wouldn't dare fly my Mini anywhere near it.

Interesting piece of info.

Hi.
Question, the high voltage line could not disturb the compass?