If you really want to be a daredevil and try to find out what happens, try this, with the ac right next to the car, calibrate the compass ;)

If that doesn't cause an issue then its like Labroides says, these Minis a more forgiving than previous models, that's not to say it wont happen though you've just been very, very lucky, you know what its like, try to do sommat to show someone and ya can never do it, don't try to do it and it will happen, its always the blooming case, the OP has proved this.

Bashy Posted at 9-4 16:19
If you really want to be a daredevil and try to find out what happens, try this, with the ac right next to the car, calibrate the compass ;)

If that doesn't cause an issue then its like Labroides says, these Minis a more forgiving than previous models, that's not to say it wont happen though you've just been very, very lucky, you know what its like, try to do sommat to show someone and ya can never do it, don't try to do it and it will happen, its always the blooming case, the OP has proved this.

I am thinking we do not know all the details of how this can happen. I have tried at least 20 times in the last two days to create the problem and have failed. So I must be missing something. I think that there is more to this type of fault than we know.

For example, last year I could not launch if there was a compass error at startup. I was forced to either calibrate or reset everything. Now I cannot duplicate that restriction. No matter the method, when the compass error message appears, it NEVER prevents me from launching. I cannot help but wonder if the changes in firmware/software in the last year have made that unnecessary. Perhaps DJI has made changes that minimize the impact. One possible solution might be to delay sync with the IMU if the fault appears. Waiting till the drone is airborne might remove the issue, so why not sync at that time? That being said, if the "compass failed" flag fails to set when there is actual interference, then the result could be like the OP experienced. I will pursue the idea that minimal compass interference might casue a problem but fail to identify it to the system by using a smaller iron object. For what it is worth, I tried to replace the 10" steel tool with a 3' cast iron crowbar today and noticed less interference than the smaller tool.

Anyway, I would like to find the real explanation and solution.

Bashy Posted at 9-4 16:19
If you really want to be a daredevil and try to find out what happens, try this, with the ac right next to the car, calibrate the compass ;)

If that doesn't cause an issue then its like Labroides says, these Minis a more forgiving than previous models, that's not to say it wont happen though you've just been very, very lucky, you know what its like, try to do sommat to show someone and ya can never do it, don't try to do it and it will happen, its always the blooming case, the OP has proved this.

If you really want to be a daredevil and try to find out what happens, try this, with the ac right next to the car, calibrate the compass ;)
The result would probably be nothing since compass calibration isn't concerned with external magnetic influences.
It's only identifying what magnetic fields are part of the drone so they can be ignored.

Trying to calibrate the compass close to a car, would either calibrate properly or fail to calibrate at all.

GaryDoug Posted at 9-4 18:48
I am thinking we do not know all the details of how this can happen. I have tried at least 20 times in the last two days to create the problem and have failed. So I must be missing something. I think that there is more to this type of fault than we know.

For example, last year I could not launch if there was a compass error at startup. I was forced to either calibrate or reset everything. Now I cannot duplicate that restriction. No matter the method, when the compass error message appears, it NEVER prevents me from launching. I cannot help but wonder if the changes in firmware/software in the last year have made that unnecessary. Perhaps DJI has made changes that minimize the impact. One possible solution might be to delay sync with the IMU if the fault appears. Waiting till the drone is airborne might remove the issue, so why not sync at that time? That being said, if the "compass failed" flag fails to set when there is actual interference, then the result could be like the OP experienced. I will pursue the idea that minimal compass interference might casue a problem but fail to identify it to the system by using a smaller iron object. For what it is worth, I tried to replace the 10" steel tool with a 3' cast iron crowbar today and noticed less interference than the smaller tool.

Are you saying that you took off with a compass warning during your test?

I really don't think that there is any more to it than what's been said, there is however something about the mini 2 that's a bit more forgiving with regards to this issue, something DJI have come up with, I ain't got a clue what mind, but less forgiving obviously doesn't mean totally eradicated as shown by the OP.

If I was to guess why its more forgiving, it could be due to where the power button is, you have to physically lift up the AC to turn it on therefore it is further away from the source of interference, whereas other drones are turned on from the top or back thus can be turned-on on the ground, car roof or park bench etc.

If this had ever happened to you in the past, you wouldn't be trying to force it, its quite scary knowing ya could easily lose your drone.

If during your tests you're seeing the compass warning message, then your tests are conclusive, the problem is, either you're not flying away far enough or the field those objects are creating is not big enough to be of any concern.

Try this, sit the drone on the steel object or car roof (not if its aluminium though as some are), then keeping the rear end on the surface lift up the front just enough so you can get ya finger under to turn it on, then quickly drop it down and let it initialise, then take off from situ and see what happens.

Bashy Posted at 9-4 20:13
Are you saying that you took off with a compass warning during your test?

I really don't think that there is any more to it than what's been said, there is however something about the mini 2 that's a bit more forgiving with regards to this issue, something DJI have come up with, I ain't got a clue what mind, but less forgiving obviously doesn't mean totally eradicated as shown by the OP.

Yes that is what I said as was clearly stated: "No matter the method, when the compass error message appears, it NEVER prevents me from launching." NEVER.

I am trying to find out why almost nobody ever reports the source of magnetic interference, including the OP. I am not convinced that we know all about this type of anomaly. My test results do not show the problem so there must be something else that is the cause. It could be something in addition to the iron nearby.

The OP has a Mini 2 which ALSO has the power switch on the bottom. I would test with an actual Mini 2 but I have given mine away.

I AM flying farther away than the OP certainly. And one source of interference was an 3' IRON crowbar 1" thick directly under the mat. I could not have done more.

You do not have to take the chances. I am doing that for us all so don't worry about it. At this point my Mini is very expendable since I never use it.

As for your last statement, I have ALREADY done that at least 3 times.

GaryDoug Posted at 9-5 06:32
Yes that is what I said as was clearly stated: "No matter the method, when the compass error message appears, it NEVER prevents me from launching." NEVER.

I am trying to find out why almost nobody ever reports the source of magnetic interference, including the OP. I am not convinced that we know all about this type of anomaly. My test results do not show the problem so there must be something else that is the cause. It could be something in addition to the iron nearby.

Yes that is what I said as was clearly stated: "No matter the method, when the compass error message appears, it NEVER prevents me from launching." NEVER.

I just wanted clarification is all

I am trying to find out why almost nobody ever reports the source of magnetic interference

Because they are not aware of the cause, I was aware when it happened to me, it was taking off from my car sunroof, I am sure Labroides will remember it...

The OP has a Mini 2 which ALSO has the power switch on the bottom. I would test with an actual Mini 2 but I have given mine away.

True, but then that's back to my theory, maybe he turned it on too close to those railings, that's where my money lays...
As you're testing with a mini and not a mini 2, the test is null and void really, but then we don't know how different they are inside I spose.
But also, you should be recording each and every test and get them on youtube so we can see it from start to finish, it would be quite interesting, certainly to me

I AM flying farther away than the OP certainly. And one source of interference was an 3' IRON crowbar 1" thick directly under the mat. I could not have done more.

Thats something at least, the problem we do have is, your not in the same position as the OP, that's where you need to do your test, like I said, we don't know how much if any, magnetic field is being given off by those objects, again, you need to use an app that measures magnetic fields and get it all on video

You do not have to take the chances. I am doing that for us all so don't worry about it. At this point my Mini is very expendable since I never use it.

Thanks

As for your last statement, I have ALREADY done that at least 3 times.

Ok then, well all we can ascertain from this is either;

• the mini is not affected by a magnetic field
• the magnetic field isn't strong enough to affect the mini
• there is no magnetic field
• It's a mini, not a mini 2
  

Should have recorded the lot lol would have been interesting.
How about you go an find a old of railings like in the OP's video and see how you get on, please, record it

Excuse me for barging in here, but does the Mini RC not have the same option as the Mavic 2 Pro where you can see the the strength of the interference from a ferrous material. On the M2P you can see the strength going  from 0 to 999. Like I hade 999 from my car's roof. It goes from Green to Yellow and lastly to Red. If not, just disregard.

kyalami Posted at 9-5 07:50
Excuse me for barging in here, but does the Mini RC not have the same option as the Mavic 2 Pro where you can see the the strength of the interference from a ferrous material. On the M2P you can see the strength going  from 0 to 999. Like I hade 999 from my car's roof. It goes from Green to Yellow and lastly to Red. If not, just disregard.

No it does not and I have both....for now.

GaryDoug Posted at 9-5 08:58
No it does not and I have both....for now.

OK thanks for that. It would have been good to have, as then you can see what is going on. Perhaps DJI may include one day???

Bashy Posted at 9-5 07:37
Yes that is what I said as was clearly stated: "No matter the method, when the compass error message appears, it NEVER prevents me from launching." NEVER.

I just wanted clarification is all

"Ok then, well all we can ascertain from this is either;

    the mini is not affected by a magnetic field
    the magnetic field isn't strong enough to affect the mini
    there is no magnetic field
    It's a mini, not a mini 2"


The above is invalid. I never said that it was not affected. I said I already did it. Not the same thing. It is not affected only AFTER initialization. To ask why is another question and may be because the IMU is already initialized and trusted more at that point. As a retired test engineer I am looking for the facts, not hearsay.

Bashy Posted at 9-5 07:37
Yes that is what I said as was clearly stated: "No matter the method, when the compass error message appears, it NEVER prevents me from launching." NEVER.

I just wanted clarification is all

"How about you go an find a old of railings like in the OP's video and see how you get on, please, record it "

Pay attention here: The next to last link to a log was to almost exactly that. A galvanized steel chain link fence and gate that was about 2 feet away at launch.

JJB* Posted at 8-31 06:47
Hi,

Had a look at your log too.

JJB: Can you explain what the Drone Heading and Yaw Reference are in your graphs? Is one the compass and the other the IMU?
Thanks

GaryDoug Posted at 9-5 18:49
JJB: Can you explain what the Drone Heading and Yaw Reference are in your graphs? Is one the compass and the other the IMU?
Thanks

Hi Gary,

Good question and i have no a good answer.....

Drone Heading is data from the compass. My FRAP 'Yaw Reference" is, well i am not sure about source data for this value....

But what i do know is that in good flights (good working compass, no errors ec) both values are equal to each other or that the start difference remains the same during flight.
How do i know ?  well, my database of more than 5000 flights shows this behaviour of the 2 FRAP heading values.
So the moment the Compass line start to differ from the ref line, 'something' is not oke with the compass.

My last flight in the chart, hand takeoff and compass functioning oke.

In DAT files, don`t have the channel names in my head now, you will see also more lines for the YAW. I did few times a cross checki between my reference line and the yaw channel line in the DAT representation ; not always in my chart the difference (so DAT more accurate), but when seen in FRAP it was there in the DAT.

cheers
JJB