Fun with M3 GPS - first flight in southern hemisphere

[Meta4]

That's all about right except ...

Essentially a myth unless you have a faulty drone

I can't imagine how a faulty drone could record a bad homepoint.

 

[Not A Speck Of Cereal]

HOMEPOINT ERRORS
Essentially a myth unless you have a faulty drone or decide to make a homepoint at point A then drive to B and not understand why your drone heads back to point A in a RTH or disconnect.

Drive to point B: That would still be a user error, not a HP error.

Faulty Drone: have there been reports of drones creating false HPs? If not, I don't think that "unless" exists either and (more importantly) continuing to mention it as if possibly it exists is just going to lead to more confusion here on the boards. Nip it in the bud.

Chris

 

[Bussty]

Drive to point B: That would still be a user error, not a HP error.

Faulty Drone: have there been reports of drones creating false HPs? If not, I don't think that "unless" exists either and (more importantly) continuing to mention it as if possibly it exists is just going to lead to more confusion here on the boards. Nip it in the bud.

Chris

This thread contains a report of an actual faulty HP. 14kms away from actual location if you are looking for reports of that happening.

 

[Meta4]

This thread contains a report of an actual faulty HP. 14kms away from actual location if you are looking for reports of that happening.

No ... it doesn't contain a report of an actual faulty homepoint.
It contains an anecdotal account of a presumed faulty homepoint.
You would need to look at the flight data to see what was the cause of what has been presumed to be a faulty home point.
It's probably not what you've accepted.

 

[alex_markov]

It never happened.
Airdata has confused you by putting that H on the map, but the drone never recorded a homepoint.
This is a common glitch with Airdata.
If you looked into the actual flight data, rather than AIrdata's picture, you'd see the lat/long data is blank.
Or post a link to the Airdata report and I can show you.

You are not entirly correct in this case - I have many flights with no GPS and Airdata shows not known position, here M3 recorded some garbage - which was fun and i shared it

 

[alex_markov]

OMG I didn't even think about the battery issue! Of course if you can't get back on remaining battery it will send you home ASAP! So assume you can in that case hit the RTH button to cancel?

What's the forum view? If your drone creates a "false" Home Point then fly's away to get to it.... you SHOULD be covered by DJI for a warranty replacement?

I cancel it so have no flyaway and have no idea about DJI - but with care refresh nowadays it's easier to fly

 

[alex_markov]

Alex, Have you also noticed in Airdata that the flight distance does not show anymore? I have done some long flights recenetly and it shows for me a either 0km or 1km.

Never have this - always shos distance when GPS was working

 

[Meta4]

You are not entirly correct in this case - I have many flights with no GPS and Airdata shows not known position, here M3 recorded some garbage - which was fun and i shared it

I'm pretty sure that I am correct.
I've been looking at flight data for several years and seen similar things with Airdata reports before.
Please post a link to your Airdata report rather than a picture of a page.
And make sure that you tick the share box.
Then I'll show you the actual data and the lat and long columns of data will be completely blank.

Those Takeoff location and last known location numbers are junk that have not come from the actual flight data.

 

[alex_markov]

No ... things like that don't happen now and didn't happen in early days of the Mavic Pro.
I'd want to see the flight data to work out what actually happened before accepting that it did.

I like statments like Never happened
but I do not try to convince You it happened in aug 2017 I recovered my MP and never happened again ... so for me it's water under the bridge

 

[Meta4]

View attachment 146551
I like statments like Never happened
but I do not try to convince You it happened in aug 2017 I recovered my MP and never happened again ... so for me it's water under the bridge

I'm not sure what you think those numbers are illustrating?
I can't see how they are relevant.

I don't like accepting an interpretation of what might have happened, without seeing evidence to confirm that interpretation.
There are other possibilities and we just don't see GPS jumping around like you described.

The closest I've seen to what you described was back in 2017 when DJI released faulty firmware for the Phantom 3 Std.
Under some circumstances, it would cause the homepoint to jump a couple of kilometres.
DJI quickly fixed that with new firmware.
That was the only time I've seen a mandatory firmware update - necessary because of the problem it was fixing.

It would be very interesting to look at the flight data from that particular flight if you still have it, to find out if the homepoint actually jumped or if there is another explanation.
If it actually jumped, I'm sure we'd have seen a lot of cases of it before now.

 

[alex_markov]

I'm not sure what you think those numbers are illustration?
I can't see how they are relevant.

I don't like accepting an interpretation of what might have happened, without seeing evidence to confirm that interpretation.
There are other possibilities and we just don't see GPS jumping around like you described.

The closest I've seen to what you described was back in 2017 when DJI released faulty firmware for the Phantom 3 Std.
Under some circumstances, it would cause the homepoint to jump a couple of kilometres.
DJI quickly fixed that with new firmware.
That was the only time I've seen a mandatory firmware update - necessary because of the problem it was fixing.

It would be very interesting to look at the flight data from that particular flight if you still have it, to find out if the homepoint actually jumped or if there is another explanation.
If it actually jumped, I'm sure we'd have seen a lot of cases of it before now.

Dear Meta,
I had no intention to dive into such an argument, just mention precious memory - now I think I understand what happened - my iPad did not have a GPS fix and for an unknown reason, MP changed the Home point to remote controller position which was last place I fly 14 km away and RTH kicked in (with 76% bat. and few hundred meters from me ) I canceled it and all was fine... a few days later MP was bricked during FW update and under care refresh, mainboard was replaced ...
I am done discussing this long-forgotten flight like it never happened
Have a nice weekend!

 

[Bussty]

I'm not sure what you think those numbers are illustrating?
I can't see how they are relevant.

I don't like accepting an interpretation of what might have happened, without seeing evidence to confirm that interpretation.
There are other possibilities and we just don't see GPS jumping around like you described.

The closest I've seen to what you described was back in 2017 when DJI released faulty firmware for the Phantom 3 Std.
Under some circumstances, it would cause the homepoint to jump a couple of kilometres.
DJI quickly fixed that with new firmware.
That was the only time I've seen a mandatory firmware update - necessary because of the problem it was fixing.

It would be very interesting to look at the flight data from that particular flight if you still have it, to find out if the homepoint actually jumped or if there is another explanation.
If it actually jumped, I'm sure we'd have seen a lot of cases of it before now.

I can't imagine how a faulty drone could record a bad homepoint.

Like I say it's a myth unless you have a faulty drone and you have just confirmed that drones can and have caused this very fault. They did it once they could do it again.

So I'm happy knowing if for some reason my drone goes flying off to some inaccurate home point (while extremely unlikely) it will never be my fault and I would be approaching the drone manufacturer for a replacement.

That's all I needed to know.

 

[Meta4]

Like I say it's a myth unless you have a faulty drone and you have just confirmed that drones can and have caused this very fault. They did it once they could do it again.

I told you of a special case that only happened a few times
It was due to faulty firmware and was fixed within about two weeks and has never been seen again.
If things like this could happen, we'd have seen incidents showing up and we haven't.

The OP has now given some more details and it turns out that he accidentally reset the home point.

So I'm happy knowing if for some reason my drone goes flying off to some inaccurate home point (while extremely unlikely) it will never be my fault and I would be approaching the drone manufacturer for a replacement.

Your drone won't go flying off to some inaccurate home point.
If you ever think it does, it's quite likely that the reason would be something that you've done.

 

[Meta4]

You are not entirly correct in this case - I have many flights with no GPS and Airdata shows not known position, here M3 recorded some garbage - which was fun and i shared it

Alexei ... can you please post the link to that Airdata report?
No-one is trying to argue anything ... just to find the real explanation for something that seems odd.

 

[EsaT]

Spoiler

So to put this all into perspective...

HOMEPOINT ERRORS
Essentially a myth unless you have a faulty drone or decide to make a homepoint at point A then drive to B and not understand why your drone heads back to point A in a RTH or disconnect.

YAW / COMPASS ERRORS
Can absolutely be user error with starting of drone in a magnetic interference area with IMU being set incorrectly to compass then evolving into an impossible YAW correction by IMU once magnetic interference no longer exists.

Can be reduced by...

• Hand launching (without watches etc to interfere with compass of drone) plus should be a long way away from rebar etc
• Making sure drone on DJI Fly app faces same orientation of actual drone

Is that correct?

Any other traps for drone pilots involving location errors/flyways?

Cheers

Already skipped some safety steps:

• Calibrate compass away from metal objects.
  
  Those can and will distort local magnetic field causing error when drone gets away from object.
  (magnetic anomalies are used to detect submarines from air)​
• Before taking into air wait for recording of home point.
  And check that it's in correct location on map and not in some lake etc.
  • Unless flying indoors/in place without GNSS reception.
    
    In which case it would be good to disable RTH.
    (sporadic reception might result inaccurate home point)​

 

[Bussty]

Already skipped some safety steps:

• Calibrate compass away from metal objects.
  
  Those can and will distort local magnetic field causing error when drone gets away from object.
  (magnetic anomalies are used to detect submarines from air)​
• Before taking into air wait for recording of home point.
  And check that it's in correct location on map and not in some lake etc.
  • Unless flying indoors/in place without GNSS reception.
    
    In which case it would be good to disable RTH.
    (sporadic reception might result inaccurate home point)​

Hi Esat

That first point makes sense but how will you know? That's why having the same compass orientation on your DJI Fly app matching where your drone is pointing is a good way to find out if you have interference or not. Example you think having your drone on a flat piece of grass with no metal within 100 meters will be the perfect compass calibration point but what you don't know is there is a lump of metal under the ground giving you interference. Checking your drone orientation on the DJI Fly app matches where your drone is pointing (so long as you know how your map should be aligned) would confirm if you have interference or not.

Your second point.. the very nature of this thread says your drone will never record a false Home Point and I know with the GPS issues on the M3 many are launching their drones first to speed up the acquisition of that point (I think height is limited and positioning is handled by the vision sensors.)

I don't know from first hand experience as I don't own a M3 yet but planning to once it is finalized.

 

[EsaT]

That first point makes sense but how will you know?


Your second point.. the very nature of this thread says your drone will never record a false Home Point...

Don't calibrate compass when on some ship/metal structure etc.
Unless you're standing on/next to some sizable iron meteorite, amount/concentration of metal in ground is such small that it causes only miniscule change to magnetic field and effect weakens fast with distance.
Well, maybe except if you're in the middle of iron ore mine...


Drone doesn't care if first coordinates it can calculate and locks onto as home point are from actual home point, or if you've already flown drone above unsuitable place to land like water/big trees etc.

Mavic 3 is of course issue DJI needs to address completely and finally.
And really should have done that already months ago, instead of following Microsoft's quality policy.
If priorities are right, it shouldn't take this long with resources of DJI size company.

 

[Meta4]

Don't calibrate compass when on some ship/metal structure etc.

You shouldn't have any need to be recalibrating the compass anywhere.

 

[Mike Mas]

Thanks for posting - Since there is erroneous GPS data on behave of the drone, as expected Air Data faults with locations and creates what is called a ghost flight!

Fly Safe - Mike