Hi all
I am still not getting this compass thing.
In the military I was "fluent" in compass using it underwater mostly, with a high accuracy.

But this spark has me beat:

First - I verified the accuracy of my cellphone compass with a needle type traditional compass.
At the site I verified the position of north using Google earth - so now I am certian where north is and that the cellphone compass is accurate.

I point Spark in the direction of north and hover it in front of me facing north as I am facing north.

The ruddy radar display does not want to play with, showing that the spark is pointing backwards and north has gone walkabout. Unless by brains are baked.

And by the way, I OFTEN fly here abd today I get a compass error, when only last week I did a compass calibration.
You may note that the DJI GO app says it is a filed error - which field was filed I know not.
But to quote a little green man "..."

huh thats weird
try reinstalling the firmware and re calibrating without any device near you

Hi,

Outer ring is not correct, gets its info from the compass of the mobile device.
So calibrate your device.
What mayby helps is to rotate device in all the 3 axis.

But if your drone was facing North, the red arrow (compass data drone) should point North at the outer ring.So drone compass is offset too.

Picture plastic compass and phone is inside, check this on the spot if your device is aligned with that manual compass.

cheers
JJB

looking at this AND your other thread that you posted on the radar / compass I still don't think you are getting one very basic concept with the radar....

The BASE of this instrument is the compass in the cell phone... ignoring the rest of the 'data fusion;' going on with this instrument like the artificial horizon and power indicators you MUST think of it this way..

the DISK is the COMPASS CARD for the PHONE and the N marker on that CARD rotates with the CARD (the entire disk) as your phone rotates and it references your PHONES idea of North.



the center of that CARD is you location and as you fly out several hundred feet the location of the red triangle is the AIRCRAFT POSITION relative to your position PLOTTED ON THE ROTATING CARD .. the ORIENTATION of the red triangle is the AIRCRAFT HEADING referenced to the CARDS N

SO moral of the story.. if you phone is not correct, or it drifts while using it.. it all falls apart...

now to your images..


in the picture the AIRCRAFT HEADING is roughly 340 .. its (red chevron) is pointing just left of the N marker..

The phone however, if you are facing north, is about 100 degrees OFF as it is telling us the phone is facing 100 degrees or ESE ..

you will notice however that the aircraft is plotted CORRECTLY RELATIVE to the INCORRECT CARD..   THE AIRCRAFT PLOT OF N IS ALWAYS RELATIVE TO THE MARKING ON THE CARD .. NOTHING ELSE..

In your indoor photos .. off the top of my head show the same thing...

the AIRCRAFT didn't change HEADING in those photos.. the red chevron had the same heading REF the N on the disk.... BUT the phones North reference SHIFTED and turned the radar display ...

This happens occasionally with me on an iPhone 6 as well.. especially if I am flying where I am turning my whole body around several times to follow it aircraft location.. I will start the flight and the North reference / streets are lined up correctly but as the flight progresses my phone will become 'off'.

additionally , my sparks and mavics internal heading will become off during the flight as well and routinely land 10-20 degrees different when they land... but my phone can be off by 100 degrees or more sometimes...

if this becomes the case during flight. if it disorients you.. discontinue use of the radar . or realize that only the aircrafts pointing to the N will be accurate but relative position will be wrong.. and on the maps.. select North up so that the map does not orient incorrectly and fly the aircraft relative to the fixed map..

hope that clears that up some...

DeuceDriv3r Posted at 11-12 05:05
looking at this AND your other thread that you posted on the radar / compass I still don't think you are getting one very basic concept with the radar....

The BASE of this instrument is the compass in the cell phone... ignoring the rest of the 'data fusion;' going on with this instrument like the artificial horizon and power indicators you MUST think of it this way..

And this is why I use the map screen and flight path for orientation. I prescribe to the KISS principle.

Yes
I must just get my head around it.

But to me it is not as Simple as it Sounds

it really is.. its not rocket science...

I calibrate my phone when I get where I am going to fly IN THE RC CLAMPS after I turn it on.... its seems to help

then once the RC is up and the iPhone app is open.. and I am happy that the map/compass of the app is matching with known headings or geographic features.. I fire up the aircraft...

once you use it a couple times.. calibrated.. you will see what is going on... if you can be taught land navigation under water this won't be difficult by comparison ...

Yep... it's all starts with a proper PHONE calibration WITH CONTROLLER attached.  If you don't do this step, then as soon as you put the phone in the controller, the magnets in the sticks throw the compass in the phone out of whack and the whole chebang goes out to lunch.

It's super easy to see just how much the controller affects the phone's compass by calibrating the phone independant of said controller and getting it to match your hand held compass.  THEN while keeping the magnet app open, attach the phone to the controller and see just how far that needle swings away from what it only seconds ago said was North.

White Ox Posted at 11-12 15:12
Yep... it's all starts with a proper PHONE calibration WITH CONTROLLER attached.  If you don't do this step, then as soon as you put the phone in the controller, the magnets in the sticks throw the compass in the phone out of whack and the whole chebang goes out to lunch.

It's super easy to see just how much the controller affects the phone's compass by calibrating the phone independant of said controller and getting it to match your hand held compass.  THEN while keeping the magnet app open, attach the phone to the controller and see just how far that needle swings away from what it only seconds ago said was North.

hey ox... pretty sure he doest care right now since it left him for another hours ago.. salt in the wound kinda thing....

Yeah... I realized that ... AFTER.... Sorry.

I wonder how many fly-aways were connected to a phone / tablet showing the wrong direction as North?  I did a flight today, and she kicked into Atti mode.  It wasn't till after the flight that I realized I was flying with the new tablet I'm setting up and I had forgotten to calibrate the compass with controller attached.  Of all my flights, I have only had it kick into Atti mode one other time.. hmmmm

White Ox Posted at 11-12 16:10
Yeah... I realized that ... AFTER.... Sorry.

I wonder how many fly-aways were connected to a phone / tablet showing the wrong direction as North?  I did a flight today, and she kicked into Atti mode.  It wasn't till after the flight that I realized I was flying with the new tablet I'm setting up and I had forgotten to calibrate the compass with controller attached.  Of all my flights, I have only had it kick into Atti mode one other time.. hmmmm

I doubt there is any 2 way positional data going back and forth between the mobile and the aircraft.. again DJI isn't exactly transparent.. but..

I have seen at least one reported fly away attempt where a guy had updated his home point in flight using 'controller position'  then hit RHT and it bolted... the waypoint was wrong... also couldn't fly back to his position because he had a max distance geofence set in his settings and he physically was outside that bubble now and the drone could not fly back to him... I think that is a rare case

I had my spark go atti on the RC not 200 ft away and maybe 100ft up after 5 minutes flight .. no warning.. close enough at least that RC and video didn't drop and flew it back manually and landed..

mavic pro flew it straight up.. turned it 90 degrees at 399ft and flew it out apr 4000 ft stopped to set up a pano and as soon as I came off the sticks it dropped the 20 sats went into atti.. again at least I had video and control and as I was flying back it regained gps.. continue the flight home and landed..
cvs analysis of that flight clearly showed yaw magyaw chasing each other for the flight with no warning on the controller until she when into atti while the logs showed instances of compass warnings x20 x50 etc.. but nothing on screen.. so DJI must be suppressing them until it reaches its limit..

the flight info on mavichelp and airdata is too superficial for a good analysis.. have to see the raw traces in csvview to really see what is going in....

i had a similar red arrow  facing  problem also . in the 4 small boxes  hit the compass box that will point the red arrow in the direction of your spark .

White Ox Posted at 11-12 16:10
Yeah... I realized that ... AFTER.... Sorry.

I wonder how many fly-aways were connected to a phone / tablet showing the wrong direction as North?  I did a flight today, and she kicked into Atti mode.  It wasn't till after the flight that I realized I was flying with the new tablet I'm setting up and I had forgotten to calibrate the compass with controller attached.  Of all my flights, I have only had it kick into Atti mode one other time.. hmmmm

Hi White Ox,

I am pretty sure that there is no link from phone compass to drone compass etc.
The only thing a mob device does is to show the N (this the full 360) compass indication in the go app to the actual North. N = N if mob device compass is ok calibrated.
Red arrow is data from compass drone, if both good calibrated the red arrow heading is set to the corect outers scale of the compass in the app.

I flew many times with a wrong mob device compass (un purpose, have 2 devices), just hard to interpretet where the red arrow is pointing (90 degrees off set to the left means visualize 90 degrees right to get it right), but one thing is always correct ; red arrow pointing to the center (HP) it will bring the craft back home. If ofcourse HP is set at takeoff.

Just check many video`s on YouTube ; many video`s shows incorrect compass outer scale indication, but no problem during flights.If outerscale is set OK but the red arrow doesn`t indicates the actual heading of the drone (within few degrees offset), oops be carefull and likely to get compass fail during flight
I tested several times a RTH with a wrongly mob device compass and they all went perfect, craft RTH ok.

see this > https://forum.dji.com/forum.php?mod=viewthread&tid=167473   red arrow pointing incorrect!

cheers
JJB

I don't know how to do multi-quotes, so I'll just say "Thanks Fellas!"

I've always enjoyed asking questions to get a deeper understanding of my hobbies.  People in the know, willing to take time to type replies helps all of us willing to take time to read.  <---Seems both of those are slowly becoming a lost art. HAHA  Hopefully, eventually, things will come full circle and I'll be doing some of the helping.

JJBspark Posted at 11-13 00:55
Hi White Ox,

I am pretty sure that there is no link from phone compass to drone compass etc.

there are only 1 think I can think of for sure that the mobile devices PUSHES to the drone.. and that is the location of the remote..

now I ASSUME that is only when we tell the drone, through the go 4 app to update the home point manually and use my present position...

HOWEVER

and I would love someone to weigh in that actually knows...

but I POSTULATE .. that the way DJI get such great range using Occusync or even WIFI on the spark it thought a combination of beam steering and beam shaping..

while we KNOW that DJI has confirmed and its obvious that these are directional antennas on the remote.. since they tell us specifically they don't radiate off the tips.. what they don't tell us is if they use beam steering. on the remote I don't think they do since there are reverences to keep the remote facing the drone ... but on the drone itself?  I don't know..

and if the drone itself IS beam steering .. does it just use the 'static;' location of the home point recorded as to where to steer the signal or is it updated at some rate?  

another reason that I think it just uses where it recorded its takeoff .. is that I have used these products with the phones GPS/cellular unavailable or off and therefore gps was off etc.. and signal range was not degrades ..

another thing that leads me to believe that beam steering might me used is that at longer ranges, when a craft goes into atti mode and thus looses heading... IE not knowing where to point ... that it is often accompanied at extended ranges with a loss of video and or RC signal .. so during an atti does the drones beam steering antenna become an omni antenna and loose gain.. unit P-GPS comes back?

again postulations and theories... anyone have something to add

DeuceDriv3r Posted at 11-13 05:08
there are only 1 think I can think of for sure that the mobile devices PUSHES to the drone.. and that is the location of the remote..

now I ASSUME that is only when we tell the drone, through the go 4 app to update the home point manually and use my present position...

Yep true yor first remark. Manually updating HP goes via mob device > app > drone, that about GPS data.
But there is no compass to compass direct line between mod device and drone.

cheers
JJB