I found the latest Spark User manual v1.6 available in the DJI GO 4 app says during calibration, the second turn should be with the camera pointing  down (ground). However, the App during calibrations says it should be on its side. Are they both valid techniques? Which one is the correct procedure? I have been having compass issues without good explanation I am wondering if I am doing something wrong during calibration.

Hi, thanks for your inquiry. You mean the instruction on the application is didn't match on the position on the manuals? Are you getting an error message while doing the compass calibration? Please check if there's a metal object or possibly interferences on your location for better calibration as well. Please keep us posted for further help. Thank you for the support.

I follow the DJI Go 4 App instructions when doing compass calibration. I found that if I am wearing my smartwatch the calibration gets affected. I turn it off while calibrating the compass and it works.

DJI Tony Posted at 11-5 12:57
Hi, thanks for your inquiry. You mean the instruction on the application is didn't match on the position on the manuals? Are you getting an error message while doing the compass calibration? Please check if there's a metal object or possibly interferences on your location for better calibration as well. Please keep us posted for further help. Thank you for the support.

1 - Correct, the instructions don’t match as seen on the pictures provided on the first post. Which is the correct instruction to follow?

2 - I've been having GPS issues lately with my spark I always get the yaw errors shown in the picture provided here. I noticed within the last month, the 3 times that I had this issue I was using the same battery. Do you know if the battery can be causing this issue? All 3 times I was flying in an open area that I highly doubt there is any magnetic interference to affect the compass. I also noticed all 3 times the Drone regained GPS on its own after being a few seconds in atti mode. Any clues? Advise? Thank you!

Hani Posted at 11-5 12:58
I follow the DJI Go 4 App instructions when doing compass calibration. I found that if I am wearing my smartwatch the calibration gets affected. I turn it off while calibrating the compass and it works.

Thank you. My issue is not during calibration. Lately I had 3 incidents with my drone going on atti mode with no good explanation. It happened once while hovering in place in a frequently known area and another time in an open space with no asphalt or metallic structures around. I am nervous to fly my spark and I have been looking at the flight records with no clear explanation of why I am getting this issue and this often. When I do my  calibration, I even take off my belt and wedding band and ask someone to hold the RC away while I rotate the spark as shown in the app. I always get a successful calibration message, but again mid flight I have been getting these GPS loses. Thanks!

Most people point the nose of the drone down when calibrating the compass. Pointing it to the side will work just as well.

djiuser_GXwSwgThfBNN Posted at 11-5 13:31
1 - Correct, the instructions don’t match as seen on the pictures provided on the first post. Which is the correct instruction to follow?

2 - I've been having GPS issues lately with my spark I always get the yaw errors shown in the picture provided here. I noticed within the last month, the 3 times that I had this issue I was using the same battery. Do you know if the battery can be causing this issue? All 3 times I was flying in an open area that I highly doubt there is any magnetic interference to affect the compass. I also noticed all 3 times the Drone regained GPS on its own after being a few seconds in atti mode. Any clues? Advise? Thank you!

yep... DJI picked up the firmware across almost their entire line since the .04.0xxx updates..

the mavic pro, air, 2 etc all showing the same issues..

in csvviewer you can watch the yaw and magyaw diverge unitil the imu gives up and goes into atti mode...

I am thinking class action law suit .... anyone an attorney around here

I point the camera down. Works just fine for me.

Hi,

I just bought a new Spark but I faced a problem. The DJI GO application always prompt me to compass calibrate in the aircraft status beside the word "normal" indication. I did calibrate the compass each time it showed in the aircraft status screen.

Can I know why this compass calibrate prompt is always appearing in the aircraft status whenever I am connected to spark on the DJI GO application? Even though I did calibrate compass and is successfully everytime, the next time of usage when I am connected, the aircraft status still reflect compass calibrate but has the word "normal" indication.

The calibrate prompt is always appearing on the aircraft status whenever I am connected to Spark through the remote controller with iPhone. I have calibrated it for many times but it just wont go away. I did as instructed each time to stay away from buildings or metallic items and rotate the Spark as shown in the compass calibration instructions. Can I check for the second phase of compass calibration vertically, do I point the nose (camera) downwards or sideway? There are many different versions to it.

Please help me on this compass calibrate problem. Thanks a lot.

Sorry, the photo as attached as I don't know why it didn't appear in my first post:

DAVE6820 Posted at 11-5 22:12
The DJI GO application always prompt me to compass calibrate in the aircraft status beside the word "normal" indication.

That's just a button to initiate the compass calibration if you'd like to do it for some reason. It's not prompting you to do it.

msinger Posted at 11-6 02:35
That's just a button to initiate the compass calibration if you'd like to do it for some reason. It's not prompting you to do it.

Thank you so much for your help. Really appreciate it as it has been bothering me for the whole day. Cheers...

DJI Tony Posted at 11-5 12:57
Hi, thanks for your inquiry. You mean the instruction on the application is didn't match on the position on the manuals? Are you getting an error message while doing the compass calibration? Please check if there's a metal object or possibly interferences on your location for better calibration as well. Please keep us posted for further help. Thank you for the support.

The question was asked which is the correct aircraft position: camera pointing down or on its side.

As usual it seems to be impossible to get a direct, simple answer when asking a simple question. Admins appear to be instructed to only ask more questions and side-step the original request whenever possible.

Why must it always be this way?

Now, WHAT IS THE CORRECT POSITION?

Chuck101 Posted at 11-6 05:30
The question was asked which is the correct aircraft position: camera pointing down or on its side.

As usual it seems to be impossible to get a direct, simple answer when asking a simple question. Admins appear to be instructed to only ask more questions and side-step the original request whenever possible.

As msinger and others above have stated, it will work either way.

my understanding .. and I am not a MEMS package designer or electrical engineer is that the multiple orientations are to get the earths inclination and declination angles (the earths magnetic fields actual lines of orientation at a given location) to pass through the sensor and multiple angles ....

I think as long as the procedure orients the chip flat.. then on edge.. it should work.. now whether which edge is presented makes a difference I can't say.... but some people swear by the nose down method...
here is a pretty good primer on the subject ...

http://geokov.com/education/magnetic-declination-inclination.aspx

Chuck101 Posted at 11-6 05:30
The question was asked which is the correct aircraft position: camera pointing down or on its side.

See post #6 above.

If you're still unsure, then go outside and try it. You'll in fact find it works either way.

djiuser_GXwSwgThfBNN Posted at 11-5 13:31
1 - Correct, the instructions don’t match as seen on the pictures provided on the first post. Which is the correct instruction to follow?

2 - I've been having GPS issues lately with my spark I always get the yaw errors shown in the picture provided here. I noticed within the last month, the 3 times that I had this issue I was using the same battery. Do you know if the battery can be causing this issue? All 3 times I was flying in an open area that I highly doubt there is any magnetic interference to affect the compass. I also noticed all 3 times the Drone regained GPS on its own after being a few seconds in atti mode. Any clues? Advise? Thank you!

Do an IMU calibration, too
Check afterwards if the Yaw Error happens less often.
I think, a yaw error (can be triggert in/with switching to sport, sometimes) is disturbing the compass.
So in my scenario the compass failure is a follow up on the yaw error, which may be an IMU problem.

Have you had "strange behavior" on take offs?
Spark not climbing, but instead bumping on the floor a couple of times?

S-e-ven Posted at 11-6 08:53
Do an IMU calibration, too
Check afterwards if the Yaw Error happens less often.
I think, a yaw error (can be triggert in/with switching to sport, sometimes) is disturbing the compass.

I don't think that its necessarily the switch to sport mode .. as much as its the INCREASE in RATES that sport mode allow and are probably not being sufficiently mathematically calculated / smothed / and corrected...

the other potential for error here in sport mode is the increased current being moved about the airframe... another potential source of induced magnetic fields that perhaps are not getting compensated for ...

it all really just points to DJI having made a change in the FC math ... probably for a new product or change in MEMS parts suppliers to 'fix' an issue that devolved into inducing this yaw issue into just about everything else...

one thing is abundantly clear in the spark.. thats the inability to properly maintain its 'horizon' math in sport FPV mode.. if flying sport down low and close in not really hitting full speed.. coordinated bank then yaw turns are fairly proper with proper 'horizon' tilt.. but get up high and get going faster and no amount of proper coordinated turn keeps the 'horizon' tilting correctly .. I get 'G' induced opposite horizon tilt on just about every turn and worse I believe to the right where right turns tilt heavy 'left turn' then jump back over .. so clearly either the rates are too high .. or more likely.. their math is wrong or the processor is too slow to make the calc in real time so that the video matches the airframe...

S-e-ven Posted at 11-6 08:53
Do an IMU calibration, too
Check afterwards if the Yaw Error happens less often.
I think, a yaw error (can be triggert in/with switching to sport, sometimes) is disturbing the compass.

I have done both IMU and compass calibrations on all 3 times that I had this issue. I take each calibration extremely serious reason why I am asking DJI for the best/proper calibration to follow.

No, I have not had any abnormal behaviors that I can perceive during take off. It just goes atti after showing me the yaw errors that I shared previously. Now, yesterday I did an IMU followed by compass calibration, then I did the experiment (see pictures) to align both my phone and spark facing north and I was surprised to see they don’t line up! Per the visual it seems the drone to be off about 5 degrees... Is this supposed to be normal??. I am not trying to argue if my phone has the correct north orientation or the drone. My point is why if they are perfectly aligned the center of the green mark coming from the front of the spark is not overlapping the white triangle? This might need another forum to discuss..

djiuser_GXwSwgThfBNN Posted at 11-6 09:10
I have done both IMU and compass calibrations on all 3 times that I had this issue. I take each calibration extremely serious reason why I am asking DJI for the best/proper calibration to follow.

No, I have not had any abnormal behaviors that I can perceive during take off. It just goes atti after showing me the yaw errors that I shared previously. Now, yesterday I did an IMU followed by compass calibration, then I did the experiment (see pictures) to align both my phone and spark facing north and I was surprised to see they don’t line up! Per the visual it seems the drone to be off about 5 degrees... Is this supposed to be normal??. I am not trying to argue if my phone has the correct north orientation or the drone. My point is why if they are perfectly aligned the center of the green mark coming from the front of the spark is not overlapping the white triangle? This might need another forum to discuss..

1 are you lining  both the phone/controller and spark to mag north or true?

2 are the controller/phone at least 10 to 20 feet apart from the aircraft?

3 if you are sitting on concrete ... good chance there is rebar in there

4 I assume that the phone compass has been calibrated too ..

I can tolerate there being a minor discrepancy on the radar as there are a lot of different measured inputs all being overlaid here so there is going to be error stacking..

how does your bird do on a long RTH ... fly it 3000 feet away... fly around a bit.. maneuver .. then RTH.. if when the aircraft snaps back around.. in the video measure the angle the aircraft is pointing vs where the homepoint really is.. that angle difference is the heading error of your drone...  using the GO 4 .. you can go into the flight log.. pull up that flight and with the satilite map pulled up.. when the drone turns to face the home point.. which it will show in the map.. becasuse it thinks that is where its pointing.. find the actual geographical feature that the drone is pointing at in the video... that is an easy way to measure the angle.. My spark is about 30 degrees rotated right on every RTH and sometimes that angle will worsen on the way back..

Is the north of your phone wrong or the north of the bird?

S-e-ven Posted at 11-6 09:26
Is the north of your phone wrong or the north of the bird?

another good point..

for example lets say that the N on the radar is 10 degrees to the left assuming mag north straight ahead and the drone is off 10 degrees right.. you will see a 20 degree split on the radar view...
thats error stacking

S-e-ven Posted at 11-6 09:26
Is the north of your phone wrong or the north of the bird?

Don’t focus on the North coordinate. In this experiment I was trying to align 3 things. 1 - North, 2- spark, 3 - phone all on the same orientation. Again, ignore number 1 (North coordinate) for a second. If my drone and my phone are perfectly aligned the white triangle which is under the North simbol in the picture, should be right in the middle perfectly align in front of the spark’s red arrow or the middle of the green mark that comes from the front of the spark. See how the white mark under the North symbol is sitting on the edge of the green shadow not in the center. Why? Is this normal? Is this an acceptable orientation tolerance? What’s the engineering released orientation tolerance?

DJI please feel free to weigh in with your coments. Thanks!

Isn't the "white triangle" a circle with a H inside, symbol for the homepoint if GPSfix happened?
You are indoors, how good is the GPS reading?

And how about the North?
Do you know exactly where north is?
Have you pointed the nose of the bird to the north?
And did you put you RC accordingly?

That is the only way, to find out which north is wrong and by how much.
Just: Are you with drone and phone enough away from metall?
Not only the compass of the drone is quickly disturbed.
Also the phones compass is it.
And Phone, RC and Spark close by is not the best configuration.
I hope, that is just a concrete flagstone under the bird, not one with steal inside!

S-e-ven Posted at 11-6 09:50
Isn't the "white triangle" a circle with a H inside, symbol for the homepoint if GPSfix happened?
You are indoors, how good is the GPS reading?

Please ignore the North Coordinate. Focus only on the white triangle that always stays right in front of your phone. My understanding is, the white triangle had to be right in front in the 90 degree vector in the middle of the green shadow. I was outside during this trial. The picture attached to this coment only to show the white triangle that I have been referring to and you can’t see on previous pictures as it was hidden under the North Symbol. Please ignore the North simbol. Thank you!

djiuser_GXwSwgThfBNN Posted at 11-6 09:58
Please ignore the North Coordinate. Focus only on the white triangle that always stays right in front of your phone. My understanding is, the white triangle had to be right in front in the 90 degree vector in the middle of the green shadow. I was outside during this trial. The picture attached to this coment only to show the white triangle that I have been referring to and you can’t see on previous pictures as it was hidden under the North Symbol. Please ignore the North simbol. Thank you!

in this picture.. you aircraft is pointing to where IT thinks north is... and your phone is pointing west.. as the red triangle is pointing at the N marker .. that is north...

in the photo above... you can see the red triangle is not pointing at the N .. so the aircraft is not lining up to north... but your phone is.. since N is on the lubber line...

DeuceDriv3r Posted at 11-6 10:33
in this picture.. you aircraft is pointing to where IT thinks north is... and your phone is pointing west.. as the red triangle is pointing at the N marker .. that is north...

in the photo above... you can see the red triangle is not pointing at the N .. so the aircraft is not lining up to north... but your phone is.. since N is on the lubber line...

https://youtu.be/ifRLlZQCqAE

I am not trying to say this is a problem. Could someone else try this experiment? Share the attitude indicator picture when you align your phone orientation (white triangle) with the drone if it’s perfectly aligned in front of the tip of the red arrow (drone). May be this is why my drone is having compass errors. May be this is why some people had lost their birds.. May be.. I wish DJI had more interaction in this forum.

djiuser_GXwSwgThfBNN Posted at 11-6 10:35
https://youtu.be/ifRLlZQCqAE

dude this isn't rocket science ..

the radar  'disk' rotates with the phone .. the N reverence is just a reverence.. if your phone is not calibrated than this N will be off.. and your phone/controller won't be pointing to the aircraft... it will be random.. but the N is ALWAYS the N reverence to which the red triangle is oriented to when plotting ITS north reference..

IE if the aircraft is indeed pointing North.. and its calibrated and no kidding pointing North it (the red triangle) will be pointed at the N on the outer edge of the radar disk ... the N is a fixed position ...  now if you rotate your phone / controller around the disk will spin.. but the red triangle will always still point at the N as long as the aircraft is still pointing North..

if your phone is calibrated so that when You are facing North holding the controller and the N on the radar is at the top of the radar display under the lubber.. your phone/controller is good... if the aircraft is aligned north now to. the red triangle will also be pointing straight up.. north as well...

I am done

DeuceDriv3r Posted at 11-6 10:41
dude this isn't rocket science ..

the radar  'disk' rotates with the phone .. the N reverence is just a reverence.. if your phone is not calibrated than this N will be off.. and your phone/controller won't be pointing to the aircraft... it will be random.. but the N is ALWAYS the N reverence to which the red triangle is oriented to when plotting ITS north reference..

I suggest investing in a real compass for this 'test'