Compass calibration thoughts

[kd4pbs]

I'm not sure if this has been mentioned before, so please pardon me if it has been discussed.
I have read where many pilots recommend removing the camera and gimbals from the H before doing a compass calibration. The reasoning behind this is sound, for anyone who has ever calibrated the compass while the camera is attached. You can see the gimbals banging away against it's stops while rotating the aircraft. The problem I see is that this would be fine if we are flying the aircraft without the camera installed, but when replacing the camera, we've suddenly introduced an electromagnetic field in close proximity with the magnetometer, as indeed the motors for the gimbals for the camera do create a measurable magnetic field. My thinking is that this could cause incorrect compass readings for the flight controller since the magnetometer was calibrated without this additional magnetic field.
I've always followed the Yuneec instructions and left the camera attached. Besides, I'm pretty sure the camera gimbals are rated and engineered to be able to survive any stop-to-stop gimbal excursions that it might experience.
I wonder if calibrating the compass without the camera attached, then reattaching the camera has caused some of the flight instability issues that some have experienced.
To test this theory, I plan on connecting the H to the GUI and noting compass readings with and without the camera attached, rotating the aircraft along the yaw axis.
Now, where's my round 'tuit... I'll certainly need that before I do any experiments.

 

[FlushVision]

I'm not sure if this has been mentioned before, so please pardon me if it has been discussed.
I have read where many pilots recommend removing the camera and gimbals from the H before doing a compass calibration. The reasoning behind this is sound, for anyone who has ever calibrated the compass while the camera is attached. You can see the gimbals banging away against it's stops while rotating the aircraft. The problem I see is that this would be fine if we are flying the aircraft without the camera installed, but when replacing the camera, we've suddenly introduced an electromagnetic field in close proximity with the magnetometer, as indeed the motors for the gimbals for the camera do create a measurable magnetic field. My thinking is that this could cause incorrect compass readings for the flight controller since the magnetometer was calibrated without this additional magnetic field.
I've always followed the Yuneec instructions and left the camera attached. Besides, I'm pretty sure the camera gimbals are rated and engineered to be able to survive any stop-to-stop gimbal excursions that it might experience.
I wonder if calibrating the compass without the camera attached, then reattaching the camera has caused some of the flight instability issues that some have experienced.
To test this theory, I plan on connecting the H to the GUI and noting compass readings with and without the camera attached, rotating the aircraft along the yaw axis.
Now, where's my round 'tuit... I'll certainly need that before I do any experiments.

Hi.
This has been discussed previously.
You make a valid point regarding the pros/cons regarding calibration with the camera on or off. Some pilots insist on calibration with the camera off and, as far as I'm aware, don't seem to have any issues with the calibration once they re-attach the camera. Other pilots calibrate with the camera on sighting that the fields generated by the camera will be taken into account by the compass. To me, then, it's all a matter of choice and what you think is good for the aircraft.

Yuneec does not give any advice to remove the camera during calibration. One can assume from that, then, is that the aircraft is meant to be calibrated with the camera attached. But since I've not seen any reports of compass errors resulting from a calibration with the camera off, then I suppose it's all down to the individual's own beliefs. After all, it's a very simple procedure to remove the camera for calibration and re-attach it...so if people are not experiencing problems then what's the harm?

 

[kd4pbs]

After all, it's a very simple procedure to remove the camera for calibration and re-attach it...so if people are not experiencing problems then what's the harm?

Exactly. Unless it causes an issue with Yuneec at some point in time. I could see them citing "Unapproved compass calibration procedure" as reason to deny warranty claims due to a crash, regardless if that was an issue or not, while I couldn't see them easily wiggling out of a warranty claim for messed up gimbals due to calibrating the compass with the camera attached, since their procedure specifically shows us to keep the camera attached while calibrating the compass.
Hopefully I'll get the chance today to test how much, if any deviation there is with it attached and unattached.
Thanks!

 

[kd4pbs]

Here are my findings.

Compass raw value reading
Direction (relative)           With Camera                  Without Camera
                                X     Y     Z               X     Y     Z
0°                            -82    187   299            -79    178   317
60°                            128   169   296             125   163   314
120°                           218  -23    317             224  -26    332
180°                           116  -179   332             113  -194   359
240°                          -94   -170   338            -85   -188   365
300°                          -196   7     326            -196  -11    350

Δ value between camera on and off
      X    Y    Z
0°    3    9    18
60°   3    6    18
120°  6    3    15
180°  3    15   27
240°  9    18   27
300°  0    18   24

I'll note that I only performed this test once, so this data is from one rotation, moving clockwise while looking from above. I did carefully place markers on the desk where I performed this test to positively align the H with camera on and with camera off, and also carefully routed the USB cord in the same manner between the two tests. I also did not move myself (other than my arms to write down the findings and to move the H) nor anything else in the environment between the two tests. The two tests were completed within 5 minutes.

I'll let the reader draw their own conclusions. As for me, I'll keep the camera on when doing a compass alignment.

 

[CraigCam]

Here are my findings.

Compass raw value reading
Direction (relative)           With Camera                  Without Camera
                                X     Y     Z               X     Y     Z
0°                            -82    187   299            -79    178   317
60°                            128   169   296             125   163   314
120°                           218  -23    317             224  -26    332
180°                           116  -179   332             113  -194   359
240°                          -94   -170   338            -85   -188   365
300°                          -196   7     326            -196  -11    350

Δ value between camera on and off
      X    Y    Z
0°    3    9    18
60°   3    6    18
120°  6    3    15
180°  3    15   27
240°  9    18   27
300°  0    18   24

I'll note that I only performed this test once, so this data is from one rotation, moving clockwise while looking from above. I did carefully place markers on the desk where I performed this test to positively align the H with camera on and with camera off, and also carefully routed the USB cord in the same manner between the two tests. I also did not move myself (other than my arms to write down the findings and to move the H) nor anything else in the environment between the two tests. The two tests were completed within 5 minutes.

I'll let the reader draw their own conclusions. As for me, I'll keep the camera on when doing a compass alignment.

I've got a crashed repair drone that's been stubborn on compass. It calibrates and is fine but gets more warnings then my other two. It's carrying a CGO3 currently as I destroyed the + and grabbed my Q500 camera and put it on this one. When I get home, I'm going to try it with the camera attached and see if it improves. Interesting numbers and something to ponder as I'm in the don't like the gimbal motor stress group and have been routinely calibrating without camera.

 

[FlushVision]

Here are my findings.

Compass raw value reading
Direction (relative)           With Camera                  Without Camera
                                X     Y     Z               X     Y     Z
0°                            -82    187   299            -79    178   317
60°                            128   169   296             125   163   314
120°                           218  -23    317             224  -26    332
180°                           116  -179   332             113  -194   359
240°                          -94   -170   338            -85   -188   365
300°                          -196   7     326            -196  -11    350

Δ value between camera on and off
      X    Y    Z
0°    3    9    18
60°   3    6    18
120°  6    3    15
180°  3    15   27
240°  9    18   27
300°  0    18   24

I'll note that I only performed this test once, so this data is from one rotation, moving clockwise while looking from above. I did carefully place markers on the desk where I performed this test to positively align the H with camera on and with camera off, and also carefully routed the USB cord in the same manner between the two tests. I also did not move myself (other than my arms to write down the findings and to move the H) nor anything else in the environment between the two tests. The two tests were completed within 5 minutes.

I'll let the reader draw their own conclusions. As for me, I'll keep the camera on when doing a compass alignment.

Interesting comparison. I acknowledge that this is just from one test under what some would argue not to be under strict scientific conditions (but how else would you do it without a laboratory), but if those figures are representative, then there may be some cause to consider the merits of leaving the camera attached...would you agree?

 

[kd4pbs]

Indeed, this is not in an RF anechoic, non-ferrous, isolated, and rated environment for such a test. That being stated, since we're not looking for absolute values in accuracy, but instead are looking at relative changes between two tests, with the removal of the camera being the only thing different, I'd have to say that in my opinion, the camera makes a difference in the direction of the sensed magnetic field, up to almost 30 degrees on one specific axis. I'd have to think that this can easily make enough of a difference to require one to calibrate without the camera if flying without the camera, and calibrate with the camera if flying with the camera.
I'd guesstimate the possible error to be +/- 2 degrees relative, since I did notice the values bouncing around a bit close to that amount while testing. So, if it were repeated and averaged several times, the values might be close to the same, slightly better, or slightly worse.
Regardless, the noticeable delta between with and without camera were very apparent.
Perhaps this is the cause of some of the "toilet bowl effect" flight characteristics some people have noticed, but certainly not everyone, since I do know that some people mentioned they calibrated without the camera and didn't put the camera back on just to test whether or not it still toilet-bowled.
I do know that mine is flying fine and correctly, calibrating the compass with the camera on and flying the same way. (knock on my wooden head again)

I guess another test would be to pan & tilt the camera in various directions while keeping the H stationary and watching the raw compass readout in the GUI just to see if the direction of the camera has some effect on the compass. If so, and it causes the same magnitude of deviation as the rotation test I did, then it'd be a crap shoot anyway, and wouldn't really matter.

I urge others to try this out just so we could get some more data points on this subject.

 

[richard bottos]

I have just competed my latest upgrade, even though all was working well, no gps errors or such.
Now trying to calibrate compass, I get an immediate gps error on the st16. I had no error condition prior to the upgrade, and repeated the upgrade twice with no problem. I moved the st16 about 20 feet away faced North, but can't get pass the gps fail on pushing the calibrate compass.
I'm fairly familiar with the h having it for almost a year. But this has me stymied. Help from anyone is welcomed.
Regards Dick Bottos