Strange Flight

[slup]

...magYaw is computed by DatCon and depends only on the 3 magnetometers and the roll and pitch data. I've never seen anything like this.

In what way correlates the TotalGyroZ & the VIOYaw to magYaw ... as they all seems to be calculated, does they all depend on the same parameters?

Another thing to note about this flight is that it was launched from a geomagnetically distorted location.

Yep, saw that & that was most probably the reason for the compass errors just before landing. Also seems as the compass was calibrated before take-off if looking on the periodic yawing there. Also the IMU & magYaw didn't have the same values when the graphs start up ... some lag before the recording started onto the SD card?

 

[BudWalker]

Interesting observation about the geomagnetically distortion, I referenced the material that is on floor of this parking garage I launch from being some weird stuff. Yesterday I calibrated the compass before launch and hand launched and didn't have any issues at all. Of course I also re-calibrated the IMU the day before so I don't know which one helped the most.

The compass calibration doesn't fix a geomagnetic distortion. You could achieve the same result by holding the MM above the distortion (1.3 m) for the time required to do a calibration. Then put on the ground and launch before the Yaw can adjust to the incorrect Yaw derived from the magnetometer data.

 

[slup]

Interesting observation about the geomagnetically distortion, I referenced the material that is on floor of this parking garage I launch from being some weird stuff. Yesterday I calibrated the compass before launch and hand launched and didn't have any issues at all. Of course I also re-calibrated the IMU the day before so I don't know which one helped the most.

The risky phase is when the Mini is powered on ... if the compass then is located in magnetic interference which deflects it ... let's say by fully 180 degrees, the IMU will be initialized to a direction that is wrong by 180 degrees.

This mainly tells the IMU that the Mini's front points in the direction that the Mini's rear does in reality.

So as soon as the Mini needs to counter a positional drift due to wind ... it will do that with the wrong motors & the positional error will grow instead of getting less & the IMU will rapidly try over & over & the Mini speeds away uncontrollably.

But if you instead power on the Mini in your hand away from mag interference ... & then put it on ground with mag interference & take off ... it will be no drama at all.

 

[BudWalker]

In what way correlates the TotalGyroZ & the VIOYaw to magYaw ... as they all seems to be calculated, does they all depend on the same parameters?


Yep, saw that & that was most probably the reason for the compass errors just before landing. Also seems as the compass was calibrated before take-off if looking on the periodic yawing there. Also the IMU & magYaw didn't have the same values when the graphs start up ... some lag before the recording started onto the SD card?

TotalGyroZ is time integrated gyroZ - so only depends on gyroZ data. VIOYaw is derived from the quaternion in the VIO record. I suspect the quaternion may have some dependency on the flight controller's Yaw data but don't actually know.

Also the IMU & magYaw didn't have the same values when the graphs start up
Could you be more specific

 

[sar104]

TotalGyroZ is time integrated gyroZ - so only depends on gyroZ data. VIOYaw is derived from the quaternion in the VIO record. I suspect the quaternion may have some dependency on the flight controller's Yaw data but don't actually know.

And it's worth explicitly noting that yaw, or more specifically unwrapped heading, depends on both gyroX and gyroY unless pitch and roll are zero everywhere.

 

[slup]

Also the IMU & magYaw didn't have the same values when the graphs start up
...Could you be more specific

I mean here just when values start to show up in this flight ... IMU & magYaw doesn't have near the same values & TotalGyroZ isn't 0.

But ... aha, this perhaps wasn't the power on ... I mean as you extracted this incident flight into your .csv?

 

[BudWalker]

I mean here just when values start to show up in this flight ... IMU & magYaw doesn't have near the same values & TotalGyroZ isn't 0.

But ... aha, this perhaps wasn't the power on ... I mean as you extracted this incident flight into your .csv?

View attachment 143726

The structure of the fc_log.log is that .DATs are placed randomly (at least to me) in the 512 Mb allocated to the fc_log.log. In this case the first part of this .DAT was over written by a subsequent .DAT. I haven't looked closely but I suspect this caused the anomalies you observed.

 

[umanbean]

The place where I launch from work is the top (3rd floor) of a parking garage. Unlike when I fly at home or other places I get a "Need compass calibration" warning EVERY time I fly. So i calibrate the compass

I referenced the material that is on floor of this parking garage I launch from being some weird stuff.

Problems arising from powering up with the Drone sitting on reinforced concrete are more likely to be caused by the steel reinforcing rebar inside the concrete rather than any kind of surface coating. It's very likely your compass did not actually need recalibrating, it just needs to be powered on away from steel rebar or any other ferrous metal.

That's why you get the 'needs calibration' warning EVERY time you fly from there.