Generally 7 seems working. I think more sensors are reported by system. However, they were not fixed. I didn't report it before, hoping BV known about the issue and will fix it.
First, the magnetic field sensor.
![Screenshot_20170220-111153[1].png](./download/file.php?id=1556&sid=852d71f65ff240599de14f3ea6d9f30a)
- Magnetic field screenshot
- Screenshot_20170220-111153[1].png (103.12 KiB) Viewed 4849 times
Above is the readout from the sensor. I waved the calibration pattern ant that changed rest values in the same position significantly. There are two bad things:
- every vawing rest values my be different
- there are drops and series of spikes on the Z axis visible also on the X axis
BV6000 can use significant amount of electricity: probably magnetic fields from power lines influence the magnetic field sensor.
Similarly, with Earth magnetic field sensor:
![Screenshot_20170220-112340[1].png](./download/file.php?id=1557&sid=852d71f65ff240599de14f3ea6d9f30a)
- Earth magnetic field sensor
- Screenshot_20170220-112340[1].png (143.21 KiB) Viewed 4849 times
Three times the phone was waived while data were being drawn. Earlier magentic field sensor had been watched and the phone was waived too. Each time the phone had been aligned with the edge of the table - so the orientation was always he same. Readout - did not.
I wander if ability to disable these sensors in system (by the user) wouldn't be good... To avoid bad readouts by programs that could use another metod of setting direction (like GPS position change). It doesn't seem to be fixable by software.
On the other way, the gyroscope could be repaired. Accelerometer might by be recalibrated. There is no such option for the gyroscope. Here are the readouts at rest:
![Screenshot_20170220-110943[1].png](./download/file.php?id=1558&sid=852d71f65ff240599de14f3ea6d9f30a)
- Gyroscope
- Screenshot_20170220-110943[1].png (95.05 KiB) Viewed 4849 times
This should not be "Gyrosope" sensor but "Gyroscope uncalibrated". Additionally, the ""Gyroscope uncalibrated" should provide drift data, which, subtracted from the readouot, should give (especially when filtered with low-pass filter) zero. "Gyroscope" should be compensated, showing rest at rest. Now it isn't.