Sensor zero state

Lennart

Hello everyone,

I am using the MetaMotionR. While it is lying completely still, the gyrometer shows ~0.6 (x), ~-0.2 (y) and ~-0.6 (z) degrees/second. Is this normal? It seems like a pretty strong drift to me, seeing that other gyrometers report something like 0.1°/min; while these values result in a full turn in 10 minutes without any movement.

If this is normal, is there anything I can do to calibrate the gyrometer? The calibration described seems to be connected to sensor fusion, not to the gyrometer.

Thanks!
Lennart

Laura

Hi Lennart,

This is not correct and should not be so.

Please see if you can replicate this and if you have another sensors, please take data from it and get back to us.

Lennart

Hello Laura,

thanks for your answer. I attached a file of gyroscope data during which the sensor was not moved (lying flat on a table). This is reproducible for different days and different locations. We sadly only have one MetaMotionR and thus no other sensors available to double check.

There is also a constant value for x (0.026 g) and y (0.009 g) for the accelerometer for the same experiment, sensor not moving at all; though I understand it is possible that it is not lying completely parallel to g, so that is not as strange.

Is it possible that the sensor is defect? Or is this regular sensor bias?

Laura

The data is constant so that looks good.
Two things:
1. Is the surface truly flat?
2. Is this board in a case?

Lennart

It's a table - though there is possibly a very slight dip as confirmed by a level. Is that possibly a reason?

Yeah, the board is in a case.

Laura

To me the sensor is clearly working because the values aren't changing. So in terms of stabilitity and readings, it looks fine.
More likely the sensor is not flat in the case nor is it on a flat table. You can take the sensor out of the case carefully and re-glue it in (please be VERY careful if you decide to do this). OR you can just take 0 readings and calibrate against it (software calibration).