From the data sheet of the ADXRS150 MEMS gyro we calculated that a change in
the the gyro temperature of 0.01degrees Celsius will change the gyro drift by
about 10 degrees/hour. Is this correct?. To what level can I rely on the
internal temperature sensor? Do you have information regarding optimal
modelling of the temperature dependence of gyro drift and what are the best
results which can be achieved? We require drifts significantly lower than the
300 degrees/hour mentioned in the data sheet.
Integrating the gyro output over time allows errors to accumulate and grow.
All gyros experience this effect. It is usually referred to as Null Stability,
and expressed in degrees/hour.
There are 2 sources of error that impact null stability over time
· Null stability over temperature
· Allen variance
Null drift due to temperature is the dominant mechanism
· A 3 point temperature compensation scheme will give you about 300
degrees/hour null stability. More points will do better.
Allen variance is an expression of the average over the sum of the squares of
the differences between successive readings of the null output sampled over the
· ADXRS150/300 Allen variance settles to about 75 degrees/hour