Estimating random walk of ISM330DLC

[APfel.1]

Hi,

I am having uncertainties when calculating displacement and change of angle effected by random walk.

I want to know proportion of failure in measurement caused by random walk.

In the datasheet of ISM330DLC random walk (ug/sqrt(h) is declared as noise density (ug/sqrt(Hz)).

I found different approaches in calculating effect of random walk effecting displacement and got confused.

Example:

noise density: 75 ug/sqrt(Hz) -> converted to ARW: 1.25 ug/sqrt(h)

displaceent after 1 min

Approach one (which I can imagine is correct):

s = 0.5*9.82(ARW * sqrt(t(h)))*t^2

= 0.5*9.81(1,25 ug/sqrt(h)*sqrt(1/60))*60^2 = 2.85 mm

Approach two I found:

s = 9.81* noise density *sqrt(t) *t

s = 9.81*(75 ug/sqrt(Hz) * sqrt(60))*60 = 0.341 m

Can anyone help and/or does have some sources with useful information?

Thanks a lot!!!

[Eleon BORLINI]

Hi @APfel.1​ , regarding your sentence

>> In the datasheet of ISM330DLC random walk (ug/sqrt(h) is declared as noise density (ug/sqrt(Hz)).

Well, the ARW is only a component of the total device noise monitored in time, and it is not declared in datasheet. I can share you some characterization values for the ARW: avg on roll axis [4.37mdps/sqrt(Hz)], avg on pitch axis [4.28mdps/sqrt(Hz)], avg on yaw axis [4.55mdps/sqrt(Hz)].

You may however refer to the velocity random walk, the VRM, calculated on the accelerometer, whose values are: avg on x axis [49.7ug/sqrt(Hz)], avg on y axis [46.4ug/sqrt(Hz)], avg on z axis [53.9ug/sqrt(Hz)]

Regarding which of the two approaches should be used for the calculation of the displacement from the noise, I believe the one showing the conversion formula from acceleration to space is the correct one:

• s = 1/2*a*t^2.

Regards