LIS3DSH Averaging Filter

[Paul Mellor]

Posted on August 01, 2016 at 15:13

Does anyone know if the LIS3DSH average samples when running at lower data rates? E.g. if the ODR is 100 Hz, will it sample at a higher rate and average the results (in order for example to reduce noise)? Or perhaps the nature of the method used to determine the acceleration in the first place will naturally average the data since the last sample?

Thanks!

Paul.

#accelerometer #filter

[Miroslav BATEK]

Posted on August 02, 2016 at 15:57

No, there is not such an averaging as you described in your post. 

You have to calculate the average in your application.

Best regards

Miroslav

[Paul Mellor]

Posted on August 11, 2016 at 21:06

Miroslav,

To be honest that does not seem to tie up with my own experiments. If I sense vibration at 1600Hz ODR and then at say 100 Hz ODR I feel I am seeing an averaging of the data at 100 Hz ODR.

If you look at Invensense ICM-20601 for example the datasheet gives details of the 'time on', i.e. the sample time of the chip, and also details of the number of such samples that will be averaged as part of the filtering, and also how the samples will be decimated to provide the selected ODR. I was wondering if such info was available for the LIS3DSH?

The state machines are fantastically useful by the way - really got good use from them!

Thanks,

Paul.

[Miroslav BATEK]

Posted on August 24, 2016 at 17:36

I’m sorry, I was wrong. I double check the functionality of LIS3DSH and it automatically does the averaging. It is not possible to change the number of samples. 

Unfortunately I can’t share the number of samples and timing on the public forum. You would need to contact your local sales representative to receive more details. 

In any case the bandwidth declared in datasheet takes in to account the effect of the averaging.

Best regards

Miroslav

[Paul Mellor]

Posted on August 26, 2016 at 13:34

Well I can't say I understand why you can't share the specification of your own product with the people who are using it... but anyway thanks for the info I will follow up as you suggest.