LSM9DS1 acceleration anti-aliasing filter

[giannisroussos0]

Posted on June 24, 2016 at 11:23

Hi all,

I have some questions regarding the LSM9DS1: 1. The Acceleration ODR choices go down to 10 Hz, however the bandwidth selection, presumable an anti-aliasing filter, is limited to 50Hz. Thus, when sampling at 10Hz or 50Hz with a bandwidth selection of 50Hz severe aliasing is to be expected. Is there a way to avoid this? 2. The accelerometed digital filter operation is not documented properly. What dow ''high resolution'' refer to? How is LPF or HPF selected? 3. Is there a block diagram for the acceleration signal available, similar to Figure 28 in the datasheet for the gyro signals?

[jacquestkirk]

Posted on July 05, 2016 at 05:51

I also have the same question regarding the anti-aliasing filter. Maybe the ODR is not the same as the sampling rate of the ADC?

I haven't tried it yet, but if the ODR is the same as the ADC sampling rate, then maybe we can use a higher data rate and decimation. It seems kind of silly that we have to do this though. But for questions 2 and 3, see figure 8.

[Petr S]

Posted on July 20, 2016 at 10:23

1)

     Hello all 

Hello all,The minimum possible bandwidth for XL analog filter is 50Hz, so no way to avoid aliasing possibility for the ODRs at 10Hz and 50Hz.

Anyway, to avoid the aliasing and limit the BW, it could be possible to go at higher ODR (e.g. 200Hz) and then add a digital external filter (averaging data).

2)

LPF and HPF selection is described in Figure 8 of the DS: it depends on HR, HPIS1 (and FDS) bits. Cut-off frequency is reported in below table:

HR

DCF [1:0]

Cut-off [Hz]

Filtering Mode

1

00

ODR/50

LP

1

01

ODR/100

LP

1

10

ODR/9

LP

1

11

ODR/400

LP

0

00

ODR/50

HP

0

01

ODR/100

HP

0

10

ODR/9

HP

0

11

ODR/400

HP

Best regards,Petr