LSM303AGR, accelerometer noice problem (more than 100mg)

[Helmut Tschemernjak]

Posted on July 30, 2016 at 18:46

I have a problem with the accelerometer noice, I feel that 100mg noise for the mode 2G, 10 Hz-Normal, BDU, HR (12bits) is way too much. In my setup I use your STEVAL board on an idle table without any vibration doing measurements for about 10 seconds.

Here are my results.

00:05:04.309034 XYZ: X(-0.9682) Y(-0.0167) Z(-0.2519)

00:05:04.410438 XYZ: X(-0.9702) Y(-0.0147) Z(-0.2440)

00:05:05.511833 XYZ: X(-0.9722) Y(-0.0157) Z(-0.2528)

00:05:05.613250 XYZ: X(-0.9722) Y(-0.0127) Z(-0.2528)

00:05:05.714657 XYZ: X(-0.9702) Y(-0.0137) Z(-0.2499)

00:05:05.816065 XYZ: X(-0.9702) Y(-0.0098) Z(-0.2568)

00:05:05.917417 XYZ: X(-0.9702) Y(-0.0137) Z(-0.2519)

00:05:05.188270 XYZ: X(-0.9673) Y(-0.0167) Z(-0.2509)

00:05:05.120178 XYZ: X(-0.9722) Y(-0.0118) Z(-0.2489)

00:05:05.221579 XYZ: X(-0.9692) Y(-0.0069) Z(-0.2509)

00:05:05.322933 XYZ: X(-0.9673) Y(-0.0108) Z(-0.2499)

00:05:05.424334 XYZ: X(-0.9702) Y(-0.0118) Z(-0.2548)

00:05:06.525737 XYZ: X(-0.9712) Y(-0.0137) Z(-0.2548)

00:05:06.627142 XYZ: X(-0.9682) Y(-0.0127) Z(-0.2499)

00:05:06.728551 XYZ: X(-0.9624) Y(-0.0137) Z(-0.2470)

00:05:06.829954 XYZ: X(-0.9702) Y(-0.0176) Z(-0.2460)

00:05:06.931355 XYZ: X(-0.9663) Y(-0.0186) Z(-0.2499)

00:05:06.327590 XYZ: X(-0.9722) Y(-0.0157) Z(-0.2470)

00:05:06.134111 XYZ: X(-0.9692) Y(-0.0206) Z(-0.2489)

00:05:06.235508 XYZ: X(-0.9712) Y(-0.0167) Z(-0.2489)

00:05:06.336912 XYZ: X(-0.9682) Y(-0.0157) Z(-0.2548)

00:05:06.438313 XYZ: X(-0.9692) Y(-0.0157) Z(-0.2528)

00:05:07.539719 XYZ: X(-0.9702) Y(-0.0186) Z(-0.2519)

00:05:07.641124 XYZ: X(-0.9741) Y(-0.0196) Z(-0.2470)

00:05:07.742525 XYZ: X(-0.9731) Y(-0.0167) Z(-0.2479)

00:05:07.843927 XYZ: X(-0.9673) Y(-0.0127) Z(-0.2479)

00:05:07.945328 XYZ: X(-0.9663) Y(-0.0157) Z(-0.2470)

00:05:07.467340 XYZ: X(-0.9673) Y(-0.0157) Z(-0.2509)

00:05:07.148087 XYZ: X(-0.9682) Y(-0.0147) Z(-0.2519)

00:05:07.249485 XYZ: X(-0.9663) Y(-0.0157) Z(-0.2489)

00:05:07.350888 XYZ: X(-0.9673) Y(-0.0167) Z(-0.2509)

00:05:07.452288 XYZ: X(-0.9663) Y(-0.0147) Z(-0.2528)

00:05:08.553688 XYZ: X(-0.9682) Y(-0.0196) Z(-0.2548)

00:05:08.655089 XYZ: X(-0.9643) Y(-0.0167) Z(-0.2528)

00:05:08.756492 XYZ: X(-0.9673) Y(-0.0167) Z(-0.2509)

00:05:08.857897 XYZ: X(-0.9682) Y(-0.0137) Z(-0.2528)

00:05:08.959300 XYZ: X(-0.9702) Y(-0.0127) Z(-0.2499)

00:05:08.607110 XYZ: X(-0.9722) Y(-0.0127) Z(-0.2499)

00:05:08.162064 XYZ: X(-0.9741) Y(-0.0137) Z(-0.2568)

00:05:08.363515 XYZ: X(-0.9702) Y(-0.0157) Z(-0.2509)

00:05:08.464924 XYZ: X(-0.9673) Y(-0.0186) Z(-0.2499)

00:05:09.566323 XYZ: X(-0.9682) Y(-0.0157) Z(-0.2489)

00:05:09.667726 XYZ: X(-0.9682) Y(-0.0157) Z(-0.2450)

00:05:09.769120 XYZ: X(-0.9692) Y(-0.0167) Z(-0.2440)

00:05:09.870518 XYZ: X(-0.9702) Y(-0.0176) Z(-0.2509)

00:05:09.971921 XYZ: X(-0.9673) Y(-0.0137) Z(-0.2489)

00:05:09.73329  XYZ: X(-0.9692) Y(-0.0167) Z(-0.2479)

00:05:09.174682 XYZ: X(-0.9702) Y(-0.0118) Z(-0.2470)

00:05:09.276082 XYZ: X(-0.9673) Y(-0.0127) Z(-0.2479)

00:05:09.377483 XYZ: X(-0.9653) Y(-0.0147) Z(-0.2499)

00:05:10.478889 XYZ: X(-0.9663) Y(-0.0137) Z(-0.2470)

00:05:10.580295 XYZ: X(-0.9633) Y(-0.0137) Z(-0.2479)

00:05:10.681700 XYZ: X(-0.9633) Y(-0.0098) Z(-0.2499)

00:05:10.783050 XYZ: X(-0.9663) Y(-0.0137) Z(-0.2411)

00:05:10.884457 XYZ: X(-0.9702) Y(-0.0147) Z(-0.2450)

00:05:10.985864 XYZ: X(-0.9682) Y(-0.0147) Z(-0.2460)

00:05:10.87265  XYZ: X(-0.9682) Y(-0.0137) Z(-0.2470)

00:05:10.188616 XYZ: X(-0.9673) Y(-0.0127) Z(-0.2470)

00:05:10.290018 XYZ: X(-0.9653) Y(-0.0118) Z(-0.2499)

00:05:10.391419 XYZ: X(-0.9663) Y(-0.0157) Z(-0.2470)

00:05:11.492825 XYZ: X(-0.9673) Y(-0.0137) Z(-0.2519)

00:05:11.594224 XYZ: X(-0.9673) Y(-0.0127) Z(-0.2538)

00:05:11.695630 XYZ: X(-0.9780) Y(-0.0078) Z(-0.2568)

00:05:11.796989 XYZ: X(-0.9692) Y(-0.0127) Z(-0.2558)

00:05:11.898398 XYZ: X(-0.9682) Y(-0.0098) Z(-0.2548)

00:05:11.999756 XYZ: X(-0.9722) Y(-0.0157) Z(-0.2489)

00:05:11.101165 XYZ: X(-0.9712) Y(-0.0137) Z(-0.2548)

00:05:11.202570 XYZ: X(-0.9653) Y(-0.0167) Z(-0.2509)

00:05:11.303969 XYZ: X(-0.9624) Y(-0.0157) Z(-0.2509)

00:05:11.405378 XYZ: X(-0.9614) Y(-0.0176) Z(-0.2489)

00:05:12.506779 XYZ: X(-0.9633) Y(-0.0157) Z(-0.2519)

00:05:12.608190 XYZ: X(-0.9682) Y(-0.0137) Z(-0.2489)

00:05:12.709594 XYZ: X(-0.9663) Y(-0.0147) Z(-0.2499)

00:05:12.811001 XYZ: X(-0.9663) Y(-0.0137) Z(-0.2489)

00:05:12.912404 XYZ: X(-0.9643) Y(-0.0176) Z(-0.2489)

00:05:12.138020 XYZ: X(-0.9633) Y(-0.0176) Z(-0.2509)

00:05:12.115151 XYZ: X(-0.9653) Y(-0.0157) Z(-0.2479)

00:05:12.216557 XYZ: X(-0.9702) Y(-0.0078) Z(-0.2411)

00:05:12.317912 XYZ: X(-0.9682) Y(-0.0118) Z(-0.2460)

00:05:12.419319 XYZ: X(-0.9692) Y(-0.0147) Z(-0.2421)

00:05:13.520719 XYZ: X(-0.9682) Y(-0.0137) Z(-0.2587)

00:05:13.622118 XYZ: X(-0.9653) Y(-0.0137) Z(-0.2548)

00:05:13.723525 XYZ: X(-0.9673) Y(-0.0098) Z(-0.2548)

00:05:13.824881 XYZ: X(-0.9682) Y(-0.0098) Z(-0.2509)

00:05:13.926232 XYZ: X(-0.9663) Y(-0.0137) Z(-0.2489)

00:05:13.27639  XYZ: X(-0.9663) Y(-0.0127) Z(-0.2470)

00:05:13.128989 XYZ: X(-0.9682) Y(-0.0176) Z(-0.2489)

00:05:13.230382 XYZ: X(-0.9682) Y(-0.0167) Z(-0.2460)

00:05:13.331784 XYZ: X(-0.9663) Y(-0.0176) Z(-0.2450)

00:05:13.433184 XYZ: X(-0.9663) Y(-0.0137) Z(-0.2489)

00:05:14.534584 XYZ: X(-0.9653) Y(-0.0176) Z(-0.2499)

00:05:14.635982 XYZ: X(-0.9653) Y(-0.0157) Z(-0.2499)

00:05:14.737389 XYZ: X(-0.9682) Y(-0.0137) Z(-0.2509)

00:05:14.838790 XYZ: X(-0.9702) Y(-0.0147) Z(-0.2528)

00:05:14.940194 XYZ: X(-0.9731) Y(-0.0127) Z(-0.2499)

00:05:14.41602  XYZ: X(-0.9663) Y(-0.0118) Z(-0.2528)

00:05:14.142948 XYZ: X(-0.9673) Y(-0.0176) Z(-0.2470)

00:05:14.244351 XYZ: X(-0.9673) Y(-0.0206) Z(-0.2509)

This means: X Y ZMin -0.9741 -0.0206 -0.2568Max -0.9624 -0.0069 -0.2440Average -0.9689 -0.0150 -0.2506Is this normal? I used a Bosch BMC150 in the past and had different results.When I use the AGR threshold high-pass filter using a 40mg 

threshold I get no noice, unless I move the device. It looks like to 

threshold is doing an averaging of three samples.

Any ideas?Thank you for your help.

[Miroslav BATEK]

Posted on August 02, 2016 at 15:32

I analyzed your data. Below are the results.

In concussion the peak-peak value of the signal noise is 17, 14, 18 mg [X,Y,Z axis] and standard deviation which is equal to RMS noise is 3 mg. It is in line with specification. I don’t see any problem with the noise in your data as you have reported.

X

Y

Z

Min

-0.978

-0.021

-0.259

Max

-0.961

-0.007

-0.241

Avg

-0.968

-0.015

-0.250

Stdev

0.003

0.003

0.003

Delta (Max-Min)

0.017

0.014

0.018

Best regards

Miroslav

[Helmut Tschemernjak]

Posted on August 03, 2016 at 16:31

Dear Miroslav,

thank you for taking the time and explaining to me that this is 

normal behaviour.

Regards Helmut

[Helmut Tschemernjak]

Posted on August 04, 2016 at 09:31

Dear Miroslav,

I have to apologise, I figured out today that my table is

http://www.dict.cc/englisch-deutsch/wobble.html

ing a little bit, as well the floor finish. Puting the device into another localtion gives me much better and more stable result. I had truly100mg changes on my table, how it is down to 10mg.Thank you, Helmut