lsm330dl accelerometer output data problem

mengsophia · ‎2012-07-02

Posted on July 03, 2012 at 06:04

i'm using the accelerometer in lsm330dl. but when i put the chip statically on the table, the xyz output i got from it is very strange,

for x, it's always between 0.2-0.45

for y, it's always around 1

for z, it's always around 16

why? i didn't enable any interrupt, is there something wrong with my register setting?

#lsm330dl #accelerometer

mengsophia · ‎2012-07-02

Posted on July 03, 2012 at 07:39

here are some part of my code,

________________

Attachments :

7.3.txt : https://st--c.eu10.content.force.com/sfc/dist/version/download/?oid=00Db0000000YtG6&ids=0680X000006I0yi&d=%2Fa%2F0X0000000biH%2F7gx3nQuTYg0yGyV.YHx9PtjQRpK4PnfYqrkewRjipV8&asPdf=false

frankmeyer9 · ‎2012-07-02

Posted on July 03, 2012 at 08:19

...,the xyz output i got from it is very strange,

for x, it's always between 0.2-0.45 for y, it's always around 1 for z, it's always around 16 I guess you mean the calculated xyz values, using float. Could you post additionally some ''raw'' acceleromete values ? Your code contains :

ACCEL_Xf = ACCEL_X * 0.002 - ACCEL_Xbias; //divide the ACCEL values by 70 to get degrees per second

You copied that probably from the gyro code. Accelerometer output is not in deg. per second.

mengsophia · ‎2012-07-03

Posted on July 04, 2012 at 05:25

sorry, i change the code, but forget to clear the comment.

i changed some register setting values, and uncomment the register 6, and the output seems meaningful.

but the z axis always show me 16, and when i flip it over, it shows me -16, i didn't find any clues in the datasheet that i should divided by 16.

ACCEL_Zf = ACCEL_Z * 0.002/16.0 ( FS bit set to 01 , 2mg/digit)

when i 've done this, the output value looks corret, ie, when i turn it 60degrees, it shows me 0.5g, and all the axies are correct. so strange.

frankmeyer9 · ‎2012-07-03

Posted on July 04, 2012 at 08:29

Admittedly I did not analyze your code in detail - I have no lsm330 to try anything.

But the results are generally meaningful. A significant z-axis output that turns negative when flipping over seems ok. When you tilt the board by 90 deg. in any direction, the corresponding axis should show an equivalent output.

I have read comments in other forums that describe the lsm330 as having a somehow noisy output, so the fluctuations of the other axes might be normal. As mentioned, I don't have any, so I can neither confirm nor refute.

The scaling is another problem, which should not be too difficult to solve if the readout code is working.

frankmeyer9 · ‎2012-07-04

Posted on July 04, 2012 at 15:22

The only thing I found was this source line, or better, definition:

#define ACCEL_VALUE_CTRL_REG4 0x98 //DISCONTINUES UPDATE#F0, LITTLE ENDIAN. 250 DPS, 4-WIRE

The value is reasonable (BDU, +-4g full-scale, 4-wire interface), but the comment is misleading and wrong. I assume the comment belongs to the CTRL_REG4_G value for the gyro part.

VALUE_CTRL_REG1 0x97 //ODR 1.344KHZ,

This is quite a high update rate. It makes sense for detection of rather fast movements.

mengsophia · ‎2012-07-07

Posted on July 07, 2012 at 15:39

thanks for suggestions.