How to convert the value read by LIS3DH

[Nats]

Hallo everyone,

I have interfaced LIS3DH with nRF52832 and reading a value as below

X    230

Y     249

Z     54

How to convert this value to standard acceleration unit?

(im using 100Hz sample rate and 2g scale.)

[hmolesworth]

Actually sometimes confusing, so maybe this would be of help. The values on the LIS2DH are signed but held in 2 x 8-bit registers which have to be handled with correct shift and sign-extension. Have a look at this (working) code to see if that helps:

typedef struct {
   uint8_t OUT_d_L;       // LIS2DH12_OUT_XYX_L
   uint8_t OUT_d_H;       // LIS2DH12_OUT_XYZ_H
}__attribute__((packed)) axis_t;

typedef union {
   axis_t   Bytes;       // LIS2DH12_OUT_X_L, H           0x28, 0x29
   int16_t  sWord;
}__attribute__((packed)) axisUn_t;

typedef struct {
   axisUn_t rOUT_X;        // LIS2DH12_OUT_X_L,H         0x28,29
   axisUn_t rOUT_Y;        // LIS2DH12_OUT_Y_L,H         0x2A,2B
   axisUn_t rOUT_Z;        // LIS2DH12_OUT_Z_L,H         0x2C,2D
}__attribute__((packed)) reg2_t;
STATIC_ASSERT((offsetof(reg2_t, rOUT_X) & 0x01) == 0);

reg2_t Sea2; // Sea of byte registers
int16_t x, y, z;

/**
 * @brief Collect signed xyz registers and correctly align for signed 16-bit
 * @param
 */
void ReadXYZ(void)
{
    ReadAccRegisters(); // <== this is your code to read the 6 consecutive 8-bit registers  forming X, Y and Z
    // Ensure we have correct sign extension, also assuming even byte alignment
    x = Sea2.rOUT_X.sWord / 16;
    y = Sea2.rOUT_Y.sWord / 16;
    z = Sea2.rOUT_Z.sWord / 16;
}

Some care is required on alignment, since 8-bit values (packed, or cuddled up) must be sent to even-byte aligned 16-bit signed result registers. You can see from the byte addresses on the LIS2DH12  that they are sequential (cuddled up). The LIS3DH is similar.

Edit: Just noticed this is 5 months old, <sigh>

[Karl Ylvisaker]

Hello @hmolesworth,

Indeed this case is 5 months old, but you reply should certainly prove useful to others coming here with a similar issue in the future!

Best regards,
Karl