We have two prototype boards each equipped with a LIS331HH.

One of the sensors seems completely broken because the Z axis shows values around 2.3g while laying flat on the table. Because of that we were now looking into self tests to check all sensors in production.

We query all the axis and divide each value by 16. The data sheet tells us that 1digit is 3mg so we expect 333 on the Z axis and 0 on X and Y. We measure X:20, Y: 35 and Z: 373, which is an offset we expect [1].

Now we start the self test in plus mode and measure X: 177, Y: 131 and Z: 828.

The self test values differ by X: 157, Y: 96 and Z: 455 from the non self test values.

In the data sheet there are values for "self-test output change". Those mean, that X and Y are supposed to differ by a max of 180 and that Z may only differ by 370?

So this sensor passes the test on X and Y and fails on Z and therefore we have to throw this sensor away, right? Or did we misunderstand the process and did it wrong?

Thanks in advance

Daniel

Yes, your understanding is completely correct.

Self-test output change in Z axis should be 220-370 LSB, if the output is 455 the sensor is broken.