AnsweredAssumed Answered

LIS2DH - wrong Z reading - has something to do with power supply

Question asked by lupal on Jun 15, 2016
Latest reply on May 21, 2017 by Gautam.Mohit
I'm working with LIS2DH on SPI, where also some other SPI slaves are connected. The supply power is 3.0V, VDD and VDD_IO are connected.
I configure it and keep reading the accelerations x,y,z just fine.
But in certain circumstances the z acceleration is wrong or the values are frozen.
I identified the following:
The behavior depends on the power supply cycle.
If I disconnect my device from power supply when the rest voltage on the capacitor did not drop bellow around 0.3V, it will work.
If the device is disconnected for some time, the voltage drops bellow 0.3V. Then the z accelerometer does not work correctly. It reads some 1g instead of -1G and 3G instead of 1G and it seems to be linear. Another device reads some 1.2G instead of -1G and some other value instead of 1G. So it is not the same as in the first case.
When connected when the value is around 0.3V the readings sometimes freeze in all 3 axis.
The mentioned 0.3V does not to be a fixed value, it is somewhere around it and it can change slightly.

I tried change ODR, full scale, run the configuration in a different time, remove the code for other peripherals etc. but it did not help.

Does anyone have a clue what could cause the mentioned problem?
Thank you.