LIS331 output is not stable

[DJohn.6]

Hi Guys,

I encountered a very strange problem and hope you can give me some suggestions!

The LIS331 is adopte to measure device tilt, we assembled the PCBA in vertical positon and this means the sensor also works at vertical status. Please see the attached picture!

While we do test, we found that Z-axis output value will decrease step by step. Howerver the X-axis is quite stable. please see test data above. The test lasted about 3 hours.

Could you please tell me the reason?

Thank you in advance!

Best Regards,

John

[Eleon BORLINI]

Hi John, in order to understand if it's a device or setup related issue... did you check if the setup is stable (e.g. with a reference inclinometer, or rotating the pcb and getting the same values)? Is the temperature stable in the room? Is the X-axis stable? Looks like a ZGL drift...

Btw if you target application is to measure the tilt angle, my suggestion is to use the IIS3DHHC device, which is dedicated exactly to this purpose.

Regards

[DJohn.6]

​Hi Eleon,

Thanks for your response!

Not only the device on my desk has this issue, many customer also feed back that has the same issue. So I don't think it a set up stability problem.

Is there a chip that feature likes the IIS3DHHC, but the cost is better?

Best Regards,

John

[Eleon BORLINI]

​Hi John, the IIS2DH is a valuable alternative solution that should cost less. Of course, accuracy will not as good as in the IIS3DHHC case... Regards

[DJohn.6]

Dear Eleon,

Thanks for your recommendation! The new devices will adopt the sensor your suggested!

Is there a way from software side to do compensation for ZGL (means:zero g level drift,right?)?

We sold and delivered many devices to different countries. It is difficult for us to call back all of them, whatever cost or time.

Best Regards,

John

[Eleon BORLINI]

​Hi John, if the drift is device-related, it could be due to the temperature or the mechanics (maybe MEMS-related). If it is a temperature drift, you can apply by SW the datasheet Zero-g level change vs temperature, i.e. 0.1mg/°C. If it is a mechanical drift, you should check if it's repeatable. if so, you could apply a negative gain vs time in post processing. Btw, which are the measure units of you excel file? Regards