Accelerometer with a reading error on the Z-axis

[anaportela]

Hello, I would like some support in understanding an issue that occurred with a product incorporating the IIS3DWB component.

 

The IoT sensor transmits global data with constant acceleration values ​​subtracted. These global measurements should not be reflecting gravity. Upon examining the raw “spectral” data—prior to any processing—it becomes clear that the value of ~1g is merely a coincidence. In reality, we are dealing with a faulty accelerometer.

What could be causing this failure, even when the sensor is stationary on a table?

 

I would like support from a specialized engineer.

Thanks.

[Andrew Neil]

I would like support from a specialized engineer.

As ​@Federica Bossi  advised in your previous thread, if you want direct support you need to raise a support case direct with ST’s online support.

[ozammo]

Hi Anaportela,

 

I am not a specialised Accelerometer Engineer however allow me to ask you a few questions to better frame the question for when the right person comes along! I have a moderate amount of experience with Accelerometers but I have alot of experience designing PCBs and housing to withstand different resonances as well as the processing of high frequency domains. 

1. Can you describe the enviroment this sensor is in?
2. What full-scale range is the sensor set to?
3. Are there any fans, motors, or vibrating machine running on or near the mounted location?
4. Is there a DC-DC switching power supply or a wireless antenna really close to the Accelerometer on the PCB?
5. Is the PCB inside housing?
6. Does the Z-axis stepping stop if you pick the sensor up in your hand or place it on a foam pad?

I appreciate that is alot of questions but understanding the enviroment is incredibly useful!

 

Kind regards,

Oz

[anaportela]

Hello ozammo,

In response to your questions:

1 - Our sensors are installed on assets in industrial environments, such as motors, pumps, gearboxes, compressors, steam turbines, and generators.
2 - The value is 16g (acceleration due to gravity).
3 - This issue occurred with a geared motor at a client’s site, which is a very noisy environment. When the unit was returned for replacement, the fault persisted even while it was sitting on an office desk.
4 - Yes. The circuit uses a u-blox NINA module for wireless (Bluetooth) communication, which is positioned directly above the accelerometer on the PCB.
5 - Yes, it is enclosed in a polyacetal housing.
6 - I left it sitting on a sponge for several hours, but the issue did not disappear; it persists.

[ozammo]

Interesting, it does sound like an event has caused permenant damage to the sensor. Here is some ideas:

 

1. Does the error persist if the NINA module is removed or disabled?
2. Is the product vibration tested to relevant standards to ensure resonance frequencies does not impede functionality?
3. Are you certain that none of the maximum limits of the sensor has been breached? While the 10,000g max allowable limit sounds high, I would not be suprised if the housing amplified the motors movements and breached the limit.