ISM330DLC misalignment measurement result

[deddy11]

Hi, I use the ISM330DLC to measure the tilting of an excavator arm (similar to an inclinometer). I have observed that some ISM330DLC accelerometer sensors exhibit misalignment in their Z-axis measurements. For example, while the X and Y axes produce values ranging from -1000 to 1000 mg, the Z-axis measurements range from -500 to 1500 mg.

To compensate for this misalignment, I apply an offset variable. However, the issue is that some sensors initially show good alignment when tested in my office, but after being installed on excavators, they develop misalignment over time, similar to what I described above.

My questions are:

1. What could be causing this misalignment to appear over time?
2. What steps can I take to prevent this from happening?

Thanks

[Federica Bossi]

Hi @deddy11 ,

The misalignment of the Z-axis measurements in your ISM330DLC accelerometer sensors could be caused by several factors, for example the mechanical stresses that an excavator arm is subjected to or the temperature variations that can affect the sensor's performance and cause drift in the measurements. The sensor might behave differently in the controlled environment of your office compared to the field conditions.

Also, the way the sensor is mounted on the excavator arm can affect its alignment. Even slight differences in the mounting angle or surface can lead to misalignment.

I suggest you protecting the sensor from extreme temperatures, moisture, and dust. Use enclosures or protective coatings to shield the sensor from harsh environmental conditions.

You can use temperature compensation techniques to account for the effects of temperature changes on the sensor's performance or implement a regular calibration routine for the sensors. This can help to identify and correct any drift or misalignment that develops over time. You can use known reference positions to recalibrate the sensor periodically.

[deddy11]

Hi Feredica, thanks for you reply.
I don’t think the mounting position is the issue, as this concerns the measurement range rather than the instantaneous values. I determine the measurement range by rotating the sensor 360° along each axis.

And interestingly, the misalignment issue only seems to affect the Z-axis. What are your thoughts on this?