My customer is using ADXL355 and was getting poor results so decided to make a self test. The sensor still works but the noise floor seems much higher than what we get from the eval board. We can only measure larger vibrations. We have looked at the spec again and made few modifications in our program by adding delay between pin activations for self test.
Two different type of setups:
1. the ADXL355 evaluation board is connected to the MCU via cables;
The result is calculated by taking the difference in readings when ST1 is enable and ST2 goes from high to low. The eval board is consistently giving similar results with different delays between activation of ST1 and ST2.
For 0.3 g expected value is: 78643. The result results we got were: 93855, 93314, 94130, 93371.
For 1.5g expected value is: 393216. The results we got were: 341890, 342291. -
2. the sensor is on PCB:
For 0.3 g we read: 87000, 172760, 102311, 116311.
For 1.5g we got 385000 and 342291.
Is there an acceptable threshold for the self test results? We suspect that the sensor was damaged during PCB assembly in the oven. Is this a possibility? Is there another way to verify our results?