Post Go back to editing

ADXL325 and ADXL326 different accelerations for same setup and same input signal

Hi,

I am using ADXL326 EVK in my designs to measure the acceleration of an Actuator. For the verification, I tried to use the other accelerometer ADXL325 to get G peak values. Strangly observed totally different results.

As per the datasheets ADXL326 requires, Vp-p/2*57 and ADXL325 requires Vp-p/2*174 to get the G peak values.

In my system, using ADXL326 measured as 625mV p-p and ADXL325 measured as 1140 mV p-pso, the effective Gpeak values are 5.48 and 3.27 respectively.

Both the chips are operating with 3V supply.

Here are my questions:

Why I am seeing the different G peak values?

Is there any way I can measure 1G/2G/3G for calibration?

Do you recommend any system available in market to get the calibration?

Thank,

Sree

Parents
  • Thanks for the question.

    In case there's any discrepancy in your input signal between your multiple tests, I would suggest to test the output without the input signal and see if the part is correctly set up and outputs as expected. Please also note that the ADXL326 and ADXL325 has different noise density which means when you use the same external filter to set the same bandwidth, the RMS noise will be different so as the peak to peak noise. However the difference should be within 20% range though based on the typ value on the datasheet.

    A quick way to calibrate the sensitivity and offset is to place the parts(axis of interest) in the +/-1g field(flip the part in gravity field) without any inputs signal for a simple two point calibration, you can use a bubble level as a reference and see if the number matches the spec on Table 1. You can refer to this link for more information regarding the 2 point calibration technique. For a 2G/3G calibration, you can also look into test fixtures like rate table that can provide constant centripetal acceleration.

    https://www.analog.com/media/en/technical-documentation/application-notes/AN-1057.pdf

    I hope this helps.

Reply
  • Thanks for the question.

    In case there's any discrepancy in your input signal between your multiple tests, I would suggest to test the output without the input signal and see if the part is correctly set up and outputs as expected. Please also note that the ADXL326 and ADXL325 has different noise density which means when you use the same external filter to set the same bandwidth, the RMS noise will be different so as the peak to peak noise. However the difference should be within 20% range though based on the typ value on the datasheet.

    A quick way to calibrate the sensitivity and offset is to place the parts(axis of interest) in the +/-1g field(flip the part in gravity field) without any inputs signal for a simple two point calibration, you can use a bubble level as a reference and see if the number matches the spec on Table 1. You can refer to this link for more information regarding the 2 point calibration technique. For a 2G/3G calibration, you can also look into test fixtures like rate table that can provide constant centripetal acceleration.

    https://www.analog.com/media/en/technical-documentation/application-notes/AN-1057.pdf

    I hope this helps.

Children
No Data