ADXL345 self test issue

Hello

         Customer use self test feature and setting high and low threshold for 3-axis. They test 55 pcs and happen 5 pcs fail due to under threshold. But they take fail unit. and re-heating ADXL345. After this action, the fail unit become good unit due to 3-axis results are between high/low threshold.   

         Why self-test results of Z axis will become pass results between HIgh/Low threshold after re-heating chip? Has any comments for this issue root cause and how to verify it? Thank you. 

           

Fail unit. 

Failed z self-test! z_sum = 107 (under 110.25 so it is fail) 

Threshold:  z_min = 110.25 z_max = 1286.25

Others axis pass setting threshold. 

BR

Patrick

Parents
  • 0
    •  Analog Employees 
    on Jun 23, 2021 7:11 PM

    Hi Patrick, 

    Can you clarify what do mean by re-heating? Do you mean re-soldering? My guess is that the soldering process induced some stress to the ADXL345 package, causing extra 0g offset due to stress. By re-soldering the board, the stress could have been released.

    I hope this helps, 

    Pablo.  

  • Hello Pablo

         Thank you. re-heating mean re-soldering. If customer re-soldering the will get pass 3 axis data between threshold. Has any suggestion to fixed the issue for customer SMT soldering on mass production? How to confirm it is soldering process induced some stress to the ADXL345 package? Thank you. 

    BR

    Patrick

  • +1
    •  Analog Employees 
    on Jun 24, 2021 2:25 PM in reply to PChen833

    Hi Patrick, 

    They should follow the recommended soldering profile on the ADXL345 datasheet, Figure 60. 

    Can you provide details of how are the min and max thresholds for self test determined by the customer? I am having a hard time figuring out why  z_min = 110.25 z_max = 1286.25. 

    Regards, 

    Pablo. 

  • Hello Pablo

               Customer follow up recommended soldering profile on datasheet. For z axis calculate min/max, they use below table. (VCC is 3.3V). 

    Zmin=75x1.47=110.25

    Zmax: 875x 1.47= 1286.25

                   Has any problem for min/max of self test calculation as above?

               Anther question,  offset calibration is also used by self test mode?  Customer do not use offset calibration on current self testing verification. Do we need to do offset calibration to solve offset is due to soldering?  Has any related offset calibration doc that customer can implement it on current testing? Thank you.     

    BR

    Patrick

  • 0
    •  Analog Employees 
    on Jun 29, 2021 3:02 PM in reply to PChen833

    Hi Patrick, 

    ok thanks for explaining that. The min/max calculation seems to be correct. I have a couple of remarks, in page 31 of the data sheet, it is recommended to average at least 0.1s worth of data at an ODR >100Hz. I would like to confirm they are following this, as well as waiting for the output to settle after self-test is enabled. 

    Also notice that in the same page, it says that " Using a range below 8g may result in insufficient dynamic range and should be considered when selecting the range of operation for measuring self test."  To avoid having this issues, what I would recommend is to change the range from 2g to 16g for the self test function, and they go back to the 2g range if necessary. 

    I hope this helps, 

    Pablo.  

Reply
  • 0
    •  Analog Employees 
    on Jun 29, 2021 3:02 PM in reply to PChen833

    Hi Patrick, 

    ok thanks for explaining that. The min/max calculation seems to be correct. I have a couple of remarks, in page 31 of the data sheet, it is recommended to average at least 0.1s worth of data at an ODR >100Hz. I would like to confirm they are following this, as well as waiting for the output to settle after self-test is enabled. 

    Also notice that in the same page, it says that " Using a range below 8g may result in insufficient dynamic range and should be considered when selecting the range of operation for measuring self test."  To avoid having this issues, what I would recommend is to change the range from 2g to 16g for the self test function, and they go back to the 2g range if necessary. 

    I hope this helps, 

    Pablo.  

Children
No Data