Post Go back to editing

ADXL213 Self-Test Response

We are running an application that performs BIT of components on the board.  As a part of this, we put the ADXL213 into self-test and check the output response to ensure it is within acceptable limits of 23%.  Roughly 5% of the time we are failing this test.  My software developer says it appears that sending the self-test discrete signal seems to make the chip do odd things momentarily.  I am of the opinion that we need to allow the chip to settle. 

We are investigating using a unit that we cannot open and look at to characterize, plus that would only give us a sample of 1.  Can anyone detail how long it could take, worst case, for this chip's self-test response to settle out?  The chip is tuned to output at 10Hz.  It would be useful to know too how quickly the chip responds to being taken out of test mode and settles into normal operation.