FIFO alignment is not correct after some time

[venkatkrishna1986]

Posted on September 26, 2016 at 23:15

We enabled GyroScope and acclerometer. We set FIFO Mode to continuous with BDU bit is set. We set ODR rate to 1.66kHz .

We are reading FIFO continuously. we are observing FIFO data mis-alignment Can somebody help on this? #lsm6ds3-fifo-alignment

[Petr S]

Posted on October 19, 2016 at 11:32

Hello Krish,

for reference you can download the X-CUBE-MEMS1 package from

http://www.st.com/content/st_com/en/products/embedded-software/mcus-embedded-software/stm32-embedded-software/stm32cube-expansion-software/x-cube-mems1.html

where besides many others you can find the ''FIFOContinuousMode'' example source code showing the usage of LSM6DS3 FIFO in Continuous mode.There will be also the extended version called X-CUBE-MEMS-XT1 available at the end of this year or latest beginning of next year. This package will support all current sensors plus selection of additional sensors in DIL24 socket.Thank you and best regards,Petr

[cyril]

Posted on February 06, 2017 at 14:33

I have the same question here. 

I read the FIFO containing 624 words based on the INT1 interrupt event (FTH).

It works well the first time, I read all the words (I check the FIFO STATUS1 and STATUS2 registers to make sure the FIFO is empty). But then, the interrupt fires a second time and I can see in the status registers that 625 words are available (when FIFO threshold is set to 624).

It means that I get the AZ value first (same as the last sample in the last flushed FIFO), then I get all the samples I was waiting for.

Do you have any explanation?