You could see the previous table ODR had +/-1% accuracy. If the ODR is 50 Hz, the data is not updated accurately for 20 ms each time.
Thank you Robin.
I had a follow up question. Currently we are using BMI 160 sample index and samplying frequency to get time stamps of certain motion events that we detect. For eg, if the event happens at 300th sample and sampling frequency is 50hz, it happened after 6seconds from a reference point.
Judging from your previous reply, can I assume that this method is not going to give an accurate answer and the error is going to accumulate as time progresses?
Also, is there any other method to locate the time stamp of an event except the CPU time?
Yes, the error is going to accumulate as time progresses.
It could be corrected periodically by referring to the local time. For example, after a few seconds, compare the local time with the accumulated time.