Batched accelerometer readings providing unexpected timestamps on Versa (sdk4.1)

[digitalparent]

Hi,

I am trying to use the batched accelerometer at 100 Hz sampling/30 reads per packet. What I am seeing is that consecutive reads are not separated by 10 ms but often 0 ms and then a jump:

index	timestamp	delta	x
0	1913440640		2.67
1	1913440640	0	2.56
2	1913440640	0	2.56
3	1913440640	0	2.58
4	1913440640	0	2.73
5	1913440768	128	2.76
6	1913440768	0	2.76
7	1913440768	0	2.94
8	1913440768	0	3.14
9	1913440768	0	3.31
10	1913440768	0	3.24
11	1913440768	0	2.99
12	1913440768	0	2.97
13	1913440768	0	2.87
14	1913440768	0	2.66
15	1913440768	0	2.23
16	1913440768	0	1.75
17	1913440768	0	4.4
18	1913440896	128	3.29
19	1913440896	0	2
20	1913440896	0	2.17
21	1913440896	0	2.02
22	1913440896	0	2.78
23	1913440896	0	3.02
24	1913440896	0	1.99
25	1913440896	0	1.86
26	1913440896	0	2.55
27	1913440896	0	3.11
28	1913440896	0	3.32
29	1913440896	0	3.31

Another issue I have seen is that a batch will include a time reading from a previous batch. Any tips on how to address these 2 problems.

thank you!