It appears like this is a duplicate post, so I am going to close this one.  Please feel free to correct me, if this is unique, with respect to this thread. 

https://ez.analog.com/mems/f/q-a/545181/adxl355-temprature-sensor-readings 

Just checking back in on this.  Were you able to figure out what the problem was? Thank you! 

Just checking back in on this.  Calibration can definitely improve on the precision, but is often easier to execute on, when you have clear requirements and an understanding for what you can trade (cost, complexity, size, power), in return for that improved…

Generally speaking the ADXL355 has excellent low-frequency noise metrics and offers the convenience of triaxial, digital functionality.  However, its measurement range and frequency response characteristics can be limiting in some vibration applications…

Thank you for your reply.
No, I didn't find a commitment for a new revision. The datasheet just says:
"an unlikely race condition may occur in products with REVID = 0x01 or earlier."
That's why I assumed that there was already a new rev…

Thank you, all three of you, for posting your question in this forum.  I am sorry that you have not received a response until now.  We have some staff turnover and are in the process of rebuilding our support team, for this forum.  Please see my answers

Thank you for posting your question in this forum.  I am sorry that you have not received a response until now.  

GBra - Are you referring to this graph?  This intends to quantify the nonlinear error, which equates to less than 0.5C of nonlinear error, over…

Thank you, John.Tuazon

Mo-Hassan: thank you for posting your question in this forum.