lsm6dsv16x temperature sensor and learning tools for mlc and fsm

[ACand.3]

Hi everyone,

I’m evaluating the LSM6DSV16X for a project and have a couple of questions:

1. Temperature Sensor Accuracy

I know the built-in temperature sensor is mainly for internal monitoring, but I’m wondering how reliable it is for actual ambient temperature readings. Specifically:

• How accurate is it in real-world conditions?
• Does self-heating significantly affect readings, especially at higher ODRs?
• Has anyone tested it in thermally insulated environments—does it tend to read higher than expected?
• Any recommendations for calibration or improving accuracy?

2. MLC & FSM Learning Resources

I’d also like to dive deeper into the Machine Learning Core (MLC) and Finite State Machine (FSM) features. I am new to the machine learning world.  Besides the application notes, are there any detailed guides, training materials that can help with:

• Training and optimizing MLC models
• Configuring FSM efficiently
• Debugging and testing implementations

If anyone has experience using these features, I’d love to hear your thoughts!

Thanks in advance!

[Federica Bossi]

Hi @ACand.3 ,

We do not provide an accuracy value on the temperature sensor because these are values that we only test on a limited number of parts and therefore cannot guarantee in production.

The internal temperature sensor is calibrated at production tester level.

The self-heating can affect the readings of the LSM6DSV16X, especially at higher ODRs, but it should be quite limited.

About FSM and MLC Resources you can find some articles in the knowledge base section of the forum.

Here some examples: 

How the edge features of LSM6DSV16X augment your s... - STMicroelectronics Community

How to save power with LSM6DSV16X by leveraging on... - STMicroelectronics Community

How to start and improve your workflow with MEMS-S... - STMicroelectronics Community