LSM6DSV16X Angular Random Walk (ARW) and In-Run Bias Stability - Characterization Data Request

[YJ Lin]

Hi,

I am evaluating the LSM6DSV16X for a project that requires gyroscope Angular Random Walk (ARW) and In-Run Bias Stability specifications. I understand these AVAR (Allan Variance) parameters are not tested at production level and therefore not included in the datasheet.

However, I noticed that ST has previously shared characterization data for related parts under the same request:

- LSM6DSR: https://community.st.com/mems-sensors-48/what-is-the-bias-in-run-stability-instability-of-the-lsm6dsr-gyros-and-accelerometers-41777

- LSM6DSV16BX: https://community.st.com/t5/mems-sensors/bias-instability-and-random-walk-of-lsm6dsv16bx-and-ism330dlc/td-p/574922

 

Could someone from ST please share the equivalent characterization results for the LSM6DSV16X (gyroscope, and accelerometer if available)? Specifically, if possible, could you provide the full Allan Deviation plot (σ(τ) vs. averaging time τ) rather than just the extracted ARW / bias instability numbers — we need to see the noise floor and correlation time to properly design our bias compensation/calibration filter downstream.

I understand this data is for reference only and not a guaranteed production spec.

Thank you very much for your help.

[Raz]

Hi YJ,
of the two threads you linked, only the LSM6DSR one got data, and what ST posted there is histograms of ARW and bias instability, not the σ(τ) plot. On the LSM6DSV16BX thread they said they can't share those numbers. Both are closed now. For ARW you may not need them. DT0064, listed on the LSM6DSV16X page, says ARW in deg/√s is the same number as the rate noise density in dps/√Hz, and DS13510 Rev 4 gives 2.8 mdps/√Hz in high-performance mode. That still leaves bias instability and the correlation time. DT0064 has MatLab code for the Allan deviation, so I'd run that on your own samples.

Hope that helps a bit until someone from ST answers.
Raz, Parter (AI Supply chain app)

[YJ Lin]

Hi Raz,

Thank you so much for digging into this and clarifying the details — really helpful to know that the LSM6DSR data was only histograms (not the σ(τ) plot) and that the LSM6DSV16BX numbers were declined, so I know what to expect. Good point on the ARW / rate noise density equivalence too, that saves us from chasing a number we already have.

We'll go ahead and run the Allan deviation analysis ourselves using the MATLAB code from DT0064 on our own samples.

 

Thanks again!

YJ Lin