Skip to main content
Bibin
Visitor II
September 15, 2026
Question

LSM6DSV80X: Measured Accelerometer 3 dB Bandwidth Lower Than ODR/2 — LPF1 Only Enabled

  • September 15, 2026
  • 0 replies
  • 19 views

LSM6DSV80X: Measured Accelerometer 3 dB Bandwidth Lower Than ODR/2 — LPF1 Only Enabled

Hello ST Community,

In our new application, we are using the LSM6DSV80X IMU, which has independent low-g (XL) and high-g (HG) accelerometer channels.

Our configuration

  • Accelerometer ODR: 960 Hz for both XL and HG.
  • Accelerometer operating mode: High-performance].
  • Accelerometer filtering: LPF1 enabled, LPF2 disabled.
  • HP_SLOPE_XL_EN = 0 and LPF2_XL_EN = 0 for the XL low-pass path.
  • Frequency-response measurement performed using a controlled vibration input.

Observed frequency response

We measured the following approximate −3 dB bandwidths:

Accelerometer channel

Datasheet expected BW

Measured −3 dB BW

High-g (HG)

ODR/2 = 480 Hz

Around 400 Hz

Low-g (XL)

ODR/2 = 480 Hz

Around 350 Hz

The measured frequency response was obtained by comparing the peak-to-peak acceleration output at different input frequencies against the low-frequency reference.

Questions

1. LPF1 bandwidth and actual −3 dB frequency

According to the LSM6DSV80X datasheet, the LPF1 output cutoff is specified as ODR/2 in the applicable high-performance, high-accuracy ODR and normal modes.

For an ODR of 960 Hz, this gives a nominal cutoff of 480 Hz.

However, our measured −3 dB bandwidth is approximately 400 Hz for HG and 350 Hz for XL.

Could ST please clarify:

  • Is ODR/2 the actual −3 dB cutoff frequency of LPF1, or does it represent another bandwidth definition?
  • What is the expected typical and maximum/minimum LPF1 −3 dB frequency at an ODR of 960 Hz?
  • Is there a difference in the LPF1 frequency response between the XL and HG accelerometer channels?
  • Could the difference be related to the internal digital filtering, operating mode, or actual internal ODR?

2. LPF1 and LPF2 filter implementation

What type of digital filters are used in the LSM6DSV80X accelerometer signal chain?

Specifically:

  • Is LPF1 implemented as FIR or IIR?
  • Is LPF2 implemented as FIR or IIR?
  • If IIR, are the filters based on Butterworth, Bessel, or another filter topology?
  • What is the filter order?
  • Are the filter coefficients or detailed frequency-response curves available?

We understand that the datasheet specifies the filter bandwidth but does not appear to provide the exact filter topology or coefficients.

3. Additional configuration or measurement considerations

Are there any other device settings that could affect the measured XL/HG bandwidth even when LPF2 is disabled?

For example:

  • HAODR configuration.
  • High-performance versus normal mode.
  • Internal averaging or low-power mode.
  • Any other internal filtering or decimation in the accelerometer signal chain.

We would appreciate any clarification from ST regarding the expected LPF1 response and the reason for the lower measured bandwidth.

Thank you.

 

 

Terms & ConditionsAccessibility statement