MEMS (sensors)

I have implemented 9D orientation calculation on the STM32L4 using the LSM6DSO16IS and LIS2MDL sensors. Since there is an initial delay when flashing the program onto the ISPU, I chose to implement the algorithm on the STM32 itself, bypassing the ISPU.The issue I'm facing is that the yaw axis value changes when I move the system and place it back in the same position—even after calibrating the magnetometer. I am using motionfx.h for both magnetometer and gyroscope calibration.I believe the problem lies in the gyroscope calibration. When I set iKnobs.LMode = 1 (static mode), the gyroscope does not calibrate during motion, and the resulting drift affects the yaw angle. To address this, I also tried setting LMode = 2 (dynamic mode). However, in dynamic mode, calibration saturates after some time and stops updating. After that, the yaw value continues to drift even when the device is stationary.Below is the code snippet I am using for orientation calculation and sen

Hi, I'm putting down a LIS2DH12 accelerometer on a new PCB. The datasheet calls for an aluminum capacitor on VDD, but the eval board (STEVAL-MKI151V1) shows a small capacitor that looks like a ceramic. I don't know for sure because I can't find the BOM for the eval board. I really don't have clearance for an electrolytic capacitor. How critical is it?

Hello STM32 Community,I’m currently working with the STEVAL-PROTEUS1 evaluation board, which integrates an STM32WB5MMG microcontroller and an IIS2DLPC MEMS accelerometer. I’m using the board together with the ST BLE Sensor mobile app, and it works great out-of-the-box for streaming sensor data over BLE.However, I’d like to change the sampling frequency (ODR) of the onboard accelerometer to suit my specific application needs. For example, I want to increase the data rate to 800 Hz, instead of the default (which seems to be 25 Hz or 50 Hz).What I’ve done So Far: int32_t IIS2DLPC_Init(IIS2DLPC_Object_t *pObj){  /* Enable register address automatically incremented during a multiple byte  access with a serial interface. */ 

Although these devices are technically "water resistant", they are sometimes described as "water proof" (even in STM materials), and the data sheet permeability of the gel is quite low. Clearly these devices can survive immersion for at least several hours, but for how long, really...There are several posts on here asking questions about this, but no quantitative answers. Has anyone (ST engineers especially) conducted experiments on device drift or lifetime for continuous immersion in water (without worrying about aggressive chemicals or surfactants)?What is the drift or lifetime for direct immersion?What about if a flexible membrane is used as an outer layer of protection (e.g. a thin layer of nitrile rubber). Polymer membranes won't be impereable, of course, but would be something which would not resist pressure change but should limit the sensor to contact with water vapour rather than liquid.

I followed most of the examples to get data such as accel, gyro, timestamp and temperature from the github repo using fifo interrupts. I also used the mlc_gym example to get data from a MLC configuration that I designed on Mems studio.  My question is, can I use both configurations at the same time? like have an interrupt for the fifo data ready and one interrupt for the MLC? As in, I am constantly reading raw accel/gyro/timestamp/temp data at a fixed 60Hz on interrupt 1 and also get a new MLC state on interrupt 2?

Hi All,I use STHS34PF80 for presence detection with embedded algorithm however it easily get stuck in presence status and cannot release from the status.Use case:People sit in from of the sensor around 30 cm, with lens on STHs34PF80.When people enter the FOV it's triggered as presence status.When people sit still for a while (e.g. 30 min), the TPresence keeps glowing.When people leave, TPresence drops however it cannot across threshold.TPresence will keep fall very slowly until it across the threshold however sometime it will never across below the threshold.Red line: presence flagOrange line: TPresence valuemy setti

Hi everyone, I'm developing a tamper detection circuit for a meter enclosure. I'm considering the STHS34PF80 infrared proximity sensor, but I'm struggling to find reference materials or application notes specifically related to tamper detection using this sensor. My goal is to use the sensor as an interrupt generator, similar to a switch. When someone opens the enclosure lid, the sensor should detect the change in proximity and trigger an interrupt signal to the microcontroller, indicating tampering. Can anyone recommend reference materials or alternative sensor options suitable for tamper detection in this application? Thanks.

Hi everyone,I'm working on a project involving the LSM6DSR sensor (very similar to the LSM6DSL), connected via SPI to a Verdin i.MX8M Plus SoM (NXP) mounted on a Dahlia Carrier Board. The firmware runs on the Cortex-M7 core (MIMX8ML8DVNLZ) using Zephyr RTOS v4.1.0.What I've done:Integrated Zephyr’s official lsm6dsl sensor sampleConfigured the sensor in the device tree (reg = <0>, spi-max-frequency = <10000000>)Enabled CONFIG_LSM6DSL=y, CONFIG_SPI=y, CONFIG_SENSOR=y in prj.confThe driver appears to initialize in the logs, but always ends up saying failed to initialize chip 0x0The issue:No SPI activity is observed on the oscilloscope.Before firmw

I'm working with the IIS3DWB high-performance MEMS accelerometer and using it to calculate both acceleration RMS and velocity RMS values.While the datasheet specifies a frequency response from DC to 6 kHz (±3 dB), I’ve noticed in practice that reliable and accurate RMS measurements only begin to appear from around 40 Hz onwards.After enabling the digital high-pass filter (HPF) as shown in the documentation (see attached figure), I understand that the minimum cutoff is ODR/800, which translates to 33 Hz at an ODR of 26667 Hz. This explains the attenuation of low-frequency components and might be contributing to the inaccuracy in the 2–40 Hz band.What is the best way to achieve accurate acceleration and velocity RMS values in the 2 Hz to 40 Hz frequency range using the IIS3DWB?

Hello ,I'm using an STM32U575 with an I2C connection to the AIS2DW12 accelerometer. The goal is to generate an interrupt when a free-fall event is detected. I followed the instructions provided in the datasheet (AN5326 - AIS2DW12 ultralow-power 3-axis accelerometer for automotive applications, STMicroelectronics).I created a function to initialize the device using the specified register values, but no interrupt is detected. There are no hardware issues—when I force the INT1 pin to a high level, the MCU correctly enters the interrupt function. Later, I found a GitHub example for free-fall detection that doesn’t use interrupts.

Hi, I am working with the LSM303AH (accelerometer + magnetometer) as a potential replacement to the LIS2DS12 (accelerometer only). The accelerometer section of both devices seems identical with the only exception being the SPI on the LSM303AH is 3 wire only. To evaluate the LSM303AH I am using the same software as has been used with the LIS2DS12 for a long time and is reliable (altered only for 3-wire SPI) but am getting strange behaviour.When reading from the FIFO in a burst read, the first 24 bytes (4 samples) come out fine but then there are a block of 16 bytes read as all zero, after this the behaviour seems erratic, some devices return to reading correct data, others return correct data but with sporadic erroneous zero bytes mixed in. If instead of burst reading all data from the FIFO in one go I read 6 bytes (1 sample) at a time with separate SPI read commands the data comes out fine.This behaviour has been replicated on 3 devices.The application notes for

Hi,my company is currently evaluating usage of the IIS2DULPX accelerometer in one of the products we are developing. Our use case requires that a wake-up interrupt is generated whenever acceleration sample exceeds a set threshold value. However, we would like to generate this interrupt with a threshold that takes into account either:an absolute value of a data sample, or a difference between a sample and zero-g offset (interrupt generated when sample > threshold - zero-g offset)a difference between a data sample and a configured, constant reference value (interrupt generated when acceleration sample > threshold +/- constant reference value).We would like to do so while keeping the ODR at 400-800 Hz and the ability to configure threshold as a value in a full +/- 16g range. From the data sheet it seems that the wake-up feature is fed by an internal slope filter, making it impossible to use wake-up feature as

Hello,I am trying to interface a Nordic MCU to a LSM6DSV16X, but I cannot get the communication to work.I am using 2 dev kits for this test setup, a nRF5340-DK and a STEVAL$MKI227KAA I got from the STEVAL-MKI227KA. VDD/VDD_IO are both 3V, CS/SPC/SDI/SDO are all used and connected correctly. SPC frequency is 1MHz and the SPI interface is set to Mode 0.Even the WHO_AM_I register is read wrong from the start. The code has a 10ms delay at startup, then it tries to read the WHO_AM_I register to verify the chip ID, but instead of reading 0x70 as expected I get 0x60, as shown on the oscilloscope screenshot below (CS is yellow, SPC is red, SDI is blue, and SDO is green) . This screenshot was taken by triggering on the first falling edge of the CS pin after switching the board from OFF to ON, so no previous interaction on the SPI bus has occurred. The screenshot shows that the MCU sends 0x8F to the IMU, requesting a read of register 0x0F (WHO_

Hi everyone!What is the benefit of setting the accelerometer ODR to 416.0f Hz when setting up the motion detection features? All motion detection features use this value. Why?An example would be the 6D Orientation detection:int32_t ISM330DHCX_ACC_Enable_6D_Orientation(ISM330DHCX_Object_t *pObj, ISM330DHCX_SensorIntPin_t IntPin) { /* ... */ /* Output Data Rate selection */ if (ISM330DHCX_ACC_SetOutputDataRate(pObj, 416.0f) != ISM330DHCX_OK) { return ISM330DHCX_ERROR; } /* ... */ }Why was this 416.0f Hz value chosen? I cannot find this in the driver documentation or anywhere else. Is this written down somewhere?Thanks!

We have purchased some LIS2DU12 and when scanning the QR code markings we see 78327KT603ES0YJ on some and 78327KT603EO0YZ on others. What is the difference? And how do we decode the QR code, didn't see anything on the datasheet?Thanks.

Hello, I have a difficulty using MEMS studio. It says "board not identified" without any more hint what could be wrong. I use STM32G474 Nucleo board with IKS01A3 on Linux (Ubuntu).I have all firmware up to date, all pins mapped correctly according to documentation but still didn't manage to establish communication. From what I saw there is some communication on UART. I am able to read the sensors from the MEMS board using my code. I can compile it and the MEMS board works... But not with the MEMS studio. Weird.  Attaching the .ioc for more context.Thank y

Hi ,I see that the bandwidth of IIS2DLPC is 3200 Hz for low power mode 1 according to the datasheet (page 22 ). Also it mentions that the maximum ODR is 200 Hz for this mode. This looks contradictory to me as 3200Hz bandwidth won't be compatible with 200Hz output data rate. Could you please explain what exactly does this mean?Best regards,Kev

I'm working with the AIS2IH accelerometer.  I've established communication via SPI, and am able to read the WHOAMI correctly, along with requesting and receiving data as expected.  However, some of the data I'm reading doesn't make a lot of sense, so I've been investigating our chip setup.  It appears that we're having issues getting some of the control registers to set correctly.Specifically I seem to be having issues with CTRL1 & CTRL6.  There may be others that I haven't discovered at this point in time, so don't assume this is a complete list.I first noticed this issue when trying to set CTRL6.  Specifically, I'm trying to set the Full-Scale Selection to +/- 16g.  According to the data sheet, I do this by setting bits 4&5 of the register to 0x3.  I'm able to send this command properly, however, when I read back the settings on CTRL6, I'm only reading 0x2 in those bits.  If I send 0x1 in place of 0x3, I receive 0x0 fro

Hello I suspect I'm about to make a dumb question but I've been strugling for the past 2 days trying to just make a simple reading of the accelerometer. Just that get those values out of X, Y and Z axis...I can get the correct responses out of the WHO_AM_I register and the status from the other setup registers. No matter what I do, the readings from registers 0x28~0x2D keep returning zero. I need help to understand what I may be doing wrong...Code below:  #include <Arduino.h>#include <Wire.h>#define i2c_sda 8#define i2c_scl 9#define x_l 0x28#define x_h 0x29#define y_l 0x2A#defin

Hello,I am developing a board with the LIS2DUXS12 accelero.During testing after assembly, I saw that there was no reaction to I2C communication. Other devices on the bus work fine, with a logic analyzer I have seen that there is indeed traffic sent to 0x18. I desoldered the component and saw that during assembly, was too little paste on the VDDIO and the SCL pin. I corrected it and resoldered the component and saw there was good contact under the microscope. I have connected the STEVAL$MKI235AA to the pins and that responds correctly.I still see the same behavior, so my first question is this:1. Are VDD and VDDIO completely isolated from each other such that VDD won't internally be connected to VDDIO in the event of a bad connection to VDDIO, the digital interface pins are in danger of being damaged?2. I also am looking for a sanity check of my schematic. I have only 1uF + 100nF on the supply as 10uF seems extremely overkill for

Is there any noise spec on VDD of LSM6DSV36S that we need to be aware of when supplying the voltage to it? I.e. concern of specific frequency of a switching regulator, ripple spec, etc. 

Hi everyone,I have set up my LIS2DW12 to recognise double taps in the Z direction, which works well with the following settings:ODR = 400 Hz (High performance mode)FS = +-2gLow Noise Mode enabledTap threshold: 0b1000 = 0.5 gLatency: 0x0F = 1200 msQuiet: 0x03 = 30 msShock: 0x02 = 40 msNow I wanted to fine-tune these parameters for my application, as some collisions that should not be detected are also being detected as double-tap events.My idea was to read the acceleration values from the output register/FIFO and send them via the UART interface of my MCU to the virtual COM port of my PC in order to recognise the characteristic waveform and derive the appropriate parameters.My questions at this point are: Which signal does the LIS2DW12 use as input for the internal double-tap detection logic, and is it possible to output the same data via the output register/FIFO? As I know fro

Good Morning,I own both a SensorTile.box and a SensorTile.Box PRO. I used the SensorTile.box to identify anomaly vibrations on a device, using the prebuilt app, with slight modifications, that performs the FFT of data coming from the accelerometer and then the FFT Compare with the reference FFT captured when no anomaly is present on the device. Now, due to some limitations of the SensorTile.box, I want to replicate it on the SensorTile.box PRO, but among the available functions offered to manipulate data, there is only FFT (DFT - Discrete Fourier Transform), but not the FFT Compare which is available on the base board. The question is: how can I recreate the app developed on SensorTile.box to operate also on the SensorTile.Box PRO? Do I need any firmware upgrade and if yes, where I can download it? It seems to me very strange a so important function is not present and I don't think it could be difficult to insert it also on this device.Could s

Hi,We currently use the LIS2DW12 in our design and have encountered an issue in some units after prolonged continuous operation (approximately 5–6 days). Specifically, the accelerometer output becomes unresponsive and remains stuck at a constant value. Attempting a soft reset does not resolve the issue.During this fault condition, all device registers remain readable, and the WHO_AM_I register continues to return the correct value (0x44). However, we observed that the BOOT bit in the CTRL2 register (0x21) remains stuck at '1', suggesting that the internal reboot process has not completed successfully. The only way to recover from this state is to complete a power cycle. Unfortunately, due to our product's design, removing the onboard battery is not feasible for end users.Could you advise on the potential root causes of this behaviour and whether any recommended solutions or workarounds do not require a physical power cycle?

Hi guys,I'm working on a new hw design and I can't find the difference between these two references.Thanks in advance,Kind regards