MEMS and sensors

Hello ST Team,We are currently building an AI-powered digital pen using ST MEMS sensors (accelerometer, gyroscope, magnetometer) and the STM32WB05KZ MCU. We’ve reviewed the st-mems-finite-state-machine repository and are exploring how the embedded FSM core in sensors like LSM6DSO32XTR can help detect gestures such as:Double tapSwipe gesturesRepetitive handwriting-like motion (“double write”)Tilt-based mode switchingWe are aiming for a low-power design, where motion recognition is offloaded to the sensor and only specific events are passed to the MCU for further processing.Could you clarify:Can the FSM core be extended to detect custom gesture patterns (e.g., two similar handwriting motions)?Is there an example or methodology to define gestures based on gyro and magnetometer data, in addition to accelerometer signals?Any tools or guidelines on blending FSM with ML or application-level classification using the sensor + STM32WB05KZ?This would greatly help validate our approach and optimiz

I am trying to build a project to detect vibration using STEVAL PROTEUS 1 (STM32WB5MMG, ISM330DHCX and BLE SENSOR). Unfortunately my sensor and BLE does not initialize despite 0 errors.I have used custom BLE Template. I have initialized sensor code in custom_app.c. My inference:Below is the main.c function MX_APPE_Init(); Custom_APP_Init();////  UTIL_SEQ_RegTask(1 << CFG_TASK_BLINK_LED_ID,while (1){   /* USER CODE END WHILE */MX_APPE_Process();   /* USER CODE BEGIN 3 */// HAL_GPIO_TogglePin(GPIOD, GPIO_PIN_2);HAL_GPIO_WritePin(GPIOD, GPIO_PIN_2, SET);HAL_Delay(100);HAL_GPIO_WritePin(GPIOD, GPIO_PIN_2, RESET);HAL_Delay(100);}/* USER CODE END 3 */}Here, the code works when I comment out MX_APP_Init() and MX_APPE_process() and i can see the LED toggle, LED constantly stays ON when I only comment out MX_App_Process() but that is not the dfesired outcome i want. I want my code to help me log vibration data via BLE, it is not happening

Hello ST team,Thank you very much for your previous help with debugging TOF. We have now selected the TOFL1CX and TOFL1CB as our distance measurement sensors. However, we have encountered a problem: both TOF models have already been soldered onto the PCB, and we cannot tell which one is TOFL1CX and which one is TOFL1CB. We would like to distinguish the two TOFs by reading their IDs, but for some reason, the IDs read out are identical. Is there any way to distinguish between these two TOF models? Their appearance also seems identical.At the same time, I am a bit confused: are these two TOFs considered as high-end and low-end versions? According to the datasheet, the main difference is the measurement range, and TOFL1CX does not support multi-target detection (but I found that running TOFL1CB’s code on the TOFL1CX also works and provides multi-target output), whereas TOFL1CB officially supports multi-target.Thank you in advance for your reply.Below is the code segment I used to get the I

Hi @John E KVAM ... I have seen several discussions about best way to use multiple VL53L4CX sensors, and I've appreciated your responses.  ("Every reboot is its own adventure!").  I'm currently implementing an unusual setup that has 5 of these sensors that could conceivably interfere with one another - we are trying to get our mounting and the view-cones to be as independent as possible, but it can't be guaranteed.   We are using the bare STM32duino library on an ESP32 (Arduino/C++) micro, which is also doing many other things.Given the physical setup of our system, it is difficult to run single lines to all the chips for XSHUT.  We would very much like to work only with the simple STEMMA/QT-style 4-conductor I2C cabling.   For this reason, we've opted to use an I2C multiplexer, rather than changing the addresses of the chips.So my questions are: 1. Is there an I2C command to tell the chip to reset itself?   In our projec

 I am trying to collect accelerometer, gyroscope, and temperature data from the ASM330LHH IMU sensor via SPI by timestamping the exact moment the data is generated. I have connected the INT1 pin of the sensor to the STM32H735RGV6 microcontroller’s PB0 pin and configured it as an external interrupt input. I enabled both the data-ready (DRDY) interrupt and the FIFO threshold interrupt by setting the appropriate registers (CTRL3_C, INT1_CTRL, MD1_CFG, FIFO_CTRL1, FIFO_CTRL3, FIFO_CTRL5, etc.). I verified that SPI communication is working correctly — I can read raw sensor data via polling mode, and the WHO_AM_I register returns the correct device ID. However, the INT1 pin does not toggle or pulse when configured for interrupt generation. I checked the INT1 pin using an oscilloscope and confirmed that it remains static.  Additionally, DWT->CYCCNT is initialized and functional for timestamping.Despite these checks, the interrupt callback (HAL_GPIO_EXTI_Callback()) never gets tri

Hi everyone,I’m developing an application that tracks the path of a device using the ISM330BX IMU and sensor fusion. I’ve based my implementation on ST’s GitHub example for sensor fusion, and the rotation outputs appear to respond correctly when I move the sensor. Pitch and roll are stable, and the orientation updates as expected.However, I’m seeing significant drift in the yaw when the device is stationary. I understand that yaw is difficult to estimate accurately with only a 6-axis sensor (no magnetometer), but I’ve observed a big difference between my application and the ISM330BX demo board (STEVAL-MKI245KA) running in MEMS Studio.For example:On my custom board, I’ve observed drift of nearly 2° per minute at a 3.84 kHz data rate.Lowering the data rate to 30 Hz helps — I now see around 0.5° per minute.In contrast, with the demo board in MEMS Studio (also at 3.84 kHz), the drift is only about 0.1° per minute.I’m trying to figure out what could be causing the higher drift in my se

Hi, I have a compulab SBC (UCMiMX8Mplus sbev) it is a dev kit. I am connecting the STEVAL-55GMBI to the board with the edge connector with our custom board. It is connected to the MIPI-CSI2 connection. I have build the vd55g1 driver with the kernel and the driver is being loaded. Although it does not create the subdevice nodes in the v4l2-ctl --list-devices. I do see errors regarding the setup : [ 10.219820] mx8-img-md: Registered sensor subdevice: vd55g1 4-0010 (1)[ 10.219837] mx8-img-md: subdev_notifier_complete error exit[ 10.219847] mxc-md 32c00000.bus:camera: Sensor register failed[ 10.219853] mx8-img-md: Unregistered all entities[ 10.219857] mxc-md: probe of 32c00000.bus:camera failed with error -515 I have attached the schematics of the adapter board that we have created to connect the SBC to the ST camera board. Could there be some issues with the driver configurations ?    

Hello,What is the maximum sustained long term acceleration that the ISM330 can handle before ruining the sensor?I see in the datasheet that the maximum acceleration for 0.2ms is 20,000 G, but I would like to know if it can handle a sustained 100G acceleration for an extended period of time.Thank you!

Hello,I have a case where my IMU (ISM330) can see peak accelerations above the 16G limit.How does this effect the output of the sensor?Could this cause aliasing of my data or would it just cause me to see a lower amplitude?Thank you! 

 Can someone please provide a working download file for "STM32Cube function pack for STEVAL-PROTEUS1 sensor datalog and file transfer over BLE and USB"?The ZIP-File Version is corrupt.It cannot be extracted neither on Windows, Linux or Mac.Thank you.@

Dear ST Support Team,I hope this message finds you well.I am currently working on a project where I aim to implement gesture recognition using the VL53L5CX sensor. However, I noticed in the official documentation that the supported cores for gesture recognition do not include the M55 (which is the core of my STM32N657 microcontroller).Could you please advise if there are alternative methods to achieve gesture recognition with the VL53L5CX on this platform? For example:Are there any software libraries or algorithms compatible with the STM32N657 that I could utilize?Would it be possible to port the existing gesture recognition functionality to the M55 core?Are there other recommended approaches or workarounds to accomplish this?I would greatly appreciate any guidance or resources you could provide to help me proceed with this implementation.Thank you in advance for your time and support.Best regards,dw1

Hello fellow users! I'm trying to implement the Fusion MOTIONFX library with the board X-NUCLEO-IKS02A1 and the nucleo board NUCLEO-U575ZI-Q so I can output via terminal all the values. So far I can output the fusion values but they are wrong as I need to define the sensors orientation on the PCB board (real world) and there is no documentation that indicates this for the IKS02A1 (I also have the IKS4A1 and such info also is not avilable) So will be nice of ST to have a:1)definitive guide or schematic showing the physical orientation of the sensor axes (ISM330DHCX, IIS2MDC and IIS2DLPC) on the X-NUCLEO-IKS02A1.2) The recommended orientation strings (acc_orientation, gyro_orientation, mag_orientation) for this specific board.3) Clarification: Does the user need to remap the sensor axes manually in code to match the declared orientation strings? Or does MotionFX internally handle axis rotation based on the strings?4) Ideally: an official reference example with working

Hello Everyone,We have a custom board that utilizes 10 x VL53L4CX sensors, length of board is about 66cm. For context I am using an NRF54L15 with Zephyr. On bootup they are all configured with a different address and an attempt is made to calibrate the sensors as well. I am using the I have been testing for a few weeks now and I am constantly running into initialization issues, all revolving around communication to sensors, especially when I am trying to run through a calibration on the sensor. I have tried different calibration methods, but all result in errors. And if I comment out the calibration I would get a Hardware Failure error after a few reads from the sensor. My initialization code flow is as follows: Shutdown all sensors -> Boot sensor N -> change I2C address -> Wait Device Booted-> DataInit -> PerformRefSpadManagement-> SetOffsetCorrectionMode(STANDARD)->PerformOffsetZeroDistanceCalibration()-> GetDeviceInfo()->SetTuningParame

Hi,I am currently working on a project using the STHS34PF80 proximity sensor from STMicroelectronics. The platform is running Android 14.0.0_1.0.0 (L6.1.55_2.2.0 BSP). I would like to know if there is a dedicated Android driver or SDK available to directly interface with the STHS34PF80 sensor—either from STMicroelectronics or open-source contributors. I have the datasheet and application notes for the sensor, and I understand its I²C communication protocol. However, before starting from scratch, I’d like to ask: 1) Any existing Android libraries or HAL implementations 2) Sensor integration examples in Android (preferably for NXP i.MX8M Plus - based platforms) Any examples or references would be greatly appreciated. Thanks in advance for your support!

Hi, I'm experimenting with VL53L4ED. it suits by requirements except that is has only 18° FoV. Is there any way to increase the FoV (upto120°) using a lens may be without hampering the laser class?

Hello.I am working with the STM32U083 and LIS2HH12TR communication via SPI. Did the SPI settings configuration in mode 3 and build a simple project in cube ide to identify and read the accelerometer data.The project also does the conversion in mg of the acceleration for all axis. When I move the axis I was expecting the values to change but there is no change as in picture below: (Debug data via uart ). Does anyone come out with this kind of problem ? Could it be sensor (accelerometer) damage, a soldering problem, or something else?Following the terminal log:Please support.Thankyou

Hi everyone, I have been struggling to work with VL53L5CX on my NRF board using Zephyr. I used the VL53L5CX_ULD_driver_2.0.1 API and changed the platform.cpp. So far, what my problem is as follows:Failed on the vl53l5cx_init, because after the firmware is successfully flashed the  _vl53l5cx_poll_for_answer always gets the wrong value as attached image Since _vl53l5cx_poll_for_answer calls VL53L5CX_RdMulti, I check whether my function is working correctly with this (https://community.st.com/t5/imaging-sensors/vl53l5cx-init-fails-in-vl53l5cx-poll-for-mcu-boot/m-p/614318/highlight/true#M3328 ). The results are that the data did not match the desired value, but it works for several reasons.My code is as follows:uint8_t VL53L5CX_WrMulti(VL53L5CX_Platform *sensor, uint16_t reg, uint8_t *p_values, uint32_t size) { uint8_t status = VL53L5CX_STATUS_OK; uint16_t current_address = reg; // Initialize current address uint32_t remaining_size

    Since we require long-distance detection, we have employed both Ranging Mode and Long Distance Mode. Additionally, due to the use of battery power, power saving is essential. Consequently, we acquire the distance reading after just two sampling cycles.    When the distance is less than 3 meters, we can successfully obtain readings without issue. However, when the distance exceeds 3 meters, acquiring a reading with only two sampling cycles becomes impossible. Instead, it requires 10 sampling cycles, or even several dozen cycles, to obtain a distance measurement. This significantly increases power consumption.    Why are multiple measurements required to detect objects beyond 3 meters? Can this be improved?    The software version we are using is STSW-IMG019: VL53L1CB_BareDriver_6.6.18.2652.    Thank you for your help.

Hi everyone,I am currently working with the STM32N6570-DK board and would like to integrate the STEVAL-56G3MAI camera module with it. The camera's overview page mentions that it supports the STM32N6 series. However, I have been unable to locate any STM32N6 header files to facilitate this integration.I did come across the STSW-IMG502 software package, which also lists STM32N6 compatibility, but it only includes drivers for Raspberry Pi and Linux platforms.I would greatly appreciate any guidance or resources (e.g., example projects, driver libraries) that could help me successfully interface this camera with the STM32N6570-DK board.Thank you in advance for your support.

We have designed and shipped a water level sensor based on the VL53L1X. It works well in most situations, but we have one issue in a particularly large tank (3m tall, 2m diameter). I believe this is due to reflections in the tank. It is a black tank, but the surface is glossy.I thought reducing the FoV might help with reflections, but the datasheet states this reduces range, but I can't really understand why. If the laser is pulsing light at the same power, and that light falls back on the SPAD, why does only looking at the center diodes affect range?I think I'm misunderstanding how it works so I thought I would ask. I'm not sure there is much I can do in this situation and it may just be unsuitable for this particular tank.Thanks

I am currently using a VL53L1X Sensor to read a distance in the range of 40cm to 130cm. I initialized the Sensor to work in the short distance mode, so the range should not be a problem. I used the VL53L1X_calibration library to get the values for offset and xtalk. In the Datasheet i read that with the function VL53L1X_GetDistance() the distance should be returned in mm. The value i get returned from that function does change when i change the physically measured distance, but it is not the expected value. For example when i want to measure a distance of 1m i get a value of about 55500. Also the measurement change on the same height in a range of 1000. For the offset and xtalk calibration i set the physical distance to 1m and gave that as a parameter in the function as 1000(mm). I copied the values i got from the function to the function setoffset and setxtalk. To calibrate i used the following diagramm:And my init in my software looks like

Hi,We are implementing a compass application using the LIS2MDL magnetometer sensor and experiencing incorrect direction readings due to the sensor's left-handed coordinate system. The raw sensor data is functioning correctly, but standard compass calculation algorithms designed for right-handed coordinate systems produce inaccurate results.Issue Summary:Sensor: LIS2MDL 3-axis magnetometerProblem: Compass calculations yield incorrect directions (primarily showing North/East instead of full 360° range)Root Cause: LIS2MDL uses left-handed coordinate system (confirmed from ST Community forum discussion)Impact: Standard compass algorithms fail to provide accurate bearingsTechnical Details:Raw sensor readings are stable and changing correctlySensor communication and initialization working properly (WHO_AM_I = 0x40)Sample raw data shows expected magnetic field variationsIssue occurs during coordinate system conversion for compass heading calculationsCurrent Environment:Platform: Embedded Linu

Hi,I was wondering if there were plans to make drivers for the ST BrightSense image sensors available for the ST STM32N6 Series μCs (these could be bare metal, FreeRTOS, ThreadX or Zephyr drivers) or for ST STM32MP2 Series SoMs (in the form of libcamera drivers)?I heard that drivers for the VD66GY image sensor are planned for the STM32N6 Series μCs but I wasn't sure about the other BrightSense image sensors.Providing drivers for the ST BrightSense image sensors for the STM32N6 Series μCs and STM32MP2 Series SoCs would allow ST customers to fully leverage the enormous power of both ST image sensors and the leading ISP and AI capabilities of STM32N6 and STM32MP2 while developing a solution entirely within the ST ecosystem. Having drivers provided by ST instead of coding our own drivers would enormously speed route to market and reduce risk.Thank you very much for your help.Will

Does ST supports multiple slave devices for single spi, I want to use mutiple VL53l8cx sensors on single spi bus using multiple chip select. I am facing challenge to run more than two slave devices (VL53l8cx sensors). I am getting failure in vl53l8cx_init() call, means failing to upload firmware in VL53l8cx sensors.

What are the best practices for controlling the exposure time of the VD55G1 in Linux?