MEMS and sensors

Hi,I have installed the v4l2 driver for the STEVAL-56G3MAI1 camera module on ubuntu 22.04.05 LTS on a RPI CM4. But I am unable to access the module through /dev/video0.I have followed the instructions in the linux start guide fro ST BrightSense but I cannot get it to work.I am not sure if I need to do something in addition to these instructions:cd en.vd56g3_1.0.3/srcmakesudo cp *.ko /lib/modules/$(uname -r)sudo depmod -asudo rebootsudo apt updatesudo apt install device-tree-compilercd en.vd56g3_1.0.3/dts/rpi1_to_4/sudo dtc pcb4188_vd56g3.dts -o /boot/firmware/overlays/pcb4188_vd56g3.dtbosudo sh -c "echo 'dtoverlay=pcb4188_vd56g3' >> /boot/firmware/config.txt"sudo reboot v4l2-ctl --list-devices gives:unicam (platform:fe801000.csi):/dev/media3 Cannot open device /dev/video0, exiting. Is there something I am missing in the setup? Should the driver be placed in a different folder?Thanks for any assistance!

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?Any suggestions or community experience on low-frequency analysis with this sensor?

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. I still imported that code block into my project as an additional initialization function :void AIS2DW12_FallDetection_Init(void) {Ais2dw_ctx.write_reg = platform_ASW12_write;Ais2dw_ctx.read_reg = platform_ASW12_read;Ais2dw_ctx.mdelay = platform_ASW12_delay;Ais2dw_ctx.handle = &hi2c3;uint8_t whoamI, rst;ais2dw12_reg_t int_route;ais2dw12_device_id_get(&Ais2

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 both devices recommend

Hello.We're having issues after installing the libcamera package for the VD56G3 sensor from https://www.st.com/en/embedded-software/stsw-img502.htmlThe ST IPA is (apparently) being treated as untrusted and we're being forced into IPA isolation, crashing the application.[10:02:34.005492925] [181325] ERROR IPCPipe ipc_pipe_unixsocket.cpp:134 Call timeout! [10:02:34.005523463] [181325] ERROR IPCPipe ipc_pipe_unixsocket.cpp:66 Failed to call sync [10:02:34.005531333] [181325] ERROR IPAProxy raspberrypi_ipa_proxy.cpp:261 Failed to call start [10:02:34.009190798] [181325] ERROR IPCUnixSocket ipc_unixsocket.cpp:192 Failed to send: Connection refused [10:02:34.009221799] [181325] ERROR IPCPipe ipc_pipe_unixsocket.cpp:66 Failed to call sync [10:02:34.009228836] [181325] ERROR IPAProxy raspberrypi_ipa_proxy.cpp:432 Failed to call mapBuffers [10:02:34.067543539] [181325] ERROR IPCUnixSocket ipc_unixsocket.cpp:192 Failed to send: Transport endpoint is not connected [10:02:34.067573243] [18132

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 I have specified above. Therefore, is it possible to either disa

Hi,I was wondering how things were progressing with writing and possibly mainlining Linux drivers for ST BrightSense and SafeSense Industrial, Consumer and Automotive Image Sensors?We're working on a new state-of-the-art open source system for conservation of Europe's most endangered dormouse, squirrel and bat species using libcamera and embedded Linux and would like to look at integrating ST image sensors.More about our work:https://new-homes-for-old-friends.cairnwater.com/and more detailed analysis of how user requirements translate to implementation decisions:https://github.com/William-Robert-Robertson/WildCameraThank you very much,Will

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_AM_I), to which the IMU responds with

We use the VL53L1X in one of our products and have been doing that for over 2+ years now. As this device is not easily solderable, we have been relying on the Chinese supply of breakout boards used in the hobby market.  Just recently we are facing range issues with entire lots of this sensor based boards. All of them exhibit range issues where the Sensor is unable to detect beyond 30 to 40cm. I thought it could be a calibration or offset issue. I've tried getting offset values using the 17% Gray sheets as recommended. Any help in this regard will be appreciated. Thanks.  

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 you for any feedback.  

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 from those bits.&nbs

I'm using a VL53L4CD as a proximity switch but with a user selectable threshold. I began developing on XNUCLEO-53L4A1 on a NUCLEO-F401 but have since migrated to my own pcb using STM32C071. I am having difficulty with poor status reports from the ToF device at distances above about 300mm on my own board. Under 300mm I get a status of 0 but anything above that and it quickly deteriorates to 1, 2 or 4 and the distance becomes very erratic.Thinking that it was down to the specifics of the pcb I reverted to the F401/XNUCLEO arrangement and I now see it's not that much better and certainly nowhere near 1.2m. I can get to about 500mm with status 0 but much beyond that leads to a status of 1 or 2 and eventually 4. My maximum range is 500mm so it's barely okay on the F401 but not on my own pcb.I'm using the standard API with some mods which don't amount to very much (threshold hysteresis mainly). It doesn't seem to matter whether the (standard) cover glass is fitted or not. The application is

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#define y_h 0x2B#define z_l 0x2C#define z_h 0x2Duint8_t addr;uint8_t outx_l;uint8_t outx_h;uint8_t outy_l;uint8_t outy_h;uint8_t outz_l;uint8_t outz_h;uint16_t X;uint16_t Y;uint16_t Z;uint16_t msb;uint8_t data;void setup() {  // put your setup code here, to run once:  Serial.begin(115200);  Serial.println("\nI2C Scanner");  delay(1000);  Serial.println ("I2C scanner.

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 pulling <50uA and adds to the power rail capacita

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 from application note AN5648 for the LIS2DU12, it uses a slope filter that is connected upstream of the

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 someone share with me the firmware upgr

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?--Edit--I'm using SPI mode on the accelerometer.Also, I forced

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

To my knowledge all the LSM6DSV sensors are 3rd gen sensors.And they have additional features to prior sensors like lower power consumption, high accuracy ODR, dual channel output mode, and the LSM6DSV80X/320X even have a second high g accel channel, a few more options/modes.Looking through all the industrial and automotive sensors, the feature set mostly aligns with the LSM6DSR/DSO variants. So I may have overlooked some, but in case not, when can we expect 3rd gen IMUs with longevity commitment and/or 105°C temperature range?

Using the LSM6 over SPI with the RP2040 Microcontroller for a mission-critical project (flight controller for UAVs).The SPI works well. Well in theory. In practice, 1 in every 20-100 reads returns corrupt SPI data (WHOAMI reads as 0x68/0x58). This issue is somewhat reduced when moving the sensor (worst when still) and when adding decoupling capacitors (reduces corruption dramatically).The project itself is ultra high volume (10k+ units), and we have already invested a very large amount of time and resources into testing.

I try  STSW-IMG037 software with  Nucleo STM32-F401RE + Expansion board with VL53L7CX+ two Brakout Boards(VL53L7CX).but it dosen't work. STSW-IMG037 software with  Nucleo STM32-F401RE + Expansion board with VL53L7CX is OK.Soft ver v2.2.29327. and 53L7A1_MultiSensorRanging.bin in X-CUBE-TOF1 is OK.What should I do?