MEMS (sensors)

can anyone provide answer.

Hi,I trying to get the wrist tilt working. I have used the example in AN5040 page 55.I using the simple example setup, nothing else is set. Write 20h to CTRL1_XLWrite 84h to CTRL10_CWrite 01h to DRDY_PULSE_CFGWhen reading the FUNC_SRC2 it always returns 0x84 which should be impossible according to datasheet. The bit WRIST_TILT_IA never goes high no matter how much the unit is tilted.I have already have accelerometer and gyroscope with FIFO and up and running so my drivers and the chip are working. However this I have disabled and are using only the 3 commands above.What am I missing, any ideas?Thanks

helloI tried to reduce the current consumption of LIS2DH12, but there was always a gap. When I measured the GSENSOR data, the current consumption was 100+uA, while it was only 3uA when not measured. So I want to know how the data in this table is obtained?

I am using​ LSM6DSRTR, and the device-tree compatible is "st,lsm6dsr".I can see two iio device node under​/sys/bus/iio/devices/iio:device1 /sys/bus/iio/devices/iio:device2But when I try to read in a python loop, it gets really slow.​Anyway better to read it? def getData(self, path): fd = open(path, "r") data = fd.read().strip() fd.close() return data def testData(self): iio_oneshot_path = "/sys/bus/iio/devices/iio:device2/in_accel_x_raw" fd = open(iio_oneshot_path, "r") while self.refresh: self.accX = self.getData('/sys/bus/iio/devices/iio:device2/in_accel_x_raw') self.accY = self.getData('/sys/bus/iio/devices/iio:device2/in_accel_y_raw') self.accZ = self.getData('/sys/bus/iio/devices/iio:device2/in_accel_z_raw') self.gyroX = self.getData('/sys/bus/iio/devices/iio:device1/in_anglvel_x_raw') self.gyroY = self.getData('/sys/bus/iio/devices/iio:device1/in_anglvel_y_raw') self.gyroZ = self.getData('/sys

Hi all,I am currently working with STEVAL-MKSBOX1V1 to estimate the angle and 3D orientation of a static device. As I am limited to using the capabilities of the lis2dw12 chip only, I am relying on the accelerometer to provide the forces in the XYZ planes in order to determine the tilt.I am interested in using this functionality, but have a few questions:Can the accelerometer measurements be used accurately for tilt estimation and 3D orientation, and do they remain stable over time without significant drift?If yes, what is the purpose of the 6D orientation functionality described in the data sheet?https://www.st.com/resource/en/design_tip/dt0097-setting-up-6d-orientation-detection-with-sts-mems-accelerometers-stmicroelectroni

Hello,I am working on an IOT project where we use motion sensors. Our prototype works with STM32WB55 + iKS01A3. We would like now to go further and replace iKS01A3 with LSM6DSO and print the PCB to put "between" the STM32WB55 and LSM6DSO. I've read articles on the internet and I think we call it an "adapter card"?However, we are a bit "lost" and we do not know exactly what to do. No tutorial fits to our situation online...Could you help us a little bit? Any advice is welcome :) Thank you,Théo

Hi there!​We are using the LSM6DSOX in our product and doing sensor fusion using xioTechnologies/Fusion.​Our hardware is mass produced and we are using the LSM6DSOX sensor with no magnetometer using 52Hz ODR and 104Hz ODR. The change in ODR doesn't seem to make any difference.​We have calibrated our sensors, especially the gyroscope. ​Using xioTechnologies/Fusion algorithm with calibrated sensor reading removes the yaw drift and gives us very stable measurements over time.​​What we noticed though, is that

I have one topic that is not super clear (or maybe it’s normal and I just don’t know something, I’m new in “sensors�?). Let’s say, range of ACC is set to 2g and we will get highest positive output(32768 in decimal), we can get proper resolution from datasheet (0.061 mg / lsb) and count that is 1998,848mg(~2g) what is less than 0.1 % difference between selected range and output.But if we get same output of gyroscope with range set to 125 dps and multiply it by resolution 4.37(from datasheet too) we get 143196,16 mdps(~143 dps) what is ~15% difference from selected range.What is the reason for this discrepancy between the configured and real range in GYRO?

I've gone through the documentation a few times and I'm not sure what I'm missing. I'm using an STM32L4 to control the lsm303agr through an i2c connection. When I run debug and probe the SCL/SDA lines on my oscilliscope, the messages look good, and from what I am seeing I am getting a slave acknowledge at the end of each transmit.I first noticed when I was collecting data to calculate hard iron compensation values that after I thought I had sent calibration data to the registers that my collected data looked exactly the same. I made a function to read out the OFFSET registers, and the lsm303agr sends back that all 6 offset registers are 0x00 (seen both in debug and oscilliscope)Am I getting the procedure to load the compensation data to the registers incorrect?Magnetometer config registers://Magnetometer configuration //CFG_REG_A_M msg[0] = CFG_REG_A_M; msg[1] = 0x88; HAL_I2C_Master_Transmit(&hi2c1, ACC_ADDR, msg, 2, H

If we use the MCU (STM32F401VE  in STEVAL-MKI109V3) to train our IMU(LSM6DSV16X) for the prediction of certain behaviour.Can we then use the IMU standalone without the MCU for the prediction?If so, Where can those predictions be read from the IMU and in what format?

Hello,Can someone please provide the ECCNs for these two STMicroelectronics products:LSM6DSMUSTRLIS2MDLTR

Hello,I was trying to setup an inactivity/activity program in Arduino using a Xiao Sense NRF52840, that has the LSM6DS3 chip on it, and with using the register i was trying to recognize inactivity and activity. If anyone has insight as to why this isnt working it would be appreciated.#include "LSM6DS3.h" #include "Wire.h"   #define CLEAR_STEP true #define NOT_CLEAR_STEP false   //Create a instance of class LSM6DS3 LSM6DS3 lsm6ds3(I2C_MODE, 0x6A); //I2C device address 0x6A uint16_t detectCount = 0;   void setup() { Serial.begin(9600); while (!Serial); if (lsm6ds3.begin() != 0) { Serial.println("Device error"); } else { Serial.println("Device OK!"); }   if (0 != config_free_fall_detect()) { Serial.println("Fail to configure!"); } else { Serial.println("Success to Configure!"); } }   void loop() { uint8_t readDataByte = 0; //Read the wake-up source register lsm6ds3.readRegister(&readDataByte, LSM6DS3_ACC_GY

I have tried this example lsm6dso_fsm.c (github stm32 STMems_Standard_C_drivers) and successfully trigger the int1 pin but found that comment the line 310 the pin does not trigger.Digging deeper I found that I need to read/poll the EMB_FUNC_STATUS_MAINPAGE (35h) register for the int1 pin to trigger on a fsm event.Is it a known behavior? Is there a way to bypass that and trigger the int1 pin without reading this register?

According to datasheet and AN5005, LIS2DH12 accelerometer can go to sleep & wake up + generate I2_ACT interrupt when acceleration goes below / above the threshold defined in ACT_THS.Which acceleration exactly is taken into account for this strategy? Maximum / minimum of X,Y,Z? Resultant? Is gravity compensated for or not?

I have been testing the internal temperature sensor on the STM32 Nucleo-L452RE recently. I'm trying to measure the ambient temperature using this sensor, thinking that it shouldn't vary more than about 2ºC compared to the internal temperature of the microprocessor, but the readings are always several degrees higher than the temperature measured by external thermometers and sensors, despite following all the steps available in the documentation and many other forum codes for configuration or calibration.For example, I am currently reading the temperature with the code below, which shows approximately 32ºC. However, I am in a room with good ventilation, the air conditioning is on at 24ºC, a thermometer measures 24.5ºC in the room, and an external temperature sensor measures 26ºC above the microprocessor.In addition, when I turn on a LED with a button or do anything else with the board, the temperature increases up to almost 34ºC.I am unsure of what else to read or

Hello,I am using the TL3GD20 gyro in the stm32F429 discovery board to develop my own driver. The output registers are fixed and they do not move, always at:I respect the 10 ms second before powering up.I activate the x y and z axis and set the PD register. I also disabled the high pass filter. I am wondering how to test that the gyro is working. do you have a method to share with me, please?I expect to have the output registers moving if I rotate the board. Despite the values, in the output register, I have to see something in those registers before starting the conversion..Here is a capture to see how I am configuring the TL3GD20 gyro:

Hello,We are going to use LIS2DE12TR accelerometer, in our ultra low power device, to wake on motion. Here is schematic for LIS2DE12TR, Our requirement :1) ultra low power consumption2) Wake MCU on Motion by INT_1 Pin Below are queries regarding schematic:1) What should be value of R48 and R49.2) Is schematic is ok?Thank You

TN0018 calls out in the Process considerations section that the board must be properly cleaned after soldering if "self-cleaning" solder paste is not used. Are there any guidelines on cleaning methods that will not damage the sensor? Can the sensor port be covered with something like Kapton tape to protect it during the cleaning process?

I have a MEMS microphone (IM72D128V01, using this eval kit) with PDM output connected to a STM32U5 NUCLEO board.It's connected to the ADF/Audo Digital Filter peripheral which I have configured to provide a 5th order CIC filter with a 32x decimation factor. ADF clocking is configured to output a 3,072,000Hz data clock to the MEMS sensor. ​Every time I start my STM32 program I have to physically power cycle the MEMS IC or else I get a DC bias in the data register - the most significant byte of the 24bits in the data regist

One of the critical measurements for the analysis of bridges is the thermal response. The 201/109 combination has an average thermal response in the FFT of 1.5 milli-g - it is slightly frequency dependent, but ignoring that the standard deviation means that below 3.5 milli-g is just Gaussian data. Anything over 3.5 is structural response, statistically speaking. The 3.5 milli-g is higher than other accelerometers we use, do you have some device with a tighter response to thermal?

hello,I am using the L3GD20 to measure the angles, I have a 10 ms interrupt and in this interrupt service routine I read each time the 6 Output registers, I am configuring the gyro like this:GyroscopeDriver.TL3GD20_Reg_Struct.Who_Am_I = I_AM_L3GD20; GyroscopeDriver.TL3GD20_Reg_Struct.CTRL_REG1.All = L3GD20_OUTPUT_DATARATE_1|L3GD20_BANDWIDTH_4|L3GD20_MODE_ACTIVE|L3GD20_AXES_ENABLE; GyroscopeDriver.TL3GD20_Reg_Struct.CTRL_REG2.All = L3GD20_HPM_NORMAL_MODE_RES|L3GD20_HPFCF_0|L3GD20_HIGHPASSFILTER_ENABLE; GyroscopeDriver.TL3GD20_Reg_Struct.CTRL_REG3.All = 0; GyroscopeDriver.TL3GD20_Reg_Struct.CTRL_REG4.All = L3GD20_BLE_MSB|L3GD20_FULLSCALE_500|L3GD20_BlockDataUpdate_Continous; GyroscopeDriver.TL3GD20_Reg_Struct.CTRL_REG5.All = 0;The spi Inspector in the following image:

Hi, I'm trying to implement sleep and wake-up function in my project. I read the AN5192 Interrupt generation, but doesn't work. My project use nrf52832 with I2C and init my gyroscope/accelerometer following the procedure below: write_register(LSM6DS0_CTRL1_XL, 0x60);      // Turn on the accelerometer  write_register(LSM6DS0_CTRL2_G, 0x4C);       // Turn on the gyroscope    write_register(LSM6DS0_CTRL8_XL, 0x09);       // Enable slope filterMy function to sleep:void imu_power_down(){ write_register(LSM6DS0_CTRL2_G, 0x00); write_register(LSM6DS0_CTRL1_XL, 0x00);}My function to wake-upvoid imu_wake_up_interrupt(){  write_register(LSM6DS0_WAKE_UP_DUR, 0x00); write_register(LSM6DS0_WAKE_UP_THS

Hello, I recently bought the STEVAL-MKSBOX1V1 card (Sensortile.box). I would like to perform emotion recognition from speech through a convolutional network of my own creation. I downloaded the FP-AI-CTXAWARE1 package and opened the project on STM32 CUBE IDE in order to modify the network called by the package since the preprocessing on the audio wave is the same. I can't figure out how to change this call because in the ai_common.c end (line 511) the ai_run function calls nn_name without ever using it. Can anyone explain to me how this call works or just how to replace the network?