SmartMotion® MEMS Motion Sensors

Hello

I just have a question concerning the eMPL library code. The function is mpu_get_int_status in the inv_mpu.c file. Is this function part of the standard API? there isn't much documentation describing the functionality. I do not want to connect the interrupt pin to a microcontroller, instead I'd prefer to query the MPU 9250 at a specific time interval for data ready (before reading the FIFO). Is this function meant to provide a way for the software to determine when data is ready to be read from the FIFO without having to use a hardware interrupt?

Thanks

Hi, I'm working with the brand new ICM-20948 but I can't find it on the website and there is not library or sample code linked to it.

I've seen during my tests that it works with the MPU9250 code for reading accel/gyro/magnet registers but I want to use the DMP(Digital Motion Processor), and it looks like the code that I was using with the MPU9250 isn't working with the ICM-20948.

Do you have a register map? A sample code?

Thanks in advance,
AB

Hi all!

We are developing a device (approx. 500 units, at first) which uses a MPU-6050 six-axis gyro/accel. We faced a problem with actual Linux kernel driver. The driver that comes with kernel is incompatible with devicetree structure.
Where can we get a devicetree-compatible driver?

Thank you, best regards.
Ivan Nikolaenko.

Hi. I'm using MPU9250 and I want to know more about DMP. but I cannot find "Programming Sequence for DMP Hardware Functions" application note.
please let me know how to get it.

Many thanks in advance.

Product: ICM 30630

Hello,

I am using Sensor Studio and following the basic tutorials such as Door Opening Detection. I have created a Custom Sensor and am now modifying the Impl code using the API documentation provided. When I try to add API GPIO calls to the Door Open Detection reference design I get unexpected errors. The steps I've followed:

1) I start with successfully compiled Door Opening Detection code.

2) I then include SensorHubExt

#include "SensorHubExt.h"

3) I create "myGPIO" of type inv_shext_gpio_struct_t

I try to operate two MPU-9250 via I2C. Somehow, it is not possible for me to access the AK8963 magnetometer of the MPU with AD0-address pin on high. I tried all possible addresses according to the datasheet of the AK8963, but none worked (could not read the wai register correctly). The other MPU with AD0 on low works fine.

Is it possible to accesse the AK8963 with the AD0 on high? Otherwise, I can just think of toggeling AD0, configuring the magnetometer to send the measurment values to FIFO and disable the bypass. Or is there a more elegant way?

Hi to all,

We are developing a sensor based on the TI SensorTag schematics with the MPU9250 to measure small angular rotations on an actuator. We already developed a fair deal of the firmware, but would like to improve our measurements. We would like to measure angular displacement from 0° to 6° with an accuracy of 0.1°. The actuator is pressed by humans.

Hi,

We are using the TI Sensortag 2 for a research with the MPU-9250. But when we apply an amount of force, let's say, drop it from a certain distance of the ground, the 9250 stops to respond and we need to wake it again.

Is this a MPU-9250 limitation, a feature we could turn off or it is a Ti Sensortag problem? If it is a TI problem, could you point some other "plug-and-play" board that we could use with this chip? We need just the accelerometer and the bluetooth to get its data in "real time".

Hi,

I have a Texas Instruments CC2650 with Debugger Devpack, and I use CCS Version 6.1.2.00015. I am trying to use the mpu9250 (built-in accelerometer) and I already managed to initialize the accelerometer configure a callback for the interrupts, and execute the code of the callback when I enable the Wake on Motion feature. However, when I read the data from the accelerometer I don't get coherent values.

hello,

I am using the Motion Driver 6.12 whith MPU9250. And I'm facing problems with changing the orientation matrix (inv_set_accel_orientation_and_scale()). Actually I observe that the Y and Z axis are sometime totaly wrong for one sample and after it goes with the good values. It appears randomly (doesn't seem to be at a specific period).