HiLetgo MPU9250 GY-9250 9-Axis 9 DOF 16 Bit Gyroscope Acceleration Magnetic Sensor 9-Axis Attitude +Gyro+Accelerator+Magnetometer Sensor Module IIC/SPI

I order this piece thinking it will piece of cake loading a library in Arduino IDE and get precise position/orientation.First the unit I got was labeled MPU-92/65 which made look for specific libraries. After several tries I gave up and wrote to hiletgo support. The replied in less than 12 hours with a library which was useful. I was able to verify that the unit was working. But some of the examples provided did not work. It made reference to a library in github that was unmaintained. But there, I was able to see that the unit works with the same library as the 9250. So I searched arduino IDE library management and found MPU9250 by hideakitai. Which worked fine and the example got all info from the unit. So, I should have tried that first.Second, there is no easy way to get positions/orientation. You need to do a bit of research on what position really means, so you know what you really want, maybe do some calibrations, take into account that electronics and magnets nearby can change magnetometer readings (really!!), and so you probably will need some way to discard some fluctuations. So if you want to learn look up for a wiki page in github by the user kriswiner regarding MPU6050 and Affordable 9 DoF Sensor.There is a catch, the SPI interface uses most of the same pins as I2C which means that pins are not clearly labeled for SPI.In summary, it works, it have response from customer support and it is cheap... It looks like a good learning experience if you want that. If not, look for some more expensive unit that gives you the results. Adafruit have one absolute position unit that seems really good out of the box.

These two purchases were great disappointments. On the first the gyro was dead, gyro registers contain zero no matter what. On second Z-axis for gyro stuck between -3.96 and -4.01 (as if the chip is spinning around Z-axis at 4 deg/sec clockwise while sitting still and does not change regardless of actual rotation). The temp sensor on second chip is 6 deg F high. Magnetometer readings between the two differ by over 500% in Z-axis.To HiLetgo's credit, after sending in the code, the manufacturer did do a 1/2 refund (they refund 1/2, you keep the bad chip and they send a new -- I'm still waiting on the replacement). I understand there can be bad runs on chips, but QA was not up to par here. (I can just picture the poor university students pulling their hair out for days wondering what they did wrong in their I2C code because they can't get the MPU9250 to respond correctly)Hopefully the third chip actually works. I should not have to build 3 sensors just to finally get one working sensor. If everything is working, this is a great little 9-DOF + temp sensor. Simple I2C interface, fast and configurable gyro rates, and would be a good value for any project. However, with any project relying on accurate position and orientation, a chip with a bad gyro, in even just one axis -- is worthless.

Fought a while to get my board working. Turns out you need to connect EDA to SDA and ECL to SCL to access the AK8963 which is the magnetic compass built into the mpu9250. After this I could see the address for the AK8963 (0x0c) and the mpu920 (0x68) but could not read data. I had to add a 4.7 pull up resistor to the sda and scl lines before I could read data. Hopefully this saves someone some time.

Worked well with Arduino and ROS. Magnetometer is garbage but that’s to be expected for this price/size.I ended up going with the sparkfun razor imu for my final product (buy from sparkfun NOT amazon) for ease of integration (it had a micro-usb) and documentation but they have the same chips so theoretically you could use the same libraries.

I purchased this item to use on my klipper installation on my Ender 3 pro, I did struggled for quite some time trying to get this working and I was not able to get it working over I2C, Klipper docs don't give you much info on this and what I end up doing is using an Atmega328p which in other words is an Arduino Uno board, so I had to flash the klipper firmware to the UNO and then connect it to my RPI 4B running my klipper install by USB and it worked, not the best way to do it and luckily I had and Arduino UNO board on hand.Device does work as advertise but I just wanted to let my Klipper friends out there know that they will be better of purchasing the ADXLs which I did and it was way easier and better connecting them directly to the PI4B.

For a school project we built an EFIS (Electronic flight instrumentation system) and needed a capable IMU (Inertial measurement unit.) I was hesitant picking up a 9 DOF after working with several MPU 6050 units thinking that the magnetometer addition would be a bit much. It worked well for our compass element in our EFIS. Of course we compensated the reading to true north using a GPS module readout as well.The real surprise to me was that a lot of my code that I used for the MPU 6050 translated to this unit perfectly. It makes sense why it would work but there was concern within our group when starting development. This unit was also my first experience in manually programming the unit itself using the I2C protocol. It was very forgiving to my mistakes.If you want a consistent reading and a reliable piece, this is your part.

Sometimes it boots with the SCL/SDA I2C bus showing the MCU9250 and the AK8963. Yay I2C bridge!Sometimes it boots with a broken I2C bridge and the AK8963 I2C is off on ECL/EDA. But you can't keep a manual bridge of the ECL/EDA to SCL/SDA wired up because when the integrated bridge decides to work, ECL/EDA trashes the bus.The measurements are reasonable though, that's why I give it an additional star.