search

bolderflight / invensense-imu

Arduino and CMake library for communicating with the InvenSense MPU-6500, MPU-9250 and MPU-9255 nine-axis IMUs.
MIT License
498 stars 211 forks source link

Accelometer Z axis #44

Closed Myykami closed 5 years ago

Myykami commented 5 years ago

The Z axis is not changing its value but I turn it at the right position and it is only changing if I turn it at the X or the Y axis then its the opposit of the x/y value. Do you know something about this problem? (The gyroscop and the magnetometer are working correctly)

Sorry for my bad english but I hope you will understand my problem.

flybrianfly commented 5 years ago

If the Z axis is pointed up or down, you'll read -/+ 9.87 m/s/s. If the X axis or Y axis is pointed up or down, Z will measure 0.

Myykami commented 5 years ago

![Uploading IMG_20190122_151839.jpg…]()

In this position the Z measure -10 and if I turn it like the arrows or in the another direction the Z is not changing anyway but I need this to let the robot turn correctly. Can you help me? Can you open the image?

flybrianfly commented 5 years ago

No, I can't see the image. But you can't use accelerometers to measure heading - only pitch and roll. You'll need to use the magnetometers to measure heading.

Myykami commented 5 years ago

Ohh ok didnt knew that. Thank you for helping me. But I saw videos where people used the accelometer for measuring the heading how is this then possible?

Myykami commented 5 years ago

ok now I tried it with the magnetometer to measure heading but new I have now other problems.

If I turn it to the left the x value goes from 0 to 20 and back but if I turn it to the right it goes from 0 to only -14(less / more).

Is this correct and why does the x value measures heading and not the z?

Myykami commented 5 years ago

is there a way of calibrating the mpu9250?

Myykami commented 5 years ago

and is there really no way to use the accelometer for heading? Cause I need this for turning a robot correctely and the magnetometer is a little bit a problem because if I hold my smartphone near to it the value is completely different than before.

flybrianfly commented 5 years ago

See the calibration methods in this library for calibrating the magnetometer. It needs to be calibrated for it to work properly. Also, any time you change its magnetic environment, you would need to re-do the calibration. Finally, the magnetometer is quite noisy, I recommend a first order IIR filter to smooth the output for use as a compass.

Gravity, which is what the accelerometer is measuring in steady conditions, only acts vertically. There is no way to get heading information from it.