MPU6050 giving constant changing erroneous data after running 15-20 hours

Hi,

I have MPU6050 chip reading euler and acceleration data.

Im running J.Rowbergs DMP MPU library.

After 15-20 hours the device starts giving erroneous angle data that changes rapidly from 0 upwards and back. Disconnecting the sensor from analog pin solves the problem and the device starts to run again properly. Is this a buffer overflow problem?

Thanks in advance

For informed help, please read and follow the directions in the "How to use this forum" post.

Rolfgren:
Hi,

I have MPU6050 chip reading euler and acceleration data.

Im running J.Rowbergs DMP MPU library.

After 15-20 hours the device starts giving erroneous angle data that changes rapidly from 0 upwards and back. Disconnecting the sensor from analog pin solves the problem and the device starts to run again properly. Is this a buffer overflow problem?

Thanks in advance

I’ve experienced similar problems while developing my balancing bot.
Here is the code I’m using. Please let me know if you find any problems with this code I believe I have eliminated the issues your describing with the MPU6050

use the calibration code to get clean values and the latest code for clean data capture
latest code uses J.Rowbergs DMP MPU library.
modify the function MPUMath() to generate euler instead of yaw pitch roll.
Z

MPU6050_Latest_code.ino (7.78 KB)

MPU6050_calibration.ino (7.64 KB)