I am working with my balancing robot project using the gyro. The two motors are the same but not in performance. The robot rotates. So, I use the gyro Z angle to keep the same direction. This is not working for me. The gyro calibrates with 500 loops to get the drift. Not good enough. So, I iterate over 5000 loops. I still get a drift.
Is it imposible to compensate for drift?
Is the gyro accurate enough to pick up on the rotation of the Earth.
Would it read by defalt one rotation per day?
Are you used to aviation gyro systems? Because the MPU-6050 is a MEMS sensor, it is noisy and drifts and it is not even a real gyro. It is by far not accurate enough to measure the rotation of the earth or measure one rotation per day. One rotation per minute is hard to do, because it is not real gyro.
The MPU-6050 3-axis gyro detects a change in rotation. If you hold the sensor in your hand and give it a quick turn, then the acceleration and deceleration of the turn will give output.
A magnetometer (compass) is used for the heading.