MPU-6050 gyro calibration

hello,
i'm playing with a mpu6050 micro-
i use the i2cdevlib mpu6050 library.

i've a problem with the gyro
even if the chip is not moving, i alwars read a small amount of gyro.
@ 2000 deg/sec raw data on X axis is between 41 and 49, Y -13 and -7, Z 27 and 31
to prevent this error, i try to calibrate this by hand substracting ia middle value to the read data
XOFFSET -42 / YOFFSET 9 / ZOFFSET -29
things goes better, but it always move a little, using DMP or mahony alghoritm.

i've only 1 chip, so i can not test this with other mpu6050.
is this a problem of my chip, or i've to calibrate the gyro, or i'm i wrong in other way.

tks :wink:

The MPU-6050 has a +/- 20 degrees/second tolerance for zero rate, but this is normal for any gyro chip.

You should be using the accelerometer readings in combination with the gyro for a zero point of reference.

Chagrin:
The MPU-6050 has a +/- 20 degrees/second tolerance for zero rate, but this is normal for any gyro chip.

You should be using the accelerometer readings in combination with the gyro for a zero point of reference.

thank you.
the internal DMP of this chip should already combine gyro and accell, or i'm i wrong?
i've made some screenshot of the behavior 16.50 / 19.20 / 20.50 time (chip is in the same position, i've not moved it)

Cattura_16.50.JPG

Cattura_19.20.JPG

Cattura_20.50.JPG

About the photos in this thread, they look like a program that takes the YPR and simulates it on your computer. Is there any kind of free programs that would do this for me on my Mac, using the MPU-6050? I've searched for motion tracking software with no luck, that doesn't seem to be the right wording...

pictures here are taken from processing.org (there is a mac version too).

for calibration, you can find the python script i've built here:
http://davidegironi.blogspot.it/2013/01/gyroscope-calibration-helper-01-for-avr.html
it should works on mac too :wink:

Thanks! I've heard of processing but I never knew what it was.