I was testing my MPU6050 with Raspberry PI and got accelerometer(X,Y,Z) and gyroscope(X,Y,Z) data forming on registers at decentely high sample rate. Looks like I've accidently fed +5v from RPI to SDA or SCL the other day and weird things started to happen with MPU6050. Looks like gyroscope got fried. Accelerometer still respond, but at very slow rate. It changes it's value about once a second no matter what clock source I set. Gyroscope stopped responding at all and writes 0 to each axis all the time. Did my MPU6050 got fried? Looks like RPI's current is fatal for i2c?
The MPU may be voltage sensitive, but most break out boards I have seen are ok at 5V. I2C itself is a current sync protocol, meaning it will generally work at any voltage, having a signal pulled to vcc and then switched to ground is what is important, not what the voltage is. In fact you only need to level shift if you have other devices on the i2c line that are voltage sensitive.
You really need to look closely at code changes, but indeed it is possible something fried. If you are really using a 6050 chip be aware you can get entire breakout boards cheaper on amazon (I know just bought a bunch myself, will be switching to using daughter cards next PCB revision as it is cheaper. 6$ and 2-day delivery, hard to argue against).
Thanks, I've figured out what problem was and that was a constantly resetting MPU6050 mode caused by wrong code. Mar 15, 2015 at 14:41
According to the MPU-6050 datasheet it accepts voltages in the range 2.375V-3.46V.
Supplying one of the pins with 5V could indeed have damaged the sensor.