I have spent the last few days trying to debug on why I get "MPU6050 connection fail" with both WriteInstinct.ino and Nybble.ino. I have not been able to calibrate the servos. The I2C switch is set to "Ar" and I have tested the MPU with this code to make sure it working and I get a bunch of scrolling negative numbers that change when ever I rotate the PCB. I don't know if this is a false positive or not, because they don's seem to be rotational angle numbers [what range of numbers should I be getting?] Re-loading the libraries does nothing. I don't know if this is a hardware problem or software. Looking at the "WriteInstinct.ino" code, it seems that setting the MPU to High [line 170, "MPU6050 mpu(0x68)" is the problem? not sure. The reason why i say this is because I ran the i2C_scanner in order to see what connected devices were using the i2c addresses. My output shows that there is one device using 0x68, but i don't know if its the MPU5060 or not.
Also, commenting out the do-while-loop from line 264 in the newest "Nybble.ino" code does nothing when it comes to initializing the MPU6050.
Any help or set of instructions on how to detect that the MPU6050 is working properly will be appreciated in order to fix this problem.
Screencaps:
WritingInstinct:
MPUTest: Trying to calibrate the servos (they have yet to be calibrated)
MPUtest: (Moving PCB, number change but hey are negative and very large)
i2C scanner: Address 0x68 is being used, but i can't confirm that it pertains to the MPU?
Nybble.ino: MPU failed initiation, when uploading the Nybble.ino
Will it be possible to get a step by step video on how exactly one calibrates the servos alone. I know that you have a video on Youtube, but an updated would be appreciated, one that includes the fix for those who have problems with the MPU. Thank you.
I did and it Worked now, thank you😋
I have got this message in my monitor, I could upload write instinct
⸮8⸮8⸮⸮j⸮8(⸮|⸮⸮⸮-8ҽ⸮
what to do I cannot calibrate now.
0x68 is the address of MPU6050.
According to your screen caps, the connection is actually good and showing rotational angle. The LED of the MPU6050 should face up in correct installation. The large negative angles (around -165 degree), are actually close to level (the 180 degree is periodical). I'm not sure why the MPU library is printing them as -165 rather than +15 degree. The calibrated accelY reading seems to be off (-32768 rather than 0. -32768 is the negative max of integer, so it might be overflow in the bytes), but is not used in the algorithm.
So the next thing to validate is uploading Nybble.ino. Ignore the "MPU6050 connection failed" message and attach the limbs close to zero state as described in the Assembly Instruction section 5.3. Use the remote (button "CH+") or serial monitor (enter "kbalance") to trigger the balance state. Nybble should stand up even without careful calibration. Push it from side and Nybble should lean and try to recover. That will prove the MPU6050 is measuring and sending the angle back to the chip.
Here's a new demo video from a user showing how Nybble should react to body rotation.
@Rongzhong Li I will update my post with my outputs later today.
Could you take screenshots of the readings? The more information the better!