Hi, I have bought several chinese boards which claim to be 9250 IMU, however most do not have magnetometer output, ie 6 axis , not 9.
Some reply on 68, others 69.
Having got a couple which do actually communicate and accepting that they are actually 6-axis sensors chips ( suitably scrubbed of identifying marks ), I'm still finding odd reversal of X and Z output.
I'm using the famous i2cdevlib for IO code and checking registers in the .h files it seems 6050 and 9250 use the same addresses for accel and gyro starting at 0x3B, with a temperature word stuck in between, so I was not expecting any surprises.
However, when I dump the data coming back I seem to find X and Z are inverted.
I'm taking X axis as forwards and Yaxis to port, Z (perpendicular to the mini PCB) up in the air.
So I'm expecting rocking the board about X (roll) to give Xgyro changes. I'm seeing Zgyro change. Similarly Z (yaw) is outputting X changes. Y is pitch and is as expected.
Maybe I'm confused about what to expect here, or is this cheap knock-off chinese clones of these famous chips that mix up the register orders?
The MPU-9250 (and the older MPU-6050) were discontinued years ago.
Aside from the fact that a great many chips purchased on the Chinese cheap IC market are outright fakes or recycled from discarded PCBs, there is no point in buying anything associated with those particular IDs.
What would you recommend using that is more recent then? Preferably where I don't need to write a IO library and can just use the same registers and code.
Any more modern sensor from a reliable supplier like Adafruit, Sparkfun or Pololu will outperform those sensors, and libraries for all them are available.