MPU9250 Self test failed

[sdrouet]

Hello Kris,

First of all thanks a lot for your great work. I am currently using the sketch PU9250BasicAHRS and I am struggling with the Self test which is far away from what it should be

Here are the values:

x-axis self test: acceleration trim within : 1.7% of factory value y-axis self test: acceleration trim within : -100.0% of factory value z-axis self test: acceleration trim within : 2.0% of factory value x-axis self test: gyration trim within : 20.5% of factory value y-axis self test: gyration trim within : -1.7% of factory value z-axis self test: gyration trim within : 0.6% of factory value MPU9250 initialized for active data mode.... AK8963 I AM 48 I should be 48 AK8963 initialized for active data mode.... Mag Calibration: Wave device in a figure eight until done! mag x min/max: 30 -1 mag y min/max: 263 232 mag z min/max: 179 147 Mag Calibration done! AK8963 mag biases (mG) 25.01, 442.68, 281.63 AK8963 mag scale (mG) 1.02, 1.02, 0.96

Output of my MPU9250: [ ]

How can I get the calibration done and corrected in order to get true pitch,roll and yaw values?

Thanks in advance for your answer.

[kriswiner]

Do you have this in your sketch?

magbias[0] = +470.; // User environmental x-axis correction in milliGauss, should be automatically calculated magbias[1] = +120.; // User environmental x-axis correction in milliGauss magbias[2] = +125.; // User environmental x-axis correction in milliGauss

then remove this or use you particular settings.

How do you know the calibration is right/wrong?

On Thu, Apr 23, 2020 at 8:08 AM sdrouet notifications@github.com wrote:

Hello Kris,

First of all thanks a lot for your great work. I am currently using the sketch PU9250BasicAHRS and I am struggling with the Self test which is far away from what it should be

Here are the values:

x-axis self test: acceleration trim within : 1.7% of factory value y-axis self test: acceleration trim within : -100.0% of factory value z-axis self test: acceleration trim within : 2.0% of factory value x-axis self test: gyration trim within : 20.5% of factory value y-axis self test: gyration trim within : -1.7% of factory value z-axis self test: gyration trim within : 0.6% of factory value MPU9250 initialized for active data mode.... AK8963 I AM 48 I should be 48 AK8963 initialized for active data mode.... Mag Calibration: Wave device in a figure eight until done! mag x min/max: 30 -1 mag y min/max: 263 232 mag z min/max: 179 147 Mag Calibration done! AK8963 mag biases (mG) 25.01, 442.68, 281.63 AK8963 mag scale (mG) 1.02, 1.02, 0.96

Output of my MPU9250: [ [image: Capture_yaw_pitch_roll] https://user-images.githubusercontent.com/12115126/80115149-ae059480-8584-11ea-9261-fde8c55a485e.PNG ]

How can I get the calibration done and corrected in order to get true pitch,roll and yaw values?

Thanks in advance for your answer.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKX5XYFSLRLC36UZZWTROBKXLANCNFSM4MPEJY6Q .

[sdrouet]

For you first question: I had it but I removed it. I added the function void magcalMPU9250(float dest1, float dest2) in order to get the magbias corrected with my value. Unfortunately it didn't change a thing.

As you can see on the output I have: AK8963 mag biases (mG) 25.01, 442.68, 281.63

To answer your second question: From what I understand on the MPU9250SelfTest function it is written on comment: +/- 14 or less deviation is a pass. So I presume that mine is a failed one as I have more than that.

Furthermore when I move my MPU9250 it gives me really crap values....

[kriswiner]

The self test is implemented incorrectly. If you look at the self test document you will see that only the 8 LSbits are to be used. This is almost never thre cause of poor results.

Poor calibration is responsible for 99% of the poor results people see.

On this topic, the calibration values you output do not seem to be correct

On Mz you see:

mag z min/max: 179 147

this is unlikely if you are moving the board in all possible directions and not just spinning the board on the table, etc.

On Thu, Apr 23, 2020 at 9:13 AM sdrouet notifications@github.com wrote:

For you first question: I had it but I removed it. I added the function void magcalMPU9250(float dest1, float dest2) in order to get the magbias corrected with my value. Unfortunately it didn't change a thing.

As you can see on the output I have: AK8963 mag biases (mG) 25.01, 442.68, 281.63

To answer your second question: From what I understand on the MPU9250SelfTest function it is written on comment: +/- 14 or less deviation is a pass. So I presume that mine is a failed one as I have more than that.

Furthermore when I move my MPU9250 it gives me really crap values....

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415#issuecomment-618492180, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKVV3UTGJYJ3L25KAJLROBSLDANCNFSM4MPEJY6Q .

[sdrouet]

So how can I implement it correctly? Can I post my code and you pin point my mistake?

[kriswiner]

Read the self test document and fix it yourself. Or look at some of the later sketches, but really, self test is not that important. Your sensor is fine.

I would pay more attention to your calibration, something not right.

On Thu, Apr 23, 2020 at 9:25 AM sdrouet notifications@github.com wrote:

So how can I implement it correctly? Can I post my code and you pin point my mistake?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415#issuecomment-618498505, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKRUNO3FBVIU2X2U3IDROBTWDANCNFSM4MPEJY6Q .

[sdrouet]

So on the calibrateMPU9250 function? I changed nothing on that function so I don't understand why it works for others when it doesn't for me?

[kriswiner]

Pilot error?

Check your results.

Collect a bunch of data on the serial monitor and plot Mx vs. My, My vs. Mz, and Mx vs. Mz. Are the points circular and centered on the origin after calibration?

https://github.com/kriswiner/MPU6050/wiki/Simple-and-Effective-Magnetometer-Calibration

On Thu, Apr 23, 2020 at 9:37 AM sdrouet notifications@github.com wrote:

So on the calibrateMPU9250 function? I changed nothing on that function so I don't understand why it works for others when it doesn't for me?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415#issuecomment-618505180, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKQQTDKZHRDD2IP7DDLROBVEZANCNFSM4MPEJY6Q .

[sdrouet]

Thanks a lot I will take a look and investiguate

[sdrouet]

Hello Kris,

I took a look on your link and indeed my MPU was uncalibrated. After some tests I manually changed the Mx, My and Mz value.

I also corrected my declination according to my location.

Unfortunately nothing really changed so far. I am using the MahonyQuaternionUpdate function btw. Here what I got:

Flat surface:

Yaw with a 360 degree turn:

Pitch with a 360 degree turn:

Roll with a 360 degree turn:

if you have any clue it would be really appreciated. Thanks again

[sdrouet]

just to add a comment my Y-acceleration is always stuck at 2000 mg and I guess it is due to the y-axis self test: acceleration trim which is within : -100.0% of factory value as stated in my first comment.

Is there any possibilities to change that value? If I am able to change it, will it solve the problem?

X-acceleration: 159.24 mg Y-acceleration: 1999.94 mg Z-acceleration: 1000.67 mg X-gyro rate: -0.092 degrees/sec Y-gyro rate: -0.069 degrees/sec Z-gyro rate: 0.031 degrees/sec X-mag field: -256.12 mG Y-mag field: -35.88 mG Z-mag field: 393.73 mG Temperature is 25.9 degrees C X-acceleration: 158.45 mg Y-acceleration: 1999.94 mg Z-acceleration: 1002.08 mg X-gyro rate: -0.130 degrees/sec Y-gyro rate: -0.023 degrees/sec Z-gyro rate: 0.000 degrees/sec X-mag field: -256.12 mG Y-mag field: -35.88 mG Z-mag field: 390.27 mG Temperature is 26.0 degrees C X-acceleration: 160.58 mg Y-acceleration: 1999.94 mg Z-acceleration: 1001.46 mg X-gyro rate: -0.069 degrees/sec Y-gyro rate: -0.038 degrees/sec Z-gyro rate: -0.015 degrees/sec X-mag field: -248.98 mG Y-mag field: -50.21 mG Z-mag field: 383.36 mG Temperature is 25.9 degrees C X-acceleration: 157.53 mg Y-acceleration: 1999.94 mg Z-acceleration: 998.41 mg X-gyro rate: -0.069 degrees/sec Y-gyro rate: 0.023 degrees/sec Z-gyro rate: 0.023 degrees/sec X-mag field: -266.84 mG Y-mag field: -46.63 mG Z-mag field: 390.27 mG Temperature is 25.9 degrees C

[kriswiner]

There is something very wrong with your accel values. I would take a look at the raw bytes without any calibration or scaling applied and see if they make any sense.

I wouldn't relay on the self test results, but accel values at rest of 2000 mgs is just wrong.

On Fri, Apr 24, 2020 at 7:17 AM sdrouet notifications@github.com wrote:

just to add a comment my Y-acceleration is always stuck at 2000 mg and I guess it is due to the y-axis self test: acceleration trim which is within : -100.0% of factory value as stated in my first comment.

Is there any possibilities to change that value? If I am able to change it, will it solve the problem?

X-acceleration: 159.24 mg Y-acceleration: 1999.94 mg Z-acceleration: 1000.67 mg X-gyro rate: -0.092 degrees/sec Y-gyro rate: -0.069 degrees/sec Z-gyro rate: 0.031 degrees/sec X-mag field: -256.12 mG Y-mag field: -35.88 mG Z-mag field: 393.73 mG Temperature is 25.9 degrees C X-acceleration: 158.45 mg Y-acceleration: 1999.94 mg Z-acceleration: 1002.08 mg X-gyro rate: -0.130 degrees/sec Y-gyro rate: -0.023 degrees/sec Z-gyro rate: 0.000 degrees/sec X-mag field: -256.12 mG Y-mag field: -35.88 mG Z-mag field: 390.27 mG Temperature is 26.0 degrees C X-acceleration: 160.58 mg Y-acceleration: 1999.94 mg Z-acceleration: 1001.46 mg X-gyro rate: -0.069 degrees/sec Y-gyro rate: -0.038 degrees/sec Z-gyro rate: -0.015 degrees/sec X-mag field: -248.98 mG Y-mag field: -50.21 mG Z-mag field: 383.36 mG Temperature is 25.9 degrees C X-acceleration: 157.53 mg Y-acceleration: 1999.94 mg Z-acceleration: 998.41 mg X-gyro rate: -0.069 degrees/sec Y-gyro rate: 0.023 degrees/sec Z-gyro rate: 0.023 degrees/sec X-mag field: -266.84 mG Y-mag field: -46.63 mG Z-mag field: 390.27 mG Temperature is 25.9 degrees C

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415#issuecomment-619037851, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKXTXMHWWOHRCJL3C7TROGNPPANCNFSM4MPEJY6Q .

[sdrouet]

Y-acceleration - raw value: 32767

So I don't understand what is wrong and why i have this value

I am using your function in order to get the raw value void readAccelData(int16_t * destination) { uint8_t rawData[6]; // x/y/z accel register data stored here readBytes(MPU9250_ADDRESS, ACCEL_XOUT_H, 6, &rawData[0]); // Read the six raw data registers into data array destination[0] = ((int16_t)rawData[0] << 8) | rawData[1] ; // Turn the MSB and LSB into a signed 16-bit value destination[1] = ((int16_t)rawData[2] << 8) | rawData[3] ;
destination[2] = ((int16_t)rawData[4] << 8) | rawData[5] ; }

[kriswiner]

Looks like your MPU is busted. But what are the individual bytes, rawData[2] and rawaDta[3]?

On Fri, Apr 24, 2020 at 8:36 AM sdrouet notifications@github.com wrote:

Y-acceleration - raw value: 32767

So I don't understand what is wrong and why i have this value

I am using your function in order to get the raw value void readAccelData(int16_t * destination) { uint8_t rawData[6]; // x/y/z accel register data stored here readBytes(MPU9250_ADDRESS, ACCEL_XOUT_H, 6, &rawData[0]); // Read the six raw data registers into data array destination[0] = ((int16_t)rawData[0] << 8) | rawData[1] ; // Turn the MSB and LSB into a signed 16-bit value destination[1] = ((int16_t)rawData[2] << 8) | rawData[3] ; destination[2] = ((int16_t)rawData[4] << 8) | rawData[5] ; }

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415#issuecomment-619085472, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKUQGRMBZ3B7PSCRIKLROGWYTANCNFSM4MPEJY6Q .

[sdrouet]

rawData[3]: 255 rawData[2]: 127

[kriswiner]

These should be ~0, so I would say your board is hosed. Try another one.

On Fri, Apr 24, 2020 at 8:57 AM sdrouet notifications@github.com wrote:

rawData[3]: 255 rawData[2]: 127

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415#issuecomment-619097278, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKWZA7Y5PUE3F2CZYI3ROGZEZANCNFSM4MPEJY6Q .

[sdrouet]

fortunately i bought 2 of them. I come back to you with my result.

Thanks again for your time

[sdrouet]

Second try with a second MPU9250.... And still bad results... I really don't know what I am doing wrong I got it perfectly calibrated.

Self test: x-axis self test: acceleration trim within : 0.6% of factory value y-axis self test: acceleration trim within : -0.4% of factory value z-axis self test: acceleration trim within : 4.2% of factory value x-axis self test: gyration trim within : 69.3% of factory value y-axis self test: gyration trim within : 20.9% of factory value z-axis self test: gyration trim within : -100.1% of factory value

On flat surface: X-acceleration: -760.86 mg Y-acceleration: -2.14 mg Z-acceleration: 994.69 mg X-gyro rate: -0.084 degrees/sec Y-gyro rate: 0.046 degrees/sec Z-gyro rate: 0.084 degrees/sec X-mag field: 122.73 mG Y-mag field: 70.74 mG Z-mag field: 401.45 mG

X-acceleration raw value: -0.76

Dont know why the X-accel is so big when it should be close to 0.

Does it mean that my other MPU is broken?

[sdrouet]

sorry the X accel raw value is X-acceleration raw value: -12479

[kriswiner]

Maybe, why don;t you try to calibrate you accel and gyro...

On Fri, Apr 24, 2020 at 1:39 PM sdrouet notifications@github.com wrote:

Second try with a second MPU9250.... And still bad results... I really don't know what I am doing wrong I got it perfectly calibrated. [image: Mx My Mz calibrated] https://user-images.githubusercontent.com/12115126/80254232-19ca2900-867b-11ea-9853-4cf743e06b89.png

Self test: x-axis self test: acceleration trim within : 0.6% of factory value y-axis self test: acceleration trim within : -0.4% of factory value z-axis self test: acceleration trim within : 4.2% of factory value x-axis self test: gyration trim within : 69.3% of factory value y-axis self test: gyration trim within : 20.9% of factory value z-axis self test: gyration trim within : -100.1% of factory value

On flat surface: X-acceleration: -760.86 mg Y-acceleration: -2.14 mg Z-acceleration: 994.69 mg X-gyro rate: -0.084 degrees/sec Y-gyro rate: 0.046 degrees/sec Z-gyro rate: 0.084 degrees/sec X-mag field: 122.73 mG Y-mag field: 70.74 mG Z-mag field: 401.45 mG

X-acceleration raw value: -0.76

Dont know why the X-accel is so big when it should be close to 0.

Does it mean that my other MPU is broken?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415#issuecomment-619228410, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKVBJRVWOBPJW3FPEWLROH2GZANCNFSM4MPEJY6Q .

[sdrouet]

I finally got it works and it was definitely the X-accel which was wrong.

I don't understand why the calibrateMPU9250 didn't do its duty correctly. I had to correct it somehow manually accel_bias_reg[0] -= (accel_bias[0]/8) -1400; // Subtract calculated averaged accelerometer bias scaled to 2048 LSB/g (16 g full scale) accel_bias_reg[1] -= (accel_bias[1]/8); accel_bias_reg[2] -= (accel_bias[2]/8);

Do you have any idea why this function doesn't manage to make it correctly?

Thanks a lot again for your time

[kriswiner]

Probably some error or other in the sketch you are using...or the way your MCU interprets int16_t, etc

On Fri, Apr 24, 2020 at 3:10 PM sdrouet notifications@github.com wrote:

I finally got it works and it was definitely the X-accel which was wrong.

I don't understand why the calibrateMPU9250 didn't do its duty correctly. I had to correct it somehow manually accel_bias_reg[0] -= (accel_bias[0]/8) -1400; // Subtract calculated averaged accelerometer bias scaled to 2048 LSB/g (16 g full scale) accel_bias_reg[1] -= (accel_bias[1]/8); accel_bias_reg[2] -= (accel_bias[2]/8);

Do you have any idea why this function doesn't manage to make it correctly?

Thanks a lot again for your time

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/415#issuecomment-619259766, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKW3OXKMYXGK46MLTE3ROIE5JANCNFSM4MPEJY6Q .