search

kriswiner / MPU9250

Arduino sketches for MPU9250 9DoF with AHRS sensor fusion
1.03k stars 472 forks source link

Calibration before final mount #392

Open Rsambeek opened 4 years ago

Rsambeek commented 4 years ago

Hi,

I am trying to use the mpu9250 to get orientation and compass data. Its all good in my test setup but eventually I have to mount this in a plane and I don't think it will be easy to rotate the plane to calibrate the compass. Is there another way I can calibrate it beforehand and apply some sort of offset to get the somewhat correct calibration.

Thanks for all the work you have already put in this it has helped me a lot with this project.

kriswiner commented 4 years ago

You can calibrate the sensors before hand.

This should work pretty well.

But yes, the plane environment will affect the offset bias and different locations will cause the magnetic declination to change. The latter can be adjusted with a lookup table and GPS. The former requires some kind of magnetic anomaly detection/correction. The algorithm should detect offset (without rotating the plane) and correct them automatically. We have such an algorithm but it is not open source.

On Thu, Oct 10, 2019 at 1:39 AM Rick van Sambeek notifications@github.com wrote:

Hi,

I am trying to use the mpu9250 to get orientation and compass data. Its all good in my test setup but eventually I have to mount this in a plane and I don't think it will be easy to rotate the plane to calibrate the compass. Is there another way I can calibrate it beforehand and apply some sort of offset to get the somewhat correct calibration.

Thanks for all the work you have already put in this it has helped me a lot with this project.

— 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/392?email_source=notifications&email_token=ABTDLKT3HWIDFW7VASG3BYLQN3S5TA5CNFSM4I7JWC72YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HQ3R2MQ, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKTXCCQAHZA62SNDSRDQN3S5TANCNFSM4I7JWC7Q .

Rsambeek commented 4 years ago

Thanks for the response. I was thinking of such an algorithm and was hoping for an easier solution but it seems like there isn't an easy fix. I guess I should start programming that then.

On that second part tho "We have such an algorithm but it is not open source." does that imply that you have a closed source one someone can buy or something. Just figuring out my possibilities.

Rsambeek commented 4 years ago

I'm still curious to this algorithm you were talking about

kriswiner commented 4 years ago

Hi Rick,

I sent a message to Greg who has this, maybe send me an e-mail at tleracorp@gmail.com in case github doesn't do forwarding correctly...

Kris

On Fri, Oct 25, 2019 at 6:14 AM Rick van Sambeek notifications@github.com wrote:

I'm still curious to this algorithm you were talking about

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/392?email_source=notifications&email_token=ABTDLKQIYI3IUBVQ7C7ZAH3QQLWJXA5CNFSM4I7JWC72YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECIJ4RQ#issuecomment-546348614, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTDLKUI5NJCYUOGXZ4GQL3QQLWJXANCNFSM4I7JWC7Q .

gregtomasch commented 4 years ago

Hi Rick,

It would be best to e-mail me at poirot@earthlink.net and we can discuss..

Best, Greg