Roll and Pitch Accuracy levels

Question

Roll and Pitch Accuracy levels

Closed this issue 5 years ago · 5 comments

Thanks for things up here.

Curious about the kind of roll and pitch accuracy levels that can be obtained with MPU-9250/ICM 20948 or any of the ST MEMS.

Generally it is believed to be <1 deg, though haven't found anything specific. Is it possible to get better than that?

Thanks.

Answer 1 · 2019-05-05T17:10:45.000Z

In all of our testing of heading we generally see pitch and roll at 1 degree error or less. We have recently procured a 2-axis goniometer and are doing some testing of heading as a function of different pitch and roll angles. But generally we would not test pitch and roll accuracy specifically. If your goal is to estimate pitch and roll only, then using an accel/gyro like the LSM6DSM with proper calibration is enough to rather easily achieve the <1 degree rms standard. How much better you could do? I do not know. What is the requirement?

…

On Sun, May 5, 2019 at 12:40 AM Srinivas ***@***.***> wrote: Thanks for things up here. Curious about the kind of roll and pitch accuracy levels that can be obtained with MPU-9250/ICM 20948 or any of the ST MEMS. Generally it is believed to be <1 deg, though haven't found anything specific. Is it possible to get better than that? Thanks. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#56>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABTDLKSLTDYXGFBJ5DE5XHTPT2FPVANCNFSM4HK2TQRQ> .

Answer 2 · 2019-05-06T16:31:32.000Z

Looking at something close to 0.1 degrees for roll, pitch and yaw. May be its a big ask!

Answer 3 · 2019-05-06T16:33:37.000Z

The real questions is, how well can you do with a good sensor suite and proper calibration methods? In other words, the answer you need will only come from actual testing.

…

On Mon, May 6, 2019 at 9:31 AM Srinivas ***@***.***> wrote: Looking at something close to 0.1 degrees for roll, pitch and yaw. May be its a big ask! — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#56 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABTDLKWDOJAQCBOJYYACNFLPUBMOLANCNFSM4HK2TQRQ> .

Answer 4 · 2019-05-06T16:37:26.000Z

Yea. That makes sense. What would be your suggestion for a good sensor suite?

Answer 5 · 2019-05-06T16:41:52.000Z

LSM6DSM

…

On Mon, May 6, 2019 at 9:37 AM Srinivas ***@***.***> wrote: Yea. That makes sense. What would be your suggestion for a good sensor suite? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#56 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABTDLKSIUZT5UTL77CHEYNDPUBNERANCNFSM4HK2TQRQ> .