W3C home > Mailing lists > Public > public-device-apis-log@w3.org > April 2017

Re: [magnetometer] Add RawMagnetometer, for uncalibrated readings.

From: Alexander Shalamov via GitHub <sysbot+gh@w3.org>
Date: Tue, 04 Apr 2017 07:36:33 +0000
To: public-device-apis-log@w3.org
Message-ID: <issue_comment.created-291419200-1491291391-sysbot+gh@w3.org>
> To build the transformation matrix, the implementation asks you to manually do the calibration.

Without external forces, ideal response of magnetometer, should be 3D sphere (when rotated about axes). By rotating device and collecting distortions caused by external objects, 'ideal response' can be reconstructed (calibrated magnetometer). For 'hard-iron effects', it is just a translation matrix, for 'soft-iron' it is more complicated, since sphere can be distorted and rotated, thus, scale and rotation needs to be applied.

Quite many HW sensors have embedded routines for calibration and store calibration data on their non-volatile memory. I have few quadrocopters and I have to re-calibrate them everytime I fly them in a new location.

> Could someone explain me why some of the hard iron calibrations are abrupt and not always continously?

Calibration process assumes that external objects are stationary and produce constant distortion. Also, sensor and it's hosting device must be fixed relative to each other during calibration.

GitHub Notification of comment by alexshalamov
Please view or discuss this issue at https://github.com/w3c/magnetometer/pull/21#issuecomment-291419200 using your GitHub account
Received on Tuesday, 4 April 2017 07:36:39 UTC

This archive was generated by hypermail 2.4.0 : Monday, 4 July 2022 12:47:53 UTC