Position and angle measurement by scanning magnetic scales or multipole wheels is described in the prior art, for instance in German Patent Disclosure DE 195 06 938 A1.
At the distances, or spacings, that result in practice between a magnetic sensor and a scale, the course of the magnetic field is not ideally sinusoidal—as assumed—but instead includes square-wave components, which can be generally well described by a third and fifth harmonic of the sine-wave function. This limits the precision of the method, however. A field course free of harmonics generally exists only at relatively great operative spacings and thus at very low working field intensities (the field intensity drops off exponentially with the distance), which makes the system very vulnerable to malfunction.
The invention proposes a sensor structure which on the basis of its geometry filters the third and/or the fifth harmonic out of the signal, so that a sinusoidal course results as the measured value, even for only slight spacings between the sensor and the scale. As a result of the aforementioned harmonic correction, the course is moreover largely independent of the working spacing of the sensor, so that even the influence of an eccentricity that may be present in the system on the outcome of measurement is minimized.
Position transducers with improved precision and vulnerability to malfunction by operation at small working spacings. Insensitive to eccentricities of the pole wheel or to its position.
It is state of the art to scan magnetic scales (such as multipole wheels) using so-called gradiometers; see also
The invention now makes use of the fact that even at a spacing d of the two half-bridges not equal to lambda/2, in most cases a sinusoidal bridge signal is available, although with a slightly reduced amplitude. The bridge for harmonics of period d furnish no signal.
For reasons of symmetry, with the aforementioned pole wheels, only odd-numbered harmonics occur. If one seeks to make a bridge that suppresses the third harmonic, then the spacing d=lambda/3 (120°), or a multiple thereof, should be selected. In that case, the phase between the half-bridges is 120°, which is precisely equivalent to the period of the third harmonic. In that case, the bridge voltage is reduced to 90%, compared to the bridge with d=lambda/2, and the bridge signal is offset by a phase of 30° compared to the fundamental of the original signal; see
The fifth harmonic can be suppressed in the same way by means of a spacing of the half-bridges of d=lambda/5, or a multiple thereof; see
In practice, sufficiently good harmonic suppression is achieved if the third and the fifth harmonics are suppressed. This can be achieved by the following arrangement and signal evaluation (see
1. The signal S1 of a bridge B1 with d=lambda/3 is averaged with the signal S2 of a second bridge B2 with d=2*lambda/3; the resultant signal is S12.
2. A signal S3 of a bridge B3 with d=lambda/3 is also available, which is phase-offset from B1 and B2 by 30°. This is accordingly the phase-corrected signal of the 120° bridge.
3. The mean value between S12 and S3 results in the field course, with the third and fifth harmonics filtered out.
Number | Date | Country | Kind |
---|---|---|---|
100 41 087 | Aug 2000 | DE | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/DE01/03190 | 8/21/2001 | WO | 00 | 7/14/2003 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO02/06878 | 2/28/2002 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
4818939 | Takahashi | Apr 1989 | A |
5430375 | Inoue | Jul 1995 | A |
5680042 | Griffen et al. | Oct 1997 | A |
5930905 | Zabler et al. | Aug 1999 | A |
6300758 | Griffen et al. | Oct 2001 | B1 |
Number | Date | Country |
---|---|---|
195 06 938 | Aug 1996 | DE |
Number | Date | Country | |
---|---|---|---|
20040012386 A1 | Jan 2004 | US |