Velocity Measurement System

1. Field of the Invention

This invention relates to speed measurements devices, and particularly to an apparatus employing magnetic means for analog measurement of the speed of an electrically conductive body, either of solid or liquid substance.

2. General Description of the Prior Art

Most known magnetic speed measuring devices are based upon eddy current effects, either changing the magnetic flux directly or by causing changes in a secondary flux path. The difficulty in the past with such devices has been that their efficiency, or signal vs. exitation energy, has been poor and that their signal linearity has generally been unsatisfactory.

Accordingly, it is an object of this invention to provide an improved magnetic type speed sensor with a good efficiency and satisfactory signal linearity which may be combined with magnetic bearings. A further object of this invention is to provide a device which permits the measurement of speed components without cross coupling effects as between components.

In accordance with this invention, an E-shaped magnetic core is positioned with its three poles closely spaced from a conductive body, the speed of which is to be monitored. A pair of Hall effect devices or generators are positioned on the surface of the end of the central pole, these generators being significantly spaced, preferably spaced a maximum distance measured along the line of movement of the body to be monitored in speed. Thus, the Hall generators are positioned where the flux density shows its greatest change resulting from the velocity of the conductive body and at points of greatest difference in flux. This difference is thus sensed by the Hall generators, and from it the speed of the body is indicated.

As a feature of this invention, means are provided to maintain a constant total flux through the magnetic circuit involved by varying an electrical input to an energizing coil supplying magnetic bias to the E-shaped magnetic core. The invention is adaptable for use for measuring the speed of bodies which are simply electrically conductive, as well as being both conductive and having a magnetic property. Further, speed components may be measured in two or more directions, it being simply necessary to orient the magnetic core or cores with Hall effect generators positioned along a line which speed or a speed component is to be measured.

Still further, this invention may solely be a discrete measurement device, or it may be combined with other magnetic circuitry, such as that of a magnetic bearing or bearings.

FIG. 1a is a schematic illustration of the basic concept of this invention.

FIG. 1b is a series of curves illustrating flux distribution resulting from the configuration shown in FIG. 1a.

FIG. 2 is a schematic illustration of a form of this invention adapted to measure the surface speed of a circular or spherical body.

FIG. 3 is a view of the magnetic core structure shown in FIG. 2 as seen when viewing the structure normal to the pole faces.

FIG. 4 is a schematic illustration of a magnetic support for a rotating spherical body.

FIG. 5 is a schematic illustration of a speed sensing system as contemplated by this invention.

Referring first to FIG. 1a, there is shown a basic illustration of this invention as employed to measure the speed of a moving body 10, moving in the direction of arrow 12. An E-shaped magnetic core 14 is energized by a winding 16 which is powered by means not shown. Typically, the power supply would be a direct current power supply. Moving body 10 is closely positioned between E core 14 and a yoke of magnetic material 18, the spacing between the core and moving body and the yoke and moving body typically being approximately one millimeter. Moving body 10 would either be of a conductive material or have a conductive surface on it. Hall effect generators H.sub.1 and H.sub.2 are positioned, as shown, on the phase of center leg 24 of E core 14, Hall generator H.sub.1 being positioned on what may be deemed the trailing flux edge of center core 24, and Hall generator H.sub.2 being positioned on what may be deemed the leading flux edge of center core 24, considering the direction of movement of moving body 10. Yoke 18 functions to provide increased flux path for greater signal output of the system, but it is not absolutely necessary for operation.

FIG. 1b is positioned under FIG. 1a to illustrate the magnetic flux density across the circuit of the device shown in FIG. 1a, and particularly to illustrate the magnetic flux density under the poles of the E-shaped magnetic core 14. Solid line 26 represents the magnetic induction when moving body 10 is at rest, and dashed line 28 represents the induction distribution occurring for a selected velocity of moving body 10. Thus, it will be noted that in accordance with this invention, Hall effect generators H.sub.1 and H.sub.2 are placed at the position of maximum flux density variation to obtain maximum sensitivity of the measuring arrangement.

Hall generators H.sub.1 and H.sub.2 are appropriately biased by a direct current source in a conventional manner, and their outputs are applied to difference circuit 32, such as a bridge, and its output, appropriately scaled, is applied to indicator 34 which displays or otherwise provides an indication of the speed of movement of body 10.

FIGS. 2 and 3 illustrate an embodiment of this invention as applied to the measurement of the speed of sphere 40, and the basic theory of operation of this invention will be explained with respect to these figures.

As shown, pole faces 42, 44, and 46 of E core 48 are curved to conform with that of the outer surface 50 of sphere 40. DC exitation is provided coil 52 from DC current source 54, controlled by current control 56, in turn responsive to flux sensed by Hall effect generator 58, positioned in the center of the face of center pole 44. Current is thus controlled to maintain a constant flux in the magnetic circuit through center pole 44.

Since it is normally not possible in such a system to employ a fixed magnetic yoke, the surface region of sphere 40 would typically be constructed of an outer layer 60 which is of a good electrical conductive material formed over layer 61, which is of a good magnetic material. Alternately, a single layer having both of these properties may be employed. As another alternate, the surface region may simply be of a conductive material, and signal amplifiers 62 and 64 are employed as shown to appropriately amplify the outputs of Hall generators 66 and 68. The outputs of amplifiers 62 and 64 are applied to and subtracted by difference circuit 70, and the resulting signal, which is proportional to velocity, is indicated by indicator 72.

To examine the operation of the system, voltages induced by the magnetic field on the rotating sphere cause electric currents to flow in the conducting surface of the sphere. These currents are oriented in such a way that they oppose magnetic flux change. The result is a decrease in magnetic flux density as seen from the rotating sphere at the leading or on-coming edge 74 of center pole 44, and an increase in flux density on its trailing or leaving edge 76. The difference in flux densities measured by Hall generators 66 and 68 is a first approximation proportional to the speed component of the sphere in the plane of the Hall generators.

A simplified analysis of the properties of the system follows:

The magnetic flux density B under the center pole of E-shaped core 48 is:

where p is the permeance of a magnetic flux tube of the magnetic flux,

AT.sub.1 are the magnetization ampere turns of winding 16, and

AT.sub.2 are the ampere turns produced by the moving armature 10,

opposing the ampere turns AT.sub.1 in the magnetic flux tube under consideration. The voltage in moving armature 10 under core 48 is for a closed electrical circuit given by the proportionality

where v is the speed of armature 10. The resulting current in the armature, caused by voltage e, is

where r is the resistance of the current path in the armature. Substituting equations (1) and (2) in (3) shows the relationship ##EQU1## The armature ampere turns AT.sub.2 acting on the flux tube under consideration are proportional to the armature currents; thus, with equation (4), it follows that ##EQU2## or resolving this equation for AT.sub.2, one obtains ##EQU3## Since AT.sub.2 causes the change in flux density .DELTA.B at the measuring points of the Hall generators as shown with equation (1), this flux change is

and with equation (5): ##EQU4## The difference of the signals from the two Hall generators is within the linearity of the Hall generators proportional to the change in flux density. Thus, with a gain factor g in the sensing loop of the Hall generators and amplifiers, the measured signal is ##EQU5## This equation shows that the signal S is proportional to the armature velocity v if v <<r/kp. For larger velocities, the signal reaches a saturation value. The non-linearity can be reduced by designing

Of the three parameters k, r, and p, k is a proportionality value (equation (5)) which cannot be changed very much, whereas r and p can be modified by proper design. Selection of a high resistance of the current path in the armature and of a low permeance, e.g., by a larger air path for the magnetic flux, are helpful. The subsequent loss in signal strength as evident from equation (8) can be compensated for by a sufficiently high gain factor g as well as by a higher magnetization of the E-shaped core.

The results demonstrated by equations (8) and (9) are quite evident if one considers that the non-linearity effect is caused by ampere turns in armature 10 which reduces the effect of exitation ampere turns provided by winding 16. Thus, the smaller the sensor signal the smaller will be the non-linear effect, and the more amplification which will be necessary to obtain a desired signal strength. In most instances, sufficient amplification at an acceptable noise figure can be obtained.

The non-linearity inherent in the electromagnetic subsystem can be compensated for by control of the magnetic flux to a constant value by the technique illustrated, that is, by controlling the energizing current to coil 52 by means of current control 56, in turn controlled by auxiliary Hall generator 58. The latter provides a control signal indicative of the magnitude of flux through E-shaped core 48, whereby, by means of conventional techniques, current control 56 controls power source 54 to provide a current input to winding 16 to maintain a constant flux.

FIGS. 4 and 5 illustrate the employment of this invention with magnetic bearings 80 and 82, which have an E-shaped cross section and which are adapted to magnetically support and position sphere 84 without any physical bearing surface. An example of such a system of support is shown in U.S. Pat. No. 3,017,777 previously issued to the applicant. The cores of magnetic bearings 80 and 82 are controlled by means of bias source 86, typically a differential type source supplying current differentially to the magnetic bearings. As will be further explained, means are provided to assure that the sum of the magnetic fluxes to the bearings is held constant in order to assure accuracy of the measurement system of this invention. As will be noted in FIG. 4, Hall effect generators H.sub.1 and H.sub.2 are positioned on extreme edges of the center pole of magnetic bearing 82, this being with respect to the direction of measurement of rotation of sphere 40. Similarly, Hall effect generators H.sub.3 and H.sub.4 are so positioned on the center pole of magnetic bearing 80.

Referring particularly to FIG. 5, it will be noted that the Hall generators are biased, in series, by means of power source 88 and series variable resistor 90, by which the bias current to the Hall generators may be adjusted. The current output of each Hall generator is fed to one of amplifiers 92, 94, 96, and 98. The sum of the outputs of these amplifiers are added by connecting them in series, and thus there is provided a sum signal which is appropriately labelled as (H.sub.1 + H.sub.2) + (H.sub.3 + H.sub.4), from which is subtracted a reference voltage obtained from voltage divider 100, in turn powered from power supply 88 furnishing biasing power to the Hall generators. By adjustment of this voltage divider, an output is obtained which is adjusted to be 0 for a condition of a selected flux sum detected by the Hall effect devices. This selected state would typically be a selected operating flux state for the magnetic bearings. In order to maintain accuracy and calibration of speed indication, it is desirable that this value of total flux be maintained. Thus, the selected output achieved by adjustment of voltage divider 100 as aforesaid is fed to terminal 101 (FIG. 5) and from terminal 101 (FIG. 4) to current control 102, and it in turn controls the total current output of bias source 86, which supplies power to magnetic bearings 80 and 82. Typically, bias source 86 would differentially apply a current to magnetic bearings 80 and 82. Discrete outputs of amplifiers 92 and 94 are subtracted in difference circuit 109, and the discrete outputs of amplifiers 96 and 98 are subtracted in difference circuit 104, and the thus obtained difference signals are summed in sum circuit 106 to provide a signal representative of (H.sub.1 - H.sub.2) + (H.sub.3 - H.sub.4). This signal obtained from the Hall generators in both magnetic bearings 80 and 82 is proportional to the speed of movement of sphere 80 along a line corresponding to a plane intersecting the Hall generators. After proper amplification in amplifier 107, the signal is fed to speed indicator 108, which provides a display or other indication of speed.