Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02330201 2001-05-09
METHOD OF OPERATING OR CONSTRUCTING A GEARTOOTH SENSOR
BACKGROUND OF THE INVENTION
Field of the Invention:
The present invention relates generally to position sensing apparatus and more
particularly to magnetic effect sensing apparatus including linear position
sensing as
well as the commonly known rotary position "geartooth sensors" wherein a
magnetically sensitive device senses a ferrous object or objects generally
projecting
from a rotating target and resembling the teeth of a gear.
Discussion of the Prior Art:
Various sensors are known in the magnetic effect sensing arts. Examples of
common magnetic effect sensors may include Hall effect and magnetoresistive
technologies. Generally described, these magnetic sensors will respond to the
change
of magnetic field as influenced by the presence or absence of a ferromagnetic
target
object of a designed shape passing by the sensory field of the magnetic effect
sensor.
The sensor will then give an electrical output which can be further modified
as
necessary by subsequent electronics to yield sensing and control information.
The
subsequent electronics may be either onboard or outboard of the sensor package
ital.
For example, geartooth sensors are known in the automotive arts to provide
information to an engine controller for efficient operation of the internal
combustion
engine. One such known arrangement involves the placing of a ferrous target
wheel on
the crank shaft of the engine with the sensor located proximate thereto. The
target
objects, or features, i.e. tooth and slot, are of course properly keyed to
mechanical
operation of engine components.
Examples of United States Patents in the related art include: 5,650,719;
5,694,038; 5,44,283; 5,414,355; 5,497,084 and 5,500,589.
It is well known in the art that the waveforms produced by the magnetic
sensor change in response to varying airgap between the target and sensor
faces. Also,
differences among the biasing magnets used in the magnetic sensor,
temperature,
mechanical stresses, irregular target feature spacing, etc., can vary the
sensor output.
Therefore, the point at which the sensor changes state, i.e. the switch point,
varies in
time, or drifts, in relation to the degree of rotation of the target. But the
mechanical
action of the engine as represented by the target does not change. That is,
there is a
"true point" on the target in angle, or degrees of rotation, related to a hard-
edge
CA 02330201 2001-05-09
-2-
transition, which represents the point at which the sensor should change state
to
indicate a mechanical function of the engine. But, due to inherent limitations
of the
sensing system, the point at which the sensor changes state will vary by some
amount
from this true point. Therefore, the sensor is losing accuracy, e.g. not
really giving a
timing signal accurately representing piston travel. Therefore, the system
controlled
by the'sensor can be inefficient. Several schemes are known in the art to
reduce this
sensor drift by providing an adaptive threshold of waveform voltage at which
to switch
the sensor. The adaptive threshold seeks to switch the sensor at a nearly
constant angle
in order to decrease switch point drift and increase accuracy of the sensor
and
to efficiency of the engine.
Various known systems for producing an adaptive threshold (AT) include
setting the adaptive threshold at a fixed level above a measured minimum
magnetic
bias signal. However, this function does not convey information proportional
to air
t 5 gap, therefore high accuracy is not achievable. Another method is setting
the threshold
at the average value of magnetic bias by using a time based integrator such as
an RC
circuit. While this method can yield high accuracy, the accuracy is not
achieved until
considerable amount of target rotation has taken place. It is more desirable
to achieve
the adaptive threshold point very quickly in the target rotation.
Other proposals, such as that proposed by United States Patent 5,650,719,
include digital schemes for tracking the voltage peak and voltage minimum of
the
output waveforms and selecting a point therebetween for the adaptive threshold
and
updating these peak and minimum values on a regular basis determined by a
selected
passage of target features.
However, all the known schemes for setting a threshold to compensate for the
sensor drift to minimize switch point deviation suffer drawbacks. Such
drawbacks may
include increased circuit complexity, leading to increased expense; extensive
target
3o rotation before the adaptive threshold is determined; and lessened overall
accuracy of
the determined adaptive threshold for the waveform variance. Compromises among
these negatives are inherent in any design. The present invention seeks to
minimize the
deleterious tradeoffs and provide a magnetic sensor which is an adequate
balance of
low cost, fast threshold acquisition time, and high accuracy.
CA 02330201 2001-05-09
-3-
SUMMARY OF THE INVENTION
The present invention discloses a method for operating a geartooth sensor. In
another embodiment, the present invention discloses a method for constructing
a
geartooth sensing system. An empirically derived first constant (M,) is
derived to
account for anticipated output voltage fluctuations inherent in the target and
the
anticipated airgap tolerance range. The M, constant is applied to the measured
peak
value (B~~) or values of selected waveforms to obtain a value Bm~. Using this
value
alone as the AT point eliminates the majority of drift in the sensor.
to A second empirically derived constant (M2) is derived and applied to a
measured or time integrated average value (Ba~J of, e.g., each wave to obtain
a low
value (Bm;,~. The average value referred to may be either a calculated
arithmetical
value or a time based value taking the form of median or mean values. Bm,X and
Bm;
are then added to obtain the final adaptive threshold value (AT) which is up-
dated at
15 the sensor circuitry to eliminate another portion of drift and define
accurate sensor
output switch points.
M, and MZ are empirically derived or modeled constants which are adapted to a
specific target configuration and duty cycle as measured or modeled over the
2o anticipated airgap tolerances of the specific application of the Hall
sensor 11 with
respect to target 13.
Because BmeX is derived from quickly acquired peak values, and because Bm~ is
the much larger value in the algorithm, the present invention synchronizes
quickly
25 while obtaining very good accuracy when the Bm;~ value is added at the
slightly later
time taken to acquire it.
Further, by utilizing the algorithm with its two empirically derived constants
applied to the two waveform values, the need for performing calibration on the
sensor
3o and, e.g. adjusting its circuitry by laser trimming or the like, to improve
sensor
accuracy is minimized.
The AT point is held in the sensor circuitry, whether analog or digital, and
may
be updated at any chosen frequency to minimize the drift of the sensor switch
points,
35 thereby minimizing sensor inaccuracy and increasing engine efficiency.
CA 02330201 2001-05-09
-4-
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention will be more fully and completely understood from a
reading of the Description of the Preferred Embodiment in conjunction with the
drawings, in which:
Figure 1 is a block diagram schematic view of the sensor according to the
present
invention.
Figure 2 illustrates a graph of plotted signal gauss waveforms (also referred
to
as magnetic transducer output waveforms) over a plurality of air gap distances
between
1 o the sensor of the present invention and a rotating target.
CA 02330201 2001-05-09
-5-
DESCRIPTION OF THE PREFERRED EMBODIMENT
Throughout the Description of the Preferred Embodiment, like components will
be identified by like reference numerals.
Referring to Figure 1, a back biased Hall element sensor or the like 11 is
placed
in proximity to a ferrous geartooth target 13. As is known the excursions
between the
teeth and slots of the geartooth target 13 will affect the magnetic bias
output of the Hall
sensor 11. The output of the Hall is then sent to an amplifier 15 which
produces a
characteristic waveform (Fig. 2) which is dependent largely upon the airgap
between
1o the Hall sensor and the target, and may be affected by temperature and bias
magnet
strength. Also the amplified Hall output signal is then sent to measuring
circuitry 17 to
determine the peak (B~,~)of the waveforms. The design of such circuitry is
considered
within the skill of the ordinary artisan and need not be elaborated here. The
amplified
output is additionally sent to circuitry 19 to determine the average value
(Beds) of the
waveforms at 19. Again, this circuitry is considered a matter of choice within
the art
and need not be detailed. The average could be an arithmetic calculation or
derived
from a time based integrator. It will of course be appreciated by the
ordinarily skilled
artisan that any variety of analog or digital implementation in hardware or
software
may be utilized to accomplish the electronic circuitry behind the Hall effect
sensor.
The measured B~~ value is then further processed at block 21 by applying a
first constant M, to the peak value in order to establish the larger portion
of the
adaptive threshold which represents the majority of elimination of drift among
the
varying signals. As seen in Figure 2, it is the peak 39 of the signal which
varies most
due to airgap variation while the minimum or bottom 41 of the waveform remains
relatively constant.
The Bag value 19 is then sent to block 23 for applying a second constant M2
to the average value to derive the value B- minimum. In analog circuitry the
M, and
3o MZ values could, e.g., be fixed on an IC resistive network by laser
trimming.
The adaptive threshold (AT)is selected to most closely approximate a line of
values through the various airgap influenced waveforms to yield a switch point
least
varying in degrees relative to target rotation. M, and Mz are constants which
are
derived from a specific target configuration and duty cycle over the
anticipated airgap
tolerances of the specific application of the Hall sensor 11 with respect to
target 13.
The M, constant is typically selected as a large percentage of the peak value
BP~~ This
value, Bp~~, conveys the greatest amount of information about airgap. B~~ is
further
the most quickly acquired value in the sensor circuitry, there being necessary
only one
CA 02330201 2001-05-09
-6-
tooth to pass the sensor in order to establish B~~. Typical values could be,
e.g., .7 to .9
for M,. It is necessary to set M, at a value to ensure that the dips 33 in the
waveform
peaks do not interfere with proper sensor switching. M1, on the other hand, is
selected
to be a much lower value, for example, .5, which is then applied to the
determined Bas,
therefore yielding a much smaller number because M, is greater than MZ and
BP~ek is
greater than Ba~~. Bas is of course acquired later in time over several
degrees of target
rotation, i.e. the passage of a plurality of tooth and slot features on the
target 13.
After the M, x B~~ and Mz x B$~gk values are established they are added, as at
to block 25, to establish the adaptive threshold value (AT) 27 which is the
point on each
air gap variant of the waveform corresponding to the least amount of drift and
the most
accurate representation of target rotation switch point for the sensor change
in output
state. The adaptive threshold (AT) is then applied to the sensor circuitry, as
by
comparison to actual sensor output represented by line 35, which is also
applied to the
15 comparator 29 in order to yield the most accurate sensor output 31.
Thus the function of the circuitry behind the Hall effect transducer and its
initial amplifier is to obtain an adaptive threshold to control the change of
state output
of the sensor by utilizing the algorithm: AT = M, x B~ek + Mz x Beg where AT
is
2o adaptive threshold, B~~ is measured maximum value of the waveform, Bas is
measured average value of the waveform, and M, and Mz are empirically derived
constants applied to Bpe~ and Beg, respectively, when taking into account the
target
design, duty cycle and expected waveform variations over a variety of
operating
conditions including expected airgap tolerance.
Because the present invention utilizes both s~,~, which contains the most
information about airgap variation signal effect, and s~,g which contains duty
cycle
information and is a highly accurate indicator of waveform switch point drift,
the
sensor
of the present invention yields a good balance of speed of adaptive threshold
acquisition, overall accuracy, and averaging consistency since measurement is
not
necessarily required of every peak and valley excursion of the waveform.
Further,
because the present invention yields a highly accurate and adaptive threshold
for the
switch point, sensor calibration of individual sensors is not necessary during
manufacture to yield predictable and accurate results.
While the present invention has been described in terms of a specific
embodiment, it will of course be appreciated that many variations will occur
to a
person have ordinary
CA 02330201 2001-05-09
_7_
skill in the art and that the present invention may be implemented in a
variety of
electrical hardware and software formats in either analog or digital domains.
