Note: Descriptions are shown in the official language in which they were submitted.
CA 02247944 1998-08-31
WO 97/33143 PCTIUS97/04136
MULTI-LOAD CELL FORCE SENSING APPARATUS
Field of the Invention
The invention is directed to a force sensing apparatus
using multiple load cells to sense an applied force. More
particularly, the invention is directed to a force sensing
apparatus in which multiple load cells may be used to
determine the position as well as the magnitude of an
applied force.
$acksrround of the Invention
Multi-load cell force sensing apparatus such as scales
have been used in various environments to measure applied
force. For example, the use of multiple load cells is
advantageous in heavy capacity weighing applications since
the applied force may be borne and sensed by a plurality of
lower capacity load cells, rather than using a single,
= large capacity load cell. Lower capacity load cells tend
to be less expensive than large capacity load cells, and
they also tend to have greater accuracy and precision.
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 2 -
The total applied force in a multi-load cell
environment is in theory the sum of the force signal
outputs of the load cells. However, it has been found that
the sensitivity of some load cells may vary from cell to
cell, and that coupling effects due to mounting may also
affect individual load cell sensitivities. As a result,
the force reading output of a multi-load cell force sensing
apparatus may vary depending upon the position of an
applied force on the apparatus.
Attempts have been made to adjust the sensitivities of
individual load cells to compensate for changes in position
of an applied force, e.g., by connecting sensitivity
reducing resistors to the output circuits of individual
load cells. U.S. Patent No. 4,804,052 to Griffen discusses
multiplying the digital outputs of load cells by load
position correction factors prior to summing the outputs to
compensate for load position. The correction factors are
determined in calibration by placing a known weight at a
plurality of positions (equal to the total number of load
cells in the apparatus), and then solving the resulting
simultaneous equations. Griffen also discusses the use of
a local area network to connect the load cells to a master
controller, such that the outputs of the load cells may be
obtained independently and free from interaction between
individual cells. User interaction with the Griffen
apparatus occurs through a separate keyboard.
Correction of a scale output for position has several
advantages, most important of which is to eliminate the
dependence of the force reading of the scale on the
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 3 -
position of a weight on the scale. However, it would also
be useful in many applications and environments to be able
to calculate the position of a force on the scale.
Therefore, a need exists in the art for a force
sensing apparatus which determines the position, as well as
the magnitude, of an applied force.
SnA=ary of the invention
The invention addresses these and other problems
associated in the prior art in providing a multi-load cell
force sensing apparatus which determines the position, as
well as the magnitude, of an applied force relative to an
operating surface to which the force is applied. Position
sensing may have many advantages in weighing applications
and the like. For example, determining the position of an
applied force may be used to sort objects, to calculate
postal rates for different postage classes or delivery
zones, to detect objects which fall outside of a desired
specification, to oversee the status of an industrial
process, etc.
Moreover, in preferred embodiments, impulse forces
applied to a force sensing apparatus may be distinguished
from static forces, thereby enabling the apparatus to
determine the position of an impulse applied to the
operating surface, e.g., due to a user applying a force to
the operating surface with the hand or foot. Such a
= feature permits an apparatus to function as a user input
device, whereby a user may select the apparatus to perform
CA 02247944 2006-02-17
79150-37
- 4 -
specified functions merely by touching the operating surface
at a specified location.
Therefore, in accordance with one aspect of the
invention, there is provided a force sensing apparatus,
comprising: an operating surface for receiving an applied
force, the applied force having a position relative to the
operating surface and a magnitude; a plurality of load
cells, each load cell receiving at least a portion of the
applied force on the operating surface and providing a force
value representative thereof; means for determining both the
magnitude of the applied force and the position of the
applied force on the operating surface based on the force
values generated by load cells; and means for outputting a
first signal corresponding the magnitude of the applied
force and a second signal corresponding to the position of
the applied force.
In accordance with another aspect of the
invention, there is provided a scale, comprising: a platter
having an operating surface defined thereon for receiving an
applied force, the applied force having a position relative
to the operating surface and a magnitude; a plurality of
load cells supporting the platter, each load cell providing
a force value representative of at least a portion of the
applied force; and a controller, coupled to receive the
force values from the load cells, the controller including:
means for determining both the magnitude of the applied
force and the position of the applied force on the operating
surface based on the force value of each of the load cells;
and impulse sensing means for detecting an impulse force
applied to the platter and for determining the position of
the impulse force on the platter.
CA 02247944 2006-02-17
79150-37
- 5 -
In accordance with an additional aspect of the
invention, there is provided a method of operating a multi-
load cell scale having a platter supported by a plurality of
load cells, the method comprising: sensing the weight of an
object placed on the platter; determining the position of a
user impulse applied to the platter, and performing a
predetermined function in response to a user impulse
received in a predetermined area of the platter.
These and other advantages and features, which
characterize the invention, are set forth in the claims
annexed hereto and forming a further part hereof. However,
for a better understanding of the invention, its advantages
and objectives attained by its use, reference should be made
to the Drawing, and the following descriptive matter, where
a preferred embodiment of the invention is described.
Brief Description of the Drawing
FIGURE 1 is a perspective view of a preferred
multi-load cell force sensing apparatus consistent with the
invention.
FIGURE 2 is a functional block diagram of the
primary components in the apparatus of Fig. 1.
FIGURE 3 is a functional perspective view of the
operating surface and load cells in the apparatus of Fig. 1,
showing X-Y coordinates overlaid thereon for illustrating
the principles of operation for the apparatus.
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 6 -
FIGURE 4 is a partially cut away and partially
exploded perspective view of the apparatus of Fig. 1.
FIGURE 5 is a functional block diagram an alternate
communication protocol for the apparatus of Fig. 1.
FIGURE 6 is a flowchart showing the load cell
controller program flow for implementing the communication
protocol of Fig. 5.
FIGURE 7 is a flowchart showing the primary program
flow for the control system in the apparatus of Fig. 1.
FIGURE 8 is a flowchart of the COMPUTE FORCE AND
POSITION routine in the program flow of Fig. 7.
FIGURE 9 is a functional perspective view of one
embodiment of the invention, having a plurality of keys
defined on an operating surface thereof.
FIGURE 10 is a functional perspective view of another
embodiment of the invention for use as a sorting device.
FIGURE 11 is a functional perspective view of another
embodiment of the invention for use as a quality control
comparison device.
FIGURE 12 is a functional perspective view of another
embodiment of the invention for use as a process control
device.
Detailed DescriAtion of the Preferred Embodiments
Turning to the Drawing, wherein like parts are denoted
by like reference numbers throughout the several views,
Figure 1 shows a preferred scale 10 consistent with the
principles of the invention. Scale 10 generally includes
an operating surface 20 defined by a platform 22, and
CA 02247944 1998-08-31
WO 97/33143 PCTIUS97/04136
7 -
supported by a plurality.of load cells 30 housed in a base
/ 24. User interface is performed through an input device
such as keyboard 28 and a display such as display 26, and
primary control is provided by a control system 50. The
input device 28, display 26 and control system 50 are shown
incorporated within base 24, although each may also be
external to the base of the scale.
Preferred embodiments of the invention are designed to
receive applied forces on an operating surface which is
supported by a plurality of load cells such that the load
cells receive at least a portion of the applied force. The
applied forces may be weights, but they may also represent
acceleration forces, impact forces, torques, angular
forces, angular acceleration forces, or any other known
force which is capable of being measured. In addition, the
applied forces may include forces initiated by a user
manually manipulating the operating surface (e.g., as a
user input, such as pressing the surface to request an
operation to be performed by the scale).
By an "operating surface", what is meant is a
structure upon which an applied force is received. An
operating surface may be disposed on a particular member,
such as platter 22 in Figure 1, and may take any contour
consistent with its function. However, an operating
surface may also be defined by force-receiving structure on
the load cells, or by any structure which is capable of
transmitting the applied force to the load cells.
Therefore, the use of a separate member for an operating
surface in the preferred embodiments should not be
CA 02247944 1998-08-31
WO 97/33143 PCTIUS97/04136
- 8 - ,
construed as requiring a separate physical object to be
used for the operating surface.
By "load cell", what is meant is a force measuring
assembly which is capable of providing a force measurement
or output in response to a force applied thereto. The
precise structure of a load cell will vary depending upon
the type of force being measured, as well as the accuracy
and precision required for the particular environment in
which the load cell is used.
Figure 2 shows a block diagram of the functional
components of scale 10. Scale 10 is controlled by a
controller or control system 50, which handles the primary
operational functions of the scale. Control system 50
drives display 26, and preferably receives some user input
from user input device 28. As will be discussed below,
additional user input may be provided by a user touching or
pressing on operating surface 20.
Control system 50 preferably handles data exchange
with a plurality of load cells 30 over a common bus 40.
This configuration is generally known as a master/slave
configuration, whereby data communication is coordinated by
control system 50 acting as a master controller. Each load
cell 30 sends or receives information across bus 40 only in
response to a command sent out by control system 50. As
will be discussed in greater detail below, several
communications protocols may be used to handle the
information exchange between control system 50 and load
cells 30.
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 9 -
Princinles of Otperation
Figure 3 is a functional perspective view of scale 10
for illustrating the principles of operation of the
preferred embodiments of the invention. Scale 10 is shown
having an operating surface 20 supported by a plurality of
load cells 30, in this case four load cells also designated
A, B, C and D. An X-Y coordinate system is shown generally
overlaying operating surface 20, with the origin (x=0,y=0)
being located at the corner of the operating surface
proximate load cell A.
Preferred embodiments of the invention rely on the
basic premise that the sum of the reactive moments
exhibited by a plurality of load cells supporting an
operating surface will be equal to the total moment applied
to the operating surface. This relationship is generally
valid whether looking at the total moments, or at the
moments only along one axis. This premise permits the
position of both the load cells and weights or forces
applied to the operating surface to be determined after
certain characteristics of each load cell are calculated.
In theory, the overall magnitude of a force W applied
to the operating surface of a multi-load cell system is
simply the sum-of the forces measured by all of the load
cells. For example, for N load cells, the overall force W
is.
N
W= F. (1)
t=1
CA 02247944 1998-08-31
WO 97/33143 PCTIUS97/04136
- 10 -
where F. is the force value output by a load cell i. In
Figure 3, load cell A corresponds to i = 1, load cell B
corresponds to i = 2, load cell C corresponds to i = 3, and
load cell D corresponds to i = 4.
However, it has been found that in actual multi-load
cell systems, some errors may exist in the force
measurements of the individual load cells due to mounting
effects, or imperfections in the installation of the load
cells, e.g., due to parallel spring effects from objects
attached to the system, unleveled mounting of the load
cells, or lever effects. These imperfections may result in
varying the overall force measurement depending upon the
location of the force applied to the operating surface.
Therefore, it has been found that each force value from a
load cell must be scaled by a span correction factor to
correct for these imperfections and thereby generate proper
overall force measurements independent of the position of a
force on the operating surface. As a result, the overall
magnitude of the force W corrected for span effects
becomes:
WFisi (2)
i=1
where sj is the span correction factor for a load cell i,
and each term Fjsi is a span corrected force value.
To calculate the span correction factors, N
calibration forces are applied to the operating surface to
generate force measurements for each load cell. Each
CA 02247944 1998-08-31
WO 97/33143 PCTIUS97/04136
- 11 -
calibration force, W. (with j=1 to N) may be used to
generate an equation:
N
W = ~ F , i si (3)
With each calibration force W. and each measured force
F,,, known, the N unknown span correction factors may be
determined by solving the N resulting simultaneous
equations. Preferably, the calibration forces are a
single, known weight placed on the operating surface at
different locations, e.g., proximate each load cell.
However, differently sized and/or placed calibration forces
may also be used.
The effective positions of the load cells along each
axis relative to the operating surface (Xi, Yj) may also be
calculated as characterizing information for the load
cells. For example, Figure 3 shows the coordinates (X,,Y,)
representing the effective position of load cell C. In
some circumstances, the effective positions of the load
cells may be determined by measuring their point of
mounting on the operating surface. However, on many types
of load cells, the effective positions of the load cells
may not be precisely at the points of mounting due to
levering effects of the load cell mounts or other effects.
Similar to the determination of the span correction
factors, N known calibration forces must be applied to the
operating surface to generate calibration data for
determining the position of each load cell. The
calibration forces Wj must be applied at known locations
CA 02247944 1998-08-31
WO 97/33143 PCTlUS97/04136
- 12 -
(xj,y'~), preferably one proximate each load cell.
Determination of the load cell positions may be performed
based upon the assumption that under stationary conditions,
the sum of the reactive moments exhibited by the load cells
along each axis will be equal to the total moment.applied
to the operating surface along each axis. Therefore, the
position x,, and y,, for each applied calibration force may
be calculated to be:
i 1 F ,sXj ~
Xj _ ( )
W.
~
l--~
y Fi, j S i Yi
j
where each W. is calculated using equation (3). Again, N
unknown equations with N unknowns result for each axis, and
the load cell positions may thus be calculated using any
suitable manner for solving simultaneous equations.
Each load cell may thus be characterized by a unique
span correction factor s, and coordinates, or position
correction factors (Xt,Yj). Preferably, calibration
process for obtaining all of this data is performed
simultaneously, using a single known weight placed at N
known positions on the operating surface. The span
correction factors may then be determined using equations
from (3), then the coordinates may be determined using
equations from (4) and (5). Other calibration weights,
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 13 -
positions, and numbers of datapoints may also be used in
the alternative.
Once the characterizing data for the load cells is
calculated, the magnitude and position of an unknown force
on the operating surface may be calculated in a relatively
straightforward manner. For example, an unknown force of
magnitude W is shown at coordinates (x,y) in Figure 3.
As above, the moments about each axis cancel and the
total supporting force is equal to the applied weight under
static conditions. The magnitude of the unknown force W is
Fsi, from equation (2) above.
Further, the axial moments, due to the supporting
force's distance from each axis, M,, and M. are:
M = ~ FjXisi (6)
_ i=1
M = ~ FiYjSi (7)
i=1
where X. and Y. are the coordinates, or position scaling
factors, for each of the load cells, and each term FfXisj
and FjYjsj are partial axial moments.
The position coordinates of the unknown force (x,y)
are determined by dividing (or taking the ratios) of the
axial moments by the magnitude of the applied force about
the relative fixed reference frame, resulting in the
equations:
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 14 -
M ~ F Xis2
x i=1
X = ~ = N ($)
~ Fisi
i=1
F Yisi
w N (9)
~ FiSi
i=1
it will be appreciated that the position coordinates
of the unknown weight are not a function of the applied
force. However, the resolution of the position calculation
may be affected by the magnitude of the applied force, as
well as the distribution of the applied weight, force, or
impulse and the time allowed for response, and further, the
number of load cells used. As few as three load cells may
be used to provide two dimensional positional
determination, although greater numbers of load cells would
provide stability in various platform configurations.
Two load cells disposed along an axis may be used to
provide one dimensional positional determination. in this
application, only one coordinate is calculated for each
load cell (e.g., using equation (4)), and only one
coordinate need be calculated for unknown forces (e.g.,
using equation (8)).
Three dimensional position determination may also be
contemplated using four or more load cells disposed in a
three dimensional spatial relationship, with an operating
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 15 -
surface that is curved to follow a three dimensional
profile. Three dimensional position determination would
have application as an accelerometer, for example, where a
directional force may be determined.
The position and magnitude of impulse forces applied
to an operating surface may also be calculated in a similar
manner as above, but instead using the relative change in
the force measurements by the individual load cells to
calculate magnitude in position. For example, by using
force change values dFi representative of the change in
force over a period of time, equations (2), (8) and (9)
become:
N
W mp = 0W OF1sZ (10)
i=1
~M OFiXasi
Ximp = W x= j N (11)
smp
OF si
2=1
OMY ~ OF Yjsi
yimp- W (12)
unp ~ 0 P,2 S 1
where W1,,,p is the magnitude of the impulse force,
are the position coordinates of the impulse force, each
dFjsi term is a span corrected force change value, each
dFiXis j and dFiYis j term is a partial change in (or delta)
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 16 -
axial moment, and 4M,, and dM, are change in (or delta)
axial moments.
The ability to separately calculate impulse magnitudes
and positions separately from a static force enables a
scale to be used as a form of keyboard or user input device
even while an object is being weighed by the scale. Using
suitable impulse detection routines (e.g., by requiring a
minimum magnitude, by requiring an impulse to occur within
a time window, and/or by requiring an impulse to return to
the static condition), an impulse as a result of a user
input may be distinguished to enable the scale to perform
certain operations consistent with its environment, as will
be discussed below.
Mechanical Confiauration
Figure 4 is a partially cut away and partially
exploded perspective view of one suitable mechanical
configuration for scale 10, with the forward portion of
base 24 (including keyboard 28 and display 26) removed for
clarity. Base 24 houses load cells 30 and much of control
system 50 (shown in phantom). To this extent, base 24 may
take any shape consistent with its function of housing and
providing a sturdy support for the components of scale 10.
Platter 22 defines support surface 20, and is
supported by load cells 30. The contour and configuration
of operating surface 20 (and platter 22) may take many
forms depending upon the types of articles weighed, or the
types of forces subjected to the operating surface in
operation. For example, for three dimensional forces, the
CA 02247944 2006-02-17
79150-37
- 17 -
operating surface may define a three dimensional object. As
discussed above, platter 22 may even be omitted, with the
load cells themselves defining the operating surface.
Load cells 30 may be any type of force sensing
structure which provides a force value representative of the
force applied to the load cell. For example, the preferred
load cells are load cell assemblies of the type disclosed in
U.S. Patent No. 5,442,146 to Bell et al., and U.S. Patent
No. 5,313,023 to Johnson. This type of load cell assembly
generally includes a load cell body 32 with an interior
aperture having a base located on one wall thereof. A load
beam spans across the aperture from the base, and a pair of
cantilevered beams extend from the base generally parallel
to the load beam. A pair of force sensors are affixed
between the ends of the cantilevered beams and the load beam
such that, upon the application of force to the load cell
body, one sensor is placed in tension and the other is
placed in compression, whereby the sensors provide
complementary outputs responsive to the applied force.
Using suitable controller electronics (shown
disposed on circuit board 34 secured to the side of load
cell body 32), the resonators are driven to oscillate at
their respective resonant frequencies, whereby the frequency
output signals therefrom may be converted to digital form.
The value of the force applied to the load cell body is
obtained by taking a scaled difference of the outputs, since
the respective resonant frequencies of the resonators will
react oppositely to an applied force. By taking the
difference of the outputs, many common mode effects, such as
CA 02247944 2006-02-17
79150-37
- 18 -
due to environmental effects, will be rejected by the
subtraction operation. In addition, temperature
compensation, as well as other additional signal processing,
may be performed to improve the accuracy and precision of
the force value output from the load cell assembly.
Moreover, a communications function may be incorporated into
each load cell controller to permit the load cell assembly
to communicate with a master controller over a common load
cell bus, as is detailed in Bell et al.
It has been found that the preferred load cell
assembly design provides extremely high resolution with
reliable rejection of many environmental interferences.
Many modifications may be made to this design, including
several modifications disclosed in the aforementioned
references, including using only one resonator, sealing the
aperture with a sealant to protect the force sensors from
environmental effects, coating the wire leads from the force
sensors with a dampening material, etc.
Other load cells may be used in the alternative,
e.g., those disclosed in U.S. Patent No. 5,336,854 issued to
Thomas H. Johnson. Many other alternatives exist, e.g.,
strain gauge designs, flexured bending beam desigris (e.g.,
by Toledo, Revere, BLH, and others), vibrating string
designs (e.g., by K-Tron and others), force motor designs
(e.g., by Sartorius, Metler, A&D, Bizerba, and others),
capacitive coupling designs (e.g., by Setra and others),
piezoelectric
CA 02247944 1998-08-31
WO 97/33143 PCTIUS97/04136
- 19 -
crystal designs, tuning fork designs (e.g., by Ishida and
others), LVDT inductive force sensors, ultrasonic-based
force sensors, magnetostrictive force sensors and force-
responsive optical devices.
As shown in Figure 4, each load cell 30 is preferably
mounted between platter 22 and base 24 using a pair of
spacers 36 affixed by fasteners 38. Spacers 36 are
preferably constructed of a shock resistant material such
as phenolic, to protect the load cell from excessive impact
forces. Spacers 36 are preferably mounted at opposing ends
of the top and bottom surfaces of each load cell body, such
that a downward force applied to platter 22 will induce a
shear force across the load cell body, which may be sensed
and calculated by the load cell controller to generate the
force signal output therefrom. Other manners of mounting a
load cell, e.g., at the ends, at opposing ends of the top
or bottom surface, etc. may be used in the alternative. In
addition, stops may be used to limit the deflection of a
load cell body to protect the force sensors thereon from
failing due to excessive force. Side force isolation
devices may also be used to minimize deflection effects
between load cells for greater accuracy.
Load cells 30 may be positioned at different points
relative to platter 22. The preferred configuration shown
in Figure 4 spaces the effective locations of the load
cells (generally proximate the mount between the load cell
and the platter) closest to the perimeter of the platter,
thereby providing a greater range of responses of the load
cells based upon differing positions of an applied force.
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 20 -
Other positions for the load cells may be used in the
alternative. Moreover, as discussed above, greater or
fewer load cells may be used to provide different accuracy,
precision and/or resolution, or to provide one or three
dimensional position detection of an applied force.
Hardware Configuration
Returning to Figure 2, the primary hardware components
of scale 10 are illustrated. As discussed above, control
system 50 provides the primary operational control of the
scale, coordinating user interfacing with display 26 and
user input 28, and handling communications with load cells
30 over load cell bus 40. Control system 50 also provides
a magnitude sensing means and a position sensing means for
computing the magnitude and position of an applied force
from the force values obtained from the load cells. In
addition, an impulse detection means and user input means
are provided to detect, process and handle any user impulse
forces applied to the scale.
Control system 50 includes a microprocessor or
microcontroller and any necessary support circuitry, such
as RAM, ROM, power supply circuitry, clocking circuitry
etc. One suitable hardware configuration is disclosed in
Bell et al., for example, although other known hardware
components may also be used.
Control system 50 also preferably includes driving
circuitry for controlling display 26. Display 26 will also
vary depending upon the particular application, and may
include numeric information, alphanumeric information,
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 21 -
separate characters or icons, etc., depending upon the
information output requirements of the application.
Suitable displays include LEDs, LCDs, CRTs, vacuum
fluorescent displays, etc. For example, Bell et al.
discloses two vacuum fluorescent displays for a counting
scale application, as well as the driving circuitry
required therefor.
Control system 50 also preferably includes driving
circuitry for a user input device such as keyboard 28.
User input device 28 will also vary depending upon the
particular application. For example, alphanumeric keys,
soft (reprogrammable) keys, and function keys may be
provided, as well as alternate input devices such as mice,
trackballs, touchpads, etc. in addition, since the
preferred scale is capable of distinguishing impulse forces
and thereby detecting user depressions of platter 22 to
enable the platter itself to receive input, a separate user
input device may be omitted altogether.
In addition, control system 50 controls communications
over load cell bus 40. To this extent, control system 50
preferably includes suitable communications circuitry,
e.g., a Universal Asynchronous Receiver/Transmitter (UART),
as well as sui-table drivers and buffers. Each load cell 30
also includes similar circuitry for receiving and
transmitting information over the bus.
Control system 50 and load cells 30 communicate via
any known protocol, e.g., the RS485 or RS232 standards,
preferably in a master/slave relationship. Each load cell
preferably has a unique address which permits the load cell
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 22 -
to be individually controlled by the control system. A
command set, known by the control system and the load
cells, must be developed to enable information and commands
to be passed therebetween. The command set may also enable
a load cell to be calibrated and characterized with
calibration data and other identifying information. One
suitable communications protocol is generally discussed in
Bell et al.; however, a discussion of the specifics of a
suitable command set is not necessary for an understanding
of the invention.
Under a master/slave protocol, control system 50
requests each load cell to capture a force reading and
output a force signal to the control system for calculating
the magnitude and position of an applied static and/or
impulse force. The control system may store the span
correction and load cell positions in its memory, or
alternatively, the span correction and load cell position
data may be stored in each load cell individually, whereby
the load cells would provide moment and span corrected
values to the control system for calculating the magnitude
and position of a force.
Alternatively, a sequential chain communication
protocol may be used, whereby each load cell contributes to
the calculation of the magnitude and position of an applied
force. As discussed above, the position and magnitude of
an applied force is generally calculated using equations
(2), (8) and (9) for static forces, and equations (10),
(11) and (12) for impulse forces. Each of these
calculations includes a summation of partial data
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 23 -
calculated from each load cell. By breaking the summations
down into individual operations, each load cell may receive
a partial result from a prior load cell, add its component,
then pass the updated partial result to the next, or
subsequent, load cell for a similar operation. The value
returned by the last load cell in series then contains the
full calculations.
For example, for any load cell n, the summations in
equations (2), (8) and (9) above may be represented as:
Wn = W -1 + nsn (13)
=M~ n _ M(n 1) + Fn'YnSn
x
n (14)
W W _1 + Fnsn
y=~ M(n-1) + F nsn
n (15)
W W _1 + F n s n
where:
W _1 = j~ Fisi (16)
n-1
Fi-Xisi (17)
i=1
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 24 -
11~ (n-1) - i 1 FiYisi (18)
Therefore, any given load cell may contribute to the
calculation of a static force magnitude and position by
calculating partial values and adding the values to the
results provided from the earlier load cells in series.
Calculation of the impulse magnitude and positions may
also be carried out in a similar manner, replacing force
readings F with change in force readings dF. Equations
(13)-(15) then become:
W mp n= Wimp(n-1) jAFnSn (19)
AM n OMX(n-1) + AFnXnSn
X1mp n Wi'mp n Wimp{n-1) + AFnSn (20)
AM},n + AFnYnSn
y1mp n Wimp n Wfmp(n-1) + L~Fnsn (21)
where:
W.imp(n-1) AFisi (22)
CA 02247944 1998-08-31
WO 97/33143 PCTlUS97/04136
- 25 -
AMxtn-1J - ~ OF3 Xisi (23)
n-1
OFiYfsi (24)
Again, any given load cell may contribute to the
calculation of an impulse force magnitude and position by
calculating partial values and adding the values to the
results provided from the earlier load cells in series.
It will be appreciated that the position variables xn,
yn, xi, n and yin,p,, need not be calculated or passed from
load cell to load cell. Rather, only the axial moments and
total force values, Wn, M,Q,, Mn,, WjPn, aM,m and dMY,, need be
10- passed from cell to cell, whereby the positional data may
be computed at any time by dividing the axial static and
impulse moments by the total static and impulse forces,
respectively, as in equations (8), (9), (11) and (12).
Figure 5 shows a functional representation of the
sequential communications protocol for a scale 10' having a
control system-50' and a plurality of load cells 30'. To
calculate static and impulse forces, control system 50'
initiates a capture by sending a command to load cell A,
passing initial partial values for the total static and
impulse forces (Wo and Wi., o) and the axial static and
impulse moments (MXo, Mo- 4Mxo~ 4M,,o) to the load cell.
Null values are initially stored in these variables,
CA 02247944 1998-08-31
WO 97/33143 PCTIUS97/04136
- 26 -
thereby enabling the implementing software for each load
cell to operate in substantially the same manner. Once the
partial values are received by load cell A, the load cell
generates a force signal and a force change signal, and
adds partial values for the static and impulse forces and
axial moments. Then load cell A sends a command to load
cell B with the updated partial values, now designated W1,
Wi-I Mic1, MYI, 4Mx1, QMy. Similarly, load cell B passes
updated values to load cell C, load cell C passes updated
values to load cell D, etc., until the final load cell in
series, n, passes total values WN, Wim N, Mx,v, My,N, dMX,,,, LiMs,N
to control system 50'. The magnitudes are known, and the
positions are calculated by control system 50' by dividing
the axial moments by the magnitudes as detailed above.
Figure 6 shows a flowchart of a suitable program
routine 150 for each load cell controller to provide the
above described sequential protocol. Routine 150 is
preferably activated by a received command requesting an
update from the particular load cell n, which will
generally be initiated by a prior load cell in series n-1.
The first step in routine 150 is to obtain the prior
static and impulse partial values from load cell n-1 in
block 151. A force value Fn and a change in force value
[1Fn are then calculated by the load cell in block 152.
Next, in blocks 153 and 154, the partial values of the
magnitudes and axial moments are updated, using equations
(13)-(15) and (19)-(21) as discussed above. Then, in block
155, the updated values are output in conjunction with a
command to the next load cell in series, designated n+1.
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 27 -
Similar routines are used in each load cell 30', with
control system 50' acting as the first and last load cells
on the network to harmonize the routines in each load cell.
With each load cell having a unique address, as well as a
unique span correction factor and unique load cell position
data, the same basic operations may be used on each load
cell to provide reliable data. Moreover, if a load cell is
added or removed from the series connection, the routines
will still provide reliable data to control system 50' if
mechanical considerations are made and the system is
recalibrated.
It will be appreciated that other communication
protocols may be used in the alternative. Further,
different aspects of the position and magnitude
calculations may be allocated between the control system
and the load cell controllers, including snap commands
which cause all of the load cells to store instantaneous
readings for final weight and position calculations.
Software Confiauration
Figure 7 shows a main routine 100 for control system
50 of scale 10. Upon power-up, or after a reset, routine
100 will enter an initialization block 102 for initializing
variables, setting up interrupts, and performing other
start up operations to initialize the scale for operation.
A main loop is represented by blocks 102-110, which is
cycled through continuously. First, in block 103, any user
input on user input device 28 may be handled by the control
system, if such a device is included on the scale. The
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 28 -
user input will vary depending upon the particular
application, e.g., entering part numbers or weights on a
counting scale, etc.
Next, in block 104, force readings are obtained from
each of the load cells, preferably by passing a command
over the load cell bus to the load cells, then obtaining
the current force readings from each load cell over the
bus. As a result of this operation, current force readings
are obtained from each load cell.
Next, a COMPUTE FORCE AND POSITION subroutine 110 is
executed to determine the magnitude and position of the
applied force to the scale. In addition, impulse detection
is performed in this routine, whereby an IMPULSE flag may
be set if an impulse is detected. This routine is shown in
greater detail in Figure 8.
In block 112 of routine 110, the magnitude and
position of the applied static force on the scale is
calculated, e.g., by inputting the force readings obtained
from the load cells in equations (2), (8) and (9) discussed
above.
In block 114, the change in magnitude of the applied
force is compared to determine whether an impulse has
occurred. This may be performed by comparing the static
force value with a stored value from an earlier time frame,
or may be performed by determining the changes in the force
readings from earlier stored values, then using equation
(10) above to determine the magnitude of the force change.
Next, in blocks 116 and 118, the force change is
analyzed to determine if the change represents a valid
CA 02247944 1998-08-31
WO 97/33143 PCTlUS97/04136
- 29 -
impulse. In the preferred embodiment, the magnitude of the
force change is compared with a minimum magnitude value,
and the duration of the force change is compared with
minimum and maximum values to determine if the change
occurred within a "window" representative of impulses.
Other manners of detecting an impulse may also be
used, e.g., determining whether the force decays back to
the original force value, integrating a change over time,
etc. In addition, a routine for filtering the impulse from
the static force value may also be used.
if a force change does not meet the criteria
established in blocks 116 and 118, routine 110 terminates
and returns to routine 100. Otherwise, if a valid impulse
is detected, control passes to block 120 to calculate the
position of the impulse, e.g., using equations (11) and
(12) above. Then, in block 122, an IMPULSE flag is set to
indicate to the main routine that an impulse has been
detected. Control then returns to routine 100.
Returning to Figure 7, once the magnitude and position
of an applied force is determined in block 110, control
passes to block 106 to handle and/or display the static
force information. This routine will in general handle the
applied force data, which usually will encompass converting
the force into appropriate units and displaying the force
on display 26. In some environments, other processing may
be performed, e.g., computing a rate in a postal scale
based upon the position and magnitude of the force,
calculating the number of pieces in a counting scale, etc.
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 30 -
For example, several applications are discussed below which
require specific processing of the force data.
Once the static force information is handled, control
passes to block 108 to determine if an impulse was detected
by routine 110. If not, control returns to block.103 to
initiate the primary program loop again. If an impulse was
detected (from the status of the IMPULSE flag), then
control passes to block 109 to handle the impulse.
An impulse (user input) is handled in block 109
consistent with the particular application or environment
of the scale, using the magnitude and/or position of the
impulse to determine the particular operations to perform.
For example, different locations on the operat=ing surface
may be designated and labeled for user input, thereby
forming a "keyboard" on the operating surface, whereby an
impulse located at one of the locations will prompt the
control system to perform a specified function (e.g., zero
the scale, enter a tare weight, store a force, etc.). An
impulse may also indicate that an object has been placed on
the scale, e.g., to prompt the scale to calculate a postal
rate, to determine the compliance of a part with
specifications, etc. Other functions may be performed in
response to an impulse depending upon the particular
application or environment of use.
After an impulse is handled in block 109, control
returns to block 103 to initiate another pass of the main
program loop. Other functions, operations, etc. may be
performed in the main program loop. In addition, many of
the aforementioned operations, e.g., communicating with the
CA 02247944 1998-08-31
WO 97/33143 PCTIUS97/04136
- 31 -
load cells over the load cell bus, driving the display,
receiving input from the keyboard, etc., may be handled in
separate interrupts. In addition, various diagnostic,
calibration, setup, etc. routines may be provided in the
control system of scale 10.
Other modifications to the preferred routines may be
made consistent with the invention. For example, the load
cells may either return the force reading data, or may
perform some partial calculations (e.g., span or position
correction) and return the partial results to the scale
control system. In addition, as discussed above, the load
cells may also perform the actual position and magnitude
calculations and return the results to the scale.
Therefore, it will be appreciated that the position and
magnitude calculations described herein may be allocated in
various manners between the load cells and the scale
control system.
Anylications
The ability to determine the magnitude and position of
static and impulse applied forces has many applications in
a number of environments. One such application is for use
as a user input device, or keyboard. The principles of the
invention may be used to provide a user input function in
addition to a weighing or force sensing function, or simply
to provide a user input function by itself.
For example, Figure 9 shows a device 200 having an
operating surface 202. A weighing area 204 and one or more
predetermined areas or keys 206 are marked on the operating
CA 02247944 1998-08-31
WO 97/33143 PCTlUS97/04136
- 32 -
surface. Each key 206 may have a label designating its
particular function, e.g., "net", "zero", "tare", "print",
etc. By defining the areas of each "key" in terms of
coordinates on the operating surface, suitable software can
determine which key was activated by an impulse force from
a user to activate the corresponding predetermined
operations on the device for the selected key. Any number,
size and shape of keys may be defined on the operating
surface as long as suitable coordinates may be developed
therefor. In fact, a separate weighing area 204 is
optional since the entire surface may be covered with keys,
as an object to be weighed may be placed anywhere on the
operating surface, independent of the keys.
The keys may be activated by pressing with the hands
or fingers. Alternatively, in the case of a truck or
platform scale, the keys may be activated by stepping on a
designated area of the scale or moving from one position to
another.
Figure 10 illustrates another application directed to
sorting devices, where different functions or modes of
operation are initiated based upon the position of an
object on the device. For example, a device 210 shown in
Figure 10 may be used as a postal sorter, with a plurality
of zones 214 defined on an operating surface 212. By
placing a package to be shipped on the appropriate zone
corresponding to its destination, device 210 may calculate
and print out a label containing the correct postage, as a
function of both the weight and the destination of the
package. Figure 10 also illustrates a two load cell
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 33 -
embodiment of the invention, where two oppositely disposed
load cells may be used to determine positioning along a
single axis.
Other sorting applications may be envisioned, such as
density or type sorters, the latter having different areas
on the operating surface defined for different parts or
pieces. Separate compartments for holding the parts may
also be provided at predefined places on the surface. By
placing parts on appropriate regions of the operating
surface, a device may be able to customize its function for
a particular part. For example, in a counting scale
application, placing a group of like parts on their
appropriate region of the surface may enable the counting
scale to automatically recall the piece weight of the parts
for calculating the number of parts in the group. Further,
a running total of different types of parts may be
maintained from batch to batch.
Another application is a comparison device for use in
quality control to determine compliance of unknown parts
with a specification. For example, device 220 in Figure 11
includes a fixture such as a plurality of fixed pins 224
disposed on operating surface 222. An unknown object 226
may be placed-against pins 224, and its center of gravity
and total weight may be determined. A comparing routine is
used to detect any deviations in weight or position (i.e.,
center of gravity) from predetermined values corresponding
to a standard piece or specification (e.g., if a piece had
misplaced components or was out of tolerance). In
addition, sealed boxes may be analyzed in a similar manner,
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 34 -
e.g., to determine if parts are missing from a box. Thus,
device 220 may operate as a non-intrusive comparative
tester. This type of scale could have many different
fixtures for positioning objects to be compared to
standards.
An additional application is in the area of process
control, as exhibited by device 230 in Figure 12. Device
230 includes process equipment disposed on an operating
surface 232. The process equipment, for example, may be
equipment for transporting material from a surge tank 237,
to a hammer mill 235 using an auger 238, and to a mixer 234
using an auger 236. By using a suitable status routine to
detect the magnitude and center of gravity of the material
in the process equipment, the material may be followed from
station to station in a non-intrusive manner. No separate
monitors or sensors are required, and the equipment
enclosures need not be opened or modified.
Preferred embodiments of the invention may have
utility in many other applications and environments of use,
e.g., in other scale/sorting applications. Further, any
combination of the above functions (e.g., center of
gravity, keyboard, sorter, center of gravity movement
detection, etc.) may also be implemented in a particular
application, and any area on a support surface may be
defined to function in any or all of these ways (e.g., a
process control system may have a keyboard entry area on
the supported structure). Thus, the invention should not
be limited by the particular applications and use
environments disclosed herein.
CA 02247944 1998-08-31
WO 97/33143 PCT/US97/04136
- 35 -
It will therefore be appreciated that the invention
may provide a number of benefits and advantages as a result
of determining the position, as well as the magnitude, of
an applied force. As various changes and modifications may
be made to the preferred embodiments without departing from
the spirit and scope of the invention, the invention
therefore resides in the claims hereinafter appended.
õ~..~~ ~'~Z/if'. ~ ~~i :1 f .1~~, 5~~~1t~~
