This invention relates to a multiple-filler filling machine operating
on a weight-base, to fill containers, such as bags, bottles and boxes, with a
predetermined amount by weight of an article in the form of, for example, liquid,
powder or particles, and, more particularly, to a weight measuring system which
may be employed for such filling machine.

BACKGROUND OF THE INVENTION
A multiple-filler filling machine of the above-described type typically
includes a plurality of fillers or filling devices for filling a plurality of containers
with an article in parallel, in series or at random. An example of weight-based,
multiple-filler filling machine for filling containers with liquid is shown in
FIGURES 1A, 1B and 1C. The multiple-filler filling machine shown includes a reservoir
2 in an upper portion of the filling machine 1, into which a liquid for filling
containers with is externally fed through a rotary joint 4, piping 6, and such.
The liquid is temporarily stored in the reservoir 2. Around the reservoir 2, a
plurality of filling pipes 8 (8-1, 8-2, ..., 8-n) are mounted at locations angularly
spaced from each other by a predetermined angle. Valves 10 (10-1 through 10-n)
are mounted at the lower ends of the respective pipes 8-1 through 8-n, for controlling
the flow rate of the liquid to be supplied through nozzles associated with the
respective ones of the valves 10-1 through 10-n. A plurality of filling platforms
14 (14-1 through 14-n) are disposed beneath the respective ones of the valves 10-1
through 10-n, around the lower portion of the filling machine 1. Each of the platforms
14-1 through 14-n is adapted for receiving a container 12 (12-1 through 12-n),
e.g. a bottle, thereon. The respective platforms 14-1 through 14-n are coupled
to respective load cells 16 (16-1 through 16-n) used in association with the respective
filling platforms 14-1 through 14-n. The load cells 16-1 through 16-n are mounted
on a load cell table 18 in the lower portion of the filling machine 1. The reservoir
2, the load cell table 18 and the filling platforms 14-1 through 14-n are adapted
to rotate at a predetermined rate about a rotation shaft 24 extending vertically
through the center of the filling machine 1 when driven by external driving means,
e.g. an electrical motor 20, through gears 21 and 22.

Empty bottles are supplied one by one from an external container
supplying machine (not shown) to a container entry section of the filling machine
1. The container entry section is on the circle along which the filling platforms
14-1 through 14-n rotate. The empty bottles are put on, one by one, the filling
platforms 14-1 through 14-n. The bottles 12-1 through 12-n are weighed, while they
are moving, and a predetermined amount of liquid is put in each of the bottles
12-1 through 12-n. The filled bottles 12-1 through 12-n are conveyed to a discharge
section of the filling machine 1, which is at a different location on the circle
from the entry section, from where they are discharged from the filling machine
1 to a conveyor in the succeeding stage.

The valves 10-1 through 10-n rotate with the bottles 12-1 through
12-n and are opened when they arrive at a predetermined location on the circle
along which the valves 10-1 through 10-n are moving. Then, the liquid is poured
into the bottles 12-1 through 12-n at a predetermined flow rate, and, at the same
time, the weight of the liquid which has been poured into each bottle 12 is measured
by the load cell 16 for that bottle. The apertures of the valves 10-1 through 10-n
are so controlled that, when the weight of the liquid poured into each bottle 12
reaches a first value slightly smaller than a target or aimed value, the flow rate
can be reduced. The valves 10-1 through 10-n are closed when the weight of the
liquid in the respective bottles 12-1 through 12-n reaches a second value which
is larger than the first value but is just below the target value. Even after the
valves 10 are closed, the liquid portions remaining in the path between the valves
10 and the bottles 12 fall into the bottles 12, and the target amount of the liquid
is poured into each bottle 12.

In order to know the weight of the liquid in each bottle 12, an analog
weight-representative signal from the associated load cell 16 must be subjected
to analog-to-digital conversion to produce a digital weight-representative signal,
and the resulting digital weight-representative signal must be subjected to weighing
arithmetic operations. For that purpose, a plurality of weighing arithmetic operation
means, e.g. arithmetic units 26 (26-1 through 26-n), are employed in association
with the respective load cells 16-1 through 16-n. The number of the arithmetic
units 26 is equal to the number of the filling platforms 14 (14-1 through 14-n).
The arithmetic units 26-1 through 26-n are mounted on an arithmetic unit table
28, which is disposed between the reservoir 2 and the load cells 16-1 through 16-n.
The arithmetic units 26-1 through 26-n are connected to the associated load cells
16-1 through 16-n with respective signal lines 28.

The arithmetic operations performed in each arithmetic unit are as
follows, for example. Description about one arithmetic unit to be given hereinafter
is applicable to all and any arithmetic units, and, therefore, in the following
description, the suffixes to the reference numerals are not used.

The weight of the object to be measured loaded on the load cell 16,
i.e. the weight of the article, e.g. liquid, and the weight of the filling platform
14 and the bottle 12 in case of the filling machine, are measured and developed
in the form of an analog weight-representative signal, and the analog weight-representative
signal is AID (analog-to-digital) converted into a digital weight-representative
signal Wa.

During adjustment of a weight measuring system employed in the filling
machine, the digital weight-representative signal developed when nothing is placed
on the platform 14 is stored as an initial value Wi in a memory in the arithmetic
unit 26. When the bottle 12 is placed on the filling platform 14 and the filling
operation starts, the digital weight-representative signal Wa starts to increase.
The sum weight Wn of the bottle 12 and the liquid which has been poured into the
bottle 12 on the filling platform 14 is calculated in accordance with an expression:
Wn = K(Wa - Wi)
where K is such a span factor determined, during the weight measuring system adjustment,
by using a reference weight placed on the filling platform 14, that the resultant
Wn can be equal to the weight of the reference weight. The span factor K is determined
based on the loaded load on the load cell 16, a voltage conversion factor, and
an analog weight-representative signal amplification factor employed in the arithmetic
unit 26. Thus, the span factor K is dependent on both the load cell 16 and the
arithmetic unit 26.

Even when the bottle 12 is not on the filling platform 14, the value
Wn is not zero if, for example, a drop of water is on the platform 14. Then, the
weight of an object on the platform 14 Is expressed by
Wn = K(Wa - Wi) - Wz
where Wz is an amount of change of the zero point. If Wn is not equal to zero (0)
when no bottle 12 is on the filling platform 14, that Wn is stored as Wz. This
procedure is called zero-point adjustment.

In the filling machine, first, only an empty bottle 12 is carried
onto the filling platform 14, and, therefore, the value of Wn in such a case is
representative of the weight Wb of the bottle 12 itself. Immediately before the
filling of the bottle 12 with a liquid is started, the weight of the bottle 12
is measured and stored in a taring memory. As the filling of the bottle 12 with
the liquid starts, Wn becomes representative of the sum of the weights of the
bottle 12 and the liquid in the bottle. Then, the operation of Wn Wb is performed
to determine the weight Wm of the liquid only. Alternatively, the zero-point adjustment
may be done with the weight of the bottle 12 also taken into account so that Wn
can represent the weight of the liquid only. In this case, Wn = Wm.

A plurality of weight levels w1, w2 and wt are set in the arithmetic
unit 26 for use in controlling the flow rate of the liquid. The weight level wt
corresponds to the target weight. The level w2 is a value close to wt, and the
level w1 is lower than w2. In the beginning of the fllling, the valve 10 is controlled
to supply the liquid through the nozzle at a flow rate of q1. The weight Wm of
the liquid in the bottle 12 gradually increases, and is repeatedly compared with
the weight level w1 at short intervals to find whether Wm > w1, When Wm becomes
larger than w1, the valve 10 is controlled to change the flow rate from q1 to q2,
where q2 > q1. The liquid supply is continued at the flow rate of q2, while
checking at frequent intervals as to whether Wm > w2. When it is determined
that Wm has become larger than w2, the valve 10 is controlled to stop the supply
of the liquid. Even after the valve 10 receives a command to close itself, a small
amount of the liquid is still supplied to the bottle 12, whereby the bottle 12
is filled with the liquid in the amount substantially equal in weight to the target
weight wt.

In weight-based filling machines, in order to determine how much liquid
has been ultimately put into the bottle 12, the liquid in the bottle 12 is weighed
again a predetermined time after the supply of the liquid to the bottle 12 is
stopped, and the measured weight is outputted as a fill weight.

The arithmetic units 26-1 through 26-n rotate with the load cell table
18 and the filling platforms 14-1 through 14-n, and the power for the arithmetic
units 26-1 through 26-n are externally supplied through a rotary connector having
a rotation axis common to the rotary joint 4. The arithmetic units 26-1 through
26-n are controlled by external, remote displaying and controlling means, e.g.
display and control unit 30, to which the arithmetic unit 26-1 through 26-n send
data including the weights of the liquid poured into the respective bottles 12.
For that purpose, the respective arithmetic units 26-1 through 26-n are connected
to the display and control unit 30 via serial communication lines so as to enable
bi-directional communications therebetween. As the number of serial communications
lines increases, the rotary connector must have an increased number of contacts
thereon, which makes the rotary connector expensive. Therefore, corresponding signal
lines led out from the respective arithmetic units 26-1 through 26-n are connected
in common on a terminal pad and connected to the display and control unit 30 by
a minimum of two lines.

The connections between the load cells (LC) 16-1 through 16-n, the
arithmetic units 26-1 through 26-n, and the display and control unit 30 are shown
in FIGURE 1B. In FIGURE 1B, each of leads 31-1 through 31-n from the respective
load cells 16-1 through 16-n shown includes analog weight-representative signal
lines and power supply lines. Terminal pads 32-1 through 32-n are used to connect
the associated leads 31-1 through 31-n to the associated arithmetic units 26-1
through 26-n. Lines 33-1 through 33-n connect the terminal pads 32-1 through 32-n
to the arithmetic units 26-1 through 26-n. The control signals for controlling
the valves 10-1 through 10-n are supplied from the associated arithmetic units
26-1 through 26-n. For that purpose, the valves 10-1 through 10-n are connected
to the arithmetic units 26-1 through 26-n by leads 34-1 through 34-n, the terminal
pads 32-1 through 32-n, and leads 35-1 through 35-n.

In order to control the filling of the bottles 12-1 through 12-n with
the liquid when each of the valves 10-1 through 10-n reaches a predetermined filling
start position, the arithmetic units 26-1 through 26-n have to find the position
of the respective valves 10-1 through 10-n. For that purpose, a position-representative
signal generating unit 36 is provided for the valves 10-1 through 10-n, The position-representative
signal is transmitted via the terminal pad 38, where they are connected in common
and coupled to the respective arithmetic units 26-1 through 26-n via wiring 40-1
through 40-n.

Corresponding communication lines 42-1 through 42-n led out from the
respective arithmetic units 26-1 through 26-n are connected in common on a terminal
pad 44, from where they are connected via two signal lines 46 through a rotary
connector 48 to the display and control unit 30.

If the filling machine handles liquid to fill containers with, it
is cleaned to keep it sanitary. The load cells 16-1 through 16-n, therefore, should
be placed within watertight enclosures 50-1 through 50-n, respectively, as shown
in FIGURE 1C. Support members 52-1 through 52-n extend out of the respective load
cells 16-1 through 16-n, and labyrinth seals 54-1 through 54-n are disposed in
the openings of the respective enclosures 50-1 through 50-n, through which the
support members 52-1 through 52-n extend out.

The above-described type of weight-based, multiple filler filling
machine includes a load cell table 18, The table 18 is relatively small in area,
and many components, including the load cells 16-1 through 16-n, and the arithmetic
units 26-1 through 26-n are to be disposed on it. Therefore, the load cell table
18 must have an area as large as required for these components. In addition, since
the load cells 16-1 through 16-n are separated from the respective arithmetic units
26-1 through 26-n, they must be connected by the relatively long wiring 31-1 through
31-n and the wiring 33-1 through 33-n via the terminal pads 32-1 through-32-n,
respectively, and the size of the wiring must be adjusted in accordance with the
size of the filling machine, which has impeded reduction of wiring costs.

If it is found that the amount of the liquid supplied to each bottle
is not appropriate, an operator may desire to adjust the machine as soon as possible.
In such a case, it is difficult to determine which are out of order, the load cells
16 or the arithmetic units 26, and, therefore, both need be replaced by new ones.
In addition, when replacement of components is made, the span factor K must be
determined anew by placing the reference weight on the filling platforms so that
the weight equal to the reference weight can be displayed on the display and control
unit 30.

If manufacturers of weight-based, multiple filler filling machine
desire to have arithmetic operations of the weight-based filling machine performed
by general information processing machines or hardware and software for general
information processing machines, they usually do not have know-how to have required
weighing arithmetic operations performed by such machines. Consequently, they must
buy dedicated filling machine arithmetic units from weighing machine manufacturers,
and manage to place them together with the load cells 16-1 through 16-n on the
load cell table 18 which has little room.

Both hardware and software of prior art arithmetic units 26 are designed
in accordance with design concepts of manufacturers of weighing machines. Users
of filling machines, therefore, are restricted in various ways when they try to
build up, based on their own concepts of filling machine users, operating procedures
and data display procedures on the display screen of the display and control unit,
for example, which must be easily operable.

Although waterproof load cells have been used in weight-based filling
machines, there have been no waterproof arithmetic units. Therefore, troubles
have to be taken to provide watertight covers and labyrinth seals for the arithmetic
units.

Therefore, an object of the present invention is to provide a weight-based,
multiple filler filling machine free of the above-discussed problems.

SUMMARY OF THE INVENTION
One solution to the above-described problems may be to integrate or
combine load cells and weighing arithmetic operation means into a single, integrated
weight measuring system so that they can be manufactured, installed and managed
integrally.

For example, weighing arithmetic operation means may be mounted on
the surface of or inside a load cell. As an alternative example, a load cell and
weighing arithmetic operation means may be put in a watertight enclosure together.

In these days, it is able to downsize electronic components significantly,
and, therefore, it has become possible to design small-sized electronic circuits
which can perform highly complicated functions, regardless of whether they are
analog or digital. On the other hand, load cells used in a weight-based, multiple
filler filling machine have a load sensing element, which, when a load is applied
to it, is deformed to a degree in accordance with the magnitude of the load. The
load sensing element has a strength sufficient to bear weight of several kilograms,
and, accordingly, the load cells have a volume sufficient to accommodate the filling-machine
weighing arithmetic operation means.

The use of integrated units including such integrated or combined
load cell and weighing arithmetic operation means increases the efficiency of
manufacturing weight-based, multiple filler filling machines.

A weight-based, multiple filler filling machine includes a large number
of load cells and arithmetic operation means, failure tends to occur very frequently.
By keeping a reserves of such integrated load cell and weighing arithmetic operation
means units, it is readily possible to replace a failing load cell and/or arithmetic
operation means with good ones. The maintenance efficiency can be further increased
by keeping integrated load cell and weighing arithmetic operation means units as
reserves for future use, with the span factor K adjusted beforehand using a reference
weight, which eliminates the need for the span adjustment during replacement.

In Japanese Patent No. 2,709,837 issued on October 24, 1997, which
corresponds to U.S. Patent No. 4,815,547 issued to Benny N. Dillon et al. on March
28, 1989, an weighing apparatus including an arithmetic operation circuit board
integrally mounted on a load sensing element, or counterforce, of a load cell is
disclosed. In the weighing apparatus disclosed in this patent, a plurality of load
cells support one table on which an article to be measured is placed, and a circuit
board with hardware including an AID converter, a CPU and serial communications
circuitry, is mounted on a side surface of each load cell. Digital weight-representative
signals output by the respective circuit boards are collected in a main controller,
from which a single measurement signal is developed as an output of the weighing
apparatus. The load cells of this weighing apparatus interfere with each other
via the table. It may be possible to arrange that the individual weight-representative
signals from the respective load cells can be read on the main controller. However,
if the single measurement signal of the weighing apparatus is abnormal due to
mechanical troubles, evaluation of the individual weight-representative signals
from the respective load cells cannot identify which one of the load cells is out
of order.

According to the present invention, each integrated unit of a load
cell and arithmetic operation means used in a weight-based, multiple filler filling
machine weighs an article in an individual one of containers and, therefore, is
not interfered by other load cells or articles in other containers. Accordingly,
it is easy to find which one of the load cells is out of order.

The load cells with the arithmetic operation means disposed on the
surface of or inside the load cells may be hermetically sealed, or each load cell
may be disposed together with an associated arithmetic unit within a hermetically
sealed enclosure. Such hermetically sealed units can be used for a weight-based,
multiple filler filling machine for filling containers with a liquid, without need
for using special watertight arrangements, such as a labyrinth.

An arithmetic operation means manufacturer may design the arithmetic
operation means for the filling machine to be operated through external hardware,
such as an external key switch having a fixed characteristic, or to output data
to be displayed in a specific form on a specific display. In such a case, even
if the arithmetic operation means is designed to be able to communicate with external
apparatuses, it is not possible for a filling machine manufacturer to produce a
filling machine based on an optimum design concept.

In order to solve this problem, in the weighing arithmetic operation
means according to the invention, instruction codes may be assigned to specific
weighing arithmetic operations, such as the zero-point adjustment, the span adjustment
and the tare weight storage. A selected instruction code is provided by the display
and control means so that the arithmetic operations corresponding to the selected
instruction code can be performed. In this case, data to be displayed generated
in the weighing arithmetic operation means, such as, for example, the weight of
an article which has been placed in a container, and the weight of the steady-state
article in the container measured after the filling has been completed, are previously
assigned with respective identification codes and outputted together with the assigned
codes so that they can be handled or processed as desired in the display and control
means. The respective codes and their meanings are open to users including filling
machine manufacturers. Accordingly, filling machine manufacturers can control
the weighing arithmetic operation means through display and control means they
have manufactured , using desired sequences and desired operating means, e.g. desired
key switches. Also, it is possibie for the filling machine manufacturers to cause
data outputted by the arithmetic operation means to be displayed in any desired
form at any desired position on the display and control means. This means that
the filling machine manufacturers can produce weight-based, multiple filler filling
machines based on their own design conceptions.

For example, communications between a plurality, N, of weighing arithmetic
operation means and single display and control means may be done through a 1:N
serial communication line. For example, the display and control means is set to
be a master device with the N arithmetic operation means set to be slave devices.
The display and control means and the respective arithmetic operation means are
connected directly to the same communication line bus, and the display and control
means, which is the master device, sequentially polls the arithmetic operation
means which are the slave devices. Thus, there is no need to use a unit, e.g. an
intermediate terminal pad, on which display data from the respective arithmetic
operation means are collected for reducing the number of connecting lines.

As for wiring for the weighing arithmetic operation means, lines for
signals outputted by the arithmetic operation means are conventionally connected
to terminals on a terminal pad for reducing the number of lines for sending signals
therethrough to the display and control means, by connecting similar signal lines
together. However, such arrangement requires the use of the terminal pad, the connections
of the respective arithmetic operation means to the terminal pad, the wiring between
terminals on the terminal pad for common connections of the related signal lines,
and so forth. This also increases the wiring to the terminal pad.

In a weight-based, multiple filler filling machine, load cells are
arranged with a small distance between adjacent ones. An example of advantageous
wiring for the weighing arithmetic operation means for a weight-based, multiple
filler filling machines is as follows. (This arrangement is more advantageous
than the above-described one in which wires from a number of arithmetic operation
means are collected together on a single terminal pad.) Each weighing arithmetic
operation means includes a printed circuit board on which one signal is outputted
at two output connectors. Similar output connectors of arithmetic operation means
disposed adjacent to each other are connected together by means of a short wiring
unit. Using similar short wiring units, a number of adjacent arithmetic operation
means are connected in series. In this case, the number of the short wiring units
is equal to the number of the serially connected arithmetic operation means minus
one. With this wiring arrangement, the amount of lines and the wiring operation
can be reduced.

The polling requires a relatively long time in communication, and,
therefore, it is difficult to employ the polling technique for controlling the
article supplying rate in a high-speed, high-performance filling scale. Accordingly,
each of the weighing arithmetic operation means itself may be arranged to have
a function to perform weighing arithmetic operations, while the display and control
means Is arranged not to be involved with the weighing arithmetic operations which
should be performed in real time. With this arrangement, the sending of a signal
to command the filling means, e.g. a valve, to reduce the article supply rate or
to stop the supply of the article, upon detection of the article reaching the set
levels in the container, which must be done at a high speed, can be made independent
of the communication rate of the communication means.

Data transferred between the display and control means and the respective
weighing arithmetic operation means include, but not limited to, parameters required
by the arithmetic operation means to perform the necessary weighing arithmetic
operations, data for controlling the supply flow rate, commands for zero-point
adjustment, span adjustment etc. and data to be displayed after the completion
of the filling operation generated in the respective weighing arithmetic operation
means, all of which do not require high-speed, real-time processing.

A weight-based, multiple filler filling machine includes a plurality
of load celts, It may be operated to fill a number of containers simultaneously.
In such a case, simultaneous, parallel processing of weight-representative signals
from a plurality of load cells is required. In addition, different from static
weighing, there are restrictions on weight-representative signal sampling and
processing times. Furthermore, in order to fill containers precisely with an aimed
amount of the article, the filling means must be controlled, too.

According to the present invention, load cells signals from which
must be processed in real time may be arranged as subordinate arithmetic operation
means to provide digital weight-representative signals. Then, superordinate arithmetic
operation means is provided for collectively processing input and output signals
from and to the respective subordinate arithmetic operation means. The superordinate
and subordinate arithmetic operation means may be connected by serial communication
lines. The superordinate arithmetic operation means need to control the article
supply rate on the basis of the weight of the article in the respective containers
provided thereto from the associated load cells. Accordingly, the filling means
may be controlled in real time, too.

As described above, by assigning to load cells, a function to measure
weights of an article in a number of containers in a parallel, real-time and high-speed
fashion, while using superordinate arithmetic operation means smaller in number
than the load cells for controlling the filling operation, the number of weighing
arithmetic operation means required for the system can be reduced, which, in turn,
results in down-sizing of the filling machine.

BRIEF DESCRIPTION OF THE DRAWINGS

- FIGURES 1A, 1B and 1C show a prior art weight-based, multiple filler filling
machine, in which FIGURE 1A is its side view, FIGURE 1B is a block circuit diagram
of the prior art filling machine, and FIGURE 1C is an enlarged view of a load cell
used in the filling machine of FIGURE 1A;
- FIGURE 2 is a block circuit diagram of a weight-based, multiple filler filling
machine according to a first embodiment of the present invention;
- FIGURE 3 is a schematic side elevational view of the filling machine of FIGURE
2, with its part removed;
- FIGURE 4A shows a load cell used in the filling machine shown in FIGURES 2
and 3, and FIGURE 4B is an enlarged view of the load cell of FIGURE 4A;
- FIGURE 5 is a plan view of another example of a load cell useable in the filling
machines of the present invention;
- FIGURE 6 is an enlarged side view of a load cell useable in a weight-based,
multiple filler filling machine according to a second embodiment of the present
invention;
- FIGURE 7 is a block circuit diagram of a weight-based, multiple filler filling
machine according to a third embodiment of the present invention; and
- FIGURE 8 is a block circuit diagram of a weight-based, multiple filler filling
machine according to a fourth embodiment of the present invention.
- FIGURE 9 shows another example of load cell module useable in the present invention,

DETAILED DESCRIPTION OF THE INVENTION
A weight-based, multiple filler filling machine 1a according to a
first embodiment of the present invention is shown in FIGURES 2, 3, 4A, 4B and
5.

As is understood from comparing FIGURE 1A with FIGURE 3, the filling
machine 1a includes, in place of the weighing arithmetic units 26-1 through 26-n
and the discrete load cells 16-1 through 16-n, arithmetic unit and load cell modules
16a-1 through 16a-n each including a load cell and a weighing function arithmetic
unit. Hereinafter, the arithmetic unit and load cell module is referred to as "load
cell module". In FIGURES 2 through 8, those components which are same as or similar
to the ones shown in FIGURES 1A, 1B and 1C are shown with the same reference numerals
as used in FIGURES 1A, 1B and 1C, and no further description thereof is given.

Each of the load cell modules 16a-1 through 16a-n includes a load
cell having a load sensing element with a cavity 81 (FIGURE 4A) formed in a portion
thereof at which the load sensing element is fixed to a base, and an associated
one of arithmetic units 26a-1 through 26a-n is placed in the cavity 81 (FIGURE
4A). The load sensing element has a cylindrical bore 82 (FIGURE 4B) for detecting
weight. The cylindrical bore 82 is deformed into an elliptic shape when a load
is applied to the load sensing element. Strain gages 83, 84, 85 and 86 are attached
on the inner surface of the cylindrical bore 82 for measuring the amount of strain.
A cylindrical member 87 is inserted into the cylindrical bore 82 to cover the inner
surface of the bore 82 and the inner surfaces of the strain gages 83, 84, 85 and
86. The cylindrical member 87 has a lid member attached to its one end, which is
placed on and welded to a peripheral step formed in the corresponding end of the
cylindrical bore 82, to thereby hermetically seal the strain gages 83, 84, 85 and
86 from the open air. Wiring between the strain gages 83, 84, 85 and 86 and the
arithmetic units 26a-1, ..., and 26a-n is done through a communication bore 88
formed in the load sensing element. The function of the arithmetic units 26a-1
through 26a-n is the same as that of the prior art arithmetic units 26-1 through
26-n. The open ends of the respective cavities are closed by welding a metal lid
80 to seal the cavities watertight and airtight. Thus, no special waterproofing
measures need be provided separately for the arithmetic units 26a-1 through 26a-n
and the strain gages 83, 84, 85 and 86. No labyrinth is required. Furthermore,
no wiring need be provided between the respective ones of the load cells 16a-1
through 16a-n to the corresponding ones of the arithmetic units 26a-1 through 26a-n.

FIGURE 5 shows another example of a load cell, in which the arithmetic
units are not disposed within the respective load sensing elements, but they are
placed in separate hermetic enclosures 89 mounted to the outer surfaces of the
respective load sensing elements.

FIGURE 2 shows how the components of the filling machine 1a are connected.
The load cell modules 16a-1 through 16a-n perform the same functions as the load
cells 16-a through 16-n and the arithmetic units 26-1 through 26-n of the prior
art filling machine. Accordingly, there is no wiring corresponding to the wiring
33-1 through 33-n shown in FIGURE 1B.

Each of the arithmetic units 26a-1 through 26a-n in the load cell
modules 16a-1 through 16a-n are provided with common signal input/output terminals
81a, 81b, 81c, 82a, 82b and 82c. (In FIGURE 2, the reference numerals 81a, 81b,
81c, 82a, 82b and 82c are shown only for the arithmetic unit 26a-2.) Signals are
outputted from these terminals 81a, 81b, 81c, 82a, 82b and 82c to output connectors
83a, 83b, 83c, 84a, 84b arid 84c, respectively. (In FIGURE 2, the reference numerals
83a, 83b, 83c, 84a, 84b and 84c are shown only for the module 16a-1.)

In a rotary or line-type, weight-based, multiple filler filling machine,
the load cell modules 16a-1 through 16a-n are equiangularly disposed. Prior to
the assembling of the filling machine 1a, the output connectors 83a, 83b, 83c,
84a, 84b and 84c of one load cell module are connected to those of adjacent load
cell modules by means of wiring units 85-1 through 85-n having the same wire length.
This can reduce the time for wiring the filling machine 1a.

Instructions, such as zero-point adjustment, tare weight storage and
span factor adjustment, from the external display and control unit 30 to the respective
arithmetic units 26a-1 through 26a-n of the respective load cell modules 16a-1
through 16a-n are provided in the form of predetermined code signals open to users.
Accordingly, a filling machine designer can determine, at his or her discretion,
operation procedures and assignment of functions to respective key switches in
the display and control unit 30. For example, a desired instruction code may be
generated by means of a software switch displayed on a screen, to thereby cause
the arithmetic units 26a-1 through 26an of the load cell modules 16a-1 through
16a-n to perform the desired instructions.

The arithmetic units 26a-1 through 26a-n output various data with
predetermined identification codes open to the users assigned thereto. The data
include steady-state weights of the article or liquid in the respective containers
or bottles measured after the completion of the filling operation, and span factors
determined in the span adjustment. Therefore, the display and control unit 30 can
judge what are meant by the identification codes attached to the data to thereby
display the data in appropriate character styles and shapes at appropriate positions
on appropriate pictures, graphically if necessary.

The load cell modules 16a-1 through 16a-n include the arithmetic units
26a-1 through 26a-n, respectively. The display and control unit 30 is designed
such that the values w1, w2 and wt, previously described with reference to the
prior art, which are used to control the flow rate of the liquid and to stop the
supply of the liquid, can be set together with predetermined codes assigned to
them in the display and control unit 30. Then, these values are automatically supplied
to memories of the respective arithmetic units 26a-1 through 26a-n and are set
therein.

Each of the arithmetic units 26a-1 through 26a-n samples an analog
weight-representative signal from the associated one of the load cells 16a-1 through
16a-n at short time intervals, computes the weight of the liquid in the associated
one of the bottles 12-1 through 12-n by performing predetermined arithmetic operations,
judges when the weight of the liquid in the bottle reaches a predetermined level,
and controls the valve as soon as the weight of the liquid reaches the predetermined
level to thereby adjust the flow rate. Performing arithmetic operations for the
liquid weight in each bottle at short time intervals makes it possible to provide
a desired control signal to the associated valve without delay. Therefore, the
load cell modules 16a-1 through 16a-n with the arithmetic units 26a-1 through 26a-n
built therein, respectively, do not adversely affect the filling precision.

The respective arithmetic units 26a-1 through 26a-n associated with
the load cell modules 16a-1 through 16a-n control the flow rates of the liquid
supplied to the respective bottles 12-1 through 12-n, and the load cell modules
16a-1 through 16a-n output control signals directly to the associated valves 10-1
through 10-n associated with the filling platforms 14-1 through 14-n. Therefore,
the leads 34-1 through 34-n for transmitting the control signals therethrough can
be connected directly to the respective valves 10-1 through 10-n, whereby the space
and time for the connections can be minimized.

FIGURE 6 shows part of a weight-based, multiple filler filling machine
according to a second embodiment of the present invention. The structure of the
filling machine according to the second embodiment is substantially the same as
the filling machine according to the first embodiment, and, therefore, only portions
different from the first embodiment are described.

Load cell modules 16b-1 through 16b-n are provided with arithmetic
units 26b-1 through 26b-n integrally mounted on the surfaces of load sensing elements
of the respective load cells. In FIGURE 6, only one load cell module, namely, the
load cell module 16b-1 and its associated parts are shown. The remaining load cell
modules 16b-2 through 16b-n are similarly arranged to the load cell module 16b-1.
The load cell modules 16b-i through 16b-n are individually placed in associated
ones of watertight enclosures 90. The load cell module 16b-1 has a member 92 with
which it is secured to the filling platform 14-1. A waterproofing member 94, e.g.
a labyrinth member, is mounted on the member 92.

Although this structure requires a waterproofing arrangement, the
arithmetic units 26b-1 through 26b-n, which are mounted on the surfaces of the
respective load sensing elements of the load cells, are also waterproofed, and
wiring can be simplified. Accordingly, the manufacturing cost can be reduced.

It should be noted that, in the first embodiment, too, the arithmetic
units 26a-1 through 26a-n can be mounted on the surfaces of the load sensing elements
of the associated load cells, and the load cell modules can be sealed to make them
watertight.

FIGURE 7 shows a filling machine according to a third embodiment of
the present invention. The filling machine includes load cell modules 16c-1 through
16c-n, each of which includes a load cell and an arithmetic operation circuit integrally
mounted on or in the load cell. Each of the arithmetic operation circuit includes
an A/D converter for AID converting an analog weight-representative signal outputted
by the associated load cell into a digital weight-representative signal, and a
communication circuit for serially transmitting the digital weight-representative
signal to an external apparatus.

Each of the load cell module 16c-1 through 16c-n includes an arrangement
for processing the analog weight-representative signal before it is converted into
the digital weight-representative signal. A reserve of load cell modules of this
type, which are span adjusted beforehand by the use of a reference weight having
the reference weight, may be kept so that a new one can be substituted immediately,
without need for any adjustment, for a malfunctioning load cell module found during
operation of the filling machine. The load cell modules according to the first
and second embodiments may be arranged similar to the ones of the third embodiment.

Another, discrete superordinate arithmetic operating unit 100 is used.
The superordinate arithmetic unit 100 receives the digital weight-representative
signals from the load cell modules 16c-1 through 16c-n, and performs various arithmetic
operations. The arithmetic operations performed by the unit 100 include the initial
tare-offsetting arithmetic operations for compensating for weight measurement variations
essentially due to variations in the mounting, geometry and the like of the respective
load cells, the weight measurements due to the respective filling platforms, and
the measurements due to the offsets in the amplifiers, the zero-point adjustment
operations, the article weight computation operations in which the weights of the
containers are subtracted, and the weight level determining arithmetic operations.
Alternatively, the arithmetic operations, except, for example, the weight level
determining arithmetic operations, may be performed in the arithmetic operation
circuits associated with the respective load cell modules 16c-1 through 16c-n.

The single superordinate arithmetic unit 100 receives the digital
weight-representative signals from all of the load cell modules 16c-1 through 16c-n,
or all of the article weight measurements, performs the weight level determining
arithmetic operations for determining the weight levels in ali of the containers
or bottles on the filling platforms, and controls all of the valves 10-1 through
10-n in accordance with the results of the arithmetic operations. The superordinate
arithmetic unit 100 determines the position of each of the valves 10-1 through
10-n from the position-representative signal from the position-representative signal
generating unit 36. The arithmetic unit 100 causes each of the valves 10-1 through
10-n to be opened when it arrives at the position where the filling operation should
be started, and, thereafter, controls the degrees of the openings of the respective
valves 10-1 through 10-n in accordance with the weight levels of the article filling
the associated containers.

The display and control unit 30 is used, separate from the arithmetic
unit 100, for effecting various settings and operations including displaying the
various values of measurements, for setting the weight levels, setting the zero-point,
and operating the various parts of the filling machine. The display and control
unit 30 may be directly connected only to the arithmetic unit 100 by a communication
line, or may be connected, as shown in FIGURE 7, to all of the load cell modules
16c-1 through 16c-n as well as the arithmetic unit 100 by a common, single, serial
communication line.

FIGURE 8 shows a filling machine according to a fourth embodiment
of the invention. Generally, it is desirable to use as few superordinate arithmetic
units as possible. If, however, the number of the load cell modules is too large
for a single superordinate arithmetic unit to perform adequate arithmetic operations
and control. In such a case, a plurality, e.g. two, of such superordinate arithmetic
units 100a and 100b are used in the filling machine shown in FIGURE 8, to thereby
share the processing of the signals from the load cell modules 16c-1 through 16c-n
between them. As for the wiring for communications, one serial communication line
may be used for each load cell module and the superordinate arithmetic unit associated
with that load cell module, with a single serial communication line used for communications
between the display and control unit 30 and the two superordinate arithmetic units
100a and 100b, as shown in FIGURE 8. Alternatively, a single serial communication
line may be used for all of the load cell modules 16c-1 through 16c-n and the superordinate
arithmetic units 100a and 100b.

The present invention has been described by means of a weight-based,
multiple filler filling machine for filling containers with liquid, but it can
be used for filling containers with powdery materials and particulate materials.

The filling machine described above is a rotary type machine with
load modules and containers rotate, but the present invention can be applied to
linear type filling machines with containers linearly moving while being filled
with an article.

The load cell module useable in the filling machines according to
the first and second embodiments includes an arithmetic unit mounted on the surface
of or inside the load sensing element. As shown in FIGURE 9, each arithmetic unit
26d may be disposed inside a support 102 supporting a load cell 100. Since the
arithmetic unit 26d is not directly mounted on the load cell 100, the arithmetic
operating unit 26d does not interfere with the bending of the load cell 100 when
load is applied to it. The arithmetic unit 26d can be waterproofed.

Alternatively, the arithmetic unit 26d may be mounted on the outer
surface of the support 102. In this case, as in the second embodiment shown in
FIGURE 6, the support 102 as well as the load cell 100 are desirably placed within
a watertight enclosure.

In a still other alternative, the arithmetic unit 26d may be placed
inside a sealing enclosure, similar to the example shown in FIGURE 5. The sealing
enclosure with the arithmetic unit 26d therein is then mounted on the support
102.

Also, like the third and fourth embodiments, an A/D converter and
an arithmetic operation circuit with a communication circuit may be placed inside
or on the outer surface of the support 102.