Digital weight measuring and computing apparatus with automatic zero correction

Abstract

This invention relates to an improved digital weight measuring and computing apparatus with automatic zero correction which automatically corrects the zero indication of the apparatus. The weight measuring apparatus comprises a scale which employs a load cell and strain gauge to develop an analog voltage representing the weight on the scale. An analog-to-digital converter is then employed to change the analog voltage into a digital representation of the load on the scale. The digital converter provides a digital indication in addition to the digital indications employed to indicate the weight on the scale. The additional digital indication represents fractions of the least weight increment indicated of the weight on the scale. The additional digital indication is then employed to control the generation of a zero correction factor and this factor is then combined with the weight indications from the scale to indicate the correct weight of articles placed on the scale. The corrected weight in combination with the unit price of the article placed on the scale is then employed to compute the value of the article on the scale.

Claims

What we claim is:

1. Improved weight measuring apparatus having a weight output composed of a predetermined number of significant weight digits and including means for automatically correcting errors in the weight output for zero weight comprising, in combination, scale means for generating a digital weight signal having at least one more least significant digit than the weight output, memory means for storing a zero correction factor, means for periodically generating a corrected weight from the algebraic sum of the generated weight signal and the zero correction factor stored in said memory means, the most significant predetermined number of digits of the corrected weight comprising the weight output, and means responsive to the corrected weight falling within a predetermined small weight range for periodically modifying the zero correction factor stored in said memory means by a predetermined amount less than the least significant weight output digit to decrease the absolute value of the corrected zero weight toward zero.

2. Improved weight measuring apparatus, as set forth in claim 1, and including means for initially storing in said memory a zero correction factor falling within a second predetermined weight range greater than the predetermined small weight range.

3. Improved weight measuring apparatus, as set forth in claim 1, and including a zero indicator, and means responsive to the corrected weight for a zero weight being accurate to within a predetermined fraction of the least significant weight output digit for illuminating said zero indicator.

4. Improved weight measuring apparatus, as set forth in claim 1, wherein said scale means generates a weight signal in which said at least one more least significant digit is a predetermined fractional increment of said weight output.

5. Measuring apparatus comprising, in combination, scale means for generating a digital weight signal having a predetermined number of digits corresponding to the weight of an article, a portion of such digits being significant, memory means for storing a zero correction factor, means for periodically generating a corrected weight from the algebraic sum of the generated weight signal and the zero correction factor stored in said memory means, means responsive to the corrected weight falling within a predetermined small weight range for periodically modifying the zero correction factor stored in said memory means by a predetermined amount less than the least significant one of the significant weight digits to decrease the absolute value of the corrected zero weight toward zero, means for computing an article value from the significant weight digits and a preselected price per unit weight, and means for printing an article label including the significant digits of the weight signal and the computed value.

6. A method for automatically correcting for any zero errors in a digital weight measuring scale having a predetermined number of output digits, said method comprising the steps of: (1) generating a digital weight signal having at least one more least significant digit than the output digits, (2) algebraically combining a digital zero correction factor stored in a memory with the generated weight signal to obtain a corrected weight, the most significant predetermined number of digits in such corrected weight forming the weight output, and (3) periodically modifying the zero correction factor stored in said memory by a predetermined small amount when the corrected weight falls within a predetermined small weight range of zero for decreasing the absolute value of the corrected weight.

7. An improved weight measuring apparatus having a digital weight output composed of a predetermined number of significant weight output digits and in which the weight output is automatically corrected comprising, in combination, scale means for generating a digital weight signal having at least one more least significant digit than the weight output, memory means for storing a correction factor, means for periodically generating a corrected weight by combining the correction factor stored in said memory means with the generated weight signal, the most significant predetermined number of digits of the corrected weight comprising the weight output, and means responsive to the corrected weight falling within a predetermined small weight range near zero for periodically modifying the correction factor stored in the memory means by a predetermined amount less than the least significant weight output digit to decrease the absolute value of the corrected weight toward zero.

8. An improved weight measuring apparatus in accordance with claim 7 further including visual indicating means, and means controlled by the corrected weight for causing said indicating means to indicate a weight less than a predetermined fraction of the weight represented by the least significant digit of the weight output.

9. An improved weight measuring apparatus in accordance with claim 8 further including delay means for delaying the actuation of the visual indicating means to prevent blinking of the visual indicating means.

10. An improved weight measuring apparatus in accordance with claim 7 further including means for changing the predetermined limits of the corrected weight within which the correction factor stored in the memory means is periodically modified.

11. An improved weight measuring apparatus, as set forth in claim 7, and including maually controlled means for storing in said memory means a correction factor which exceeds the predetermined small weight range.

12. An improved weight measuring apparatus comprising, in combination, scale means for generating a digital weight signal comprising a first part having a plurality of significant digits and a second part representing a fraction of the weight represented by the least significant digit of the first part of the digital signal, computer means interconnected with said weight measuring apparatus for (1) storing a weight correction factor, (2) for combining the weight correction factor with the digital weight signal to obtain a, corrected weight signal, and (3) for periodically modifying the stored weight correction factor by a predetermined amount less than the least significant digit of the first part of the digital signal when the corrected weight signal falls within a predetermined weight range to decrease the absolute value of the corrected weight signal toward zero.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic block diagram of apparatus embodying the principles of the present invention for weighing, computing a value and printing a label showing the weight, price per unit weight and computed value of each of a plurality of successive articles;

FIG. 2, composed of FIGS. 2A - 2J, shows a flow diagram illustrating the operating sequence of apparatus embodying the principles of the present invention;

FIG. 3 is a schematic circuit diagram showing a switch arrangement for entering data into apparatus for weighing, computing a value and printing a label for each of a plurality of successive articles;

FIG. 3A shows a multiplex interface between the analog-to-digital converter and the microcomputer multiplex input interface.

FIG. 4. composed of FIGS. 4A - 4C, is a schematic logic diagram of circuitry for controlling the operating sequence and computing in apparatus for weighing, computing a value and printing a label for each of a plurality of successive articles;

FIG. 5 is a schematic logic diagram showing multiplexing circuitry for supplying data to a digital weight display;

FIG. 6 is a block diagram of a printer for use in apparatus according to the present invention; and

FIG. 7 is a diagram of one possible memory location arrangement in the random access memory in the microcomputer of the exemplary apparatus of the present invention.

DESCRIPTION OF AN EXEMPLARY EMBODIMENT

Referring now to the drawings and particularly to FIG. 1, a block diagram is shown of apparatus 10 embodying the principles of the present invention for weighing articles and, for each weighed article, for computing a value based upon a predetermined price per unit weight and subsequently printing an article label bearing the weight, the price per unit weight and the computed value for the article. The weight of each article is measured by a scale 11. Although the scale 11 may be of various known designs using mechanics, optics and electronics, it is preferably an electronic scale of the type shown in U.S. Pat. No. 3,709,309 which issued on Jan. 9, 1973 to Williams, Jr. et al, and the disclosure of such patent is incorporated herein.

The scale 11 generally comprises a load cell 12 which generates an analog output signal proportional to the weight of an article placed on a platter or weight platform 13. The analog output from the load cell 12 is applied to an analog-to-digital converter 14 which has a digital output corresponding to the gross weight of the article on the platter 13. The digital weight signal from the converter 14 is preferably in a binary coded decimal (BCD) format which is particularly desirable for use in computing an article value. The binary coded decimal (BCD) format comprises four binary digits for each decimal digit. The weights given to the binary digits are 1, 2, 4, 8 respectively of each decimal digit. The converter 14 may be arranged to provide any desired number of decimal digits, depending upon the maximum weight to be weighed on the scale and also upon an increment represented by the least significant decimal digit. In the exemplary embodiment described herein, as well as in the exemplary embodiment set forth in the above-identified patent to Williams, Jr. et al, it is assumed that five decimal digits will be supplied by the analog-to-digital converter 14.

The above-identified patent to Williams, Jr. et al discloses circuitry for automatically canceling unwanted direct current signals from the direct current or analog signals from the load cell, sensor, transducer, or strain gauge. In addition, this patent discloses circuitry for automatically zeroing the analog portion of the system while the transducer output is momentarily interrupted and for filling in gaps in the analog signal resulting from interruption of the transducer or strain gauge output.

As explained in the above-identified patent to Williams, Jr. et al, it is sometimes desirable to indicate a negative weight from the overall arrangement from the scale load cell through the converter and digital control arrangement. Thus, when the scale has been corrected to read net weight and the container and the commodity are both removed from the platter, the scale should indicate a negative weight equal to the weight of the container which is the tare weight for which the scale has been adjusted. The overall arrangement in accordance with the exemplary embodiment of this invention is accordingly arranged to indicate a negative weight under these circumstances and also under other conditions described herein.

However, the arrangement for making the various analog corrections described in the above-identified patent to Williams, Jr. et al do not prevent the zero output display of the scale from varying or wandering in a random manner.

In accordance with the present invention, the digital output or display from the scale mechanism is further corrected in the manner described herein to correct for the random wandering of the zero display of the scale.

A digital weight signal from the scale 11 is applied to a control unit 15 through interface circuits of FIG. 3A and circuits 24 of FIG. 4B. The control unit 15 includes an input/output buffer and memory 16 which receives the digital weight data through the interface circuits of FIG. 3A and the circuits 24 from the scale 11. Data input switches 17 are also connected to the input/output buffer and memory 16. The data input switches 17 include a manually operated keyboard for entering price data, a printer mode switch and tare weight switches. The input/output buffer and memory 16 functions as an interface with an arithmetic logic unit 18. A sequence controller 19 causes the arithmetic logic unit 18 and data memory or registers 20 to compute the value of each weighed article and to supply such value through the input/output buffer and memory 16 to a printer 21. The arithmetic logic unit 18, the sequence controller 19 and the data registers 20 are preferably included in an integrated circuit microcomputer, as will be discussed in greater detail below.

The value is computed from a price per unit weight which is obtained either from the data input switches 17 or from a commodity plate inserted into the printer 21. The commodity plate automatically supplies price information to the input/output buffer 16 in a manner similar to that described in U.S. Pat. No. 3,459,272 which issued to Susor on Aug. 5, 1969 and the disclosure of such Susor patent is incorporated herein. The commodity plate includes raised type for use in printing the commodity or article name on a label. Price per unit weight information is encoded on the commodity plate by means of the presence or absence of a plurality of holes or notches at predetermined locations. An optical or other suitable type of reader is provided in the printer 21 for sensing the presence or absence of the holes and for converting the price per unit weight information to a BCD output. The weight data from the scale is corrected for any tare weight and zero error by the arithmetic logic unit 18 and supplied to the printer 21 and to a digital weight display 22. After a steady state weight reading is received by the logic unit 5 from the scale 11, the sequence controller 19 causes the arithmetic logic unit 18 to compute an article value. The computed article value, the net article weight and the price per unit weight information are used by the printer 21. The sequence controller 19 controls data output to the printer and initiates printing a label.

After a label is printed, the sequence controller 19 will normally inhibit the arithmetic logic unit 18 until the label has been removed from the printer 21 and a motion -- no motion cycle has appeared on the scale 11 to indicate that the weighed article has been removed and a new article has been placed on the scale 11. The sequence controller 19 may also inhibit the arithmetic logic unit 18 in the event of the occurrence of various conditions. For example, if the price or tare information is changed, the apparatus 10 is adapted to go into a "lock" condition which prevents printing a label until a "lock" switch is manually actuated to extinguish an indicator light. Such an interlock prevents an accidental change in price or tare weight data, as when an operator accidentally bumps one of the switches 17. Another interlock may be provided to prevent printing a label if data print wheels are not properly set up to the correct value, weight and price data. Still another interlock may prevent printing an erroneous label in the vent that either the maximum weight capacity of the scale 11 or the maximum value capacity has been exceeded.

For convenience, the apparatus 10 is adapted for operating in several different modes. The different modes of operation affect the manner in which the printer 21 prints a label. The data input switches 17 include the mode switch which permits selecting either "single", "demand" or a "continuous" mode of operation for printing labels in which the value is computed for each weighed article. In the single mode of operation, the apparatus 10 must be manually actuated for each label which is printed. In the demand mode of operation, a new label is printed each time a printed label is removed from the printer 21 and the scale 11 has gone through a motion -- no motion cycle to cause the computation of a new value. In the continuous mode of operation, the printer 21 will continuously print labels all bearing the same weight, price per unit weight and value. The mode switch may further include "price by count" modes of single, demand and continuous. In the price by count mode of operation, the printer 21 will print labels bearing a count or factor of the number of pieces in each article or package to be labeled and the total price for this number of pieces. For example, a grocery store may package six oranges or six pears together. If the package price is, for example, six for $ 1.29, then the label will be printed bearing the legend "$1.29" in the place of the article value and "6/$1.29" in place of the price per unit weight.

In the following description of an exemplary embodiment of the present invention, it is assumed that the automatic zero correction feature will work within a range of -0.005 pounds and +0.005 pounds. However, these limits may obviously be changed to any desired value. In addition, an expanding range switch or button is provided which will allow the zero correcting feature to operate within other weight limits. For example, by operating this expanded range button under certain circumstances, the automatic zero correcting arrangement may be employed to correct in the range from 0.105 pounds and 0.095 pounds. However, the arrangement in accordance with this invention is also arranged to prevent any operation of the zero correcting arrangement if the indication of the weight on the scale platter or pan exceeds 0.6 pounds.

Also, a ready lamp or other indicator is provided which is turned on when indication from the scale is within +0.002 pounds and -0.002 pounds for a predetermined interval of time. The lamp is employed to indicate that the scale is properly corrected and in condition to weigh another object or commodity. The automatic zero wander correction does not stop at 0.002 pounds, but continues on to be corrected to zero accurately. The automatically operated correcting feature operates sufficiently rapid so that correction will all be completed prior to the time the ready lamp is turned on and thus prior to the time an object or commodity to be weighed is placed on the platter or pan to be weighed.

It is obvious that the predetermined limits of -0.002 and +0.002 pounds, +0.005 pounds, and -0.005 pounds and the limits of 0.6 and -0.6 pound have been arbitrarily set or selected and that any other set of suitable limits may be selected and provided by obvious minor changes in the control equipment in accordance with the present invention.

Turning now to FIGS. 2A through 2J, a flow diagram is shown for an exemplary operating sequence of the apparatus 10 for measuring the weight of an article, computing from the measured weight and a price per unit weight the article value and printing an article label. The flow diagram consists of a series of diamonds or rhombuses and rectangles. Each diamond corresponds to a question having either a yes or a no answer which may be obtained by conventional methods. Each rectangular block corresponds to the performance of a specific function such as causing a label to be printed. The numbers placed in the circles to the top and left of the blocks represent input locations. For example, an "A2" in a circle on the left of the flow diagram in FIG. 2A represents an input to the second block from the top in sheet A of FIG. 2. The numbers in the circles to the right of the blocks in the flow diagram represent an output connected to a different location in the flow diagram. For example, the first or uppermost block in FIG. 2A has an output to "E1" if the answer to the question is no. This indicates that if the answer is no, a jump is made to E1 or the input on the first block in sheet E of FIG. 2.

For convenience, FIG. 2 has been separated into parts A through J. Generally speaking, FIG. 2A covers the basic cycling of the apparatus 10. Upon the occurrence of a predetermined sequence of conditions, the cycle is completed with a print pulse. FIG. 2B represents the logic of a check of various interlocks for the occurrence of new data. FIG. 2C shows the sequence of operations for the operation of the lock switch which must be pressed when data is changed and the printer 21 is in the single mode of operation. FIG. 2D shows the logic for checking the setting of the service switch which affects the weight display 22 and for checking for the presence of excess weight and excess value. FIG. 2E generally represents the operation of the scale motion detector. FIG. 2F shows the operation of the "expand" switch which permits checking the accuracy of scale zero and of the auto-zero capture range expand switch. FIG. 2G shows the operation of the auto-zero correction circuitry, of the price by count logic and of the value computation. FIG. 2H shows the manner in which the condition of value, price and weight blanking switches are checked and the sequence in which the outputs are actually blanked. FIG. 2I shows the manner in which tare weight is manually entered into the apparatus 20. FIG. 2J shows the sequence of operations for transferring weight signals from the analog-to-digital converter to the microcomputer.

As previously indicated, FIG. 2A shows the basic operating cycle for the apparatus 20. Each new cycle is initiated at the input A1 to a block 25. If motion has been present on the output from the scale 11 during the previous cycle of the apparatus 10, a motion flage MOTNF will be set. The motion flag MOTNF may, for example, consist of a bit stored in a memory or the state of a flip-flop or a latch. If the motion flag MOTNF is not set, the logic jumps to the E1 input while if the motion flag MOTNF is set, the logic proceeds to a block 26. The logic must proceed through the block 26 before a label can be printed. Thus, the block 25 requires a motion -- no motion cycle before a label is printed. In the block 26, the presence of a motion flag MOTNF is again checked. If the flag is still present, the cycle again turns to the E1 input while if the flag is not present, the logic proceeds to a block 27 which checks to see if the printer is in a demand or continuous mode of operation as set on a mode switch. If the system is operating in a demand or continuous mode, the logic proceeds to the A5 input, while if it is in a single mode, logic proceeds to a block 28.

In the block 28, a check is made on whether or not the printer 21 is set up to inhibit the recognition of a motion -- no motion function if a previously printed label has not been removed from the printer. If the motion detector is not inhibited, logic proceeds at A7 while if it is inhibited by the printer 21, a check is made in block 29 on whether or not a new weight measurement is required by the printer 21 due to a lack of verification in the data supplied to the printer 21. If a reweigh is required, the logic proceeds to A7 while if reweigh is not required, a block 30 checks to see if print data is stored in the printer 21. If print data is stored, the system returns to A1 while if print data is not stored, it proceeds to a block 31 which checks on whether or not a "no tare weight" key has been pressed. If the no tare key is pressed, a block 32 checks to see if the price per pound data is equal to zero. If the price per pound data is equal to zero, a block 33 outputs a print pulse to the printer 21 for printing a label and the cycle returns to A1. This function is provided to allow printing of labels when the weight is equal to or less than 0.1 pound, for testing.

If either the no tare key is not pressed or the price per pound is not equal to zero, the system proceeds from the blocks 31 or 32, respectively, to a block 34 which checks to see if the net weight from the scale 11 is greater than 0.1 pound. If the net weight is not greater than 0.1 pound, it is assumed that an article is not present on the scale platform 13 and the system returns to A1. If an article greater than one-tenth pound is present on the scale platform 13, a block 35 checks to see if a SET latch or flip-flop is true. If it is true, it indicates that the system interlocks have been broken and the cycle returns to A1. If SET is not true, a block 36 checks to see if the net weight is positive. If the net weight should for any reason be negative, such as when a tare weight is entered into the apparatus 10 and a package or article has not yet been placed upon the scale platform 13, the cycle returns to A1. If the net weight is positive, a block 37 checks to see if an OVERCAP latch or flip-flop is true. OVERCAP is trued when the weight capacity of the apparatus 10 has been exceeded.

For the following description, it will be assumed that the apparatus 10 is capable of measuring up to thirty pounds so that OVERCAP will be true if a weight greater than thirty pounds is placed upon the scale platform 13. If OVERCAP is true, the cycle returns to A1. If the capacity of the apparatus 10 has not been exceeded, a print pulse is outputed at a block 38. From block 38 control is returned to A1. When the control is transferred to A15, block 39 causes the system to read all external inputs including the mode switch, the tare weight switches, the price switches, the price by count switches, the auto-price input from the printer and any weight input from the scale 11. After the external inputs are read, a block 40 checks to see if the apparatus 10 is in a price by count mode. If the apparatus 10 is in the price by count mode, logic proceeds to B3 while if it is not in this mode, a block 41 checks to see if auto-price data has been received from the printer 21. As previously indicated, the auto-price data may be supplied from a commodity plate inserted into the pritner 21. If auto-price data is received from the printer 21, a block 42 checks to see if the price data during the last cycle was also auto-price data received from the printer 21. If not, the system proceeds to B7. If the last price was also from the printer, a block 43 reads and stores the auto-price data inputs from the printer 21. The system then proceeds to B1.

Turning now to FIG. 2B, the logic is shown for checking the various interlocks. From the input B1, a block 46 checks for correct parity in auto-price data received from the printer. If there is a parity error, the logic proceeds to B7. If there is no parity error, the logic proceeds to a block 47. If the printer 21 is operating in a manual mode, the diagram in FIG. 2B is entered through B2, wherein a block 48 checks on whether or not the last cycle was also in the manual code. If not, the cycle proceeds to B7 while if the last cycle was also manual, the logic also proceeds to the block 47. The block 47 checks on whether or not the price data input has changed. If no change has occurred, a block 49 checks an interlock on the printer door. If the printer door is open, the logic proceeds to B7 while if it is not open, a check is made by a block 50 on whether or not an auto-zero inhibit signal is received from the printer 21. If no inhibit signal is received, the cycle proceeds to B8.

Returning to the block 47, if the price input data has changed, at a block 51 the new price input data is moved to a price per pound output memory, hereinafter referred to as "P/LB OUT". The cycle then proceeds to a block 52 and an interlock flag INTF is set. The cycle also proceeds to the block 52 from the block 50 if an auto-zero inhibit is received from the printer 21. The interlock flag INTF may consist of the setting of a latch or a flip-flop or a bit stored in a memory. The presence of the flag indicates the occurrence of an interlock such as a change in the printer mode, the opening of the printer door or the presence of an auto-zero inhibit signal. After the interlock flag INTF is set at the block 52, a block 53 checks to see if an "auto-tare" switch has been actuated. If not, the logic proceeds to I1 for manually entering a tare weight. If the auto-tare switch has been pushed, a block 54 updates the tare weight by storing the current measured weight in a TARE memory and again sets the interlock flag INTF. A block 55 then checks to see if the print mode has changed. If not, the logic proceeds to B12 while if the mode has changed, a block 56 updates a "print mode out" signal which indicates when the printer 21 is in the demand or continuous modes of operation. A check is then made at a block 57 to see if the interlock flag INTF is set. If not, the logic jumps to C1. If the interlock flag INTF was set, a block 58 clears the interlock flag INTF and sets an "initialize" flag INITF. Thus, the initialize flag INITF is set whenever either the price input or the tare weight input has changed. This flag inhibits the printing of a label until the lock switch is manually actuated, thus preventing an erroneous change in the tare weight or price data. After the initialize flag INITF is set, a block 59 trues the SET latch to indicate that the interlocks have been broken and the system proceeds to C5.

FIG. 2C shows logic relating to operation of the lock switch which must be actuated when the initialize flag INITF is set. A block 62 checks to see if the initialize flag INITF is set. If not, the cycle jumps to C5. If the initialize flag INITF is set, a block 63 checks to see if the mode switch is in the demand or continuous mode. If the mode switch is in either of these modes, the cycle again proceeds to C5. If not, a block 64 checks to see if the lock switch is pressed. If not, the cycle proceeds to C5. If the lock switch is pressed, a block 65 clears the initialize flag INITF and clears the SET latch and continues with the C5 input to the block 66. The block 66 returns the cycle back to either A1 or A2, whichever was the original point of origin for arriving to the block 66.

The logic of FIG. 2D checks the setting of a service switch which permits isolating price and value errors in printed labels between the printer and the logic unit which calculates and stores this data, and also checks for the presence of either an excessive weight or an excessive value which may cause an error in the output from the apparatus 10. From the D1 input, a block 69 checks to see if the service switch is in a display price position. If so, a block 70 transfers the price data stored in the P/LB OUT memory into a WEIGHT OUT memory wherein such price data is displayed on the digital weight display 22. If the service switch was not in the display price position, a block 71 checks to see if the service switch is in the "display value" position. If not, the cycle proceeds to D3 while if it is in this position, the cycle proceeds to a block 72 which causes a transfer of value data stored in a VALUE OUT memory into the WEIGHT OUT memory for displaying the value data on the digital weight display 22. Thus, if an error is present in the price shown on the printed label and the service switch is in the display price position, a comparison may readily be made between the price appearing on the digital weight display 22 and the price appearing on the printed label. If these prices are not identical, the serviceman will know that the error is due to a fault in the printer 21. If the data is identical, the serviceman will know that the error is caused in the logic unit 15. A similar check may be made on the computed value.

The D3 input as well as the outputs from the blocks 70 and 72 are applied to a block 73 which turns on a "ZERO" light. The ZERO light indicates that the scale is properly zeroed to within 1/4 of the least significant displayed weight digit. A block 74 then outputs all data and the printer mode to the printer 21. Subsequently, a block 75 checks to see if the maximum value capacity of the apparatus 10 has been exceeded. If the maximum value has been exceeded, a block 76 will set an OVERVALUE latch or flip-flop or store a bit in a memory location. After the block 76 has set OVERVALUE or if the logic jumped to D7 because the maximum value capacity was not exceeded, a block 77 checks to see if the maximum weight capacity of the scale 11 has ben exceeded. If not, the cycle will proceed to C5 where it is returned to either A1 or A2. If the maximum capacity of the scale 11 has been exceeded, a block 78 sets an OVERCAP latch or flip-flop or stores a bit in a memory location and an OVERCAPACITY flip-flop or latch is set to turn on an indicator light. The cycle then returns through C5 to either A1 or A2.

In FIG. 2E, a check is made to see if weight data read from the scale 11 is legitimate, i.e., no analog-to-digital conversion is occurring in the converter 14. If not, a check is made on whether or not a zero expand switch is actuated. If weight data is legitimate, a motion check is made to see if the data has been consistent or steady for a predetermined number of cycles.

Fron an input E1, a block 81 checks to see if a T1 flag is set. The T1 flag is set whenever new weight data has been read from the scale 11 into the logic unit 15. If the T1 flag is not set, a block 82 checks to see if a clock signal T1 is true. A true T1 represents a time interval during which weight data is not permitted to change. The apparatus 10 may, for example, operate on a 200 millisecond cycle. T1 may be set true for an arbitrary time interval, such as true for 60 milliseconds out of each cycle and false for the remaining 140 milliseconds. If T1 is true, logic proceeds from the block 82 to E4, while if T1 is not true a block 83 sets the T1 flag and the logic proceeds to E4. At E4, a block 84 checks to see if a "zero expand" switch is actuated. If not, the logic returns to A15 while if the zero expand switch is actuated, a block 85 sets the interlock flag INTF and the logic returns to A15. The zero expand switch is used for checking the accuracy and adjusting the weight zero when no weight is present on the platform 13 on the scale 11. When the zero expand switch is closed, the weight displayed on the digital weight display 22 is shifted by one decimal point. Thus, if the scale normally has a maximum reading of 30.00 pounds, the displayed weight willl be shifted over to display X.XXX pounds or to display as low as 1/1000th of a pound.

If the T1 flag T1F is set when the logic reaches the block 81, a block 86 checks to see if T1 is true. If not, the logic returns to E4. If T1 is true, a block 87 clears the T1 flag T1F. A block 88 then causes a weight reading to be entered into the logic unit 15. Next block 96 causes the tare timer state to be read out from the tare timer storage space in the RAM 187. Then in accordance with block 97, if the count in the tare timer is not zero it is reduced by one as indicated in block 97A and the control advances to block 89. If the count recorded in the tare timer is zero the control advances directly to block 89. After the weight is read out and the tare timer decremented if required as described above, block 89 checks the memory or switches for the setting of the motion detector count and the motion detector band. The motion detector count is the number of cycles or repetitions which must occur with no motion present before a label is printed. For example, switches may be set to establish that the apparatus 10 must cycle without motion at least twice or at least three times before a label can be printed. The band is the amount of permissible change during a no motion condition. For example, it may be determined that it is desirable to have a weight reading maintained within plus or minor 0.005 pound for a no motion condition. This results in a range of 0.01 pound for the motion detector regardless of the measured weight. This arrangement eliminates prior art problems in obtaining a uniform motion detector sensitivity for all weight measurements. The prior art optical motion detectors have not had a uniform sensitivity. After the count and band for the motion detector are determined, a block 90 calculates the weight minus the sum of the target plus the band. If this value is positive, then motion is present. The block 90 also determines the weight minus the sum of the target minus the band. If this sum is negative, then motion also is present. The target is taken to be the last weight reading.

If the block 90 calculates that there is motion, then a block 91 transfers the logic to F1 while if motion is not present, a block 92 checks to see if a motion flag MOTNF was set during the last cycle of the apparatus 10. If the motion flag is clear, logic proceeds to F2 while if a motion flag is present, a block 93 increments by one a no motion counter. After the counter is incremented, a block 94 compares the total count with the no motion count determined at the block 89. If the contents of the counter is not equal to or greater than this count, the cycle proceeds to F2 while if it is greater than or equal to the count, the motion flag MOTNF is cleared at a block 95. The cycle then proceeds to F2.

In FIG. 2F, sequence of operation is shown for the operating sequence of the zero expand, the auto-zero capture range expand and the weight overcapacity check. The F1 input, which is entered if motion is present, is applied to a block 98 which clears the no motion counter and sets the motion flag MOTNF. After the flag is set, a block 99 updates the target and subtracts a digital weight of eight pounts. The output from the load cell 12 and the scale 11 preferably is always positive since it is in a digital format. Some types of analog-to-digital converters 14, such as a dual slope integrating converter, are more accurate if used in a range wherein they only have a positive output. This may be accomplished by offsetting the output from the scale 11 to fall within the range of from 8 pounds to 8 pounds. Thus, the zero is arbitrarily offset by 8 pounds. The block 99 substracts an initial weight of eight pounts from the scale reading to zero the weight signal when no weight is present on the platter 13. After the eight pounds is subtracted, a block 100 checks to see whether or not the zero expand switch is actuated or true. If the zero expand switch is actuated, a block 101 moves the four least significant digits of the measured weight, or the digits X.XXX pounds, to the WEIGHT OUT memory and clears an AUTO-ZERO correction counter. Thus, the digital weight display 22 will now show the true zero weight setting of the apparatus 10 to within one-tenth of a normal weight graduation. Since the auto-zero operation is inhibited, a block 102 turns off the ZERO light and logic jumps to D1.

If the zero expand switch is not closed, sequence proceeds from the block 100 to a block 103 which checks to see if an "auto-zero capture range expand" switch is closed. If not, sequence proceeds to F6. Normally, the weight reading used by the apparatus 10 for calculating a value is automatically zeroed if the weight from the scale 11 is less than 0.005 pound when no weight is present on the platter 13. However, when the auto-zero capture range expand switch is closed, a block 104 will check to determine if the weight recorded in the raw weight register is equal to or less than 0.6 pound.

If the weight reading is greater than 0.6 pound, the sequence jumps to G7. If it is less, block 106 causes this weight, up to 0.6 pound, to be recorded in the zero error correction register in the AUTO-ZERO register. The weight is then corrected at a block 105 by the contents of the AUTO-ZEO register and moved to the WEIGHT OUT memory.

After the zero is corrected, a check is made to see if the maximum weight capacity of the apparatus 10 has been exceeded. Such maximum capacity has arbitrarily been set at thirty pounds which is generally sufficient for apparatus of the type described when used in the meat department of a grocery store. A block 107 checks to see if the corrected weight stored in the WEIGHT OUT memory is greater than thirty pounds. If it is, a block 108 sets an OVERCAP latch, flip-flop or similar memory device. If the maximum weight is not exceeded, a block 109 clears OVERCAP. Blocks 108 and 109 are connected to a block 110 which checks to see if a motion flag MOTNF is present. If not, logic proceeds to G1, while if motion is present, a block 111 checks to see if SET is true. If SET is true, the logic proceeds to G7 while if it is not true, a block 112 blanks the weight output and the logic proceeds to D1.

FIG. 2G shows the sequence of operations of the auto-zero correction circuitry, of the price by count mode of operation and of the value computation if a mode other than price by count is selected. A correction is automatically made to errors in the scale zero when the absolute value of the previously corrected weight is less than 0.005 pound. A total correction may be made up to 0.6 pound in 0.001 pound increments, or in other suitable incremental values. Furthermore, gross zero corrections of up to 0.6 pound may be made by use of the zero capture range expand switch which stores correction weight up to 0.6 pound to be stored in an AUTO-ZERO memory.

The G1 input to a block 115 is compared to see if the absolute value of the corrected weight is less than or equal to 0.005 pound. If not, a block 116 clears the 1/4 graduation flag to turn off a light which indicates that the weight reading used by the logic unit 15 for computing a value is within 1/4 of one graduation displayed on the digital weight display 22. From the block 116, sequence of operations proceeds to G7, skipping any changes in the contents or correction factors stored in an AUTO-ZERO correction register. If the absolute value of the corrected weight is less than or equal to 0.005 pound, a block 117 checks to see if the absolute weight is less than or equal to 0.002 pound. If not, a block 118 sets the zero count register to twelve which in turn causes the 1/4 graduation indicator light to be turned off later in the cycle.

If the absolute value of the corrected weight is less than 0.002 pound, a block 120 indicates that the count stored in the zero count register is read out of this register but also remains stored in this register. In accordance with block 119, the zero count register is checked to determine if zero is recorded in this register. If the count is zero, block 132 checks the weight to determine if the weight is zero. If it is, the sequence jumps to G7. If the weight is not zero, the sequence continues to block 121. If the count in the zero count register is not zero, block 114 indicates that the count in this register is incremented by one. Thereafter, the sequence goes to block 132 and then as described.

After the 1/4 graduation flag is set at the block 120, or after the logic has jumped to G5, a block 121 checks to see if the total auto-zero correction factor stored in the AUTO-ZERO register is less than or equal to 0.6 pound, the maximum permissible correction factor. If the correction factor is greater than or equal to 0.6 pound, no change is made in the correction factor and the logic jumps to G7, while if it is less than 0.6 pound the factor stored in AUTO-ZERO is modified by 0.001 pound at a block 122. After the sequence of operation jumps to G7 or after the AUTO-ZERO correction factor is modified in the block 122, a block 123 causes the tare weight stored in the TARE memory to be subtracted from the corrected scale weight and the four most significant weight digits are moved to WEIGHT OUT. Thus, it will be apparent that an auto-zero correction is made regardless of the fact that a tare weight may have previously been entered into the apparatus 10.

After weight data is stored in WEIGHT OUT, a block 124 checks to see if the printer mode switch has been set to a price by count mode of operation. If not, the logic jumps to G12 and subsequently a value is computed. If the apparatus 10 is in a price by count mode, a block 125 causes a factor or count received from the price by count switch to be stored in the P/LB OUT memory for supplying to the printer. A block 126 then checks to see if the price by count factor has changed from the last cycle of the apparatus 10. If a change has occurred, a jump is made to A15 and all external inputs are again read. If no change has occurred, a block 127 causes the price per pound data entered through the price switches to be stored in the VALUE OUT memory. If the logic has jumped to G12 and price per pound data from the switches or from the printer is present, the price per pound data is moved to the P/LB OUT memory location at a block 128 and zeros are forced into a register which stores the price by count factor read from the input switches 17. A block 129 then computes a value by multiplying the contents of the P/LB OUT memory by the contents of the WEIGHT OUT memory and stores the answer in the VALUE OUT memory. After the value is computed and stored, a block 130 compares this value with $100, the maximum value capacity of the apparatus 10. This maximum value has been arbitrarily selected to limit the number of print wheels required by the printer since measuring apparatus of this type, when used for labeling meat in the meat department of a grocery store, will normally not be required to exceed $100. If the computed value is greater than $100, a block 131 writes zero in the VALUE OUT memory and the logic jumps to C5 while if the maximum permissible value has not been exceeded, the logic jumps to H1.

In some instances, an operator of the apparatus 10 may wish to print a label which does not include all three of the price, the weight and the value. The weight, for example, is blanked whenever the apparatus 10 is operating in a price by count mode. Or, it may be desired to print a label bearing only the weight. This may be desirable in a wholesale operation where the wholesale purchaser will reprice the article for retail sales. Therefore, the apparatus 10 may be provided with manual switches which permit selectively blanking the price, the weight and the value from the printed label. The logic for performing these functions is shown in FIG. 2H.

The H1 input is connected to a block 134 which checks to see if a "blank price" switch has been actuated. If the switch is actuated, a block 135 causes blanks or numbers which index the print wheels to blank spaces to be stored in the P/LB OUT memory location. If the blank price switch is not actuated, or after blanks have been stored in the P/LB OUT memory, a block 136 checks to see if a "blank weight" switch has been actuated. If the blank weight switch is actuated, a block 137 checks to see if the apparatus 10 is operating in a price by count mode. If the apparatus 10 is in the price by count mode, the weight will already have been blanked and logic jumps to H6 while if it is not in the price by count mode, a block 138 blanks the WEIGHT OUT memory. If the blank weight switch was not actuated, a block 139 checks to see if the scale is in the price by count mode. If not, logic jumps to H6 while if it is in this mode, the block 138 will blank the WEIGHT OUT memory. If the WEIGHT OUT memory has been blanked or the logic has jumped to H6, a block 140 checks to see if a blank value switch has been actuated and, if not, the logic jumps to D1 while if it is actuated, a block 141 blanks the VALUE OUT memory and the logic then jumps to D1.

FIG. 21 shows the manner in which the tare weight is manually entered into the appartus 10. The tare weight, which is stored in the TARE memory, is subtracted from the measured gross article weight for obtaining a net weight used in computing a value. A check is made at a block 144 to see if a "no tare" switch has been pushed. If the no tare switch is pushed, the TARE memory is cleared and the interlock flag INTF is set at a block 145 and the logic then jumps to B10. If the no tare switch was not pushed when the logic was at the block 144, a block 146 checks to see if any other tare switch is pushed. If not, the logic jumps to B10 while if a tare switch is pushed, the block 147 checks to see if a three second time interval has elapsed since the last tare switch was pushed. If three seconds has elapsed, a block 148 clears the TARE memory and, subsequently, the tare weight represented by the pushed tare switch is stored in the TARE memory and the interlock flag INTF is set. If the three second interval has not elapsed when the block 147 is reached, the contents of the TARE memory are updated by adding the tare memory value represented by the pushed tare switch to the contents of the TARE memory. Thus, the tare weight stored in the TARE memory will then represent the accumulation of tare weights from two tare switches. For example, an operator may push a 0.10 pound tare switch and a 0.06 pound tare switch within a three second time interval and the total tare weight stored in the TARE memory will equal 0.16 pound. From the block 149, the logic returns to B10. The three second time interval was selected on the basis that most people can select and push two switches in this interval. Of course, a different time interval may be used.

FIG. 2J shows an exemplary sub-routine or sequence of operations for causing weight signals received from the analog-to-digital converter to be entered in the weight registers of the microprocessor employed in the present invention. The subroutine or sequence of operations shown in FIG. 2J are represented in FIG. 2E by block 88.

The remaining drawings show details of logic and circuitry of an exemplary embodiment of the invention for performing the functions described in the description of FIG. 2. Turning first to FIG. 3, the input switches 17 including the price, printer mode and tare weight data input switches are shown in detail. Price by count data, price per unit weight data, tare weight data and printer mode data is all supplied to the input/output buffer and memory 16 in the control unit 15 on nine switch buses 155 which represent the digits one through nine. It will be apparent that no bus is needed for a zero entry which corresponds to the absence of a signal on any of the nine buses 155. When the apparatus 10 is operated in a price by count mode, a count or factor is entered through a price by count switch 156. The price by count switch 156 is as a two-wafer or two-pole rotary switch having eleven contacts for selectively entering a count two through a count twelve. It will be appreciated that a count of one would not normally be used nor is there normally a need for a count greater than twelve, although this may be accomplished by providing additional contacts on the switch 156. A count of two through a count of nine is entered into the control unit 15 by setting the switch 156 and strobing or grounding a strobe line 157 while a count of ten, eleven or twelve is entered by setting the switch 156 and strobing a strobe line 158. When the price by count switch 156 is set to a factor between two and nine, a signal on the strobe line 157 causes an appropriate output on one of the nine switch buses 155. Similarly, when a signal is applied on the strobe line 158 and the price by count switch 156 is set to a factor of ten, eleven or twelve, a signal will also appear on one of the switch buses 155.

The apparatus 10 is designed for calculating an article value from price per unit weight data having three significant digits or from $0.01 up to $9.99 per pound. The price per unit weight data is manually entered through three switches 159-161. Each of the price switches 159-161 is a ten-contact rotary switch. A contact representing zero is not connected while contacts on the switches representing the digits one through nine are connected to corresponding ones of the nine switch buses 155. The switch 159 is used for entering pennies, or the least significant digit of the price data. A strobe line 162 is connected for providing a signal on the common terminal of the penny switch 159. When a signal is applied on the strobe line 162, the penny price data is entered on the connected one of the switch buses 155. The switch 160 is provided for entering dimes price data when a signal is received on a dime strobe line 163. Similarly, the switch 161 is connected for supplying dollar price data when a signal is received on a strobe line 164. Thus, price data up to a maximum of $9.99 per pound may be entered through the three price switches 159-161. Of course, it will be apparent that the number of price switches may be varied to meet other requirements for the apparatus 10 and the monetary units represented by the price switches may be changed to the local currency where the apparatus 10 is used.

Tare weight is entered into the logic unit 15 by means of nine momentary contact tare switches 165 which enter tare weight in 1/100th pound increments from 0.01 pound to 0.09 pound and a switch 166 which enters a tare weight of 0.10 pound. A signal is periodically applied by the logic unit 15 on a strobe line 167 which is connected to the 1/100th pound tare switches 165. These switches 156 are normally open push button switches. In the event that one of the switches 165 is pushed when a signal appears on the strobe line 167, an output appears on the associated one of the switch buses 155. A strobe line 168 is connected to the 0.10 pound tare switch 166, also a push button switch, and to an "auto-tare" switch 170 and a "no tare" switch 171. In the event that any of the switches 166, 170 or 171 is pushed when a signal appears on the strobe line 168, an output will appear on an associated one of the switch buses 155.

The apparatus 10 is designed for operation with a keyboard tare weight ranging from a minimum of 0.01 pound to a maximum of 0.19 pound. This is accomplished by providing a timing circuit in the control unit 15 which is initiated whenever one of the switches 165 or 166 is actuated. If one of the switches 165 is actuated and within the measured time interval the switch 166 is actuated, the tare weights for the two switches are summed. Similarly, if the switch 166 is actuated first and within the measured time interval one of the switches 165 is actuated, the total of the two tare weights is again summed. If two of the tare switches 165 are actuated within the time interval, only the most recent value is entered. The time interval may, for example, be about three seconds which should afford sufficient time for an operator to select and actuate two of the tare switches 165 and 166. If a greater time elapses, only the most recent are weight entered through a switch 165 or 166 is accepted by the control unit 15.

A printer mode switch 72 also supplies data over the switch buses 155. The printer mode switch 172 is a rotary switch having a common terminal connected to a printer mode strobe 173. When a signal is received over the strobe 173, an output appears on one of the switch buses 155, depending upon the setting of the mode switch 172. In an exemplary embodiment of the invention, the printer mode switch permits selecting between a single mode, a demand mode, a continuous mode, a price by count-single mode, a price by count-demand mode and a price by count-continuous mode. In either of the single modes of operation, a single label is printed each time an article is weighed. The printer is actuated each time the scale goes through a motion -- no motion cycle, a value is computed and the previous label has been removed from the printer 21. In the demand modes of operation, labels are printed as previous ones are removed from the printer. In the continuous modes of operation, the printer will continuously print a series of labels having the same weight, price per unit weight and value until printing is manually terminated.

The switches 17 also include an auto-price switch 174. When price information is to be supplied automatically from a commodity plate in the printer, the auto-price switch 174 is actuated. When the switch 174 is actuated, an indicator light 175 is automatically illuminated to annunciate this fact. A switch 176 is provided for expanding the capture range of the auto-zero circuitry from a normal range of zero plus or minus 0.005 pound to up to plus or minus 0.6 pound. The switch 176 is a momentary push button switch. A switch 177 is provided for inhibiting operation of the auto-zero circuitry to permit government inspectors to check the weighing accuracy of the apparatus 10. Finally, a switch 178 is provided for expanding the weight reading shown on the digital weight display 22. The zero expand switch is normally used by maintenance people in calibrating the zero weight setting for the apparatus 10. When the zero expand switch 178 is actuated, the displayed weight which is normally in a format of XX.XX pounds is shifted over by one digit to display a weight reading of X.XXX pounds. This permits calibrating the scale zero to within 1/1000th of a pound. Finally, the switches 17 include a lock switch 179. Any time there is a change in data entered into the apparatus 10 other than weight data, the apparatus 10 moves from what is normally referred to as a "ready" state to a "set" state to indicate that an interlock has been broken. The lock switch 179 must be manually actuated to return the apparatus 10 to the ready state. Thus, an operator cannot accidentally bump one of the tare switches 165 and 166 or the price switches 159-161 after the apparatus 10 is in the ready state and print erroneous labels.

FIG. 3A shows an exemplary arrangement for multiplexing the output decimal digits from the analog-to-digital converter of the above-identified Williams, Jr. et al patent so that this information may be transmitted over four data conductors or leads. In addition, four address leads are required to indicate which of the decimal digits is being transmitted over the four common leads at any particular instant or interval of time.

While both the above-identified patents to Williams, Jr. et al and the present exemplary embodiment of the invention show five decimal digits, it is obvious that any suitable number of decimal digits may be employed merely by increasing the number of counterstages and latches and related equipment.

The arrangement shown in FIG. 3A is controlled by a source of control signals 350. This source of signals may be an oscillator or any other source of control or clock signals which may be derived from the analog-to-digital converter such as a clock souce of this converter which would be divided down to a much slower pulse or clock rate.

The clock source 350 is further divided by five by equipment 351 which may be of any suitable form. The output of this dividing circuit as shown in FIG. 3A comprises a binary output having conductors 1, 2 and 4. This is the weight of the signals output from the frequency divider 351. These signals are then applied to the multiplexing interface circuits 352, 353, 354 and 355 which may be all the same. These devices are arranged to switch the five input leads shown to the one output lead under control of the input binary address signals on conductors 1, 2 and 4. Thus, when the conductors 1, 2 and 4 are all zero, the input leads A1, A2, A4 and A8 are connected respectively to WT. DATA 1, WT. DATA 2, WT. DATA 4 and WT. DATA 8 leads. When a one signal is applied to the conductor 1 input address conductor and the other two address conductors are zero, then the B1, B2, B4 and B8 input leads are connected to the respective WT. DATA 1 lead, the WT. DATA 2, WT. DATA 4 and WT. DATA 8 conductors, respectively. In a similar manner, the other input conductors of these switching devices are connected to the output weight data leads.

The input leads A1, A2, A4 and A8 are connected to the four binary output leads from the first decade of latches 136 of the above-identified patent to Williams, Jr. et al. Similarly, the input conductors B1, B2, B4 and B8 are connected to the four binary output leads from the second decade of latches 137. The remaining input leads are similarly connected to the corresponding binary leads of the other output decade latches 138, 139 and 140 of the above-identified Williams, Jr. et al patent.

In addition, the output binary coded signals over conductors 1, 2 and 4 from the divider 351 are connected to a translating circuit arrangement 356 such that when all zeros are applied on the leads 1, 2 and 4, zero will be also applied to the digit selector leads DIG SEL B, DIG SEL C, DIG SEL D AND DIG SEL E thus indicating that the first decade signals appear on the weight data leads WT. DATA 1, WT. DATA 2, WT. DATA 4 and WT. DATA 8. When a one is applied to the number one lead, a one or voltage signal will be applied to the DIG SEL B conductor thus indicating that the B decade signal will be transmitted over the WT. DATA 1, WT. DATA 2, WT. DATA 4 and WT. DATA 8 leads. Similarly, a one signal will be applied to the digit selector leads C, D and E when the corresponding weight data signal of these decimal digits is applied to the WT. DATA 1, WT. DATA 2, WT. DATA 4 and WT. DATA 8 leads.

The switching devices for switching any one of the input lines to an output line are similar to the eight-line to one-line decoders 198, 199, 200, 202 of FIG. 4B except that the last three of the eight-line conductors are not connected. Similarly, the dividing arrangements 351 and the translating arrangement 356 are well-known and commercially available.

Details of an exemplary control unit 15 are shown in FIG. 4. FIG. 4 consists of FIG. 4A, FIG. 4B and FIG. 4C which are arranged as shown on the first sheet of the drawings. In an exemplary embodiment of the invention the control unit 15 is a microcomputer 185 which functions to compute the value of each weighed article and to control the operating sequence of the apparatus 10. The microcomputer 185, for example, may be of a type commercially available in integrated circuits and in the exemplary embodiment described herein the microcomputer 185 comprises a Model MCS-4 Microcomputer Set manufactured by Intel Corporation of Santa Clara, California. Such a microcomputer employed in the exemplary embodiment of the present invention described herein includes a central processing unit (CPU) 186, a random access memory (RAM) 187 and five read only memories (ROM) 188-192. In the exemplary arrangement described herein, the CPU 186 is an Intel Type 4004 integrated circuit, the RAM 187 is an Intel Type 4002 integrated circuit and the ROM's 188-192 are Intel Type 4001 integrated circuits. However, it will be appreciated that other commercially available integrated circuit microcomputers or other types of commecially available computers will operate in accordance with the principles described herein.

These various CPU, RAM and ROM units are interconnected in the manner shown in FIG. 4A, as required in order for these units to cooperate one with another as required by the circuit configurations of these standard commercially available units. The connections are clearly described in the Users Manual for the MCS-4 Microcomputer Set published by the Intel Corporation. Briefly, the ROM's, RAM and the CPU are all interconnected in parallel by the data bus system shown at the top of FIG. 4A. These connections are in accordance with the requirements of the computer as commercially available. These connections permit the cooperation between the ROM's, the RAM and the CPU. Thus, the CPU will transmit an address over the bus system which defines a storage space in one of the ROM's, for example. The ROM having this storage space has internal control circuitry which will respond to this address and in turn cause information stored at the designated address to be transmitted back over the data bus system to the CPU which then responds to this information in the usual manner.

The storage portion of the RAM, the ROM's and the sequence controlling portion of the CPU comprise the sequence controller 19 shown in FIG. 1. Arithmetic unit 18 of FIG. 1 comprises the arithmetic unit of the CPU 186. The data registers 20 also comprise registers in the CPU 186 and the registers in the RAM 187. Input/output ports of the ROM's 188-192 and the RAM 187 and the related equipment comprise the input/output buffer and memory 16 shown in FIG. 1.

The read only memory units referred to herein as ROM's 188-192 store fixed data and also store a series of control orders or instructions for controlling the operating sequence of the entire apparatus 10. These orders or instructions, as is well understood by persons of ordinary skill in the programming and computer art, control the central processing unit CPU 186. These orders or instructions are readily obtained by persons of ordinary skill in the programming and computer art from the flow charts of FIG. 2A-2J by translation of the flow charts into computer language as required by the particular microcomputer and set forth in the instructions in the users manual for the respective computer. A program listing for performing the operations specified in FIGS. 2A-2J is attached as an Appendix to this specification. The program in the Appendix is in the language required for an Intel MCS-4 microcomputer, as specified in the Users Manual for the MCS-4 Microcomputer Set.

FIG. 7 shows the storage areas of the RAM 187 and the storage areas assigned to various registers for controlling input and output data, as well as process data, so that the data control apparatus will operate to automatically correct the zero indication of the scale in accordance with the present invention.

The control orders or instructions control the CPU 186 so that it will obtain the necessary fixed information from the ROM's 188-192 as well as the required control orders and instructions and obtain the data from the RAM 187 and from the various input devices so that the correct weight of each weighed article will be accurately determined and then, after various conditions are met, its value computed, and after the value is computed and various other conditions are met, causes a label to be printed.

The data used by the microcomputer 185 consists of data from the switch buses 155, data from the scale 11, data from the printer 21 and data from various interlocks. Four address outputs 193 from the RAM 187 are connected through inverters 194 to four address buses 195. Address information and other data supplied from the RAM 187 to the address buses 195 determines the data supplied to the microcomputer 185 and the data supplied from the control unit 15 to the printer 21. External data from, for example, the switches 17, is supplied to the microcomputer 185 on four input data buses 196 connected to the ROM 188. At the proper time interval, the external input data on the buses 196 passes though the ROM 188 onto four input/output data buses 197 connected in parallel with the five ROM's 188-192, the RAM 187 and the CPU 186.

The external input data is multiplexed onto the buses 196 by means of four 8-line to 1-line decoders 198-201 and a decimal-to-binary coded decimal (BCD) decoder 202. Each of the four 8-line to 1-line decoders 198-201 has a single output connected to a different one of the external input data buses 196 connected to the ROM 188. Three of the four data address buses 195 from the RAM 187 are connected in parallel to the four decoders 198-201 for selecting the inputs to the decoders 198-201 which are connected simultaneously to the buses 196. Thus, corresponding ones of the eight inputs to the four decoders 198-201 are connected to the outputs for such decoders. The zero inputs for each of the four decoders 198-201 are connected to the four outputs from the decimal-to-BCD decoder 202. When a zero address is supplied to the data select inputs of the decoders 198-201, the nine switch buses 155 are connected in series through the decimal-to-BCD decoder 202 and the line zero inputs of the decoders 198-201 to the four data input buses 196 to the microcomputer 185.

While the switch buses 155 are connected to the microcomputer 185, signals are sequentially applied to the price switch strobes 162-164, the price by count strobes 157 and 158, the tare switch strobes 167 and 168 and the mode switch strobe 173 for entering this data into the microcomputer 185. The strobe signals are applied on these strobe lines from a BCD-to-decimal decoder 203. The ROM 190 has four data outputs 204 connected respectively through four inverters 205 to four address buses 206. The address buses 206 supply address data to the BCD-to-decimal decoder 203 for sequentially scanning the ten outputs which strobe the price switch strobes 162-164, the price by count strobes 157 and 158, the tare switch strobes 167 and 168 and the mode switch strobe 173.

When price per unit weight data is received in an automatic mode from a commodity plate in the printer 21, such data is received over four lines 210-213. The four lines 210-213 are connected, respectively to the line one inputs on the 8-line to 1-line decoders 198-201. Four weight digit selection inputs 214-217 are connected to the line two inputs to the decoders 198-201, respectively, and four weight data lines 218-221 are connected, respectively, to the line three inputs to the decoders 198-201. The data appearing on the digit selection inputs 214-217 identifies which weight digit is present on the weight data lines 218-221. The weight data appearing on the lines 218-221 at any given time is a single digit of weight in a BCD format. If five digits of weight are to be received from the scale 11, the five weight digits are sequentially read by alternately receiving the digit selection data on the lines 214-217 and the actual weight data on the lines 218-221.

Interlock information is supplied to the microcomputer 185 by means of the line four through line six inputs of the decoders 198-200 and the line four through line seven inputs of the decoder 201. The line four input to the decoder 198 receives data from the lock switch 179, the line five input is connected to a motion detector inhibit output from a manual print switch in the printer 21 and the line six input receives a "reweigh" signal from the printer 21. The line four input to the decoder 199 is connected to a price contact on a service switch which, when actuated, connects the line four input to ground. When the service switch is actuated, price data is shifted into a weight memory for displaying on the digital weight display 22. The line five input to the decoder 199 is connected to a lead 222 which receives a signal when a label is printed, as will be discussed in greater detail below. The line six input to the decoder 199 is connected to receive an acknowledgment signal from the printer 21 when print data for a label has been stored.

The line four input to the decoder 200 is connected to a value contact on the service switch which, when actuated, causes the computed value to be shown on the digital weight display 22. The line five input to the decoder 200 is connected to the auto-price switch 174 and through an amplifier 223 to the auto-price indicator 175. When the apparatus 10 is operated with an auto-price received from the printer 21, a parity check is made to verify the accuracy of the price per unit weight data. An auto-price parity signal is applied on a line 224 to the line six input to the decoder 200. The printer door has an interlock switch connected to the line four input to the decoder 201. This interlock prevents accidental actuation of the printer while an operator has the door open for changing commodity plates or for maintenance. The line five input to the decoder 201 is connected to the zero expand switch 178 (FIG. 3), the line six input is connected to the T1 clock (not shown) which provides a pulse signal to indicate the time interval during which weight data may be read from the scale 11. As previously indicated in the discussion of FIG. 2, the T1 clock may have a 200 millisecond cycle, comprised of 60 milliseconds in which weight data may be read from the scale 11 and 140 milliseconds in which new weight data is measured by the scale 11. Finally, the line seven input to the decoder 201 is connected to the capture range expand switch 176.

The line seven inputs to the decoders 198-200 are connected to three switches 225-227, respectively. One of the switches 225-227 is closed to establish the number of sequential hits or no motion cycles of the motion detector before a label is printed. The switches may, for example, indicate that only a single hit is required if the switch 225 is closed, two hits are required if the switch 226 is closed and three hits are required if the switch 227 is closed.

Output data from the microcomputer 185 is stored within a random access memory 228. Weight, price and computed value data is supplied to the RAM 228 from the RAM 187 in the microcomputer 185 over the buses 195. Address information for storing data in the RAM 228 is supplied from the ROM 190 over the buses 206. The buses 206 are connected through an address selector 229 to address inputs on the RAM 228. The ROM 189 is connected over a line 230 to an input to the address selector 229 for connecting the buses 206 to the RAM 228 for supplying a data storage or write address or for connecting four buses 231 to the RAM 228 for supplying a readout address. Thus, when weight data or price data is supplied to the microcomputer 185, such data is also stored in the RAM 228 and when a value is computed, the computed value is also stored in the RAM 228. The ROM 189 enables writing or storing data in the RAM 228 by applying a "write enable" signal through an amplifier 236 to a line 237. The RAM 228 has four output buses 232 which are connected in parallel for supplying data to the printer 21 and to the weight display 22. The output buses 232 are also connected through three exclusive OR gates 233-235 for generating a parity bit from the output data. A BCD signal is generated in the printer corresponding to the setting of each print wheel in the printer 21. A parity bit generated from the setting of each print wheel is compared with the corresponding parity bit from the gates 233-235 for verifying the accuracy of the printer setup. If there is a lack of parity, a "reweigh" signal is applied from the printer 21 to the line six input to the decoder 198.

As indicated above, an output address is supplied to the RAM 228 over the readout address buses 231. The output address buses 231 are also connected to supply an output data address to the weight display 22 and are connected through amplifiers 240 to supply address data on outputs 240' to the printer 21. An address is applied on the buses 231 from a four bit address counter 241. A clock signal is applied from a clock source (not shown) to a decade counter 242. One output of the counter 242 is applied through an inverter 243 for counting up the four bit address counter 241. A different output from the decade counter 242 is applied through an inverter 244 for supplying a clock signal to the printer 21.

An output 245 from the ROM 189 in the microcomputer 185 is connected through an inverter 246 to a line 247 which enables quad bistable latches 248. When the latches 248 are enabled, data present on the address and data buses 195 from the RAM 187 is set into the latches 248. One output 249 from the latches 248 is a memory update request. The memory update request line 249 is connected to a NAND gate 250. The NAND gate 250 has an output connected through an AND gate 251 to an enable input on the four bit address counter 241. The address selection line 230 connected from the ROM 189 is also connected to a second input of the AND gate 251. A signal is applied on the line 230 to the AND gate 251 at the same time the readout address lines 231 are connected to the RAM 228. If the output of the NAND gate 250 is high at the same time, the AND gate 251 will enable the four bit address counter 241 for supplying a sequence of addresses for reading data from the RAM 228. The NAND gate 250 also has four inputs connected to the address buses from the counter 241 and a printer off input from the printer which is high when the printer is off. Thus, when the bistable latches 248 are set to apply a signal on the memory update request line 249 or when the printer is off, the address counter 241 will cycle whenever a signal is received from the ROM 189 on the line 230. Once a cycle is started by a signal on either the memory update request line 249 or by a pulse on the printer off line, the address counter 241 will continue to cycle until all of the address buses 231 go to a logic zero.

The bistable latches 248 also include an output 252 which indicates when the weight reading is below zero. The output 252 is connected through an inverter 253 to supply a MINUS sign output 254 to the printer 21 and also to the weight display 22. Still another output 255 from the latches 248 is connected through an inverter 256 to an output 257 which illuminates an out-of-range indicator lamp. A fourth output 258 from the latches 248 is connected through a NAND gate 259 to illuminate a ready indicator lamp. A "print stored" input 260 from the printer 21 is connected through an inverter 261 to a second input of the NAND gate 259. The output 258 from the latches 248 is also connected through an inverter 262 to a line 263 which supplies a SET signal to the printer. The line 263 is also connected through an amplifier 264 for illuminating a SET indicator lamp. The printer also supplies the REWEIGH signal on a line 265 through an amplifier 266 to illuminate a "weigh again" indicator lamp, a "take label" signal over a line 267 through an amplifier 268 to a take label indicator lamp and an "add label" signal over a line 269 through an amplifier 270 to an add label indicator lamp. An indicator is also provided for indicating when the apparatus 10 is operating within the weight zero limit or within a predetermined fraction of a weight graduation of zero. This has previously been referred to as the 1/4 graduation lamp. The ROM 192 has a single output 271 which is connected through an inverter 272 to a line 273. The line 273 is connected through an inverter 274 for energizing the zero limit indicator lamp.

During normal operation of the microcomputer 185, the ROM 189 applies a periodic pulse on the line 247 for resetting the bistable latches 248. This pulse will appear once each time the microcomputer 185 goes through a complete program cycle. In typical operation of the microcomputer 185, the pulse will appear at about 0.2 second intervals. However, it is possible for a noise pulse or some other disturbance to cause the central processing unit 186 to end up at an incorrect or nonexistent address. In such event, the microcomputer 185 becomes "hung up" and is in effect "dead". When this condition occurs, the microcomputer 185 must be restarted before a label can be printed. Timing circuitry is provided for automatically restarting the microcomputer 185 in the event that two sequential periodic pulses are missing from the line 247.

The line 247 is connected through an inverter 281, a current limiting resistor 282 and a diode 283 to the input of a threshold or level detector 284. The input of the threshold detector 284 is also connected through a high value resistor 285 to a voltage source and through a capacitor 286 to ground. When the voltage on the input of the threshold detector 284 is below a predetermined level, the threshold detector 284 will have a high or positive output. However, if the input of the threshold detector 284 exceeds a predetermined voltage level, the output of the threshold detector 284 will go negative. The capacitor 286 is charged at a relatively slow rate through the resistor 285. Each time a cycle pulse appears on the line 247, the inverter 281 will have a low output for rapidly discharging the capacitor 286 through the diode 283 and the current limiting resistor 282. Under normal operation of the microcomputer 185, the cycle pulses on the line 247 maintain the charge on the capacitor 286 below the threshold level of the detector 284. However, in the event of two sequential cycle pulses failing to appear on the line 247, the capacitor 286 will become sufficiently charged as to cause the threshold detector 284 to change states.

When the output of the threshold detector 284 goes negative, a capacitor 287 is dischared through a diode 288 and a current limiting resistor 289. A decrease in the voltage on the capacitor 287 causes a second threshold detector 290 to change from a negative to a positive output which is applied through a resistor 291 to the base of a transistor 292. The transistor 292 then switches states of conduction for applying a recycle or restart signal on a line 293 connected in parallel to the CPU 186, the RAM 187 and the ROM's 188-192 in the microcomputer 185. A resistor 294 and a diode 295 are connected in series between the input and the output of the threshold detector 284. After the threshold level is reached and the output of the detector 284 goes negative, the capacitor 286 is discharged through the resistor 294 and the diode 295 until the output from the detector again becomes positive. The capacitor 287 is then charged through a resistor 296, thereby causing the outputs of the detector 290 and the transistor 292 to change. If two more cycle pulses are absent from the line 247, the capacitor 286 will again become charged sufficiently for the outputs of the detectors 284 and 290 and the transistor 292 to change states, applying another clear and restart signal to the microcomputer 185.

FIG. 7 is a diagram showing one of many possible memory location arrangements in the random access memory RAM 187 of the exemplary arrangement of the apparatus of the present invention.

As indicated in FIG. 7, the RAM is provided with four memory register areas. Each of these memory register areas is arranged to store four binary digit words which in the exemplary embodiment described herein usually are coded to represent a decimal digit. Each memory area 16 is arranged to store 16 of these four binary digit numbers or other information. The register areas are selected by address designated "4 HIGH" above each one of the areas. Thus, the address of the first area is 0000. The address for the other three memory register areas is shown above each of these register areas. In addition, each of the four binary digits forming a word or number in each of the areas is assigned an address, which address is shown to the left of the first memory area. As indicated, the address for the memory areas comprise the first four high numbers, or binary digits of the address, while the address of the individual words or numbers within each area is designated by an address designated "4 LOW". Thus, at least certain of the same address symbols may be employed to designate both memory areas and also the words within the memory areas. Thus, two addresses are distinguished by their location in the addresses as is well known by persons of ordinary skill in the art of microcomputer operation.

The RAM 187 also includes four status registers shown in the lower part of FIG. 7. Each of the status registers has an address similar to the corresponding memory register area as indicated above each of the status registers. The rectangles in the status registers represent a storage space for a single binary digit or bit. Thus, each of the status registers may store four four-bit binary words. In addition, each of the bits of each of the words may be employed to store a binary bit which is independent of the other binary bits of a particular word at the particular address. In other words, as indicated, the zero correction sign bit, 714, is used to store the signal of the zero correction. This bit is stored in this bit space independently of the information stored in the other three bits spaces of the last word of the first status register.

As indicated in FIG. 7, the first five word spaces 710 in the first register space are employed to store the five binary coded decimal digits of the raw weight received from the weighing apparatus and are designated raw register. The next five four-bit register word spaces 735 are employed to store the five decimal digits of the motion target weight. The next word, or register space 726, is employed to store the count or number of hits employed to determine whether or not there is motion upon the platter of the scale. The last five word spaces 711 are employed as a zero correction register and store the five BCD digits employed to correct the zero indication of the scale. The other register spaces are designated to indicate the manner in which the particular register spaces are employed.

The various status register spaces shown at the bottom of FIG. 7 are similarly designated with the name of the bit or bits employed to record the various information required to provide the various features of the present invention, as described herein.

OPERATION OF THE SYSTEM

As described in the above-identified patent to Williams, Jr. et al, the sensor controlled by the load cell 12 of the scale provides an output voltage which represents the load on the load cell, which in turn is a function of the load on the scale. This output voltage is then amplified and processed so as to remove an unwanted direct current component and to reduce or remove unwanted variations in this voltage so that the voltage accurately represents the load on the load cell and the load on the scale. This analog voltage is then employed to control an analog/digital converter 14.

Also the analog-to-digital converter employed in the exemplary system described herein, in addition to generating a digital signal representing the weight on the scale to be displayed also generates a digital signal representing a fraction of the weight represented by the least significant digit of the displayed weight. In the exemplary embodiment described herein, it is assumed that this additional signal represents tenths of the weight represented by the least significant digit displayed. However, this additional digital signal may represent any other suitable or desirable fraction of the weight represented by the least significant digit displayed.

The analog-to-digital converter in effect samples this corrected analog voltage at repeatedly recurring instants of time. These sample voltages are then employed to control the output of the digital converter. Thus the output of the digital converter is a digital signal which accurately represents the load on the load cell and thus the load on the scale. In the exemplary arrangement described in the above-identified patent to Williams, Jr. et al, and in the exemplary arrangement described herein, the analog-to-digital converter requires a cycle of about 200 milliseconds. This cycle is thus repeated approximately five times a second. Near the end of each cycle of operation of the analog-to-digital converting apparatus, this apparatus transfers digital signals representing the analog weight input to a set of latches. These latches then maintain these digital signals for a predetermined interval of time. During a portion of this interval of time, the analog-to-digital converter also applies an output signal to the T1 lead. This T1 signal is obtained from device 100 of FIG. 2 of the above-identified Williams, Jr. et al patent. Thus during the time the output signal is applied to the T1 conductor, output from the latches remains constant so that the input to the multiplex switching device 352, 353, 354 and 355 of FIG. 3A remains constant. Consequently, the output signals applied in sequence to the weight data conductors WT. DATA 1, WT. DATA 2, WT. DATA 4 and WT. DATA 8 represent the respective decimal digits of the output weight from the analog-to-digital converter. The frequency or speed of the clock pulses from source 350 are such that the equipment of FIG. 3A will operate through a plurality of cycles during the time an output signal is applied to the T1 conductor. In other words, each of the binary representations of the five decimal digits representing the weight will be applied in succession to the output weight data conductors WT. DATA 1, WT. DATA 2, WT. DATA 4 and WT. DATA 8 a plurality of times during the time an output signal is applied to the T1 conductor.

Concurrently with the application of the respective decimal digits to the weight data lines WT. DATA 1, WT. DATA 2, WT. DATA 4, and WT. DATA 8, corresponding signals are applied either to none or to one of the data selecting conductors B, C, D and E indicating the specific of the five decimal digit signals applied to the weight data conductors WT. DATA 1, WT. DATA 2, WT. DATA 4 and WT. DATA 8 at each of the instants of time.

Assume now that power has been applied to the system so that the analog-to-digital converter will start to develop an analog conversion and provide a digital output signal after an interval of time. At the same time, the control equipment 15 will perform an initializing operation in the usual manner and cause zeros to be stored in each and every register space in the RAM 187. Thus, zeros will be stored in all of the register spaces of this RAM 187 shown in FIG. 7.

After zeros are stored in all of the register spaces, the control of the system is transferred to the main program A1 shown in FIG. 2A. Block 25 causes the memory space 718 of FIG. 7 to be tested to determine whether a one or a zero is recorded in this space. This space, which is designated "MOTNF", represents a motion flag and under the assumed conditions, a zero will be recorded in this space with the result that the program is now transferred to E1 of FIG. 2E. As a result, the system then tests the storage space T1F 717 to determine whether a one or zero is stored in this space. Under the assumed conditions, a zero will be stored in this storage spot. This storage space is designated a T1 flag. Since a zero is stored in this storage space, the program proceeds from box 81 to box 82 where the condition of the T1 clock lead (see the bottom of FIG. 4B) is tested to determine whether one signal or a zero signal is present on this lead.

Under the assumed conditions, a zero will be present on this lead so that the program proceeds to block 83. With the zero present on this lead, the analog-to-digital converter indicates that a satisfactory output signal is not present on the output leads from the analog-to-digital converter.

In block 83 the computer is directed to change the information stored in the storage bit space 717 to a one, indicating that the analog-to-digital converter is in the process of providing a valid output digital signal, but that such signal is not present at this time on the output leads from the analog-to-digital converter.

From block 83 the control then goes to block 84 where the zero expand switch is tested. A flag or bit stored in a status register bit space 719 indicates the setting of the zero expand switch. Since this switch is assumed to be off, the zero expand is not true and the control is thus transferrred to A15 of FIG. 2A at which time the control equipment successively reads all of the various external inputs and stores their condition in the third one of the storage registers having an address 0010 of FIG. 7. After all of these switches have been tested and their condition recorded in the corresponding register spaces assigned to them, the control then proceeds to block 40 where the price by count mode switch is tested. Under assumed conditions, this switch will be conditioned so that price by count is not the mode of operation. Consequently, control is then transferred to block 41.

Again, assuming this system is not operating in the auto price mode, a zero will be stored in the auto price bit storage location in the status register 0001 and, as a result, the control of the program is transferred to block 48 of FIG. 2B via transfer B2. As a result, block 48 tests the status register storage space 738 designated auto price in the status register 0001. Assuming that the last cycle was not auto price and that a zero bit is stored in this register indicating that the last cycle was also a manual price entry, the control is transferred from block 48 to block 47. Block 47 determines whether or not the price input has changed.

Thus, block 47 compares the information just recorded in the storage spaces SW-.OP through SW-.PO with the previous price output registered in the output price register 739 in the register 0011. Under the assumed conditions, the prices may or may not be the same. If the manual switches are all set on zero, then these prices will be the same and the price will not have changed so far as the control equipment is concerned. As a result, the control goes to block 49 of FIG. 2B where a safety switch on the printer door is checked, since this will be closed, control then goes to block 50 where the condition of an auto zero inhibit switch is tested. If this switch is operated, it is desirable to disable to the auto zero correction operations as described herein.

Assuming that the auto zero inhibiting switch is off, then the control is transferred to block 53 over the transfer B8. Block 53 tests the auto tare switch.

Under the assumed conditions, the information stored in the storage space SW-.T in the storage register 0010 indicates that the AUTO-TARE switch is not pushed with the result that the program now transfers to block 144 via transfer I1 of FIG. 2I.

Block 144 causes the computer to test the storage areas SW-.T and SW-.OT and if none of the tare switches has been operated, the program is transferred to block 145 which causes the tare register area 720 of FIG. 7 to be all returned to zero, of zeros again recorded in area under the assumed conditions. In addition, a one is recorded in the INTF status register space 736 of the status register 0010. Then the control is transferred to block 55 via B10 of FIG. 2B.

Block 55 causes the computer to compare the printer mode recorded in space 722 of register 0010 with the recordings in the status register bits 723 and 724. Since it is assumed herein that the printer mode will be the manual mode, these registers will not have changed so the operation will be transferred to block 57 via transfer B12. In block 57, the condition of the interlocked flag in the status bit register 736 is interrogated and as pointed out above, this flag has been set to a one so that control then goes to block 58 where the interlock flag is changed from one to zero and the initialized flag INITF stored in space 727 is set to a one. Next, the set bit in space 737 of the status register 0011 is set and the control transferred to block 66 via C5. Since this cycle of operations started from block 25, block 66 causes the computer to transfer control back to block 25 via transfer A1.

Since the motion flag is not set, that is, since a zero is still stored in the status register storage space 718, the control will be transferred to block 81 via transfer E1. At this time, the block 81 causes the computer to again test the T1F flag, that is, the T1F storage space 717. During the previous cycle, as described above, a one was stored in this block indicating that T1F is set. Consequently, at this time the control will be transferred to block 86, instead of 82, as described in the previous cycle. However, under the assumed conditions, the signal on the T1 clock lead is still zero so that control will be now transferred to block 84 via the transfer E4. Thereafter, block 84 tests the zero expand switch which, as assumed to be unoperated, so that control will then be transferred to block 39 via transfer A15. Thereafter the remainder of the second cycle is substantially the same as described above assuming that none of the various keys or switches have been actuated.

At the end of the second cycle, control will be transferred to the block 25 via A1 transfer in the manner described above and each succeeding cycle of the main program will be repeated in accordance with the above description of the second cycle so long as the signal from the analog-to-digital converter on the T1 clock lead remains zero. During each of these cycles, the T1F or T1 flag signal is set in a one state by a one being stored in the storage space 717 and a zero signal remains on the T1 clock lead from the analog-to-digital converter.

NO MOTION DETECTION

Assume now that during some one of the above-described cycles of the main program, the analog-to-digital converter 14 completes a conversion and supplies digital output signals to the control equipment 15. Consequently, the analog-to-digital converter 14 will also apply a one signal to the T1 clock lead. As a result, when a control is transferred to the block 86 during the next main program cycle, as described above, the T1 lead will test one or true so that the control is now transferred to block 87 where the T1F or T1 flag is cleared. In other words, a zero is now stored in the T1F status register space 717 instead of a one.

Thereafter, the control goes to block 88 where the weight output signals from the analog converter 14 are read into the raw weight register 710 of FIG. 1.

FIG. 2J shows a flow chart of an exemplary sub-routine for entering weight signals from the analog-to-digital converter in the raw weight register 710 of FIG. 7.

As described above when the analog-to-digital converter is in condition for transferring digital signals to the raw weight register 710, an output signal is applied by the analog-to-digital converter to the T1 lead. As a result the control sequence proceeds from block 86 through block 87 to block 88 as described above. Within block 88 the control is transferred to a block 370. This block indicates that the RO register within the CPU unit is set to eleven. The P1 register is set to zero and the R3 register also set to zero. These registers being located within the CPU unit.

The RO register is employed to indicate when all of the digits of the weight are transferred to the raw weight register 710. The P1 register is employed to direct the digits to the proper register spaces within the raw weight register. The R3 register is employed to select the desired decimal digit to be transferred from the analog-to-digital converter to the raw weight register 710. Thus initially with these registers set as described above, the block 371 to which the control is transferred from block 370 first reads the R3 register and finds a zero recorded therein. Block 370 then translates this zero to indicate that the zero should be read from each one of the leads DIG SEL B, DIG SEL C, DIG SEL D and DIG SEL E. The translation order or sequence is designated KBP which translates the binary digit in the R3 register into a one-out-of-four code. Next block 372 tests the digit select leads B, C, D and E and if any signal condition on these leads other than zero on all of them is found, the control is transferred back to transfer point J2 and the above cycle repeated. The small loop including blocks 371 and 372 are then repeated until zero is found on all four of the digit select leads DIG SEL B, DIG SEL C, DIG SEL D and DIG SEL E.

When a zero is found on all of these leads, the control is then transferred to the block 373 where the value of this digit is read into the raw weight register area selected by the P1 register in the CPU unit. This will be the first digit space in the raw weight register 710 since a zero was previously set in the P1 register in the CPU unit.

After the first digit is thus transferred to the raw weight register 710, the zero in the R3 register in the CPU unit is again translated by the KBP operations into a one-out-of-four code and the signals on the digit select conductors DIG SEL B, DIG SEL C, DIG SEL D and DIG SEL E compared with this code. If under the assumed conditions zeros are found on all of these leads, thus corresponding to the zero in register R3 after being translated to the one-out-of-four code, the control is advanced to block 376.

If on the other hand the signals on the weight data conductors WT. DATA 1, WT. DATA 2, WT. DATA 4 and WT. DATA 8 change during the time the weight is being read into the raw weight register 710, then when the signals on the digit select leads DIG SEL B, DIG SEL C, DIG SEL D and DIG SEL E are again compared with the setting of the R3 register by block 375 these signals will have changed so that the control is then transferred back to transfer point J1 and thus to block 370 instead of block 376. Under these assumed conditions wherein the signals on the B, C, D and E digit select conductors change during the reading in of a digit, it is assumed that the digit transferred to the raw weight register will be in error. Consequently, the transfer operations are started over again and proceed as described above. Namely, the control will be first transferred to block 370 where the registers RO, P1 and R3 are set as described above. Then the control is advanced to block 371 and the setting of the R3 register translated to a one-out-of-four code and again compared with the B, C, D and E select conductors. This cycle is then repeated until the signals on these conductors are the same as the translated setting of the R3 register at which time the digit on the weight data conductors 1, 2, 4 and 8 is again transferred to the raw weight register.

Assume that the signals on these digit select conductors do not change so that when they are again tested by the block 375 they will again correspond to the translated setting of the R3 register. Consequently, the control is advanced to block 376 where the setting of these three registers P1, R3 and RO are all incremented by one. The control then advances to block 377 where the number stored in the RO register is tested. Under the assumed conditions this number is now twelve (eleven plus one). Since this number is not zero, the control is transferred to the J2 transfer point and thus to block 371 where the above cycles of operation are repeated until the second of the decimal digits of the raw weight are transferred from the analog-to-digital converter to the raw weight register. The control is then again transferred to block 376. Assuming that the signals on the B, C, D and E digit select conductors have been changed during the recording of the value of the second digit in accordance with block 376, the registers P1, R3 and RO are again incremented so that the next or number three digit will be transmitted from the interface equipment of FIG. 3A to the raw weight register 710.

After the fifth digit has been thus selected and transferred to the raw weight register 710, the RO register in the CPU unit will have recorded in it fifteen. This is incremented by the sub-routine of block 376. As a result, the setting of this RO register will now be zero. Since this is a four digit binary register when the register is incremented with fifteen stored in it, it is restored to zero. With the result that the control is now transferred from reading the weight to the block 89 of FIG. 2E.

From block 88 the control sequence advances through blocks 96 and 97 and at times through block 97A which blocks are employed to advance the tare timer in a well-known manner. Briefly, the block 96 causes the count stored in the tare timer space 728 to be read out. Then block 97 determines if the count read out is zero. If it is zero the control sequence advances to block 89. If the count read out is not zero then the control sequence advances to block 97A where one is subtracted from the count read out and the new count restored in the tare timer storage spaces 728.

Next, in block 90, the computer determines whether or not there has been motion or a change in weight on the scale platform. The arrangement, in accordance with the exemplary embodiment of this invention, is arranged so that if the weight changes or is less than a predetermined amount, that is within a predetermined band, it is assumed that there is no motion of the platform scale. If, on the other hand, a weight change exceeds the predetermined band, then it is assumed that there is motion of a scale platform. In accordance with the present invention, the bandwidth may be predetermined to any desired value, which value is stored in the pre-assigned storage spaces in one of the ROM's.

Under the assumed conditions, zeros will be stored in the motion target weight register 735. Assume now that the band is 0.003 pound, plus or minus, and that a weight of 0.004 pound is read into the raw weight register 710 from the analog-to-digital converter 14 during this cycle of the control equipment 15. Then in accordance with block 90, the zero weight from the motion target weight register 735 is subtracted from the weight in the raw weight register 710 and then the band 0.003 subtracted from the difference which leaves +0.001. Since the final difference is positive, it indicates that motion is present. Next, the weight in the target weight register is subtracted from the weight in the raw weight register and the band of 0.003 pound added to the difference, and since this sum is not negative, it does not indicate motion. However, motion was indicated by the first calculation. Thereafter, in response to the motion indication from block 90, block 91 causes the program to transfer to block 98 via the F1 transfer. Block 98 first clears the no motion or hit counter 726 (i.e. causes all zeros to be stored in this counter space) and sets the motion flag MOTNF. That is, it causes a one to be stored in the status register 718. Then block 99 causes the weight in the motion target weight register 735 to be changed to the 0.004 pound recorded in the raw weight register. The weight of 0.004 pound remains recorded in the raw weight register 710 at this time. Next, an initial weight of 8 pounds representing the weight of the scale platter on the load cell is subtracted.

The program then proceeds through blocks 100 where zero expand switch is tested and found not to be operated. At block 103 the zero range expand switch is tested and found not to be operated, so the program then jumps to block 105 via transfer F6 where the weight recorded in the raw weight register 710 is corrected by the auto zero weight stored in the zero correction register 11, which is zero at this time under the assumed condition. The program then proceeds to block 107 where the control equipment determines that the weight on the platter is not greater than 30 pounds. Consequently, the control then advances to block 109 where a zero is recorded in the over capacity status register space 725.

Next, the control equipment 15 is advanced to the block 110 where the motion flag is tested. That is, the status register space 718 is read out and since a one is recorded in this space, the program then advances to block 111 where the one recorded in the set status register space 737 is tested. Since a one is assumed to be stored in this space, the program is then transferred to G7 so that the block 123 subtracts the tare weight in the register 720 from the weight in the raw weight register 710, causes round off operation to be performed and restores the thus corrected weight in the raw weight register and moves the four most significant digits from the raw weight register 710 to the weight out register 712.

Next, the program advances through blocks 124 transfer G12, blocks 128 and 129, 130 and 131 and transfer C5 to block 66.

Since this cycle of operation of the control equipment 15 started from block 25, as described above, the block 66 now causes the program to be transferred back over transfer A1 to block 25 and another cycle of the control equipment 15 is initiated.

At this time the motion flag MOTNF is set, i.e. the status register space 718 has a one stored in it as described above, so the control sequence advances to block 26 instead of being transferred to block 81 via transfer E1.

In block 26, the status of the MOTNF status register space 718 is again tested and since a 1 is recorded in this space, this flag is not clear; consequently, the program is transferred by transfer E1 to block 81.

At this time the T1F flag is not set. In other words, a zero was recorded in the status register space 717 on the last cycle of the control equipment 15. Consequently the program now proceeds to block 82 where the condition of the T1 clock lead from the analog-to-digital converter is tested. At this time it is assumed that a 1 signal is on this lead with the result that the program now transfers over the E4 transfer to block 84. Since the zero expand switch is not operated, this zero expand is not true with the result that the program is transferred over transfer A15 to block 39. The remainder of this cycle of the control equipment 15 then is as described above through blocks 40, 41 transfer B2 to blocks 48, 47, 49, 50, 52, 53 transfer I1 to block 144, then block 146 and transfer B10 to block 55. From block 55, control is transferred to block 57 over transfer B12 and then through block 57 transfer C1 to block 62, transfer C5 to block 66.

Since this cycle was initiated through block 26, block 66 now returns the control to block 26 over transfer A2 and the above cycle of operations then repeated. This cycle of operation is then continuously repeated so long as a 1 signal is applied to the T1 clock lead by the analog-to-digital converter.

Next, when the analog-to-digital converter removes the 1 signal from the T1 clock lead, the program during the next cycle of the control equipment 15 is subsequently transfered to blocks 81 and 82 in the manner described above, the T1 clock lead will not be true, in other words it will test 0 so the program is then advanced to block 83 where the T1 flag is again set; that is, the 1 is recorded in the T1F storage space 717 in the status registers. Thereafter, the above cycles of operation of the control equipment will continue as described above initially except that the block 66 of FIG. 2C will return the control to block 26 over the A2 transfer since the cycle started from this block instead of block 25 as described initially.

By the use of the T1F flag and the testing of the T1 clock control lead as described above, the output from the analog-to-digital converter is transferred to the control equipment only once during each of the 200 millisecond cycles of the analog-to-digital converter 14. This transfer occurs in the next succeeding cycle of the control equipment after 1 signal is applied to the T1 clock control lead. Thereafter the 1 signal has to be removed from this lead and re-applied by the analog-to-digital converter 14 before a subsequent weight will be transferred from the analog-to-digital converter to the control equipment 15.

Since the control equipment 15 operates in the abovedescribed manner for the various cycles, when weight information is not transferred from the analog-to-digital converter, the operation of the control equipment 15 during such cycles will not be repeated. Instead the operation during only those cycles during which weight information is transmitted from the analog-to-digital converter to the control equipment 15 will be described. However, it is to be understood that control equipment 15 advances through numerous of these cycles in the manner described above between each of these cycles during which weight information is read out from the analog-to-digital converter and stored in the raw weight register 710.

Assume now on the next weighing cycle, that is the next time a weight is received from the analog-to-digital converter and stored in the raw weight register 710, the weight stored in this register will not have changed but will be .004 pound. Consequently when the calculations at block 90 have performed in the sequence as described above, the calculations will indicate that no motion is present. No motion would also be present if the weight received from the analog-to-digital converter does not vary more than 0.003 pound from the 0.004 pound previously received from the register.

Under these assumed conditions with the calculations performed in accordance with block 90 indicating no motion, block 91 will then cause the control to be advanced to block 92 when the motion flag MOTNF is read out and checked. This time the motion flag will not be zero since a 1 was previously stored in the status register space 718 and has not been changed. As a result, control is advanced to block 93 where the no motion counter or hit counter 726 of FIG. 7 is incremented by one. Under the assumed condition, 0 has been previously stored in this counter so the counter will now indicate a count of 1. Control is then advanced to block 94 where the count is this counter 726 is read out and since it is less than some predetermined value, the count will not be greater than this value with the result that the control is transferred via F2 to block 99 and the remaining portion of the cycle repeated as described above.

On the next cycle during which weight indication is again received from the analog-to-digital converter in the manner described above, if no motion is still determined by block 90, then the hit counter 726 will again be incremented and have a count of 2. Thus each cycle during which no motion is present, this counter is incremented by 1. If, however, during any of these cycles the calculations of block 90 indicate that motion is present, then block 91 will transfer the control to block 98 via transfer F1 at which time the hit counter 726 is cleared, i.e zeros are stored in the counter spaces, and the motion flag 718 is again set to 1; in other words, a 1 is written over the 1 already in this storage register space.

After the predetermined number of no motion indications have been obtained from calculations of block 90 and the control advanced through blocks 91, 92, 93 to 94, it will be determined that the counter is in excess of the required number of counts.

Thus, to obtain a no motion indication, it is not only necessary that the scale remain within a fixed band indication of weights, but also that it remain in this band for a predetermined interval of time. This interval of time can be predetermined within increments of 0.200 of a second, the cycle period of an analog-to-digital converter. Assuming that five such increments are required thus requiring that the weight indication remain within 0.003 pound for one second when the control is transferred to block 94 and a count of five stored in the counter 726, block 94 will advance control to block 95 where the motion flag MOTNF is cleared. In other words, a 0 is now recorded in status register space 718.

Thereafter the control is transferred to block 99 via transfer F2 and then advanced through the blocks 99, 100, 103, transfer F6 to block 105, block 107, block 109 and to block 110.

In block 110, the motion flag is again tested and since a 0 is now stored in the status register 718, the control is transferred to block 115 over the transfer G1 and thereafter the control is transferred as described hereinafter under the heading of "Automatic Zero Correction". On the next cycle of the control equipment 15, after the no motion flag is cleared as described, control is returned to the block 26. In accordance with this block, motion flag is read out and tested and since it is clear, control is advanced to block 27 where it is determined whether the scale is arranged to operate in the demand mode or continuous mode. Under the assumed conditions, the scale is arranged to operate in the manual mode and not in the demand or continuous mode so that the control continues to a block 28. At the block 28, a check is made to see if a "motion detector inhibit" switch is closed to disable the motion detector portion of the apparatus 15. It is assumed that the switch is not closed and the control continues to block 29. Block 29 determines that a reweigh is not necessary at this time so control is transferred to block 30. In accordance with block 30, the control equipment under the assumed conditions will determine that print data is not stored so the control sequence advances to block 31. Block 31 determines that the no tare key is not pressed so that control is advanced to block 34. If the weight is greater than 0.1 pound then the control is advanced to block 35 where the set status register space 737 is tested. This space 737 is assumed to be true so that the control is then transferred back to block 25 via transfer A1. If the weight as assumed is not greater than 0.1 pound then block 34 causes the control to be transferred back to block 25 via transfer A1. In either case the above-described cycle then repeats starting with block 25 for each of the various cycles described so long as the motion flag is not set; that is so long as zero remains recorded in the status register 718.

If the weight is greater than 0.1 pound, the set is not true, the weight is not minus, and is not over the capacity of the scale, then the control sequence advances through blocks 34, 35, 36 and 37 to block 38. Block 38 causes an output print pulse which causes the weight to be printed. Then the sequence returns to block 25 via transfer A1. Thereafter, the above-described sequences are repeated.

AUTOMATIC ZERO CORRECTION

Assume now that either when the scale is first turned on or that after the previous object has been weighed and removed from the scale, the scale has not returned accurately to zero. Assume, for example, that the scale has returned to 00.0004. also assume that zeros are entered in the zero correction register 711 and after the various operations referred to above, have been completed, 00.004 pound is entered in the raw weight register 710. A 1 will be registered in the weight sign register 715 and 1 in the zero correction sign register 714.

The one in the sign registers indicate that the sign of the weight or zero correction is plus and zeros entered in these registers represent a negative sign or weight indication. The manner in which the various signs are determined and entered into the respective registers will be apparent from the following description.

Thus, on the first cycle of the control equipment 15 after the previous weight has been removed from the scale and the scale returned to idle condition or the power turned on, 00.004 will be entered into the raw weight register 710 and a 1 indicating a plus will be entered in the weight sign register 710 and a 1 indicating a plus will be entered in the weight sign register space 715. Next during this same cycle of operation of the control unit 15, the weight in the zero correction register 711 is subtracted (block 105 FIG. 2F) from the weight in the raw weight register and the difference is then re-entered in the raw weight register 710. Under these assumed conditions, zero will be subtracted from the 00.004 recorded in the raw weight register and the difference 00.004 rerecorded in the raw weight register 710.

Next the automatic zero correction limit of 00.005 pound is subtracted (block 115 FIG. 2G) from the 00.004 pound in the raw weight register and since the 00.004 pound is less, the weight indication of the scale is within the automatic zero correcting range so the control equipment will then follow additional steps required to automatically correct the 00.004 pound. Next the 00.004 pound in the raw weight register is compared with the zero or 1/4 graduation range of 00.002 pound (block 117 FIG. 2G).

Both this comparison and the previous comparison may be made in any desired manner such as, for example, by means of threshold circuits of any well known and suitable type. However, in accordance with the present invention, this comparison is made by the central processor unit CPU 186 of the microcomputer or control unit 15.

Briefly, a series of orders or instructions direct the central processing unit 186 to predetermined locations in the ROM units where the limits of 00.005 and 00.002 are stored. These instructions direct the central processing unit 186 to obtain these limits and then subtract them from the weight in the raw weight register. As a result, the central processing unit 186 causes the limit 00.002 obtained from the predetermined address in the ROM to be subtracted from the 00.004 stored in the raw weight register 710. The result of this subtraction is 00.002 which is a positive number thus indicating that the magnitude of the weight in the raw weight register is not equal to or less than 0.002 pound. As a result, 12 will be entered in the zero count register 713 (block 118, FIG. 2G).

Next, the reading of 12 in the zero count register 713 is read out by the control equipment (block 120) and since this reading is less than 16, the zero count register 713 does not return to zero, the count in this register is incremented by one so that a 13 is now stored in the zero count register 713 (blocks 119 and 114, FIG. 2G).

Next, since the weight stored in the raw weight register is not 00.000 (block 132, FIG. 2G) the weight recorded in the zero correction register 711 is compared with the 0.6 pound limit for the operation of the zero correction (block 121, FIG. 2G). In performing this comparison, as above, any suitable type of comparing or threshold circuits may be employed, however, in the specific arrangement described herein, the central processing unit CPU 186 will obtain, under control of orders or instructions stored in the ROM's 188-192, the 0.6 of a pound from a predetermined location in the read only memory and then subtract the weight in the zero correction register 711 from this value. If the value in the zero correction register 711 is less than the 0.6 of a pound the zero correcting operation will continue. This limit of 0.6 pound and also the other limits of 00.005 and 00.002 pound may be changed or determined by the value stored in the read only storage devices ROM's at the predetermined locations employed for storing these values.

Since under the assumed conditions the 00.000 stored in the zero correction register is less than 0.6 pound the weight stored in the zero correction register 711 is augmented or changed by 00.001 pound (block 122, FIG. 2G) and the sign in the zero correction sign register 714 is changed to a plus.

Next a round-off operation is performed by adding 00.005 to the weight now recorded in the raw weight register 710. As a result, the weight of 00.009 is rerecorded in the raw weight register 710 (block 123, FIG. 2G).

Next the four most significant digits recorded in the raw weight register 710 which are all zeros is transferred to the output weight register 712 so that they will be available for controlling the output indication of 00.00.

During a subsequent portion of this first cycle, the 13 stored in the zero count register 713 is again read out and since this number is not 16, i.e. not 0, the 1/4 graduation lamp is turned off if it was on or maintained off if it was already off. Thereafter, during the remaining portion of this first cycle of the control unit 15, other operations may be performed. However, the values stored in the various registers described above will remain until the next operating cycle of this control unit 15.

During the next cycle of operation of the control unit 15, the 00.004 weight will be registered again in a raw weight register 710 in the same manner as during the first cycle. Then the value in the zero correction register 711 which is now 00.001 pound will be subtracted from the weight in the raw weight register 710 and the difference is rerecorded in the raw weight register 710. Thus 0.003 pound is now stored or recorded in the raw weight register 710.

Next the 00.003 pound stored in the raw weight register 710 is compared with the 00.005 pound operating limit of the automatic zero correcting operation which is obtained from the read only storage units of the system. Since the result of subtracting the 00.003 pound from the 00.005 pound is positive, the plus sign remains stored in the zero correction sign register and the system continues to perform additional operations required for the automatic zero correction in accordance with the present invention.

Next the 00.003 pound stored in the raw weight register 710 is compared with the zero or 1/4 graduation range in the manner described above. Thus the 00.002 pound, obtained from the read only memory at the predetermined location assigned to this limit, is subtracted from the 0.003 pound in the raw weight register 710. The result of this subtraction is a positive number, namely 00.001 pound.

Since the weight in the raw weight register is not less than 00.002 pound the control unit causes 12 to be again entered in the zero count register 713 and then this register read out as described above. Since the count is not 16, i.e. not 0, the count in the zero count register 713 is incremented by one so that 13 is now recorded in the zero count register 713. In addition, since the weight in the raw weight register 710 is not 00.000 the weight stored in the zero correction register 711 of 00.001 pound, is compared with the limits of 0.6 pound of the zero correction range by subtracting the 00.001 from the 0.6 pound. The result is positive so the weight in the zero correction register 711 is augmented by 00.001 pound.

As a result 00.002 pound is now recorded in the zero correction register 711. Next, a roundoff operation is performed by adding 00.005 pound to the 00.003 pound in the raw weight register 710 and the result 00.008 pound recorded in the raw weight register 710.

Next the four most significant digits in the raw weight register, namely 0000, are transferred to the output weight register so they will be available for controlling the display of 0000.

Also during this second cycle, the 13 recorded in the zero count register is again read ou