GB1604841A - Method of and apparatus for determining a dimension of an article - Google Patents

Description

(54) A METHOD OF AND APPARATUS FOR DETERMINING A DIMENSION OF AN ARTICLE (71) We, GKN FASTENERS LIMITED, a British Company of, Cranford Street, Smethwick, Warley, West Midlands, B66 2SA, do hereby declare the invention for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement:- This invention relates to a method of and apparatus for determining a dimension of an article.The expression "determining the dimension of an article" is to be deemed herein to mean any of the following, namely determining the numerical value of the dimension, determining whether the dimension conforms or does not conform to a predetermined value or lies above or below the predetermined value or lies between predetermined values, and upon performance of any of these determinations indicating and/or recording the result and/or performing an operation consequential upon the result.

The invention has been developed primarily in relation to a requirement to determine at least one dimension of each of a succession of articles contactlessly, that is to say without the boundaries of the article which present the dimension to be determined being brought into physical contact with a measuring means for the purpose of such determination. The articles may be blanks for components for making other articles which are not yet manufactured or fully manufactured to a final form, or may be the fully manufactured articles themselves. A particular example of articles for which the apparatus in accordance with the present invention has been primarily developed is blanks for forming screws.Such blanks may each comprise a shank portion of circular cross-sectlon and a head portion also of circular cross-section in a plane perpendicular to the axis of the shank portion but of larger diameter than the diameter of the shank portion, and integrally connected to the latter at one end.

One of the dimensions which may be required to be determined is the diameter of the head portion and another dimension may be the length of the shank portion, but it is of course to be understood that, even as applied to this restricted class of article, other dimensions may require to be determined, for example the diameter of the shank portion.

Articles of relatively light weight, such as screw blanks, can be fed to the apparatus at relatively high feed rates, for example of the order of 1000 articles per minute, there being nevertheless a considerable variation in the feed rate depending upon the precise form of the article and other considerations such as the number of the dimensions to be determined, the accuracy with which such dimensions are to be determined, and the degree of positional control (if any) which it is necessary to exert upon the articles to ensure that such determination shall be accurate. Typically, however, feed rates for screw blanks may be in the range of 140 to 1400 articles per minute.

Besides the basic requirement of providing accurate dimension determination, whether the articles be screw blanks or other small blanks or components, or fully manufactured articles, it is further desirable that there shall be reliability in identifying the determined dimension with a particular article and providing the requisite indication or operation (for example displacing reject articles not conforming to dimensional specification from the normal feed path to a reject station or container) irrespective of variations in the rate of feed.Further, it is desirable that if the articles undergoing dimension determination in the apparatus are of a form such, or become oriented to a position such, that there is obstruction to or jamming of continued operation of the apparatus, the dimension determining operation as a whole (in relation to the continuous succession of articles supplied) shall not be brought to a halt indefinitely (in the absence of supervision).

The principal object of the present invention is to provide a method of, and apparatus for, determining the dimensions of articles which more nearly meets these requirements than those hitherto available.

According to a first aspect of the invention, a method of determining a dimension of an article comprises the steps of: a. scanning the article to generate gating signals representative of the boundaries of said dimension, b. generating a succession of electrical encoder signals representative of respective increments of said dimension, c. gating the encoder signals by the gating signals and determining the number of encoder signals gated, d. storing a number representative of the encoder signals which would pertain to an article having a reject value of said dimension, e. comparing the number of encoder signals determined by said gating for the article undergoing measurement with said stored number, and f. developing an output signal in response to said comparison.

In one manner of carrying out the method, the encoder signals may be generated by a shaft encoder having relatively rotatable elements, the gating signals may be generated by photo electric means subjected to relative scan by the article, and the relative rotation of the shaft encoder elements and the relative scanning movement may be coordinated to establish that an increment of one of said relative movements (i.e. that between the shaft encoder elements) is always accompanied by an increment of the other of said relative movements (i.e. scanning of the photo electric means by the article) maintaining a constant ratio of magnitudes of these increments.Alternatively the encoder signals may be generated by photo electric means subjected to relative scan by the article, and the gating signals may also be generated by photo electric means subjected to relative scan by the article (such photo electric means for the two functions may be separate from each other or elements of the photo electric means may be common to the two functions).

According to a second aspect of the invention, an apparatus for determining a dimension of an article comprises: a. a body or supporting structure affording a feed path for the article, the feed path passing through a station (herein called the gauging station) at which the dimension of the article is to be determined, b. feed means for advancing the article through the gauging station in an orientation such that opposite boundaries of the dimension to be determined are spaced apart along the feed path, and drive means for the feed means, c. encoder means for generating a succession of electrical pulses representing increments of travel along the feed path and occurring in respect of the article at least over a distance in the region of the gauging station, d. sensing means responsive to passage of the article through the gauging station to provide electrical boundary signals representing respectively opposite boundaries of the dimension to be determined, and e. electrical circuit means (herein called the measuring circuit) associated operatively with the encoder means and the sensing means and including, means to receive. said pulses and said boundary signals and to operate arithmetically thereon to determine the number of pulses occurring between said boundary signals, means for storing a number representative of the encoder pulses which would pertain to an article having a reject value of said dimension, comparator means for comparing the number of pulses produced by the article undergoing measurement with the number stored, and means for developing an output signal in response to said comparison.

The output signal may be a simple pass or reject signal for each article and/or may contain digital data representing the numerical value of the dimension determined.

The encoder means may comprise a rotary shaft encoder which is driven in timed relation with the feed means so that each increment of rotation of the shaft encoder represents a corresponding distance travelled along the feed path for the article undergoing dimension determination at the gauging station.

Convenicntly, but not essentially, the feed means may comprise a rotary feed member having article receiving formations spaced apart peripherally around its axis of rotation and serving to advance the articles through the gauging station, and such rotary feed member would preferably be connected to a rotary input element of the shaft encoder directly. This eliminates any inaccuracy in counting due to backlash which otherwise might exist between the feed member and the rotary input element of the encoder means.

The sensing means may comprise photo electric means so mounted on the body or supporting structure of the apparatus as to be subjected to a travelling image of a profile of the article, such profile containing the dimension to be determined.

Alternatively the encoder means may comprise an array of photo electric cells extending with the length, or major dimension, of the array along the feed path through the gauging station, said array of cells including a group of closely spaced cells for generating the encoder signals, and said sensing means comprising at least one photo electric cell for generating said boundary signals.

When a plurality of cells are provided for generating said boundary signals, there may be two such cells situated respectively at opposite ends of said closely spaced group of cells. In a preferred arrangement further boundary signal generating cells (additional to the two at the ends of said group) are provided at intervals along said feed path beyond the downstream end (having regard to the direction of feed) of said group at spacings which are multiples of the length of said group.

Optical means may be provided for projecting the image onto the photo electric means (whether provided for encoder pulse generation or for boundary signal generation) with magnification of the dimension.

The apparatus may, and preferably does, include sorting means for displacing selected articles from the feed path (such articles being those for which dimension determination results in a reject status or a pass status as may be convenient), said sorting means being responsive to the output signal from the electrical circuit means aforesaid. In forms of apparatus in which the feed means includes means for varying the speed at which it effects advancement of the articles along the feed path, the sorting means may have electrically energised operating means but may nevertheless be of a character having an intrinsic or unavoidable delay time before it can be brought into operation to be effective to displace the selected article after incidence of an electrical initiating signal.

A further object of the invention is to overcome or reduce the risk of malfunction from the cause above mentioned.

According to a further feature of the invention the apparatus may include: a. sorting means operable in response to a reject signal derived from the output signal of the measuring circuit to separate articles for rejection (herein called reject articles) which are outside the limit or limits of tolerance of the dimension of articles to be passed (herein called pass articles from the pass articles), said sorting means being operable in response to presence or absence of the reject signal but after a delay dependent upon operational characteristics of said sorting means, b. means for varying the speed at which the articles advance along the feed path, and c. compensating means for bringing the sorting means into operation to effect displacement of a selected article at a sorting station situated at a substantially fixed position along the feed path from the gauging station, so that the selected article is properly operated upon by the sorting means irrespective of speed variation of the feed means.

Thus, the sorting means may include, an electrically energised operating means controlled by a control circuit which is connected to receive a speed signal representing speed of advancement of the articles and may embody a subcircuit or circuit element presenting an electrical parameter representative of the delay time of the sorting means, arithmetical circuits also being provided to operate on the speed signal and delay time parameters to develop an output signal for initiating operation of the sorting means at such time, in advance of arrival of the selected article at the sorting station, as will ensure proper operation thereon by the sorting means.

It may happen from time to time that the articles undergoing measurement become displaced while being caused to travel along the feed path by the feed means, with the result that the operation of the feed means is stopped due to the article becoming jammed between a movable element of the feed - means and a fixed element such as part of the body or supporting structure of the apparatus.

A further object of the invention is to ameliorate the effects of such malfunctioning.

According to yet another feature of the invention the apparatus may include: anti-jamming means comprising a circuit (herein called the anti-jamming circuit) for temporarily reversing forward movement of the drive means in response to the occurrence of a jammed condition and thereafter restoring said drive means to forward movement.

In a preferred form the measuring circuit include: inhibiting means for preventing a false count of encoder pulses occurring during the interval of reversed and restored forward drive of the feed means until the latter again attains the position at which the jammed condition initially took place.

The anti-jamming circuit may further include means for temporarily reversing and restoring the drive means to forward operation a predetermined number of times to attempt to clear a jammed condition. A shut-down circuit may further be provided responsive to attainment of said predetermined number to shut down the apparatus upOn failure to clear the jammed condition after such number of cycles of reversed and restored forward movement.

The invention will now be described, by way of example, with reference to the drawings accompanying the provisional specification wherein: Figure 1 is a perspective view of the complete apparatus; Figure 2 is a plan view of the apparatus in accordance with the invention for carrying out the method thereof; Figure 3 is a fragmentary view in end elevation looking in the direction of the arrow A of Figure 1; Figure 4 is a view in side elevation and in vertical cross-section on the line Il of Figure 2; Figure 5 is a fragmentary view in front elevation showing the screen upon which a "shadow" of each article under examination is formed, and the arrangement of photo-sensitive elements from which dimensional data is derived; Figure 6 is a fragmentary view in section on the line X-X of Figure 5;; Figure 7 is a schematic circuit diagram of the apparatus; Figure 8 is a circuit diagram of the circuit card for developing logic to determine the diameter dimension of each article and developing a reject signal; Figure 9 is a circuit diagram of an input circuit card for developing logic, e.g.

detecting articles having shank lengths which are too short and too long; Figure 10 is a circuit diagram of an interface circuit card for feeding encoder (diameter measuring) pulses to the circuit of Figure 8; Figure 11 is a circuit diagram of a circuit for providing the reject signal to operate an air blast device with appropriate delay dependent upon feeding speed for the articles; Figure 12 is a circuit diagram of a circuit card providing antijamming facilities, e.g.

reversed drive of the feed disc and thereafter restored forward drive for a predetermined number of times; Figure 13 illustrates the time relationships between signals developed in the various circuits.

The invention is further described with reference to the drawings accompanying this specification which are numbered sequentially from those above referred to and wherein: Figure 14 illustrates diagrammatically for a second embodiment of the apparatus, a photo cell array which serves to replace the 4000 p.p.r. output of the shaft encoder in the circuit of Figure 10 of the first embodiment, and also to replace the diameter measuring photo transistor of the previous embodiment, and to provide an output for the generation of encoder signals and effective gating signals; Figure 15 is the circuit diagram of an input card for use in the second embodiment of apparatus to receive the output of the photo transistors of Figure 14 and develop the encoding and effective gating signals;; Figure 16 is a circuit diagram of a circuit card for developing logic to determine the diameter of each article and developing reject signals and effectively replaces Figure 8 of the first embodiment.

Referring firstly to the mechanical layout of the apparatus, the apparatus comprises a body or supporting structure 10 on which is mounted a container Il in the form of a hopper for the articles to undergo examination. As now described the apparatus is used for examining screw blanks each having a shank portion and a head portion connected integrally thereto, the apparatus being designed to provide numerical determination of shank diameter and determination of whether shank length lies between a short limit and a long limit. It will, of course, be understood that the form of the feed means and arrangement of sensing means can be modified for examination of other forms of article within the scope of the invention.

Feed means, hereinafter described in more detail, is provided for advancing the blanks through a gauging station 12 (Figure 2) and thereafter through a sorting (reject) station 13, and finally to a pass station 14.

The feed means comprises a first part 15 in the form of a guideway affording a longitudinally extending slot between guide rails, the latter being driven in a vibratory mode and inclined downwardly from the hopper 11. From an outlet on the hopper blanks are fed into the slot so that their heads re ;t on the upper surfaces of the guide rail and the blanks move downwardly under the combined effects of gravity and the vibratory movement.

The feed means further comprises a second part 16, also including guide rails 18, between which is defined a slot 19 for the progression therealong of the blanks, one of the rails having obliquely arranged passageways 20 into which compressed air may be fed for establishing air jets to urge the blanks in a forward direction. Further, the feed means comprises a third part in the form of a disc 21 (Figure 4) mounted for rotation about a vertical axis 22 and having equally angularly spaced slots 23 extending to its periphery, for example twenty in number, which receive respective blanks and, by rotation of the disc, advance these through the stations 12, 13 and 14 already mentioned.

The disc is carried by a vertical shaft 24 supported in a bearing structure 25 mounted in the top wall of the body or supporting structure 10 and carrying a drive pulley 26 adjacent to its lower end driven by a belt 27 from a pulley 28 on the shaft 29 of an electric driving motor (not shown), drive being established by the pulley 28 through the intermediary of a friction clutch 30. The latter reduces shock load on the drive mechanism in the event of a blank becoming jammed between the disc and some fixed part of the apparatus.

For producing an optical image (in profile) of each blank as viewed in a direction at right angles to its axis, the apparatus further comprises a light source 31 providing a light beam havinarallel rays directed along the axis 32 of the light source and incident upon a blank at the gauging station 12. A lens system 33 on the opposite side of the gauging station produces an image in the form of a shadow, of which the periphery represents the profile of the blank, and which is reflected by mirror elements 34a, 34b contained in the body 10 to be incident at a screen 35 arranged in a forwardly and upwardly presented attitude at one end of the apparatus.

The shadow images presented on the screen are seen in Figure 5 and include a shadow image 36 due to a portion of the feed disc, an image 37 due to the shank portion of the blank, and an image 38 due to the head portion of the blank.

At the outer side of the screen photosensitive means is provided responsive to passage of the image 37 across the screen, e.g. from left to right as seen in Figure 5.

The photosensitive means is supported by mounting means providing for positional adjustment of the photo-sensitive means to suit various forms of articles to undergo examination.

Typically guide elements in the form of bars 39 and 40 parallel to each other are provided at the front of the screen and on these are mounted respective photosensitive means 41, 42. 41 ma comprise a single photo transistor and associated indicator 43, whereas photo-sensitive means 42 may comprise two photo transistors and associated indicator elements 44 and 45.

Photosensitive means 43 provides effectively for gating of shank diameter, while elements 44 and 45 provide collectively signals determining whether the shank is within short and long limits, the difference between these limits being dependent upon the spacing of the photo cells and the degree of magnification provided by the optical system which typically is ten times. The guide bars may be mounted for adjustment of position in directions at right angles to their lengths and the photo-sensitive elements may be mounted for adjustment both along and, if desired, laterally of the guide bars.

Driven in synchronism with the feed disc 21 is a means for generating encoder pulses.

conveniently this is in the form of a shaft encoder of known construction and is mounted as seen at 46 at the lower end of the shaft 24 and has its rotary output element 47 coupled to the shaft 24. The shaft encoder is arranged to generate, for each revolution of the disc, 1000 primary pulses. As hereinafter described three outputs are developed from the 1000 primary pulses, these being 2000 p.p.r.

(pulses per revolution) for developing a signal to ensure operation of the sorting (reject) device at the proper time, notwithstanding feed disc speed variations; 4000 p.p.r. (for shank dimension measurement); 20 p.p.r. (for controlling operation of circuits as hereinafter described) and 1 p.p.r. (a marker pulse likewise for controlling operation of the circuits).

Generation of the pulses by the shaft encoder may be effected in any suitable manner, for example by photo-sensitive elements incorporated in the shaft encoder which are successively masked from a light source by a rotor. Other systems may, however, be employed in suitable cases, for example electro-magnetic generation.

In general terms, the various subordinate circuits of the apparatus, which collectively form the whole electrical circuit, provide the following functions: (a) Numerical determination of shank diameter; (b) Non-numerical determination of shank length; (c) Development of a pass or reject signal according to whether these quantities are within the specification or not;; (d) The use of the reject signal to trigger operation of a sorting (reject) device for separating blanks having a reject status from blanks having a pass status and with compensation as to timing to ensure that the reject device operates when reject status blanks are properly positioned in its field of operation, notwithstanding the possible variations in the speed of rotation of the feed disc 21 to suit different types of blank or other requirements, and preferably with the ability to adjust the "width" of the field of operation of the reject device according to the character of the blanks;; (e) Anti-jamming facilities in the form of temporary reversal and restoration of forward movement of the drive motor (with inhibition of dimension determination by the encoding pulses to avoid false determination during the antijamming procedure) and with facilities for controlling the number of reversals and restorations which can occur before shut down of the apparatus.

Referring now to Figure 7, the devices enclosed by the boundary 50 substantially perform the function of determining shank diameter numerically and effecting a nonnumerical determination of shank length; those enclosed by boundary 51 substantially control operation of the reject device, i.e.

an electro-magnetically controlled air blast nozzle as hereinafter mentioned, and those enclosed by boundary 52 substantially comprise the anti-jamming circuits.

In general terms the schematically shown units of Figure 7 operate as follows. Sorter logic and reject logic circuits 54 and 55 respectively receive a shank diameter signal on terminal tl of a width representative of shank diameter, and also receive shank long and shank short signals on terminals t2, t3 in the event of either of these conditions being detected. A "set function" unit 56 allows the operator to select dimensional tolerances for reject in respect of diameter, and to select either short or long reject, or both. Diameter sample unit 57 provides for programming of the logic circuits by passage of a "good" sample through the apparatus.

Encoder pulses at 4000 p.p.r. are fed to the unit 54 and are gated by the diameter width signal at terminal tl to provide numerical determination of shank diameter.

For physical convenience the reject station 13 is spaced angularly from the gauging station 12 by four pitches (twenty pitch marker signals are developed for a complete revolution of the feed disc) so that the actual angular spacing is 72 . Actuation of a reject device, for example an air nozzle in which the air supply is controlled by a solenoid operated valve, takes place at a time determined by AND gate unit 58 but requires to be delayed at least by the time corresponding to four pitches.

Consequently the reject signals fed on channel 59 to the AND gate 58 are routed through a four-bit shift register 60 which is stepped on by pitch marker pulses.

Unless the feed disc 21 is moving at its fastest speed, onset of the air blast requires to be further delayed to ensure that it will be "centered" on the blank generating the reject signal at the time at which such blank reaches the reject station. For this purpose a timing logic circuit 61 provides a second output to AND gate 58 and is itself programmed by inputs from a pitch marker generator (developing pulses at the rate 20 p.p.r.) and a disc speed sensor 62 which measures the number of 2000 p.p.r. encoder pulses in a predetermined interval (15 milliseconds). The timing logic circuit 61 provides also for manual selection of the "width", i.e. duration, of the air blast.

In the event of a jammed condition occurring causing the feed disc to stop or slow down to below a predetermined speed, a "missing pulse" detector 64 sensing discontinuance of the 4000 p.p.r. encoder pulses, provides an initiating signal to a reversal logic circuit 65 controlling the configuration of drive motor input circuit, and hence the direction of drive of the motor driving the feed disc. Establishment of current to the motor by way of this circuit configuration is controlled by timer circuits 66, 67 and an OR gate device 68, so that the motor reverses in an attempt to clear the jam, and after a predetermined reverse rotation is restored to forward drive.

During this operation the reversal logic circuit develops an inhibit signal inhibiting the supply of 4000 p.p.r. encoder pulses to the sorter logic circuit 54 so as to prevent false counting of shank diameter.

A manually settable attempt index circuit 69 is provided so that the apparatus can be programmed to shut down after a predetermined number of jamming clearance attempts, e.g. one to four such attempts.

The attempt index circuit 69 is reset through OR gate device 70 by the synchronising pulses generated once per revolution of the disc to prevent attempts registering cumulatively for faulty blanks handled in successive cycles of rotation of the disc. Initial reset of the attempt index is also brought about by operation of a start press button switch connected to terminal t4.

With respect to the remaining inputs and outputs shown in Figure 7, terminal t5 is connected to a counter for recording or displaying total count of blanks examined, terminal t7 to a device for recording or displaying total reject blanks, t8 to the disc shaft motor relay or contactor for establishing energisation, terminal t9 to the supply circuit switch system of the motor for determining direction of rotation when energised, and terminal t10 to a solenoid vale controlling air supply to the air blast nozzle forming the reject device.

Referring now to Figure 10, this shows a circuit for developing the various pulses required in operation of the apparatus, namely (a) 4000 p.p.r. (pulses per revolution of the feed disc and encoder rotor) for encoding the sorter logic circuit; (b) 2000 p.p.r. for feeding to a further circuit establishing operation of the reject device and the correct time and as a basis for generating pitch marker pulses 20 p.p.r.: (c) 1 p.p.r. for use as a synchronising pulse.

The shaft encoder provides two outputs which are connected to terminals tll and t12 and which provide substantially square wave pulses each having a mark space ratio of substantially one, and in quadrature with each other, both being generated at a rate of 1000 p.p.r. (complete cycles).

EXCL. OR gates 70 and 71 have their input terminals connected, as shown, between positive voltage terminals energised from a low voltage source conveniently 15 volts, as shown at t13 through resistor chains Rl, R2 and R4, R3 to earth and are connected via capacitors Cl to C4 to the terminals tll, t12.

References "1" or "0" adjacent relevant terminals indicate the initial states and (1) and (0) indicate states occurring during operation by incidence of rising and falling wave fronts in the input signals at tl 1, tl2.

Thus the outputs of the EXCL. OR gates (70,71) deliver negative-going "0" pulses each at 2000 p.p.r., with those delivered from the output of 71 occurring midway between those delivered from the output 70.

The two pulse trains are added in NAND gate 73 and appear as a single pulse train having positive-going pulses "1" at the output terminal of EXCL. OR gate 74 connected to terminals tl4, tl5 at which the pulse train at the rate of 4000 p.p.r.

The "0" pulses 2000 p.p.r. transmitted through NOR gate 75 to terminal tl6 will appear as a positive-going 2000 p.p.r. pulse train.

The shaft encoder provides a separate output to terminal tl7 at the rate of 1 p.p.r.

This is fed via C5 through two NAND gates 76, 77 connected as invertors to terminal t18. Also shown in Figure 10 is a driver circuit for the solenoid valve controlling air blast. The input terminal tl9 would be connected to t10, Figure 7, as shown in brackets thereon. Transistors Tl, T2 provide a current output at terminal t20 to the winding of the solenoid air valve, the other end of which winding is connected to a suitable source of positive voltage. Zener diode ZD1 limits the positive voltage appearing across the collector to emitter of T2, thus controlling the back e.m.f. of the solenoid at switch-off.

Referring now to Figure 8, which shows the logic sorter circuit for determining the numerical value of the shank dimension of the blank in comparing it with "good" preset values, and also for receiving and analysing "short" and/or "long" shank length signals, the primary diameter parameter or gating signal produced as a result of scanning the image of the shank portion by the photosensitive element 43 is fed in at terminal t23. Initially it is required that such signal shall be generated by a "good" blank, i.e. one of which the shank diameter is within the tolerances specified.

To conduct this operation the operator presser a "take sample" switch. which is connected to terminal t21 changing the state of this terminal from "1" derived from the positive voltage terminal t22 to "0" and setting NAND latch 79 and 80 to change state. Thus an initial "0" output at the outlet terminal of unit 80 becomes a "1", which is applied to the lower input terminal of NAND gate 81, and appears as a "0" at its output terminal.

A "0" signal of which the width represents the shank diameter is applied to terminal t23 from hoto-cell element 43, and after operating chmidt trigger invertor unit to produce a sharply defined "1" at its outlet appears as a "I ' at the outlet of the second of two NAND gates 81, 82 for application to the control terminal 2 of counter 84a, 84b.

The lower input terminal of NAND gate 81 is held at "0" by operation of NAND latch 79, 80 (upon operation of the "take sample" switch), which also feeds a "0" input to a timer unit 83 providing a 1.2 millisecond delay pulse. The trailing edge of this pulse provides a "0" signal at the output terminal 10 of unit 83 and is transmitted via capacitor C3 to the lower inlet of NAND gate 82, thereby allowing the "diameter" gating "0" to pass to the outlet.

The purpose of the time delay unit 83 is to ensure that the operative output "0" occurs significantly after manual closure of the switch, so that contact faults, vibration or the like will not give rise to spurious pulses at the lower inlet terminal of NAND gate 82 and provides the initial "I" count after the counter reset pulse (from gate 92).

Counter 84a and 84b each provide a binary count 16, and when connected as shown provide a total count of 256. Counter 84a, 84b is required to store numerical data corresponding to the higher tolerance of a "good" sample. The initial " I " pulse applied to control terminal No. 2 is in fact registered as a count, so that this counter holds a value (n+l) where n will be the count of an actual sample.

After operation of the "take sample" switch, encoder pulses of the 4000 p.p.r.

series from terminal tl can pass into the counter 84a, 84b to be stored therein which will eventually store a number representative of shank diameter of the "good" sample+ 1 (arising from operation of the "take sample" switch). The encoder pulses are obtained from terminal t72 (Figure 11) connected to the output terminal of OR gate 112a. Input terminal t71 thereof receives encoder pulses of the 4000 p.p.r. series fed in at terminal t71 from tl4 (Figure 10) and such pulses will appear at the output of gate 112a and hence t72 unless an inhibit signal "I" is applied to terminal t70.

The trailing edge of the shadow signal appearing at outlet terminal No. Il of TRIGGER-NAND unit 78 passes through capacitor Cl and resets the NAND latch 79, 80 thereby holding the count stored in counter 84a, 84b at the value previously stated.

Counter 85a, 85b, which is similar to 84a, 84b, is used to store data corresponding to the lower tolerance of a "good" sample and is programmed from counter 84a, 84b by "subtraction" of a predetermined count brought about by setting of switch 86.

Typically the number subtracted may be 7 so that the count stored in 85a, 85b will be n6 where n is the count corresponding to "good" shank diameter.

After storage of the (n+l) and (no6) data in the reference counters 84a, 84b and 85a, 85b "control" terminal No. 2 of counter 84a, 84b will be "0" and no further count will be stored therein. Further encoder pulses, due to passage through the gauging station of blanks to be examined, will be counted in counter 87a, 87b, of which the control inlet terminal No. 2 is connected directly to outlet terminal of unit 78, so that this counter is enabled whenever a "I" gating pulse representing shank diameter of such a blank occurs at this terminal.

The count registered in 87a, 87b is compared with both that registered in 84 a, 84b and in 85a, 85b, such comparison being effected in respective comparators 88a, 88b and 89a, 89b. The outlet terminals No. 13 and No. 12 respectively of these comparators remain at "1" until, in the case of 88a, 88b, an oversize diameter blank is detected, and in the case of 89a, 89b an undersize blank is detected, in which case the relevant outlet goes "0".

NOR gates 90 and 91 have respective input terminals connected to terminals t26 and t27 which are normally at "0" but either or both of which can be switched to "I" by operation of selector switches respectively for oversize rejection or undersize rejection, or both. Incidence of a "0" at the other input terminals of NOR gates 90 and 91, due respectively to an oversized or undersized blank diameter, will produce reject signals "1" at terminals t28, t29.

Reset and indicator facilities are incorporated in the circuit shown in Figure 8. In respect of operation of the photosensitive element 43, the indicating LED (light emitting diode) associated therewith is connected to terminal t24. Transistor Tl is caused to conduct when terminal No. 11 of TRIGGER-NAND unit 78 has gone to the "1" state. The LED thus conducts when the associated photo sensitive element 43, which may conveniently comprise a phototransistor, is in shadow. This simply provides an induction to the operator of proper functioning up to the outlet terminal of unit 78. Terminal t5 will go to a "1" for each screw blank passing through the gauging station (irrespective of shank diameter) and can be connected to a counter for recording the total of such blanks passing.

NAND gate 92 normally provides a "0" at its outlet. When the "take sample" switch is operated, however, capacitor C4 momentarily produces a "1" at this terminal and resets the reference counters 84a, 84b and 85a, 85b. This takes place on the leading edge of the pulse produced in timer 83 by operation of the "take sample" switch so that these counters are reset before incidence of the single pulse giving rise to the +l in the reference (n+l) generated from the trailing edge of the pulse produced in timer 83.

Terminal t31 is connected to the positive voltage supply and terminal t32 receives pitch marker signals 20 p.p.r. (generated in the circuit of Figure 10 hereinbefore described), which latter are fed to timer 93 producing a "1" pulse at its outlet of 0.12 milliseconds. The purpose of the timer 93 is to produce a delay in resetting the counters 87a, 87b to allow any reject signals produced from the comparators 88a, 88b and/or 89a, 89b to be incident at terminals t28, t29 (for operating the shift register 125 in the circuit of Figure 11 hereinafter referred to) before the data is lost by the resetting. From the trailing edge of the pulse generated in timer 93, a reset pulse transmitted through NAND gate 94 to terminals 7 and 15 of the counter 87a, 87b after each pitch (18 of disc rotation) Circuits are also provided for energising indicator lamps to inform the operator that a sample has been correctly taken and for indicating to the operator that a sample requires o be taken.

Terminal No. 14 of counter 85b will be at "1" so long as the end around carry, i.e.

subtraction between the count, e.g. (n+l), in 84a, 84b and that (n-6) in 85a, 85h is positive. This will be the case when a sample has been taken correctly. NOR gate 85 will then be "0" at its output terminal and will turn on transistors T3 and T2 providing a current supply at terminal t33 to a "sample taken" indicator lamp, and cancelling the indicator "press and take sample".

If a sample has not been taken, the count in 84a, 84b will be equal to the count in 85a, 85b, terminal 14 of 85b will be at "0", and transistors T2 and T3 will be turned off and the "sample taken" lamp connected to terminal t33 will be turned off. The "1" present at outlet terminal of NOR gate 85 will produce a "0" outlet terminal of NAND latch 96, 97 (which is initially steered to establish a "1" at this terminal bv connection to earth of outlet terminal of NAND unit 97 through capacitor C7).

Transistors T5 and T4 will conduct and current will be supplied at terminal t34 to a further indicator lamp labelled "press to take sample".

Indicator lamps are also connected to terminals t26 ("high" tolerance on shank diameter) and t27 ("low" tolerance on shank diameter) but are not illuminated because NOR gates 90 and 91 do not draw enough current. Terminals t26, t27 are, however, connected to inputs of NAND gate 98, and, on operation of teach switch will produce "0"s at these inputs and causing a "0" to appear at outlet terminal of (invertor) NAND gate 99.Unless a "0" is present at the lower input terminal of NAND gate 97 (sample not taken) the NAND latch 96, 97 cannot set in a condition to energise transistors T4 and T5, and consequently the "press to take sample" lamp connected to terminal t34 cannot be illuminated to instruct the operator to perform this operation unless the diameter tolerance selector switches connected to terminals t26 and t27 have already been operated. Pressing the appropriate tolerance switch will light the respective lamp by grounding it.

Referring now to the circuit diagram shown in Figure 9, this relates to a logic circuit for developing reject signals in the event of shank length being short or long compared with tolerance.

The states of the two photo-sensitive elements 44 and 45, which may comprise photo-transistors, for checking shank length are shown in the table below.

Photo Photo Output element element signal 44 45 required Shank standard 0 1 pass Shank short 1 1 reject Shank long 0 0 reject In both cases photo-sensitive elements in shadow produce a "0" output and illuminated produce a "1" output.

In the trigger connected to terminal t35 is contained a TRIGGER-NAND unit 100, a NOR latch 101, 102 a TRIGGER-NAND unit 103 and a NOR gate 107, the states of the terminals t35, t40 are indicated by "0" or "1" and the symbol in brackets represents the fault condition, that is photosensitive unit 44 illuminated by reason of a "short" shank.

In the channel connected to terminal t36 there is contained TRIGGER-NAND unit 104, a NOR latch 105, 106, a TRIGGER NAND unit 108 and a NOR gate 109. In this channel again the states of the terminals t36, t41 are indicated "1" or "0", the symbol in brackets relating to fault condition that is of photo-sensitive unit 45 in shadow by reason of a "long" shank.

The NOR latches 101, 102 and 105, 106 are periodically reset by incidence of a pitch marker signal (20 p.p.r.) at terminal t39 fed to a timer 110 producing a 0.12 millisecond pulse which is fed via capacitor C2 to NOR unit Ill. The terminal states of this unit are shown as "1" or "0" with the bracketed symbol pertaining to the state immediately following the trailing edge of the pulse generated in unit 110. Accordingly the NOR latches 101, 102 and 105, 106 are reset after fault states appearing as "1" at terminals t40 or t41 have been transferred to the shift register 125 (Figure 11) hereinafter described.

Selector switches connected respectively to terminals t37 and t38 allow the operator to set the circuit to detect shank lengths shorter than standard and/or longer than standard as may be required. In the event of non-selection, the upper input terminal of NOR gate 107 and the lower input terminal 8 of NOR gate 109 remain in the "1" state holding a permanent "0" (pass) in the output terminals of the gates 107, 109, as the case may be. Terminals t37, t38 are steered with lamp indication in a manner identical to the previous case in respect to diameter sorting.

Transistors TI and T2 provide drive current to LEDs connected to terminals t42, t43 and associated respectively with the photo-sensitive elements 44, 45 to indicate a shadowed condition of the associated photo-cell.

Terminals t46 and t47 are connected to suitable positive voltage supplies.

Reference is now made to the reject device, a control circuit for which is shown in Figure 11, and to time relationships of signals generated in this circuit and shown in Figure 13.

Firstly, however, referring back to Figure 1 to 4, it will be understood that any suitable nozzle means may be provided having outlet orifices either above or below the plate 23, situated on the axis 13a of the reject station and directed to produce radial outward displacement of a blank to be rejected.

Outwardly of the periphery of the feed disc 21 a maintaining nozzle providing a counter air jet may be provided to counterbalance centrifugal forces on the screw blanks generated by rotation of the disc, this counterbalancing air jet being switched off when the reject blast is in operation by the two-way solenoid valve.

Referring firstly to the time relationships involved, as illustrated in Figure 13, section (a) shows a square wave, of which one complete cycle represents one pitch (18 of feed disc movement.

For measuring the speed of feed disc rotation, and at the beginning of each pitch, a fixed duration timer pulse is generated, e.g. 15 milliseconds, as shown in section (b), and section (c) illustrates measurement by gating of the number of encoding pulses occurring within the timing pulse in successive pitch cycles. For a constant speed disc this number will be constant but will vary (as illustrated) if disc speed varies.

In two successive cycles it is shown as being nl and n2.

Midway through the first pitch cycle the pulses nl counted in the 15 millisecond period are copied into a second counter, and simply remain there until the end of the first cycle. From the beginning of the second pitch cycle encoder pulses start to be counted into the second counter starting from the value nl already copied into it, as illustrated in section (d) of Figure 13. At some stage in the second cycle the count will reach a predetermined value, say 35 as shown, and this stage may be immediately at the beginning of the second cycle if the copied count has been 35 (corresponding to the fastest disc speed) or later (as shown) for slower disc speed. In practice the numerical value of the predetermined count is set by switches controlling a comparator so that anything equal to, or greater than, this value will produce the required output.

A third counter, of which section (f) is illustrative, counts continuously from the beginning of each delayed reject cycle and has set into it a predetermined value, for example 100 representing one pitch of electrical energisation of the solenoid valve controlling air supply to the reject device nozzle. This counter controls the blast width and the reject signal. In section (g), Figure 13, blast width termination will be seen to be coincident with this count viz 100.

Sections (h) and (1), Figure 13, show another example with the counter set to terminate blast at a count of 80 (80% pitch).

The output from the second counter comparator requires to be effectively gated with output from the shift register 125 so that a reJect signal derived from the blank which had passed through the gauging station will take effect to operate the reject device at the proper time. Provided that a reject signal illustrated in section (e) is furnished, the combined output from the comparator and the reject signal serve to bring the air blast nozzle into operation for a period equal to one pitch cycle as illustrated in section (g) of Figure 13. In practice the reject signal shown in section (e) will have been initially generated from pitch cycles earlier (corresponding to the physical angular displacement between the gauging and reject stations) but will be delivered from shift register 125 during the pitch cycle in which the reject blank concerned starts to move through the reject station.

It may, however, be desired to shorten the duration of the air blast for various reasons, for example the particular character of the articles undergoing examination, or to avoid a succeeding "pass" article being inadvertently engaged by the final part of the air blast, and this operation has already been discussed.

Referring to the circuit diagram of Figure 11, pitch marker pulses at the rate of 20 p.p.r. with a 1:1 ratio are developed in Johnson counter circuits 112, 113, of which the first is fed into terminal t48 with "1" pulses from the shaft encoder circuit of Figure 10 terminal tl6 at the rate of 2000 p.p.r. Each counter unit provides division by 10 and also receives a synchronising pulse of 1 p.p.r. from terminal t49 connected to terminal tl8, Figure 10.

NAND gates 114 and 115 used as inverters and having inputs respectively set initially to "0" and "1" sense rising and falling wave fronts. The output terminal of NAND gate 115 provides one train of "1" pulses at 20 p.p.r. and the output terminal of NAND gate 116, used as an inverter, provides another train of "1" pulses at 20 p.p.r. 180C out of phase with those delivered from NAND 115. The pulses of the train delivere i from NAND gate 116 each coincide with the beginning of a pitch cycle and those from NAND gate 115 occur at the mid points of pitch cycles.

A timer 117 which is reset at the beginning of each periodic pitch cycle by pulses from NAND gate 116 provides a pulse used as a gating signal of 15 milliseconds width which enables unit 1 18a of a first counter 118a, 1 18b at its control terminal 1. Encoding pulses 2000 p.p.r. are fed from NAND gate 120 and are stored in the first counter 118a, 1 18b for the time interval of 15 milliseconds. Such pulses are those numbered nl, n2 in section (c) of Figure 13.

At the half pitch instant the second counter 11 9a, 11 9b is enabled by a half pitch pulse delivered from NAND gate 115 to its control terminal 1 and the pulses counted in 118a, 11 8b are copied without loss from 118a, 118b into 119a, 119b.

At the beginning of the next (second) pitch cycle timer 117 is reset by a pitch pulse from NAND gate 116 and the second counter 119a, 1 19b starts to store encoder pulses at the rate of 2000 p.p.r. These encoder pulses are added to those already stored in 119a, 1 19h and transferred continuously to the comparator 120a, 120b which is preset, by manually operable switch 121 a, 121b, to deliver an output signal at 12 at a predetermined count, e.g.

35.

This count is equal to the number of pulses at the fastest disc speed contained in the time period set by timer 117, which in turn is equal to the delay time for the air blast to reach an article at the reject station after the ejector device has been electrically energised (assuming no delay in the electrical system). Account can also be taken of any angular displacement of the axis 13 of the reject device from the exact multiple (four) of pitches between the gauging and reject stations to provide the requisite advance or delay necessary to achieve impingement of air blast on the reject article immediately it reaches the reject rotation.

In the particular example now described the fastest disc speed is assumed to be 70 r.p.m. equivalent to passage of 1400 blanks per minute past the reject station, and the delay period of the solenoid is assumed to be 15 milliseconds. For 100 encoder pulses per pitch cycle (2000 p.p.r.) the preset count to be set on switch 121a, l 21 b is given by the relation:: solenoid delay pitch transit time at max disc speed 15 0.4285 =35 For slower speed disc speeds the count stored in counter 1 l8a, 1 18b and copied into counter 119a, 119b will be lower and consequently when counter 119a, 119b is enabled to take additional encoder pulses during the second pitch cycle it will take longer for this counter 119a, 119b to reach the preset number, e.g. 35, and coincidence between the counts registered in 1 19a, 119b and 120a, 120b will occur later in the second cycle.

At the instant at which such coincidence takes place an output signal is developed at terminal No. 12 of comparator 120b and is fed through inverter 122 to the lower input terminal of NAND 123. If a reject signal had been present at any of terminals t50 to t53 (the first two being diameter reject signals and the last two shank length reject signals) it would be entered into the shift register 125 during the pitch in which it was generated after passing through OR gate 124 and would be presented at the outlet 22 of the shift register 125 four pitch cycles later. It is required to be presented at the upper input terminal of NAND gate 123 during but not earlier than the beginning of this fourth pitch cycle, so that the reject device can be energised at least half a pitch before the beginning of the fifth pitch cycle.

This condition is ensured by operation of NAND gate 126, the lower input terminal of which receives a pitch marker at the beginning of each latch period from NAND gate 116. Consequently NAND latch 127, 128 will be set at that instant (subject to there being a reject signal at output terminal 22 of the shift register 125) to produce a "1" at the upper input terminal of NAND gate 123.

This will then deliver an output from its output terminal earlier than the beginning of the fifth pitch cycle by an amount determined by disc speed, but not earlier than the beginning of the fourth pitch cycle.

The sequence of events is as follows: At the fourth pitch marker if there is a reject in shift register 125 a reject pulse emerges from gate 126 which sets latch 127,128 and this enables gate 123. An output from comparator 120a, 1 20b (with delay as appropriate) will (a) reset latch 135, 136 stopping counter 119a, l19b, (b) reset latch 127, 128, through inverter 122, (c) set latch 129, 130 which releases the reset on counter 131a, 131b allowing it to count encoder pulses, (d) initiates blast signal at t54.

The counter 131a, 131b and associated switch can be set to provide energisation of the solenoid valve for a full pitch period from the onset of blast energisation or for any desired percentage of such period.

Encoder pulses are fed to terminal 10 of counter 131a, 131b from NAND gate 120 (subject to the absence of an inhibit signal at terminal t49a), and, when the count registered equals that determined by the setting of switch 132a, 132b all inputs of NAND gate 134 have the same level "1" producing an output "0" at its output terminal which resets NAND latch 129, 130 rests same to its original condition replacing the energisation signal at terminal t54 by a "0". Counter 13 it, 131 b is itself reset at terminal 15 at the beginning of each reject energisation signal from terminals 2 and 4 of NAND latch 129, 130.

NAND latch 135, 136 performs the function of enabling the counter 119a, 119b at terminal 5 to count at the beginning of each pitch cycle, and receives its controlling input from NAND gate 116 at the beginning of each pitch cycle via NAND gate 137 acting as an inverter. Latch 135, 136 is reset to stop the second counter 1 19a, 1 19b by the parity signal from the comparator 120b terminal 12.

Terminal t55 provides a pitch marker output to other circuits already referred to.

Terminal t56 provides a driver pulse to a reject counter.

Reference is now made to Figure 12 showing the circuit providing anti-jamming facilities.

Encoder pulses 4000 p.p.r. are fed in at terminal t60 from the appropriate terminal tls, Figure 10, and are fed to terminal No.

11 of a retriggerable monostable circuit unit 140. The output at terminal No. 10 thereof remains at "1" so long as the feed disc continues to rotate and the encoding pulses are generated, but the unit detects a "missing" encoder pulse at terminal 10 and will then go to "0" upon failure to retrigger.

In general, the states of certain of the terminals of the units incorporated in Figure 11 are shown on the circuit by a figure 1 or 0 representing normal operation, i.e. there being no slowing or jamming of the feed disc by a blank trapped mechanically between the feed disc and some stationary part of the apparatus, and by a figure (1) or (0) when such a jammed condition occurs. Any variation in this convention will be referred to in the following description.

Upon occurrence of a jam, NAND unit 141 starts timer unit 142, terminal 9 of which then immediately goes to "0" and deenergises transistors T3, T4 and cuts off current to a motor drive relay connected to terminal t6 1. This action wi 1 also occur if the jammed condition, instead of producing a complete stoppage of the disc, has set up an obstruction which causes the feed disc to rotate slowly (for example at about half the speed of the lowest setting of feed rate).

Concurrently with the incidence of a "0" output from the "missing" pulse detector 140, NAND latch 146, 147 will have changed over, causing transistors Tl, T2 to energise and reverse the motor connections via a relay connected to terminal t62 (the motor having been de-energised by the "0" output from terminal 9 of the timer unit 142 to transistors T3, T4).

The delay period timed by unit 142 is selected to be sufficient, before it times out, to allow the motor, once de-energised, to come to rest, and on timing out, terminal No. 10 of unit 142 reverts to "1" and via NAND gate, 144 feeds a "0" input to terminal No. 4 of reverse timer unit 145 which starts then to time out.

Also, concurrently with generation of a "0" signal at terminal No. 10 of the "missing" pulse detector 140, NAND latch 148, 149 changes over and provides inhibit signals at terminals t63 "0", t64 "1". The former inhibits operation of the solenoid for vibrating the first part 15 of the feed means already mentioned. The latter provides an inhibit signal to terminal t70 (Figure 11) to inhibit development of half pitch pulses by holding terminal No. 13 of counter 112 at "I".

When timer unit 142 times out and starts the reverse timer unit 145, terminal No. 9 of timer unit 142 reverts to the "1" state and re-energises transistors T3, T4 and hence establishes motor drive so that the disc starts to be driven in the reverse direction.

Concurrently the output from terminal 10 of the timer unit 142 causes a "1" output to appear at the lower input terminal of NAND gate 143 while its upper input terminal is held at "0" by the output from latch 146, 147. A "1" output is thus applied from NAND gate 143 to input terminal 10 of counter 147a, 147b to start this counting "down", encoder pulses being received at terminal 15 due to reversed rotation of the feed disc.

When counter unit 147a, 147b reaches a predetermined "down" count in the reverse direction, e.g. 16 (a "1" at pin No. 6 of unit 147a causes reverse timer unit 145 to revert to its initial state providing a "1" at terminal No. 6).

This resets NAND latch 146, 147 thereby reversing the motor connections via terminal t62 and simultaneously removes drive (energisation) to the motor at terminal t61 via NAND gate 141 and timer unit 142.

When the latter has again timed out drive (energisation) is re-established at terminal t61 by reversion of terminal No. 9 of the timer unit 142 to the "1" state.

If therc. is a jam in the reverse direction timer unit 145 overrides counter unit 147a and a "1" is established at terminal No. 6 of unit 145 after a predetermined delay and restoration of forward drive takes place as already described through resetting of NAN latch 146, 147.

Concurrently with the motor being reenergised in the forward direction, NAND gate 143 receives two "1" inputs one from latch 146, 147, and one from terminal 9 of timer unit 142, and counter unit 147a, 147b then receives a "0" at terminal 10 from the output terminal of NAND gate 143 and is thereby set up to count in the direction (thereby cancelling the "down" count which will have taken place during reverse rotation of the motor and feed disc) and when the original zero position is reached, corresponding to the point at which the jam occurred, a zero count is recorded. NOR gate 150 then reverts to a "I" at terminal No. 13. This resets NAND latch 148, 149 via inverter 151 and cancels the inhibit signals at terminals t63, t64 so that "diameter count" is able to be resumed, assuming that the jammed condition is cleared.

It should be noted that the direction of up/down counter 147a, 147b is changed only after the rest period, provided by counter 142, when the disc direction is in the process of being changed. This is accomplished by NAND gate 143 and ensures that any runon of the disc particularly in reverse, with the motor de-energised, will be recorded in counter 147a, 147b correctly sensed.

If there is no clearance of the jammed condition the above described sequence is repeated.

Each time repetition takes place, shift register 152 is stepped on by the output from terminal No. 10 of NAND gate 143.

Switch 153 can be set manually to provide any number of "attempts" at clearance up to 4.

In the circuit diagram the switch is shown set to terminal No. 2. In this case, after the second attempt, terminal No. 2 will go "1" and through inverter 154 and diode D4 a "0" will be supplied to terminal t66.

Terminals t65 and t66 are connected to push-button "on" and "off" switches. t65 normally has a "1" condition and goes "0" when its "on" switch is pressed and t66 normally has a "1" condition and goes "0" when the "off" switch is pressed.

NAND latch 155, 156 controls the energisation of transistors T5 and T6 and hence terminal t67 which is connected to a high duty relay or contactor controlling or forming the main isolator which will switch the apparatus "on" and "off". Accordingly a "0" at the outlet terminal of inverter 154 will bring about switching "off" of the apparatus. Diode D4 prevents a short circuit on this inverter in the event of the manual "off" switch connected to terminal t66 being operated.

The shift register 152 is reset on each revolution of the feed disc by a synchronising pulse fed to the upper input terminal of NAND gate 157. At start up, the normal state of terminal t65 is "1" as is also that of terminal t68 so that a reset pulse (1) will be delivered from the output terminal of NAND gate 157 through shift register 152 during each revolution of the feed disc, and upon depressing the start button.

The speed at which articles are fed by the first part of the feed means, and by the feed disc, is selected to optimise feeding efficiency, i.e. to reduce to a minimum the reversals of the feed disc. In general, as the lengths, widths, or weights of screws or blanks increases the optimum feeding speed is reduced.

Referring now to Figure 14, this shows a portion of a screen 210 corresponding to the screen 35 of the embodiment thus far described, but the photo transistor 41 providing the gating signal and circuits generating 4000 p.p.r. encoder signals from the shaft encoder are effectively replaced by an array of photo transistors comprising a first group 211 followed by a number of successive groups, four of these 212, 213, 214, 215 being shown by way of example.

A portion of the image 216 of the shank of an article such as a screw blank, of which the shank diameter is required to be measured, is shown travelling across the screen in the direction of the arrow 217. The array of photo transistors may conveniently be mounted at the rear or underside of the screen and will, when in shadow and within the area of the shank image, provide a "1" output, and when illuminated will provide a It Output. appreciated that it would be It will be appreciated that it wouId be possible to provide an array of photo transistors at equal spacings continuously across the screen between the extreme lefthand and extreme right-hand position, and to generate the first encoder signal and the start of the gating signal when the extreme left-hand photo transistor such as CO were masked, and generate the end of the gating signal by the unmasking of the photo transistor CO so as to provide a count of encoder signals within that gating interval.

Such a method, although within the scope of the invention, would require the provision of photo transistors at positions spaced apart uniformly across the entire width of the screen at pitch intervals such as those shown for the group 211.

The requisite generation of encoder signals and effective gating signals can, however, be achieved more economically by the arrangement shown.

The group 211 contains a predetermined number of photo transistors (16 in the present case) at pitch spacings determined by the accuracy of the measurement required (in the present case typically 0.005 inches), and gating and generation of encoder signals starts simultaneously when the extreme left-hand photo transistor or the control cell CO is uncovered as a result of passage of the image 216 to the right.

The encoder signal count then generated by the associated circuit shown in Figure 15 is then equal to the number of intervals between the photo transistors of the array 211, namely 16, and the encoder count is actually a count down, namely 16, 15, 14 etc.

The function of the cell groups 212 to 215 is in effect to generate a "take count" signal which corresponds to the end of the gate, and the encoder signal then subsisting will represent the diameter of the shank.

Were only single "take signal" cells A, B C, D is included in each of the groups, the accuracy of diameter determination would then be equal to the pitch spacing between two successive cells of the group 211, e.g.

between cell No. 11 and cell No. 10.

However, a further cell is provided in each group 212 to 215, namely Al to Dl, which is spaced from its companion cell A to D by half the pitch between the cells of the group 11.

When the cells such as A, Al or B, B1 are successively masked, the "take reading" signal effectively consists of two signals which can conveniently be referred to as primary and secondary "take reading" signals.

If, in the interval between the primary and secondary "take reading" signals, no further cell in the group 211 has been unmasked, then this signifies that at the instant at which the primary "take reading" signal occurred, e.g. cell B was masked, the trailing edge of the image 216 had travelled less than half the distance from cell 11 and cell 10, and so an encoder count corresponding to cell 11 being unmasked is the closest approximation to the blank diameter.If, on the other hand, at the instant at which the secondary "take reading" signal is generated, i.e. cell Bl is just masked, and a further encoder signal has been generated by unmasking of cell 10 in the group 211, this would indicate that the trailing edge of the image 216 had travelled more than half the distance from cell 11 to cell 10 at the time at which the primary "take reading" signal was generated, and, therefore, an encoder count corresponding to the unmasking of cell 10 would represent the best approximation of shank diameter.

In effect, therefore, the provision of the "take reading" cells A, Al, to D, Dl each at a half pitch spacing from each other improves the accuracy by a factor of 2.

Moreover, the arrangement illustrated requires only an array of closely spaced cells to be provided in respect of the group 211 since incidence of a "take reading" signal generated by cells B, Bl or C, Cl can provide for the addition of an encoder count equal to that produced by the whole of the group 211, or a multiple thereof, plus the encoder signal (down to which a count has been achieved) in respect of the group 211.

In the example illustrated, the count would be 16 plus 11. If cell 10 had been unmasked before cell B1 were masked, an additional "+" count put into the counting system concurrently with unmasking of the control cell CO would be cancelled, whereas, if cell B I were masked before cell 10 were unmasked, the additional "T" count put in by the unmasking of the control cell CO at the beginning of count down would be left in to be recorded in the system as hereinafter described.

Thus, the following table illustrates typical determination of blank diameter.

It will be understood that although for identification purposes a "+" count is referred to, it is actually a count of "1", and aggregates, if present, to the subsisting total, or not if not present, so that the resulting whole number is the nearest approximation to the decimal that would actually represent the dimension measured.

Count at primary and Whether "+" Encoder Possible Blank secondary signal left signal Diameter='D' "take reading" in or reading (figures to be signals cancelled multiplied by p/m) B (I) B1 (2) (3) (4)=1+3 (5) 16+11 16+11 left in 27+ 28.0 > D > 27.5 16+11 16+10 cancelled 27 27.5 > D > 27.0 16+ 10 16+10 left in 26T 27.0 > D > 26.5 16+10 16+9 cancelled 26 26.5 > D > 26.0 where p is the pitch between cells in group 211.

where m is the magnification of the optical system.

Referring now to Figure 16, the control cell CO produces a "0" when unmasked by the trailing edge of the image 216 and this passes through inverter 218, R < network 219, and inverter 220. This produces a "0" pulse at terminal 12 of NOR gate 221 and a "1" output at its output terminal unless an inhibit signal applied to terminal t201 is present at its upper input terminal.

An output signal from NOR 221 sets data latch 22 (which is a D type flip-flop circuit). At the same time the "T" count is set in to data latch 222 and will appear in the form of a "1" signal on terminal t202 (representing 2-1 data).

Cells in the group 211 Nos. 15 to 1 then successively provide outputs "0" which are sup lied to encoders 223, 224 and via exclusive OR gates appear as a binary coded count on terminals t203 to t206 (connected respectively to the 20 to 23 lines of the 4-bit output).

For simplicity only the "take reading" cells D, Dl are shown. Readings, however, from the cell group A to D are fed to encoder 228 and thence, in the case of some outputs therefrom, via OR gates 229 and 230 to encoder 231 providing outputs at terminals t207 to t209. These three terminals are connected to lines representing counts of 24, 25, 26, the lowest of these being thus 16 representing the count to be added when a primary "take count" signal is generated by cell B.

Outputs of the "take reading" cells Al to Dl are fed via resistor, capacitor, and diode network 232 to a NOR gate latch 233 which provides a "1" output in the interval between primary and secondary "take reading" signals and is fed to terminal No. 9 of the data latch 222. If during this interval the 20 line connected to terminal t203 provides an output (an additional cell in the group 211 is uncovered) then output terminal of exclusive OR gate 234 produces a "I" signal which is fed to the data latch 222 via OR gate 235 and will then cancel the "t" count signal established on terminal t202 at the time at which control cell CO was unmasked.

The data latch 222 provide, in effect, the gating signal on terminal t210 which begins when the control cell CO is unmasked and ends with incidence of a "take reading" signal from one of the cells in the group A to D.

Referring now to Figure 16, the general manner of operation is that encoded count data is fed in continuously as it occurs at terminals t21 1 to t217 connected respectively to terminals t203 to t209 (Figure 15), and is applied to the input terminals at the lower side of counter unit 236 and to the inputs at the left-hand side of counter units 237, 238.

Counter unit 236 registers a value at its output terminals at its upper side which changes only between incidence of the control signal from cell CO and the primary "take reading" signal from one of the cells A to D, this being achieved by feed-in of the "gating" signal at terminal t218 connected to terminal t210 (Figure 15). Upon incidence of the primary "take reading" signal, the count appearing at the output terminals of 236 is compared with a preset high count (which may be correct value plus I) set into comparator units 239, 240 and with a predetermined low count (which may be correct -value minus 2) set into comparators 241, 242 as a result of feeding a test sample (known to be to specification as regards the diameter of its shank) through the apparatus.

For this purpose the operator presses a "take sample" switch applying a momentary "1" signal at terminal t219 and the input of inverter 243, thereby setting data latch units 244 and 245 and a 6mS timer 246. A fault interlock connection is provided so that if one or the other of the high or low reject switches connected to terminals 220, 221, and which are held mechanically closed when operated, are not operative this prevents the taking of a sample under these conditions. The interlock connection is made between the output terminal of AND gate 247 and terminal 5 of data latch unit 244. Terminal 222 is connected to an indicator lamp circuit signalling "press and take sample" which is extinguished when data is presented to the adder units 254 and 255 as the sample is measured, such signal being applied via OR gate 248 and NAND latch 249.The same signal produces an output on terminal t223 "sample taken".

With the data latch 243, 244, 245 set, counters 237 and 238 are cleared by the signal at terminal CLR from output terminal of NAND gate 250, this being effected by the input at the lower input terminal of gate 250 from t219. The gating signal incident at terminal t18 when the control cell is uncovered by the trailing edge of the sample image enables counter 237 via NAND gates 245 and 251 and a count is registered at the input terminals of counters 237 and 238 in consequence of the inputs to terminals t211 to t217.

The purpose of the delay timer 246 is to increment counters 237 and 238 by a count of 1 and to latch the + information, if present at terminal 225, via data latch 252 to the comparator 239 and is frozen until the next take sample reading sequence is completed.

The outputs of counters 237 and 238 are applied to the upper set of terminals of comparator units 239 and 240 and this value is frozen until a clear signal is applied to counters 237 and 238. This occurs when the "press and take sample" button is depressed, producing an input T219 or via a "power on reset" signal is applied to T228.

These signals also clear counter 236.

The count output of value +1 is also applied to the inputs at the upper sides of adder units 254, 255 which perform a subtraction function, the count to be subtracted being applied by a switch unit 256. The example in figure 16 shows a value of 6 (corresponding to three pitches between cells of the group 211).

Upon comparison between the count value frozen for each article examined at the output terminals of counter unit 236 with the count values set in to comparators 239 and 240 and comparators 241, 242, reject signals will be generated at terminal 13 of the former and terminal 12 of the latter if the sample is respectively over-sized by a count of 1 or more or undersized by a count of at least 2.

Such reject signals are incident respectively at inputs to OR gates 257 and 258 and produce reject signals at terminals t226 and t227 respectively and will change during the period between the control cell being uncovered and "take reading" signal which then freezes their state, but this is immaterial since all the reset signals are gated with the 20 PPR signal for entering into the shift register.

"+" count information is supplied from terminal t225 to the inputs of comparators 239 and 241 unless cancelled under the conditions previously described, so that acceptance or rejection of the particular article takes place in the improved conditions of accuracy resulting from production of this signal.

In the further embodiment of the invention described with reference to Figures 14 to 16 various consequential modifications require to be applied to other circuits previously described in connection with the first embodiment when used with the further embodiment.

These will be capable of being made without difficulty by persons skilled in the art but the following are nevertheless mentioned.

In the circuits of the first embodiment requiring 4000 p.p.r. input of encoder signals, such input will no longer be derived from the shaft encoder for the purposes of diameter count but pulses of lower frequency generated by the shaft encoder will be used as before.

The single adjustably mounted photo transistor provided in association with the screen of which the image of the article under measurement is displayed, and which was formerly utilised for providing a gating signal, will no longer be required since such gating signal is derived in the array of photo transistors above described with reference to Figure 14.

A further modification which may be applied to either embodiment is that a reject station and pass station 14 may be inter changed, nozzle means being utilised to remove pass articles instead of reject articles from the slots of the rotary disc, and appropriate modifications being made in the circuit of Figure 11 energise the nozzle solenoid for each article except those in respect of which a reject signal is generated.

The provision of the anti-jamming means herein referred to is the subject of our copending application 8103652 (Serial No.

1604842) divided from the present application.

WHAT WE CLAIM IS: 1. A method of determining a dimension of an article comprising the steps of: a. scanning the article to generate gating signals representative of the boundaries of said dimension, b. generating a succession of electrical encoder signals representative of respective increments of said dimension, c. gating the encoder signals by the gating signals and determining the number of encoder signals gated, d. storing a number representative of the encoder signals which would pertain to an article having a reject value of said dimension, e. comparing the number of encoder signals determined by said gating for the article undergoing measurement, with said stored number, and f. developing an output signal in response to said comparison.

2. A method according to claim 1 wherein: a. a pass article having a dimension equal to that required is subjected to said scanning and said encoder signals are generated as aforesaid, b. the resultant number of encoder signals is stored modified by the addition and/or subtraction of a predetermined number, representing the limit of dimensional tolerance, to produce a high tolerance number and/or low tolerance number, c. comparison is then effected between said high and/or low tolerance stored number and the number of encoder signals resulting from said gating in respect of subsequent articles subjected to the method.

3. A method according to claim 2 wherein: a. the high and low tolerance numbers of signals are stored in respective stores, b. the number stored in each of the stores is compared with the number of encoder signals resulting from said gating in respect ot each of said subsequent articles to generate either high reject, or low reject, signals when said articles are not within the limits of tolerance.

4. A method according to any one of claims 1 to 3 wherein: a. the encoder signals are generated by a shaft encoder having relatively rotatable elements, b. the gating signal is generated by photo

**WARNING** end of DESC field may overlap start of CLMS **.

Claims (32)

1. **WARNING** start of CLMS field may overlap end of DESC **.

   of 6 (corresponding to three pitches between cells of the group 211).

   Upon comparison between the count value frozen for each article examined at the output terminals of counter unit 236 with the count values set in to comparators 239 and 240 and comparators 241, 242, reject signals will be generated at terminal 13 of the former and terminal 12 of the latter if the sample is respectively over-sized by a count of 1 or more or undersized by a count of at least 2.

   Such reject signals are incident respectively at inputs to OR gates 257 and 258 and produce reject signals at terminals t226 and t227 respectively and will change during the period between the control cell being uncovered and "take reading" signal which then freezes their state, but this is immaterial since all the reset signals are gated with the 20 PPR signal for entering into the shift register.

   "+" count information is supplied from terminal t225 to the inputs of comparators 239 and 241 unless cancelled under the conditions previously described, so that acceptance or rejection of the particular article takes place in the improved conditions of accuracy resulting from production of this signal.

   In the further embodiment of the invention described with reference to Figures 14 to 16 various consequential modifications require to be applied to other circuits previously described in connection with the first embodiment when used with the further embodiment.

   These will be capable of being made without difficulty by persons skilled in the art but the following are nevertheless mentioned.

   In the circuits of the first embodiment requiring 4000 p.p.r. input of encoder signals, such input will no longer be derived from the shaft encoder for the purposes of diameter count but pulses of lower frequency generated by the shaft encoder will be used as before.

   The single adjustably mounted photo transistor provided in association with the screen of which the image of the article under measurement is displayed, and which was formerly utilised for providing a gating signal, will no longer be required since such gating signal is derived in the array of photo transistors above described with reference to Figure 14.

   A further modification which may be applied to either embodiment is that a reject station and pass station 14 may be inter changed, nozzle means being utilised to remove pass articles instead of reject articles from the slots of the rotary disc, and appropriate modifications being made in the circuit of Figure 11 energise the nozzle solenoid for each article except those in respect of which a reject signal is generated.

   The provision of the anti-jamming means herein referred to is the subject of our copending application 8103652 (Serial No.

   1604842) divided from the present application.

   WHAT WE CLAIM IS: 1. A method of determining a dimension of an article comprising the steps of: a. scanning the article to generate gating signals representative of the boundaries of said dimension, b. generating a succession of electrical encoder signals representative of respective increments of said dimension, c. gating the encoder signals by the gating signals and determining the number of encoder signals gated, d. storing a number representative of the encoder signals which would pertain to an article having a reject value of said dimension, e. comparing the number of encoder signals determined by said gating for the article undergoing measurement, with said stored number, and f. developing an output signal in response to said comparison.
2. 2. A method according to claim 1 wherein: a. a pass article having a dimension equal to that required is subjected to said scanning and said encoder signals are generated as aforesaid, b. the resultant number of encoder signals is stored modified by the addition and/or subtraction of a predetermined number, representing the limit of dimensional tolerance, to produce a high tolerance number and/or low tolerance number, c. comparison is then effected between said high and/or low tolerance stored number and the number of encoder signals resulting from said gating in respect of subsequent articles subjected to the method.
3. 3. A method according to claim 2 wherein: a. the high and low tolerance numbers of signals are stored in respective stores, b. the number stored in each of the stores is compared with the number of encoder signals resulting from said gating in respect ot each of said subsequent articles to generate either high reject, or low reject, signals when said articles are not within the limits of tolerance.
4. 4. A method according to any one of claims 1 to 3 wherein: a. the encoder signals are generated by a shaft encoder having relatively rotatable elements, b. the gating signal is generated by photo

   electric means subjected to relative scan by the article, c. the relative rotation of the shaft encoder elements, and the relative scanning movement, are coordinated to establish that an increment of one of said relative movements is always accompanied by an increment of the other of said movements maintaining a constant ratio of magnitudes of these increments.
5. 5. A method according to any one of claims 1 to 3 wherein: a. the encoder signals are generated by photo electric means subjected to relative scan by the article, b. the gating signals are generated by photo electric means also subjected to relative scan by the article.
6. 6. A method according to claim 5 wherein: a. the photo electric means is subjected to different degrees of illumination (one of which may be zero) by relative traverse of the photo electric means by an image of the article, b. the scanning is produced by scanning an array of photo electric cells extending with the length or major dimension of the array along the path of travel of the image, c. the array of cells including a group of closely spaced cells for generating said encoder signals, d. said array also including at least one cell situated at an end of said group for generating the gating signal.
7. 7. A method according to claim 6 wherein generation of the gating signals is performed by cells at opposite ends of said group.
8. 8. A method according to claim 7 wherein generation of the gating signals is performed by a cell at that end of said group at which the image is first incident, and by any one of a plurality of further cells situated respectively at the other end of said group and at distances beyond said other end of said group which are multiples of the length of said group.
9. 9. An apparatus for determining a dimension of an article comprising: a. a body or supporting structure affording a feed path for the article, the feed path passing through a station (herein called the gauging station) at which the dimension of the article is to be determined, b. feed means for advancing the article through the gauging station in an orientation such that opposite boundaries of the dimension to be determined are spaced apart along the feed path, and drive means for the feed means, c. encoder means for generating a succession of electrical pulses representing increments of travel along the feed path and occurring in respect of the article at least over a distance in the region of the gauging station, d. sensing means responsive to passage of the article through the gauging station to provide electrical boundary signals representing respectively opposite boundaries of the dimension to be determined, and e. electrical circuit means (herein called the measuring circuit) associated operatively with the encoder means and the sensing means and including, means to receive said pulses and said boundary signals and to operate arithmetically thereon to determine the number of pulses occurring between said boundary signals, means for storing a number representative of encoder pulses which would pertain to an article having a reject value of said dimension, comparator means for comparing the number of pulses produced by the article undergoing measurement with the number stored, and means for developing an output signal in response to said comparison.
10. 10. Apparatus according to claim 9 wherein the electrical store means provides for storing a predetermined variable number of encoder pulses representing high and low limits of tolerance of the size of articles to be measured.
11. 11. Apparatus according to either of claims 9 and 10 wherein: a. the measuring circuit includes means for storing in the store means a number of pulses produced by measurement of a pass article having the required dimension, b. associated with the store means is a means for effecting addition to and/or subtraction from the number stored therein, c. the comparator means is so connected or arranged as to effect comparison between, on the one hand the aggregate number stored in the store means resulting from operation of the means referred to in subclauses a. and b., and on the other hand the number of pulses occurring between the boundary signals of an article undergoing measurement.
12. 12. According to claim 11 wherein: a. the store means includes irrespective stores for storage of a high tolerance number representing the dimension of an oversized article to be rejected, and for storage of a low tolerance number representing the dimension of an undersized article to be rejected, b. the comparator means includes comparators associated respectively with these stores and generating reject signals upon coincidence between the number of pulses occurring between the boundary signals and fed into the comparators and the number stored in the stores.
13. 13. Apparatus according to any one of claims 9 to 12 wherein the encoder means comprises a rotary shaft encoder which is driven in timed relation with the feed means so that each increment of rotation of the shaft encoder represents a corresponding distance travelled along the feed path for the article undergoing dimension determination at the gauging station.
14. 14. Apparatus according to any one of claims 9 to 13 wherein the feed means comprises a rotary feed member having article receiving formations spaced apart peripherally around its axis of rotation and serving to advance the articles through the gauging station.
15. 15. Apparatus according to claim 13 and 14 wherein the rotary feed member is connected to a rotary input element of the shaft encoder directly.
16. 16. Apparatus according to any one of claims 9 to 12, or to claim 14 as appendant directly thereto, wherein: a. the encoder means comprises an array of photo electric cells extending with the length, or major dimension, of the array along the feed path in the region of the gauging station, b. said array of cells includes a group of closely spaced cells for generating the encoder signals, c. said sensing means comprises at least one photo electric cell of the array for generating said boundary signals.
17. 17. Apparatus according to claim 16 wherein cells for generating said boundary signals are situated respectively at opposite ends of said closely spaced group.
18. 18. Apparatus according to claim 17 wherein further boundary signal generating cells are provided at intervals along said feed path beyond the downstream end (having regard to the direction of feed) of said group at spacings which are multiples of the length of said group.
19. 19. Apparatus according to either of claims 17 and 18 wherein: a. the first boundary signal is generated by means responsive to unmasking of the boundary cell at the leading end of said group, b. the second boundary signal by means responsive to the masking of the next succeeding boundary cell to be masked following said unmasking.
20. 20. Apparatus according to any one of claims 17 to 19 wherein: a. at each boundary cell position downstream of the first, a pair of boundary cells are provided spaced apart by half the spacing of the cells of said group to generate successive primary and secondary boundary signals in addition to an initial boundary or control signal generated by the boundary cell at the upstream end of said group, b. means are provided in the electrical circuit for selecting a number of encoder pulses between the first occurring and the next occurring boundary signals and registered consequent upon said arithmetical operations, such number corresponding to the nth encoder pulse or to the (n+l) encoder pulse according to whether an encoder pulse does or does not occur between the primary and secondary boundary signals.
21. 21. Apparatus according to any one of claims 9 to 20 wherein the sensing means comprises photo electric means so mounted on the body or supporting structure of the apparatus as to be subjected to a travelling image of a profile of the article, the profile containing the dimension to be determined.
22. 22. Apparatus according to claim 21 wherein optical means are provided for projecting the image onto the photo electric means (whether provided for encoder pulse generation or for boundary signal generation) with magnification of the dimension.
23. 23. Apparatus according to any one of claims 9 to 22 wherein the apparatus includes: a. sorting means operable in response to a reject signal derived from the output signal of the measuring circuit to separate articles for rejection (herein called reject articles) which are outside the limit or limits of tolerance of the dimension of articles to be passed (herein called pass articles), from the pass articles said sorting means being operable in response to presence or absence of the reject signal but after a delay dependent upon operational characteristics of said sorting means, b. means for varying the speed at which the articles advance along the feed path, and c. compensating means for bringing the sorting means into operation to effect displacement of a selected article at a sorting station situated at a substantially fixed position along the feed path from the gauging station, so that the selected article is properly operated upon by the sorting means irrespective of speed variation of the feed means.
24. 24. Apparatus according to claim 23 wherein the sorting means includes an electricaíly energised operating means controlled by a control circuit which is connected to receive a speed signal representing speed of advancement of the articles, and includes a sub-circuit or circuit element presenting an electrical parameter representative of the delay time of the sorting means, arithmetical circuits also being provided to operate on the speed signal and delay time parameters to develop an output signal for initiating operation of the sorting means at such time, in advance of arrival of the selected article at the sorting station, as will ensure proper operation thereon by the sorting means.
25. 25. Apparatus according to any one of claims 9 to 24 wherein the apparatus includes anti-jamming means comprising a circuit (herein called the anti-jamming circuit) for temporarily reversing forward movement of the drive means in response to the occurrence of a jammed condition and thereafter restoring said drive means to forward movement.
26. 26. Apparatus according to claim 25 wherein the measuring circuit includes inhibiting means for preventing a false count of encoder pulses occurring during the interval of reversed and restored forward drive of the feed means until the latter again attains the position at which the jammed condition initially took place.
27. 27. Apparatus according to either of claims 25 and 26 wherein the anti-jamming circuit includes means for temporarily reversing and restoring the drive means to forward operation a predetermined number of times to attempt to clear the jammed condition.
28. 28. Apparatus according to claim 27 further including a shut-down circuit responsive to attainment of said predetermined number of reversals and restorations and failure to clear the jammed condition to shut down operation of the apparatus. ~~~~~~~~~~~~~~~~~~~
29. 29. A method of determining the dimension of an article substantially as herein described with reference to and as illustrated by Figures 1 to 13 of the drawings.
30. 30. A method of determining the dimension of an article substantially as herein described with reference to and as illustrated in Figures 14 to 16 of the accompanying drawings.
31. 31. Apparatus for determining the dimension of an article substantially as herein described with reference to and as shown in Figures 1 to 11 of the drawings.
32. 32. Apparatus for determining the dimension of an article substantially as herein described with reference to and as shown in Figures 14 to 16 of the accompanying drawings.