US 3646353 A
A flying spot scanner is employed to inspect a web which rides over a supporting surface. Optical signals are produced at the edges of the web, and at the edges of the surface; and optical web defect signals are produced while scanning the web. To blank signals which are not web dependent, the invention proposes to split scan-produced signals into first and second halves thereof, whereby web-edge defining signals may be isolated, and employed to operate a blanking circuit. To prevent the registering of defects which are proximate the edges of the web, each pair of blank-defining signals produced during a given scan are shifted timewise to a subsequent scan, one of each such pair being shifted a little more, and the other of each such pair being shifted a little less, than a scan duration.
Claims available in
Description (OCR text may contain errors)
United States Patent Bhullar et a1.
 FLYING SPOT SCANNER BLANKING  Inventors: Pushpinder S. Bhullar; Robert A. Wright,
both of Rochester, NY.
 Assignee: Eastman Kodak Company, Rochester,
 Filed: Oct. 19,1970
 Appl. No.: 81,969
[ Feb. 29, 1972 Primary Examiner-Ronald L. Wibert Assistant ExaminerEdward S. Bauer At!orneyWalter O. Hodsdon and Robert F. Cody  ABSTRACT A flying spot scanner is employed to inspect a web which rides over a supporting surface. Optical signals are produced at the edges of the web, and at the edges of the surface; and optical web defect signals are produced while scanning the web. To blank signals which are not web dependent, the invention proposes to split scan-produced signals into first and second halves thereof, whereby web-edge defining signals may be iso lated, and employed to operate a blanking circuit. To prevent the registering of defects which are proximate the edges of the web, each pair of blank-defining signals produced during a given scan are shifted timewise to a subsequent scan, one of each such pair being shifted a little more, and the other of each such pair being shifted a little less, than a scan duration.
6Claims,8Drawing Figures 20 22 24 A DEFECT REG. 7 USING C/(T BLANK CKT Qo e1:
'PATENTEDFEBZS I972 3,646,353
SHEET 1 [1F 3 A TTOR/VE VS RT m @E l a? 3 ww all .2 an m ,@\.u 4m J 57W C wv i w w QEQEMQ m m 1 mw Q \A a kwwm m= WW I l AT w Q m R i .2 4 0 EB M Ex q m J WK mEPfiwQ w Q x a 396 Q Q 5% b9 @E J 5}: Q q u x w .9 m Wwm 9 mt mw/ 55 m q Q 11 355m H N y .WQQIQ Q Tl l PATENTEUFEB29 m2 SHEET 2 OF 3 A TTORNEYS FLYING SPOT SCANNER BLANKING BACKGROUND OF THE INVENTION 1. Field of the Invention This invention relates in general to flying spot scanners, and in particular to such scanners as are useful for inspecting webs and the like.
2. Description Relative to the Prior Art Consider a web which, while being subject to the spot of a flying spot scanner, passes over a wide web-supporting surface. Modulation of the spot occurs as the spot (1) first passes onto the web-supporting surface, (2) passes from such surface onto the edge of the web, (3) passes over defects in the web, (4) passes from the edge of the web back onto the web-supporting surface, and (5) moves off the web-supporting surface. Generally, only spot modulation caused by web defects is useful for determining web quality; and thus, it is the prior art practice to blank, by one means or another, spot modulation which is not defect-dependent.
It frequently obtains that web defects are created, near the edges of a web, as a result of slitting, and otherwise handling, the web; and thus, blanking must also cancel edge-defect modulation to prevent inadvertent registration of defects which have no material bearing on the quality of the web.
US. Pat. No. 3,198,951, indicates a blanking method which is dependent on placing photopickups proximate the edges of a web, respectively to turn-on and turnoff a defect detection circuit.
US. Pat. Nos. 3,360,651 and 3,510,664, both, apparently, recognize that the system of US. Pat. No. 3,198,951 (1) cannot tolerate weaving of a web over a supporting surface, (2) requires repositioning of photopickups to accommodate different size webs, and (3) cannot be effective, without modification, for inspecting webs of nonuniform width. U.S. Pat. Nos. 3,360,651, and 3,510,664, suggest that spot modulation, as caused by web edges, be used for blanking purposes: the former producing a gate-on pedestal pulse in response to the lead edge of a web; and the latter employing a counting technique for gating on and off, and registering defects.
US. Pat. No. 3,360,651, is considered closest, in concept, to the present invention: U.S. Pat. No. 3,360,651, teaches that spot modulation corresponding to the edges of the web may be used to define a pedestal pulse for gating a defect detector into and out of operation. In addition, US. Pat. No. 3,360,651, teaches the effective narrowing of such pedestal pulse to prevent defect detection at the edges of the web under inspection. Pedestal narrowing is accomplished in US. Pat. No. 3,370,651, by pulse delay techniques involving feedback controls; and whereby the system of US. Pat. No. 3,360,651, automatically corrects for webs of varying widths.
So long as the surface over which the web of U.S. Pat. No. 3,360,651, passes is unproductive of signals as the flying spot of the scanner in question moves onto and OK such surface, the system of US. Pat. No. 3,360,651 will undoubtedly be effective for its intended purpose. Consider, however, the production of scan signals having the above-mentioned five occurrences of modulation during a given scan. Needless to say, were signals of such type to be applied to the Schmitt trigger circuit of US. Pat. No. 3,360,651, such circuit would undesirably operate two extra times during each scan of the flying spot, thereby preventing the proper operation of its pedestal defining circuit. Such extra triggering occurs because scan signals, of the type to be handled by apparatus according to the invention, produces two positive and two negative spike pulses for each flying spot scan, whereas scan signals which are handled by the apparatus of US. Pat. No. 3,360,651, produce only two spike pulses for each flying spot scan.
SUMMARY OF THE INVENTION The invention principally concerns a way to distinguish between pulses which define the edges of a web, as opposed to pulses which define the edges of a web-supporting surface.
Distinguishing between such pulses, according to the invention, is predicated on the use of clock pulses: A first clock pulse corresponding to a first portion (half) of the scan signal gates signals of one sense to an edge blank circuit (or pedestal producing circuit); and a second clock pulse corresponding to a second portion (half) of the scan signal gates signals of the opposite sense to the edge blank circuit (or pedestal producing circuit). In one version of the invention, each scan signal has, during each of its halves, first and second pairs of steep positiveand negative-going signal changes. By using a first clock to gate the positive-going signal change of the first scan signal half, and by using a second clock to gate the negativegoing signal change of the second scan signal half, the inner and outer bounds of the web may be defined. In other words, the present invention proposes to split the scan signal into halves, by means of the aforesaid clock pulse technique, thereby to distinguish the pulses which define the bounds of the web-supporting surface, and of the web itself.
Two other aspects of the invention obtain: (1) a technique for producing a pulse that appears to occur in response to, but before another pulse, and (2) a technique for distinguishing between large defect signals during a flying spot scan, and signals which result from spot modulation caused by the websupporting surface.
OBJECTS OF THE INVENTION 1. To provide blanking in a flying spot scanner system employed to inspect webs that are supported on a signal-producing surface;
2. To simulate the generation of a pulse a predetermined time before the occurrence of a signal which appears to trigger such pulse, etc.
The invention will be,described with reference to the V figures, wherein:
FIG. 1 illustrates in block form a web inspection system adapted to accommodate the invention,
FIG. 2 illustrates a signal form of a type which may be processed by means of the invention,
FIG. 3 is a schematic block diagram of a circuit according to the invention,
FIG. 4 is a plan view of a web illustrating scan dimensions according to the invention in a presently preferred form thereof,
FIG. 5 is a block diagram of a circuit which may be connected into the circuit of FIG. 3 to provide the scan dimensions indicated in FIG. 4,
FIG. 6 illustrates pulse diagrams useful for describing the combined apparatuses of FIGS. 1, 3, and 5,
FIG. 7 is an overall block diagram of apparatus embodying the various features of the invention, and
FIG. 8 illustrates pulse diagrams useful for understanding the operation of the circuit of FIG. 7.
Referring to FIG. 1, a flying spot scanner 10 causes a beam of radiation to sweep across a web 12 which rides over a websupporting roller 14. The roller 14 has a reflectivity different from that of the web 12; and light reflected from the web 12 and roller 14 is collected by optics l6 and directed to a photodetector 18. Each flying spot scan produces an optical signal like that shown in FIG. 2.
FIG. 2 indicates that the optical signal applied to the photodetector 18 increases (C) as the flying spot first goes onto the (shiny) surface of the roller 14, drops (D) as the spot goes onto the web 12, varies in response to web defects, increases (A) when the spot goes back onto the (shiny) roller, and drops (B) to an ambient level when the spot goes off such roller.
Output signals from the photodetector 18 are amplified (20) and applied to a gate circuit 22 which, in turn, applies such signals to a device 24 for registering the occurrence of defect representative signals. Operation of the gate circuit 22 depends on the occurrence of an edge blank pulse from a circuit 26, which pulse inhibits the operation of the gate circuit 22. The edge blank circuit 26 receives the amplifier 20 output signals, and signals from a photodetector 28 (and amplifier 30) that defines the point at which the flying spot first starts to sweep across the roller-and-web, thereby to produce the edge blank pulses. The photodetector 28 output signals have a scan duration (ST) as indicated.
The edge blank circuit 26 of FIG. 1 is indicated in detail in FIG. 3; note contacts P,Q,R:
A differentiator circuit 32 is adapted to receive the flying spot scan signal (FIG. 2); and produces therefrom a pair of spike pulses corresponding to the signal changes C,D and A,B. Since the signal duration which occurs between D and A corresponds to the period of scanning which is of interest, the invention proposes to isolate the signal occurrences D and A, each by means of a respective clock pulse. Such isolation is necessary because of the signal producing characteristics of the web-supporting roller 14 which is to say that positiveand negative-going spike pulses are produced during both the first and second halves of each scan (ST). The negative-going signal (D) that defines the lead edge of the web is isolated from the negative-going signal (B) that defines the trailing edge of the web-supporting roller 14 by means of a polarity detector 33 and a clock from afirst half-cycle detector 34, and the positive-going signal that defines the trailing edge (A) of the web 12 is isolated from the positive-going signal (C) that defines the lead edge of the web-supporting roller 14 by means of a polarity detector 35 and a clock from a second half-cycle detector 36. The half-cycle detectors 34,36 may take the form of counter leads on a monostable multivibrator that is responsive to the output of the roller-edge defining photodetector 28; such a multivibrator having a cycle of operation corresponding to the duration ST/2.
A pair of AND-gates 40,42 respectively receive the negativeand positive-going spike pulses, and the first and second half cycle clocks; and attendantly, such AND-gates 40,42 respectively produce the web-scan defining pulses D and A which turn a flip-flop 44 off and on, thereby respectively to uninhibit and inhibit the operation of the gate 22. The crosshatched area of the signal output of the flip-flop 44 represents an edge blank pulse so produced.
As stated above, it frequently obtains that a web is to be inspected across less than its full width (see dimensions Z-X, FIG. 4), thereby to prevent the ragged edges 46 of the web from producing defect signals. This means the blank pulse must be widened; which at first blush suggests that the edges A and D of the flip-flop 44 output are to be respectively advanced and delayed. However, since the occurrence of such edges are unpredictable, i.e., they depend on the nature of the scan signal, the matter of edge advancing, as opposed to edge delaying, is an impossibility. To simulate edge-advancing, the invention proposes a signal delay technique: the signal (A) to be advanced being delayed for a little less than the duration ST; and the signal (D) to be delayed being delayed a little longer than the duration ST. Thus, in accordance with the invention, signals which are produced during one scan are employed for generating a widened blanking pulse for use during a subsequent scan. FIG. 5 shows a circuit adapted to be connected to contacts L,M,R of FIG. 3 to implement the widening of an edge blank pulse, i.e., the circuit of FIG. 5 indicates appropriate delay devices 48,50 cooperative with the flip-flop 44 Reference should now be had to FIG. 6 which graphically illustrates how blank-defining pulses A, and D, effect a blanking pulse that overlaps (by At and At) blank-defining pulses A and D which is to say that blanking, for a given flying spot scan, is determined during an earlier scan and such blanking will be effective for ordinary rates of web travel.
The following description, relating to apparatus of FIG. 7, indicates a working arrangement for the circuit techniques discussed above in connection with FIGS. 3 and 5: Similar characteristic notations are employed in all such figures for corresponding components. See also the waveforms of FIG. 8:
A monostable multivibrator, responsive to the output of the photodetector 28, produces a pair of complementary clock pulses (34,36), each having a duration ST/2. For polarity detectors 33,35, Schmitt triggers respectively responsive to predetermined negative and positive signal level outputs from the differentiator 32 are employed; and their respective outputs are applied to the AND-gates 40,42 together with respective clocks 34,36, thereby to isolate the signals D and A.
To delay the signal D, as required, a pair of multivibrators 48, and 48, are employed. The multivibrator 48, produces a pulse of duration ST/Z; and the multivibrator 48 responsive to the trailing edge of the pulse from the multivibrator 48,, produces a pulse of duration ST/2+At.
Similarly, the signal A is delayed by a pair of multivibrators S0, and 50 the former for a duration ST/2, and the latter for a duration of ST/2-At.
By means of the D and A delays 48,50, the flip-flop 44 (bistable multivibrator) is respectively reset and set-in response to a given pair of D and A signals-a little more, and a little less, than a scan duration after such D and A signals are produced, thereby causing the multivibrator 44 to appear to reset in response to, and after, a signal which has just occurred, and to set in response to, and before, a signal which has not yet occurred.
The bistable multivibrator 44 serves to inhibit the gate 22 in the manner described above.
For certain web defects, such, for example, as a break or hole in the web being inspected, steep large amplitude signals S, sufficient to trigger the (positive) Schmitt trigger 33, are applied to the differentiator 32. See scan 3, FIG. 8. When a signal S occurs during the second half of a flying spot scan, it is gated through the AND-gate 42 and will, absent the improvement to be described presently, effectively widen the blanking pulse so that the defect S will go undetected. (The second half of a flying spot scan, as opposed to the first half, is of concern for the reason that only during the second scan half are positive thresholds of interest.) To avoid so widening the blanking pulse, in response to a large defect signal, that such signal may go undetected, the invention proposes to change circuit timing in response to large defect signals S, so that the signal A will turn off the flip-flop 44 (i.e., so inhibit its operation that it remains in its reset state), thereby to cause defect registration (24) to be in response to signal outputs produced from the web-supporting roller 14.
To this end, an AND-gate 60 is adapted to receive the signal A and the delayed half cycle pulse of the multivibrator 50, and, since such signals are usually of opposite polarity, the AND-gate 60 usually has no output. As soon, however, as the multivibrator 50, has the timing of its output pulse shifted by a signal S, the signal A gets gated out of the AND-circuit 60. A multivibrator 62, the purpose of which is to complement the half-cycle delay of the multivibrator 50, so that the effect of the isolated signal A will be a full scan duration after its occurrence, sets a bistable multivibrator 64 so that such multivibrator may inhibit the operation of the blank pulse producing multivibrator 44. Thus, the occurrence of a large defect signal S during the second half of a scan duration effectively turns off the multivibrator blanker 44, the gate circuit 22 being thereby left open so that signals from the web-supporting roller may be registered.
The invention has been described in detail with particular reference to preferred embodiments thereof, but it will be understood that variations and modifications can be effected within the spirit and scope of the invention. For example, whereas half-cycle clocks are indicated, such clocks may be longer or shorter durationwise than a half-cycle, so long as they complement occurrencewise the pulses which they are intended to isolate; the durations At and At may be alike or different; the delays afforded by the components 48,50 may be respectively a little more and a little less than multiples of the duration ST, etc.
What is claimed is:
l. A system for inspecting webs which are conveyed over a means of the output of said first AND gate means by a duweb-supporting surface, said surface and said web having subration slightly longer than a given multiple of the time stantially different abilities to modulate incident radiation, that it takes to sweep radiation across said surface and comprising: said web, and
a. a flying spot scanner for beaming and sweeping, at a given 5 b. second means for delaying the application to said bistable cyclic rate, a source of radiation across said surface and means of the p t f aid e nd AND gate means by a said web, duration slightly less than said given time multiple.
b. means for collecting radiation from said surface and said The apparatus of Claim 21 b f producing i l corresponding to respective a. wherein said first delay means comprises first and second di ti sweeps, l0 monostable multivibrators for producing respective c. gate circuit means adapted to receive signals produced by Square Wave Signals, first multivibrator being p Said radiation collecting means, and sive to the output of said first ANl) gate means and said means f producing gating signals for controlling the second multivibrator being responsiveto the trailing edge operation of said gate circuit means, said means for f the output 9 sald first multfvlbratorj the producing gating Signals comprising: bined durations of sa d square wave signals being slightly 1. means for producing first and second sets of clock greater than the Sam Sweep of Sam flymg Spot signals, the signals of said first set corresponding durationwise with the first half of each said radiation sweep, wherem second i means compnses. thud and and the signals of said second set corresponding durafijurth monostable .mumvlbrinors. for profhfcmg i tionwise with the Second half of each Said radiation two square wave signals, said third multivibrator being Sweep responsive to the output of 531d second AND gate means 2 f d ti ti l and said fourth multivibrator being responsive to the trailmeans or recelymg l 3 Sal S'gna S ing edge of the signal output of said third multivibrator, producfzd Sald radlanon. collefmng means the combined durations of said square wave signals from pmducmg pans of edge'ckfimng Splke P l dunng said third and fourth multivibrators being slightly less the first and second halves of each radiation sweep, than the Said Sweep time ofsaid flying Spot seamen said spike pulsescorresponding to radiation modula- 4 The apparatus ofclaim 1 including: non Sald radlanon sweeps onto and Off Sald a. circuit means for distinguishing between said edge definporting surface, ing spike pulses and spike pulses produced in response to 3. first and second AND gate means for receiving respeclarge web defects and tively said first and second sets of clock signals, b. means responsive to defect-produced spike pulses for 4. means for applying the spike pulses, corresponding to placing said gate circuit means in its signal passing state.
radiation modulation as said radiation sweeps from said h apparatus Of Claim 1 wherein: Surface onto said Web, to said first AND gate means, a. said web has a photographically sensitized surface, 5. means for applying th spike l corresponding to b. said web-supporting surface is a roller drum, the axial radiation modulation as said radiation sweeps from said dimension of which is Wide! than Said and b onto id Surface, to id Second AND gate means, c. said roller drum, at least proximate its extremities, has a d mirrorlike surface. 6. bistable means cooperative with said gate circuit apparatus of f h 5 iflcludingi means, and responsive to the pulse outputs of both said 40 F l means for dlstlllgulshmg bfitwee" l g AND gate means, thereby to place said gate circuit "8 spike Pulses and Spike Pulses Preduced response to means in a signal pasing state while said radiation large web defecisi and Sweeps across Said web b. means responsive to defect-produced spike pulses for The apparatus ofclaim 1 including: AND gate placing said gate circuit means in its signal passing state. a. first means for delaying the application to said bistable