claimed is: 20 1. An apparatus for the inspection of a runnin- sheet for defects protrudin.- alon- a normal to said sheet comprisin- in combination a pair of sighting paths disposed transverse said sheet closely adjacent thereto in planes substantially coparallel to the plane including said run- 25 ning sheet and spaced apart in the direction of sheet travel a distance equat to the minimum length of defect it is desired to base rejection upon, a light source directed at said sheet in the vicinity of said sighting paths, and electro-optical means responsive to light reflected along 30 said sighting paths from said defects consisting of a pair of individual photomultiplier tubes each responsive to light from one of said sighting paths solely, a voltage amplifier in electrical circuit with the output of each of said photomultiplier tubes individually, a - coincidence 35 circuit including a vacuum tube having two triode sections, the grid of one of which triode sections is in electrical circuit with the amplified voltage output of on-, said photomultiplier tube while the grid of the other of said triode sections is in electricll circuit with the ampli- 40 fied voltage output of the, other of said photomultiplier tubes, a thyratron tube having its control grid in electrical circuit with the combined output of both of said triode sections of said vacuum tube in said - coincidence circuit, and means responsive to the output of said 45 thyratron tube indicating the existence of individual ones of said defects. 2. An apparatus for the inspection of a running sheet according to claim 1 wherein said light source consists of a fluorescent tube operated on direct current having 50 means periodically reversin.- the polarity of said direct current across the terminals of said fluorescent tube and a switch momentarily rendering said electro-optical means indicating the existence of defects in said sheet inoperative during the time interval in wh-eh said polarity is in r)5 course of reversal. 3 ' An apparatus for the inspection of a running sheet for defects protruding along a normal to said sheet comprising in combination a pair of sighting paths disposed closely adjacent to and transverse said sheet in planes 60 substantially coparallel to the plane including said running sheet, each of said sighting paths being split into a portion disposed above said sheet and another portion disposed below said sheet, the corresponding portions of each sighting path being spaced apart in the direction 65 of sheet travel a distance equal to the minimum length of a defect it is desired to base rejection upon, a light source directed at said sheet in the vicinity of said sighting paths, and electro-optical means responsive to light reflected along said sightin.- paths from said defects indi- 70 cating the existence of individual ones of said defects which obtrude coincidentally in said pair of sighting paths.. 8 4. An apparatus for the inspection ot a running sneet for defects protruding along a normal to said sheet comprising in combination a pair of sighting paths disposed transverse said sheet directed toward the region closely adjacent to said running sheet in planes substantially coparallel to the plane including said running sheet, and spaced apart in the direction of sheet travel a distance equal to the miniinum length of defect it is desired to base rejection upon, a light source directed down each of said sighting paths, and clectro-optical means responsive to changes in the amou.Tit of light transmitted down each of said sighting paths due to the presence of a defect in said paths indicating the existence of individual ones of said defects which obtrude coincidentally in said pair of si.-hting paths. 5. A method for the inspection of a running sheet for defects protruding along a normal to said sheet, comprising scanning the space closely adjacent to said runnin.- sheet in planes substantially coparallel to the plane including said running sheet, detecting the presence of a defect in each of a pair of discrete regions spaced apart lengthwise of said sheet a distance equal to the minimum length of a defect it is desired to base rejection upon, and obtaining an indication of any defects responsive to coincidental detecting in said pair of discrete regions. 6. A method for the inspection of a running sheet according to claim 5 wherein detecting of said defects in said discrete regions is conducted on opposite sides of said sheet. 7. An apparatus for the inspection of a running sheet for defects protruding along a normal to said sheet, comprising in combination means scanning the space closely adjacent to said running sheet in planes substantially coparallel to the plane including said running sheet, means detecting a defect in said sheet in a pair of discrete regions spaced apart lengthwise of said sh@-et a distance equal to the minimum length of a defect it is desired to base rejection upon, and means responsive to said means detecting said defect indicating the existence of any defects which are detected coincidentally in said pair of discrete regions. 8. An apparatus for the inspection of a running sheet according to claim 7 wherein said means detecting a defect in said sheet in said pair of discrete regions spaced lengthwise of said sheet are adapted to detect said defects on opposite sides of said sheet. 9. An apparatus fo@r the i@ispection of a rumiing sheet for defects protruding along a normal to said sheet, comprising in combination a pair of sighting paths disposed transverse said sheet closely adjacent thereto in planes substantially coparallel to the plane including said running sheet, and spaced apart in the direction of sheet travel a distance equal to the minimum length of defect it is desired to base rejection upon, a light source directed at said sheet in the vicinity of said sighting paths, and electro-optic-,tl means responsive to light reflected along said sighting paths from said defects indicating the existence of individual ones of said defects which obtr-ude coincidentally in said pair of si.-hting paths. References Cited in the file of this patent UNITED STATES PATENTS 2,233,483 Metcalf ---------------- Mar. 4, !1941 2,246,501 Bradner et al - ---------- June 24, 1941 2,295,327 Bendz ----------------- Sept. 8, 1942 2,429,331 Sachtleben ------------- Oct. 21, 1947 If17 2,640,152 Hartwig --------------- May 26, 1953 2,721,942 Friel et al. -@ ----------- Oct. 25, 1955 2,753,464 Stone ------------------ July 3, 1956 2,791,931 Summerhayes ---------- May 14, 19@57 2,806,401 Demuth et al - ---------- Sept. 17, 1957 UNITED STATES PATENT OFFICE - Certificate of Correction Patent No. 2,961,94:7 Williai,n M. Flook, Jr. It is hereby certified tl3-at error appears in the above iiuiiibered p,,tteiit requiring correctioli alid that the said Letters Patent should read " corrected below. In the drawiiigs, Sheet 2, coiitaiiiiiig Fi.g. 5, shotild tppear -bs shoivii belolv iiistetd of as in tl-ie patent: Jan. 10, 1961 W. M. FLOOK, JR 2,967,947 SHEET INSPECTION MET[IOD AND APPARATUS Filed Oct. 4, 1956 3 Stieets-Sheet 2 January 10, 1961 INT VE NT OR WIL LIA M M FLO OK, JR. BY A T T O R N E I ' columii 3, line 389 for "patr" read -part-; coluinn 5, line 53, for "tempQr,,tily" re,,td -temporarily-.
2 @ 9 6 7 9 9 4 7 Uni*ted States Patent Office P ' a t e n t e d J a n . 1 0 , 1 9 6 1 2,967,947 SH,EET INSPECTION METHOD AND APPARATUS@ 5 i*iiiikm M. Flook, Jr., GreenvilIe, Del., assignor to E , L@ du Pont de Nemours and Comp@ny, Wilniiiigton, Del.@ a ccirporation of DelitWaie Filed Oct. 4, 1956; Ser. No. 613,9267@ 10 9 Claims. (Cl. 250-219) invention relates to a method and apparatus for 15 the inspection of a running sheet such as a textile yatn wai-p'sheet, woven textile fibric, paper or the like, and p-arti6ularly to !a method and apparatus for the coiitinuous inspection of a warp sheet of heavy denier:yarn iri,ade up of textile tire cord yarn or similar strands. 20 Textile tire cord is typically mad6 up fronit-,visted fila,ments of yarn of about 840 denier. It is conveni6nt to beam the yarn on a large com-mon spool whicli often contains @a inultiplicity, typically 160, of ends wound adj acent- one another, so that each end can be ag@in: un25 wound without interference @from otl-ier ends by the piircha@er who utilizes the yarn in the fabrication of tire c6rd. A@ number of smal-I paelcages of yarn are commonly wound on the beam for each yarn end, the yarn from one package being spliced to that of a succeeding 3 0 pack-age with weaver's knots, which @s an acceptable procedure not affecting the quality of the product. An a,ver,age of 300 passable knots may occur in the processiiig c@f '160' individual yarn ends in the course of an hour of iiormal production, therefore, aiy inspection proredure 3 5 nifist@ discriminate successfully in favor of passage of the knots as non-objectionable phenomena over true defects. Ahother anomaly which exists, but which is not presently classified in @the trade as a cause for rejeotion, is the existence of loops, which may result from one or more 40 monofilaments being longer than the rest and projecting from the yarn. A more serious defect is the so-called "fluff ball," which usiaally consists of a small twisted mass in the yarn resulting from a break in one or more monofilaments at 45 one point, or the colleetion of substantial lengths of such bioken monofilaments at one point in the yam. This defect is objectionable because it usually is accompani@ d by a, loss of strength, if caused by the breakage of one or more filaments, and by difficulties in subsequent handling due' to, interference occasioned by the existence o,f the fluff ball in the product. Accordingly, it is the pre@ent practice to subject the running yarn to visual in.spection, stopping the yam in its travel whenever an objectionable number of fluff bahs is determined to exist, 55 cutting out @the objectioniable area, retying the severed ends a ' nd passing the product on to the beam. Visual inspection is tedious, vulierable to human failure, and reduces pro,dtiction output, since the yarn could normally be wound at relatively high speeds if it were not for the 60 necessity for slow speeds imposed by the inspection procedure. The following description is particularly directed to the inspection of yarn warp sheets; however it @will be understood that inspection of other sheet-like materials is equally feasible by the use of this invention 65 and the term "sheet" as utilized herein is intended to comprehend opaque and transparent webs, such as paper, pl@astic, w6ven textilo, fabric and the like, a@ well as a niultiplicity of running yarn ends disposed in generally 7 0 parallel relationship in the manner of a textile warp. M @object of this invention is to pro@ide an aiitomati@ inspectio-n procedure fdi running sheets traveling at very high@ speeds. Other objeots of this invention are the provision of an electro-optical inspection method and apparatus 'for a warp sheet of runndng yarns which is capable of ' discriminating between passable and nonp,)Ssablelahomalies in the yarn to effect evaluation on tho prop&r basis, and which can operate rapidly enough to de,,Cct def6cts at substantially greater operating speeds than:@are now possible. Further objects of this invention are @ th6@ provision of a very stable sheet inspection apparatus Nvhi6h is capable of counting defects and giving a warning signal whenever a rejectable defect is enc@ountered, together with @automatic stopping of windup, if @ desired, @; which are achieved as brou.-ht out in the detaileddescription and the drawings, in which: Fig@, l'is a partially schematic side elevation of a multipl6@ nd:yarn beaming installation, Fig, 2 is -a top plan view of the instaration of Fig. 1, Fig'-. 3 is a schematic represen-tation in sectional plan of a @referred embod-iment of the optical arrangement of the- inspection 4apparatus of tliiis invention, Fig.@ 4 is a section taken transverse the sighting paths on litie- 4@@4 of Fig. 3, Fig. 5 is @a schematic representation of an c@lectrical eircu-it@ adapted for use in conjunction with the optical ar'tatigement4 of Figs. 3 and 4, Fig. 6 is a schematic representation transverse the sighting p@--ths af a split path einbodimeiit of this invention, and Fig. 7 is a scl,.ematic representation in sect,,tonal plafi of an optiral arrangement of inspection apparatus ac'-' cordin.- to this invention wherein hght beams are employed to scan the running yarn warp sheet. Generally, this inven-tion consis-ts of a method and apparatus for sensing a defect in a runmng sheet in a pair o@f discrete regions spaced lengthwise of the sheet a sufficient distance apart so that rejectable de'Ler-ts solely are sensed coincidentally for both of the regions, and obt-aining an indication of any defects responsive to coinoidental sensing in the two reg,'@Ons. A -reat variety of transducers are adap@ted to se7nse the existence of individual defects in a selected region of a running sheet typical examples being a d ' flectible wire stretched @acr@,, th, sh,,t :@lo,,,Iy adjacent thereto and connected in circuit w,@'th a pulsing m-icrophone, con-, ventional capacitometers of various types, and other devices; however, it is preferred to utilize electro-optical means for conducting inspection according to this inve@ntion. In the following detailed description relating to electro-optical inspection s@stems particularly, the, term "sighting path" is employed as a species com@rehended within the generic term "discrete region@ of defect sensing." Referring to Figs. I and 2, yarn to be beained i :supplied from a multi@licity of packages mounted in a creel indicated generallyat 5, which deliver a plurality of yarri' ends 6 for windup on a beam 10 driven by a conventional beanung machine 11. The y@rh w@r]@ sheet 7 cbnsisting of approximately 160 sepaiate strands, is' formed as a flat sheet of individual yarn strands- disposed. closely adjacent lone another by conver@ence -Uide 8 which@ rece@ives ends 6 from the creel. In a typical;case, warp 7 may be 54" wide and convergence' guide @8 is spaced away from idler roll 9 on beaining inachine 11' a distance far enough to permit stopping the y , arn windup process, if desired, upon detecfion of a defect within a short enoiigh time interval to remed@ mattets befor6 t]@e yarn is wound up on beam 10. i A prdf6rred -@rraiigement of a@varaius for 'effe&ting' nspection according to this inventi-on uiilizing a p@ir of' uiiitary disciete iegions of defect -sensing is shown- m 3 Figs. 3 and 4. All of the optical components are preferably mounted within a common light-tight housing 15 which is provided with an opening 16 oriented transverse warp 7. Opening 16 is sufficiently wide to accommodate two sighting paths 17 and 18 normal to warp 7, path 17 (refer Fig. 4) preferably being disposed on the opposite side of the yarn from path 18, although this is not essential. Paths 17 and IS terminate in light trap 19, which is simply a box closed at all points except for apertures in line with the sighting paths and painted on the inside with light absorptive black paint to provide a constant r,- ference for operation of the apparatus and, at the same time, reduce the ambient light normary incident on photomultiplier detectors 27 and 34- hereinafter described. Sighting path 17 is spaced from sighting path 18 a sufficient distance aic)ng the direction of yarn travel so that only rejectable defects will lie across both channels simultaneously and, thereby, give a coincidence signal, as hereinafterdescribed, representative of anomalies whicb constitute a basis for rejection. That is, the spacing between sighting paths 17 and 18 in the direction of yam travel is equal to the minimum length of defect it is desired to base rejection upon. Normally, a spacing of 1/4" is satisfactory to achieve the necessary discrimination. The ends of the si.-hting paths within housing 15 terminate at silvered reflective 90' prism 23, one face of which is disposed transverse sightin.- path 17 and the other face of which is disposed transverse sightin.- path 18. Reflected light from path 17 is passed through collimating lens 25, which is mounted behind light mask 24, the light from lens 25 being thereby focused through the aperture of niask 26, disposed at the focal point of lens 25, on the light-sensitive cathode of pliotomultiplier cell 27. The aperture of mask 26 in a typical case was 0.010" wide, the lens 25 used therewith having a focal length of 8". A completely symmetrical optical counterpatr consisting of light mask 31, collimating lens 32, mask 33, and photomultiplier cell 34 is provided for the sighting path 18. Referring to Fig. 4, the light for the system is provided by a fluorescent tube bulb 38, which may be a General Electric type 90/TI7 operated on direct current with periodic reversal of polarity as hereinafter described, mounted in housing 39 open at 41 along the side adjacent yarn warp 7. Although light source 38 is mounted beneath the yarn warp, it serves to also illuminate the region above the warp as well as the region below the warp, because light passes freely through the spaces between the running yarn strands. Smooth, wear-resistant steady rests 40 are preferably provided in abutment with the yarn warp at the opening of housing 39 to prevent cyclic vibrations or fluttering of the warp sheet, such as might occur from windage or the like. Where the distance between the inspection apparatus and guide 8 is relatively great, steady rests 40 are preferably supplemented by passing yarn warp 7 in an S-wrap about a pair of freely rotating rollers before the yarn passes to rests 40. One conventional circuit which has performed satisfactorily in conjunction with the inspection apparatus of this invention is detailed schematically in Fig. 5. As shown, light reflected from prism 23 is impressed on the cathode elements of photomultiplier tubes 27 and 34 individually, the output of photomultiplier 27 being repre-sentative of conditions existing along sighting path 17, whereas the output signal of photomultiplier 34 is repre-sentative of conditions existing along sighting path 18. The power supply for photomultipler tubes 27 and 34, indicated generally at 45, may deliver a thousand v. D.-C. filtered output, as does a Powell.high vqltage unit, which is u ' tilized in association With a suitable auxiliary, such a I s a Model 28 Lambda power supply, . The pow6r input to supply'45 is derived from @ower supply line 55 through load 43, 44 being an elcqtrical ground, and t@e output is 2,907,947 4 delivered through lead 46 and the tap of resistor 47, to the cathodes of tubes 27 and 34 in conventional manner. The aniplified electrical output signals representative of the light amounts impressed on the respective cathodes are passed through leads 50 and 51 connected to the anode elements of the photomultiplier tubes to A.-C. adjustable gain voltage amplifiers 52 and 53 (operative in the 201000 c. range approximately), respectively. A.-C. amplifiers are utilized because the signals processed by the 10 apparatus are electrical pulses and, accordingly, steady state D.-C. conditions have no effect on operations. Operating power is supplied to ariiplitier 52 by direct connection with B+ 190 v. power supply line 55, while amplifier 53 is powered from the same source by connec15 tion through lead 56. The circuit past the amplifiers is also powered from line 55, the electrical circuit being completed through common ground 54. The output of amplifier 52 is impressed on grid 60 of the first section of double triode section coincidence cir20 cuit tube 61 (which may be a 12AX7 type) through tapped resistor 62 (typically 5 megohms) connected in series with fixed resistor 63 (which may be 22 megohms) to power lead 55. Similarly, the output of amplifier 53 is impressed on grid 65 of the other triode section of 25 tube 61 through tapped resistor 66 in series with fixect resistor 67, also connected to lead 55. The cathode elements of both of the triode sections of tube 61 are grounded at 68, while the plate elements of both of the triode sections are connected together to line 55 through 30 fixed resistor 69, which may have a magnitude of about 270K. The output signal of tube 61 is derived through coupling capacitor 70 (typically 0.01 mfd.) and impressed on the control grid 72 o4l thyratron tube 71 (which ma-,@ be a type 2D21) through series resistor 73 (which may 35 b-. a 100K resistor). Grid resistor 74 (of about 220K value) is connected to ground 54. Thyratron cathode resistor 75 and resistor 80 toether constitute a voltage divider to provide the appropriate cathode bias. The suppressor grid 79 is connected to the cathode element 40 of tube 71 as is conventional practice. The output signal from thyratron 71 is drawn from the plate element thereof through lead 81 connected to one conta@t of normally closed pushbutton switch 82, the other contact of which is connected in series with nor45 mally closed relay contacts C, operated by relay C, hereinafter described. The circuit to power supply line 55 is completed tlirough normally closed relay contacts El of relay E, which are connected in series with the relay coil and fixed resistor 83. Relay E may be a three con50 tact type of 15 ma. size provided with a holding capacitor 78 of approximately 8 mfd. capacitance. Two auxiliary circuits are provided to: (1) obtain a count of rejectable defects as these occur from time to time and (2) halt windup of the yarn by beaming ma55 chine 11, at the operator's option, so as to permit cutting fluff balls or other defects from the ends and retying them prior to viinding product on beam 10. The counter circuit, indicated generally at 84, consists of a counter actuation coil 85 connected in series through 60 normally open relay contacts E3 of relay E witn 115 v. A.-C. power lines 86 and 87. An operative condition verifyin.- flash lamp 88 is preferably connected in shunt with coil 85, and an R-C arc suppressor network 89 is connected in shunt with relay contacts E3- 65 A practical contro@' eireuil for beaming machine 11 utilizes a manually operated double pole, double throw switch Sl, the left-hand side of which is effective, wheil ciosed on the upper pair of switch contacts, to shunt the output signal from thyratron 71 around relay contact.-, 70 El. T'ne right-hand side in upper switch closure positioli is simultaneously effective to complete the circuit through control leads 94 and 95 in the power supply circuit (not further shown) of beaming machine 11. An interlock is provided through normally open relay contacts E of 75 relay E, which are connected in series with leads 94 and 95. The lower pair of contacts of switch S, are open eircuited. The circuit includes rel,"y C, which is connected in series with resistor 96 froim power supply line 55 to ground 54 through normally open switch contacts D3. Contacts D3 are closed neriodically at predetermined intervals of an hour, or somewhat longer, by a conventional timing cam, indicated aene rally at 100, throu,-h the driving connection shown in broken line representation. Relay C is provided with additional contacts, not further detailed herein, which operate in the power suppl3i 101 of fluoresceiit ttibe bulb 38 to thereby reverse the polarity thereon, as hereinbefore mentioned. This polarity reversal avoids localized accelerated deteriorat;on of the phosphor lining of 39 with accompanying objectionable uneven illumination of the yarn. In operation, it will be understood that a fluff ball or other rejectable defect in the running yarn will obtrude @eyond the plane of yam warp 7 either above or below the yirn, where it reflects light incident on it froni fluorescent tube bulb 38 along the sighting paths. To constitute a registered defect, as hereiiibefore brought out, light must be reff@eted sirnultaneously along botli sighting paths 17 and 18, so as to produce a signal of appropriate potential by actuation of coincidence circuit tube 61. Tube 61 is operated so that each of its triode sections are normally conductin.@ to saturation. Under these circumstances, the negative signal derived at the gridresponsive to light impinging on a photo-multiplier tube, such as 27 as an example, immediately etits off current flow through the associated triode section and raises the potential at point A in the plate circuit a finite amount. In the typical circuit described, the parameters are chosen so that cutoff of the left triode section results in a rise of potential at point A from an ambient level of about 40 v. to about 60 v., a rise of 20 v., whereas cutoff of botli sections raises the potential of point A to the level of the B+ supply, or approximately 190 v. 'thyratron 71 in this instance was adjusted to fire on a sudden voltage rise of about 30 v., so that the thyratron will fire only when there is coincidental cutoff of both of the triode sections of tube 61. Since cutoff of both sections oectirs only in the presence of a rejectable defect, it will be understood that discrimination is thereby achieved against all anomalies except those which are known to be of rejectable nature. If this criterion is met, ,@rid 72 of thyratron tube 71 has impressed thereon a positive signal, which causes the thyratron to conduct with delivery of an output pulse to relay E. Assuming that switch SI is in open position as shown in Fig. 5, the output pulse from thyratron 71 causes relay E to close contacts E3 in the counter circuit, which peri,nits current to flow temporaily through coil 85, thus giving a-single count. Relay E at the same time acts to ppen relay contacts El,, and to hold them open for a finite interval by the time delay action of capacitor 78. This time interval is proportioned to permit sufficient time for counter 84 to make a count while, at the same time, restoring the circuit to detection condition as rapidly as practicable to safeguard against registration of a plurality of defects as only a single defect. 'Opening of relay contacts El terminates conduction through thyratron 71, whereupon the tube is restored to inspecting condition. With the foregoing operation,, inspection is limited soldly to the counting of rejectable defects. Where it is desired that beaming. halt upon the occurrence of a defect, the operator closes switch SI on the upper set of contacts. Switch S, then shunts contacts El while, at the same time completing the . electrical circuit through control leads 94 and 95, except for relay contacts E2If the apparatus now detects a defect, contacts E2 close and an appropriate signal passes through leads 94 and 95, halting windup by beaming ma@-hine 11 until the operator can remove the defect from warp 7, or other2@9671947 wise take corrective action@ Opening of relay contacts El does @ not, with this operation, halt cbnduction@ thr6uih thyratron 71 and the operator miist open switch 82 manually to clear the circuit for further inspection service, thus insuring that attention has been given to matters before resumption of beaming. The counter circuit is, of course, actuated through one count corresponding t6 each individual defect regardless of the settin-, of switch S1. . Periodic reversal of polarity of fluorescent tube bulb 10 38 under control of timing cam 100 wiR give a spurious i-ejection signdi unless positive safeguards are provided, because some light is always passed to the photomultiplier tubes due to strary light effects, which light drops to zero level when bulb 33 goes off during polarity reversal. Re15 lay C averts the possibility of spurious sign@ls by opening relay contacts C, upon closure o,f switch contacts D3 a slight interval in advance of polarity reversal. With contacts C, open relay E remains unenergized and there is no defect registration by the ap@aratus for the brief 20 i-@iterval contacts D3 remain closed. Since polarity reveral requires only a few m@'lliseconds tinie, and occurs but once during an hour or more of operation, this is no serious disadvantage. The gains of amplifiers 52 and 53 are adjustable, as 25 hereinbefore described, which permiis adjustment of the sensitivity of the inspection appar@tus with respect to density of defectsencountered. In some iri-stances improved defect discrimination is obtained by the use of a pair of split path regions of de30 fect sensing, such as the sighting paths represented in Fig. 6. Here each sighting path is divided into two separate portions, the leading path consisting of 17a, above warp 7, and 17b, below the warp. Similarly, path 18 is divided into an upper part 18a and a lower part 35 18b, each disposed along the line of warp travel with a spacing of aboui 1/4" from their counterparts 'm path 17, as hereinbefore described for the embodiment of Figs. 3 and 4. The cross sectional areas of 17a and 17b may each be about equal to that of 17 previously describ,ed, 40 Iga and I$b each then equalling 18. The optical system for the arrangement of Fig. 6 is identical with Fig. 3@ it being understood that radiation received along the two split portions of each sight,',ng path now passes to a commoii detector for each path and that operation is other45 wise identical with that hereinbefore described. The increased reliability of the scanning arrangement of Fig@ 6 is probably due to the fact that each sighting path now views areas both above and below the warp. Unsymmetrical defects which may proj6ct, for example, 50 more above the warp than below it are less likely to fail to yield c6incidence signals. While it is preferred to conduct the inspection with defect sensing based on light reilected from individual d6fects, it is also practicable to employ selective inter55 ruption of light transmission in the sensing, a preferred optical arrangement for which is detailed in Fig. 7. Here the detection side of the system is identical with that shown in Fig. 3; however, the opposite side corresponding to what was formerly light trap 19 is now housing 60 106 for the light source, wliieh may be two independent 50 watt incandescent light bulbs 107 and 103, each disposed back of individual masks 109 and 110, respectively, in line with collimating lenses 111 and 112, respectively. The two light beams are deflected along sighting paths 65 17 and 18 by silvered, reflective 90' prism 113 and con.stitute the defect sensing means by virtue of selective interruption of light transmitted therealong responsive to the presence of rejectable defects. It will be understood that the same electrical circuit as shown in Fig. 5 can be 70 employed in conjunction with the embodiment of F - ig@ 7, except that it is now necessary to adjust th(@ gain conttol suitably to accommodate for the changed ambi@,-nt'lightlevel. When the arrangement of Fig. 7 is utilized,;fliio@ rescent tube bulb 38 is disnensed with. 75 The split sighting path arrangement of Fig. 6 rnay, if 2,967,947 7 desired, be employed with defect sensing by the apparatus of Fig. 7, conventional light dividers being then employed to define the separate light beams Nvhich are directed along paths 17a, 17b, 18a, and 18b. It will be understood that the several embodiments 5 hereinbefore described for the inspection of warp sheets are equally applicable to the inspection of sheet materials generally, such as paper, plastic, textile fabrics and the like, except that it is, of course, then necessary to use an individual light source on opposite sides of the sheets 10 wherever the sheets are light-opaque. From the foregoing it will be understood that my invention provides a method and apparatus capable of inspecting running sheet material which is adapted to distinguish between rejectable defects and non-rejectabl-. 15 anomalies rapidly and dependably, and that nusierous alterations can be made therein without departure from the essential spirit, wherefor it is intended to be limited only by the following claims. What is