GB1601324A - Reproduction scanning systems and methods - Google Patents

Description

PATENT SPECIFICATION ( 11) 1 601 324

C ( 21) Application No 29130/80 ( 22) Filed 9 Mar 1978 ( 19)( cq ( 62) Divided out of No 1601322 ( 31) Convention Application No 776321 ( 32) Filed 10 Mar 1977 in i'' t o ( 33) United States of America (US) I O ( 44) Complete Specification Published 28 Oct 1981 ( 51) INT CL 3 H 04 N 1/02,.

( 52) Index at Acceptance H 4 F DB 525 L 525 R 525 S 527 A 1 527 B 527 F 527 H 527 R 2 527 S 530 H 530 K 542 R 1 583 X 58959 54951 569 X 570 ( 72) Inventor: WILLIAM F GUNNING ( 54) REPRODUCTION SCANNING SYSTEMS AND METHODS ( 71) We, XEROX CORPORATION of Xerox Square, Rochester, New York, United States of America, a corporation organised under the laws of the State of New York, United States of America, 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: 5

The invention relates to reproduction scanning systems and methods.

Laser scanning techniques for writing or printing on a medium sensitive to the laser beam have been disclosed in the prior art, as shown, for example, in U S patent 3,922,485 In general, the laser beam is intensity-modulated in accordance with information to be printed on a receiving medium, the modulated laser beam being directed to a rotating scanner, or 10 reflector, such as a multi-faceted polygon The rotating scanner in turn causes the modulated laser beam to scan, in sequence, across a sensitive medium located a distance away from the scanner The information contained in the intensitymodulated laser beam can be directly written on the medium if the medium is sensitive to the laser beam, or in an alterntive embodiment, the laser beam can selectively discharge a charged insulating or 15 semiconducting surface in accordance with the intensity of the beam In the alternative embodiment, the degree of charge dissipation corresponds to the information contained in the intensity of the laser beam The areas of the medium which are not discharged by the laser beam are subsequently developed, for example, by standard xerographic techniques.

Present day copiers which are commercially available and which utilize the xerographic 20 process include a platen upon which the document to be reproduced is placed, the platen being flat or curved The document is generally flood illuminated or scanned with light and the reflections therefrom are imaged via a copy lens to a charged photoconductive medium to discharge the medium in accordance with the image formed on the document.

The Telecopier (Trade Mark) 200 a facsimile transceiver manufactured by the Xerox 25 Corporation, Stamford, Connecticut, directs reflections from a laserscanned document onto a photosensitive transducer, the electrical signal output thereof being transmitted to another location and used to modulate a laser beam to reproduce the scanned document.

However, the Telecopier 200 is generally not considered a copier type device because, inter alia, a scanning platen and other copier features are not available 30 Although copiers now commercially available are not adapted to utilize scanning techniques to scan line-by-line a document placed on the copier platen to produce a serial bit stream corresponding to the scanned information (i e a raster type scanning system), it would be advantageous if such copiers could be modified to incorporate the laser printing technique disclosed, for example, in the aforementioned patent, the modified copier thus 35 requiring a system which provides for two-dimensional raster input scanning A system for two-dimensional raster input scanning which utilizes a laser, as described, for example, in U.S Patents 3,970,359 and 4012,585, provides a flying spot scanning system which is capable of scanning an unmodulated beam to a reading station for reading a stationary document and a modulated beam to an imaging station for, inter alia, reproducing the 40 2 1 601 324 2 scanned document thereat.

The availability of a copier which utilizes raster-type input scanning of a document placed on a platen and laser scanning techniques for writing on a lasersensitive medium would provide many advantages inherent with the use of lasers and raster type input scanning, such as increased copying speeds and resolution In particular, it would be advantageous if 5 an intermediate storage medium were provided between the input and output scanning stations, to allow for manipulation and storage of the scanned information, and, in particular, to provide for electronic precollation which electronically arranges representations of images to allow collated sets of documents to be reproduced Other desirable features of such a copier would include input scan reversal for alternate bound pages during 10 bound volume scanning, synchronization of the system by a clock associated with the storage member, a synchronous system reducing the size and cost of a synchronising buffer associated therewith, input/output interleaving with a print interrupt feature, image centering and edge faceout for image reduction, and independent magnification/ demagnification by separately-adjustable raster spacing 15 According to the invention there is provided a scanning system which is as claimed in the appended claims.

The following specific description with reference to the drawings relates to reproduction apparatus incorporating several inventions which are subject matter of: application 9392/78 (patent 1601322); application 8029129 (patent 1601323), and application 8029131 (patent 20 1601325).

Embodiments of the invention will now be described by way of example with reference to the accompanying drawings, in which:Figure 1 shows in simplified form an optical arrangement; Figure 2 is a simplified block diagram of the overall system; 25 Figure 3 illustrates a disc surface and a typical recording pattern formed on the disc; Figure 4 is a more detailed block diagram of the system; Figure 5 illustrates in more detail the operation of the synchronizing buffer which comprises a portion of the system; Figure 6 is a more detailed block diagram of an output shift register; 30 Figure 7 illustrates how a reduced image may be centered on an output medium, and Figures 8 to 10 are linking diagrams illustrating operation of part of an embodiment of the invention.

Referring now to Figure 1, light sources 10 and 12 provide original beams 14 and 16, respectively, for utilization by the scanning system Light sources 10 and 12 are preferably 35 lasers which provide collimated beams of monochromatic coherent radiation, laser 10 comprising a helium-cadmium laser which generates blue laser light at a wavelength of 3316 A, and laser 12 comprises a helium-neon laser which generates red laser light at a wavelength of 6328 A The use of the two laser beams ensures that the document scanner is not insensitive at the wavelengths of lasers 10 or 12 and hence, the system is suitable for 40 detecting light fluxes reflected from multi-colored documents in addition to the fact that a choice of laser beams is available for forming information on a radiationsensitive medium.

Light beam 14 is incident upon beam splitter 18 which directs a portion of light beam 14 to dichroic mirror 20 Light beam 16 is also incident on dichroic mirror 20, which is positioned to reflect the flux in beam 14 as a combined beam 22 (combined with transmitted beam 16) 45 Beam 22 is incident upon pre-image cylinder lens 24 which transmits the beam to mirror 26 which directs the beam to a rotating scanner 28 via a split doublet 30 The portion of beam 14 transmitted by beam splitter 18 is incident on modulator 32 which may either be an acousto-optic or electro-optic type device, the output thereof being incident on scanner 28 via pre-image lens 34, mirror 36 and split doublet 30, the split doublet 30 allowing the 50 separate beams incident thereon to be focused on the platen 62 or drum 76 Rotating scanner 28, shown as comprising a polygon having a plurality of reflecting facets 38, is driven by motor 40 via drive shaft 42.

Scanner 28 rotates in the direction of arrow 44 causing the laser spot (combined laser beam) incident thereon to deflect in the x direction at mirror 43, the output beam being 55 directed to a movable scanning assembly 45, shown in a simplified representational form, which comprises mirror 46, cylinder lens 48, mirror 50, bidirectional motor 52 having a stepped pulley 53 on its output shaft, cables 54 and 55 and pulleys 56 and 57 Elements 43 and 50 are rigidly fixed to cable 54, element 46 being fixed to cable 55, element 46 being driven at half the speed of elements 48 and 50 to maintain a constant focal length between 60 the platen 62 and mirror 43 This technique is generally referred to as half-rate mirror scan, such a technique being disclosed in U S Patent 3,970,359 A scan spot 58 is produced which moves along scan line 60, formed in the x-direction at platen 62, as scanner 28 continues to rotate Although not shown in the figure, a document, or a page in a bound volume, to be scanned is placed face down on the top surface of transparent platen 62 Since motor 52 is 65 1 601 324 bidirectional, the direction of y scan is selectable by an operator by appropriate activation of buttons formed on an operator's panel 92 (shown schematically in Figure 2) which in turn causes a system-controlling microprocessor 90 (Figure 2) to generate the appropriate control signals As will be set forth hereinafter, the particular scan direction selected is determined by the type of input being scanned, alternate pages of a bound volume generally 5 requiring reversal of the normal scan direction.

When a document is placed face down on platen 62, it is scanned by the two color laser beam spots 22, the document reflecting the incident radiation flux in accordance with the document information being scanned A fraction of the reflected flux is detected by one or more photomultiplier tubes (or other photosensitive device) represented by a single 10 photomultiplier tube 66 located under the platen 62 via mirror 64 The photomultipliers convert the variation in intensity of the reflected laser beam into electrical information signals which may be transmitted to an intermediate storage device 96 via a synchronizing buffer 98 (shown in Figures 2 and 4) and thence to a recording device via the intermediate storage device, synchronizing buffer and modulator 32 for producing a copy of the 15 document scanned as will be explained hereinafter The scanner 28 and scan system 45 are arranged to scan the material on platen 62 in a manner whereby a plurality of scan lines 60 is generated across the width of platen 62 such that the material on the transparent platen is completely scanned.

As shown in the Figure, the single beam reflected from mirror 36 is also incident on the 20 facets 38 of scanner 28 and caused to scan mirror 70 which directs the beam to mirror 72, mirror 72 in turn scanning the incident beam on cylinder lens 74 Cylinder lens 74 focuses the beam on a recording member 76, such as a xerographic drum, rotating in the direction of arrow 78 A plurality of scan lines 80 is formed on the surface of drum 76 in a similar spatial relationship (the reproduction not being accomplished in time synchronism because 25 the output from the photomultiplier tubes is initially directed to the intermediate storage device 96 via a synchronizing buffer 98, in the preferred embodiment) with the information being scanned on platen 62 thereby to reproduce a copy of the image on drum 76 in a manner as described in the aforementioned U S Patent 3,922,485 A start-ofscan detector 82 is provided adjacent to mirror 72 to provide a signal when the scan on drum 76 (a portion 30 of the xerographic processor 77 shown in Figure 2) is initiated, and an end-of-scan detector 84 is provided adjacent mirror 72 to provide a signal when each scan line is completed It should be noted that although a single polygon scanner is shown for both input and output scanning separate polygon scanners which are synchronously driven may be utilised.

Reference is made to US Patent 4,012,585 which describes, inter alia, scanning an 35 unmodulated laser beam at a reading station for reading a stationary document thereat and directing a modulated laser beam to an imaging station for reproducing the document image thereat and which utilizes a single scanner element.

Figure 2 is an optically-simplified version of Figure 1 and further shows, in a simplified form, the electronic input scanning signal processing, storage and output scanning functions 40 of the present invention.

The tangential speed of drum 76 of xerographic processor 77 is assumed to be 12 "/second for purposes of the calculations to follow, but other speeds may be used The paper feed for both simplex (printing on one side of the output paper) and duplex operation printing on both sides of the output paper) is provided, for example, by the Xerox (registered 45 trademark) 4000 copier manufactured by the Xerox Corporation, and is initiated on demand under control of the system microprocessor controller 90 It should be noted that the function of microprocessor 90 is that of system management and, when properly programmed, controls the operating sequence of the entire system It also sets up the appropriate operating parameters derived from user controls on panel 92, such as 50 magnification ratio, mode of operation normal or reverse scanning mode, etc In general, the system controller issues appropriate commands to xerographic processor 77, receives status signals therefrom, issues scan and storage control parameters and the start-of-scan signals receives status signals from the rest of the system and, of course, interacts with the user panel 92 Any properly programmed microprocessor, such as the Intel 8080 or the 55 Motorola 6800, or minicomputers such as the Nova series manufactured by the Data General Corporation, Southboro, Massachusetts, can perform these functions Since the present invention is directed to the general interrelationship of the system elements, a specific description of the microprocessor system controller 90 and the operating software therefor is not set forth herein 60 It should be noted that the dimensions and the calculations that follow are approximate and are set forth for illustration purposes only.

In the preferred embodiment, input scanning is provided on a flat platen 62 ( 14 " x 17 " for example) the scan mirrors moving across the short ( 14 inch) dimension of the platen as shown in Figure 1 to provide for v scanning 65 1 601 324 The X direction scanning in the long ( 17 inch) dimension, is produced by a multifaceted rotating polygon 28 having 26 facets The actual total length of scan is 17 85 inches, which provides 0 43 inches over-scan at each end of the 17 " platen, which allows the scan clock generator 94 to be resynchronized prior to the start of the next scan line.

Resolution in the X and Y directions of scan is assumed to be equal That is, the bits/inch 5 (pixels/inch) in the X direction equals the lines/inch in the Y direction for both input and output scanning.

For a given output paper size, the output scan density (this refers to resolution and not optical density) is constant, with reduction in image size being accomplished in accordance with the present invention, by reducing the input scan density (resolution), the number of 10 pixels per output page being independent of the reduction ratio selected Reducing the input scan density in the Y direction is accomplished by increasing the Y scan mirror velocity by operator selection of a desired magnification value (the range, for example, being from 1 0 to 0 61), the input scan density in the X direction being reduced by decreasing the number of bits/inch in the X-scan direction by varying the scan clock 15 generator 94 by the magnification ratio selected This allows independent control over the reduction/magnification in the X and Y directions and makes good use of the capacity and bandwidth of the storage system 96 the storage system preferably utilizing a magnetic disc 97.

Images scanned and readout by photodetector 66 are stored in uncompressed, binary, 20 digital format preferably on a dual platter, four track parallel, moving arm magnetic disc system 96 via synchronizing buffer 98 the tracks preferably being formed in a helical pattern The total capacity of the disc 96 preferably is a pproximately 8 x 108 bits, which allows storage of 48 8-1/2 x 11 " impressions (pages) scanned at approximately 423 lines/inch The average data bit rate to the disc system 96 (the system including a magnetic 25 disc 97, positioning arms disc drive etc) is assumed to be 23 59 megabits/second.

Synchronization in the scanning system is derived from the disc system itself A primary clock rate of approximately 28 62 megabits/sec is formed by the timer block 100 in conjunction with the disc and is used to control the recording of information thereon from buffer 98 This clock rate, which will also be synchronous with data read from the disc 97 30 (since the second disc is identical to disc 97, only disc 97 will be referred to hereinafter), is counted down in timer (or clock) 100 to produce appropriate two-phase AC signals to drive a synchronous scanner motor 40, the Y scan mirror drive motor 52 and appropriate clock signals to synchronizing buffer 98 The clock signal from scan clock generator 94 (used to control the timing of data that modulate the laser beam on output scanning and to sample 35 the photodetector signals on input scanning) is generated in bursts, under the control of the start-of-scan and end-of-scan photodetectors 82 and 84, respectively The scan clock generator 94 is therefore slaved to the speed of polygon 28, which in turn is derived from the disc system 96, the scanning system timing therefor being synchronized with the disc speed The input scan speed relationships are chosen to cause data to be generated at an 40 average rate equal to the ability of the disc 97 to store it If the rotational speed of disc 97 were to change slightly, the scanner 28 and scan clock 94 would follow the change This synchronous system timing method allows the size of the synchronizing buffer 98 to be significantly reduced in size (and cost) and substantially less than the capacity of one helical turn on the disc 97 (as will be set forth hereinafter, one turn of the disc 97 is capable of 45 storing 4 (surfaces) x 48 (sectors per turn) x 4096 (bits per sector), which is 48 times less than the size of the synchronizing buffer which preferably will be utilized) Synchronizing buffer 98 is required because the peak data rate during input scan is approximately 38 megabits/second, which is higher than the rate that disc 97 can accept the input data (approximately 28 megabits/second) The average bit rates over a number of scan lines, 50 however, will be approximately equal Further, synchronizing buffer 98 smooths out any gaps between sectors on disc 97 the sectors including 4096 data bits, when the reproduction system is in the print mode, the system controller 90 preventing gaps (and sector headings, labels, etc) from being stored in the synchronizing buffer during the print mode of operation 55 The time to scan an original on the input platen 62 (see Figure 1) is selected to be the same as the time required to expose the xerographic drum 76 in xerographic processor 77 to reduce the time required for output and the size of synchronizing buffer 98.

The following relationships are given to provide an indication of system performance.

The following definitions are useful 60 1 601 324 ABR BPS BPP CPPS DBC DPC DR Lp L, M N S Di S Do, SLS SPB Ri Vd VP V Vp Vv Wp = Average bit rate for magnetic disc 97 (bits/sec) = Bits per scan line = Bits (pixels) per page = Clock pulses per sector for the magnetic disc 97 = Disc data capacity in bits = Disc data capacity in pages = Divide ratio for generating polygon drive frequency = Output paper length (in) (Parallel to axis of xerographic drum 76).

= Input platen scan length including overscan (in) (Assumed to be 17 885 in) = Magnification ratio ( 1 O to 0 61) = Number of facets on the polygon scanner 20 (assumed to be 26) = Input scan density (lines/inch or bits/inch) = Output scan density (lines/inch or bits/inch) = Scan lines per second = Peak input scan bit rate (bits/sec) = Xerographic drum surface velocity (in/sec) (assumed to be 12 ips) = Polygon angular velocity (rpm) = scan velocity (in/sec) = Output paper width (in) Therefore, from the geometries and characteristics obtained:

of the system, the following are (a) Output scan density S Do = lABR/((Vd)Lp))l 1/2 (b) Polygon rpm required for output scan Vp = 60 (SD,) (Vd)/N (c) Bits (pixels per output page) BPP = (S Do,) Wp) (S Do,) (LP) (d) Input scan density S Di = (SD,,) (M) or M = S Di/SD),, (e) Input Y scan velocity Vy = Vd/M or M = Vd/Vy: S Di being inversely proportional to Vy.

(f) Scanner rpm required for input scan Vp = 60 (SLS)/N= 60 (S Dj) (V,)/N= 60 (Sd,) (Vd)/N.

(g) Scan lines per second SLS = N(Vp)/60 = (SD,) (Va).

(h) Peak input scan bit rate SPB Ri = (SLS) (L) (S Di) = (M) (L,) (Vd) (S Do) 2.

(i) The total number of pages that may be stored on the disc DPC = DBC/BPP = DBC/(SD,)Y(Wp) (LP) The following summarizes some of the system characteristics for 8-1/2 x 11 " output paper.

1 601 324 TABLE I

Average bit rate (mbps) 23 59 Output scan density (lpi) (as determined 5 by the speed of drum 76 and the disc clock rate) 422 77 Output scanner velocity (rpm) 11,707 10 Bits/11 " line 4,650 47 Megabits/output page 16 71 Storage Capacity of Disc in pages 48 19 15 Peak input rate (mbs) 38 30 If a maximum reduction factor of 0 61 is assumed, the input scan density in lines/inch and bits/inch is reduced from 422 77 to 257 89 The output copy from xerographic processor 77 20 is still produced at the maximum scan density of 422 77 scan lines per inch The total number of pixels per output page is constant and independent of magnification and therefore allows for a simple and effecti',e way of controlling magnification by controlling input scan density.

The input Y direction scan mirror velocity is increased from 12 inches/second to 19 67 25 inches/second for the 0 61 magnification ratio The peak input bit scan rate accordingly drops from 38 30 megabits/second to 23 36 megabits/second.

When larger output paper is used, the scan line density and disc page storage capacity are reduced Table II lists system characteristics wherein 10 12 " x 14 33 " output paper is used.

Note that since the bit rate is fixed, and the paper area is larger than in the Table I example, 30 the output scan density will be less.

TABLE II

Average bit rate (mbps) 23 59 35 Output scan density (lpi) 370 41 Output scanner velocity (rpm) 10,257 40 Bits/14 33 " line 5,307 91 Megabits/output page 19 89 Storage capacity in pages 40 47 45 Peak input rate (mbs) 29 40 At a reduction ratio of 0 61, the input scan density becomes 225 95 lines/inch with output scanning remaining at 370 41 scan lines per inch 50 Although the embodiment described herein is preferably utilized to provide for electronic precollation (precollation being provided in simplex operation by copying the number of input originals in sequence onto the disc 97 and printing a predetermined number of copies of each sequence via the xerographic processor 77), it should be noted that, by changing control parameters and the software used by the microprocessor 90, many 55 additional features may be provided e g providing a small alphanumeric display for interactive guidance for the system user; a small portion on the large disc capacity can be used to store statistics on system use: the disc could be used to store software diagnostic routines to be used by the microprocessor 90 for trouble diagnosis; a scan density compatible with easy conversion to facsimile could be selected, etc 60 The disc 97 to be utilized is assumed to comprise two platters (four surfaces) recorded and read in parallel, one surface 99 of which is illustrated in simplified form in Figure 3 The data are recorded, for example, in 1024 discontinuous sectors 101 within angular area 102, 48 such angular areas being formed in band area 103 around the disc circumference (approximately 50,000 sectors thereby being provided) Each sector 101 is subdivided into 65 1 601 324 three main sections The first section contains a space 104 for a fixed header identifying the sector number The second section 105 is a rewritable control area of 128 useful bits identified as "label" The third section 107, separated from section 104 by gap 111, is the normal data area of 4096 data bits There are 48 such sectors per turn, each sector being separated by gaps 111 Information is preferably recorded in a spiral pattern (similar to a 5 phonograph record) with a total of 1024 active data turns The spiral type pattern (track) allows data to be read continuously with the disc read/write heads 116 following the track as in a phonograph record The header area 104 of each sector may be arranged to contain a pattern that is used to servo control the radial position of the recording-playback head to allow it to follow the spiral data path 10 The number of circumferential clock periods (not shown in the Figure) required in each sector for gaps, header and label (error detection and correction bits may be provided if desired) is assumed to be 872 Therefore, the total sector length is 4968 clock periods Table III summarizes typical performance characteristics for the disc system 96:

15 TABLE III

Data bits/sector (each surface) 4,096 Clock periods/sector 4,968 20 Data bits/sector (four surfaces) 16,384 Sectors/turn 48 25 Turns/surface 1,024 Data bits/turn (on each surface) 196,608 Data bits/turn (four surfaces) 786,432 30 Average data bit rate/surface (mbs) 5 89 ( 30 x 48 x 4096) wherein the disc rotation rate is 30 revolutions per second 35 Average bit rate for four tracks (mbs) 23 59 Peak bit rate/surface (mbs) 7 15 ( 30 x 48) ( 4096 + 872) 40 Total peak bit rate (mbs) 28 61 Total data capacity (bits) 805,306,365 45 Although not considered part of the present invention, it should be noted that the large size of the data blocks in this system make the use of isolated and burst error detecting and correcting codes efficient and attractive.

The seek operation, wherein the radial disc arms seek the starting sector on the disc 97, is defined by specifying a unique sector number out of the total of 49, 152 sectors along the 50 spiral track by the system controller 90 and having a controllerspecified acceleration motion to enable the disc arms to locate the correct sector New information (representing images in this system) is written directly over old data without a separate erase pass to save system time.

Referring now to Figure 4, it should be noted that signals to and from the microcode 55 programmed microprocessor system controller 90 are indicated in the Figures by circles adjacent to a label of a function entering or coming from a particular electronic subsystem block.

Direction control device 121 receives an input (Y scan drive frequency) on lead 122 from reduction counter 130, the system controller 90 introducing the signal "start scan" on lead 60 124 The velocity and direction of the Y scan motor 52 are set up by the system controller The scan velocity (Y scan drive frequency) is determined by the "magnification ratio" control parameter on lead 128 specified by an operator via panel 92 (Figure 2) which is used to determine the clock frequency division ratio in the system timing counter 129 (logically a set of counters, the count ratio being changed by the selected magnification ratio) and the 65 1 601 324 reduction counter 130 The magnification ratio signal is applied to reduction counter 130 via lead 127, a reduced clock signal being applied thereto from system timing counter 129 via lead 119 The direction control device 121 causes a Y scan pass to be initiated after the 'start scan' signal and direction information are provided by the system controller 90, the direction information being initially set up by an operator via panel 92 Logic circuits within 5direction control device 121 determine the proper polarity of the Y scan drive waveform applied to motor 52 for the correct direction of scan In the normal (noninverted mode) it is assumed that the Y direction of scan is in the +Y direction from an initial position 61 (Figure 1), whereas in the inverted mode of operation the Y direction of scan is in the -Y direction from initial position 63 10 The start of Y scan time is derived by the system controller 90 from information it has about the starting sector number for the next page of information to be entered into the disc 96 during input scanning The system controller 90 receives information about where the disc 96 is as it rotates from the header and check logic block 131 on output lead 132 The controller 90 checks the Y scan status from the direction control block on lead 126 prior to 15 initiating a start-scan command, to be sure the scanner is in the correct home, or initial position The correct home position obviously is dependent upon whether scanning is to proceed in the normal or reversed directions.

It should be observed that, because the input scan line density must be changed (reduced) to vary the magnification ratio, it is preferable to change the size of the scan spot for input 20 scanning, in order that the scanning spot cover the entire area of the document, thereby maintaining the optimum ratio of scanning aperture size to scan line density To increase the Y dimension of the scan spot optically (anamorphically), optical aperture control 133 is utilized during input scanning, the aperture control increasing the size of the scanning spot (in the Y direction) associated therewith via a signal from system controller 90 on lead 135 25 On input scanning, the effective X dimension of the spot (in the direction of high speed scan) may be controlled by changing the electronic bandwidth of the aperture control 134 following the photodetector 64 via a signal from system controller 90 on lead 135 During output scanning, the effective size of the spot in the X direction (which is maintained essentially constant because the output scan line density is maintained constant) is 30 controlled by the timing of signals supplied to the acousto-optic modulator 32 via lead 125 under control of the scan clock generator 94.

As set forth hereinabove with reference to Figure 1, a blue and red laser 10 and 12 are used for input scanning to avoid color blindness which would occur if monochromatic illumination were used In the system shown, both lasers are used, for input scanning and 35 the blue laser is used for output scanning.

Although it would be cost-effective to use a single polygon X scanner 20 for both input and output scanning, it may be preferable to use a second polygon which utilizes a separate two-phase synchronous drive motor.

In the single scanner design, one pair of scan synchronizing detectors will normally suffice 40 i.e end-of-scan detector 84 and start-of-scan detector 82 Signals from these two devices allow the generation of precisely-controlled streams of "bit clocks" for sampling the signal from photodetector 66 on input scanning, or controlling the timing of image data fed to the laser modulator 32 on output scanning It should be noted that the system mode of operation (whether input scanning or output printing) is determined by the operator via 45 panel 92 The scan clock frequency is controlled by phase detector 136, start-stop control device 137, voltage controlled oscillator 138, linearizer 140, and bits/inch counter 142.

The voltage-controlled oscillator 138, oscillating at a preset frequency, does not operate continuously, but is released to start oscillating on each scan by the start-of-scan pulse and is stopped at the end of scan via start/stop control 137 The phase comparison in phase 50 detector 136 is also initiated when the start of scan pulse is received via lead 141 The count down ratio of the bits/inch counter 142 is set by the system controller 90 according to the operator selected magnification ratio and output paper size utilized The preferred range is from approximately 423 bits/inch to approximately 226 bit/inch (input scan onto 14 33 " paper at magnification of 0 61) When the preset number of bits (voltage cycles) (bits/inch 55 times the input scan length including overscan) from oscillator 138 have been counted in the bits/inch counter 142, a pulse is coupled to the phase detector 134 via lead 144 If the average signal frequency from oscillator 138 is correct, a pulse will be received from the end-of-scan detector 84 at the same time If, for example, the polygon 28 had speeded up slightly, the end-of-scan pulse will arrive at the phase detector before the bits/inch counter 60 pulse on lead 144 This will cause the phase detector 136 to generate a voltage error signal to increase the frequency of oscillator 138 Note that there are 26 such samples of scanner rotation rate for each rotation of the scanner 28, because it has been assumed that scanner comprises 26 facets.

The bits/line counter 146, synchronized by oscillator 138 via lead 145, counts down from a 65 1 601 324 preset count which corresponds to the various sizes of output paper to which the developed image formed in the xerographic processor 76 is transferred by standard techniques in the preset mode The range (count) is 4656 to 5312 bits/line, which is less than the range for counter 142 because the lattr count is preset on the basis of the input platen scan line length and including overscan These numbers are slightly larger than those listed in Tables I and 5 II in order to be compatible with the operation of the synchronizing buffer 98, the number of bits/line being rounded upward to the nearest multiple of 16.

The linearizer 140 generates a second input to oscillator 138 via lead 147 to correct for non-uniform velocity of the scan spot, the bits/line counter 146 providing a signal to linearizer 140 via lead 149 to provide an indication where in the scan line the spot is located 10 at any instant It has been observed that the instantaneous scan velocity normally is higher at the edges of a scan than at the center of the scan Even though the input and output scan nonlinearities might compensate each other, electronic linearity correction of the image data stored in the disc by scan clock variation may be preferable to allow later coupling between machines with different scan geometries 15 The scan clock gate 148 releases precisely-timed bursts of clock pulses on lead 200 at the start of its countdown cycle, ranging in frequency from 38 30 to 17 93 megabits/second as determined by the system controller 90 (output paper size and magnification ratio) The number of pulses in the clock burst is determined by the countdown ratio set in counter 146.

The scan clock gate 148 is used to control the timing of loading the synchronizing buffer 20 assembly 98 with signals from the photodetector 66 in the input scanning mode, the unloading of the synchronizing buffer 98 to the disc system 96 for input scanning being under the control of the disc clock, to be described hereinafter.

The threshold detector 150, with its input control parameter on lead 151, is used in simple signal-processing operations to produce, in effect, extremely high gamma A threshold 25 slicing level may be modified under user control to help remove background and otherwise clean up inferior originals Existence of the image information in electronic form makes possible a wide range of image enhancement techniques.

The timing of the entire scanning system is slaved to the disc clock On input scanning, signals from the photodetector 66 will come in bursts because (for 11 " paper) the active scan 30 time is only 11/17 855 of the total scan line period for the case of no reduction This produces a peak input scan bit rate (SPBR;) of 38 30 megabits/second Similarly, the disc input and output data flow in bursts to compensate for the overhead necessary for sector gaps, headers and labels The peak disc data rate is 28 62 megabits/sec Therefore, total peak instantaneous bit rate for the synchronizing buffer is 38 30 plus 28 62 megabits/ 35 second The average input rate is equal to the average output rate for most modes of operation and is equal to 23 59 megabits/second An exception occurs when the reduced image of the 1 " x 17 " input platen is smaller than the output paper size, as determined by the operator-selected magnification ratio and paper size In that case, "white border bits" are generated to fill the output page as is described hereinafter 40 Figure 5 shows some of the functional blocks enclosed in the dotted outline corresponding to the synchronizing buffer 98 of the block diagram of Figure 4 The buffer storage 170 required to accommodate the bursts of data is assumed to be made up of 16, 1 K random access memory (RAM) chips Each input and each output operation of the RAM handles 16 bits in parallel It is assumed that chips operating at 200 nanoseconds full cycle 45 time will be utilized This will provide a peak rate of 80 megabits Serialto-parallel shift register 172 and parallel-to-serial shift register 174 make the necessary conversions at input and output respectively for the random access memory 170.

For the non-inverted first-in, first-out operation mode of operation, a load address counter 180 selected by address selection gates 181, sequences through the 1024 addresses 50 in RAM 170, sequentially and circularly to load data therein from the threshold detector in the input scanning mode of operation Similarly, an unload address counter 182 provides sequential unload addresses for the RAM 170 under control of address selection gates 181 when data are to be unloaded to the disc 97.

The data selection gates 186 contain parallel/digital gates that switch the input and output 55 bit streams to and from the synchronizing buffer 98 For input scanning, the peak input scan bit rate clock on lead 200 controls the input shift register 172 via the shift register clocks on lead 206 and load address counter 180 timing via the load/unload clocks on lead 204 The peak bit rate disc clock on lead 201 controls output shift register 176 via lead 206, and unload address counter 182 timing via lead 204 The threshold detector 150 (Figure 3) is the 60 input data source to the data selection gates 186 via input shift register 172 and holding register 173, the output image data from RAM 170 going to disc 97 Similarly, for output scanning (printing) the disc clock on lead 201 controls the input to RAM 170 via shift register 172 and load timing via load address counter 180, while the scan clock on lead 200 controls the output of RAM 170 via output shift register 174 and the unloading address 65 1 601 324 10 counter timing via counter 182.

When a bound volume is placed on the input platen 62, successive pages of the volume may be placed upside down on the platen to make use of the book edge feature incorporated in copiers commercially available In order to reverse the image so that all pages will be right side up when the output is generated, the X and Y scan directions both 5 must be reversed (scan inversion is accomplished by operator selection of a "Scan Invert" button (not shown) on panel 92 Note that if only the Y scan direction were reversed a mirror image of the document scanned would be reproduced) Although the Y scan direction can be changed by appropriate control of the Y scan direction control device 121, thereby resetting the initial start position and direction of scan mechanically changing the X 10 scanning direction is not feasible due to the inertia and high operating speeds of the scanner 28 The X-scan direction is therefor reversed electronically as follows: For an 8-1/2 x 11 " input document, it is assumed that approximately 291 sixteen bit words comprise one scan line in the 11 inch X-scan direction During the input scan (the system is assumed to be in the inverted input scanning mode) load address counter 180 via address selection gates 181 15 causes the input scan information from photodetector 66 ( 291 sixteen bit words) to be stored in sequence, for example in storage locations 0 to 290 in RAM 170, at least one complete scan line being stored therein Lead 230 is appropriately energized to allow storage to be accomplished when the store mode of operation is selected Preset address counter 179 is caused to be set to a first preset address 290 in the inverted mode of 20 operation, a signal on lead 177 causing the unload address counter 182 via address selection gates 181 to count down sequentially from storage location 290 (i e 289, 288) such that the scan line information is read out word by word in the reverse order in which it was stored, an appropriate control signal being applied to lead 230 to enable RAM 170 to be read out The information read out is coupled to output shift register 174 via lead 175, data 25 selection gates 186, and output holding register 183 and thereafter to disc 97 As shown in Figure 6, the output shift register 174 is coupled to the 16-bit output holding register 183 and comprises four shift registers 240, 242 244 and 246 When information is to be recorded on discs 97, appropriate control signal from system controller 90 is applied to register 174 on lead 250 to enable the information to be read out in four-bit blocks to be applied to the disc 30 write block 222 and thereafter to be applied to the four recording surfaces of the discs 97 via write amplifiers 223 (Figure 4) When the information read out from RAM 170 is to be applied to modulator 32 and thereafter reproduced by xerographic processor 76, the signal on lead 250 enables the information to be read out serially on lead 125 In a similar manner, although not shown in the figure input shift register 172 is adapted (via a signal from system 35 controller 90 on lead 251) in the input scan mode, to convert the input serial data stream into 16-bit parallel format and to convert the four-bit word from the discs 97 via amplifiers 225 and data recovery circuits 220 into 16-bit parallel words in the print (write) mode.

The next scan line is recorded in locations 291 through 580 in RAM 170 and the preset address counter 179 is set to address 580, the data in these addresses being read out in a 40 manner as described hereinabove with reference to locations 0 through 290.

In the inverted mode of operation, the bits in output shift register 174 are shifted from left to right and read out on lines 239, 241 243 and 245, whereby each bit in the scan line is transposed for reverse scanning In the normal (non-inverted) mode of operation, the bits in each scan line are shifted right to left and read out on lines 237, 247, 249 and 253 with no 45 transposition of the bits comprising the scan line occurring In other words, shift register 174 is bidirectional, data bits being shifted out right-to-left in the inverted mode of operation, whereas the data bits are shifted left-to-right in the normal FIFO (first in first out) mode of buffer operation It should be noted that input shift register 172 need not be bidirectional because in the print mode of operation, the transposed bits stored on the discs 50 97 will be in the correct sequence when read out.

When the system is in the print mode, as determined by operator energization of a PRINT" button on panel 92 (not shown), the output from discs 97 is read out via read pre-amplifiers 225 and initially stored in memory 170 in the address specified by load address counter 180, counter 180 being selected by address selection gates 181 to store 55 information in RAM 170 To unload data to the modulator 32, unlead address counter 182 is selected by gates 181 and caused to transfer the information in RAM 170 via data selection gates 186 and output holding register 183 to output shift register 174 It should be noted, as set forth hereinabove, that because the scan lines have already been reversed prior to being stored on disc 97 unload address counter 182 is not caused to count down by 60 a signal from buffer control 202 on lead 177 The data which are being read out therefore are electronically reversed in the X-scan direction.

The scan clock on lead 200 is utilized to control the timing of loading the RAM 170 with signals from the photodetector 66 on input scanning, the unloading of the RAM 170 being controlled by the clock signal derived from the disc system 96 on lead 201 For output 65 11 1 601 324 11 scanning, the loading of the RAM 170 is controlled by the clock signal from disc system 96, whereas the unloading of the RAM 170 is controlled by the scan clock signal on lead 200.

The load and unload address clocks are applied to lead 204 and shift register clocks are applied to lead 206 via synchronizing buffer control 202.

The header and check logic 131 (Figure 4) is connected to the shift registers 172, 174 via 5 leads 227 and 228 to enable the acquisition and loading of header and control information from the data stored in the shift registers The system controller 90 will supply header and check logic 131 with the following parameters: Lines/page, bits/line, and page start sector number, which in turn modifies the data stored in the RAM 170 with this information prior to loading the discs 97 Since four surfaces of the disc are used in parallel, the basic disc data 10 block is 4 x 4096 = 16,384 data bits which corresponds to the timing of one disc sector.

Because the largest number of bits in a scan line may be greater than 4096 data bits, the start of successive scan lines may not occur at sector boundaries It is assumed that the first scan line of each page may start at a sector boundary identified by the page start sector number 15 The label information associated with each sector may identify the number of lines remaining in the current page and the location of the boundaries between successive scan lines for each sector This information can be thought of as completely defining the format and other relevant information about the data to follow.

The header and check logic block 131 will check sector identification and will preferably 20 also verify data integrity by generating and comparing error detection and correction redundancy patterns by standard computer techniques, although this does not form part of the present invention Sector number checking is aided by the availability of the current sector position of the disc derived from the system timing counter 129 of Figure 4 which supplies sector pulses per disc revolution to sector counter 240 via lead 241 (approximately 25 50,000 total for 1024 turns) As shown pulses from timing counter 129 are also applied to the buffer control 202 (approximately 28 2 megabits/sec) and header and disc logic 131 (one index pulse per disc revolution) via leads 201 and 242, respectively The clock for disc data recovery circuit 220 is derived from the recorded data during a read operation, the clock for the disc write logic circuits 222 being derived from the system timing counter 129 during 30 recording Each of the four independent data recovery circuits 220 will generate its independent read timing clock, although the disc system timing clock controls the combined output data stream as it is passed to the main synchronizing buffer 98.

The header and check logic 131 will issue sector number commands to the seek control block 206 via lead 224 that controls the positioner (not shown) for disc arms 115 'Seek 35 complete' status is indicated to the system controller 90 via lead 207 when the commanded sector has been acquired by the seek control 206 The system controller 90 can then issue the start-scan signal to the seek controller 206 to allow the disc heads to follow the spiral track, either for recording or playback of the disc data.

The position detector 210 generates radial head position error signals (i e radial 40 deviation from the spiral track) from the playback voltage on lead 211, which may be generated by the position control pattern permanently recorded in the fixed header segment of each sector Timing for this operation is derived from the system timing counter 129 via lead 228.

The gear clock PLL 212 is a phase-locked loop frequency multiplier used to generate the 45 28.62 megabit/second basic system timing signal The input for this block is derived from a multi-toothed gear mounted to the disc drive hub, (a plurality of teeth corresponding to each of the 48 sectors per turn) a magnetic detector pickup mounted on the disc support structure generating a pulse as each tooth rotates therepast, a pulse stream thereby being generated having a frequency proportional to the rotational speed of disc 96 A typical 50 input to gear clock phase-locked loop 212 is 192 pulses/second In order to provide the required maximum system pulse rate of 28 62 Mbs, gear clock 212 multiplies the input pulse rate by a factor of approximately 5500 The detector is separated from the recording discs 97 and is always available whether the disc system 96 is reading or writing It is to be noted that system timing counter 124 supplies a plurality of pulse signals, including pulse rates 55 reduced in frequency from the 28 62 Mbs input on its output counter leads to provide appropriate timing signals to the various system elements For example, a frequency of 100 Hz is generally required to drive motors 40 and 152 The count ratio of counter 129 is varied by the magnification ratio on lead 128.

Three basic modes of operation are involved in the operation of the present system The 60 first is a preparatory one noted as 'job set up', the second is 'input scanning' where originals are scanned and written on the disc, the third is 'output scanning' where copies are produced xerographically.

During the 'job set up', the system controller 90 furnishes a starting sector number for the first page The disc seek control 206 will find that sector issued by header and check logic 65 12 1 601 324 12 131, and then set up the idle mode holding pattern and indicate a 'seek complete' condition to the system controller 90 on lead 207 Similarly, the proper timing ratios will have been issued to cause the scanner 28 rpm to be selected and stabilized The scan clock phase-locked loop will be generating the correct number of bits/inch and bits/scan line for the selected magnification ratio and output page size, the proper scan clock thereby being 5 applied to lead 200 The header control logic 131 will have been set up with the bits/scan line and scan lines/page parameters The controller 90 will generate the sector number to start each page, and these will be provided sequentially to the seek control 206 as the job progresses in order to allow for electronic precollation The controller 90 has been given the number of pages/book and the number of books (copies)/job by the user through the 10 control panel 92.

The controller 90 may derive or be told by the operator of the simplex/duplex status of each output page and computes appropriate page start sector numbers to provide the optimum sequence for duplex output production (if the xerographic processor 77 is capable of duplex operation) 15 After the job is set up, the input scanning operation can proceed The operator places his first original on the platen and pushes either the "Normal" or "Invert" scan button on panel 92 This causes the system controller 90 to initiate a scan in either the + Y or -Y direction (Figure 1) at an initial starting position The Y scan motor 52 will start with a lead time (with respect to the arrival of the page start sector number of the disc) to allow the Y scan 20 mirror to accelerate and stabilize at the selected velocity (as determined by the selected reduction ratio) and depending on normal or reverse scan direction, both parameters being operator initiated As was mentioned hereinabove for reverse scanning, one or more complete scan lines must be loaded into the synchronizing buffer 98 prior to the arrival of the page start sector at read heads of the disc At this time, the disc system 96 will demand 25 output from the buffer 98 in inverted (or LIFO) mode The data flow into the buffer 98 from the photo-detector 66 is timed according to the scan clock synchronization circuits and is not determined by the position of the Y scan drive motor 52 Variations in the position of the Y scan mirror at the start of electrical scan are equivalent to a shift in the position of the original on the platen (in the Y direction) and do not affect the synchronizing buffer A 30 position detector can be provided to check the timing of this operation to allow the system controller 90 to adjust the lead time parameter.

The system runs to the end of the page and the disc system 96 seeks the next page start sector number If the input scanning is being done for simplex output printing, the next page will start at the next sector following the last sector used in the previous page For 35 duplex output, appropriate page start position interlace will have been generated by the system controller 90 That is, the sequence of pages along the spiral track on disc 96 will be arranged during input scanning for the benefit of high throughput output.

Operation during the third mode, output scanning, is similar In the idle condition, the disc system 96 acquires the page start sector The paper feed from either the duplex 40 recirculation paper path or the normal paper supply from xerographic processor 77 can be triggered on demand from the system controller 90 Collation is then done electronically as each page is read from the disc in sequence to form a book, the number of books that will be generated being dependent on operator selection of the appropriate buttons on panel 92.

Interleaved input and output may be required for example, when a job requiring 25 45 copies of a 13-page original has been loaded and the system is in the output (print) mode.

The operator then wishes to load a new job This fact, plus the other normal job set up quantities, are entered via the control keyboard 92 and the first original of the new job is placed on the platen 62 When the start button is pushed, the system controller 90 finishes printing the output page in process and then momentarily interrupts the output printing 50 operation The system controller 90 resets the scan clock rate and an input scan takes place.

The system then immediately resumes output printing while the operator changes to the next original on the input platen the process being repeated until the first job is completed and all the originals of the new job have been scanned.

The following sets forth an analysis of some of the factors that may be utilized to 55 determine the size of synchronizing buffer 170 and the system timing relationships and considers the case of input scanning, using 8-1/2 X 11 " output paper size and the normal (no reduction) mode This appears to place the most stringent demands on the size of buffer Table IV hereinbelow lists some data, (times being in micro-seconds and bit rates in megabits/second) for the system described hereinabove 60 1 601 324 TABLE IV

Total sector time 106/30 x 48 694 44 Active sector time 16,384/28 61 572 55 S Inter sector time (assumed gap time) 121 89 5 Total scan line time 60 x 106/(N) (VP 197 11 Active scan line time 4656/38 30 121 58 Inactive scan line time 75 53 10 Total bits/scan line 4656 Total bits/sector 16,384 Number of scan lines/sector 3 51 Peak bit rate to disc 28 62 15 Peak bit rate from scanner 38 30 The most stringent demands made on synchronizing buffer is in the inverted page mode where at least one complete scan line must be loaded into the buffer memory 170 prior to removal of information for the disc 96 The minimum lead time for information supplied to 20 the buffer memory from the input scanner that is required to prevent the disc unload requirements from overtaking the data available in the buffer should be determined.

Time will be measured, in the following calculation, with respect to the instant, time to, that data bits must be supplied to the 96 disc from the buffer 170 The time to load the 4656 bits of the first scan line into the disc 96 is 25 4656/28 62 = 162 71 microseconds.

The disc therefore accepts a line of data in less than the 197 11 microseconds total scan linetime Therefore, when disc 76 is ready to receive the beginning of the fourth scan line near 30 the end of the first disc sector, which will occur at t 4 = 3 x 162 71 = 488 12 microseconds after to, the input scanner at t 4 must have loaded four complete scan lines into the buffer 35 The time required to load N scan lines into the buffer is given by n ( 197 11) -75 53 If TL denotes the lead time in microseconds with respect to the start of the data block (to), 40 4 ( 197 11) 75 53 -TL = 488 12, TL = 224 79 45 This is shown in Figure 8, in which 1, 2 3 represent scan line numbers (both disc and scan), and x represents inactive time.

Thus TL represents the latest start time at which input scan signals may start to enter the 50 synchronizing buffer 170 measured with respect to time t 0, the initiation of the unload to the disc, the unload initiation process being controlled by buffer control 202.

The earliest start time as shown in Figure 9, is determined by the upper limit on the size of buffer 170 After feeding three scan lines into the buffer without removing any information for the disc 96 there will be 55 ( 16,384 3 x 4656) = 2416 bit positions left in the buffer 170 The start of transfer to the disc 96 from the buffer (to) will occur at some time during the loading of the fourth scan line into the buffer 170 Prior to t, 60 the net input rate of the buffer 170 will be 38 30 megabits/second input The differential input rate after to will be 38.306 -28 62 = 9 68 megabits/second.

14 1 601 324 14 The active time for scanning the fourth scan line ( 121 58 microseconds) can be divided into two intervals, t, + t 2.

tl + t 2 = 121 58 5 The total net increase in bits contained in the buffer 170 during the input of the fourth scan line cannot exceed the remaining capacity of the buffer 170 ( 2416 bits) Therefore, t, x 38 30 + t 2 x 9 68 = 2416.

10 t, = 43 30 microseconds.

The earliest lead time that the scanning input can start is, therefore, TE = 3 x 197 11 + 43 30 = 634 63 microseconds 15 The optimum lead time with respect to to normally would be considered to be the average of the earliest and latest lead times, i e 430 microseconds However, the scan line start times precess with respect to the disc sector start times The optimum enable time for allowing the input scanner to start loading will leave the buffer 170 equal margins before the 20 earliest allowed time and after the latest possible occurring time (after enable) These possible data load start times are separated by one total scan time or 197 11 microseconds.

Thus, if m = margin time, 2 m + 197 11 643 63 224 79 25 m 110 87 Therefore, for the case of 8-1/2 x 11 " output paper size inverted scanning, no reduction, the optimum time to initiate input scan loading of the synchronizing buffer 170 is 30 TE -m = 532 77 microseconds before to.

These relationships are shown in Figure 10.

An example of how the normal first-in first-out operation might function during 35 non-synchronous interlaced load and unload cycles is set forth hereinafter Assume again the 8-1/2 x 11 " output paper, no reduction, input scan case 16-bit words will be available in the input data holding register 173 (Figure 5) at intervals of 16/383 = 0 4178 microseconds 40 This information must be loaded into the RAM 170 at some time before the next 16-bit data word is assembled in the input shift register, i e before 417 8 nanoseconds have elapsed.

Similarly, the output shift register 174 will require a new 16-bit word from its output holding register 183 at intervals of 45 16/28 62 = 0 55905 microseconds.

If there is a coincidence in the time at which an input word is ready and an output word can be accepted, input is given priority, since inputs come faster, when simultaneous 50 requests for RAM operation occur Table III illustrates (in simplified terms neglecting logic delays of a few nanoseconds) a possible sequence of events For this example, it is assumed that an internal sync buffer logic clock on lead 201 running at 57 24 mega-pulses/sec, instead of the 28 62 megabits per second set forth hereinabove, is made available by the gear clock phase-locked loop 212 55 Therefore, internal events can be initiated only at the times of occurrence of these clock pulses, or about every 17 47 nanoseconds Each RAM memory cycle (either store (load) or read (unload)) is assumed to take 200 ns Assume that the memory cannot be recycled until at least the second clock pulse occurs following the completion of any memory cycle, or after any new non-synchronous memory cycle request is generated The times listed for 60 completion of memory cycles, and also for the availability of input words, are not synchronous with the internal buffer clock and are designated as "NS" in the table In this arrangement, RAM output requests will occur synchronously at intervals of 32 internal clock periods.

For purposes of identification, the input words being loaded are designated as 101, 102, 65 1 601 324 etc, while the words being unloaded are 1, 2, 3 etc.

TABLE III

RAM (FIFO) Timing Example Internal Clock Pulse Number 1 NS 14 NS NS 32 NS 39 NS NS 53 NS 65 NS NS 78 79 NS Time nsec 17.5 217 5 227 1 244 6 417 8 444 6 454 2 471 7 559 0 671 7 681 3 698 8 835 6 898 8 908 4 925 9 1118 1 1125 9 1135 6 1153 0 1253 4 1353 0 1362 7 1380 1 1580 1 No RAM Operation resync.

start load 101 end load 101 resync.

start unload 1 Holding Register Sta Input Output 101 ready 1 ready empty 102 ready empty end unload 1 resync.

start load 102 2 ready end load 102 resync.

start unload 2 empty 103 ready empty end unload 2 resync.

start load 103 3 ready end load 103 resync.

start unload 3 empty 104 ready end unload 3 resync.

start load 104 end load 104 activity waiting for empty empty request empty 161606241 RAM (FIFO) Timing Example Internal Clock Pulse Time Holding Register Sta 5 Number nsec RAM Operation Input Output NS 1671 2 105 ready 10 96 1677 1 resync 4 ready 97 1694 6 start load 105 NS 1894 6 end load 105 empty 15 109 1094 2 resync.

1921 7 start unload 4 20 NS 2089 0 106 ready The point to notice is that the FIFO sequence catches up with the combined input and output tasks at 1580 1 nsec after the start of the example It waits for the generation of a new request which comes at 1671 2 nsec when a nonsynchronous load request is generated, 25 and the pattern starts to repeat.

An input scan timing problem occurs when the reduction ratio causes the reduced image of the input to be smaller than the output paper size The size of the original(s) on the platen 62 is of no concern if the cover is closed The video signal variation because of the difference in reflectivity of the platen cover and the unmarked areas of the paper can be set 30 below the slicing level of the threshold detector 150 and should not be noticeable.

Figure 7 (a) is a representation of a reduced image 270 formed on output paper 272 (this can also correspond, for example, to the electrostatic dot pattern formed on drum 76 within the xerographic processor 77) As can be seen, in order to center the image 270 on output paper 272 the left-hand and right-hand borders (as viewed from the paper) 274 and 276, 35 respectively, and the upper and lower borders 278 and 280, respectively, must be appropriately generated to center the image 270.

Figure 7 (b) shows apparatus which may be utilized to center the image 270 shown in Figure 7 (a) The system controller 90 via leads 280 and 282, loads registers 284 and 286, respectively, with appropriate data (dependent on magnification ratio and output paper 40 size) relating to the borders 274, 276, 278 and 280 For the X input scan direction, a problem arises if 17 M<Lp For 11 " paper, this is M< 0 65 (M< 0 84 for 14 33 paper) In these cases, there would be fewer input bits available than is required for one output scan line (SD (L) the input scan bit rate being less than the average disc bit rate.

Register 284 is therefor loaded with appropriate data corresponding to borders 274 and 45 276, the output of register 284 being compared in comparator 290 with information regarding the X position of scan from bits per line counter 146 Register 286 is similarly loaded by microprocessor 90 via lead 282 and is compared with the Y position of scan from Y scan position counter 294 (i e compares the scan position with the known border conditions) When 17 (M)<Lp (determined by system controller 90), the necessary "white 50 margin zeros" are split equally between the beginning and end of each scan line, the output on line 126 being correspondingly controlled Referring to Figure 4 A, the output on lead 126 is coupled to a logic device 300 which comprises AND gates 301 and 303 The output on lead 126 is coupled to one input of AND gate 301 and to an inverting input of AND gate 303 The output from data selection gates 186 is applied to the other input of AND gate 301, 55 whereas a voltage Vc is applied to the other input of AND gate 303 When lead 126 is low AND gate 303 is enabled and passes the voltage Vc to the modulator 32 to cause the laser 10 to generate the necessary white margins (the beam from laser 10 discharges the appropriate margin areas of drum 96) If lead 126 is high, gate 303 is disabled, gate 304 is enabled and the data signals on lead 125 pass to modulator 32 to modulate the laser light from laser 10 to 60 reproduce collated pages in xerographic processor 77.

Although not shown in the Figures, the Y scan position counter 294 is adapted to cooperate with the shaft of motor 52 in a known manner to provide signals representing the Y position of the scan line.

Similarly, for 14 (M)<Wp the width of the platen, as reduced, is less than the output 65 1 601 324 17 1 601 324 17 paper width, when M< 0 61 for 11 " paper (or M < 0 72 for 1433 paper) For this situation register 286 is appropriately loaded with data corresponding to borders 278 and 280 a string of completely-blank scan lines being generated both before and after the Y scan starts and finishes, producing valid data within the width of the image on drum 76.

These procedures will center the reduced image of the platen area on the output page 5 The surrounding white borders will be electronically generated by causing the laser to perform the function of an adjustable fade-out lamp.

It should be noted that the drive frequency for the two-pole polygon motor 40 is Vp/60 Hz In order to generate a two-phase quadrature motor drive signal, a quadruple frequency clock rate is required The correct value will cause scan bits to be generated at the average 10 data rate of the disc Then (BPS) (N) (Vp)/60 = ABR, wherein BPS is the bits per scan line rounded upwards The peak bit rate of the disc 96 is 15 related to the average bit rate by the ratio of the number of clock pulses/sector, CPPS, to the data bit times per sector or CPPS/4096 The polygon drive frequency divide ratio, DR is selected such that l(CPPS) (ABR)/4096)l/DR = 4 (Vp)/60 = 4 (ABR)/(BSL) 20 (N) DR = (CPPS) (BSL) (N)/16,384.

with CPPS = 4968, BSL = 4656, N = 26, DR = 36,707.

25

Claims (1)

1. WHAT WE CLAIM IS:-

   1 A system for copying image-bearing originals, including:

   means for reading an original in two orthogonal directions to produce electrical signals corresponding to a known number of picture elements per unit length of scan; means for storing the said signals; 30 means for accessing the stored signals to write picture elements on a copy medium at a chosen rate, and means for adjusting the effective rates of reading in both directions, relative to the rate of writing, to adjust the size of the copy relative to the size of the original.

   2 A copying system as claimed in claim 1, wherein the adjustable reading speed is able 35 to be increased above a value which corresponds to a unity ratio, wherein the number of pulses per unit length produced by said reading means is proportionally decreased, whereby an image reduced in size compared with the original is reproduced on the copy medium, and wherein the adjustment ratios may be different in the two orthogonal directions.

   3 A copying system as claimed in claim 2, wherein the number of picture elements read 40 in said first and second directions are equal.

   4 A copying system as claimed in any preceding claim, in which the reading means uses at least one beam of coherent radiation.

   A copying system as claimed in claim 4, in which the copy-producing means uses a write beam of coherent radiation 45 6 A copying system substantially as described with reference to, and as shown in, the accompanying drawings.

   7 A method for copying image-bearing originals, including the steps of:

   reading an original in two orthogonal directions to produce electrical signals corresponding to a known number of picture elements per unit length of scan; 50 storing the said signals; accessing the stored signals to write picture elements on a copy medium at a chosen rate, and adjusting the effective rates of reading in both directions relative to the rate of writing, to adjust the size of the copy relative to the size of the original 55 For the Applicant, K.B WEATHERALD, Chartered Patent Agent.

   Printed for Her Majesty's Stationery Office by Croydon Printing Company Limited, Croydon Surrey, 1981.

   Published by The Patent Office, 25 Southampton Buildings, London, WC 2 A IAY, from which copies may be obtained.