Scanning Beam Control System and Method With Bi-Directional Reference Scale

1. Field of the Invention

This invention is directed generally to scanning systems which must precisely locate image picture elements (pixels) throughout a flat format during writing or reading operations. More particularly it relates to the positioning of image pixels in both X and Y directions on a format so as to correct height errors as well as spacing errors with a high order of precision compatible with extremely small spot size.

2. Description of the Prior Art

In high speed scanning systems such as precision plotters, printers, and the like, a number of problems are encountered in exactly locating image pixels throughout an object field format, particularly a flat format. The scanning beam, which is moved transversely across an object plane (here called the X direction) is equally of non-constant velocity and must traverse non-uniform path lengths. Errors can thus occur in the X direction due to optical distortion, scanner speed, projection of the scan onto a flat format, or simply because of vibration of the equipment. Discussion of such problems can be found in an article entitled "Generation of Precision Pixel Clock in Laser Printers and Scanners" by Gerald Toyen, published in SPIE, Volume 84, Laser Scanning Components and Techniques (1976), pp. 138-145.

In the aforementioned article by Toyen, it is proposed to split a laser beam prior to modulation and direct an auxiliary, unmodulated, beam in parallel with a main beam so as to be reflected off a polygon and through an optical grating which functions as a positional reference. Thus by synchronizing data to a pixel clock derived from the reference, one precisely locates pixels in the X-direction whether reading or writing, although the systems may otherwise be substantially different.

The aforementioned auxiliary beam clock does not, however, correct or eliminate height (Y-direction) errors on the format. These errors can occur due to scan inaccuracies such as non-parallelism of the faces of a scanning polygon, vibration, table drive errors, and the like. Since such errors affect the linearity as well as placement of a line of pixels, they determine the vertical resolution of the system, and should be minimized to a degree consistent with the system application. Mechanical and optical systems can only be made highly precise at inordinate expense, while at the same time modern scanning systems may use focused beams of the order of 15 .mu.m in diameter and pixels each 0.0005". Systems of the prior art do not provide a basis for economically achieving such resolutions in both X and Y directions.

U.S. Pat. Nos. 4,306,242, issued Dec. 15, 1981 to E. A. Jeffery; 4,310,757, issued Jan. 12, 1982 to F. T. Check, Jr. et al; 4,311,384, issued Jan. 19, 1982 to W. H. Keene; and 4,311,385, issued Jan. 19, 1982 also to W. H. Keene; disclose additional examples of laser recording systems and the like, which employ an auxiliary or reference beam for clocking purposes. Another example of laser scanning apparatus having beam position means can be found in U.S. Pat. No. 4,279,472, issued July 21, 1981, to G. S. B. Street, while other examples of clocking arrangements for laser beam recording systems, and the like, can be found in U.S. Pat. Nos. 4,212,018, issued July 8, 1980 to M. Ohnishi et al; 3,389,403, issued June 18, 1968 to J. M. Cottingham; 4,285,012, issued Aug. 18, 1981 to Y. Ohara et al; 4,293,202, issued Oct. 6, 1981 to M. Ohnishi et al; and 4,307,409, issued Dec. 22, 1981 to N. L. Greenig et al. U.S. Pat. Nos. 4,274,703, issued June 23, 1981 to T. Fisli discloses an optical system which provides a flat scanning line, and 4,245,228, issued July 13, 1981 to G. F. Cook, discloses a laser plotter construction wherein a modulated light beam is deflected in order to scan a light sensitive medium transversely while sweeping a high resolution raster perpendicular to the scan direction in order to form points of intersection where data are to be located.

In accordance with the present invention, a scanning system incorporates a single reference pattern to provide highly precise positional references for the scanning beam in two directions. The apparatus may include means for scanning a modulated light beam, such as a rotatable polygon deflector arranged for deflecting the light beam through an optical system onto a flat object format, along at least one closely parallel but spaced apart reference beam. Prior to the scanning system, both the modulated principal beam and the reference beam pass through a beam height deflector device which is controllable by applied signals to deflect both beams in equal or proportionate amounts in the Y direction. Subsequent to the imaging optics and adjacent the flat field format the reference beam alone is deflected through a reference pattern disposed adjacent to the writing path for the principal beam. The reference pattern modulates the scanning reference beam to provide a pulse width modulated signal whose periodicity synchronizes a reference clock for the X-axis direction and whose pulse width controls the height deflector.

The reference beam system includes a beam splitter system for bypassing the modulator with a given proportion (generally minor) of the principal beam, but reestablishing both beams in parallelism, preferably collinearly, before entering the height deflector. The optical reference pattern includes a reference scale provided with regularly spaced indicia or spokes having selected geometries in two directions. An evenly spaced set of edges on one side generates leading edges providing a data clock referenced to beam position for the X-direction. Spoke width, however, which varies linearly as a function of Y over a given length, generates pulse widths defining a direction error signal, which in turn is fed back to eliminate height error by controlling the height deflector. The reference indicia are spaced at some fraction of the pixel rate and used to synchronize a higher speed clock, so that an extremely fine reference pattern need not be used. The height deflecting assembly can be a vertically deflecting acousto-optic deflector.

Further, a reference beam is advantageously generated by rotating the plane of polarization when bypassing the modulator. Thereafter, after controlled height adjustment, a polarizing beam splitter in the optical path directs the reference beam into a training mirror path so as to impart the desired small divergence between the reference and principal beams. Both beams pass into the optical system for scanning and focusing prior to subsequent separation of the reference beam just prior to the flat format.

An advantage of the present invention is that pixel lines can be precisely located in both X and Y directions on a flat format for either a plotter or scanner. Furthermore, straight lines can be scanned because the system compensates for a number of errors in a scanning optical system.

Another advantage of the present invention is that vertical or height correction or location of pixels can be achieved as a function of pixel clock synchronization.

FIG. 1 is a diagrammatic, perspective representation of a plotting or scanning apparatus employing a reference scale and height deflecting system according to the present invention;

FIG. 2 is a diagrammatic, plan view showing a specific example of a reference scale according to the present invention; and

FIG. 3 is a flow chart showing a pixel location control apparatus according to the present invention.

Referring now more particularly to FIG. 1 of the drawings, an apparatus 10 according to the present invention for precisely locating a scanning beam in two dimensions comprises a system for writing successive lines across a film F disposed on a flat object field or format. A concentrated light beam 12, such as a coherent laser beam from a monochromatic (in this example) laser 14, is directed through an acousto-optic modulator 16 and onto a rotating polygon scanner 18 for scanning in the X-direction (transversely across the film F along a predetermined scan line). The scanning beam preferably but not necessarily is passed through imaging optics in a flat field optical system 19 of which the polygon scanner 18 is a part, as described in a previously filed patent application entitled "Scanning Imaging System And Method" by Michael Nagler, Ser. No. 459,230, filed Jan. 19, 1983, and assigned to the assignee of the present invention. This optical system provides diffraction limited focusing across a wide flat field (here 171/2") and a uniform beam spot size of approximately 17 .mu.m. A clock generator 20 responsive to X axis reference signals provides a clock signal, so that incoming data applied to a register 22 may be clocked out at times related to beam position so as to precisely position the pixels. Driver amplifiers for the modulator 16 have been omitted for simplicity and brevity, as have specific details of the optical system which are conventional and immaterial to the inventive concepts. The film F is advanced in the Y direction by a carriage (not shown) which can move continuously or incrementally and it will be understood that the entire system is enclosed within a light tight housing if film sensitive to visible light is employed.

A reference beam assembly includes a conventional beam splitter 24 disposed between the laser 14 and the modulator 16 for deflecting a minor proportion (e.g., 10%) of the principal beam in a perpendicular direction. This reference beam 25 is successively reflected off a pair of angled mirrors 26, 27 so as to bypass the modulator 16 and then be returned toward the principal beam path. In the bypass path, however, the reference beam is rotated 90.degree. in polarization by a rotator 29. Upon return to the principal beam path the reference beam is returned to collinearity by a beam recombiner 30.

An acousto-optic beam deflecting device 32 is disposed subsequent to the beam recombiner 30 to adjust the height (Y-direction) of the beam as pixels are being written on film F. Y-axis control signals derived as described below control the device 32. (In a scanning system in which the object document is being read for data, the principal beam is not modulated and reflected light must be detected, but the clocking and servoing schemes are fundamentally alike.) Both orthogonally polarized beams are equally deflected in the X-axis direction by the device 32.

A polarized beam splitter 33 responsive to the polarization of the reference beam, slightly deflects the reference beam 25 from the principal beam 12 following the beam deflector 32. A pair of training mirrors 34, 35 in the path of the reference beam redirect it into a path of very slight divergence relative to the principal beam, with the beams crossing at an intermediate pupil transfer or imaging device 36. Although the beams diverge at a small angle they may still be referred to as substantially parallel because they have a very small spacing after traveling. Both beams enter the flat field optical system 19 via a pair of beam folding mirrors 17, 17' and a scanning polygon 18 and subsequently are directed downwardly (as seen in FIG. 1) toward the scan line along film F.

The scanning principal beam 12 impinges on the flat film F format, while the correspondingly scanning reference beam 25 only is deflected horizontally by an angled mirror 37 onto an optical reference scale 38 that is disposed substantially parallel to and adjacent the scan line on the film F. Varying opacity gradations (discussed below with respect to FIG. 2) on the reference scale 38 modulate the intensity of the reference beam 25 during scanning, and this is sensed by a photodetector 39 disposed behind the reference scale 38.

Referring now to FIG. 2, the reference scale 38 modulates the reference beam as a function of both the scan line position along the X axis and the lateral deviation (Y axis error) in the main beam 12. More specifically, reference scale 38 includes regularly spaced and identical indicia forming spokes 40 of opaque or low transmissivity in contrast to a transparent or translucent background, and having two sets of geometrical properties. The first is a set of evenly spaced leading edges 41, which when scanned by the beam generate the leading edges of a series of pulses. These leading edges define a data clock exactly referenced to a position of the principal beam 12 in an X-direction of the format. The second geometrical property is a variation in spoke width, which varies as a function of beam position along the Y axis. Spoke width, which is shown as nonsymmetrical about a line 42 extending perpendicularly to and generally bisecting spokes 40, is defined as the distance between a leading edge 41, which can be considered a leading spoke edge, and an associated trailing edge 44. As can be seen, the spacing between a pair of edges 41, 44 is greater above line 42 than below same, with a linear transition being made in a zone 45. A reference scale 38 of about 171/2" in length and having about 4400 pairs of edges 41, 44 has been found satisfactory.

The different widths between the edges 41 and 44 formed on the reference scale 38 cause signals to be generated by photodetector 39, as the length of the scale 38 is being scanned by the impinging reference beam 25. This signal is fed back in conventional servo circuits 50 (not shown in detail) to control the beam height deflecting device 32 to adjust the main beam 12 and reference beam 25 positions in the Y direction to seek the center line 42 on the reference scale 38. If the pulse width generated for a beam height position between edges 41 and 44 is less than a predetermined time interval, the deflector 32 is driven to raise beam 12. If, on the other hand, the beam position crosses edges 41 and 44 to provide a greater than nominal pulse width, the deflector 32 lowers the beams and thus the associated scan line.

An example of a beam height deflecting device 32 is an acousto-optic deflector driven by an adjustable radio frequency (RF) source. The driving source itself may be controlled by suitable known servo circuitry 50 as a function of signals received from the photodetector 39. The pulse width may, for example, be converted to a variable RF signal by signal integration or time duration counting techniques (together with D/A conversion) used to control a variable frequency oscillator.

Referring to FIG. 3 of the drawings, a flow diagram is shown setting forth the various elements making up a preferred apparatus for carrying out the present invention.

Starting with the light source, which can be a conventional continuous wave laser 110, preferably polarized, a beam is fed through coupling or training mirrors 112 to a conventional beam splitter 114. One portion of the split beam, referred to above as the main beam, is sent directly to an acousto-optical (A/O) modulator section 116 where the beam is modulated as it propagates onward to a beam recombiner 118. The auxiliary reference beam created by splitter 114 is fed to a training mirror 120, rotated 90.degree. in a polarization direction by a rotator 122, and then trained by a second mirror 124 into the beam recombiner 118 to be superimposed with the main beam emitted from modulator 116. There now are in the system two collinear beams superimposed, one modulated and one not, and having orthogonal polarizations.

The recombined beams are now sent into an A/O height servo 126, preferably a glass block acoustically modulated by a lithium niobate piezoelectric crystal, which deflects the superimposed beams up or down in response to beam positional error. They then pass to a polarizing beam splitter 128, which sends one beam directly to a pupil imaging device 132 and the other beam through the same device 132 via a first training mirror 130 which is positioned to offset the reference beam at a small angle relative to the principal beam. A second training mirror 134, after the pupil imaging device 132 is used in establishing the desired small angle of divergence between the beams as they enter an imaging and deflecting system 136, directs the main beam so as to scan a suitable flat format F. The modulated reference beam impinges on a beam folding mirror 137 and is reflected to a reference scale 138 constructed in the manner of reference scale 38 in FIG. 2. The output of a photodetector 139 behind the reference scale 138 is fed to a data clock synchronizer 140 which provides synchronizing signals to data clock circuits 142 operating at a basic multiple of the reference indicia rate so as to multiply the number of image pixels accordingly. The data clock is input to a data-in storage register 144 that controls operation of the principal beam modulator section 116 through a conventional driver 146.

The output from the photodetector, which will be in the form of pulses detected by photomultiplier circuits feeding an amplifier, is also fed to height error signal circuits 148 which times the distance between the on and off going spoke edges on the reference scale, as described above, and controls position of the acousto-optic height servo 126 through a suitable RF driver 150.

As can be readily understood from the above description and from the drawings, the present invention provides in a simple and economical manner a pair of feedback loops, one for the X-direction of a flat format and the other for the Y-direction thereof, which assures precise locating of pixels on any desired format. Further, the arrangement can be employed with either printing or scanning as desired. The use of a single reference beam to correct location errors in both the X and Y directions simplifies construction of the apparatus, in that scan line position and parallelism are assured with lower cost optics than would otherwise be needed. In addition the system assures close control of position in both directions, although it is to be understood that the Y-direction loop can be used alone, or in conjunction with any other kind of pixel clock.

While different alternatives and modifications have been suggested, it will be appreciated that the invention is not limited thereto but encompasses all forms and variations falling within the scope of the appended claims.