Adaptive Corrector of Facet Errors in Mirror Scanning Systems

1. Field of the Invention

This invention relates generally to photographic recorders of the ray type and more particularly to mirror scanning systems compensating for various orientations of mirror facets by corresponding alterations in the direction of light rays emanating from an acousto-optic modulator.

2. Description of the Prior Art

Optical scanning systems are utilized in a variety of situations including non-impact printers and facsimile devices. Typically, such systems employ a multi-faceted mirror which is rotated about a stationary axis to provide for a scanning of a light beam within a plane, usually the horizontal plane. For the imprinting of alpha-numeric characters or other symbols on a recording medium, the medium may be formed of photo-sensitive material and positioned along the periphery of a drum. The axis of the drum lies in the scanning plane. The drum is rotated by a predetermined increment between successive scans so that each scan produces one line of printed material on the recording medium.

For modulating the intensity of the light beam so as to produce light and dark regions on the recording medium, many systems are now employing acousto-optic modulators wherein an interaction between an acoustic wave and a light wave results in a splitting off of a portion of the optical energy from the main beam, or zero order beam, into a diffracted, or first order beam. The scanning mirror and the drum are positioned to intercept the diffracted beam. Thus, by activating and deactivating the acoustic wave, the diffracted beam is amplitude modulated.

In the use of the foregoing modulators, the angle of diffraction, or Bragg angle, is dependent on the frequency of the acoustic wave and on the speed of propagation in the medium of the modulator as these two parameters determine the wavelength of the acoustic wave. Such frequencies are on the order of radio frequencies, RF, and may approach 100 MHz (megahertz). Accordingly, in the construction of such scanning systems, the locating of the optical components is dependent on the selection of the acoustic frequency and the Bragg angle so that the diffracted beam will be intercepted by the scanning mirror and reflected to the site of the scan line on the recording medium.

A limitation arises in the construction of such systems in that great precision is required to insure that a succession of scan lines will appear to be uniformly presented on the recording medium, and, in particular, that the scan lines will be evenly spaced. However, due to the geometry of the optical paths and the optical elements, a slight misalignment in the vertical orientation of a mirror facet relative to a second facet results in a noticeable variance in the spacing between lines of printed matter on the recording medium. Accordingly, it has been necessary to maintain high tolerance on the manufacture of the scanning mirror. It would be most advantageous if the scanning system would be less dependent on the precision of the manufacture of the scanning mirror for increased accuracy in the spacing of scan lines even in the presence of discrepancies in the orientations of the mirror facets.

This limitation is compounded by considerations which become apparent during an attempt to build the scanning system with features which might compensate for discrepancies in the facet orientations. For example, the use of fixed compensation, established individually for each mirror facet as by perturbations in the beam direction, requires a time consuming alignment process. Also, the use of an optical sensor for an automatic compensation for the facet errors introduces the considerations of electronic component drift which would introduce much greater errors than those of the facet orientations.

The foregoing illustrates limitations of the known prior art. Thus, it is apparent that it would be advantageous to provide an alternative directed to overcoming one or more of the limitations as set forth above. Accordingly, a suitable alternative is provided including features more fully disclosed hereinafter.

In one aspect of the present invention, this is accomplished by providing an apparatus for correcting facet errors in mirror scanning systems including a rotatable assembly of optical elements for scanning a beam in a plane. The beam is directed to the rotatable assembly by an acousto-optic modulator which modulates the beam with data and deflects the beam as a function of the frequency of an acoustic signal which drives the modulator. A sensor senses the position of the beam as directed by the rotatable assembly. A controller is responsive to signals of the sensor for controlling the frequency. The controller includes a memory for storing frequency values for scans by successive ones of the optical elements, and means for altering the frequency.

The foregoing and other aspects will become apparent from the following detailed description of the invention when considered in conjunction with the accompanying drawings. It is to be expressly understood, however, that the drawings are not intended as a definition of the invention but are for the purpose of illustration only.

In the drawings:

FIG. 1 is a diagrammatic view of a mirror scanning system incorporating an embodiment of the invention for correction of facet errors in the mirror;

FIG. 2 shows a portion of the system of FIG. 1 including an elevation view of the mirror with a schematic presentation of light rays for different orientations of a mirror facet in the vertical plane;

FIG. 3 is a block diagram of an acoustic generator in the system of FIG. 1;

FIG. 4 is a block diagram of a frequency controller of the system of FIG. 1, a variation in the frequency altering the vertical orientation of a light ray in accordance with the invention;

FIG. 5 is a block diagram of an averaging unit in FIG. 4;

FIG. 6 is a diagram, partially schematic, of electrical circuitry of a sensor interface in the controller of FIG. 4;

FIG. 7 is a schematic diagram of an alternative embodiment of the sensor interface; and

FIGS. 8 and 9 illustrate, respectively, an algorithm and comparator simulation of the algorithm for a microprocessor to be used in an alternative embodiment of the controller of FIG. 4.

The present invention is generally related to the optical sweeping system disclosed in the commonly-assigned U.S. Pat. No. 4,307,409 entitled, "Multi-Aperture Feedback System for a Laser Scanner", invented by Nelson L. Greenig & Richard M. Shelton, and issued on Dec. 22, 1981, which is expressly incorporated herein by reference.

Referring now to FIG. 1 there is shown a scanning system 20 incorporating a preferred embodiment of the present invention. The system 20 includes a drum 22 supporting a recording medium 24 of photosensitive material which is illuminated by the light of a laser 26 for the printing of data, such as alphanumeric characters or other symbols, upon the medium 24. The system 20 includes an acousto-optic modulator 28, focusing optics shown as a lens 30, which is typically a multi-element lens, and a multi-faceted mirror 32 which is rotated by a motor 34 for scanning a beam of the laser light upon the medium 24. Each mirror facet 36 provides a separate scan line upon the medium 24.

During the generation of each scan line of printed matter upon the medium 24, the drum 22 is advanced by an incremental amount of rotation designed by a motor 38, shown mechanically coupled to the drum 22 by dashed line 40. Synchronization of the rotations of the motors 34 to 38 is provided by a system timing unit 42. Timing for the mirror motor 34 is provided by signals conducted along line 44 from the timing unit 42. Symchronization signals from a shaft angle encoder 38A of the drum motor 38 are coupled to the timing unit 42 by a terminal T for synchronizing the timing unit 42 with rotations of the drum 22. Accordingly, the drum motor 38 may be energized directly from an electric power line.

The system 20 further comprises an RF (radio frequency) generator 46, a frequency controller 48, a data source 50 and a linear sensor 52 of the laser light. The output RF signal of the generator 46 is coupled to a transducer within the modulator 28 to provide an acoustic wave at a radio frequency for driving the acousto-optic modulator 28. The amplitude of the RF signal and, hence, of the acoustic wave is modulated by the data source 50, a corresponding modulation appearing on the light ray incident upon the recording medium 24. The frequency of the RF signal or electric wave and, hence, the acoustic wave, is established in response to a frequency command signal from the controller 48, the command signal being digitally formatted and coupled by a data bus 54.

The sensor 52 is positioned alongside the drum 22 where the scan begins and is oriented to the scanning plane of the mirror 32 so as to intercept the beginning of a scan line. The sensor 52 provides electrical signals, in a manner to be described hereinafter, which signal the location of a scan line to indicate whether the scan line is too high or too low. The electrical signals are coupled by a line 56 to the frequency controller 48 whereby the controller 48 can order a change in the frequency of the scan lines.

Referring also to FIG. 2, the effect of the acoustic-drive signals within the modulator 28 is more readily seen. In the absence of the acoustic-drive signal, the light of the laser 26 is undeflected as it passes through the modulator 28 and the lens 30 to provide the zero-order or undeflected beam 58 shown incident upon an exemplary stop 60. In the presence of the acoustic wave, a portion of the radiant energy of the beam 58 is deflected as a first-order or deflected beam 62. An exemplary three such deflected beams are shown in FIG. 2 and are further identified by the legend A, B, and C to illustrate specific ones of these beams 62. The angle of the deflection is dependent upon the frequency of the acoustic wave so that, by varying the frequency in a known manner, either one of the deflected beams 62A, 62B or 62C can be obtained.

In the instant system 20, only the deflected beams 62 are to be utilized in producing a scan line upon the recording medium 24. Since the undeflected beam 58 is not utilized, the stop 60 absorbs the energy thereof. Alternatively, as is done in other scanning systems (not shown), the energy of the beam 58 may be utilized to measure the instant scan angle, as by allowing the beam 58 to impinge upon a portion of the scanning mirror 32.

As best seen in FIG. 2, the orientation of the mirror 32 is precisly set with reference to the motor shaft 64, suitable bearing and supporting structure (not shown) being attached thereto to insure the precise orientation of the mirror 32 relative to the modulator 28 and the drum 22 for the precise positioning of the scan line. Nevertheless, slight perturbations in the location of the scan line may occur due to small variations in the orientations of respective ones of mirror facets 36 with respect to the central axis 66 of the scanning mirror 32. The dashed lines 36A and 36C represent two possible deviant orientations of a mirror facet 36B. Three deflected beams 62A-C are provided with the requisite angles of deflection for impinging, respectively, on the mirror surfaces represented by the three lines 36A-C for further reflection towards the drum 22 to impinge upon a common scan line at 68. Thus, for each perturbation in the alignment of the set of mirror facets 36, there is some angle of deflection of first order beam 62 which can compensate for the misalignment in each facet 36 to insure that the beams 62 are reflected by the mirror 32 to the proper location at the drum 68.

Since, as has been noted hereinabove, such angle of deflection of the beam 62 is dependent on the frequency of the acoustic wave within the modulator 28 by the generator 46, misalignments of the mirror facets 36 are compensated by the selection of the appropriate frequency command signal on bus 54 for establishing the requisite frequency and wavelength to the acoustic wave. The foregoing procedure in the compensation is applicable for laser light both within the visible spectrum as well as outside of the visible spectrum. The beam of radiant energy produced by the laser 26 is defracted in accordance with a phenomenon of Bragg defraction in the medium of the modulator 28 in the presence of the acoustic wave.

Referring now to FIG. 3, the R.F. generator 46, previously seen in FIG. 1, comprises a phase-locked loop 70, a crystal-controlled source 72 of a reference frequency for use by the phase-locked loop 70, and an amplifier 74 having a gate signal terminal connecting with line 76. A transducer 78 for converting electrical energy into sonic energy is located in modulator 28. The phase-locked loop 70 includes a digital frequency synthesizer responsive to a digitally-formatted frequency command signal on bus 54 for providing a sinusoidal electrical signal having a frequency equal to that of the desired acoustic signal. Digitally-controlled frequency synthesizers are well known and available commercially and, accordingly, such circuitry is readily implemented within the phase-locked loop 70.

The sinusoidal signal produced by the phase-locked loop 70 is amplified by the amplifier 74 to a suitable power level for driving the transducer 78. In response to the signal on line 76 from the data source 50 (FIG. 1), the amplifier 74 may be placed in a state of conduction or non-conduction to the passage of the signal from the loop 70 to the transducer 78. Thereby, in response to the signal on line 76, the acoustic wave and the resultant deflected beam 62 (FIG. 2) may be established or discontinued to provide, respectively, a light or dark area on the recording medium 24 upon the drum 2. The coupling of the R.F. Drive signal from the amplifier 74 to the modulator 28 is provided by line 80. The successive bursts of the drive signal on line 80 may be either of the same frequency or of different frequencies depending on the value of the frequency command signal on bus 54.

Referring now to FIG. 4, the frequency controller 48 comprises a holding register 82, an electronics unit 83 comprising an averaging unit 84 and a sensor-interface unit 86, an address counter 88, a memory 90, and a counter 92 for correction of the frequency command words stored in the memory 90. The memory 90 stores the value of the frequency of the R.F. drive signal for the modulator 28 (FIG. 1), the values of the frequency being represented by digitally-formatted command words or numbers. The memory 90 is addressed by the counter 88, the counter 88 counting clock signals applied by the timing unit 42 (FIG. 1) via line 94 to the clock-input terminal, C, of the counter 88. A clock signal on line 94 appears at the completion of each scan line so as to ready the controller 48 for the next scan line. In this embodiment, there are 15 facets on the polygon mirror 32. There are also 15 counts within counter 88, the counter 88 counting modulo-N where N=15 is the number of facets. Therefore, after each revolution of the mirror, the counter 88 returns to its original state. Prior to the generation of each scan line, the requisite frequency command word is loaded from the memory 90 into the register 82 and held within the register 82 during the scan so as to provide for the required frequency of the acoustic-drive signal throughout the generation of the scan line.

In accordance with a feature of the invention, the digitally-formatted frequency command is represented as a number, and is parallel loaded into the counter 92 at P to preset the counter 92, acting through the interface unit 86, transmits a signal which either increments or decrements the count of the counter 92 to increment or decrement the count is based on the sensed position of a scan line relative to a reference value of position. Thus, the counter 92 continues to increment this count during subsequent scans by an individual one of the mirror facets 36 as long as the scan line is sensed as being to one side of the reference location, the counter 92 thereafter decrementing when the sensed position of the scan line is on the opposite side of the reference location.

The signal commanding the incrementing (counting up) or decrementing (counting down) is applied by the interface unit 86 to the up/down, U/D, terminal of the counter 92. A signal indicating that a terminus of a scanning beam has been detected by the sensor 86 is applied to the clock terminal, C, of the counter 92 to alter the count by the value of one. The updated, or corrected, value of the frequency command, as stored in the counter 92, is then loaded into memory 90 in place of the previous value of the command word. Thereby, the value of the frequency command time is steadily altered from revolution to revolution of the scan mirror 32 to optimize the value of the frequency and wavelength of the acoustic wave in the modulator 28. Once the reference position of the scan line has been attained, the counter 92 oscillates about the desired value of the command word by an amount no greater than unity, the magnitude of the oscillation being sufficiently small relative to the magnitude of the number representing the command word so as to introduce no more than a negligible error in the angle of deflection of the scanned beam. The unit 84 provides a correction signal to the interface unit 86 for compensation of component defects in accordance with a further feature of the invention as will now be described.

Referring now to FIG. 5, the averaging unit 84 comprises a counter 100 and an integrator 102, the integrator 102 including a resistor 104, a capacitor 106 and an operational amplifier 108. The capacitor 106 connects the output terminal of the amplifier 108 with a summing point at the junction of the resistor 104 and the negative (or inverting) input terminal of the amplifier 108 to provide the well-known integrator circuit. With respect to the output terminal of the counter 100, only the sign bit thereof is utilized, the sign bit being coupled via the resistor 104 to the amplifier 108. Thus, the signal at the output terminal of the counter 100, on line 110, has the appearance of a pulse-width modulated signal, as portrayed in the exemplary waveform 112, and having a duty factor dependent on the magnitude of component drift, as will be described below. The signal on line 110 is then integrated by the integrator 102 to provide the correction signal on line 114 for the interface unit 86. The correction signal on line 114 has a substantially constant (or slowly varying) amplitude depending on the duty factor of the waveform 112.

In operation, the counter 100 counts clock pulses, the counting being up or down dependent on an up/down command signal provided by the interface unit 86. The counter 100 is preloaded with zero at the same time that the facet parameter locations in the memory 90 are preloaded with the parameters for the desired average frequency. A well-known bias circuit 115 is coupled to the input terminal of the amplifier 108 to offset the voltage of the sign bit so that, for a waveform 112 having equal durations of the logic 1 and logic 0 states, the integrator 102 provides a value of zero at its output termination line 114. The time constant of the resistor-capacitor circuit of the integrator 102 is much longer than the period of rotation of the scan mirror 32 to insure accurate integration of the successive counts of the counter 100.

Thereby, the averaging unit 84 produces the signal on line 114 with an amplitude proportional to the average value of the command words. In the event that the average value of the command words becomes excessively large or excessively small, this indicating a drifting in the parameter value of the components in the interface unit 86 and the sensor 52, the duty cycle of the square-wave signal on line 114 serves as a correction signal for altering the values of certain components in the interface unit 86, as will now be described, to compensate for the foregoing drift.

Referring now to FIG. 6, the diagram shows the electric circuitry of the sensor 52, the interface unit 86 and their interconnection. The sensor 52 comprises a first electrode 116 and a second electrode in the form of a resistive element 118, both of which are imbedded in photosensitive material. In the presence of electrical excitation and incident light, electric current flows between the terminal 116 and the element 118. The element 118 terminates in lines 121 and 122, the current of the sensor 52 dividing between the two lines 121-122 in accordance with the location of impact of the light beam 62 (FIG. 1) upon the sensor 52.

Voltage excitation, -V, to the plate 116 is applied via line 123 from a power supply 126. The lines 121-123 are included within the lines 56 (FIG. 1) connecting the sensor 52 with the frequency controller 48. However, the interface unit 86 includes two field-effect transistors 128 and 130 and comparators 132 and 134, resistors 136 and 138 and a source 140 of a reference voltage for the comparator 134.

The source 130 may be readily fabricated by a resistive divider network (not shown) connected across the terminals of the supply 126. Each of the transistors 128 and 130 are utilized as resistive elements in arms of a bridge circuit 142, their remaining two arms being formed by the two portions of the element 118 through which the sensor current divides. The lines 121 and 122 connect with the input terminals of the comparator 132 which senses a null in the bridge circuit 142. A comparator 132 is responsive to the polarity between the lines 121 and 122 for providing an output logic signal having a high or low logic level corresponding to a logic 1 or 0.

On line 144, depending on the location of the point of impingement of the light beam 62 upon the sensor 52, the signal on line 144 is applied, as shown in FIG. 4, to the up/down terminal of the counter 92. The magnitudes of the voltages on the lines 121 and 122, with reference to ground, are summed together by the summing resistors 136 and 138, the sum being applied to the input terminal of the comparator 134. In the absence of the optic signal of the light beam 62, the magnitude of the sum signal presented by the resistors 136 and 138 is less than the magnitude of the reference voltage of the source 140. In the presence of the optical signal, the magnitude of the sum signal is greater than the magnitude of the reference voltage. A logic level of the output signal of the comparator 134 on line 146 is, thus, responsive to the presence or absence of the light beam 62, the signal on line 146 being coupled, as shown in FIG. 4, to the clock terminal of the counter 92.

In accordance with a further feature of the invention, an interface unit 86 includes further circuitry for the compensation of drift in the parameter values of the components of the bridge circuit 142. The circuitry of the interface unit 86 further comprises four resistors 151, 152 and 153 and an amplifier 156. The amplifier 156 is an inverting amplifier and the resistor 153 is coupled between the input and output terminals of the amplifier 156 to establish a predetermined gain thereto. The resistors 151-152 serve as summing resistors for combining the correction signal on line 114 with a fixed bias at the input terminal of the amplifier 156. The fixed bias is obtained by virtue of the connection of the resistor 152 to the power supply 126, the two resistors 151-152 serving as a voltage divider of the bias signal.

In operation, the resistance of the transistors 128 and 130 in combination with the resistances of the respective portions of the element 118 serve as the arms of the bridge circuit 142. Each of the transistors 128 and 130 is of the depletion N-channel, field-effect type transistor wherein the resistance between the respective source and drain terminals is dependent upon the magnitude of the voltage at the gate terminals. Thereby, any variation in the resistance of the arm of the bridge circuit 142, as may occur through aging, or heating, or by a variation in the position of the sensor 52 relative to the mean value of the scanning beam, can be compensated by adjusting the voltages at the respective gate terminals, labeled #1 and #2. The correction signal on line 114 is a substantially constant or slowly varying analog voltage and is applied directly to gate #1, only a fraction of the integrated signal being coupled to the gate #2 or via the network of resistors 151-153 and the amplifier 156. The value of the resistor 152 is initially set in conjunction with a nominal value of the signal on line 114 to provide for values of resistance which substantially balance the bridge circuit 142.

Further adjustment of the balance of the circuit 142 is accomplished by virtue of the unequal coupling of the correction signal to the gate terminals of the transistors 128 and 130. As is depicted in the graph 164, appended adjacent the amplifier 156, an increase in the amplitude of the signal on line 114 results in an increase in the magnitude of the voltage applied to gate #1 and a decrease in the magnitude of the voltage applied to gate #2. The decreasing voltage at gate #2 results, in part, from the inverting function of the amplifier 156. Since the resistance values of the transistors 128 and 130 are dependent upon the values of the voltage at the respective gate terminals, it may be seen with reference to the graph 164 that, by an appropriate selection of the amplitude of the signal on line 114, a desired relationship between the resistance values of the transistors 128 and 130 may be established for counteracting any shift in the bridge parameters such as a shift in the position of the sensor 52 relative to the drum 22 in FIG. 1. In operation, therefore, and with the reference also to FIG. 4, a requisite frequency for the R.F. generator 46 is established by an iterated process wherein the frequency command word is continually updated by the counter 92 in response to signals of the interface unit 86 and, furthermore, wherein the circuitry of the interface unit 86 is adapted automatically to compensate for drifts due to component aging and due to shifting position of the sensor 52. The averaging unit 84 continually updates the average value of the frequency command words to sense whether any drifting has occurred. The resultant correction signal on line 114 is utilized to re-establish balance of the bridge circuit 142 so as to insure proper operation of the comparator 132 in transmitting its output signal along line 144 to the counter 92.

Any misalignment in the location of the sensor 52 would result in excessive unbalance in the bridge circuit 142 with a resultant signaling by the comparator 132 to drive the correction counter 92 to excessively high values of frequency command words on bus 54. Such words might be driven outside the range of storage capacity of the memory 90. However, the signal on line 114 maintains the bridge circuit 142 substantially balanced so as to insure that the command words on bus 54 remain within the range of the memory 90 and within the range of the counting capacity of the counter 92.

It should be noted that such a shifting in the values of all of the command words or all of the mirror facets 36 of FIG. 1 does not introduce an error in the spacing of the scan lines but, rather, introduces a fixed offset of all of the printed lines on the medium 24 from the bottom and top edges of the medium 24. For example, in the event that the sensor 52 were repositioned by a small amount, such as a tenth of a millimeter above or below the anticipated sight of the scan line, then the frequency controller 48 would command a new value of frequency which would alter the angle of deflection of the beam 62 so as to reposition the scan line in accordance with the position of the sensor 52. Thereby, the system 20 in FIG. 1 accomplishes the adaptive correction for errors in the orientation of the facets of the mirror 32.

While the foregoing system has been provided with features required for the adjustment of the location of a scan line to compensate for a misalignment in the facets of the rotating scanning mirror, it is to be understood that the system is generally applicable for the deflection of a beam of radiant energy in any situation wherein the position of the beam is to be sensed and adjusted.

Referring now to FIG. 7, there is shown a schematic diagram of a sensor interface unit 170 which is an alternative embodiment of the interface unit 86, shown in FIG. 6. The interface unit 170 precedes an input signal on line 114, as does the interface unit 86, and provides two output signals on line 144 and line 146 as does the interface unit 86. In addition, the three lines 121 and 122 and 123 are coupled to the sensor unit 52 appear with both the interface units 86 and 170. Thus, the interface units 170 may be substituted for the interface unit 86 for connection with the other components of the frequency controller 48, as shown in FIG. 4.

The interface unit 170 is of dual-channel construction, as is the bridge circuit of the interface unit 86 of FIG. 6. One channel of the interface unit 170 comprises amplifiers 172 and 174. Each of the amplifiers 172 and 174 are operational amplifiers and have feedback resistors, respectively, 176 and 178. Similarly, the other channel comprises operational amplifiers 180 and 182, there being a feedback resistor 184 connected between the input and output terminals of the amplifier 180 while the field effect transistor 186 serves as the feedback element of the amplifier 182. The output terminals of the amplifiers 174 and 182 are coupled to the input terminal of a comparator 188. The dual amplifiers 172 and 174 are serially connected between the line 122 and the comparator 188, the junction of the two amplifiers 172 and 174 being coupled by a line 190 and a summing resistor 192 to an input terminal of an operational amplifier 194. The two amplifiers 180 and 182 are serially connected between the line 121 and the comparator 188, the junction of the two amplifiers 180 and 182 being coupled by a line 196 and a summing resistor 198 to the input terminal of the amplifier 194. The operational amplifier 194 is provided with a feedback resistor 200 and is also connected to one input terminal of a comparator 202. The comparator 202 compares the magnitude of the output signal of the amplifier 194 with a reference signal provided by the reference source 204. The drift correction signal on line 114 is coupled by a resistor 206 to the gate terminal of the transistor 186. The electric power for energizing the sensor 52 is provided from a source of voltage (-V) coupled to the sensor 52 via a resistor 208 and the line 123.

In operation, the serial connection of the amplifiers 172 and 174 provides a voltage at one input terminal of the comparator 188 while the serial connection of the amplifier 180 and 182 provides a second voltage at the second input terminal of the comparator 188. The two voltages at the input terminal of the comparator 188 may be of equal amplitude in the event that the laser beam of FIGS. 1 and 2 strikes the central portion of the sensor 52 during a scanning by the beams. Alternatively, the voltages at the input terminals of the comparator 188 will differ in the event that the laser beam impinges upon the sensor 52 at a point which is of unequal distance between the sensor terminals coupled to the lines 121 and 122. Thereby, the comparator 188 provides a signal on line 144 which may be either of two logic states, mainly, a logic 1 or a logic 0 depending on whether the site of impingement of the laser beam dependent on the relative magnitude of the sum signal and the reference signal.

In the event that there is drifting in the electrical parameters of the sensor 52 or in the circuitry coupled thereto, or in the event that there is a mechanical shift in the position of the sensor 52 relative to the rest of the optical system of FIG. 1, then the signal on line 144 will be subject to an error due to the foregoing drift. The drift connection signal on line 144 corrects for the foregoing drift error as follows. The resistance of the transistor 186 is varied by a variation in the amplitude of the signal 114, which signal is applied to the gate terminal of the transistor 186. A change in the resistance results in the gain of the amplifier 182 and the consequent shift in the balance point between the two channels of the interface units 170. By shifting the balance point, the parameters of the circuit are thereby altered to compensate for the error associated in the foregoing drift, the error thereby being corrected. Thus, it is seen that the interface unit 170 provides the same function as does the interface unit 86 of FIG. 6.

Returning to FIG. 4, it is noted that, by way of alternative embodiment, the electronic unit 83 comprising the averaging unit 84 and the sensor interface unit 86 or 170 may be conveniently fabricated by the use of a microprocessor circuit configuration in the form of a micro-circuit on a chip. Such a microprocessor would then be suitably programmed to perform the functions of the averaging unit 84 and the interface unit 86 or 170. Such a microprocessor would provide the requisite signals on the lines 144 and 146 for operations of the correction counter 92 to update the parameters stored in the memory 90 as was disclosed previously with reference to FIG. 4. In addition, the microprocessor would be responsive to the difference between the sensor signals on the lines 121 and 122 for responding to the position of impingement of the laser beam upon the sensor 52.

In the use of the miroprocessor, all of the facets of the scanning unit 32 are provided with a desired count representing the requisite center frequency for the modulator 28 of FIG. 1, the desired count being stored in the memory 90 as has been disclosed previously with reference to the system 20. The operation of the microprocessor is accomplished by the use of an algorithm and a program for carrying forth the algorithm. A simplified bock diagram of the facet correction algorithm is shown in FIG. 8 with a computer simulation thereof being shown in FIG. 9.

The algorithm is explained as follows. The comparator compares two analog Levels A and B giving a true digital output when B.gtoreq.A. The A input is an analog representation of the facet perpendicularity error. The B input is an analog representation of the beam position wherein the higher the beam, the higher the voltage B. There are two loops that tend to change the analog levels at A and B. If B.gtoreq.A is true, the beam is too high and the beam tends to be lowered. This lowering, however, is delayed one revolution before being effective on the beam. If there is a voltage at A which represents desired beam height, then the B loop would tend to bring the beam from all facets to the same point.

The voltage at A could be supplied by a potentiometer, and experimentally works over a short time periods. However, due to actual conditions, there is a tendency for the "fixed" height of the beam to drift up and down. This drift is not sufficient so as to deteriorate printing, however, it does present a risk that the beam level can drift so far that the extreme facets reach the correction limit of the design. If this happened, the correction loop would be saturated resulting in only partial correction. This is where the bias comes into play which, in effect, multiplies the facet parameters (high/low crossover points) by a "constant" so that the A input is always held at a level which does not allow the B loop to saturate.

The A (bias) loop consists of a slow integrator, one whose output increases at a constant rate with high input and decreases at the same rate when the input is low. The integrator input is derived from the sign bit of a common register which is incremented or decremented by 1 at each facet time at the same time one of its individual facet registers is updated. Whenever this common register sign bit changes, the integrator reverses. In actual practice, the total integrator output variation is generally less than 100 MV during operation.

Referring again to FIG. 8, it can be seen that whenever (B.gtoreq.A) is true, the A loop tends to increase A and the B loop tends to decrease B. This tends to invert (B.gtoreq.A) creating a condition that continuously recycles and is always self-correcting. At power turn on, however, the A loop is in saturation with (B.gtoreq.A) true and remains so for any condition of B. To purge this saturation condition, a start-up mode is built into the microprocessor which does six things:

(1) Starts its bias integrator with a negative slope (since general clear drove it into position saturation).

(2) Fixes the VCO at a midrange frequency.

(3) Counts (B.gtoreq.A) compares per revolution.

(4) Loads all facet registers with the desired central frequency parameter.

(5) Clears the common register to 0.

(6) Exits to the operating loop when the number of compares per revolution reach 7 or 8.

This control loop has been simulated on a programmable calculator, ignoring loop saturation, with the following results. Good resolution was achieved provided the integrator rate was slow enough. The steady state average frequency was equal to the frequency at start as demonstrated with two sets of random numbers used as facet parameters. Set number II was generated by adding a constant to set number I which was, in turn, generated by a normal distribution random number generator contained in a canned calculator program. Sets I and II both reached the same steady state regardless of starting bias voltage.

The foregoing algorithm has been implemented with an Intel 8748 microprocessor chip which contains all the facet registers, the common register, and the history record of the correction made during the previous eight revolutions. A description of this chip and its operating instructions are found in the booklet entitled, "Intel MCS-48 User's Manual", April 1979, published by Intel Corporation, Santa Clara, California, which is expressly incorporated herein it its entirety by reference. An "Instruction Set" is found on pp. 4-1 through 4-5 of the reference document. The computer program listings are presented below.

The foregoing has described an apparatus for correcting facet errors in mirror scanning systems.

It is anticipated that aspects of the present invention, other than those specifically defined in the appended claims, can be obtained from the foregoing description and the drawings. ##SPC1##