JP2005169784A - Image recorder - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress unevenness of image while enhancing reproducibility of fine lines by suppressing unevenness of image occurring when horizontal registration is shifted at a print position due to distortion or the like of an encoder in case of wide width print or twin head at an LFP thereby reducing shift of horizontal registration over the entire region of print width. <P>SOLUTION: Entire region of print width can be regulated with a varied horizontal registration by making the regulation value of horizontal registration variable during print, especially during movement of the carriage in the main scanning direction. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to an image recording apparatus for recording an image, and particularly to control (so-called registration) for superposing a plurality of ink colors.

  2. Description of the Related Art Conventionally, among image recording apparatuses, particularly ink jet recording apparatuses, a recording head having a plurality of recording elements (nozzles) arranged in a direction parallel to a conveyance direction (sub-scanning direction) of a recording medium (paper, etc.) is mounted on a carriage. The carriage moves (scans) in a direction (main scanning direction) orthogonal to the sub-scanning direction, ejects ink droplets with a recording head to form an image on the recording medium, and then forms the image on the recording medium after forming the image. There is a so-called serial type recording apparatus in which the image of the next scan is formed in the same manner.

  In such a serial type recording apparatus, as shown in FIG. 2, each recording head has a structure in which a plurality of nozzle rows are arranged in the main scanning direction, and each nozzle row has each ink color (for example, (Black: Bk, cyan: C, magenta: M, yellow: Y, etc.) ink is supplied, each ink color is ejected from each nozzle row, and these ink droplets are superimposed on the recording medium. A color image is formed. In a serial type recording apparatus using such a recording head, the positions of the ink color nozzle arrays in the recording head are different in the main scanning direction as shown in FIG. If the landing position is not adjusted, a registration error occurs, resulting in an unclear image with blurred color shifts and line drawings.

  Therefore, it is necessary to adjust the landing position (registration: hereinafter abbreviated as “registration”) of each conventional dot, which is performed in the recording apparatus. The parameters necessary for registration adjustment are hereinafter abbreviated as registration adjustment values.

  In the initial recording device, printing is performed with a time delay from the printing of the first nozzle row that reaches the recording medium first in the main scanning direction by the time obtained by dividing the interval between the nozzle rows of each ink color by the carriage moving speed. By doing so, the cash register is adjusted. This time delay method is based on the premise that the carriage moving speed is constant. In general, however, the carriage moving speed has a certain amount of fluctuation and is not necessarily constant, so that a registration shift may occur.

  Therefore, in a current recording apparatus, a scale for a linear encoder is provided in the main scanning direction, an encoder sensor for reading a slit pattern (transmission, non-transmission) of the scale is provided on the carriage, and each position is detected by detecting the position of the carriage. Registration adjustment is performed by delaying and printing the position (distance) corresponding to the interval between the ink color nozzle rows. Scales include optical patterns (transmission / non-transmission, reflection / non-reflection, etc.) and magnetic patterns (N pole / S pole), and an encoder sensor suitable for the scale is used. Note that the pitch of the scale pattern is a known value and is generally equal to the print resolution in the main scanning direction or 1 / n (n is a natural number) of the print resolution. The distance can be measured by counting the rising and falling edges of the encoder signal. In particular, in the case of 1 / n, there is a recording apparatus that uses a method of counting up to the print resolution internally. On the other hand, since the interval between the nozzle rows of the recording head is also a known value such as a value designed at m times the printing resolution (m is a natural number), the edge count number of the encoder signal corresponding to the delay distance is calculated in advance. The registration adjustment value can be obtained by counting the edge of the encoder signal during actual printing and starting printing when it becomes equal to the registration adjustment value.

  As another method for obtaining the registration adjustment value, there is a method in which a registration adjustment pattern is actually printed on a recording medium or the like, and a registration adjustment value is determined by allowing the operator to select a pattern with little color misregistration. As another method, there is a method of determining a registration adjustment value by printing a registration adjustment pattern on a recording medium or the like and detecting a deviation amount of each ink color by a sensor. These can also correct an error with respect to the design value of the interval between the nozzle rows of the recording head.

  Further, as a registration adjustment method, there is a method in which the above two methods (time delay, distance delay) are combined. Specifically, the registration adjustment in units of print resolution is performed with a distance delay based on the encoder signal, and the registration adjustment finer than the print resolution is performed with a time delay. This is effective for correcting that the nozzle row interval of each ink color of the recording head actually has an error from the design value (m times the printing resolution). Further, when obtaining a registration adjustment value from an actual registration adjustment pattern, registration adjustment with higher resolution can be performed. Further, there is a method in which registration adjustment finer than the printing resolution is performed based on a signal obtained by multiplying an encoder signal instead of a time delay.

  The reference position of the carriage includes a position where the carriage is stopped in a standby state (home position: abbreviated as HP), a reference position of a linear scale, a position of a separately provided sensor (start sensor, etc.), and the like. Used as a base point for counting encoder signals.

  These register values are set at the time of shipment or installation of the recording apparatus, before use by the user, or the like. In addition, the registration adjustment value is set by the above-described method or the like immediately after the recording head is replaced by a user or a service person. In addition, the user may readjust before printing.

  However, during printing, these registration adjustment values are fixed and are not changed. That is, the registration adjustment value is the same in any area of the printed image.

  The registration adjustment values are provided independently for each ink color nozzle row, and when performing reciprocal printing in the main scanning direction, registration adjustment values for the forward path and the backward path are provided, and the registration adjustment values are set for the forward path and the backward path. The registration adjustment value is fixed during movement in the main scanning direction.

  There is an inkjet image forming apparatus that dynamically corrects the timing of ink discharge pulse generation according to the distance between the recording head and the printing surface of the recording medium (hereinafter referred to as a head gap). This invention corresponds to the fact that the gap dynamically changes during the scanning of the carriage. The dynamic change of the head gap is caused by the floating of the recording medium due to various causes such as the properties of the recording medium and ink absorption (see, for example, Patent Document 1).

  However, the cause of misregistration is not only the dynamic change of the head gap due to the floating of the recording medium, but also various mechanical causes.

  In particular, the accuracy of the linear scale serving as a reference for registration adjustment may be caused by registration deviation.

  In general, the slit pitch of the linear scale has a certain amount of error, and a high-precision linear scale in which this error is reduced is generally expensive. Also, the longer the linear scale, the more likely this error will occur and the more expensive it will be to reduce the error. This linear scale error can be a position delay error in registration adjustment.

  Also, an ink jet recording apparatus in which not only four color images (Bk, C, M, Y) but also other special colors (red, green, blue, etc.) and light colors (light cyan, light magenta, etc.) are added. In addition, the arrangement of the nozzle rows of the respective ink colors becomes longer in the main scanning direction, and the position delay and time delay for registration adjustment also become longer, and there is a possibility that the amount of error and the chance of occurrence of errors increase. Also, the carriage becomes longer in the main scanning direction, and there is a high possibility that an error will occur in mechanical accuracy.

  In addition, even in a recording apparatus in which a plurality of recording heads having a plurality of nozzle rows are arranged in the main scanning direction and a recording apparatus in which a plurality of recording heads having one nozzle row are provided, the arrangement of the nozzle rows for each ink color is the main scanning. Since it becomes longer in the direction, the position delay and time delay for registration adjustment are also increased, and the possibility of occurrence of an error increases. Also, the carriage becomes longer in the main scanning direction, and there is a high possibility that an error will occur in mechanical accuracy.

  In addition, as an ink jet recording apparatus, a large format printer capable of printing on a wide recording medium (for example, 36 inches, 42 inches, 60 inches, etc.) has become widespread. Such a large-format printer must move at least by the width of the moving distance in the main scanning direction. During this long movement in the main scanning direction, the registration adjustment value is fixed. Therefore, the linear scale needs to be long, and the mechanical mechanism needs to be long in the main scanning direction, so that there is a high possibility that an error that causes registration deviation occurs in various mechanical accuracies.

  This large format printer is actively used as a graphic printer for printing high-definition images. Therefore, there is a demand for a clear image having no registration misregistration (color misregistration) over the entire area of a large print image. In particular, when the degree of registration (the degree of registration is the same, that is, the same meaning as the amount of registration deviation after registration adjustment) changes, the color of the image tends to change. This occurs because the degree of overlap of dots of each ink color changes as the degree of registration changes.

  In Patent Document 1, dynamic correction of the timing of generating an ink ejection pulse is performed according to a change in head gap with respect to an initial semi-fixed ejection timing time (delay time from a linear scale signal). A configuration is disclosed in which the correction timing value (delay time) is added and printing is performed at the true print timing. However, in the configuration of Patent Document 1, since the moving speed of the carriage in the main scanning direction is generally not constant or due to an error of the linear scale, an error may occur in the delay time or an absolute print position.

In addition, in the configuration of Patent Document 1, it is described that the correction of the print timing is performed by a high-speed processing circuit unit that is executed for each minimum interval unit of the linear scale. Requires a discharge timing (heat time), a semi-fixed discharge timing time (delay time from the linear scale signal), and a correction timing value (delay time) corresponding to the change in the head gap. When the correction timing becomes a large value, the heat time may overlap with the heat time of the next column. If the heat time is sufficiently shorter than the minimum interval of the linear scale or the time interval corresponding to the printing resolution, the occurrence of this overlap is reduced. However, at present, with the increase in the print resolution and the speed of movement of the carriage, the time interval corresponding to the minimum interval of the linear scale and the print resolution becomes shorter. When the number of nozzles in a row is increased and the nozzle is divided and driven (divided drive), the total heat time of divided drive per column becomes longer, and the possibility of overlapping becomes higher. there is a possibility.
JP-A-11-2914746

  In large-format printers, with a fixed registration adjustment value, the registration may not be able to be partially adjusted over the entire printing width, and the degree of registration deviation in the main scanning direction varies depending on the printing location, resulting in uneven color or blurred images. There is a problem that may occur.

  In particular, this problem may become prominent depending on the accuracy of the linear scale.

  In addition, when the device becomes wider, the length of the linear scale and the mechanical configuration become longer, and the possibility that a problem will occur further increases.

  Further, in a recording apparatus equipped with a plurality of recording heads in the main scanning direction, the position adjustment and the time delay of registration adjustment become longer, so that the occurrence of errors may increase, and this problem may become prominent.

  Further, in the conventional dynamic printing timing correction, there is a possibility that the correction accuracy may be deteriorated due to the printing speed, and that the heat time may overlap and the image may be disturbed.

[Description of partial misregistration]
The occurrence of this partial misregistration will be described below with reference to FIG.

  In FIG. 3, two recording heads a and b are mounted on a carriage, and position detection is performed by one encoder sensor. The mounting interval between the recording head a and the recording head b is L, and the ideal slit interval of the linear scale is d. While the carriage moves in the direction of the arrow in FIG. 3, the recording heads a and b eject ink droplets.

  This printing during movement will be partially explained.

First, printing is performed with the recording head a at the position X0, printing is performed with the recording heads a and b at the position X1, and the printing of the recording head a at the position X0 and the printing of the recording head b at the position X1 are overlapped. Therefore, when the registration adjustment value R1 is set, Expression 1 is established.
Formula 1: L = P1 * d + R1
Note that P1 is the number of encoder pulses counted from the position X0 to X1, the ideal delay from the recording head a of the recording head b with respect to the distance L is P1 × d, and R1 is the actual error. Correction of minutes.

Similarly, when printing is performed with the recording heads a and b at the position X2, and the printing of the recording head a at the position X1 and the printing of the recording head b at the position X2 overlap, Expression 2 is satisfied.
Formula 2: L = P2 * d-P1 * d + R1
P2 is the number of encoder pulses counted from positions X0 to X2.

That is, Formula 1 and Formula 2 to Formula 3 hold.
Formula 3: P2 = 2 × P1
Equation 3 corresponds to the linear scale between X0-X1 and X1-X2 being equal. That is, the linear scale has no distortion or error such as accumulated error.

However, if there is a cumulative error Δd (X1, X2) between X1 and X2,
Formula 4: P2 × d = 2 × P1 × d + Δd (X1, X2)
And the right side of Equation 3 is
Formula 5: P2 * d-P1 * d + R1 = 2 * P1 * d + [Delta] d (X1, X2) -P1 * d + R1
= P1 × d + R1 + Δd (X1, X2)
= L + Δd (X1, X2)
≠ L
This is a state in which the printing of the recording head a at the position X1 and the printing of the recording head b at the position X2 do not match, that is, a registration error has occurred.

  Thus, in the actual linear scale, the slit interval d includes a slight error, and since this error is not always the same in a partial range of the linear scale, the accumulated error Δd (Xm, Xn) is Since it differs depending on the location of the linear scale, the registration adjustment cannot be performed with the same registration adjustment value R1. Note that m <n, m, and n are integers.

  The accumulated error also changes depending on the width of the section Xm-Xn. When this width is small, the accumulated error can be almost ignored. Therefore, in a printing apparatus having a small carriage width and a narrow printhead interval L, the registration error due to the accumulated error is not caused. Inconspicuous. However, when the recording head interval L is increased, the accumulated error may not be ignored, and the registration deviation becomes conspicuous.

  Accordingly, in order to solve the above problems, an image recording apparatus according to the present invention includes a moving unit that relatively moves a recording medium and a recording head in a nozzle row arrangement direction, and an image formation that forms an image based on image data during the movement. An image recording position of each nozzle row at a certain image position, and a recording position adjusting means for adjusting an image recording position of each nozzle row of the recording unit in the image forming means with respect to a moving direction. Static position control control means for adjusting and controlling so as to overlap, and dynamic position for changing the recording position in the position adjusting means by a correction amount within a preset range at a plurality of image positions being moved by the moving means An image recording apparatus having a control means.

  The image position may be a print start position or a print end position during the relative movement.

  In the image recording apparatus, the plurality of moving image positions are cycles set in advance, and in particular, an integral multiple of a synchronization signal (HT_TRG) that is a reference for the printing timing of one column. is there.

  The correction amount in the preset range is a value smaller than the difference between the cycle of the synchronization signal (HT_TRG) serving as a reference for the printing timing of one column and the printing time of one column (HIGH width of HT_WIN). This is an image recording apparatus.

(1) In an image recording apparatus in which a recording head records an image on a recording medium while the recording section having a plurality of nozzle rows in which a plurality of nozzles are arranged and the recording medium move relatively in the nozzle row arrangement direction.
Moving means for relatively moving the recording medium and the recording head in the nozzle row arrangement direction;
Image forming means for forming an image based on image data during the movement;
A recording position adjusting unit that adjusts an image recording position of each nozzle row of the recording unit in the image forming unit with respect to a moving direction;
Static position control control means for adjusting and controlling the position adjustment means so that the image recording positions of the nozzle rows overlap at a certain image position, and recording at the position adjustment means at a plurality of image positions during movement by the movement means. An image recording apparatus comprising dynamic position control means for varying the position by a correction amount within a preset range.

  (2) The image recording apparatus according to (1), wherein the certain image position is a print start position or a print end position during the relative movement.

  (3) The plurality of moving image positions are cycles set in advance, and are in particular integral multiples of a cycle of a synchronization signal (HT_TRG) which is a reference for printing timing of each column. The image recording apparatus according to (1).

  (4) The correction amount in the preset range is a value smaller than the difference between the cycle of the synchronization signal (HT_TRG) serving as the reference for the printing timing of each column and the printing time of one column (HIGH width of HT_WIN). (1) The image recording apparatus according to (1).

  According to the image recording apparatus of the present invention, dynamic registration adjustment can be performed while the carriage is moving, and partial registration adjustment can be supported over the entire printing width, and the degree of registration misregistration varies depending on the printing location in the main scanning direction. This has the effect of suppressing the occurrence of color unevenness and blurred images due to the above.

  Further, by reflecting the characteristics of the linear scale and the mechanical characteristics in the dynamic registration adjustment amount, the above effect can be exhibited even for a printer having a wide print width.

  Further, by making the dynamic registration adjustment amount correspond to a plurality of nozzle rows and recording heads, the above-described effect can be exhibited even for a recording apparatus equipped with a plurality of recording heads in the main scanning direction.

  Further, since the reference of the printing timing in the dynamic registration adjustment is based on the linear scale, there is an effect that the correction accuracy does not deteriorate due to the printing speed.

  In addition, by providing a restriction on the difference between the registration adjustment values in the preceding and succeeding columns in the dynamic registration adjustment, there is also an effect that the heating time of each column is prohibited from being overlapped, and printing disturbance is prevented.

  In addition, because the dynamic registration adjustment range is not limited to one pixel or less, it takes into account processing of carry and carry when it extends over one pixel or more, so dynamic registration adjustment for a wide range of distortion factors There is also an effect that is possible.

  First, the principle of the present invention will be described.

In order to eliminate the registration error occurring in Equation 5, a dynamic registration adjustment value ΔR (X1, X2) is introduced,
Formula 6: L = P2 × d−P1 × d + R1 + ΔR (X1, X2)
And the printing of the recording head a at the position X1 and the printing of the recording head b at the position X2 are overlapped (registration is matched). This equation 6 is a new value obtained by adding the dynamic registration adjustment value ΔR (X1, X2) to the registration adjustment value R1 for matching the printing of the recording head a at the position X0 and the printing of the recording head b at the position X1. Therefore, the registration adjustment value R2 is set to a proper value.

When this ΔR (X1, X2) is obtained, from Equation 6 and Equation 5,
Expression 7: L = (2 × P1 × d + Δd (X1, X2)) − P1 × d + R1 + ΔR (X1, X2)
= P1 × d + R1 + Δd (X1, X2) + ΔR (X1, X2)
Further, from Formula 1 Formula 8: ΔR (X1, X2) = − Δd (X1, X2)
It becomes. This indicates that the dynamic registration adjustment value ΔR (X1, X2) is a value that cancels the accumulated error Δd (X1, X2) of the linear scale.

  When the dynamic registration adjustment value ΔR (Xm, Xn) is introduced between every section Xm-Xn, a new registration adjustment value Rn = R1 + ΔR (Xm, Xn) in each section is obtained.

  The accumulated error Δd (X0, X1) is also present in the section X0-X1 and can be expressed as the registration adjustment value R1 = R0 + ΔR (X1, X0). However, the registration adjustment value R0 of the previous section is required and is extended. Therefore, it is desirable to introduce a value after actual registration adjustment for R1. Therefore, ΔR (Xm, Xn) corrects the relative cumulative error Δd (Xm, Xn) with respect to the distance (P1 × d + R1) between the sections X1 and X2.

  By sequentially updating a new registration adjustment value Rn including such a dynamic registration adjustment value during carriage movement, it becomes possible to adjust the registration between the sections Xm-Xn.

  Next, embodiments of the present invention will be described with reference to the drawings.

  In the embodiment, an ink jet printer (hereinafter referred to as a printer) is described as an image forming apparatus.

[Main internal configuration of the printer]
FIG. 4 is a block diagram of the main internal configuration of the printer according to the embodiment of the present invention.

  In FIG. 4, reference numeral 1 denotes a host computer such as a personal computer that sends instructions such as image data to be printed, print parameters, and print commands. Reference numeral 2 denotes a printer according to an embodiment of the present invention. An interface unit (hereinafter abbreviated as I / F unit) 3 that controls the interface with the computer 1 and an image data processing unit that controls image processing of image data from the I / F unit 3 and writing and reading of image data 4 (inside this, an image memory unit 5 for storing image data and an image data transfer unit 6 for transferring the image data as print data to the recording heads 8 and 9) and a recording medium according to the print data Two recording heads 8 and 9 for ejecting ink to form an image, a carriage 7 mounted with the recording heads 8 and 9 and moved to the main scanning method, The linear scale 14 stretched over almost the entire scanning direction, the encoder sensor 10 that reads the pattern written on the linear scale 14 during the main scanning movement, the keys for operating the various printers 2, the operating status and the state An operation unit 11 having an LCD or the like for displaying, a sequence control unit 12 for performing various sequence controls (inside, although not shown, a ROM including a CPU and a program, a RAM for work, and various I / Os), A printing position control unit 13 that controls the printing position of the recording heads 8 and 9 to control the printing position, a carriage motor 19 that is a drive source for reciprocating the carriage 7 in the main scanning direction, and the rotation and position of the carriage motor 19. A carriage control unit 17 for controlling the recording medium, a conveyance motor 18 as a drive source for feeding the recording medium in the sub-scanning direction, and the conveyance motor 1 A conveyance control unit 16 for the rotation and position control, in the sub-scanning encoder 15 for detecting the movement amount in the sub-scanning direction of the recording medium, mainly constituted. In addition to these configurations, although not shown in FIG. 4, the printer 2 includes a power supply unit, an ink tank that supplies ink, a tube that is a path for supplying ink from the ink tank to the recording heads 8 and 9, and the like. The ink supply unit, the wiping unit that cleans the nozzle surfaces of the recording heads 8 and 9, the capping unit that contacts the nozzle surfaces of the recording heads 8 and 9 to protect the nozzle openings, and the ink from the recording heads 8 and 9 There is a suction part (a wiping part, a capping part, and a suction part are collectively referred to as a recovery unit in the future), and a mechanical that converts and transmits the rotation of the carriage motor 19 to the movement of the carriage 7 in the main scanning direction. A mechanical mechanism (hereinafter referred to as a scanning mechanism) or a mechanical mechanism (hereinafter referred to as a scanning mechanism) that converts and transmits the rotation of the conveyance motor 18 to the movement of the recording medium in the sub-scanning direction. Has also referred) etc. transport mechanism.

[Overview of print sequence]
An outline of the print sequence in the configuration of FIG. 4 will be described.

  First, when the power of the printer 2 is turned on, the sequence control unit 12 moves the carriage motor 19 via the carriage control unit 17 to move the carriage 7 to the reference position (home position) and restores it as an initialization process. Processing such as controlling the unit to perform recovery processing and driving the transport motor 18 via the transport control unit 16 to move the paper to the sub-scanning reference position of the platen (position where printing can be performed by the recording heads 8 and 9). The I / F unit 3 is set so as to be ready for printing and to receive a command from the host computer 1.

  Each command from the host computer 1 is received via the I / F unit 3, and the sequence control unit 12 controls each unit in the printer 2 in response to each command. At this time, in particular, when a print command is received, the image data following the print command is activated by the instruction of the sequence control unit 12 and stored in the image memory unit 5 in the image data processing unit 4 for image data processing. The image data in the image memory unit 5 is processed through the unit 4, converted into print data suitable for the recording heads 8 and 9, and stored in the image memory unit 5 again.

  Next, when the print data is stored in the image memory unit 5 by a predetermined amount (the amount required for the recording heads 8 and 9 to print one scan), the sequence control unit 12 passes the carriage motor via the carriage control unit 17. To move the carriage 7 in the main scanning direction.

  The movement of the carriage 7 generates pattern signals ENC_A and ENC_B (corresponding to ENC_A and ENC_B in FIG. 6) of the linear scale 14 from the encoder sensor 10, and the carriage control unit 17 uses the pattern signals ENC_A and ENC_B. The servo control of the motor 19 is performed, and the print position control unit 13 finds the print start position of each ink color nozzle row in each of the recording heads 8 and 9 and reads the print data from the image memory unit 5 to the image data transfer unit 6. The recording heads 8 and 9 are sent to head drive units (inside the recording heads 8 and 9) corresponding to the respective ink color nozzle rows, and ink droplets are ejected from the recording heads 8 and 9 in accordance with the drive data. An image for one scan is formed in the main scanning direction.

  Next, the conveyance control unit 16 drives the conveyance motor 18 to convey the paper by a predetermined amount, and stops the paper at a position for printing in the next scan.

  During this time, the image data processing unit 4 similarly stores the image data corresponding to the print data of the next scan in the image memory unit 5, processes the image data in the image memory unit 5, and prints suitable for the recording heads 8 and 9. It is converted into data and stored in the image memory unit 5 again.

  When the print data for the next scan is stored in the image memory unit 5 by a predetermined amount, the sequence control unit 12 controls each unit to perform printing for the next scan in the same manner as described above.

  When these printing sequences are performed until the last scan (until the end of the page) and all printing is completed, the sequence control unit 12 moves the carriage 7 to the home position and prepares for the next printing command or the like. .

[Outline of print position control unit]
An outline of the print position control unit 13 in FIG. 4 will be described with reference to FIG.

  The print position control unit 13 includes a reference signal generation unit 30 that generates various reference signals (HT_TRG, REZI_INT, BLK_TRG, D_CLK, D_LAT, HT_ENB, etc.), a registration adjustment unit 31 that performs registration adjustment for each pixel, A registration adjustment unit 32 that performs registration adjustment within one pixel is provided.

  Each reference signal here will be briefly described. HT_TRG is a signal serving as a reference for the print position corresponding to the resolution of the recording heads 8 and 9 in the main scanning direction, and REZI_INT is used when dynamic registration adjustment is performed. It is a signal that serves as a dynamic registration change trigger, BLK_TRG is a signal that serves as a reference for divisional driving of the recording heads 8 and 9, D_CLK is a clock for driving data transfer to the recording heads 8 and 9, and D_LAT is: The recording heads 8 and 9 are signals for latching drive data in the register, and HT_ENB is a heat signal for causing the recording heads 8 and 9 to discharge according to the driving data.

  The sequence control unit 12 also supplies various signals (registration change interval N, resolution ratio DPI, block division number Nb, etc.) and a CLK signal to the reference signal generation unit 30, and prints the print width W and each nozzle to the registration adjustment unit 31. Variables such as a registration adjustment value Rp for each pixel for each column and the number of heat blocks Nh are supplied, and a registration adjustment value Rt within one pixel for each nozzle column is supplied to the registration adjustment unit 32.

  First, various reference signals and related signals and operations in the reference signal generation unit 30 will be described with reference to FIG. All signals are expressed in positive logic.

  When the carriage 7 starts to move in the main scanning direction (outward) for printing, encoder signals ENC_A and ENC_B are output from the encoder sensor 10 as shown in FIG. 6 and supplied to the reference signal generator 30. The encoder signals ENC_A and ENC_B are signals having a phase difference of 90 degrees, and each period T is the pitch d of the linear scale 14 and the moving speed V of the carriage 7 in the main scanning direction, and T = d / V. Although the period T changes when the speed V varies, the positions of the carriage 7 are substantially equal at the encoder pitch d between the rising edges of the encoder signals ENC_A and ENC_B (that is, the period T). After synchronizing the encoder signal ENC_A with the system clock CLK, a one-shot signal ENC_CLK is generated with reference to the rising edge of ENC_A. Since the period of CLK (for example, 10 nsec) is sufficiently small with respect to the period T of ENC_A (for example, d = 300 dpi, T = about 66 μsec when V = 50 inch / sec), the period of ENC_CLK is the period of ENC_A. And can be used as a signal representing the position reference of the carriage 7. Numbers 1 to 16 given to CLK in the figure are numbers entered for convenience in measuring the period T of ENC_CLK, and even larger numbers are required depending on the ratio of the actual period of CLK and the period T of the encoder signal. . For example, if the period of CLK is 10 nsec and the period T of ENK_CLK is 66 μsec, the number increases to 6600. The system clock CLK is supplied from the sequence control unit 12 to each unit such as the reference signal generation unit 30.

  Next, in the reference signal generation unit 30, ENC_CLK is multiplied by DPI (in FIG. 6) according to the ratio DPI (described with DPI = 2 in FIG. 6) of the printing resolution in the main scanning direction and the resolution of the linear scale 14. The signal HT_TRG having a finer period is generated by being multiplied by 2 and supplied to the registration adjusting unit 31. Therefore, the HT_TRG signal is a signal serving as a reference for the print position corresponding to the resolution of the recording heads 8 and 9 in the main scanning direction. The resolution ratio DPI is a variable notified from the sequence control unit 12 to the reference signal generation unit 30. In the following description, the linear scale 14 has a resolution of 300 dpi, DPI = 2, and a print resolution of 600 dpi. In the double multiplication method, here, the period T of ENC_CLK is measured by counting CLK, and HT_TRG is generated in half of the period. Note that the generation of the reference signal HT_TRG signal having a resolution higher than the resolution of the encoder includes a method using both edges of ENC_A and a method using both edges of ENC_A and ENC_B, and the method need not be limited here.

  Further, in the reference signal generation unit 30, when the carriage 7 starts to move in the main scanning direction (outward) for printing and ENC_CLK and HT_TRG are generated by the encoder signals ENC_A and ENC_B, the period of the HT_TRG signal is divided by Nb. A signal is generated and supplied to the registration adjustment unit 32 and the image data transfer unit 6 in one pixel. The BLK_TRG signal is a signal used to divide and discharge the nozzles in the nozzle row (hereinafter referred to as divided driving). In the registration adjusting unit 32 in one pixel, the BLK_TRG signal is registered in one pixel. This is a reference signal for adjustment. When the block division number Nb = 16, the BLK_TRG signal becomes a signal obtained by dividing the period of the HT_TRG signal by 16, and the registration adjustment within one pixel can be performed with the accuracy of dividing one pixel into 16 parts. That is, here, the registration adjustment can be performed in units of 1/16 of 600 dpi.

  In the reference signal generation unit 30, reference signals D_CLK, D_LAT, and HT_ENB for transferring print data to the recording heads 7 and 8 are generated based on BLK_TRG and supplied to the image data transfer unit 6. The D_CLK signal is a clock for serial transfer, and data is transferred at both rising and falling edges, and the number of edges (20 in FIG. 6) is the same as the number of data to be transferred (DATA & BLK signal is expressed in 20 bits). is there. The D_LAT signal is a latch signal, and is generated during the Low period of the D_CLK signal (in FIG. 6, it is synchronized with BLK_TRG). HT_ENB is a heat pulse signal, and is generated so as to fall within the period of the BLK_TRG signal in the Low section of the D_LAT signal.

  Next, the operations of the registration adjustment units 31 and 32 will be described separately for the case before the registration adjustment, the case where the static registration adjustment is performed, and the case where the dynamic registration adjustment is performed.

[Print timing before register adjustment]
First, the case before registration adjustment (Rp = 0, Rt = 0) will be described with reference to FIGS.

  FIG. 7 is a timing diagram for explaining details of printing timing for one nozzle row, and FIG. 8 is a timing diagram for explaining a relationship of printing timing for each nozzle row of the recording heads 8 and 9. . FIG. 5 shows internal and peripheral blocks of the print position control unit 13 and signals between them.

  In the registration adjustment unit 31 for each pixel to which the HT_TRG signal generated in the reference signal generation unit 30 is supplied, the rising edge of the HT_TRG signal is counted by an internal counter (not shown), and the count value of FIG. When CNT is output and the comparator (not shown) in the registration adjustment unit 31 matches the count value CNT with a certain set value ST + Rp, the WIN signal is set to HIGH, and another comparator (not shown) in the registration adjustment unit 31 is output. ) Matches the set value END + Rp (= ST + Rp + W), the WIN signal is set to LOW. Here, since it is before registration adjustment, Rp = 0, and before registration adjustment, Rp is ignored.

  The set value ST indicates a print start position for one scan, and the set value END indicates a print end position for one scan. Note that the variable W corresponds to the print width. The set values ST and END are notified from the sequence control unit 12 to the registration adjusting unit 31 before printing. In FIG. 7, ST = 1000, END = 7000, W = 6000 (corresponding to 10 inches at 600 dpi). Accordingly, the WIN signal is a signal representing a printing section of an ink color nozzle row of the recording heads 8 and 9, and printing is performed at the nozzle row while the WIN signal is HIGH.

  As shown in FIG. 8, the WIN signal exists for each ink nozzle row of the print heads 8 and 9, and from the BK nozzle row to the Y nozzle row of the print head 8 and from the Y nozzle row of the print head 9 to BK. Assuming that the WIN signals to the nozzle rows are WIN_BK1, WIN_C1, WIN_M1, WIN1, WIN_Y2, WIN_M2, WIN_C2, and WIN_BK2, in the nozzle row arrangement as shown in FIG. 2, there are intervals d1 to d7 between the nozzle rows. When moving in the forward direction, with reference to the BK nozzle array of the recording head 8, WIN_BK1, WIN_C1, WIN_M1, WIN_Y1, WIN_Y2, WIN_M2, WIN_C2, and WIN_BK2 correspond to the BK nozzle array of the recording head 8, respectively. Rise by the interval between nozzle rows (ie Become one scan of the print start) is delayed it. However, since the print width W is common to each nozzle row, the HIGH width of the WIN signal of each nozzle row is the same, and the trailing edge (that is, the end of printing for one scan) is similarly delayed.

  In the case of the return pass, if the BK nozzle row of the recording head 8 is used as a reference, the WIN signal of each nozzle row advances by the interval between the nozzle rows with respect to the WIN signal WIN_BK1 of the BK nozzle row of the recording head 8. Become a direction.

  In generating each WIN signal, the set values ST and END exist independently for each nozzle row of each recording head 8 and 9, and are denoted as ST_ * and END_ *. In addition, * here represents each of each nozzle row BK1, C1, M1, Y1, Y2, M2, C2, BK2.

  Similarly, the one-pixel unit registration adjustment value Rp and the one-pixel registration adjustment value Rt are also provided independently for each nozzle row of the recording heads 8 and 9, and are provided independently for the forward path and the backward path. Rp _ * _ FW is the forward registration adjustment value for each pixel, Rp _ * _ BW is the backward registration adjustment value for each pixel, Rt _ * _ FW is the backward registration adjustment value for each pixel, and Rt _ * _ BW is the backward registration adjustment value for each pixel. (* Also represents each of the nozzle rows BK1, C1, M1, Y1, Y2, M2, C2, and BK2). These registration adjustment values are all initialized to 0 before registration adjustment.

  In order to generate each WIN signal of each nozzle row, set values ST and END are also provided for each nozzle row (specifically, provided as a register of the one-pixel registration adjusting unit 31), and intervals d1 to d7 between the nozzle rows. In addition, the sequence controller 12 sets the set values ST and END in consideration of the print width W and the forward / return path, and a comparator for generating each WIN signal is also provided in the registration adjustment unit 31 for each nozzle row. Note that one counter is sufficient to count HT_TRG. Basically, the ST and END values set in the forward path are set to END by adding the variable α in the ST value in the return path, and the variable α in the END value is added to the ST and set to ST. The cash register can be roughly adjusted. Note that the variable α corresponds to a registration deviation amount between the forward path and the backward path due to a delay in a certain section (synchronized with HT_TRG) of processing in the printer 2 after the WIN signal rises (recording heads 8 and 9 and the printer 2). Does not include individual differences).

  In order to simplify the detailed description of the print timing controlled by the registration adjusting unit 32, the WIN signal described for one nozzle row will be described with reference to FIG.

  In FIG. 7, the WIN signal is expressed for one nozzle row (for example, BK1), and the registration adjustment unit 31 causes the WIN signal to rise at ST = 1000 when the registration adjustment value Rp = 0 for one pixel. ing. The WIN signal is supplied to the registration adjustment unit 32, and a counter (not shown) in the registration adjustment unit 32 counts rising edges of the BLK_TRG signal and outputs a count value (REZI_CNT). When the counter (not shown) compares the count value (REZI_CNT) with the registration adjustment value Rt (Rt = 0 in FIG. 7), the TRNS_WIN signal is set to High. A block in which another counter (not shown) and a comparator (not shown) in the registration adjusting unit 32 count BLK_TRG after the TRNS_WIN signal becomes High and output a count value (BLK_CNT), and print it with the comparator. When the number of heat blocks Nh (12 in FIG. 7), which is the total number, coincides with the count value (BLK_CNT), the TRNS_WIN signal is set to Low. Note that the REZI_CNT counter is cleared when the HT_TRG signal is High, and the BLK_CNT counter is cleared when the TRNS_WIN signal is Low, and is enabled when the TRNS_WIN signal is High. The heat block number Nh, which is the total number of blocks to be printed, is a variable supplied from the sequence control unit 12 to the registration adjustment unit 32.

  In FIG. 7, since the registration adjustment value Rt = 0 in one pixel (before registration adjustment) is shown for the WIN signal shown for one nozzle row, the TRNS_WIN signal is high when REZI_CNT = 0. And BLK_CNT = 12, which is Low.

  Thus, the TRNS_WIN signal generated by the registration adjustment unit 32 is supplied to the image data processing unit 4 (particularly the image data transfer unit 6).

  When the TRNS_WIN signal is in the high state, the image data transfer unit 6 sequentially generates the address signal ADDR in which the print data & block number data (DATA & BLK) is stored and generates the read signal RD, and the image memory unit 5 Are sequentially read out from the print data & block number data (DATA & BLK), stored in the shift register (not shown) of the image data transfer unit 6, and sent to the corresponding nozzle row of the recording head 8 by the D_CLK_A signal. Data reading and transfer for the number of heat blocks Nh to be printed while the TRNS_WIN signal is High are performed.

  Note that the D_CLK_A signal and the DCLK_B signal are signals common to the print heads 8 and 9 through the same signal as the D_CLK signal in two paths. The RD signal, the ADDR signal, the DATA & BLK signal, and the HT_ENB signal are individual signals for each nozzle row of the recording heads 8 and 9, and are distinguished as RD_ *, ADDR_ *, DATA & BLK_ *, and HT_ENB_ *. To do. In addition, * represents each of each nozzle row BK1, C1, M1, Y1, Y2, M2, C2, BK2. 5 and 7, “*” is represented by A and B, which indicates that A corresponds to the nozzle row with the recording head 8 and B corresponds to the nozzle row with the recording head 9. Is shown.

  Further, the image data transfer unit 6 generates an HT_WIN signal that is delayed by one cycle of BLK_TRG from the TRNS_WIN signal, and outputs the HT_ENB_A signal only when the HT_WIN signal is in the High state (specifically, the HT_ENB signal is input to the HT_WIN signal). Gate on).

  The image data transfer unit 6 supplies the above-described signals (DATA & BLK_A or DATA_ & BLK_B, D_CLK_A or D_CLK_B, D_LAT_A or D_LAT_B, HT_ENB_A or HT_ENB_B) to the recording head 8 (or recording head 9).

  In the recording head 8 (or recording head 9), the DATA & BLK_A (or DATA & BLK_B) signal is captured by the D_CLK_A (or D_CLK_B) in the shift register (not shown), and the data DATA & BLK_A (or DATA & BLK_B) serially transferred by the D_LAT_A (or D_LAT_B) is recorded. A plurality of nozzles latched in a register (not shown) in the head 8 (or recording head 9), and the latched print data (DATA_A or DATA_B) is ON (1) and the corresponding block (BLK_A or BLK_B) In contrast, while HT_ENB_A (or HT_ENB_B) is High, a heater (not shown) in the recording head 8 (or recording head 9) is energized, and ink droplets are ejected.

  While the HT_WIN signal is High, an image to be printed corresponds to one column (one row) of a nozzle row of the recording head 8 (or recording head 9), and these operations are performed while the WIN signal is High. By repeating the same, one-scan printing is formed.

  In each of the recording heads 8 and 9, there are the same shift register and latch for each nozzle row, and the processing is performed in parallel for each nozzle row. At this time, DCLK_A and D_LAT_A are commonly used in each nozzle row in the recording head 8, and DCLK_B and D_LAT_B are commonly used in each nozzle row in the recording head 9. DATA & BLK_A (or DATA & _BLK_B) and HT_ENB_A (HT_ENB_B) are independent signals in the print heads 8 and 9 and independent in each nozzle row in the print head.

  Although the WIN signal for one nozzle row has been described above, the other nozzle rows operate in the same manner. Before the registration adjustment, the registration adjustment value Rp _ * = 0 for each pixel and the registration adjustment value Rt _ * = 0 within one pixel, so the same operation is performed for WIN_ * of each nozzle row.

  Since each nozzle row performs printing for one scan during each WIN_ * High and before registration adjustment, the printing of each nozzle is slightly different due to mechanical differences or recording head differences. Printing is performed with the cash register misaligned.

[Print timing: With static registration adjustment]
Next, the case of performing static registration adjustment will be described with reference to the timing chart of FIG. In order to simplify the description here, a nozzle row of a certain recording head will be described with reference to FIG.

  First, the registration adjustment for each pixel will be described. In FIG. 9, the registration adjustment value Rp = 1 for one pixel, the print start position ST for one scan ST = 1000, the print end position END for one scan 7000, and the print width W = 6000, the registration adjustment unit 31 for each pixel supplied with the HT_TRG signal generated in the reference signal generation unit 30 counts rising edges of the HT_TRG signal with an internal counter (not shown). 9 count value CNT is output, and when the comparator (not shown) in the registration adjustment unit 31 matches the count value CNT with a certain set value ST + Rp = 1001, the WIN signal is set to HIGH, and the other in the registration adjustment unit 31 When the comparator (not shown) of FIG. 5 matches a certain set value END + Rp (= ST + Rp + W) = 7001, the WIN signal is set to LOW.

  Accordingly, the signal WIN signal representing the print section of the ink color nozzle row is delayed by the registration adjustment value Rp (= 1 pixel) of one pixel as compared with FIG. 7, and this WIN signal is converted into one pixel. To the cash register adjustment unit 32.

  The set values ST, END, and Rp are notified from the sequence control unit 12 to the registration adjustment unit 31 before printing. Although the registration adjustment value Rp for each pixel has been described as 1 this time, this is not limited to this, and Rp is an integer, -1, -2,. . . It is also possible to take negative values such as In addition, there is a case where the range of Rp as the expected range of machine registration is limited to an integer such as -n to n.

  Next, registration adjustment within one pixel will be described.

  When the registration adjustment unit 32 in one pixel receives the WIN signal in which the registration adjustment in one pixel unit is performed by the registration adjustment unit 31 in one pixel unit, a counter (not shown) in the registration adjustment unit 32 displays the BLK_TRG signal. And the count value (REZI_CNT) is output, and the comparator (not shown) in the registration adjustment unit 32 compares the count value (REZI_CNT) with the registration adjustment value Rt (Rt = 1 in FIG. 9). If they match, the TRNS_WIN signal is set to High. A block in which another counter (not shown) and a comparator (not shown) in the registration adjusting unit 32 count BLK_TRG after the TRNS_WIN signal becomes High and output a count value (BLK_CNT), and print it with the comparator. When the number of heat blocks Nh (12 in FIG. 9), which is the total number, coincides with the count value (BLK_CNT), the TRNS_WIN signal is set to Low. Note that the REZI_CNT counter is cleared when the HT_TRG signal is High, and the BLK_CNT counter is cleared when the TRNS_WIN signal is Low, and is enabled when the TRNS_WIN signal is High. The heat block number Nh is a variable supplied from the sequence control unit 12 to the registration adjusting unit 32.

  In FIG. 9, since the registration adjustment value Rt = 1 in one pixel (static registration) is shown for the WIN signal shown for one nozzle row, the TRNS_WIN signal is high when REZI_CNT = 1. And REZI_CNT = 13 (BLK_CNT = 12) and Low. Therefore, compared with FIG. 7, the TRNS_WIN signal delayed by the registration adjustment value Rt = 1 (block) in one pixel is generated each time in each column (in the cycle of HT_TRG) while the WIN signal is High. .

  In this way, the TRNS_WIN signal generated by the registration adjustment unit 32 is supplied to the image data processing unit 4 (particularly the image data transfer unit 6), and the image data transfer unit 6 performs the TRNS_WIN signal similarly to the description of FIG. Thus, the HT_WIN signal delayed by one cycle of BLK_TRG is generated, and the HT_ENB_A signal is output only when the HT_WIN signal is in the High state (specifically, the input HT_ENB signal is gated by the HT_WIN signal).

  The image data transfer unit 6 supplies the above-described signals (DATA & BLK_A or DATA_ & BLK_B, D_CLK_A or D_CLK_B, D_LAT_A or D_LAT_B, HT_ENB_A or HT_ENB_B) to the recording head 8 (or recording head 9).

  The recording head 8 (or recording head 9) performs ejection based on a signal sent in the same manner as described in FIG.

  In FIG. 9, since Rp = 1 and Rt = 1, the discharge in FIG. 9 that has undergone registration adjustment is delayed by 1 pixel + 1/16 pixel with respect to the discharge in FIG.

  The in-pixel registration adjustment value Rt need not be limited to 1, and a value from 0 to 15 can be taken in the case of FIG. In addition, although it is possible to set Rt to a negative number, it is also possible to convert Rt to a positive number by reflecting it in Rp as a carry. In addition, when Rt exceeds 15, it is possible to make Rt 15 or less by reflecting it in Rp as a carry.

  Although the resolution of the registration adjustment value Rt in one pixel is 1/16 and FIG. 9 has been described with reference to FIG. 9, it is not limited to this, and Nb = 24, 20, 15, etc. may be used. It must be larger than the number of blocks Nh. This condition is particularly necessary when the division driving cycle and the resolution of registration adjustment within one pixel are combined.

  Here, the division drive cycle and the registration adjustment resolution are combined, but this is not restrictive. For example, REZI_CNT can be changed to a divided driving cycle by performing registration adjustment in units of 1/8 image so that the rising edge of BLK_TRG is counted twice.

  Further, the number of heat blocks Nh = 12 for division driving need not be limited to this value, and can be changed by the structure of the recording heads 8 and 9 or can be treated as a variable.

  Further, even when the recording heads 8 and 9 are not divided and driven, registration adjustment within one pixel unit and within one pixel is possible. In this case, the registration adjustment for each pixel is the same as in FIG. 9, but the registration adjustment within one pixel can be similarly performed using REZI_CNT. It should be noted that the TRNS_WIN signal can be created by defining the High length not by the number of divided blocks but by, for example, the image transfer size, and can be processed in the same manner after reading and transfer of image data.

  In addition, registration adjustment values Rp and Rt within one pixel unit and one pixel are provided independently for each nozzle row of each recording head 8 and 9 for each nozzle row, and a forward registration adjustment value for each pixel row for each nozzle row is obtained. Rp _ * _ FW, Rp _ * _ BW is a return registration adjustment value in one pixel unit, Rt _ * _ FW is a forward registration adjustment value in one pixel, and Rt _ * _ BW is a return registration adjustment value in one pixel. Each of the nozzle rows BK1, C1, M1, Y1, Y2, M2, C2, and BK2) can be used to perform the above-described one-pixel unit and registration adjustment within one pixel. At this time, it is possible to have the registration adjustment units 31 and 32 for the forward path and the return path for each nozzle row. However, one registration adjustment unit 31 and 32 for the forward path and the return path may change between the forward path and the return path. In addition, the variable (Rp, ST, END, Rt, etc.) set from the sequence control unit 12 to the registration adjustment units 31 and 32 is changed between the return path value and the forward path value between the return path and the forward path. It is also possible to respond by setting.

[Print timing: With dynamic registration adjustment]
Next, the case of performing dynamic registration adjustment will be described with reference to the timing chart of FIG. In order to simplify the description here, a nozzle row of a certain recording head will be described with reference to FIG.

  First, registration adjustment in units of one pixel will be described. Similarly to FIG. 9, in FIG. 10 as well, one pixel registration adjustment value Rp = 1, one scan print start position ST = 1000, one scan print end position END = 7000 Assuming that the print width W is 6000, the rising edge of the HT_TRG signal is detected by an internal counter (not shown) in the registration adjustment unit 31 of one pixel unit to which the HT_TRG signal generated in the reference signal generation unit 30 is supplied. 10 is output, and when the comparator (not shown) in the registration adjustment unit 31 matches the count value CNT with a certain set value ST + Rp = 1001, the WIN signal is set to HIGH, and the registration adjustment unit When another comparator (not shown) within 31 matches a set value END + Rp (= ST + Rp + W) = 7001, the WIN signal is set to LOW.

  Accordingly, the WIN signal representing the print section of the ink nozzle row is delayed by the registration adjustment value Rp (= 1 pixel) of one pixel as compared with FIG. 7, and this WIN signal is included in one pixel. To the cash register adjustment unit 32.

  9 and FIG. 10 are largely different from each other in the dynamic registration adjustment, a signal called REZI_INT signal is generated from the reference circuit generation unit 30, and when the REZI_INT signal is received by the sequence control unit 12, the registration adjustment value in one pixel is obtained. This is the point where Rt is communicated to the registration adjustment unit 32. The cycle of the REZI_INT signal is set at a registration change interval N. In FIG. 10, the sequence control unit 12 notifies the reference signal generation unit 30 that N = 2, and the reference circuit generation unit 30 detects the rising edge of the HT_TRG signal internally. The counter controller 12 generates a one-shot high signal for the REZI_INT signal every N (= 2) cycles of the HT_TRG signal, and the sequence control unit 12 sets the rising edge of the REZI_INT signal to the internal CPU (not shown). In the interrupt processing such as FIG.), The registration adjustment value Rt in one pixel is updated to the registration adjustment unit 32. When receiving the notification, the registration adjustment unit 32 reflects the updated Rt in the next HT_TRG cycle, and performs registration adjustment within one pixel as in FIG.

  In FIG. 10, the registration adjustment value Rt within one pixel is updated when the count value CNT = 1000 of the HT_TRG signal is Rt = 1 (1 is set before the start of printing for one scan), CNT = 1003 Is updated to Rt = 0, CNT = 1005 is updated to Rt = 1, and is roughly (similarly updated every two cycles of the HT_TRG signal, the value is not shown), and CNT = 6999 is updated to Rt = 2. This is held until printing for one scan is completed and settings for the next scan are made. With respect to the update of Rt, the reflection of Rt in the registration adjusting unit 32 is actually reflected with a delay of one cycle of the HT_TRG signal. Therefore, Rt = 1 when CNT = 1001 to 1003, and Rt when CNT = 1004 to 1005. = 0, CNT = 1006 to 1007, Rt = 1, medium, CNT = 7000, Rt = 2.

  The registration adjustment within one pixel at CNT = 7000 will be described with reference to FIG. 10. Rt = 2 notified to be updated at CNT = 6999 is validated as a new Rt immediately before CNT = 7000. Specifically, two registers for Rt (not shown) are prepared in the registration adjusting unit 32, one is used as the current Rt, and another is used as the new Rt, and the falling edge of TRNS_WIN is used. It can respond by switching. After switching, the new Rt register becomes the current Rt and the free Rt register is used for the next update. A flag register (not shown) indicating whether or not to switch is also prepared in the registration adjusting unit 32, and the flag is set when Rt is updated from the sequence control unit 12, and the flag is cleared after switching the register. Thus, the reflection of the new Rt after the update of Rt occurs.

  When Rt = 2 is reflected, since WIN signal = High when CNT = 7000, as in FIG. 9, the counter (not shown) in the registration adjustment unit 32 counts the rising edge of the BLK_TRG signal and counts it. (REZI_CNT) is output, and the comparator (not shown) in the registration adjustment unit 32 compares the count value (REZI_CNT) with the registration adjustment value Rt (Rt = 2 in FIG. 10). Set to High. A block in which another counter (not shown) and a comparator (not shown) in the registration adjusting unit 32 count BLK_TRG after the TRNS_WIN signal becomes High and output a count value (BLK_CNT), and print it with the comparator. When the number of heat blocks Nh (12 in FIG. 10), which is the total number, and the count value (BLK_CNT) match, the TRNS_WIN signal is set to Low. Note that the REZI_CNT counter is cleared when the HT_TRG signal is High, and the BLK_CNT counter is cleared when the TRNS_WIN signal is Low, and is enabled when the TRNS_WIN signal is High.

  Since Rt = 2 when CNT = 7000, the TRNS_WIN signal becomes High when REZI_CNT = 2, and is Low when REZI_CNT = 14 (BLK_CNT = 12). Accordingly, a TRNS_WIN signal delayed by the registration adjustment value Rt = 2 (block) in one pixel is generated as compared with FIG.

  The following operations are the same as those in FIG. 9, and the TRNS_WIN signal generated by the registration adjustment unit 32 is supplied to the image data processing unit 4 (particularly the image data transfer unit 6), and the image data transfer unit 6 Similarly to the description, the HT_WIN signal delayed by one cycle of BLK_TRG from the TRNS_WIN signal is generated, and the HT_ENB_A signal is output only when the HT_WIN signal is in the high state (specifically, the input HT_ENB signal is gated by the HT_WIN signal). ).

  As described above, the registration adjustment value Rt in one pixel is updated and reflected at the registration change interval N, so that the registration adjustment in one pixel is performed at a different Rt while the WIN signal is High.

  It is not necessary to limit to the registration change interval N = 2 used here, it is an arbitrary variable, and can be realized by setting in advance for each machine or automatically or manually when actually used. .

  The dynamic registration adjustment is to perform registration adjustment with different registration adjustment values for one nozzle row during one scan as shown in FIG. 10, and the same for one nozzle row during one scan as shown in FIG. Performing the registration adjustment with the registration adjustment value is referred to as static registration adjustment.

  In FIG. 10, the dynamic registration adjustment for one nozzle row has been described. However, it can be similarly performed by updating Rt by generating a REZI_INT signal for the other nozzle rows. At that time, the registration update interval N is updated with the smallest registration update interval N among the plurality of nozzle rows because the hardware scale of the circuit or the like becomes smaller if the registration update interval N is handled in common for all nozzle rows. We can cope by preparing. If the circuit scale is allowed to increase, the registration update interval N can be provided for each nozzle row, or for each group of nozzle rows (for example, for each print head 8, 9).

  Also, instead of making the registration update interval N constant during one scan, it is possible to make N variable, but the variable state is stored by a number of columns in the RAM, mathematical expressions on the program, etc. It is necessary to keep.

  Further, Rt to be updated can be dealt with by storing it in a RAM, a ROM or the like in the sequence control unit 12 and sequentially reading it by an interrupt process activated by a REZI_INT signal, but this is not restrictive. It is also possible to store the Rt to be updated in advance in a storage unit such as a RAM in the registration adjustment unit 32 and update it with the REZI_INT signal. In addition, since Rt to be updated corresponds by being provided independently for each nozzle row, the registration adjustment unit 31 for each pixel, the registration adjustment unit 32 in one pixel, and the image data processing unit 4 are provided for each nozzle row. It is provided independently, and the reference signal generation unit 30 corresponds by using it in common. In addition, Rt is stored as a data string in the storage unit. However, the present invention is not limited to this, and Rt can be mathematically stored from the distortion characteristics of the linear scale.

  In FIG. 10, Rt used for dynamic registration adjustment is 1, 0, 2, but this is not the case, and takes the value of the registration adjustment range 0 to 15 for one pixel similarly to the registration adjustment value Rt of FIG. 9. It is possible. The registration adjustment range 0 to 15 need not be limited to this range, and varies depending on the number of block divisions Nb for dividing one column cycle (cycle of the HT_TRG signal). In this description, since Nb = 16, the range of Rt is 0 to 15. If Nb = 8, the range of Rt changes from 0 to 7.

  However, the printing timing of the current column (High period of the TRNS_WIN signal) cannot overlap the timing of the previous column or the timing of the subsequent column. Therefore, the difference ΔRt = Rt (−1) −Rt between the registration adjustment value Rt (−1) used in the previous column and the registration adjustment value Rt used in the current column is a cycle of one column (cycle of HT_TRG) − One column heat time (HT_WIN HIGH width) cannot be exceeded.

Note that one column cycle (HT_TRG cycle) is a block division number Nb × 1 block cycle (BLK_TRG cycle), and −1 column heat time (HT_WIN HIGH width) is the number of heat blocks Nh × 1. When the cycle of the block (the cycle of BLK_TRG) and the unit of the registration adjustment value Rt in one pixel is also a cycle of one block,
ΔRt = Rt (−1) −Rt ≦ Nb−Nh
There is a relationship.

Similarly, for the difference ΔRt between the registration adjustment value Rt used in the current column and the registration adjustment value Rt (+1) of the subsequent column, ΔRt = Rt−Rt (+1) ≦ Nb−Nh
There is a relationship.

  Even in the case of a recording head that does not perform block division driving, the difference ΔRt of the registration adjustment value Rt does not exceed one column cycle (HT_TRG cycle) -1 column heat time (HT_WIN HIGH width) as described above. Can not.

  The limitation on the difference ΔRt of the registration adjustment value Rt will be described with reference to FIG.

  [A] in FIG. 11 is a case where there is no dynamic registration adjustment (in the case of static registration adjustment), the registration adjustment value Rt = 2 in the current column, the previous column, and the subsequent column, and the block division number Nb = 16, the number of heat blocks Nh = 12, and therefore the difference ΔRt of the registration adjustment value Rt is always 0. The cycle for one column is a time corresponding to the number of block divisions Nb = 16 minutes, and the heat time for one column is a time corresponding to the number of heat blocks Nh = 12 minutes. Since time = 4 blocks, the relationship of “registration adjustment value Rt difference ΔRt ≦ 1 column cycle-1 column heat time” is always satisfied.

  [B] in FIG. 11 is a case of dynamic registration adjustment, and the number of block divisions Nb = 16 and the number of heat blocks Nh = 12 are the same as [A], but registration adjustment values Rt = 2, 0, 1 Therefore, the difference ΔRt of the registration adjustment value Rt changes to 2 and −1, but always satisfies the relationship of “registration adjustment Rt difference ΔRt ≦ 1 column cycle-1 column heat time”. ing.

  [C] in FIG. 11 is also a case of dynamic registration adjustment, where the block division number Nb = 16 and the heat block number Nh = 12 are the same as [A], but the registration adjustment value Rt = 2, 5, 0. Since the difference ΔRt of the registration adjustment value Rt changes to −3, 5 because it is varied, the relationship of “registration adjustment Rt difference ΔRt ≦ 1 column cycle-1 column heat time” is not satisfied ΔRt = 5 has occurred. The black shaded portion in the figure shows a state where the heat time of the current column and the heat time of the subsequent column overlap.

  By paying attention to such a difference of the registration adjustment value Rt, it is also possible to handle the correction data of the dynamic registration adjustment value.

  The handling of the correction data will be described with reference to FIG.

  In the figure, the horizontal direction represents the carriage movement direction in the forward path, the left side is the printing start side, the vertical direction represents the registration displacement amount, the upward direction is the direction in which the printing position is delayed, and the vertical axis is within one pixel. Scaling with register adjustment value. The curve represents the characteristic of correcting the registration deviation amount by the dynamic registration adjustment obtained from the error of the linear scale 14 or the like. The points obtained by sampling this curve in the cycle N in which dynamic registration adjustment is performed (specific numerical values are not shown here) are indicated by black dots. The lower dotted line represents a registration adjustment value Rp (a specific numerical value is not shown here) in units of one pixel, and the middle dotted line is the smallest registration adjustment value Rmin in the curve, and the upper dotted line Is the largest registration adjustment value Rmax in the curve, and is represented here by Rmin = 4 and Rmax = 8. Here, the registration adjustment range within one pixel will be described with the same block division number Nb = 16 as described above.

  The dynamic registration adjustment values Rn at the sampled black dots are R0 = 6, R1 = 6, R2 = 7, R3 to R6 = 8, R7 = 7, R8 = 6, R9 = 6, R10 = 5 from the left. R11 = 5, R12-R16 = 4, R17 = 5, R18 = 5,. . . . Therefore, the difference ΔRt between the registration adjustment values of the current column and the previous column is 0, 1, 1, 0, 0, 0, −1, −1, 0, −1, 0, −1. , 0, 0, 0, 0, 1, 1,. . . It becomes. When static registration adjustment is performed with the registration adjustment value for each pixel being Rp and the registration adjustment value within one pixel being R0 = 6, the current column and the previous column are used as correction amounts for performing the dynamic registration adjustment value. The difference ΔRt between the registration adjustment values in the column can be used. The registration adjustment value difference ΔRt here satisfies the relationship of −1, 0, 1 and “registration adjustment Rt difference ΔRt ≦ 1 column cycle−1 column heat time”. The registration adjustment value difference ΔRt is stored, ΔRt is read in accordance with the registration adjustment value change cycle N, and ΔRt is added to the registration adjustment value Rt within one pixel of the previous column, and the next column A registration adjustment value Rt within one pixel is obtained and sequentially sent to the registration adjustment unit 32 within one pixel to perform dynamic registration adjustment.

  In particular, when the correction amount is defined so that the difference ΔRt between the registration adjustment values is limited to −1, 0, 1, and 3, the registration adjustment value Rt is not obtained, and the value of the difference ΔRt is −1, 0. In response to 1, the TRNS_WIN signal of the next column can be advanced by one block, not changed, or delayed by one block.

  As shown in FIG. 1, the description has been given so far in the case where the fluctuation range Rmin to Rmax of the registration adjustment value for performing dynamic registration adjustment is 4 to 8 and within the registration adjustment range 0 to 15 within one pixel. However, it is not necessary to limit to this, for example, when Rmin = 13 and Rmax = 17 as shown in FIG. 12 and the next column or when Rmax = 2 and Rmin = −2 as shown in FIG. It is also possible to cope with the case where it extends over, or the case where the amount of shift is one pixel or more and spans a plurality of columns, as shown in FIG.

  The correspondence to the above three examples in FIGS. 12 to 14 will be described below.

  First, in the case of FIG. 12, since the values of the dynamic registration adjustment are all 0 or more, the processing can be performed in the same manner as described above, but when the dynamic registration adjustment value changes from 15 to 16, Special processing is required when changing to.

  The case where the dynamic registration adjustment value changes from 15 to 16 will be described with reference to FIG. 15. When the dynamic registration adjustment value is 15, when the dynamic registration adjustment value is 15, the registration adjustment value in one pixel Rt = 15. When the rising edge of the HT_TRG signal is 1) the TRNS_WIN signal is generated with the counter value REZI_CNT = 15 of the base counter (not shown), the column is printed, and when the dynamic registration adjustment becomes 16, the HT_TRG One rising edge 2) of the signal is ignored and one is ignored (the rising edge of the HT_TRG signal is delayed by one) is stored in the sequence control unit 12, and the registration adjustment value Rt within one pixel is stored. Is updated to 0 (= dynamic registration adjustment value 16−block division number 16), and TRNS_WI is set when REZI_CNT = 0 with respect to the rising edge 3 of the next HT_TRG signal. And it generates a signal to perform the printing of the column. This corresponds to the absence of generation of the TRNS_WIN signal in the HT_TRG cycle of 2). After this, the start of the column is that the rising edge of the HT_TRG signal is delayed by one. Thereafter, while the dynamic registration adjustment value exceeds 16, the dynamic registration adjustment value minus the block division number 16 The recalculated value is used as the updated value as the one-pixel registration adjustment value Rt.

  Next, the time when the dynamic registration adjustment value changes from 16 to 15 will be described with reference to FIG. 16. When the dynamic adjustment value is 16, a place where the rising edge 1) of the HT_TRG signal is originally used as a reference is 1 Before starting operation with the registration adjustment value Rt = 0 in one pixel at the rising edge 2), it is determined whether or not a value of 15 or less comes to the dynamic registration adjustment value in the next update of Rt. In the case of coming, in order to generate another TRNS_WIN signal at the rising edge 2), a comparator (not shown) for comparing with REZI_CNT and a counter (not shown) for counting blocks and making BLK_CNT are compared with BLK_CNT. One more series of comparators (not shown) are provided to operate at the rising edge 2) at the registration adjustment value Rt = 15 in one pixel, and at the rising edge 2) the registration adjustment value Rt in one pixel. = 0 By simultaneously processing the operation, the two TRNS_WIN signals can be generated within one cycle of the HT_TRG signal, and two columns can be printed. When the inside of the sequence control unit 12 storing that the rising edge of the HT_TRG signal is delayed is cleared, and thereafter the rising edge 3) is used, and the dynamic registration adjustment value is 15 or less , Can work with the same process until now.

  In this way, by delaying or advancing the TH_TRG signal cycle by one, the one-pixel registration adjustment value Rt rises from 15 to 16 and becomes 0, or it falls from 16 to 15 digits. doing.

  Next, the case of FIG. 13 (Rmin = −2, Rmax = 2) will be described. The registration adjustment value Rp for each pixel is set to one less so as to advance forward in time, and the amount of advancement is determined. It is possible to cope with this by adding 16 to the dynamic registration adjustment value and canceling it by delaying it, and assuming that Rmin = 14 and Rmax = 18 and performing the same processing as in FIG. Also at this time, it is necessary to cope with the fact that the registration adjustment value Rt within one pixel rises from 15 to 16 and becomes 0, or falls from 16 to 15 digits.

  Next, the case of FIG. 14 (Rmin = −2, Rmax = 17) will be described. Similarly to FIG. 13, the registration adjustment value Rp for each pixel is set to one less so as to advance in time. The advanced amount is offset by adding 16 to the dynamic registration adjustment value and delayed in time, and it is possible to cope with this by performing the same processing as in FIG. 12 assuming that Rmin = 14 and Rmax = 33. However, the carry occurs twice from 15 to 16, and 31 to 32, and the carry occurs twice from 32 to 31, and from 16 to 15. Therefore, in order to cope with this, the rising edge of the HT_TRG signal is sent to the sequence controller 12. This can be dealt with by counting how many parts are stored in which the edge is delayed by one.

  Correspondence between carry and carry of the registration adjustment value Rt within one pixel is not limited to the above description, and it is also possible to cope with a hardware configuration or the like in the registration adjustment unit 32 within one pixel. . For example, the counter (not shown) for generating the count value REZI_CNT can be counted up to a number larger than the block division number Nb = 16, and has a plurality (the number allowed for carry), so that the rising edge of each HT_TRG This can be dealt with by starting counting with the standard. At this time, a comparator (not shown) with the counter value REZI_CNT, a counter for generating the count value BLK_CNT (not shown, the number of count stages may be the same as before), and a comparator with the counter value BLK_CNT (not shown). It is also necessary to have multiple (as many as the carry is allowed). The counter (not shown) that generates REZI_CNT needs to be cleared by the TRNS_WIN signal generated corresponding to the counter that generates BLK_CNT when REZI_CNT is generated, rather than high of the HT_TRG signal. In addition, it is necessary to perform processing for assigning the hardware of the series for which the generation of the TRNS signal is completed to the next rising edge of the HT_TRG signal.

  The dynamic registration adjustment amount as shown in FIG. 1 is obtained in advance up to the maximum print width for each nozzle row, so that the necessary portion can be determined from the position of the image to be printed, the print width, the nozzle row to be used, etc. By cutting out and using the dynamic registration adjustment amount, various print formats can be supported. Basically, the forward and return passes can be handled by using the dynamic registration adjustment amount from the left or the right, but the dynamic registration adjustment amount is determined independently for the forward and return passes. It is also possible to switch the use depending on the forward or backward path of printing.

  The method for obtaining the dynamic registration adjustment amount is not the main point of the present invention, and is not limited. Although details are omitted, the distortion characteristics of the linear scale and the mechanical design such as the print head, the linear scale, and the maximum print width are provided. A method of calculating from a value and a method of actually measuring and measuring a registration adjustment pattern with the maximum print width and the use print width are also conceivable.

  In this embodiment, it is possible to perform only static registration adjustment with or without using the above dynamic registration adjustment. Since the dynamic registration adjustment mode for determining whether or not to perform dynamic registration adjustment can be set by the sequence control unit 12, static registration adjustment as shown in FIG. 9 or dynamic registration adjustment as shown in FIG. 10 can be performed. Is possible. The dynamic registration adjustment mode is set from the operation unit 11, the method corresponding to the time of shipment, the printing width, the printing position, and the printing mode (standard mode, high-definition mode, etc.). It is possible to associate whether or not to do so, and it is also possible to adjust the dynamic registration adjustment amount.

The figure explaining the dynamic registration adjustment value which concerns on embodiment of this invention Diagram for explaining nozzle row of recording head Diagram for explaining partial misregistration 1 is a block diagram of a main internal configuration of a printer according to an embodiment of the present invention. The block diagram of the internal structure of the printing position control part 13 which is the main structures of the printer which concerns on embodiment of this invention. Timing chart of signals in the reference signal generation unit in the print position configuration unit 13 which is the main configuration of the printer according to the embodiment of the present invention. FIG. 3 is a timing chart for explaining details of printing timing for one nozzle row in the printer according to the embodiment of the present invention. In the printer according to the embodiment of the present invention, a timing diagram (before registration adjustment) for explaining the relationship of printing timing with respect to each nozzle row of the recording heads 8 and 9 FIG. 6 is a timing chart for explaining details of printing timing for one nozzle row when static registration adjustment is performed in the printer according to the embodiment of the present invention. FIG. 4 is a timing chart for explaining details of print timing for one nozzle row when dynamic registration adjustment is performed in the printer according to the embodiment of the present invention. [A], [B], [C] FIGS. 7A and 7B are diagrams for explaining limitations on the difference between dynamic registration adjustment values in the preceding and succeeding columns in the printer according to the embodiment of the present invention. The figure explaining the case where there is a carry-over in the dynamic registration adjustment range in the printer according to the embodiment of the present invention. The figure explaining the case where there are a carry down and a negative numerical value in the dynamic registration adjustment range in the printer according to the embodiment of the present invention. FIG. 7 is a diagram for explaining a case where there are a plurality of carry-downs and negative values in the dynamic registration adjustment range in the printer according to the embodiment of the present invention. The figure for demonstrating the carry-over process in the dynamic registration adjustment range in the printer which concerns on embodiment of this invention. The figure for demonstrating the process of a carry-down in the dynamic registration adjustment range in the printer which concerns on embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Host computer 2 Printer 3 I / F part 4 Image data processing part 5 Image memory part 6 Image data transfer part 7 Carriage 8, 9 Recording head 10 Encoder sensor 11 Operation part 12 Sequence control part 13 Print position control part 14 Linear scale 15 Sub-scanning encoder 16 Carriage control unit 17 Carriage control unit 18 Carriage motor 19 Carriage motor 30 Reference signal generation unit 311 Registration adjustment unit 321 for each pixel Registration adjustment unit in a pixel

Claims (4)

In an image recording apparatus in which a recording head records an image on a recording medium while the recording unit having a plurality of nozzle rows in which a plurality of nozzles are arranged and the recording medium move relatively in the nozzle row arrangement direction,
Moving means for relatively moving the recording medium and the recording head in the nozzle row arrangement direction;
Image forming means for forming an image based on image data during the movement;
A recording position adjusting unit that adjusts an image recording position of each nozzle row of the recording unit in the image forming unit with respect to a moving direction;
Static position control control means for adjusting and controlling the position adjustment means so that the image recording positions of the nozzle rows overlap at a certain image position, and recording at the position adjustment means at a plurality of image positions during movement by the movement means. An image recording apparatus comprising dynamic position control means for varying the position by a correction amount within a preset range.   2. The image recording apparatus according to claim 1, wherein the certain image position is a print start position or a print end position during the relative movement.   The plurality of moving image positions are cycles set in advance, and in particular, an integer multiple of a cycle of a synchronization signal (HT_TRG) serving as a reference for printing timing of each column. Image recording device.   The correction amount in the preset range is smaller than the difference between the cycle of the synchronization signal (HT_TRG) serving as a reference for the printing timing of each column and the printing time of one column (HIGH width of HT_WIN). The image recording apparatus according to claim 1, wherein:

Images