JPH03146367A - Image formation device - Google Patents

Abstract

PURPOSE:To raise precision of correction by varying a pulse width with respect to density data in a direction wherein correction is performed and by narrowing a correction width of the density data by a method wherein the density data to be inputted from a host computer is integrated for a specified integration period, and the integrated value is compared with a threshold. CONSTITUTION:Density data is received from a host computer. Since a threshold TH is constant, a pulse width is not proportional to an input data. Therefore, the density data is corrected with a correction part 1 so that the pulse width varies linearly to the density data. A D/A converter 2 converts this corrected data to an analog voltage IN. An integrator 3 integrates this inputs voltage IN for a specified integration period. Integration voltage INTG is compared with the threshold TH fixed with a comparator 4 within this integration period and consequently, an LDON signal indicating light emission time of a semiconductor laser is sent to a driver 5 of the semiconductor laser. The pulse width of the LDON signal is modulated according to the density data, and a printing area of one dot is varied according to the density data.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an image forming apparatus that prints images such as halftone images in response to image data sent from a host such as an image reader.

(Prior Art and Problems to be Solved by the Invention) When printing a halftone image with a printer, density data (gradation signals) of each pixel is sent from the host to the printer. The printer generates a print signal in response to this density data.

Conventionally, the print signal has been controlled by a dither method or the like to determine whether or not to print dot by dot. Therefore, the size of the l-dots was the same.

However, in recent years, pulse width modulation has come to be used to express halftones. In this method, the density of one dot is represented by, for example, an 8-bit gradation signal, and the pulse width of the print signal is modulated in accordance with the density to change the printing area of 1 dot, producing a halftone image (256 gradations). Reproduce. Therefore, halftone images can be printed without deteriorating resolution. For example, in a laser printer, a semiconductor laser emits light during an output period of this pulse width to form a dot on a photoreceptor with a size corresponding to the pulse width.

To generate a print signal using this pulse width modulation method, for example, there is a method in which density data is compared with a fixed pattern signal (for example, a silver wave) having a fixed amplitude at a fixed period. In this harmonic method, a silver wave having a constant amplitude and period is input to one input terminal of a comparator.

One cycle of this silver wave corresponds to the output cycle of density data. A gradation signal (e.g., 8 bits) input from the host is converted into an analog voltage during one cycle of the silver wave, and is input as density data to the + input terminal of the comparator. Therefore, a pulse signal for printing is output during a period in which the density data is larger than the silver wave within one cycle. This pulse signal is
The dot is output in a period proportional to the density data, and the printing area of one dot is modulated in accordance with the pulse width.

Further, in the triangular wave method, which is a modification of the harmonic method, in order to improve gradation, a triangular wave with a constant amplitude and a constant period is used instead of a silver wave, and one pulse is output at a two-dot period.

This harmonic method is simple and has a simple circuit configuration. However, when it is desired to output pulses with a dot period that is a non-integer multiple, there are problems such as changing the output period of the silver wave.

On the other hand, the present invention uses a density data integration method, which is a novel pulse width modulation method. This density data integration method has various advantages over the harmonic method, such as the ease of outputting pulses with a dot period that is a non-integer multiple, and the ease of performing γ correction of density data at the same time.

An object of the present invention is to provide an image forming apparatus equipped with a novel pulse width modulation method.

(Means for Solving the Problems) A first image forming apparatus according to the present invention generates an image recording signal having a pulse width that is modulated in accordance with density data, and generates an image recording signal having a pulse width that is modulated in accordance with density data. An image forming apparatus that prints dots of a certain size includes an integrating means for integrating density data every predetermined integration period, and an integral value of the integrating means after the start of integration within the predetermined integration period. The present invention is characterized by comprising a comparison means for comparing with a predetermined threshold value and generating an image recording signal having a pulse width determined according to the comparison result.

The second image forming apparatus according to the present invention further includes timing generating means for generating the timing of integration in the integrating means so that the predetermined period of the integrating means is a period different from the output cycle of the density data. It is characterized by

The third image forming apparatus according to the present invention further includes a selection means for selecting an integration period of the integration means, and a timing generation means for generating an integration timing in the integration means in accordance with the integration period selected by the selection means. It is characterized by having the following.

(Operation) The concentration data input from the host is integrated over a predetermined integration period by the integrating means, and this integrated value is compared with a predetermined fixed threshold value by the comparing means. Therefore, the pulse width of the output signal of the comparing means is modulated in accordance with the density data.

In this case, in the timing generating means, the integration period can be made different from the output period of the concentration data.

Further, the timing generating means can switch the integration period of the integrating means in accordance with the selection by the selecting means.

(Embodiments) Hereinafter, embodiments of the present invention will be described in the following order with reference to the accompanying drawings.

(a) Light emission time control using density data integration method <a-1> Configuration of density data integration method <a-2> Integration and correction of density data <a-3> High-speed two-system method <a-4> Integration time Switching and density data addition <a-5
> Stabilization of gradation at low or high density (b) Overall configuration (c) High-speed two-system configuration (d) Addition of density data of multiple pixels (e) Correction of density data (f) Sub-scanning direction Gain switching (g) Integrating section and comparator section The present invention is particularly relevant to section (a).

Margin below (a) Light emission time control using density data integration method <a-1> Configuration of density data integration method In the laser printer according to this embodiment, density data (gradation signal ) A semiconductor laser is emitted to form a latent image on the photoreceptor, and printing is performed using an electrophotographic process. The emission time of the semiconductor laser is changed for each l dot in accordance with the concentration data.

In this embodiment, the light emission time is controlled by the concentration data integration method described below.

FIG. 1 schematically shows the basic configuration of semiconductor laser light emission control in this embodiment. Here, density data (8 bits) is received from the host. Next, the concentration data is
Correction is performed in the correction unit l in order to perform correction as will be explained later. The D/A converter 2 converts this corrected data into an analog voltage IN, and the integrator 3 integrates this input voltage IN in a predetermined integration period. The integrated voltage INTG is compared with a fixed threshold value TH by a comparator 4 within this integration period, and as a result, an LDON signal indicating the emission time of the semiconductor laser is sent to the semiconductor laser driver 5. In this way, the semiconductor laser emits light during the output period of the LDON signal. That is, the pulse width of this LDON signal is modulated in accordance with the density data, and the printing area of one dot is changed in accordance with the density data.

Now, to simplify the explanation, the concentration data received from the host is directly sent to the integrator 3 without being corrected by the correction unit 1.
Assume that the input voltage is IN. An example of the integral waveform in this case is shown in FIG. 2(B).

Integrator 3 is reset before starting integration.

Since the input voltage IN is constant within the integration period (while the INTGT signal is being output), the integration voltage INTG increases linearly with time. This integrated voltage INTG is compared with a threshold value TH by a comparator 4. In this case, the connections to both input terminals of comparator 4 are opposite to the example in FIG.
The output voltage INTG of the integrator 3 is connected to the input terminal. When the integrated voltage INTG reaches the threshold value TH, the comparator 4 raises the LDON signal. This LDON signal is generated during the period in which the integrated voltage is greater than the threshold TH, ie, until the end of the integration period. Therefore, the semiconductor laser emits light during a period corresponding to the concentration data within the integration period.

In this way, the pulse width of the LDON signal indicating the light emission period is modulated according to the concentration data. Note that the integrator 3 is reset after the integration period. Actual input voltage I to integrator 3
N is not the density data itself, but a value that has been appropriately corrected by the correction unit l.For details of the correction, see (a-2
> section.

When emphasis is placed on gradations at low densities, it is preferable to steepen the slope of the integrated voltage INTG at low densities, as will be explained in detail in section (a-5>). For this purpose, assume that the correction unit 1 takes the conglomerate of the concentration data received from the host, and that value is set as the input voltage IN to the integrator 3 (taking the conglomerate is also considered a type of correction). ).Therefore, when the concentration data is small, the input voltage is large, and when the concentration data is large, the input voltage is small.An example of the integral waveform in this case is shown in the second example.
This is shown in Figure (A). Integrator 3 is reset before starting integration. Unlike the previous example, the polarity of the connection between the input voltage to the comparator 4 and the threshold value is as shown in FIG. Comparator 4 outputs the LDO at the same time as the integration starts.
Raise the N signal.

On the other hand, since the input voltage IN is constant within the integration period,
The integrated voltage INTG increases linearly with time. This integrated voltage INTG is set to a threshold value TH by a comparator 4.
When the integrated voltage INTG reaches the threshold value TH, the comparator 4 stops generating the LDON signal. The lower the concentration, the steeper the slope of the integrated voltage and the shorter the pulse width. Thereafter, the integrated voltage INTG increases until the end of the integration period, so that no LDON signal is generated. Therefore, similar to the previous example, the pulse width of the LDON signal is modulated in response to the concentration data.

Note that even in this case, the actual input voltage IN to the integrator 3 is a value appropriately corrected by the correction section 1.

Although analog circuits are used for the integrator 3 and comparator 4, digital circuits may also be used.

(a-2> Integration of density data and correction pulse width (semiconductor laser light emitting time) It is desirable that the NvR of the image printed by light emission is proportional to the density data as shown in straight line B in Figure 3. .

However, in general, if no correction is made in the correction section l,
When the concentration data or a combination thereof is input to the integrator 3, the pulse width changes non-linearly in an upwardly convex manner as shown by curve A with respect to the concentration data. This is because in the concentration data integration method, the integrated waveform is compared with the threshold value TH and the LDO
This is because although the slope of the integral waveform itself is proportional to the input data when the N signal is generated, the threshold value TH is constant, so the pulse width is not proportional to the input data.

Therefore, in the correction section l, as shown by curve C in FIG.
The density data is corrected so that the pulse width changes linearly with respect to the density data. Note that when the threshold value TH approaches O, the curve A in FIG. 3 approaches the straight line Bj.

In reality, it is also necessary to correct the γ characteristics of the printer. Therefore, the correction section 1 also performs γ correction at the same time. Although the gamma characteristic of a printer generally changes non-linearly as shown by the curve in FIG. 5, it is desirable to correct it so that it changes linearly with respect to the gradation density data as shown by curve E.

Therefore, the density data may be corrected by the correction section 1 as shown by the curve F in FIG. 6.

Therefore, as shown by curve G in FIG. 7, it is sufficient to perform the correction in a combination of FIG. 4 and FIG. 6 at once.

In this example, the emphasis is on gradation stability at low density (
(See section <a-5>), the input voltage IN to the integrator 3 is generated using correction data that has been corrected for congruence of the concentration data. Therefore, the correction section 1 performs correction so that the input voltage IN becomes large at low concentrations, as shown by curve G in FIG.

Note that the correction takes into consideration the influence of the threshold value of the comparator unit 30, which will be described later. The correction is performed using a table in the ROM of the correction unit l (see section (e)).

The concentration data corrected by the correction section l is then converted into an analog voltage IN by the D/A converter 2 and supplied to the integrator 3. The integrator 3 performs integration as shown in the example of FIG.

That is, at the beginning of each integration period, the integration capacitor is discharged and the integration voltage returns to zero. The integration period begins, and the integration of the input voltage IN is started, and at the same time, the signal LDON indicating the emission period of the semiconductor laser becomes H level, and the emission starts. Since the integration constant of the integrator 3 is set large, the integrated voltage increases linearly in proportion to the integration time. The integrated voltage is set to the threshold value TH by comparator 4.
compared to When the integrated voltage becomes larger than the threshold value TH, the signal LDON goes to L level and the semiconductor laser stops emitting light.

After the integration period has passed, the integrator 3 is reset and the integrated voltage returns to O. The shorter the integration period, that is, the higher the speed, the more the effect of reset cannot be ignored, so in the example of FIG. 8, a reset period is provided between the integration periods corresponding to each dot, which will be described in detail later using FIG. 34. *i is set so that the LDON signal is not output during the reset period.

It is also possible to cancel the effect of the reset by incorporating the effect of the reset on the signal LDON into the correction.

Furthermore, the slope of the integral waveform of the integrator 3 becomes gentler as the input voltage becomes smaller (that is, as the concentration data becomes larger). If the integrated voltage at the end of the integration period does not reach the threshold TH, the emission period will be the same as the integration period;
It stops changing. Therefore, the correction data at the maximum concentration (density 255) is set so that the integrated voltage reaches the threshold value TH at the end of the integration period, as shown in FIG.

Furthermore, when stability at high density is important, for the same reason, the correction data at the minimum density (density O) is set to a value other than 0, as shown in FIG.

As can be seen from FIGS. 3 and 5, at high concentrations, the deviation from linearity due to the concentration data integration method tends to compensate for the deviation from linearity due to the γ characteristic. Therefore, the degree of correction can be reduced and the accuracy of correction can be increased.

It is also possible to do without correction.

(a-3> High-speed two-system system In the example shown in Figure 8, the integration period and reset period are provided alternately. However, when high-speed processing is required, the specific steps shown in Figure 18 will be used later.) In addition, another system of integrator and comparator may be provided in parallel, and integration and reset may be performed alternately in both systems.

FIG. 9 shows an example of the integral waveform in this case. In this figure, the solid line and the dashed line of the integral waveform indicate the waveforms of integrators of different systems, respectively. According to the configuration shown in FIG. 8, there is no effect of the output voltage at the time of resetting the integrator, and the time loss caused by providing a reset period is also eliminated, so that high speed operation can be achieved.

<a-4> Switching of integration period and addition of density data High resolution is required for character originals, and high gradation is required for photographic originals. However, it is difficult to achieve both high resolution and high gradation. If the density data from the host is printed as is, the resolution of the original density data can be maintained. However, in photographs where gradation is important, gradation is insufficient.

On the other hand, if continuous density data is printed as one unit, the dynamic range increases and therefore the gradation can be increased, but in this case the resolution decreases.

The pulse width modulation method of this embodiment generates a pulse by comparing a waveform obtained by integrating density data with a predetermined threshold value. Therefore, the threshold is fixed. Now, if the integration period is lengthened while keeping the concentration data output clock the same, a plurality of concentration data will be integrated within the same integration period, and a pulse can be generated using the plurality of concentration data as one unit. For example, FIG. 1A schematically shows an example of pulse width modulation by integration with an integration period that is twice the basic integration period. In this case, the threshold value TH is also doubled. Furthermore, it is also possible to make the integration period a non-integer multiple of the basic period. For example, Figure 10 (B)
indicates a 2.5 times integration period. Furthermore, by configuring the integrator-comparator system so that the integration period can be switched between the basic period and a longer period, the operator can select between documents for which high resolution is to be prioritized and documents for which high gradation is to be prioritized. The integration period can be selected accordingly. For example, for reasons to be explained later, when priority is given to resolution, the integration period is set to be 1.5 times the basic period, and when priority is given to gradation, the integration period is set to 2.5 times the basic period. For selection, a selection key may be provided on the operation panel, for example.

In response to the operator's selection, the generation timing of the INTGT signal indicating the integration period in one integrator-comparator system is switched, and the integration period of the integrator and the threshold voltage of the comparator are simultaneously switched. Alternatively, two integrator-comparator systems having different integration periods may be provided in parallel so that one can be selected.

In reality, if gradation is important, within the integration period,
If the concentration data is simply integrated one by one in units of basic periods, only the first concentration data contributes at the beginning of the integration, so the weight of the concentration data integrated first becomes larger. In order to emphasize gradation, it is better to make all related density data contribute to the integral waveform on average. Therefore,
In this embodiment, in addition to simply switching the integration period,
All related concentration data are added in advance, and the added value is used as the input voltage to the integrator 3. Figure 11 (A).

(B) shows an example in which the added value of density data is integrated as an input voltage in the cases of FIGS. 10 (A) and (B).

Note that if the average value is calculated and used as the integral input voltage instead of the added value itself, there is no need to switch the threshold voltage. However, in this case, the gradation does not increase.

Also, the stability of the electrophotographic process must be considered. In this embodiment, one dot time is a basic period of 56 ns, and it is difficult to obtain stable gradation in the electrophotographic process by pulse width modulation within this one dot time. Therefore, even when giving priority to high resolution suitable for character manuscripts, it is better to take an integration period longer than the basic period. Therefore, in order to ensure a certain degree of stability of the electrophotographic process and a certain degree of resolution, an integration period that is 1.5 times the basic period is adopted in the standard mode. Then, 1.5 pieces of concentration data are added to form the input voltage to the integrator 3.

On the other hand, in a photo mode that prioritizes high gradation, an integration period that is 2.5 times the fundamental period is used. Then, the 2.5 times the concentration data is added and used as the input voltage to the integrator 3. Note that the reason why a non-integral multiple period is used in the photo mode is to prevent false lines from occurring. Details will be explained in section (d). Since you can switch between standard mode and photo mode, you can select the gradation and resolution for each individual document.

<a-5> Method for stabilizing gradation at low density or high density, stabilization of gradation at low density will be explained. Figure 12 (A
), (B), there is actually some noise on both the integral waveform and the threshold value TH. Therefore, the intersection of the two causes variations.

As is clear from the figure, the steeper the slope of the integral waveform, the shorter the crossover time and the smaller the variation S.

Therefore, when emphasis is placed on gradation stability at low concentrations as in this embodiment, for example, as shown in FIG. 7, the integrator 3
It is sufficient to increase the input voltage IN to make the integral waveform steep. Therefore, the input voltage IN is increased when the concentration is low, and is decreased when the concentration is high. Correspondingly, L
The output period of the DON signal must become longer as the concentration data increases, so the LDON signal rises when the integrator output falls below the threshold, and falls when the integrator output falls above the threshold. (See Figure 2 (A)).

On the other hand, if gradation stability at high density is important, the first
As shown in FIG. 3, the concentration data may be corrected so that the input voltage is small at low concentrations and the input voltage IN is increased at high concentrations. Then, the light emission period starts at the intersection of the integral waveform and the threshold value, and ends at the end of the integral period (see Fig. 2).
See B).

One advantage of this method is that by setting the threshold voltage appropriately, the influence of noise in the background part of the image can be removed.

In this method of increasing the input voltage IN at high concentrations, the influence of the reset waveform when resetting the integrator is particularly large at low concentrations, and the pulse width does not faithfully reproduce the concentration data. Therefore, it is better to remove the influence of reset from the fall of the LDON signal.

FIG. 14 shows an example of an integral waveform and an LDON signal in the case of a high-speed two-system type in which the input voltage IN is increased at high concentration.

It is preferable that the amplitude of the integral waveform of the integrator be as small as possible from the standpoint of removing unnecessary noise. With this integration method, it is possible to secure the amplitude in important concentration portions (low concentration or high concentration), and the amplitude is low in other areas, making it possible to lower the average level of noise.

(b) Overall configuration FIG. 15 is a block diagram showing the overall configuration of a laser printer according to an embodiment of the present invention. Density data and print control data from a host such as an image reader are received by the interface control section II. In addition, the operation panel 12
The interface control unit 1 receives various instructions from the operator.
1 and displays the printer status. Based on these instructions and data, the interface control section 11 sends density data to the print head control section 13 and control signals to the electrophotographic control section 14. In response to this,
The print head control section 13 rotates the polygon motor of the print head section 15 in accordance with the density data, causes the semiconductor laser to emit light, and forms an electrostatic latent image on the photoreceptor of the electrophotographic processing section 16 . On the other hand, the electrophotographic control section 14 controls the electrophotographic processing section 16 in response to this image writing to print an image on plain paper using an electrophotographic process.

FIG. 16 shows details of the operation panel 12.

Here, 21 to 24 are input keys, and 30 to 40 are display elements. Key 21 is a PAUSE key for temporarily stopping the printing operation. The key 23 is a shift key, and when pressed at the same time as the key 22, it becomes a CANCEL key to interrupt printing. Key 24 is a key for selecting standard mode and photo mode. In addition, the display element 39 is an LED that lights up in the photo mode.
FIG. 7 shows a block diagram showing the configuration of the print head control section 13. As shown in FIG. The interface section 51 receives an image signal (V
IDEO), a synchronization signal (SYNCK), and an image area signal (VD) indicating an image area in the sub-scanning direction are received and sent to the linebacker memory 52. On the other hand, a horizontal synchronization signal (H3YNC) is sent to the interface section 51.

The line pack memory 52 uses a clock (approximately 13M) of the image data (VIDEO) received from the image reader.
Hz) and the printing clock on the printer side (approximately 17MHz)
) This is a buffer memory for absorbing the difference in data, and uses a commercially available first-in, first-seven-first-out (FIFO) memory.

The adding unit 53 is a block that adds 1.5 dots or 2.5 dots of image data (VIDEO) from the line buffer memory 52 to improve gradation.

The correction unit 54 uses the image data (A
VIDEO), this block performs correction and γ correction for the integration in the integrating section 59, which will be described later, and performs the correction using a lookup table in the ROM.

The D/A converter 55 converts the digital image data (IDATA) corrected by the corrector 54 into an analog voltage (DAOUT).
).

The sub-scanning gain switching unit 56 is a block that switches the gain in the sub-scanning direction every l line in order to improve the gradation up to high density. ) is output as an analog voltage (AMP) by switching the gain for each line.

The sample and hold unit 57 is a block that samples and holds this analog voltage (AMP) and outputs an analog voltage (SHO) in order to stabilize the input voltage to the integration unit 59.

The gain offset adjustment section 58 is a block that adjusts the gain and offset of the entire system with respect to the input voltage (SHO) and outputs the result at a stage before the integration section 59.

The integrating section 59 is a block that integrates the input voltage (IN) from the gain offset adjusting section 58.

The input value unit 60 calculates the output value of the integration unit 59 (■NTG
) is compared with a threshold voltage to obtain a pulse signal (CMP) for activating the semiconductor laser 62 of the print head.

The LD driver section 61 is a block that drives the semiconductor laser 62 of the print head section 15 in response to a pulse signal CMP (LDONM) from the comparator section 60. Note that a circuit for automatically controlling the output of the semiconductor laser 62 is also provided, but its explanation will be omitted. LD driver section 6
1 also includes a light emitting area control unit that outputs a permission signal for light emission in the main scanning direction and the sub scanning direction, and
Although a forced light emission control unit is provided that forces the semiconductor laser 62 to emit light in order to cause light to enter the OS sensor (not shown), detailed description thereof will be omitted.

Note that the SO3/H5YNC unit 63 receives synchronization signals SO8 and H5 from the detection signal 5ENSOR from the SOS sensor.
This is a block that generates YNC.

Further, the clock generator 864 is a block that generates various timing signals to each block from the signal generated by the oscillator 65, the SO5 signal, and the PH0T○ signal. As will be explained later, various timing signals are supplied at different timings in the standard mode and the photo mode. That is, in response to the PH0TO signal set by the key 24, timings such as integration are generated in the standard mode and photo mode, as shown in FIGS. 21, 22, 31, and 32. Here, the clock generator 64 generates DREQ for data reading based on the clock signal of the oscillator 65.
A signal (such as that shown in FIG. 21) is generated, and various timing signals are generated in response to the DREQ signal using a logic circuit using a known hard logic element.

(c) High-speed two-system type configuration When generating the semiconductor laser emission time using the concentration data integration method, integration periods and reset periods must be provided alternately in the integration section 59 in order to eliminate the influence of reset.

Therefore, as explained in section <a-3>, speeding up can be achieved by providing two integrator-comparator systems. That is, it is sufficient to perform integration alternately in each system.

FIG. 18 shows a two-system part when part of the overall configuration of FIG. 17 is made into two systems (various signals of both systems are 1.
2 to distinguish them. ). That is, after the sub-scanning gain switching section 56, a sample hold section 57, a gain offset adjustment section 58, an integration section 59, and a comparator section 60 are installed.
A sample hold section 71, a gain offset adjustment section 72, an integration section 73, and a comparator section 74 are provided in parallel to the . Although the two systems differ in the timing of integration (see FIGS. 31 and 32), they operate in the same way in other respects.

The output signals of both humbrate sections 60 and 74 are sent to the semiconductor laser driver 61 via an OR gate 75. This enables high-speed operation as shown in FIGS. 9 and 14.

(d) Addition of density data of multiple pixels In this embodiment, the density data of 400 DPI stored in the line buffer memory 52 is sequentially read out using a clock (DREQ) with a period of 56 ns. The data stored in the line buffer 7 memory 52 is received from the host and is based on the image readout unit (for example, a linear CCD).
This is density data for each pixel of the sensor. In printing, if an image is written on the photoreceptor in units of this received data, the resolution will not deteriorate. However, it is difficult to obtain stable gradations in the electrophotographic process by controlling the light emission within the above-mentioned light emission period within one clock period (56 ns).

To solve this problem, if we add the density data of two consecutive pixels and print in units of 2 clocks, the gradation will be stable, but the resolution will be half that of printing in units of 1 clock. .

Therefore, in order to ensure both gradation and resolution to some extent, printing may be performed at a cycle of an appropriate length between one clock cycle and two clock cycles. Therefore, in this embodiment, printing is performed in units of 1.5 clocks in the standard mode. In this case, the resolution is approximately 10 lines/opening, which can be secured to some extent.

In the standard mode, printing density data must be created in units of 1.5 clocks from the density data read from the line buffer memory 52. Therefore, for three consecutive pieces of density data, the middle density data is halved and added to the first and last density data. And thus 3
The two summed data obtained from the density data of m are 1.5
It outputs at the cycle of the clock and prints a dot corresponding to its size. Due to this addition, the gradation appears to be 1
．． It becomes five times. This addition process is repeated in units of l every three clock cycles.

There is also a mode suitable for expressing midtones (called photo mode).
is provided to enable stable gradation to be expressed in the electrophotographic process. In this case, it is conceivable to add a plurality of pieces of density data and print at a clock cycle that is multiple times that number (for example, three clock cycles). However, if a halftone image is expressed with a 3-bit period, a vertical line with a 3-bit period will appear. To avoid this line, for example, it is necessary to shift the screen by one dot in the sub-scanning direction to create a screen angle of 45''.

To solve this problem, if the period is not an integral multiple, for example, 2.5 clock periods (resolution is 6.4 lines/screen), the vertical lines will no longer be visible.

Therefore, in the photo mode, data processing is performed in units of 5 degree data. That is, the third density data among the five consecutive density data is halved and added to the sum of the two density data before and after that data. Then, two pieces of added data obtained from the five pieces of density data are outputted at a cycle of 2.5 clocks, and a dot corresponding to the size is printed.

The operator can press the key 24 on the operation panel 12 to select the two addition modes, standard mode and photo mode, or select the number of data to be added. In response to this selection, a PH0TO signal is generated, and in response to this, the clock generating section 64 sends an ADDLA signal for selecting the number of data to be added to the adding section 53.

Next, addition of density data of a plurality of pixels will be specifically explained.

FIG. 19 shows the connection of the line buffer memory 52.

The line buffer memory 52 consists of a FIFO memory 101 and two latches 102 and 103, and is used to buffer the difference between the data transfer clock from the host and the printing clock on the printer side. Interface 51
The 8-bit image data (VIDEO) from is latched by the latch 102 and then written to the FIFO memory 101 at the timing of the signal 5YNCK from the host.

Further, data (VIDEO) read from the FIFO memory lot at the timing of the clock DREQ is also latched once by the latch 103 and then output. The successive clocks DREQ are supplied from the clock generator 64.

FIG. 20 is a block diagram of the adding section 53.

The adding unit 53 adds the image data (density data) sent from the line buffer memory 52 in two addition modes. In the standard mode, 1.5 pieces of data are added, and in the photo mode, 2.5 pieces of data are added. The mode is set by the PH0TO signal, and in standard mode P
H0TO- “H”, PH0TO- in photo mode
It is “L”.

The 8-bit density data VIDEO from the line buffer memory 52 is sent to the latch memories 111, 112, 113 as signals DATALA1, DATALA2.
It is latched at the timing of DATALA3. The output data VIDEO2 (8 bits) of the latch memory 112 and the output data VIDEO3 (7 bits) of the latch memory 113 are added by an adder 114. At this time, the least significant bit of the input data VIDEO of the latch memory 113 is not input to the adder 114, but is shifted to the lower side by one bit and is input, so that the data is 1/2 of the value latched in the latch memory 113. will be added. Adder 1
The addition data VIDEOAi (9 bits) of No. 14 is further added to the output data VIDEOI (8 bits) of the latch memory Ill in an adder 115, and the value after the addition is VIDEOAi (9 bits).
I DEOA2 (10 bits) is latch memory 11
6, it is latched at the timing of the signal ADDLA.

FIG. 21 shows a timing chart of addition in photo mode. The signal HIA is a signal indicating that the area is within the image area in the main scanning direction during scanning in the main scanning direction, and falls after a certain period of time after the SO5 signal is output by the SOS sensor, and corresponds to the image area after that. When the time has elapsed, it will stand up. When the signal HIA becomes "L" in the image area, the data reading clock DREQ from the line buffer memory 52 is supplied to the line buffer memory 52, and the density data VIDEO is output at its rising edge. Clock DR as shown in the figure
At the rising edges after the second rising edge of EQ, density data is output in order from address 0. In the figure, data at address n is indicated by ■. Each timing signal for latching this output data VIDEO is 5 as shown in the figure.
The clocks DREQ are generated at different timings with each cycle as one cycle. The timing signal DATALA1 for the latch memory Ill is outputted so as to be able to latch the 5m, 5m+4 (m is O or a positive integer) density data, and the timing signal DATALA2 for the latch memory 112 is outputted as 5m+1.5m+3. It is output to the latch memory 113 so that the th density data can be latched.
The timing signal DATALA3 is outputted so that the 5m+2nd density can be latched. Further 2.
A timing signal ADDLA for the latch memory 116 that latches the added value of 5 data is output every 2.5 clocks.

To explain the first cycle, the data at address 0 is latched into the latch memory 111 by the timing signal DATALA1, and then the data at address 1 is latched by the timing signal DATALA1.
The data at address 2 is latched into the latch memory 112 by TALA2, and then the data at address 2 is latched into the latch memory 113. As a result, the added value of the data at address 0, the data at address 1, and the data at (2 address)/2 is added to the adder l.
It is output from 15. Therefore, the added value (VIDEOA2) of this 2.5 data is latched in the latch memory 116 by the timing signal ADDLA, and the signal VIDEOA
is output as Next, the data at address 3 is latched by the latch memory 112 using the timing signal DATALA2, and then the data at address 4 is latched by the timing signal DATALA2.
It is latched into the latch memory 111 by TALA1.

As a result, the added value of (data at address 2)/2 data, data at address 3, and data at address 4 is added to the adder 115.
is output from. Therefore, the added value of 2.5 data is added to the latch memory 116 by the timing signal ADDLA.
is latched and output as signal VIDEOA.

Thereafter, similar addition is performed for every five clocks DREQ.

FIG. 22 shows a timing chart of standard mode addition. In the standard mode, since addition of 1.5 data is performed, the latch memory 111 is not used, but only the other two latch memories 112 and 113 are used. Therefore, the timing signal DATALAI for the latch memory 111 is always kept at "L" level. On the other hand, the timing signal DATALA2 for the latch memory 112 is 3m, 3m+2
The timing signal DATALA3 for the latch memory 113 is outputted so as to latch the (3m+1)th data (m is O or a positive integer), and l.

The timing signal ADDLA of the latch memory l16 which latches the added value of 5 data is 2.

Output every 5 clocks.

To explain about the first cycle, the data at address 0 is transferred to the latch memory 111 by the timing signal DATALA2.
The data at address 1 is then latched into the timing signal D.
It is latched into the latch memory 113 by ATALA3. As a result, the data at address 0 and (data at address 1) /
The added value of the two data is output from the adder 115. Therefore, the added value of this 1.5 data is the timing signal DA
The signal is latched into the latch memory 116 by TALA and output as the signal VIDEOA. Next, the data at address 2 is latched into the latch memory 112 at the timing of the timing signal DATALA2.

As a result, the sum of the data of (data at address 1)/2 and the data at address 2 is output from the adders 114 and 115. Therefore, the added value of 1.5 data is latched into the latch memory 116 by the timing signal ADDLA, and is output as the signal VIDEOA.

Thereafter, similar addition is performed for every three clocks DREQ.

(e) Correction of density data The addition unit 53 adds t; the density data is then corrected in the correction unit 54 in consideration of γ correction etc. as explained in section <a-2> (7th (see Figure 13).

The density data is corrected not only in the standard mode and the photo mode, but also in the AIDC mode (a mode in which the amount of toner is controlled so that the image density is constant). Note that the AIDC mode will not be explained below.

FIG. 23 is a block diagram of the correction section 54.

Concentration data V I DEOA output from the adding section 53
(10 bits), the ROMs 131 and 132 are accessed. The two ROMs 131 and 132 each store the lower 8 bits and upper 2 bits of correction data. Therefore, the iodide and concentration data IVIDE shown in FIG.
OA to address terminals A0 to A of both ROMs 131 and 132
, and also input the PH0TO signal to the address terminal A1° to select the mode, and input the AIDC signal to the address terminals A and H to select the AIDC mode.
Input the signal. That is, in ROM131.132, as shown in the address map of FIG.
Correction data for the photo mode is stored at addresses 400H to 7FFH, correction data for the standard mode is stored at addresses 800H to FFFH, and correction data for the AIDC mode is stored at addresses 800H to FFFH. The output data IDATA of the ROMs 131 and 132 are latched in the D-FFs 133 and 134 using the timing signal ROMLA, respectively.
It is output to the D/A converter 55. Here, the latch timing signal ROMLA is output every time data IDATA is output from the ROMs 131 and 132. The timing of the output is shown in FIGS. 21 and 22.

(f) Switching the gain in the sub-scanning direction In this embodiment, in order to obtain faithful gradations up to high gradation levels,
The gain is switched for each line in the sub-scanning direction.

FIG. 26 schematically shows an example in which three points of low density, medium density, and high density are printed over three lines. The gain of the center line is smaller than that of the upper and lower lines, so that it does not overlap with the points of the upper and lower lines even at high concentrations. Therefore, as shown by the curved line in FIG. 27, faithful gradations can be obtained up to high gradations. On the other hand, if each line has the same gain as in the past, overlap will occur between the lines at high densities, and there will be little change in gradation at high densities.
The gradation is not expressed faithfully (for example, see M in FIG. 27).

FIG. 28 shows a circuit diagram of the sub-scanning gain switching section 56.

The output voltage DOUT from the D/A converter 55 is a voltage 7 lower voltage 151 for minimizing the output current of the D/A converter 55 and widening the dynamic range of D/A conversion.
is output as a signal vl. In order to accommodate the difference in the processing voltage of the sample and hold unit 57 in the next stage, this signal Vl is generated using op-amps 152 and 153;
and 80%) and output to the analog switch 154. This switch 154 is connected to the HIA for each line.
It is switched by the EXC signal, and the output voltage of each differential amplifier is alternately output as the signal AMP. Using the circuit constants shown in FIG. 28, the output of this differential amplifier is as follows for the OP amplifier 152, and the gain can be adjusted by the resistance value. Note that the line switching signal HIAEXC can be generated by switching the output of, for example, a monostable multivibrator with the ADDLA signal (see FIGS. 21 and 22).

(g) Integrating section and comparator section FIG. 29 shows a circuit diagram of the integrating section 59. The voltage IN whose offset has been adjusted by the gain offset adjustment section 58 is
It is integrated using an OP amplifier 171. In this integrating circuit, an input voltage IN is inputted to one input terminal via a resistor R1, and the input terminal is grounded via a resistor R2. A capacitor C is connected between one input terminal and an output terminal, and a FET switch and a resistor R are connected in parallel to the capacitor C.
3 is connected. On the other hand, a constant voltage ST is input to the input terminal. This FET switch is activated during reset (INT
The GT signal becomes H and the signal C level, and the charge in the capacitor C is discharged. During integration, the INTGT signal is set to L level, the FET switch is turned off, and integration is performed.

At this time, using the circuit constants shown in the figure, the output voltage INTG
is as follows.

(Here, [ represents the integration time.) Figure 30 shows the integral waveform P when the integral input voltage IN is equal to the reference voltage ST, that is, at the maximum concentration, and when the integral input voltage IN is greater than the reference voltage ST. , that is, shows the integral waveform Q when the concentration is smaller than the maximum value. In addition,
T)I is the threshold value of the comparator 60 at the subsequent stage.

When the integral waveform is lower than the threshold value TH, the laser driving pulse LDON is output. Here, for maximum concentration, the integrated waveform reaches the threshold TH at the end of the integration period. If the concentration is below the maximum concentration, the integrated waveform Q will exceed the threshold TH by the end of the integration period.

In addition, when emphasizing gradation at high density, the integral waveform P
applies to the case of concentration O corresponding to the minimum input voltage (see FIG. 13).

In the case of high-speed two-system type (see Figure 18), the integration section 59.7
3, the integrator shown in FIG. 29 is used, and the comparator section 60
．． The comparator shown in FIG. 829 may be used as 74.

Here, a control signal is generated so that integration is performed alternately in both systems.

FIG. 31 and FIG. 32 show timing charts in photo mode and standard mode, respectively. Integral part 59.73
are the signals INTGTI and IN, respectively, indicating the integration period.
Integration and reset are performed alternately by TGT2.

FIG. 33 shows a circuit diagram of the comparator section 60. The output signal INTG of the integrating section 59 is input to one input terminal of the comparator 181 via a resistor. Further, the collector terminal of a transistor 183 whose emitter is grounded is connected to this − input terminal, and the integrating section 59 is connected to its base terminal.
The <INTGT signal is input based on the INTGT signal output from the clock generator 64 except for the reset period. On the other hand, the threshold voltage TH is input to the ten input terminal of the comparator 181. Note that this threshold voltage TH is
The switch 182 is actuated by the PH0TO signal (l: standard mode, O: photographic mode), and the value is changed to the value corresponding to the addition mode. The comparator 181 compares the re-input voltages and outputs a comparison signal CMP for creating pulse radiation. During the reset period of the integrating section 59, the INTGT signal is supplied to the gate of the transistor 183, the transistor 183 becomes conductive, a forced high-speed reset becomes possible, and the voltage at the input terminal is set to OV. Therefore, the comparison signal CMP is output faithfully to the density data.

In order to eliminate the influence of resetting the integrating section 59, the third
As shown in FIG. 3, signal processing is facilitated by setting the input of the comparator to 0 during the reset period, but as shown in FIG. 185, and the output signal of the AND gate 185 may be used to cancel during the reset period.

By the way, according to the method of increasing the input voltage IN to the integrating section at low concentrations, the LDON signal is output until the integrated voltage exceeds the threshold TH, as shown in FIG. 2 (A). No matter how small the input voltage is (that is, no matter how large the input voltage to the integrating section is), the LDON signal will always be output.For example, even in the background part of the image, the semiconductor laser will turn on slightly, causing the image to become dirty. There is a possibility that it will appear. Therefore, Fig. 8 1
In the two-system circuit shown above, the comparator section 60
．． 74 is configured as shown in FIG.
By inputting the light to the LD driver through the gate, the above-mentioned problem at the lowest concentration can be solved.

The circuit shown in FIG.
The G signal is sent directly to the comparator 18 without going through a transistor.
1, and the threshold voltage TH is connected to one terminal of comparator 1.
It is input to the child terminal of 81.

That is, according to this configuration, the CMP signal is input to the above-mentioned AND gate also during the reset period of the integrating section, corresponding to the period in which the integrated voltage INTG signal is lower than the threshold voltage TH.

Therefore, the integrated voltage and LDON output from the AND gate
As shown in Figure 36, the relationship with the signals is that when one integrator is reset, the LDON signal rises when the integral waveform falls to the threshold TH, and the other integral waveform exceeds the threshold TH. At this time, the LDON signal falls. In this case, even at the lowest concentration, there is some time from the beginning of the integration period until the integrated waveform exceeds the threshold TH, but until the reset waveform of the other integrator falls below the threshold TH, the LDON The signal does not go to H level.

Since the time it takes for this reset waveform to fall below the threshold TH becomes longer as the density data becomes smaller, the above-mentioned problems such as stains on the background of the image can be solved.

(Effects of the Invention) When expressing gradations and reproducing halftones by modulating pulse width, pulse width modulation can be performed using a new density data integration method.

Since the integral value is compared with the threshold value, the pulse width changes with respect to the density data in the direction of correcting the γ correction, so the correction width of the density data becomes narrower, and the accuracy of correction can be increased.

[Brief explanation of the drawing]

FIG. 1 is a schematic block diagram of the basic configuration of pulse width modulation using concentration data integration method. FIGS. 2A and 2B are diagrams of integral waveforms, respectively. FIG. 3 is a diagram showing changes in gradation of a printed image corresponding to density data. FIG. 4 is a diagram of correction data for density data. FIG. 5 is a diagram of γ characteristics. FIG. 6 is a diagram of correction data for the γ characteristic. FIG. 7 is a diagram schematically showing actual correction data. FIG. 8 is a diagram of an integral waveform in an integrator. FIG. 9 is a diagram of an integral waveform in high-speed two-system processing. FIGS. 1A and 1B are diagrams showing pulse generation with an integration period longer than the basic period, respectively. Figures 11 (A) and (B) are respectively Figure 10 (A)
, (B) is a diagram showing pulse generation when an addition value is input. FIGS. 12A and 12B are diagrams showing the influence of noise on the comparator, respectively. FIG. 13 is a diagram of correction data in a modified example. FIG. 14 is a diagram of an integral waveform in high-speed two-system processing in a modified example. FIG. 15 is a block diagram of the overall configuration of the laser printer. FIG. 16 is a diagram of the operation panel. FIG. 17 is a block diagram of the print head controller. Figure 18 shows two high-speed two-system print head control units.
It is a block diagram of a system part. FIG. 19 is a block diagram of the line buffer memory. FIG. 20 is a block diagram of the adder. FIG. 21 is a timing chart in photo mode. FIG. 22 is a timing chart in standard mode. FIG. 23 is a block diagram of the correction section. FIG. 24 is a diagram of addresses of the ROM of the correction section. FIG. 25 is an address map of the ROM of the correction section. FIG. 26 is a diagram of an example of printing of three lines. FIG. 27 is a graph of the dependence of gradation on density data. FIG. 28 is a circuit diagram of the sub-scanning gain switching section.~FIG. 29 is a circuit diagram of the integrating section. FIG. 30 is a diagram of an example of an integral waveform in the integrating section. FIG. 31 is a timing chart of integration in photo mode. FIG. 32 is a timing chart of integration in standard mode. FIG. 33, FIG. 34, and FIG. 35 are all circuit diagrams of the comparator section. FIG. 36 is a diagram of an integral waveform in high-speed two-system processing. l...correction section, 3...integrator, 4...
Comparator, 12... Operation panel, 13...
・Print head control unit, 24... Addition mode selection key, 53... Addition unit, 5
6... Sub-scanning gain switching section, 59... Integrating section, 60
... Comparator section, 62 ... Semiconductor laser,
73...Integrator section, 74...Comparator section.

Claims (3)

[Claims] (1) In an image forming apparatus that generates an image recording signal having a pulse width that is modulated in accordance with density data, and prints dots of a size corresponding to the pulse width of this image recording signal, an integrating means that integrates for each integration period, and a pulse width that compares the integral value of the integrating means with a predetermined threshold value after the start of integration within the predetermined integration period, and determines the pulse width according to the comparison result. 1. An image forming apparatus comprising: comparing means for generating an image recording signal having the following characteristics: (2) In the image forming apparatus according to claim 1, timing generating means generates the timing of integration in the integrating means so that the predetermined integration period of the integrating means is a period different from the output cycle of the density data. An image forming apparatus comprising: (3) The image forming apparatus according to claim 2, further comprising a selection means for selecting an integration period of the integration means, and a timing generator for generating the timing of integration in the integration means in accordance with the integration period selected by the selection means. An image forming apparatus comprising: means.