US 3828354 A
An ink drop printer for correcting drop-to-drop interactions by controlling the charge applied to a drop being formed based on the position to which the drop is to be deflected on the print medium, the charge placed on a selected number of previously formed drops and the charge to be placed on a selected number of drops to be formed.
Claims available in
Description (OCR text may contain errors)
United States Patent  Hilton Aug. 6, 1974 INK DROP CHARGE COMPENSATION METHOD AND APPARATUS FOR INK DROP PRINTER  Inventor:
Howard T. Hilton, San Jose, Calif.
International Business Machines Corporation, Armonk, NY.
 Filed: Sept. 27, 1973  Appl. No.: 401,330
 References Cited UNITED STATES PATENTS 3,631,511 12/1971 Keur et a1. 346/75 3,789,422 l/1974 Haskell et al. 346/75 Primary ExaminerJoseph W. l-lartary Attorney, Agent, or FirmOtto Schmid, Jr.
 ABSTRACT An ink drop printer for correcting drop-to-drop interactions by controlling the charge applied to a drop being formed based on the position to which the drop is to be deflected on the print medium, the charge placed on a selected number of previously formed drops and the charge to be placed on a selected number of drops to be formed.
9 Claims, 8 Drawing Figures sa CLOCK PRIMARY J40 PULSE DEFL. SIGNAL GENERATOR GENERATOR e52 DATA DATA FLIGHTCORRECT'N 54 SOURCE STREAM SIGNAL 4 4 MEANS MEANS GENERATOR INDUCTION -46 CONTROL CORRECTION MEANS SlGNAL GENERATOR PAIENIEDMIB 61 14 3.828.354
sum 1 or 5 VOLTAGE INK PRESSURE SOURCE SYSTEM DATA INPUT FIG-1 40 v. CLOCK PRIMARY PULSE DEFL SIGNAL I GENERATOR 1 GENERATOR 52 54 36, F r 54 DATA DATA FLIGHTCORRECT'N ADDER SOURCE STREAM ,SIGNAL 481 MEANS MEANS MEANS GENERATOR 712 mnucnow 6 CONTROL CORRECTION MEANS SIGNAL f GENERATOR 52 PAIENIEDMIB w 3.828.354
sum 2 or 5 PAPER 2 monow R 7 6202428 22 26 50%??? i6 2U24'28'Y 2226'30' v 1 SB RRR R- Rl-R- 1 l l l 1 v I 2 3 4 5 6 7 8 9 10 H 12 15 14 i5 16 I? 18 19 20 21 22 23 24 25 26 27 28 29 50 FIG. 3
ERROR TAGE VOL
i 5 R) 15 2b 25 5'0 MATRIX FIG 4 SECOND LEAD(Le2Ei) POSITION PATENIEDAUB 61w 3.828.354
SHEET 5 BF 5 SIGNAL GENERATOR CLOCK COUNT (5 BITS) CLOCK DEGODE SUM CHARGE VOLTAGE Lei EiA LeiEiB LeiEiC Le2E1A -Le2E1B L01 E2 I Le1E2A Le1E2B Le 2 EM Le2E2B Le4E2A FIG.8
INK DROP CHARGE COMPENSATION METHOD AND APPARATUS FOR INK DROP PRINTER CROSS REFERENCE TO RELATED APPLICATIONS Non-sequential Ink Jet Printing Ser. No. 325,494, filed Jan. 22, 1973, related to the matrix scanning order described in this application.
BACKGROUND OF THE INVENTION This invention relates to an ink drop printing aparatus and more particularly to improved apparatus for producing'more accurate drop deflection and placement on a print medium.
In the prior art it has been known to utilize selectively charged ink drops to print characters. One problem which arises in this type of printer is distortion in the printed characters due to interaction between drops. To correct this distortion there has been tried in the past the use of so-called guard drops in which at least one uncharged drop is placed between eachof the charged drops. This technique iseffective to a'certain extent, but the technique has the disadvantage that the printing process is slowed due to the unavailability of the uncharged drops for printing operations. Another technique which has been used in the prior art is to compensate the charge applied to a drop being formed based on a determination that the previous drop has been charged. This technique has the advantage that all drops are available for printing. However, this technique does not completely eliminate the distortion problem due to drop-to-drop interactions, particularly at high printing rates encountered in a line printer operation.
SUMMARY OF THE INVENTION It is therefore the principle object of this invention to provide a method and apparatus wherein all drops can be used for printing accurately formed characters at high printing rates.
It is another object of this invention to provide an improved method and apparatus wherein the charge voltage for each drop is compensated for both previously charged drops and the charge to be placed on drops to be formed.
Briefly, according to the invention there is provided an ink drop printer in which ink drops are charged according to an information signal and deflected for proper placement on'the print medium to produce the designated character. The appropriate charge voltage for each drop is generated by algebraically combining a positional signal defining a particular drop placement position, a signal generated in response to the charge placed on a predetermined number of previously formed drops and a signal derived from the charge to be placed on a selected number of drops to be formed.
BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagrammatic schematic" view of the ink jet printing apparatus embodying the invention;
FIG. 2 is a schematic block diagram of the charge voltage generator with correction according to the invention;
FIG. 3 is a diagram showing possible ink drop placement according to one matrix scanning order;
FIG. 4 shows a series of graphs of matrix position vs. error voltage for induction error correction;
FIG. 5 shows a series of graphs of matrix positions vs. error voltage for correction of flight errors;
FIG. 6 is a schematic block diagram of a specific em bodiment of the control means utilizing digital techmques;
FIG. 7 is a schematic block diagram of an alternate embodiment of the control means using analog techniques;
FIG. 8 is a schematic block diagram of a third embodiment'of the control means utilizing analog techmques.
DESCRIPTION OF THE PREFERRED EMBODIMENTS Referring to FIG. 1, there is shown an ink drop print? ing apparatus in accordance with the invention. This apparatus has an ink pressure system 10 suitably connected so that ink is driven through a plurality of nozzles 11 at high pressure to form a plurality of jets 12 each of which is controllable to print character's intwo adjacent print positions in the print line. An' electromechanical transducer 14 is provided and signal source 15 is utilized to energize transducer 14 to vibrate the jet systems so that the jets break up in synchronism with the impressed vibration to form a series of uniform spaced ink drops 16. As each ink drop 16 is being formed, a charge electrode 18 in the vicinity of the jet stream can be used to induce a charge on that drop. When breakoff of the drop occurs, the charge is trapped so that the drop is subject to the action of a subsequent electrostatic field. Charging of each drop is accomplished by the instantaneous value of the voltage produced at the instant of drop breakoff by control means 20. After charging, the drop is deflected by a fixed electrostatic field produced by deflection means 22 when energized by a suitable high voltage source 24. The amount of deflection produced by the electrostatic field is proportional to the amount of charge that has been induced by the charging electrode on the drop so that a control exists as to where a droplet may be placed on print medium 26. The electrostatic field is operable to produce a one-dimensional line scan to form line sections of a character and the transverse motion of the print medium 26 produces another dimension so that repeated line scans can generate a total character. In printing, many areas do not receive ink drops and unwanted drops are deflected by a suitable charge so that they enter into sump means 28 for recirculation within the ink system.
Thus, without correction the voltage V applied to the charge electrode is determined by the position to which the drop is to be deflected. However, in the present invention the voltage V is calculated as before, but before being applied to the charge electrode is modified based on whether or not the surrounding drops have been charged.
A general equation for drop deflection correction can be written as follows:
a n m mn V V+ (1) V =Nominal voltage to deflect to position X with no other drops deflected; C 0, l, to represent the logical condition of surrounding drop deflections; B,,,,, the Mth correction term for drop N. This equation would be used directly if one was implementing a read only memory (ROM) correction scheme. In this case the C would be made mutually exclusive (only one could exist) and the value B would be stored at location (n, C =1). This would be a total unreduced approach, i.e.,'while many of the B,,,,, might be equal, there is no, attempt to utilize this fact.
.The economy of ROMs is utilized in this case.
In the simplest implementation, signals proportional to V,, and fixed values can be added to orsubtracted from V,, to yield the charge voltage. This technique yields a corrected signal V, of the form:
c n -1) -i n -1) -2) n +(C+2) (A+2 B+2)- Where v C, 0, 1 digital value Bit X;
' A, the proportional correction;
B, the fixed correction.
The first requirement for developing a specific correction is that a sequential matrix scanning order be defined whichidentifies for each clock period the position to which a drop will go if deflected. Table 1 gives this sequence both in position number (ascending. order from the sump) and the nominal charge electrode voltage for a particular matrix scanning order chosen for purposes of example.
TABLE 1 Nominal Clock Period Matrix Position Charge Volts l 42 1 I6 I32 2 9 90 3 24 I80 4 2 48 5 I7 I38 6 I0 96 7 25 186 8 3 54 9 I8 144 l0 I1 102 ll 26 I92 12 4 6O l3 I9 I50 I4 I2 I08 15 27 198 I6 5 66 I7 I56 l8 I3 I14 19 28 204 20 6 72 21 21 I62 22 I4 I20 23 29 210 24 7 78 25 22 I68 26 I5 I26 27 2l6 28 8 84 29 23 I74 FIG. 3 shows the drop location on the paper of sequential drops according to the matrix scanning order of Table l. The matrix scanning order given for purposes of example is described in greater detail and claimed in the above-identified application Ser. No. 325,494. In this matrix scanning order, the drops issuing sequentially from nozzle 11 are never printed adjacent to one another. This placement maximizes the inflight distance between successive ink drops thereby reducing drop interaction. The total print coverage of the head shown in the specific embodiment is two character positions of 30 matrix position numbers as shown in FIG. 1. The slight misalignment in the'printing which results from the non-sequential printing can be. corrected by displacing the deflection plates 22 a sufficient angle ()5 with respect to the print line.
In conjunction with the data stream, Table l defines the charge electrode voltage for each period for the case when an isolated drop is deflected. When the data stream is a zero for the subject period, the charge electrode voltage is zero volts. When the data stream is a l for the subject period, the charge electrode voltage assumes the voltage in the table (204 volts for period 19, for example). Without correction, the data stream and the matrix scanning order is. all that is needed to define the charge electrode voltage.
However, when this is done, significant errors in drop position are encountered. These errors come from two primary sources. The first and generally largest source of erroris chargeinduction. When a dropis generated in the presence of other charged drops, its charge is a function of the charge electrode voltage and the charge on adjacent previously formed drops The magnitude of this effect is determined by several factors such as the charge electrode spacing, wave length-to-jet diameter ratio, centering in the charge electrode, etc. and FIG. 4 represents the magnitude for the matrix scanning order shown in FIG. 3. The number of drops considered for the correction is determined by the print pa rameters and the residual allowable error. Plotted in FIG. 4 is the required correction voltage to achieve the corrected charge on the drop versus the matrix position number for the drop for a'number of drops chosen to meet assumed error specifications. These correction voltages can be determined experimentally or by calculations from theoretical models of drop charging and flight interaction. These corrections are needed whether or not the drop is deflected and are linear with the affecting drops charge. The correction for the first lead (previously formed) drop and second lead drop are shown. For this matrix scanning order, the third lead drop was considered to be below the error threshold so no correction is included for this drop. The ertors are designated as Lel El and Le2 E1. The E1 identifies the error source as charge induction and Lel and Le2 refer to the first lead drop and second lead drop respectively.
The second error source is flight interaction. In flight two charged drops interact through their charges (electrostatic repulsion) and through the lead drop wake (aerodynamic drag reduction). Both of these effects cause errors in the landing position of a drop. However, in a well-chosen matrix pattern and deflection geometry, these errors can also .be corrected by charge electrode voltage modification. The correction values for these interactions are shown in FIG. 5 for the chosen matrix scanning order. These errors are designated Le4 E2, Le2 E2, Lel E2, Lal E2 and U12 E2. The E2 designates flight errors and the prefix refers to the relative drop sequence number as Le for lead drop (previously formed) and La for lag (to be formed) drop. The flight corrections are needed only when the subject drop is deflected. For the matrix scanning order given in the example (FIG. 3) and the specific drop parameters and flight geometry, correction for four lead and two lag drops are required to meet the assumed error specifications.
A general block diagram of control means according to the invention is shown in FIG. 2. A control means 32 is provided to supply signals to coordinate the transfer of data in the charge signal generating means with other circuits and components of the printing system such as head phasing, control of paper, ink pressure etc. The data to be printed is supplied by data source means 34 and the data is supplied on a serial basis to data stream means 36. Clock 'pulse generating means 38 provides a series of pulses in synchronism with ink drop formation. Clock pulse generating means 38, data stream means 36 and control circuits 32 may be provided as hard wired circuits, as a special purpose com- Puter, or by rz lx wsra l sisev a .M2229 stored in a Read Only Storage device 62. The correction provided by Read OnlyStorage device 62 corresponds to the error voltage shown in FIG. 4 for induction errors and in FIG. 5 for flight errors for thespecitied matrix position. A partial Read Only Store map is.
provided in Table II to indicate the correction stored in Read Only Store 62 for the matrix scanning order shown in FIG. 3 and the correction voltages shown in FIGS. 4 and 5. The primary deflection signal and the combined correction signals are combined in adder 64 to produce a digital signal indicative of the charge to be placed on the charge electrode 18 for the drop being formed. This signal is converted to an equivalent analog signal in digital-to-analog converter 66 and this signal at terminal 68 is coupled to charge electrode 18 to produce the proper charge on the drop being formed to produce the drop deflection for the characterspecified by'the input data.
TABLE 2 PARTIAL ROS MAP Data Correction Matrix Position Stream Invoked l 8 l5 I6 2 l 0l24 0 0 0 0 0 0 None 0 0 O 0 0 0 0 0 0 0 l None 0 0 0 O 0 000100 LelEl 24 29 23 5 l7 0 0 l 0 0 l Le4E2 0 2 4 2 7 0 01 I00 LelELEZ l 32 26 5 l5 0 0 l l l l LelEl,LelE2.Le2El 6 42 ll 30 Le2E2,Le4E2 l l 0 0 0 0 None 0 0 0 0 0 l l l 0 0 0 LalE2,La2E2 0 3 4 2 0 l l l l l i All 6 45 44 13 30 Memory Contents for DS Matrix Position computer. A primary deflection signal generating means 40. is provided to generate the primary deflection signal substantially as shown in Table I for the selected matrix scanning order. A first correction signal generating means 44 is provided to generate an induction correction signal on line 46 and a second correction signal generating means 42 is provided to generate flight correction signals online 48. Adder means 50 is provided to produce an algebraic sum of the primary deflection signal on line 52, the flight correction signal on line 48 and the induction correction signal on line 46. The output of adder means at terminal 54 is coupled to the charge electrode 18 to produce the proper charge on the drop being formed so that the drop is de-' flected for accurate placement on print medium 26 to produce the character specified by the input data.
An implementation of the charge signal generating means which utilizes digital circuits is shown in FIG. 6. The clock signal generating means comprises a five-bit clock counter 56. The data stream means comprises a seven-bit register 58. In the embodiment shown register 58 is a serial-in, parallel-out shiftregister which has data at all times for the last four drops formed (1 to 4), the drop being formed (0) and the next two drops to be formed (+1, +2). The primary deflection signal generator 60 produces a digital output corresponding to the deflection signal specified in Table I for the 30 counts provided by the clock counter 56 to represent the matrix positions. In this embodiment the flight correction signal generator and the induction correction s al snsratm a .wmbinsd and, th siafo matiqnis In FIG. 6 the primary deflection signal generator comprises logic circuits which transform the clock count to a binary weighted code which corresponds to the voltage values required for the primary deflection voltage. All the correction is generated by accessing the correction ROS 62. The memory address register 66 for ROS 62 utilizes address data from two sources, six bits from the data stream and the five bits from clock counter 56. At each address in ROS 62 is the correction demanded by the bit pattern from the data stream and the clock count. Based on the magnitudes of the voltages in the specific example disclosed, six bits are required to cover the total correction range and be directly scaled with the primary deflection. Table 2 shows the partial contents of ROS 62. The left hand column shows the data stream values and the second column indicates the correction segments (from FIGS. 4 and 5) that each pattern invokes. Representative matrix position values are shown which designate the correction value stored at the address designated by the clock count and the data stream. When addressed by the clock count and the data stream, the output of ROS 62 assumes a primary weighted code equal to the contents of that address. The primary deflection voltage in binary code and the correction voltage from ROS 62 are algebraically added in adder 64 to produce a binary code corresponding to the corrected voltage for charge electrode 18. These signals are applied to DAC 66 to produce an analog signal which is proportional to the required charge voltage, and this signal is applied directly to the charge electrode amplifier.
The'capacity of ROS 62 for the embodiment described above is 2,048 six-bit words. One simple variation in this control circuit is to extend the capacity of ROS 62 to 8 bits which permits the storage of both the primary deflection code and the correction code thereby eliminating the deflection code generator 60 and adder 64. A second variation of this control embodiment is to break up the correction into independent parts (segments which can be linearly summed) by adding a register to the adder and multiple access to the ROS. The corrected voltage code could be generated by summing the component corrections. In this case, the number of addressbits for the ROS would be reduced from II to 8 bits clock count, 3 pattern), but
the complexity of the circuit would be increased and the time to generate the corrected voltage would be increased as well. The embodiment shown in FIG. 6 is attractive since the correction hardware can be multiplexed between many heads with an all-digital interface. On' the input side the data stream for each head is multiplexed to the data stream input and in this case the data would be entered into register 58 in parallel instead of serially. This arrangement would also require a register assigned to each head to latch up the output of the adder by latch 65 (FIG. 6) and supply the input to its DAC 67'.
The embodiment of the charge signal generating means shown in FIG. 7 hasthe same inputs to shift register 58 and the primary'deflection code generator 60 is identical. The addressing is also accomplished by clock counter 56. The primary deflection code is generated by generator 60 and coupled directly to DAC 70 to produce the primary deflection signal on line 7 1. The primary deflection signal is also coupled to the inputof lead I register 72. The lead I register 72 and lead 2 register 74 are parallel-in and parallel-out registers and these registers are clocked synchronously with the data stream and clock counter 56. The output of lead I register 72 is coupled to a second DAC 76 and to the inputs of the lead 2 register 74. The lead 2 register outputs are coupled to a third DAC 78. The contents of the lead I register and the lead 2 register are the deflection code for drops n-1 and n-2 respectively, With the appropriate gain settings on DACs 76 and 78, the DAC outputs generate the LelEl term on line79 and the Le2El term on line 80. In this embodiment, the E2 terms are generated by logically decoding each segment in FIG. 5 and gating an analog signal which approximates that segment. A segment refers to a correction which can be implemented by a term AV which is approximately equal to A+BV,,, or in other words, the correction voltage equals a constant and a term proportional to the nominal voltage. The logical AND circuits 80 combine the present bit DS(O) and the interferring conditions DS(fl) to identify the data de- 5 pendent correction terms needed. The clock is de-' coded into groups of matrix positions which correspond to the correction segments. These two sets of inputs are combined in AND circuits 82 to produce the correction logic gates. Each AND gate corresponds to one of the correction segments and is the gating signal to one of the analog gates 84. The analog gates have as their input the primary deflection term on line 71 and a DC level with the appropriate impedances toproduce the required A+l?tV, term when the gate is open. To produce the correct charge voltage, the primary deflection term, the two E1 terms and the 8 E2 terms are summed by summing amplifier 86 to produce the proper charge voltage at terminal 88.
The embodiment shown in FIG. 8 is similar to the previous embodiments in the generation of the primary deflection signal. The primary deflection signal is coupled to DAC 90 and this signal is gated by analog gate 92 to produce the primary deflection voltage. In this manner, the nominal deflection voltage is available for use in generating the E1 correction terms- Five more. terms are required for the E1 terms and the AND gates 94 are not conditioned by DS(O). Generation of all the correction terms is now the same as the generation of the E2 terms in the previous embodiment. The primary deflection voltage and the 13 correction terms are summed in summing amplifier 96 to produce the corrected charge electrode voltage.
While the invention has been particularly shown and described with reference to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in the form and details may be made therein without departing from the spirit and scope of the invention.
What is claimed is:
1. The method of printing in an ink drop character printer of the type wherein a stream of individual ink drops are each selectively charged responsive to character information signals for selective placement on a record member according to a predetermined scanning order, the method comprising the steps of:
generating a primary deflection signal responsive to said character information signals and the position within the scanning order of the drop being formed; generating a first correction signal based on the relative position in said scanning order and the information signals relative to a first predetermined number of previously formed ink drops to compensate for charge induction errors;
generating a second correction signal based on the relative position in said scanning order and the information signals relative to a second predetermined number of previously formed drops and a third predetermined number of drops next to be formed to compensate for flight errors; and
combining the primary deflection signal, the first correction signal and the second correction signal to apply to the charging electrode so that said drop being formed can be deflected in a constant electrostatic field to the place on the record member to produce characters designated by said character information signals.
2. In an ink drop character printer of the type wherein a stream of individual ink drops are each selectively charged responsive to character information signals for selective placement on a record member according to a predetermined scanning order, the improvement comprising:
means for generating a primary deflection signal responsive to said character information Signals and the position within the scanning order of the drop being formed;
means for generating a first correction signal based on the relative position in said scanning order and the information signals relative to a first predetermined number of previously formed ink drops to compensate for charge induction errors;
means for generating a second correction signal based on the relative position in said scanning order and the information signals relative to a second predetermined number of previously formed drops and a third predetermined number of drops next to be formed to compensate for flight errors; and
means for combining the primary deflection signal,
the first correction signal and the second correction signal to apply to a charging electrode so that said drop being formed can be deflected in a constant electrostatic field to the place on the record member to produce characters designated by said character information signals.
3. The apparatus according to claim 2 comprising control means for producing signals to coordinate the generation of said deflection signal and the other com ponents of said printer.
4. The apparatus according to claim 3 wherein said control means comprises:
means for producing a stream of data signals representative of the data to be printed, and
means for producing a series of clock pulses in synchronism with the formation of said ink drops.
5. The apparatus according to claim 4 wherein said first correction signal and said second correction signal are computed and placed in a storage means, and
means for selectively accessing said previously stored correction signals responsive to said clock pulses and said stream of data signals.
6. The apparatus according to claim 4 wherein said induction error correction signals comprise a predetermined proportion of said prior primary deflection signal, and
means for selectively gating predetermined flight error correction signals responsive to said clock pulses and said stream of data signals.
7. The apparatus according to claim 4 comprising means for generating a plurality of induction error correction signals and a plurality of flight error correction signals, and
means for selectively gating said correction signals responsive to said clock pulses and said stream of data signals.
8. The apparatus according to claim 4 wherein said means for producing a stream of data signals representative of the data to be printed comprises a shift register.
9. The apparatus according to claim 5 wherein said means for accessing said correction signals is shared by a plurality of printing heads.