United States Patent 3866237

Phase control device for an ink jet printer in which the phase between ink droplet forming means and the droplet charging means is adjusted and maintained by using the reset signal of a ring counter of predetermined capacity and changing the number of pulses in a train producing cycling of the counter. The counter is normally supplied with clock pulses at a frequency which is an integral multiple of the frequency of droplet formation and counter capacity is such that reset signals nominally occur at the same frequency as the droplet formation so that charging signals can be applied in phase with the formation of droplets. For unwanted droplets, a sawtooth wave is applied to the charge plate so that the charge on the droplet depends on the exact moment of drop break-off. The magnitude of the sawtooth wave is such that all unwanted droplets strike a gutter. The location of droplet impact on the gutter is sensed and the pulse train to the counter is changed by either adding or deleting a pulse to thereby change the time of the reset and initiation signals by a fraction of the droplet formation frequency.

Application Number:
Publication Date:
Filing Date:
International Business Machines Corporation (Armonk, NY)
Primary Class:
Other Classes:
International Classes:
B41J2/115; G06K15/10; H04N1/23; (IPC1-7): G01D18/00
Field of Search:
View Patent Images:
US Patent References:

Primary Examiner:
Hartary, Joseph W.
Attorney, Agent or Firm:
Johnson, Kenneth P.
What is claimed is

1. In an ink jet printer having a transducer for vibrating a nozzle at a predetermined frequency to produce droplets having a selected phase relation with applied, variable, input information charges induced thereon by an input information source energizing a charging electrode for correspondingly varying deflection of said droplets in an electric field between deflecting electrodes, apparatus for varying the phase relation between said transducer vibrations and said applied charges comprising:

2. Apparatus as described in claim 1, wherein said control means includes means for generating sampling signals for enabling change of said phase relation and gating means connected to said clock pulse means and sampling signal means, being operable for selecting certain said output signals to be effective to change said count cycle time.

3. Apparatus as described in claim 2, wherein said gating means is rendered operable by the occurrence of a sampling signal in predetermined relation with one of said clock pulses.

4. Apparatus as described in claim 3 wherein said sampling signals occur at a frequency less than said calibration signal frequency.

5. Apparatus as described in claim 1, wherein said detecting means has two sensor locations and produces respective output pulses peculiar thereto; and

6. Apparatus as described in claim 1 wherein the output signals of each said location alter the number of pulses in said train in a particular manner.

7. Apparatus as described in claim 1 wherein said predetermined calibration signal frequency is the same as said nozzle vibration frequency.

8. Apparatus as described in claim 1, wherein said detecting means has first and second sensor locations and produces respective first and second output signals; and

9. Apparatus as described in claim 1, wherein the frequency of said clock pulses is an integral multiple of the frequency of said calibration signal frequency.

10. Apparatus as described in claim 1 wherein said calibration signal frequency is the same as said nozzle vibration frequency and said clock pulse frequency is an integral multiple thereof.

11. In an ink jet printer system wherein ink under pressure is delivered to a nozzle which is vibrated at a predetermined frequency by a transducer means to produce a stream of drops and a charging electrode is positioned adjacent said nozzle to charge said drops in a selected phase relation with information signals from input information signal means applied to said electrode for producing deflection of said drops in accordance with said information input signals as said drops move in an electric field between a pair of deflecting electrodes, apparatus for varying the phase relation between the formation of said drops and said information input signals comprising;

12. In an ink jet printer system in which pressurized ink issues from a nozzle vibrated by a transducer means to produce a stream of droplets that receive in a selected phase relation from a charging electrode at the time of formation a charge representative of information from an information input means to establish a subsequent marking or discard position upon passing through an electric field between a pair of deflection electrodes, apparatus for varying the phase relation between said transducer vibration and said applied charge comprising:


A portion of the apparatus herein disclosed is part of the subject matter of a patent application entitled "Phase Control for Ink Jet Printer," Ser. No. 253,065, filed May 15, 1972 by J. W. Haskell, now abandoned and refiled as a continuation, Ser. No. 380,641 on July 7, 1973 and assigned to the assignee of the instant application.


This invention relates generally to ink jet printers and more particularly to apparatus for maintaining the proper phase relationship between the droplet forming means and charging voltage for each droplet.


The phase relationship between the applied charging voltage and the droplet formation is critical to the accurate placement of the droplets upon a printing surface. When the droplet is formed during the transition from one charging voltage to another, deflection cannot be predicted and drops are misplaced on the printing surface, generally resulting in deformed characters. In the past, attempts to insure the proper phase relationship have employed usually an analog circuit to change the relationship or have employed coarse adjustments such as 180° or 90° of phase shift which then becomes an approximation if the phase should have been corrected by different amounts.

During operation of the prior art phase control circuits, there is a tendency for voltage levels to gradually shift so that accurate phase maintenance deteriorates. Also in the prior art, the correction signal which is determined to be necessary is used to control a delay circuit which is satisfactory for the coarse adjustments but makes fine adjustments in the phase relationship nearly impossible unless elaborate delay networks are employed.

Some of the correction circuits used in the past also set aside special phase checking times during operation. However, the above mentioned application, Ser. No. 253,065, discloses a scheme wherein each unused ink droplet is given a calibration charge and the impact region of that discarded droplet can be detected and appropriate corrections made. This permits the phase correction to occur at random times without waiting until special testing cycles can be made.

It is accordingly a primary object of this invention to provide a new and novel phase control system for an ink jet printer which employs digital signals to attain the necessary phase shift.

Another important object of this invention is to provide apparatus for phase shift control for an ink jet printer in which the increments of adjustment can be of nearly any desired size.

A further object of this invention is to provide a phase control system for an ink jet printer in which a ring counter of predetermined capacity is continuously supplied with clock pulses and the charge application to ink droplets depends upon the time of reset of the ring counter.

A still further object of this invention is to provide a phase control system for an ink jet printer in which phase correction can occur over a wide range of time intervals.

It is also an important object of this invention to provide a phase control system for an ink jet printer in which phase change is bi-directional and in which the system will automatically find a region of optimum performance and then hunt by discrete steps about the optimum point.


The foregoing objects are attained in accordance with the invention by providing a clock pulse generator that has a frequency that is an integral multiple of the frequency of the ink droplet forming means. The droplet forming means is operated through a frequency divider and the clock output is directly connected to a ring counter through gating logic circuits. The capacity of the ring counter is equal to the integral multiple of clock pulses per droplet generated. The output of the ring counter during reset provides an initiation signal for both the character generator circuit and the sawtooth calibration charging circuit for discarded droplets, which in turn controls the voltage application to the charging plates at the time of droplet formation.

Discard droplets charged by the ramp voltages will fall into one of a pair of impact locations in a gutter according to the ramp voltage level at the time of droplet formation. Droplet impact is detected and a signal is generated as to which impact area the discard droplets are impinging upon. These impact signals are used to control the above mentioned logic circuits to either suppress one count or add one count to the pulse train which is supplied by the clock to the ring counter. Thus, the time at which the reset signal occurs in the ring counter varies at each correction, either approximately one clock pulse period earlier or one clock pulse period later.

The timing of the pulse train alterations is controlled through a sampling pulse which can be supplied at regular or randomly selected intervals. The sampling pulse activates gating circuits which will permit the addition or deletion of a clock pulse to the ring counter. The gating circuits also assure that pulse deletion or addition will occur at the proper time to avoid interference with the regularly generated clock pulses.

This invention has the advantage of permitting smaller increments of adjustment than heretofore possible. For example, in the embodiment to be described, the clock pulse frequency is 1 MHz while the droplet formation is at 125 KHz. Thus, incremental adjustments are 45° for each change. Only one alteration of the normal clock train is permitted in any one counting cycle. By using discard drops as a means for testing the phase relationship, correction can be made at various intervals. For example, the sampling pulse, if desired, can be produced about as frequently as each millisecond or at greater time intervals such as 15 to 20 milliseconds, depending upon the amount of control necessary for proper drop deflection. In the embodiment disclosed, the discard drops do not impact a correct position in the gutter. Instead when impact occurs in one of two gutter positions, the correction circuits change the direction of phase shift to move the discard drop impact into the other gutter location.

This phase control circuit uses only digital signal levels, and hence avoids the need for close discrimination as required by analog circuits. The control circuit is simplified in arrangement and does not require complex special circuits.

The foregoing and other objects, features and advantages of the invention will be apparent from the following more detailed description of preferred embodiments of the invention as illustrated in the accompanying drawing.


In the drawing:

FIG. 1 is a schematic circuit diagram of a phase control system for an ink jet printer constructed in accordance with the invention;

FIGS. 2-5 are timing diagrams of signals illustrating various embodiments during operation of the phase control system of FIG. 1.


Referring to FIG. 1, reference numeral 10 denotes ink droplet generating and deflecting apparatus usually employed in an ink jet printer system. Ink from supply 11 is brought through duct 12 and pressurized by pump 13 to issue from nozzle 14 which is vibrated by transducer 15 to form individual droplets in the vicinity of a pair of charging electrodes 16. The voltage applied across the nozzle and electrodes 16 is varied to induce differing charges on the droplets as formed, which then pass between deflecting electrodes 18 connected to a high voltage source providing a constant electric field. The droplets are thereby caused to move in a trajectory determined by the individual charge each carriers as it passes through the field so as to impact a printing surface 19 or a gutter 20. Gutter 20 collects the discarded droplets and is connected to a source of vacuum, not shown, which removes the ink and usually returns it to supply 11 for recirculation.

During operation, the formation of droplets does not always occur at the same break-off point between charging electrodes 16 and hence, may receive an inappropriate charge causing erratic deflection when passing between electrodes 18. The variation in droplet break-off time may be due to change in ink viscosity or temperature or for other reasons. It is therefore important that the phase relationship between the applied charge and nozzle vibration causing the formation of droplets be closely controlled to attain accurately defined printing trajectories.

In accordance with the invention, a pulse generator designated as clock 25 produces pulses at a relatively high frequency, such as 1 MHz. These pulses are supplied to frequency divider 26, preferably at a ratio of 8 to 1 so that the output of the frequency divider is 125 KHz. These signals are supplied to the vibrator drive circuit 27 which activates transducer 15 to thereby produce droplets at the same frequency.

During printing, charges are applied to droplets at electrodes 16 in accordance with signal voltages produced by character generator circuit 28 through charging electrode driver 29. When droplets are to be discarded, they are charged in accordance with a ramp voltage from discard drop charging circuit 30 which also supplies its output to driver 29. Driver 29 permits the application of a discard charge only in the event that no charging signal is received from character generator 28. In this manner, each droplet not used for printing a character is given a charge which will ultimately cause deflection into gutter 20. The output wave form from driver 29 illustrates the composite of two wave forms from character generator circuit 28 and discard charging circuit 30.

Printing of a character is initiated upon receipt of a character selection signal from a source not shown. However, the generator circuit and discard charging circuit are both inoperative until receipt of an initiation signal from a counter 31. Counter 31 is an eight position ring counter which provides an output as an initiation signal upon each reset of the counter at a nominal rate of 125 KHz. The nominal counter reset frequency is the same as the output frequency of divider circuit 26 which is supplied to the nozzle vibrator. Counter 31 is advanced by pulses from clock 25 along line 32 which supplies both of two branches 33 and 34. Clock pulses on line 33 pass through logical OR circuit 35 and serve as one of two inputs to logical AND circuit 36. When AND 36 is fully conditioned the clock pulses are supplied through OR 37 to ring counter 31. The clock pulses on line 34 are supplied to AND 38 which, when fully conditioned, produces pulses at AND 39. When the latter is fully conditioned, then clock pulses will appear at OR 37 for application to ring counter 31.

It is thus noted that counter 31 cannot be operated until the proper gating signals are applied along either line 33 or line 34. The necessary gating signals are obtained from gutter 20 which receives the discard ink droplets. Gutter 20 is formed with two compartments 20a and 20b and is constructed in accordance with the principles disclosed in the aforementioned patent application by J. W. Haskell. Compartment 20a, forming one impact location for discard droplets, contains therein a pair of spaced electrodes 45, indicated schematically in the drawing, which are adapted to be wet with the ink droplets falling into that compartment. Compartment 20b, forming a second impact location for discard droplets and separated by a knife edge 46, has no contacts. Both compartments are, of course, connected to the vacuum removal system for the accumulated ink. Contacts 45 form the sensors to detect into which compartment the ink droplets are currently falling. These contacts are connected to a voltage divider and filter network 47 to provide an electrical output signal indicative of the presence or absence of the ink droplets. This detection network comprises a pair of resistors 48 and 49 connected between a positive voltage source and ground. A capacitor 50 parallels resistor 49 and contacts 45 from ink discard compartment 20a.

When ink droplets, which are electrically conductive, fall into compartment 20a, they complete a circuit between the two contacts 45 to thus produce a low impedance path to ground, bypassing resistor 49, and thus establish approximately a ground potential as one input to comparator circuit 51. However, should the discard droplets be impacting compartment 20b, contacts 45 will be essentially open so that the input signal to comparator 51 will be that voltage appearing across capacitor 50 and resistor 49. Comparator circuit 51 has as its second input a suitable reference voltage source and upon comparison with the detection circuit input will provide either a low level output, indicating droplet impingement in compartment 20a, or a high level output, indicating impingement by the discard droplets in compartment 20b. The output signal of comparator 51 is applied to both AND 39 and inverter circuit 52.

Assume for the time being that the output level from comparator 51 is high, and will serve as the conditioning signal for AND 39 which has a second input from AND 38. It will be recalled that AND 38 has one input clock pulses which are supplied along lines 32 and 34. The second input to AND 38 is from a logical NAND circuit 53, to which one input is clock pulses on line 54, and the second which is along line 55 from a sampling pulse circuit to be described subsequently. At present, it is sufficient to assume that no pulse is present on input line 55. Therefore, with each clock pulse, NAND 53 will not be operable since it lacks a signal on line 55 and its output will therefore be at a high level. This will fully condition AND 38 to thereby gate clock pulses from lines 32 and 34 to AND 39. Hence, with the high level present from comparator 51, AND 39 will be effective to transmit clock pulses to OR 37 and then to ring counter 31. As the ring counter is cycled, each reset provides an initiation signal at its output which starts character generator circuit 28 and discard charging circuit 30 as described above.

If it is now assumed that the output level from comparator 51 is low, it will provide a blocking signal at AND 39. However, inverter 52 will reverse the level and the high level signal will fully condition AND 36 so that clock pulses appearing on lines 32 and 33 and passing OR 35 will be gated through AND 36. The output of AND 36 is supplied to OR 37 and thence to ring counter 31.

At this point it will be seen that ring counter 31 will receive regular clock pulses through either AND 36 or AND 39, depending on whether discard droplets are impacting respective locations 20a or 20b of the gutter. No change, however, occurs in the cycle time of the ring counter, since it will be regularly receiving clock pulses at the 1 MHz frequency through one of the alternate circuits.

A change in the phase relationship between the drop formation by transducer 15 and drop charging by circuits 28, 29 and 30 is controlled through the application of a sampling signal. This signal may be produced at regular intervals or intermittently as desired or as necessary to properly maintain the phase relation. In FIG. 1, a sampling signal is regularly obtained at approximate 17 ms. intervals through a transformer 60 connected across an alternating current source 61 for example, 60 Hz. The output of the secondary winding is coupled to a pulse sharpening circuit 62 whose output is connected to latch circuit 63. The latch output serves as one input to AND 64 which has a second input applied from inverter 65 to which clock pulses are supplied on line 66. Thus, AND 64 will provide a high level output during the negative portion of the clock pulse when latch 63 is set. This output of AND 64 sets another latch 67 and is also used to reset latch 63. The setting of latch 67 serves as a conditioning input for AND 68 which has a second input clock pulses on line 66. Therefore, latch 67 will be set during the negative portion of the clock pulse and AND 68 will produce a high level output on line 69 during the next positive portion of the clock pulse. The high level output on line 69 is used to reset latch 67 and is supplied as a sampling signal for activating single shot 70 and conditioning NAND 53 as mentioned above.

Timing relationships between the clock pulses and initial sampling signal to provide an output sampling signal at AND 68 are shown in the waveforms of FIGS. 2 and 3. FIG. 2 illustrates the condition when the sampling signal initially occurs during the time the clock pulse is low (waveforms a and b). Latches 63 and 67 are both immediately set because of the high level output of inverter 65 (waveforms c, d, e, and f). Latch 67 being high conditions AND 68 so that a sampling signal occurs at the time the clock pulse level next becomeos high (waveforms f and g). Latch 67 is reset at the fall of the clock pulse.

In FIG. 3, if the sampling signal initially arrives during the time the clock output is high, latch 63 will be set but latch 67 cannot be since inverter 65 blocks AND 64 (waveforms a - e). Latch 67 is set when the clock output next becomes low, but AND 68 blocks any output until the next clock high level occurs (waveforms e - g).

When a sampling signal appears from AND 68 via line 69 at single shot 70, the single shot is triggered by the down stroke of the sampling signal to produce a short output pulse, the down stroke of which triggers a second single shot 71. The sequence of two single shots is used to insure that signals from single shot 71 occur at approximately the center portion of the down level of a clock pulse on line 33. (See waveforms d - g of FIG. 4). The output from single shot 71 serves as an extra input pulse for ring counter 31. When a signal appears at the output of single shot 71, it passes immediately through OR 35 to AND 36. If at that time, comparator 51 is at low level, then inverter 52 and AND 36 produce a signal which is supplied through OR 37 to ring counter 31 as an extra pulse to advance the ring counter one position in addition to the advances normally resulting from the clock pulses. Thus, ring counter 31 will reach its reset point one clock period sooner and produce an initiation signal to the character generator 28 and discard charging circuit 30 that much earlier. AND 36 was conditioned by the comparator circuit that indicated that discard drops were impacting at location 20a of gutter 20. With the earlier initiation signal supplied to circuits 28 and 30, a charge will also be applied earlier to droplets at charging electrodes 16. This will cause discard droplets to move to impact location 20b because the ramp voltage is started earlier and thus is higher at the instant of drop formation.

Assume this time that comparator 51 provides a high level output (waveforms, FIG. 5). Thus, AND 36 is blocked because of inverter 52 and no pulses reach counter 31 via that route (waveform d). Further, returning now to NAND 53, when the sampling pulse is high on line 55 and a clock pulse is at high level on line 54, NAND 53 will provide a low level output to block AND 38 (waveforms f and g). Thus, during the existence of a sampling signal on line 69 and 55, no clock pulses will pass AND 38. Although conditioned by the high level output from comparator 51, AND 39 will produce no output to OR 37 (waveform h). Thus, one of the clock pulses is barred from reaching ring counter 31. Upon the removal of the sampling pulse on line 69, however, clock pulses will resume and continue to increment the ring counter. Sampling pulse 69 is limited to being effective for only a single clock pulse. Since counter 31 has missed one of the regular clock pulses, it will produce an initiation signal to the circuits 28, 29, and 30, one clock period later. This results in driver circuit 29 applying a delayed charging signal to the droplets. As a result, discard droplets that are charged will now impinge on impact location 20a.

For operation of the phase compensation device, the ink jet nozzle is originally aimed so that with no signal applied to the change electrodes 16 in FIG. 1, the droplets will impact the gutter at compartment 20a with detection electrodes. Thereafter, the control system is energized and every droplet that is not needed for printing goes to the gutter and gets a charge depending upon the relationship of the break-off time with the ramp voltage produced from discard charging circuit 30. During the first moments of operation, the relationship is not defined, but there is a ramp for each discard droplet. The first droplets may thus hit the part of the gutter without the terminals, or the part with the terminals, but all droplets will impact the same part, omitting for the time being the case where the drops are split by the divider 46. If the droplets hit the part of the gutter without the electrodes, resistance will be high, as described above, so that upon the next sampling pulse, one clock pulse will be suppressed and ring counter 31 will be filled approximately one microsecond later. The ramp for the discard droplets will also start that much later in relation to the excitation voltage applied to vibrator transducer 15. The discard droplets will thus see a slightly lower potential at separation time and the point of impact on gutter 20 will shift towards the divider. This process will repeat itself with each sampling pulse until the droplets cross the divider. At that instance, the resistance between the electrodes in compartment 20a decreases and the output signal of comparator 51 becomes low. Thereafter, a pulse will be added to the pulses coming from the clock to the counter through AND 36, causing the ramp to start 1 microsecond sooner with respect to the instant of droplet break-off, thus the charge voltage on the discard droplets will be higher, causing the impact point to shift back toward compartment 20b without ink sensing terminals. This process of hunting will continue during the operation of the system.

If, for any reason, the instant of droplet break-off shifts with respect to the excitation voltage of the vibrator transducer 15, the system will add or subtract pulses at the rate of one per sampling period until the normal hunting mode is re-established.

It can readily be seen that when droplets hit the divider, such operation does not impair the functioning of the system. At worst, the extreme amplitude of hunting is plus or minus one-eight of the period of excitation of the vibrating transducer 15, an amount well within tolerable limits. It is to be noted also that although a frequency of 1 MHz has been disclosed, this frequency can be changed to provide a different frequency at vibrator transducer 15 or ring counter 31. The ring counter may require twice the frequency and the capacity to produce smaller increments of adjustment during operation.

Although a ring counter has been disclosed as the means to produce an initiation signal, a shift register may readily be substituted for the counter.

From the foregoing description, it will be apparent there is provided a system for effecting close control over the phase relations in an ink drop printer during printing operation. The printer is corrected as to the phase relation through the utilization of all discard drops which can occur during or at the end of a character or line. Further, adjustments can be made in smaller increments than heretofore possible and use highly accurate digital techniques.

While the invention has been particularly shown and described with reference to a preferred embodiment thereof, it will be understood by those skilled in the art that various chages in form and details may be made therein without departing from the spirit and scope of the invention.