GB2618807A

GB2618807A - Methods and apparatus for droplet deposition

Info

Publication number: GB2618807A
Application number: GB2207216.9A
Authority: GB
Inventors: Gouch Martin
Original assignee: Xaar Technology Ltd
Current assignee: Xaar Technology Ltd
Priority date: 2022-05-17
Filing date: 2022-05-17
Publication date: 2023-11-22
Also published as: WO2023222999A1; GB202207216D0

Abstract

The present invention relates to methods and apparatus for depositing droplets of fluid onto a medium, particularly in the field of printing. The apparatus 10 comprises an array of chambers 12, each of which can be actuated to release a droplet of fluid (an active firing chamber), actuated without depositing a droplet of fluid (an active non-firing chamber), or not actuated at all (an inactive non-firing chamber) in a given actuation cycle. The methods include receiving input data, the input data being split into a grid of pixels, each row of the grid corresponding to an actuation cycle and each column of the grid corresponding to a chamber in the array. The method further includes considering the history of deposition in the system and the impact thereof on a given chamber. Based on this assessment, that chamber is allocated to one of the three categories to address problems associated with non-uniform droplet deposition.

Description

METHODS AND APPARATUS FOR DROPLET DEPOSITION

Field of the Invention

The present invention relates to methods and apparatuses for depositing droplets of fluid onto a medium with reproducible results each time. For example, the apparatus may include a printhead.

Background to the Invention

Droplet deposition heads are now in widespread usage, whether in more traditional applications, such as inkjet printing, or in materials deposition applications, such as 3D printing and other rapid prototyping techniques, and the printing of raised patterns on surfaces, e.g. braille or decorative raised patterns. In such materials deposition applications, it may be desired to deposit a relatively large amount of fluid on a medium using droplet deposition heads. In some cases, the fluids may have novel chemical properties to adhere to new mediums and increase the functionality of the deposited material.

Recently, inkjet printheads have been developed that are capable of depositing inks and varnishes directly onto ceramic tiles, with high reliability and throughput. This allows the patterns on the files to be customized to a customer's exact specifications, as well as reducing the need for a full range of files to be kept in stock.

In still other applications, droplet deposition heads may be used to form elements such as colour filters in LCD or OLED displays, e.g. as used in flat-screen television manufacturing.

It will therefore be appreciated that droplet deposition heads continue to evolve and specialise so as to be suitable for new and/or increasingly challenging deposition applications. Nonetheless, while a great many developments have been made in the field of droplet deposition heads, there remains room for improvements in the field of droplet deposition heads. :30

As background to the present work, a mechanism by which droplets of fluid may be ejected from an array of fluid chambers is illustrated in Figure 1. This shows an array 10' of fluid chambers 12 forming part of a droplet deposition head, and, underneath, a simplified representation of the same array. The chambers are bounded on one side by a substrate 15.

Neighbouring fluid chambers 12 are separated by actuable side walls 14 formed of a piezoelectric material such as lead zirconate titanate (also known as PZT). The chamber 12 on each side of each piezoelectric wall 14 is coated internally with a metal layer that acts as an electrode, for applying a potential difference across the respective wall. That is to say, in this early example, within a given chamber 12 the metal electrode layer extends from the internal wall on one side of the chamber to the internal wall on the other side of the chamber (as better shown in other examples discussed herein, see e.g. Figure 2). This, however, is by no means the only electrode configuration that can be used. For example, each electrode extending from the internal wall on one side of the chamber to the internal wall on the other side of the chamber may be cut (e.g. by laser) along the centre of the fluid chamber, effectively dividing the electrode into two independently addressable electrodes.

If the same potential is applied to the electrodes on either side of a given wall, such that there is no potential difference across the wall, the wall remains stationary. On the other hand, if different potentials are applied to the electrodes on either side of a given wall, such that a potential difference exists across the wall, the wall moves by virtue of the reverse piezoelectric effect, which transforms potential difference into movement. The walls that move may be termed "active" walls, while the walls that remain stationary may be termed "non-active" walls.

Figure 1 illustrates a simplified representation of an array of chambers where two chambers experience a decrease in their volume due to the inward movement of their walls. As a consequence, the pressure in those two chambers increases (denoted by "+"), and the pressure in neighbouring chambers decreases (denoted by "-"). If the potential difference applied across the walls is high enough (e.g. to overcome surface tension effects), a droplet of fluid is forced out of the chamber that is under increased pressure ("+"), through a nozzle 16. Such chambers are referred to as "firing" chambers herein, because they eject ("fire") a droplet of fluid. Figure 1 also shows two chambers (on the far right of the diagram) that experience no change in volume because their walls remain stationary. These chambers are called "non-firing" chambers because they do not eject a droplet of fluid. It should be noted that the chambers denoted by "-" may be either firing chambers, since they are also capable of firing later in the same actuation cycle, or non-firing chambers if, in the same actuation cycle, their walls do not move in a way that causes ejection. As a particular example, where during an actuation cycle only one wall of a chamber moves, this may not cause ejection of a droplet, yet where an actuation cycle causes both walls to move, a droplet may be ejected. In another example, a droplet may be ejected when only a single wall of a chamber moves during a given actuation cycle, for example where a potential difference is applied across one wall which is large enough to cause a correspondingly large volume change for that chamber to eject a droplet.

The chambers 12 are formed as channels enclosed on one side by a cover member 17 that contacts the actuable walls; for each chamber in this example arrangement a nozzle 16 for fluid ejection is provided in this cover member 17. The cover member 17 may comprise a metal or ceramic cover plate, which provides structural support, and a thinner overlying nozzle plate, in which the nozzles are formed. Alternatively, a relatively thin nozzle plate might be used on its own as a cover member.

In the example of Figure 1 (and indeed throughout the present disclosure), each of the actuable piezoelectric walls 14 may comprise an upper half and a lower half, divided in a plane defined by the array direction (left to right in Figure 1) and the channel extension direction (into the page in Figure 1). The upper and lower halves of the piezoelectric walls may be poled in opposite directions perpendicular to the channel extension and array directions so that, when a potential difference is applied across the wall perpendicular to the array direction, the two halves deflect so as to bend towards one of the fluid chambers; the shape adopted by the deflected walls resembles a chevron and this may therefore be referred to as a "chevron mode" of actuation (once more this is shown in more detail in the different example shown in e.g. Figures 2A and 2B). Alternatively, each of the actuable piezoelectric walls may be poled in a unitary manner in a single direction (i.e. not as upper and lower halves poled in opposite directions) so that, when a potential difference is applied across the wall, the wall deflects in a "shear mode" of actuation. In some cases, the actuation of a given chamber may include a "draw-release" process. Here the volume of the chamber is first increased to draw additional fluid into the chamber, and the volume of the chamber is subsequently decreased (back to the initial volume, or sometimes to a yet smaller volume) to expel a droplet of fluid.

In any case, it is apparent that an actuation cycle is a time period in which each chamber in the array has been addressed and instructed to fire or not fire (in accordance with input printing information) as appropriate. This results in a series of dots of fluid being deposited on the medium at locations corresponding to firing chambers, the dots of deposited fluid forming a line on the medium. The line so formed may be unbroken in some cases, and in other cases, the line may be formed of smaller line segments separated by gaps corresponding to locations of non-firing chambers. Each time period in which each fluid chamber has been addressed and actuated if needed (and not actuated if not needed) is a single actuation cycle. Between adjacent cycles, the relative positioning of the array and the medium onto which droplets are deposited is typically changed so that the chambers are aligned with a new part of the medium and a new (once more, possibly broken) line can be formed on the medium in the next cycle.

A development of this basic idea is shown in Figure 2A. This Figure illustrates an array of fluid chambers 10 arranged to operate in a multi-cycle printing mode (specifically, a 'shared-wall' 3-cycle mode printing mode is shown), in which the chambers 10 are first divided into three groups: A, B, and C, (labelled in Figures 2A and 2B) operating in different printing sub-cycles (sub-cycles A to C, respectively, in Figure 2B). In a 3-cycle printing mode, chambers in different groups cannot eject simultaneously on the same sub-cycle. For example, chambers in group A can only fire in sub-cycle A, chambers in group B can only fire in sub-cycle B, and chambers in group C can only fire in sub-cycle C. There is therefore a time delay between firing the chambers of groups A, B and C, indicated by the droplets ejected from the A chambers being higher (were ejected longer ago) than those from the B chambers, which are themselves higher (were ejected longer ago) than those from the C chambers. A complete firing sequence of A -> B -> C defines one actuation cycle for a given line of printing to be deposited on a medium. Notwithstanding this asynchronous firing scheme between the three groups, the formation of a horizontal line on the medium can be achieved by staggering the nozzles/chambers, or inputting a waveform that changes the velocity of the droplets fired in different cycles.

For the next line a new set of chamber activations is instructed based on the desired printing pattern and the corresponding chambers from the A group fire, followed by the corresponding B, then C chambers. In other words, the full actuation cycle is formed by the aggregate of sub-cycles A, B and C. In this way, as in the previous description of a full actuation cycle, each chamber in the array is addressed once during the full actuation cycle, and correspondingly is able to eject a drop if this is appropriate in view of the input print data, and not eject a drop if this would be appropriate in view of the input print data.

In a little more detail, it can be seen that the input data (at the bottom of Figure 2A) indicates that chambers la, 3a, 4a, 5a, and 6a should eject a drop (input data is black), while chambers 2a and 7a should not eject drops (input data is white). As noted above, the array has been divided into A, B and C groups, meaning that the appropriate chambers can be actuated in each sub-cycle. Specifically in the example shown, chambers la and 4a are A chambers, so the A sub-cycle is updated to designate these chambers as firing chambers, while chamber 7a is designated as a non-firing chamber. Similarly, chamber 5a is a B chamber so the B sub-cycle is updated to designate this chamber as a firing chamber, while chamber 2a is designated as a non-firing chamber. Finally, chambers 3a and 6a are C chambers, so the C sub-cycle is updated to designate these chambers as firing chambers. There are no non-firing chambers in the C sub-cycle in this actuation cycle.

With the sub-cycles allocated as set out above, the printing occurs as shown in Figure 2B. Here the A sub-cycle is presented as being enacted first, followed by the B sub-cycle and finishing with the C sub-cycle (although this is arbitrary, either the sub-cycles could be re-labelled in any order, or the actual actuations could be re-ordered without affecting the overall picture). In this way, a complete actuation cycle is shown in Figure 2B. Note that for each sub-cycle, the chambers belonging to other groups are indicated with an X, highlighting that those chambers cannot fire on that sub-cycle. Chambers which are part of the sub-cycle, but which are designated as non-firing are indicated with a 0 -see e.g. 2a on the B sub-cycle and 7a on the A sub-cycle. The aggregate effect of the three printing sub-cycles (i.e. the complete actuation cycle) is to provide the pixel line illustrated at the bottom of Figure 2B. As would be expected this matches the input data.

As will be apparent from the disclosure of the invention in this document, the novel methods and apparatus are applicable to both single sub-cycle systems such as that in Figure 1, or to multi sub-cycle systems, such as the specific example of a 3-cycle system in Figures 2A and 2B.

An alternative arrangement of walls and electrodes is shown in Figs 2C and 2D. Here a similar process is followed, in that chambers 12 are separated by walls 14. Each side of each wall 14 is provided with an electrode (solid and dashed lines). However, instead of each chamber 12 having a single electrode as in Figure 2A, each wall (labelled 0 to 7) has two electrodes -a first electrode la, 2a, etc., shown as a solid line located on the "first direction" side of the wall (left in Figure 2C) and a second electrode 1b, 2b, etc., shown as a dashed line located on the "second direction" side of the wall (right in Figure 20). This arrangement opens up more possibilities for controlling the motions of the walls 14.

This is achieved as shown in Figure 6. In addition to a set of laser cuts X to separate the electrodes of one chamber 12 from the electrodes of neighbouring chambers 12, a further set of laser cuts Y is implemented to cut the electrode for each chamber 12 into two electrodes. Each of the two electrodes for a given chamber 12 are therefore deposited on each of the two walls 14 defining the chamber 12 and separating the chamber 12 from its neighbouring chambers 12. As shown the chambers 12 are once again each provided with an aperture 16 (sometimes referred to as a nozzle or outlet) for ejection of drops of fluid.

This arrangement can be used in the single cycle and 3-cycle modes discussed above, by actuating the walls in a corresponding manner. It will be appreciated that the systems described herein may operate with a single cycle mode, a 2-cycle mode, a 3-cycle mode, or indeed any higher number of sub-cycles forming a single actuation cycle. In cases where the chambers are provided with two separate electrodes, this can provide more flexibility in controlling the operation of the device (i.e. in controlling the wall motions), albeit at the cost of greater complexity in the control electronics.

As used herein, for example in the description below, "image line" may be used interchangeably with "pixel line", and image rows or columns instead of lines or pixels On this situation, the word pixel takes on the meaning of an individual element in an image). For example, the pixels of an image which are printed in a single actuation cycle (optionally comprising a plurality of sub-cycles) may be referred to as a row of pixels. Pixels of an image which are printed by the same chamber at different times (i.e. in different actuation cycles) may be referred to as a column of pixels. Collectively the full set of rows is equivalent to the full set of columns, and defines the full image area, A problem exists in the general class of printing systems discussed above. Under some conditions in which a given chamber is actuated infrequently, the droplet deposition procedure results is the generation and release of drop that has non-uniform characteristics. The release of the non-uniform drop leads to various problems, including: * Faster First Drop: this drop impacts the medium sooner than the surrounding drops, appearing displaced on the medium in the direction of movement of the medium relative to the array of chambers.

* Smaller Drop than the surrounding drops: this causes inaccuracy when calculating ink consumption by simply counting the number of drops fired.

* Drop Election is not Clean: this means that small droplets or mist may be generated and get deposited on surfaces such as the nozzle guard, causing a pool of ink which can eventually fall from the place it has collected and drop onto the medium below and creating an unwanted artifact of an uncontrolled splash of ink on the medium which may not cure or dry correctly. To make matters worse, different colours/inks may drip at different rates.

Various solutions have been attempted. As noted above the problem is linked to infrequent chamber actuation. Consequently, to reduce the effect of the problems above, a waveform to almost fire a drop without actually firing the drop may supplied to the printhead in cases where no droplet is to be ejected. That is to say that in each actuation cycle, each chamber is supplied with a control signal to either eject a droplet, or to move the walls without ejecting a droplet. This 'white space waveform' causes the chambers to actuate their walls as if a drop ejection was intended but without actually ejecting the drop. In other words, the white space waveform imparts resonances in the chamber and its neighbouring chambers just as though a drop had been fired without giving the correct conditions to eject a drop.

In one example, the white space waveform may reduce the time duration of the wall actuation such that not enough energy is produced to eject a drop or, alternatively, it may reduce the magnitude of the wall actuation such that not enough energy is produced to eject a drop.

In another example, when using "draw-release" white space waveform, the "release" part may be omitted such that no ink will be ejected. In this example, the walls simply return to the nominal position and the ink in the chamber is disturbed and residual energy is left in the ink in the chamber. Alternatively, "draw-release" white space waveform may have the "release" action shorter than the ejection waveform such that it does not produce enough energy to eject the ink through the nozzle.

Whichever method is used, the 'white space waveform' actuates chamber walls that otherwise would be stationary causing the following issues: * Significant increase of heat produced by the printhead because, typically, a printed image will have a significant number of chambers not printing dots.

* Significant increase on the printhead's wear because chambers are actively activated, and walls actuated despite no drop ejection being needed. This is particularly true if the head is left running printing nothing whilst in standby.

There is therefore a desire to overcome the above limitations of the various printing systems and achieve a more energy-efficient manner of printing, that is also able to eject single drops when required to do so, and to achieve simultaneous ejections of drops with a fine-grained resolution.

Summary of the Invention

Aspects of the present invention are set out in the appended independent claims, while particular embodiments of the invention are set out in the appended dependent claims.

Disclosed herein is method for depositing droplets of fluid onto a medium utilising a droplet deposition head, the droplet deposition head comprising: an array of fluid chambers separated by interspersed walls formed of a piezoelectric material, each fluid chamber communicating with an aperture for the release of droplets of fluid, each of said walls separating two neighbouring fluid chambers, and each fluid chamber being defined by a first wall in a first direction relative to the fluid chamber, and a second wall in a second direction relative to the fluid chamber, the second direction being opposite to the first direction; wherein each chamber has at least one electrode disposed across surfaces of the first and second walls internal to the chamber, the electrodes internal to each chamber being providable with actuation voltages independently of actuation voltages supplied to the electrodes of other chambers; wherein each of said walls of each chamber is actuable to change the volume of the chamber from a neutral configuration such that, in response to the application of difference of potential between a first electrode and a second electrode located on opposed sides of a wall separating a first chamber and a second chamber adjacent to the first chamber, the wall moves into a deformed configuration relative to the position of the wall in the neutral configuration, the method comprising, the steps of: receiving input data, the input data being split into a grid of pixels, each row of the grid corresponding to an actuation cycle and each column of the grid corresponding to a chamber in the array; and for each actuation cycle, assigning, based on said input data, all the fluid chambers within said array as either firing chambers for depositing droplets in the actuation cycle or non-firing chambers which do not deposit droplets in the actuation cycle, and wherein each non-firing chamber is further assigned as either an active non-firing chamber or an inactive non-firing chamber wherein each non-firing chambers is assigned as an active non-firing chamber or an inactive non-firing chamber based both on information on preceding actuation cycles and on information on the next actuation cycle; and actuating at least one of the walls of each active non-firing chamber without depositing a droplet of fluid onto the medium and actuating at least one of the walls of each firing chamber such that each firing chamber releases at least one droplet the resulting droplets forming dots disposed on a line on the medium, the dots being separated on the line by gaps corresponding to the non-firing chambers.

Note that the array may form part of a larger printhead in the sense that the array may comprise only a portion of the total number of chambers. In other examples, of course, the array may include each chamber forming part of the overall printing apparatus. In some examples, there may be multiple rows of chambers. In such cases, the novel features discussed herein may be applied to each row, or only to some rows, or indeed only to portions of some rows. As noted above, the system may include a single electrode per chamber or the two walls may have electrically separate electrodes located thereon. In either case, the actuation of a particular chamber (i.e. if it is assigned as a firing chamber or a n active non-firing chamber) may involve one or both walls being actuated.

As noted above, the walls are actuable by controlling the potential difference across each wall, by virtue of the reverse piezoelectric effect. The exact form of the motion depends on specific details of the materials, for example, some piezoelectric materials move towards the higher voltage, others move towards the lower voltage. While this naturally affects the format of the control signals to achieve a desired effect, the general principles are that, when a first difference of potential is applied (a first side of a wall is at a higher potential than a second side), the wall moves in a first direction, when applying a second difference of potential (different from the first difference of potential; the first side of the wall is at a lower potential than the second side) the wall moves in a second direction. This can be used, in conjunction with knowledge about the behaviour of the material from which the walls are formed, to control the walls to cause the volume of chambers to be increased or decreased as needed to eject a droplet, consistently with the various ejection protocols discussed above. It will be apparent for example that the relevant effect is the use of wall motions to cause chambers to increase or decrease in volume to cause the drop ejection, by moving the walls of a chamber respectively away from or towards the centre of the chamber. For the avoidance of doubt, applying no potential difference across a wall typically leads to the wall remaining in (or reverting to) a neutral, undeformed position.

The input data may be the full bitmap file (discussed in more detail below), or it may just be the subset required to implement the method. For example, the method disclosed above includes considering the activity of the same chamber in the preceding cycles. In effect, the inventors have identified that there are certain criteria which correlate strongly with the production of non-uniform droplets. In particular, the actuation history of a given chamber (i.e. recent history of when that chamber was last actuated) and the actuation history of neighbouring chambers (i.e. how recently the chambers on either side of the chamber were actuated) affects the status of that chamber.

This is because a discovery underlying the present invention is that long periods of inactivity of a given chamber can lead to the problem of non-uniform droplet deposition because a chamber which is inactive for a period of time receives less energy and the fluid does not flow correctly out of the chamber through the nozzle when actuated as a consequence. It cannot be assumed that walls of non-firing chambers move during an actuation cycle, and consequently chambers adjoining those non-moving walls are liable to suffer from this problem if they do not move for a sequence of one or more consecutive actuation cycles.

The terms "awake" and "asleep" will be used herein to differentiate between a given chamber being, respectively, able to deposit a droplet in the regular manner, and being likely to deposit a non-uniform droplet. The terms "wake up", "fall asleep" etc. shall be construed accordingly. Therefore, an assessment is made as to whether the chamber itself has been actuated sufficiently recently to avoid this problem. Moreover, the inventors have realised that actuation of adjacent (i.e. neighbouring) chambers can also act to avoid these problems, and a separate (albeit related) assessment is made as to whether either of the adjacent chambers has been actuated sufficiently recently that the chamber in question is awake.

In other words, in an actuation cycle, the chambers may be assigned one of three designations, each having a corresponding action. In cases where the chamber is due to fire (deposit a droplet) in that actuation cycle, the chamber is assigned as a firing chamber and a normal droplet deposition signal is sent to the relevant chamber walls to cause ejection of a droplet, for example in the manner discussed above. In cases where no droplet is intended to be deposited, the chamber is assigned as a non-firing chamber and no droplet is ejected. However, based on the input data of surrounding deposition patterns in preceding and successive rows, and including adjacent chamber activity), the chamber may be assigned as either an active non-firing chamber, in which the walls are supplied with a signal which causes a non-deposition actuation of the chamber walls (the walls move but no droplets are ejected/deposited) or as an inactive non-firing chamber, where the walls do not move at all, and no droplet is ejected/deposited.

For the avoidance of doubt, as used herein, where a given chamber is described as having walls which do not move at all, this means specifically that due to the consideration of the status of that individual given chamber, no actuation of the walls is provided. This specifically does not preclude the walls of that given chamber moving during the same actuation cycle.

For example if the consideration of a chamber adjacent to the given chamber causes one or more walls of that adjacent chamber to move, then the walls of the given chamber will also move in that actuation cycle. Note that in the following text unless the situation specifically requires only one wall or specifically requires both walls to move, references to moving the wall or the walls of a chamber shall be interpreted as covering either situation. :30

One method of determining whether a given chamber is awake or asleep is to measure or calculate (e.g. by counting cycles or sub-cycles elapsed and multiplying by the cycle frequency) the time which has elapsed since that chamber was last actuated and compare this time to a threshold. Similarly, the elapsed time since either adjacent chamber last actuated can be calculated or measured and compared to a different threshold. If both (or in some cases either) of the thresholds is exceeded, then the chamber is determined to be asleep, and remedial action is taken. As noted above, the remedial action is to wake the chamber up in an appropriate manner. Usually this takes the form of actuating one or both of the chamber walls enough to cause the chamber to wake up, without ejecting a drop. Note the assumption here is that chambers only need waking up towards the end of a string of actuation cycles in which that chamber has not been instructed to deposit a droplet, since if the chamber had been actuated recently, then it would still be awake and would not need waking up.

Typically, action is taken to wake the chamber up one actuation cycle prior to the next cycle in which the chamber is required to deposit a droplet. This ensures that a uniform droplet can be ejected on demand even if a given chamber has been inactive for a long time. Note that it is beneficial to introduce the non-deposition actuation into a regular actuation cycle to ensure that the chamber actuations remain synchronised. It is apparent that the specific design of the printhead (i.e. chamber dimensions, nozzle size, fluid characteristics, etc.) all affect the time taken for the chamber to transition from awake to asleep. Therefore, the thresholds set out above may be changed depending on the fluid in use and the characteristics of the printhead As noted above, the non-deposition actuation (i.e. the signal supplied to the active non-firing chambers) is supplied during the regular actuation cycle, and for a 3-cycle shared wall system it is supplied during the sub-cycle corresponding to the group to which the relevant chamber belongs. Given that the process involves actuation in a manner which is synchronised to the actuation cycles (and/or the actuation sub-cycles), it is often appropriate to measure the time period in the natural time scale of the system -the period of a single actuation cycle (or actuation sub-cycle). In other words, it is possible to phrase the thresholds set out above in terms of whether the chamber in question has actuated in the preceding M actuation cycles (or has occurred in the past time period of length M') or whether either adjacent chamber has actuated in the preceding N actuation cycles (or has occurred in the past time period of length N'), where typically N<M (or equivalently N'<M), since the effect of actuating adjacent chambers is less strong than directly actuating the chamber itself. This has the advantage that it is a simple threshold, easily determined empirically for a given set up, and inherently results in a non-deposition actuation schedule which is synchronised with the actuations cycles.

In any case, the threshold (whether couched in terms of number of cycles or elapsed time) will depend on: print frequency, fluid type, chamber design and so forth. Therefore, while certain numbers of cycles are provided in this document, the frequency, fluid type, head design being used etc. may change these specific values.

By forming a protocol to account both for the previous actuations of a given chamber and the actuations of the chambers adjacent to the chamber, it is possible to greatly reduce or even eliminate the production of non-uniform droplets, by using the non-deposition actuation to keep chambers awake (or wake them up) so that they are ready to eject droplets (i.e. awake) an advance of their being required to deposit droplets, thereby solving the problem of non-uniform droplet deposition.

Note that since the method above is applicable to each chamber in the array, and can be applied to each actuation cycle, the entire image (the entirety of the input data) can be processed in this way. Because the method means that non-deposition actuation signals may be applied only when needed, and not e.g. all the time as in earlier examples, it is apparent that there are several advantages to this approach, including significant energy savings, a reduction in printhead heating, and reduced wear of the printhead, leading to increased printhead lifetimes.

The input print data may include just a predetermined number of preceding lines/actuation cycles (sufficiently many preceding cycles to determine reliably whether the chamber is awake), the current line/actuation cycle, and the subsequent cycle, allowing the protocols set out below to determine from this subset of the data whether there is a need to awaken any given chamber or not. As noted above, in some cases, there may be less information received at the input stage than this, or more, depending on the specific method used. For example, in some cases, the whole image may be pre-processed to encode each pixel accordingly (using three states, and therefore a minimum of two bits per pixel), but then only a single line need be sent at a time, thereby trading off a more costly encoding scheme for the data against a reduced buffer requirement on the printhead.

Optionally, during an actuation cycle each non-firing chamber is assigned as an active non-firing chamber or an inactive non-firing chamber based at least in part on: locations of firing chambers or active non-firing chambers in one or more preceding actuation cycles; and/or locations of firing chambers in a successive actuation cycle. Specifically, the method may include, during an actuation cycle, each non-firing chamber being assigned as an active non-firing chamber, or an inactive non-firing chamber based at least in part on whether that chamber was a firing chamber or an active non-firing chamber in either of the two preceding actuation cycles. Additionally or alternatively, the method may include, during an actuation cycle, each non-firing chamber being assigned as an active non-firing chamber, or an inactive non-firing chamber based at least in part on whether that chamber or either directly adjacent chamber was a firing chamber or an active non-firing chamber in the preceding actuation cycles. Optionally, for a given chamber assigned as a non-firing chamber, whether that chamber is assigned as an active non-firing chamber or an inactive non-firing chamber depends at least in part on whether adjacent chambers to the given chamber in the array have been assigned as firing chambers, active non-firing chambers, or inactive non-firing chambers for that actuation cycle.

As noted above, actuations in preceding actuation cycles can wake up a chamber to inhibit unwanted non-uniform droplet deposition. In some cases, the presence of a non-deposition actuation (i.e. an active non-firing chamber) in preceding rows and adjacent columns can be sufficient to render a chamber awake. Note that in this case, the threshold time or threshold number of preceding rows above which the chamber is deemed to be asleep may be lower than (and is usually at most no higher than) the threshold for other manners of assessing whether the chamber is awake or asleep. In other words, if the chamber in question can be deemed to be awake if a non-deposition actuation has occurred in an adjacent chamber within the preceding P actuation cycles (or has occurred in the past time period of length P') and the thresholds M and N discussed above have the following relationship with P, the newly introduced threshold: P*1<M, or equivalently P'1\1'<M'.

Finally, it should be noted that the decision as to whether to assign the chamber as an active or an inactive non-firing chamber can be based in part on the next row to be printed. This is because if a chamber is deemed to be asleep, this is not inherently a problem if that chamber would not need to print in the next actuation cycle. Instead, the chamber may be left asleep with the associated saving in energy use and reduction in printhead wear. In order to address the problem of non-uniform droplet deposition, each chamber may be woken up only in the actuation cycle preceding an actuation cycle in which it is next due to deposit a droplet, thereby minimising energy usage and printhead wear, and focussing any non-deposition actuations to the situations in which they are needed.

Note that, as used herein, the terms "preceding actuation cycle" and "successive actuation cycle" mean the actuation cycle directly preceding or directly following the current actuation cycle, respectively. That is if the current actuation cycle is cycle n, then the preceding actuation cycle is cycle n-1 and the successive actuation cycle is cycle n+1. VVhere the term "preceding {two, three, etc.} actuation cycle" is used, this means the two, three, etc. consecutive actuation cycles directly preceding the current cycle. That is to say that if the current actuation cycle is cycle n, then the preceding two actuation cycles is the set of actuation cycles {n-2, n-1}, and the preceding three actuation cycles is the set of actuation cycles {n-3, n-2, n-1}. This can of course be generalised to any number of preceding cycles.

In particular, the inventors have found suitable thresholds for commonly used systems to be that a given chamber is likely to be asleep if that chamber has not been actuated for the preceding four actuation cycles AND if there has been no actuation of an adjacent chamber in the preceding three actuation cycles (in the language above, the thresholds are M = 3, N = 2, meaning that if these thresholds are exceeded, the chamber is likely to eject a nonuniform droplet the next time it is actuated, unless action is taken to wake the chamber up.

There are various ways in which the non-deposition actuation may be implemented to avoid the undesirable effects of non-uniform droplet deposition. These are set out below in more detail, but broadly the purpose of these specific examples is that they each reduce both the likelihood of non-uniform droplet deposition and the energy usage of, and wear on, the system. These improvements are seen when the presently disclosed systems and methods are compared with those which either never actuate the walls unless the chamber is due to print and those systems which always actuate the walls, providing a non-deposition actuation when no droplet is due to be printed, and providing a full actuation to deposit droplets when this is the desired outcome.

In some examples, therefore, for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if either: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid. This example provides a manner of ensuring that any chamber which is due to print in the next actuation cycle will be awake when that cycle begins because it reliably detects, using a small number of simple rules, situations in which the chamber is liable to be asleep. However, the simplicity of the rules comes at a cost and there is a trade off in that this example sometimes causes a non-deposition actuation of chambers which do not need such an actuation.

A further example is one in which, for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; or either of the two adjacent chambers was or will be assigned as a firing chamber or an active non-firing chamber in the preceding or current actuation cycle; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid. This example avoids some of the unneeded non-deposition actuations which can occur in the previous example, by supressing non-deposition actuations if there has been (or will be) an actuation in at least one of the adjacent chambers in the same line, as this will lead to the chamber under consideration being awake.

In another further example, for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; or either of the two adjacent chambers was or will be assigned as a firing chamber or an active non-firing chamber in the preceding or current actuation cycle; or either of the adjacent chambers is to be assigned as a firing chamber in the successive actuation cycle; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid. This example extends the method in the previous example to consider whether, in the next actuation cycle, the adjacent chambers will have been actuated. This would have the effect of waking up the chamber being considered, and consequently, any non-deposition actuation which would otherwise have occurred can be suppressed, thereby saving energy.

As noted above, the methods set out herein can be applied to a multi sub-cycle actuation protocol. In some cases, this may be a three-cycle shared wall actuation, as shown schematically in Figures 2A and 2B. In other cases, there may be as few as two sub-cycles.

It will be appreciated that more than two or even more than three sub-cycles can be used in some examples, using the same general principles. In examples in the present disclosure which make use of the three-cycle shared wall printing mode, the chambers of the array are is arranged in three sub-arrays, such that any three adjacent chambers includes at least one chamber which is a member of each sub-array, and wherein the chambers of the array are arranged in a repeating pattern of chambers belonging to each of three sub-arrays arranged in the same order to from the repeating pattern; wherein each actuation cycle includes the actuation of three actuation sub-cycles, each actuation sub-cycle including: the assignment of the chambers of a different one of the sub-arrays as firing chamber, active non-firing chambers, or inactive non-firing chambers; the actuation of (at least one of) the walls of the firing chambers of that sub-array to release at least one droplet; the actuation of (at least one of) the walls of the active non-firing chambers without releasing a droplet; and wherein each actuation cycle includes the actuation of the three actuation sub-cycles in order of a first actuation sub-cycle, followed by a second actuation sub-cycle, followed by a third actuation sub-cycle. Arranging the system/method in this way allows the use of multiple sub-cycles and advantageously leads to the following two further examples.

The first further example in one in which, for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; or either of the two adjacent chambers was or will be assigned as a firing chamber or an active non-firing chamber in the preceding or current actuation cycle; or that chamber is part of a second or third sub-array, configured to be actuatable as part of the second or third actuation sub-cycles respectively and the adjacent chamber forming part of the preceding sub-cycle will be assigned as a firing chamber or an active non-firing chamber in the successive actuation cycle; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid, including where that chamber is part of a first sub-array, configured to be actuable as part of the first actuation sub-cycle. In this example, the method takes account of whether the adjacent chambers in the next actuation cycle will be part of an actuation sub-cycle which is enacted before the actuation sub-cycle of the chamber being considered. If either adjacent chamber is indeed part of an earlier sub-cycle and that chamber is due to be actuated, then the chamber being considered will be awake by the time its actuation sub-cycle occurs. This means that if the chamber in question is due to print, it is unlikely to suffer from non-uniform droplet deposition issues.

The second further example in one in which, for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; or either of the two adjacent chambers was or will be assigned as a firing chamber or an active non-firing chamber in the preceding or current actuation cycle; or that chamber is part of a second or third sub-array, configured to be actuatable as part of the second or third actuation sub-cycles respectively and the adjacent chamber forming part of the preceding sub-cycle will be assigned as a firing chamber or an active non-firing chamber in the successive actuation cycle; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; if that chamber is assigned as a non-firing chamber and will be assigned as a firing chamber in the successive actuation cycle and neither that chamber was assigned as a firing chamber or an active non-firing chamber nor were or will either of the adjacent chambers assigned as firing chambers or active non-firing chambers in the preceding or current actuation cycle and that chamber is part of the second or third sub-array, configured to be actuable as part of the second or third actuation sub-cycles respectively and the adjacent chamber forming part of the preceding sub-cycle will not be assigned as a firing chamber or an active non-firing chamber in the successive actuation cycle; then the adjacent chamber forming part of the preceding sub-cycle is assigned as an active non-firing chamber in the successive actuation cycle; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid, including where that chamber is part of a first sub-array, configured to be actuable as part of the first actuation sub-cycle. In this example, the non-deposition actuation signal is deferred to the last possible time, if it is needed. As noted, in some cases, this allows the non-deposition actuation signal not to occur in the current actuation cycle but in a preceding actuation sub-cycle in the next actuation cycle in some cases. This in turn means that chambers woken up in this manner remain awake for one actuation cycle longer than if they had been woken up in the current actuation cycle, thereby improving efficiency yet further by reducing the need for subsequent non-deposition actuations.

In some cases, the assessment of whether to categorise non-firing chambers as active or inactive is made based on a binary pixel map in the vicinity of the chamber in question 0.e.

the position of the chamber in the array, and the actuation cycle relative to preceding and successive actuation cycles. For the examples set out above in which the thresholds of M=3 and N=2 are used, the input data may be processed to assign, for each actuation cycle, each chamber in the array as a firing chamber, an active non-firing chamber or an inactive non-firing chamber based on whether that chamber has been assigned as a firing chamber or an active non-firing chamber in the preceding two actuation cycles or whether either adjacent chamber is or will be assigned as a firing chamber or an active non-firing chamber in the current or preceding actuation cycle. In particular, for each pixel in the grid, the assignment as a firing chamber, an active non-firing chamber or an inactive non-firing chamber may be performed using a pixel-wise AND operation on a bitmap version of the input data and a vicinity mask formed by considering the previous two actuation cycles of each chamber and the preceding and current actuation cycles of each adjacent chamber.

That is to say that pixels in the vicinity mask have a value of "1" in an "active region", i.e. if they are up to two rows behind the current pixel, or adjacent and in the current or directly preceding row, 0 otherwise. Similarly, the bitmap representation has a 1 if that pixel is to be printed, and a 0 if not. This means that the result of the AND operation is to output a 1 anywhere in the grid which is both a pixel to be printed and exists in the active region and is 0 otherwise. It is then easy to identify that the pixel under consideration is awake if there are any 1s in the resulting grid.

As noted above, the methods set out herein can be applied to a multi sub-cycle actuation protocol, for example a three sub-cycle actuation cycle. In such cases, the method may be provided as follows. The chambers of the array are arranged in three sub-arrays, such that any three adjacent chambers includes at least one chamber which is a member of each sub-array, and wherein the chambers of the array are arranged in a repeating pattern of chambers belonging to each of three sub-arrays arranged in the same order to from the repeating pattern; wherein each actuation cycle includes the actuation of three actuation sub-cycles, each actuation sub-cycle including: the assignment of the chambers of a different one of the sub-arrays as firing chamber, active non-firing chambers, or inactive non-firing chambers; the actuation of at least one of the walls of each firing chamber of that sub-array to release at least one droplet; the actuation of at least one of the walls of each active non-firing chambers without releasing a droplet; and wherein each actuation cycle includes the actuation of the three actuation sub-cycles in order of a first actuation sub-cycle, followed by a second actuation sub-cycle, followed by a third actuation sub-cycle; and where the pixel is in the second or third actuation cycle, the vicinity mask includes a consideration of the adjacent chamber in the preceding actuation sub-cycle in the next actuation cycle. This arrangement allows a non-deposition actuation to be suppressed in cases where the chamber in question would be woken up by an adjacent chamber in the next actuation cycle, when the chamber in question is due to print in the next actuation cycle.

An optional development of this idea is to provide the bitmap version of the input data formed by a pixel-wise OR combination of a first bitmap representation of whether a chamber was assigned as a firing chamber in preceding actuation cycles with a second bitmap representation of whether a chamber was assigned as an active non-firing chamber in preceding actuation cycles. This allows the assessment to be made based on whether the chambers in the active region were either firing chambers or active non-firing chambers, which in turn impacts whether the chamber in question is currently awake.

The bitmap version of the input data and/or the vicinity mask is a 3 x 4 grid with the dimension having extent 3 representing the current pixel and both adjacent pixels and the dimension having extent 4 representing the current actuation cycle, the next actuation cycle and two preceding actuation cycles. This limits the scope of the analysis which is performed for each pixel to the region which typically affects the status (awake or asleep) of the pixel under consideration, since only adjacent chambers and a certain number of preceding cycles can affect the status of the chamber. As noted above, the thresholds may change depending on various parameters, in which case the size of the grid will change accordingly.

An alternative method includes processing the entire image to be printed as a whole. Taking this holistic view allows the use of various optimisation algorithms to designate active non-firing chambers in the most efficient manner possible, i.e. that which results in the fewest active non-firings while ensuring that any chamber due to print in an actuation cycle is awake at the time it is due to print. This can ensure that the smallest number of non-firing actuations is implemented across the entire printing process.

Optionally, prior to a first actuation cycle, an initiation cycle is executed, the initiation cycle comprising actuating at least one of the walls of one or more chambers without depositing a droplet of fluid onto the medium. This procedure may be applied to the first two, three, etc. actuation cycles (the number being dependent on the size of the thresholds M and N), in order that any chamber which is due to print, but which has not yet been woken up by any preceding actuation cycle can be woken up in time to print, if needed. In some cases, the initiation cycle includes actuating at least one of the walls of all of the chambers in the array without depositing a droplet of fluid onto the medium. This initiates the array with all chambers being awake. In some cases, the initiation cycle may actuate at least one of the walls of every other chamber or of every third chamber in the array without depositing a droplet of fluid onto the medium. Optionally, the initiation cycle includes actuating at least one of the walls of each chamber within a subset of the chambers, wherein the subset comprises only those chambers which are to be assigned as firing chambers in the first three actuation cycles.

Also disclosed herein is a droplet deposition apparatus comprising one or more fluid chambers, the apparatus being configured to carry out any of the methods set out above.

The droplet deposition apparatus may further comprise a computer in data communication with said one or more fluid chambers, wherein the computer is programmed to carry out the assigning step based on the input data. Optionally, the computer is further programmed to send instructions to the one or more fluid chambers, so as to cause them to carry out the steps of actuating the walls in accordance with the outcome of the assigning step.

Also disclosed herein is a computer program comprising instructions to cause the droplet deposition apparatus to execute any of the methods set out above.

Each variant of the method set out above may be carried out on any of the variants of the apparatus described above. This brings the corresponding advantages of that specific method to the apparatus.

Brief Description of the Drawings

Embodiments of the invention will now be described, by way of example only, and with reference to the drawings in which: Figure 1 illustrates an array of fluid chambers forming part of a droplet deposition head, with some walls of the chambers having been actuated, and beneath, a simplified representation of the same array with the same actuated walls; Figure 2A shows an end view of an array arranged to operate in a 3-cycle mode; Figure 2B shows a schematic of the printing process for a single actuation cycle of the array of Figure 2A; Figure 2C shows another example of an array in which each chamber has two independent electrodes for use in independently controlling the two chamber walls; Figure 2D shows an example of a laser cutting pattern to form the electrodes shown in Figure 2C; Figure 3A illustrates a first criterion for determining whether a given chamber is asleep or awake; Figure 3B illustrates a second criterion for determining whether a given chamber is asleep or awake; Figure 4 is a flow chart illustrating the initial steps of several of the methods disclosed herein; Figure 5A is a flow chart illustrating a first method for determining the status of a chamber; Figure 5B illustrates the relationship of the current pixel to pixels in its vicinity as enacted in the method shown in the flow charts of Figures 4 and 5A; Figure 6A is a flow chart illustrating a second method for determining the status of a chamber; Figure 6B illustrates the relationship of the current pixel to pixels in its vicinity as enacted in the method shown in the flow charts of Figures 4 and 6A; Figure 7A is a flow chart illustrating a third method for determining the status of a chamber; Figure 7B illustrates the relationship of the current pixel to pixels in its vicinity as enacted in the method shown in the flow charts of Figures 4 and 7A; Figure 8A is a flow chart illustrating a fourth method for determining the status of a chamber; Figure 8B illustrates a relationship of the current pixel to pixels in its vicinity as enacted in the method shown in the flow charts of Figures 4 and 8A; Figure 8C illustrates a further relationship of the current pixel to pixels in its vicinity as enacted in the method shown in the flow charts of Figures 4 and 8A; Figure 9A is a flow chart illustrating a fifth method for determining the status of a chamber; Figure 9B illustrates a relationship of the current pixel to pixels in its vicinity as enacted in the method shown in the flow charts of Figures 4 and 9A; Figure 9C illustrates a further relationship of the current pixel to pixels in its vicinity as enacted in the method shown in the flow charts of Figures 4 and 9A; Figure 10A illustrates a sixth method for determining the status of a chamber; and Figure 10B is a detailed view of method step 1028 in Figure 10A.

In the figures, like elements are indicated by like reference numerals throughout.

Detailed Description of Preferred Embodiments

The present embodiments represent the best ways known to the Applicant of putting the invention into practice. However, they are not the only ways in which this can be achieved.

The presently described embodiments relate to a printing mode in which non-firing chambers are selectively supplied with non-deposition actuation signals to reduce or eliminate the deposition of non-uniform droplets.

In particular, the inventors realised that, while periods of inactivity lead to non-uniform droplet deposition, to avoid a significant increase in heat and wear of the printhead (i.e. the array of fluid chambers), it is not necessary to almost fire a drop on all no fire conditions (where no drop is scheduled to be deposited). Instead it is sufficient to almost fire a drop in the no fire print location just before the drop required to be fired if there are no fire assignments in certain key places in the history of the printhead. This may vary with specific printhead design, printhead deposition frequency, and with the physical properties of the ink, for example. However, as a specific example, for certain printheads in use today, the inventors have found that the likelihood of non-uniform droplet deposition correlates strongly with situations in which there has been no actuation in the same chamber for the previous four actuation cycles; OR in the two directly adjacent chambers for the previous three actuation cycles.

These conditions are represented schematically in Figures 3A and 3B respectively. Both Figures show a grid of pixels for printing. Each column is associated with a chamber in the array and each row is associated with an actuation cycle and pixels and chambers/actuation cycles are used interchangeably with this connection being implied throughout. This view will be used consistently throughout this document. In each Figure, a pixel which is considered to make the assessment is denoted with 23. Pixels which are not part of the consideration are denoted with 20. The pixel currently under consideration (often referred to as the current pixel) is denoted 21, and in each case the above conditions have not been met, so the pixel under consideration has been illustrated as depositing a non-uniform droplet 22.

In Figure 3A, whether or not there has been any actuation in the same chamber for the previous four actuation cycles. Therefore the four central pixels 23 above the current pixel 21 are considered. Since no actuation has occurred in any of these, the first criterion is failed -that is to say that an analysis under the first criterion does not result in the current pixel 21 being deemed awake. Note that in this analysis, neither the pixels of the left column 20 nor those of the right column 20 play a part.

In Figure 3B, the analysis considers whether an actuation has occurred in the two directly adjacent chambers for the previous three actuation cycles. This means that the pixels of the left and right columns for the three preceding rows 23 are considered, while the middle column 20 is not considered at all. The left and right pixels of the lowest row 20 are also not considered. This is because they have not necessarily happened yet, and there is no guarantee that (even if they are due to print on the current cycle) they will actuate before the current pixel 21 does. This situation is addressed in more detail below, however. Finally the left and right pixels of the upper row 20 are also not considered in this analysis because even if they did represent an actuation of their corresponding chamber, this would have occurred too long ago, and the chamber corresponding to the current pixel 21 (sometimes referred to as the current chamber, for conciseness) would have fallen asleep again. Since no actuation has occurred in any of the relevant pixels 23, the chamber corresponding to the current pixel fails the second criterion. That is to say that an analysis under the second criterion does not result in the current pixel 21 being deemed awake. Note that if the situation is as shown collectively in both of Figures 3A and 38, then both criteria will have been failed and the chamber corresponding to the current pixel 21 will be deemed to be asleep.

The solution to this problem set out herein is to introduce the non-deposition actuation at times which are calculated to avoid the above situations. In particular, the methods and apparatuses disclosed herein are able to assign to chambers in the array one of the three following conditions (using the specific example illustrated in Figures 3A and 3B -different compositions of printhead and fluid may need to consider different numbers of preceding rows): 1. Do not actuate the walls and do not release a drop: used for the majority of printing when the fluid chambers have been assigned as non-firing chambers (no fire condition with walls not actuated).

If no drop is required in a pixel on the medium, use this no fire condition on the corresponding chamber in the corresponding actuation cycle, such that the walls are not actuated.

2 Actuate the walls but do not release a drop: this is a no fire condition, but with walls nevertheless actuated. This is used when the fluid chamber will be assigned as a firing chamber in the next actuation cycle but is asleep because: - the same fluid chamber was assigned as non-firing chamber in the previous four consecutive actuation cycles; AND - the two directly adjacent fluid chambers (i.e. one on each side) of the fluid chamber were assigned as non-firing chambers in the previous three consecutive actuation cycles.

When there is a pixel on the medium with no drop to be ejected but the following line requires a drop to be ejected on the same pixel, and if the chamber is asleep (i.e. there have been no fires in the previous three actuation cycles in the same chamber or in the previous two actuation cycles in the adjacent chambers, then use the no fire condition so as not to eject a droplet, but nevertheless walls are actuated to wake up the chamber prior to its being required to print.

3 Actuate the walls and release a drop: used when the chamber is assigned as a firing chamber (fire condition).

In other words, where a drop is required in a pixel on the medium, use the fire condition as normal.

These wall actuation protocols can be transmitted and processed by printhead control apparatus to cause appropriate wall actuations in broadly two ways: * Encoding conditions 1 to 3 directly into the image data stream: this requires the image data for each pixel to be composed by two bits (tri-level image data). This means that the information of only four pixels can be sent per byte transferred to the printhead.

* Selecting the waveform such that conditions 1 to 3 are met, while each pixel is encoded with only either 0 or 1 (bi-level image data): 1 for printing a dot, 0 for do not print a dot. This means that the information of eight pixels can be sent per byte transferred to the printhead, allowing to maximise each printhead's printing speed for the available bandwidth.

At this point, one can conclude that: * The data transfer rate of the tri-level image data must be at least twice the data transfer rate of the bi-level image data for a given print frequency; and * Any example that transmits the information of conditions 1 to 3 to the printhead using bi-level image data, can be easily modified to use tri-level image data instead.

Therefore, in order to be able to determine which of the three conditions should be used at any given moment, the circuitry that either encodes the conditions or selects the waveform should be able to store the input bi-level data of at least: * the previous two pixel lines; * the current pixel line to be printed; AND * the input data for the following pixel line.

In order to implement this protocol, at the highest level, the methods set out herein all include steps of: receiving input data which is split into a grid of pixels, each row of the grid corresponding to an actuation cycle and each column of the grid corresponding to a chamber in the array. For each actuation cycle, an assignment is made (based on the input data) for all of the fluid chambers within said array as either firing chambers for depositing droplets in the actuation cycle or non-firing chambers which do not deposit droplets in the actuation cycle. Moreover, each non-firing chambers is further assigned as either an active non-firing chamber or an inactive non-firing chamber. These assignments are made with a view to avoiding chambers which need to print in a successive actuation cycle from being asleep when they fall due to print. It is possible with any given print system to empirically derive data from the specific print system being used to calibrate the specific criteria for when to wake up a given chamber. For example one set of test prints can be made with a variety of numbers of directly preceding rows of non-actuated pixels located directly behind a test pixel, and others with a variety of numbers of directly preceding rows of non-actuated pixels in columns offset from a test pixel (in either direction) by an integer number of chambers. By analysing this data and looking in particular for evidence of non-uniform droplet deposition, the thresholds to be applied for each criterion can be extracted based on knowledge of the number of rows (or equivalently, time elapsed) since the most recent actuation.

From here, it is possible to assign non-firing chambers as active or inactive based on information on preceding actuation cycles and on information on the next actuation cycle, as well as the derived thresholds for the criteria. Then at least one wall of the active non-firing chambers are actuated without depositing a droplet of fluid onto the medium and the walls of firing chambers are actuated such that each firing chamber releases at least one droplet.

Note that the array may form part of a larger printhead in the sense that the array may comprise only a portion of the total number of chambers. In other examples, of course, the array may include each chamber forming part of the overall printing apparatus. In some examples, there may be multiple rows of chambers. In such cases, the novel features discussed herein may be applied to each row, or only to some rows, or indeed only to portions of some rows. :30

In other words, in an actuation cycle, the chambers may be assigned one of three designations, each having a corresponding action. In cases where the chamber is due to fire (deposit a droplet) in that actuation cycle, the chamber is assigned as a firing chamber and a normal droplet deposition signal is sent to the relevant chamber walls to cause ejection of a droplet, for example, in the manner discussed above. In cases where no droplet is intended to be deposited, the chamber is assigned as a non-firing chamber and no droplet is ejected. However, based on the input data of surrounding deposition patterns in preceding and successive rows, and including adjacent chamber activity, the chamber may be assigned as either an active non-firing chamber, in which the walls are supplied with a signal which causes a non-deposition actuation of the chamber walls (the walls move but no droplets are ejected/deposited) or as an inactive non-firing chamber, where the walls do not move at all, and no droplet is ejected/deposited.

Typically, action is taken to wake the chamber up one actuation cycle prior to the next cycle in which the chamber is required to deposit a droplet. This ensures that a uniform droplet can be ejected on demand even if a given chamber has been inactive for a long time. Note that it is beneficial to introduce the non-deposition actuation into a regular actuation cycle to ensure that the chamber actuations remain synchronised. It is apparent that the specific design of the printhead (i.e. chamber dimensions, nozzle size, fluid characteristics, etc.) all affect the time taken for the chamber to transition from awake to asleep. Therefore, the thresholds set out above may be changed depending on the fluid in use and the characteristics of the printhead.

As noted above, the non-deposition actuation (i.e. the signal supplied to the active non-firing chambers) is supplied during the regular actuation cycle, and for a 3-cycle shared wall system it is supplied during the sub-cycle corresponding to the group to which the relevant chamber belongs. Given that the process involves actuation in a manner which is synchronised to the actuation cycles (and/or the actuation sub-cycles), it is often appropriate to measure the time period in the natural time scale of the system -the period of a single actuation cycle (or actuation sub-cycle). In other words, it is possible to phrase the thresholds set out above in terms of whether the chamber in question has actuated in the preceding M actuation cycles (or has occurred in the past time period of length M') or whether either adjacent chamber has actuated in the preceding N actuation cycles (or has occurred in the past time period of length N'), where typically N<M (or equivalently N'<M'), since the effect of actuating adjacent chambers is less strong than directly actuating the chamber itself. This has the advantage that it is a simple threshold, easily determined empirically for a given set up, and inherently results in a non-deposition actuation schedule which is synchronised with the actuations cycles.

In any case, the threshold (whether couched in terms of number of cycles or elapsed time) will depend in general on: print frequency fluid type, chamber design and so forth. Therefore, while certain numbers of cycles are provided in this document, the frequency, fluid type, head design being used etc. may change these specific values.

As noted above, actuations in preceding actuation cycles can wake up a chamber to inhibit unwanted non-uniform droplet deposition. In some cases, the presence of a non-deposition actuation (i.e. an active non-firing chamber) in preceding rows and adjacent columns can be sufficient to render a chamber awake. Note that in this case, the threshold time or threshold number of preceding rows above which the chamber is deemed to be asleep may be lower than (and is usually at most no higher than) the threshold for other manners of assessing whether the chamber is awake or asleep. In other words, if the chamber in question can be deemed to be awake if a non-deposition actuation has occurred in an adjacent chamber within the preceding P actuation cycles (or has occurred in the past time period of length P') and the thresholds M and N discussed above have the following relationship with P, the newly introduced threshold: PI\I<M, or equivalently P'51\I<M'. :30

There are various ways in which the non-deposition actuation may be implemented to avoid the undesirable effects of non-uniform droplet deposition. These are set out below in more detail, with specific reference to the relevant Figures. Broadly the purpose of these specific examples is that they each reduce both the likelihood of non-uniform droplet deposition and the energy usage of, and wear on, the system. These improvements are seen when the presently disclosed systems and methods are compared with those which either never actuate the walls unless the chamber is due to print and those systems which always actuate the walls, providing a non-deposition actuation when no droplet is due to be printed, and providing a full actuation to deposit droplets when this is the desired outcome.

In Figure 4, the initial stage 400 of many of the protocols set out herein is shown. In effect this captures conditions 1 and 3 above -actuate the walls sufficiently to print a dot, or do not actuate the walls at all, respectively. In what follows, the specific example discussed above will be used as an example only. That is the methods will refer to a chamber being asleep if it has not printed in the preceding four activation cycles AND if its neighbours have also been inactive for the three preceding actuation cycles. It will be appreciated that the general approach is broader than this for the reasons given above, and that the disclosure herein extends to cover any number of preceding actuation cycles for either criterion, so long as the choice is made in the light of data specifically delineating the likelihood that the chamber will deposit a uniform droplet from the likelihood that the chamber will deposit a non-uniform droplet.

The method starts at step 402 in which input data are received. This may be the entire pixel map to be printed, but as will be apparent from the discussion herein, the data may be a condensed version, considering only pixels in the vicinity of the current pixel. The input data often includes information on at least a number of preceding rows and adjacent columns to the current pixel, in order to assess whether the current pixel is awake or asleep. In addition, the current row and the successive row are usually included. Given this, the current row is usually already present because it will have been transferred as part of the preceding actuation cycle (at which time it was the successive row).

The printhead circuitry usually therefore receives the input bi level data of four pixel rows (current, two preceding and one successive) and, based input bi level data of each pixel line, the printhead circuitry assigned all fluid chambers into firing chambers and non-firing chambers based on whether any given chamber corresponds to a location where a dot is to be printed or not.

Next, at step 404, the pixel data may then be divided into four pixel rows such that each row includes the data of each pixel and comprises input bi-level data (e.g. 0: do not release a drop; 1: release a drop).

The method proceeds by considering each pixel in the current row (equivalent to considering what each chamber in the array should do), as noted in step 406. The first consideration of this pixel by pixel process is to ask, at step 408, whether the pixel is due to print in the current actuation cycle. If the answer is YES, then the output is simple -the walls are actuated to eject a droplet. This is indicated by step 414 -"print a dot". Note that the YES option in response to step 408 assumes that a suitable procedure has been in place up to the current actuation cycle, such that the current chamber is awake. In other words, the methods set out herein have been performed in preceding actuation cycles so that a non-uniform droplet is not inadvertently deposited.

In cases where the current row is one of the first M pixel rows in the image, it will not be possible to perform the full analysis set out above since there is insufficient information available. In some cases, these rows may be printed in the medium without pixel-by-pixel analysis, with the analyses set out herein being performed for the third row onwards.

Other options for resolving this issue are to, prior to a first actuation cycle, execute an initiation cycle. The initiation cycle may comprise, for example, actuating the walls of one or more chambers without depositing a droplet of fluid onto the medium. This procedure may be applied to the first two, three, etc. actuation cycles (the number being dependent on the size of the thresholds M and N), in order that any chamber which is due to print, but which has not yet been woken up by any preceding actuation cycle can be woken up in time to print, if needed. In some cases, the initiation cycle includes actuating the walls of all of the chambers in the array without depositing a droplet of fluid onto the medium. This initiates the array with all chambers being awake. In some cases, the initiation cycle may actuate the walls of every other chamber or of every third chamber in the array without depositing a droplet of fluid onto the medium, thereby providing good coverage of the array while reducing the energy expended by a factor of 2 or 3. In yet further examples, even more energy is saved by using an initiation cycle which includes actuating at least one of the walls of only those chambers within a subset of the chambers, wherein the subset comprises only those chambers which are to be assigned as firing chambers in the first M actuation cycles.

Should the answer to the question in step 408 be NO instead, the method then considers at step 410 whether the current chamber is scheduled to print in the next actuation cycle. If not (NO at step 410) there is no need for the current chamber to be awake for the next actuation cycle, since it will not be printing, so the issue of non-uniform deposition does not arise. This leads to the method to conclude with step 418-no actuation of the walls at all.

If instead the current chamber is scheduled to deposit a droplet in the next actuation cycle (YES at 410) then it is necessary to consider whether the current chamber is awake by virtue of the first criterion discussed above. Therefore step 412 asks whether the current chamber printed in the preceding two rows (i.e. actuation cycles). If the answer is YES, then there is no need to take further action as the current chamber is awake, so the walls are not actuated, and the method terminates again at step 418.

At this point, the initial stage method 400 has separated the situations into those where a droplet is definitely to be printed, and those in which no wall actuations are to occur. Should the answer to the question in step 412 be NO, then there are various ways to proceed from here. These are set out in more detail in Figures 5A to 9C, this being indicated by step 416 which indicates that the method should continue with any one of steps 520, 620, 720, 820 or 920.

Broadly, each of these continuation methods represents a trade-off between the complexity of implementation and efficiency due at its core to simpler protocols having a higher likelihood that non-deposition actuations are deployed when they are not strictly necessary.

Each can be employed depending on the specific circumstances envisaged, e.g. including consideration of the data capabilities of the printhead, the number of rows which need considering, the relative number of non-deposition actuations which are anticipated for a given print cycle, etc. The simplest option for proceeding 500 is shown in Figure 5A, which begins at step 520 with a continuation of the initial stage method 400. In this case, the method 500 acknowledges that the clear cases where a deposition should be made or no actuation should be made have already been identified. Therefore the method proceeds to step 522 and actuates the walls (i.e. ensures that at least one of the walls of the current chamber is actuated), but without depositing a droplet. This wakes up the current chamber, ready for the current chamber to print, which is scheduled for the next actuation cycle by virtue of the answer to question 410. It will be apparent that this is not a particularly nuanced approach to introducing active non-firing chambers, and as might be expected, this example pays for this simplification to the printhead circuitry logic with the expense of actuating walls of the non-firing chambers when it may not be required, increasing the heat generation and the wear on the printheads beyond a hypothetical lower bound.

Figure 5B shows an example of the pixels in the vicinity of the current pixel 3b (labelled pixel being considered). Shown in black are pixels which do not contribute to the analysis (1a-4a; lc-4c). The only pixels to be considered, in line with the combination of Figures 4 and 5A, are the two preceding ones in the same column (1 b, 2b), the current pixel (3b -assessed to decide whether it should print in this cycle in step 408) and the pixel in the same column as the current pixel on the next row (4b) is assessed as to whether it will be printed (question 410). While this approach does allow for long strings of inactive pixels to be corrected when they occur in the same column, it can be seen that even if there were to be a dot printed in 2a, 3a, 2c, or 3c (meaning that the current chamber is awake), a non-deposition actuation would be scheduled, potentially leading to overactuafion and increased wear and energy expenditure.

An improvement 600 to this protocol is therefore shown in Figure 6A, which once more follows Figure 4. In general, portions of one protocol which overlap with portions of another protocol which has already been described will not be described in detail again. For this reason, steps in different protocols which nevertheless have the same effect are labelled with reference numerals which share the same two final digits, with the first digit representing the Figure number.

In Figure 6A, an additional step 624 is included which considers whether chambers adjacent to the current chamber were actuated in the preceding actuation cycle or will be in the current actuation cycle. If the answer to this question is YES (there has been such an actuation), then the current chamber will be awake on the next actuation cycle (on which it is due to print) according to the protocol, because by the next actuation cycle there cannot have been a string of three or more non-actuations in the adjacent chambers if the answer to 624 is YES. Because the current chamber is already awake in this case, there is no need for further action and the method proceeds to terminate at step 618, in which no wall actuations are instructed. However, if the answer to question 624 is NO, then by virtue of the path taken through the flow charts in Figures 4 and 6A, neither of the relevant criteria have been met and the chamber is therefore deemed to be asleep. Therefore the method proceeds to step 622 and actuates one or more walls of the current chamber to wake it up.

Figure 6B shows a modified version of Figure 5B, highlighting that additional pixels are considered in this approach. This is highlighted by pixels 2a, 3a, 2c and 3c being white (indicated as being considered in the analysis). This is in line with the protocol set out above in respect of Figure 6A. The remaining pixels in Figure 6B share their allocations with the equivalent ones in Figure 5B, for the same reasons. It is apparent that this example avoids some of the unneeded non-deposition actuations which can occur in the previous example (Figures 5A and 5B), by supressing non-deposition actuations if there has been (or will be) an actuation in at least one of the adjacent chambers in the same row, as this will lead to the chamber under consideration being awake.

Yet a further development 700 of the general protocol is shown in Figure 7A, which once more follows on from step 416 in Figure 4. This protocol follows the same route as that in Figure 6A, at step 724. However, should the answer to this question be NO, instead of directly deploying a non-deposition actuation, the protocol asks a further question at step 726. Specifically the protocol 700 asks at step 726 whether either chamber adjacent to the current chamber is due to eject a droplet in the next actuation cycle. If the answer to this question is YES, then the protocol 700 proceeds to terminate at step 718 with no wall actuation being required. Only if neither of the adjacent chambers to the current chamber is due to print in the next actuation cycle does the protocol 700 proceed to terminate with step 722, in which at least one wall of the current chamber is actuated without depositing a droplet. This is because actuating a chamber adjacent to the current chamber in the next actuation cycle will result in the current chamber being awakened in that next actuation cycle.

Figure 7B shows a development of Figure 6B including this new information. The only difference is that pixels 4a and 4c are now white, indicating that they form part of the consideration of whether the current chamber is asleep or not, in accordance with the flow chart of Figure 7A On conjunction with that of Figure 4).

This protocol my occasionally result in a chamber requiring a non-deposition actuation, but not receiving it, for example in cases where the only adjacent chamber(s) which actuate(s) in the next actuation cycle actuate(s) after the current chamber has actuated to print a droplet in cases where the system uses a three-cycle actuation. In these cases, the current chamber risks still being asleep and there is a risk of a non-uniform droplet being deposited.

This situation is addressed in a further development 800 of the general initial stage protocol 400, shown in in Figure 8A. In these examples, the array is assumed to have a multi-cycle actuation protocol, such as the three-cycle protocols set out above in detail, and as shown schematically in Figures 2A and 2B. This will not be described again in detail, but for the purposes of the following discussion, it is assumed that the protocol is a three-cycle shared wall actuation scheme and that the chambers of the array are arranged in three sub-arrays (or groups), such that any three adjacent chambers includes at least one chamber which is a member of each sub-array, and wherein the chambers of the array are arranged in a repeating pattern of chambers belonging to each of three sub-arrays arranged in the same order to from the repeating pattern. Moreover, each actuation cycle is assumed to include the actuation of three actuation sub-cycles, each actuation sub-cycle including: the assignment of the chambers of a different one of the sub-arrays as firing chamber, active non-firing chambers, or inactive non-firing chambers, and the corresponding actuation (or not) of the walls of the chambers of that sub-array in accordance with the assignment. As noted above, each actuation cycle includes the actuation of all three actuation sub-cycles in order of a first actuation sub-cycle, followed by a second actuation sub-cycle, followed by a third actuation sub-cycle. Arranging the system/method in this way allows the use of multiple sub-cycles and advantageously leads to the following two further examples in Figures 8A to 8C and 9A to 9C.

The following discussion (Figures 8A to 10) assumes that the A sub-cycle occurs first in an actuation cycle, followed by the B sub-cycle, then finally the C sub-cycle is executed. Since up to this point the specific labelling of the sub-cycles is arbitrary, this retains the flexibility that the physical location of the chambers of each group, relative to the chambers of the other groups (essentially, is the array ordered from left to right as...ABCABC... or...CBACBA...) is not determined. This is not a problem as the sub-cycles can be rearranged/relabelled arbitrarily without loss of generality.

However, it is important to note that Figures 8B, 8C, 9B, 9C, and the images which form part of steps 1034 and 1036 imply that the chamber to the left of the current chamber is earlier in the actuation cycle. In other words there is an assumption that if the sub-cycles arranged to occur in the order A -> B -> C, then if the current chamber is a B chamber, the chamber to the left is an A chamber, if the current chamber is a C chamber then the chamber to the left is a B chamber, meaning that the chamber to the left actuates earlier in the actuation cycle than the current chamber. This rule is broken if the the current chamber is an A chamber, in which case the chamber to the left is a C chamber, hence the A chambers (referred to interchangeably as the first sub-cycle in the actuation cycle) are called out as a special case.

It will be apparent that, while the Figures assume that this is the case, it is equally possible that the chamber to the right is due to actuate earlier in the actuation cycle than the current chamber (i.e. the array is arranged in a CBA pattern). For this reason, while the specific case of ABC ordering is discussed, the following should be interpreted as also covering the mirror image case (CBA), where the exact same principles apply, except with a mirroring effect required to transfer the teaching across.

Turning now to the detail of Figure 8A, the steps 724 and 726 of Figure 7A are repeated as steps 824 and 826 and will not be discussed again. However, rather than going straight to implementing a non-deposition actuation, the protocol 800 leverages the three-cycle system, and specifically the knowledge of the order in which the sub-cycles are implemented. This leads to question 828, which ascertains whether the current pixel is a pixel of the first sub-cycle (for ease of reference, referred to as A hereinafter). If YES, there is no way for the chamber to be woken up in the next actuation cycle prior to printing, and so the method 800 proceeds directly to terminate at step 822, whereby at least one wall of the current chamber is actuated in the current actuation cycle without depositing a droplet.

If, however, the answer is NO, then the protocol 800 proceeds to step 830, in which a further question is asked: Is an adjacent chamber in a preceding sub-cycle due to print in the next row -in figures 8a and 9A "preceding" does not mean directly preceding but means "occurs at some point earlier in the same actuation cycle". Note that from question 828, we know that the present chamber is part of either actuation sub-cycle B or C (the second or third actuation sub-cycles, respectively). If it is a B chamber, then only the A chamber precedes it in the next actuation cycle, while if the current chamber is a C chamber then both the A and B chambers precedes it in the next actuation cycle (i.e. both adjacent chambers will actuate if so instructed prior to the present chamber). Given this, the question at section 830 should be interpreted as containing two sub-questions: 1. if the current chamber is a B chamber, is the adjacent A chamber due to actuate in the next actuation cycle? 2. if the current chamber is a C chamber, is either of the two directly adjacent chambers due to actuate in the next actuation cycle? If, given the designation of the current chamber, the answer is answered YES, then no action need be taken (method 800 terminates at step 818) as the actuation of at least one of the adjacent chambers will occur prior to the time when the current chamber is due to print, and therefore the current chamber will be awake at that time. Conversely, if the question is answered NO, then the chamber must be specifically awoken so the method proceeds to terminate at step 822, in which at least one wall of the current chamber is actuated without depositing a droplet.

The adaptations to Figure 7B implied by this method are shown in Figures 8B and 80. Figure 8B shows the example of a B sub-cycle (i.e. the current chamber is a B chamber), in which only actuations in the next actuation cycle which form part of the A cycle are included in the analysis. This means that contrary to Figure 7B, pixel 4c is not considered here, since it is due to actuate Of it does indeed actuate at all) after pixel 4b would have already printed.

This means that it cannot assist in waking up pixel 4b, so need not be considered. Conversely, the situation for when the current chamber is a C chamber is exactly as shown in Figure 7B, since it is known that the C sub-cycle occurs last in any given actuation cycle, so both adjacent chambers can result in the current chamber being awoken.

In Figure 8C, no adjacent chamber in the next actuation cycle is considered. This is because Figure 8C represents the situation when the current chamber is in the (chronologically) first sub-cycle. In such cases, there can be no action in the successive actuation cycle which wakes up the current chamber and therefore these pixels do not feed into the analysis. Note that this means that the map of pixels which are considered in this case is identical to that shown in Figure 6B.

A further development 900 of this idea is shown in Figure 9A. Here the process is very similar to the process 800 in Figure 8A. Here, however, there is no equivalent to step 826, and the process proceeds directly from a determination of NO at step 924 (equivalent to steps 724 and 824) to step 928 and 930 which asks whether the chamber is an A chamber, exactly as in step 828.

If the answer to question 928 is YES, then there can be no actuation of the current chamber in the next actuation cycle in time for the current chamber to print, so the method 900 proceeds to step 922 where it terminates by actuating at least one wall of the current chamber in the current actuation cycle, without ejecting a droplet. If the answer to question 928 is NO, then the method proceeds to step 930 which asks whether an adjacent chamber in a preceding sub-cycle is due to print in the next actuation cycle (identical to question 830).

As before, if this question 930 is answered YES, then there is no need to take further actuation (and the method terminates at step 918) since the scheduled printing of an adjacent chamber earlier in the next actuation cycle will cause the current chamber to wake up prior to its own scheduled printing in the next actuation cycle.

However, if the answer to question 930 is NO, then the method proceeds to step 932 in which, in the next actuation cycle, at least one wall of an adjacent chamber which is part of an earlier sub-cycle is actuated. This in turn causes the current chamber to wake up prior to its scheduled droplet deposition.

This is shown schematically in Figures 9B and 9C. In Figure 9B, the situation is broadly as in Figure 8B, but here, the non-deposition actuation is applied to pixel 4a, rather than to the current pixel (3b). Note that this is for the case where the current chamber is a B chamber, so only has the adjacent A chamber which precedes it in the next actuation cycle. In cases where the current chamber is a C chamber, then either one of the two adjacent pixels (4a, 4c) could be actuated without droplet deposition to achieve the desired effect. In some cases, both adjacent pixels could be actuated if desired. Note that there is no inherent advantage to actuating either pixel 4a or pixel 4c in these cases. Depending on the wider print pattern, it may be beneficial to select one or other of these, since that may also wake up another pixel which is not shown in Figure 9B.

The situation in Figure 9C is identical to the version shown in Figures 8C and 6B. As before, this reflects the fact that where the current chamber is an A chamber, there can be no awakening derived from actuations in a successive actuation cycle, since the A sub-cycle occurs first in the actuation cycle. In this example, the non-deposition actuation signal is deferred to the latest possible time, if it is needed at all. As noted, in some cases, this allows the non-deposition actuation signal not to occur in the current actuation cycle but in a preceding actuation sub-cycle in the next actuation cycle in some cases. This in turn means that chambers woken up in this manner remain awake for one actuation cycle longer than if they had been woken up in the current actuation cycle, thereby improving efficiency yet further by reducing the need for subsequent non-deposition actuations.

Turning now to Figures 10A and 10B, in which an alternate approach 1000 is taken. Here the initial steps 1002, 1004, 1006, 1008, and 1010 are directly parallel to steps 402, 404, 406, 408, and 410, and result in the same outcomes. These will therefore not be discussed again in detail here. :30

Should the answer to question 1010 (will the current pixel print on the next row) be YES, then the method proceeds to determine, based on the sub-cycle to which the pixel belongs, whether there have been or will be suitable actuations in adjacent chambers in recent pixel rows or in the upcoming row. This is performed by way of a pixel-wise AND operation using a suitable mask for the sub-cycle to which the pixel belongs, as indicated by step 1028. The details of this step are set out in more detail in Figure 10B.

In Figure 10B, it can be seen that step 1028 includes a step 1034, in which the question is asked as to which sub-cycle the pixel belongs to. The possible answers are A, B and C in this example indicating a three-cycle system, wherein A is the chronologically first sub-cycle to occur, B comes second, and C is the final sub-cycle (as noted above, these cycles are convenient labels for illustration, and alternative ordering of the cycles may be implemented).

The method proceeds much like method 800 in Figure 8A, but instead of polling the history of recent actuations to identify the status of the current chamber, binary bit masks are layered using bitwise AND operations using specialised pixel masks for each of the A, B and C sub-cycles and the input data, in which the instruction to print a dot at a particular pixel is recorded as a 1 at that pixel and is 0 otherwise. In this example, 3x4 pixel bitmasks are used to delineate the active area (shown in white in steps 1036a-c) from the area which cannot affect the status of the current pixel (black regions in steps 1036a-c). Pixels outside of this region may either not be defined, or may be filled with Os since far away and/or far in the future and/or long ago printing actions do not affect the status of the current pixel.

In cases where the current chamber is an A chamber, the method proceeds at step 1036a to perform a pixel-wise AND operation between the image data and the "A" mask indicated in step 1036a, which restricts the active region (discussed above) to only preceding actuation cycles and the current actuation cycle. Once again, this is because where the current chamber is an A chamber, there can be no waking up effect provided by adjacent chambers in the next actuation cycle, since the A chamber will need to print before any such effect could be felt.

As an alternative, the current chamber may be a B chamber. In this case, the method proceeds at step 1036b to perform a pixel-wise AND operation between the image data and the "B" mask indicated in step 1036b, which restricts the active region (discussed above) to only preceding actuation cycles, the current actuation cycle and the adjacent A chamber in the next actuation cycle. Here this is shown as being to the left of the current chamber (indicating an ABC ordering of the chambers), but it will be appreciated that the mirror image could be used instead. Once again, this is because where the current chamber is an B chamber, there can be no waking up effect provided by an adjacent C chamber in the next actuation cycle, since the B chamber will need to print before any such effect would be felt. Conversely an A chamber in the next actuation cycle which is due to print and is adjacent to the current B chamber will do so before the current chamber is due to print, and so will wake up the current chamber in time to print.

Finally, if the chamber is a C chamber, then either adjacent chamber may wake up the current chamber before it needs to print. This is captured in step 1036c, in which a "C" mask is illustrated having an active region (white) which includes both adjacent chambers in the next actuation cycle. If the current chamber is a C chamber, then both adjacent pixels in the next actuation row will wake up the current chamber in time for the current chamber to print if they do indeed print in the next cycle.

In each of the examples shown in steps 1036a-c, the result of the AND operation will be a grid which has is in it only where the input image data instructs a print which aligns with the white portions of the "A" mask, the "B" mask or the "C" mask, as appropriate for the group to which the current chamber belongs. The next step in the process is therefore to assess whether any such 1s exist, as set out in step 1038. If the answer to this question is NO, then the current chamber is deemed to be asleep and a non-deposition actuation signal is sent to wake up the chamber, causing the method 1000 to terminate at step 1022.

On the other hand if even a single 1 exists in the result of the pixel-wise AND operation (if the YES branch is taken), then the current chamber is deemed to be awake and no further action need be taken, so the method proceeds to terminate at step 1018. It will be apparent that this approach replicates, using a different methodology the steps of previously discussed approaches, thereby providing the associated advantages of reduced heating and increased printhead lifetime.

Note that in the above example a different mask is used for each of the A, B and C sub-cycles. This ensures that walls of chambers are only supplied with a non-deposition actuation signal in cases where the walls need actuation. However, in cases where a small amount of superfluous non-deposition actuation signals is tolerable, the process can be simplified by using the same mask for different sub cycles. In general, this means that a given sub-cycle can use either its own specific mask or the mask for any preceding sub-cycle. This means that the A cycle can only use the A mask, the B cycle can use the A or B masks, and the C cycle can use any of the three masks (A, B or C).

Two particular implementations of this are of particular interest. The first is to always use the A mask, which simplifies the implementation, but can lead to over-actuation of the walls and a reduced efficiency. The second is to use the A mask for the A sub-cycle and to use the B mask for the B and C sub-cycles, thereby reducing the likelihood of implementing unnecessary non-deposition actuation signals by using the correct mask for two out of three of the sub-cycles (and using a mask for the third sub-cycle which still captures on average half of the situations in which a non-deposition actuation signal is not needed), while still simplifying the implementation in the sense that fewer different masks are needed.

Note also that in this example, "print" and "don't print" were assigned a 1 and 0 value respectively in the input data. Likewise the bitmasks were set up with 1s in the locations which are to be considered (the active region) and Os in those which are not to be considered. It will be apparent that either or of these encoding regimes may be switched so that 0 and 1 encode the opposite information in each case. It is trivial to make this change, although the pixel-wise combining operation may need to use a different logical operation.

As a simple example consider a situation in which "print" and "don't print" are assigned a 0 and 1 value respectively in the input data and the bitmasks are set up with Os in the locations which are to be considered) and 1s in those which are not to be considered (i.e. the exact opposite encoding to that illustrated in Figures 10A and 10B). In this case it would be necessary to highlight pixel locations in which both the input data and the bitmask have a 0 (where previously we were looking for pixel locations in which both the input data and the bitmask have a 1). For this task we could use a pixel-wise NOR operation, which outputs a 1 only if both inputs are 0. Step 1038 can therefore still ask whether and 1s remain to achieve the equivalent effect in this example. Other encodings will of course result in different logic.

As noted above, the present disclosure extends to (although not expressly depicted in the figures) a droplet deposition apparatus comprising one or more fluid chambers, the apparatus being configured to carry out any of the methods set out above. The droplet deposition apparatus may further comprise a computer in data communication with said one or more fluid chambers, wherein the computer is programmed to carry out the assigning step based on the input data. Optionally, the computer is further programmed to send instructions to the one or more fluid chambers, so as to cause them to carry out the steps of actuating the walls in accordance with the outcome of the assigning step.

Claims

Claims 1. A method for depositing droplets of fluid onto a medium utilising a droplet deposition head, the droplet deposition head comprising: an array of fluid chambers separated by interspersed walls formed of a piezoelectric material, each fluid chamber communicating with an aperture for the release of droplets of fluid, each of said walls separating two neighbouring fluid chambers, and each fluid chamber being defined by a first wall in a first direction relative to the fluid chamber, and a second wall in a second direction relative to the fluid chamber, the second direction being opposite to the first direction; wherein each chamber has at least one electrode disposed across surfaces of the first and second walls internal to the chamber, the electrodes internal to each chamber being providable with actuation voltages independently of actuation voltages supplied to the electrodes of other chambers; wherein each of said walls of each chamber is actuable to change the volume of the chamber from a neutral configuration such that, in response to the application of difference of potential between a first electrode and a second electrode located on opposed sides of a wall separating a first chamber and a second chamber adjacent to the first chamber, the wall moves into a deformed configuration relative to the position of the wall in the neutral configuration, the method comprising, the steps of: receiving input data, the input data being split into a grid of pixels, each row of the grid corresponding to an actuation cycle and each column of the grid corresponding to a chamber in the array; and for each actuation cycle, assigning, based on said input data, all the fluid chambers within said array as either firing chambers for depositing droplets in the actuation cycle or non-firing chambers which do not deposit droplets in the actuation cycle, and wherein each non-firing chamber is further assigned as either an active non-firing chamber or an inactive non-firing chamber wherein each non-firing chamber is assigned as an active non-firing chamber or an inactive non-firing chamber based both on information on preceding actuation cycles and on information on the next actuation cycle; and actuating at least one of the walls of each active non-firing chamber without depositing a droplet of fluid onto the medium and actuating at least one of the walls of each firing chamber such that each firing chamber releases at least one droplet the resulting droplets forming dots disposed on a line on the medium, the dots being separated on the line by gaps corresponding to the non-firing chambers.
2. The method according to claim 1 wherein, during an actuation cycle, each non-firing chamber is assigned as an active non-firing chamber or an inactive non-firing chamber based at least in part on: locations of firing chambers or active non-firing chambers in one or more preceding actuation cycles; and/or locations of firing chambers in a successive actuation cycle.
The method according to claim 2 wherein, during an actuation cycle, each non-firing chamber is assigned as an active non-firing chamber or an inactive non-firing chamber based at least in part on whether that chamber was a firing chamber or an active non-firing chamber in either of the two preceding actuation cycles.
4. The method according to claim 2 or claim 3 wherein, during an actuation cycle, each non-firing chamber is assigned as an active non-firing chamber or an inactive non-firing chamber based at least in part on whether that chamber or either directly adjacent chamber was a firing chamber or an active non-firing chamber in the preceding actuation cycle.
The method according to any one of claims 2 to 4 wherein, for a given chamber assigned as a non-firing chamber, whether that chamber is assigned as an active non-firing chamber or an inactive non-firing chamber depends at least in part on whether adjacent chambers to the given chamber in the array have been assigned as firing chambers, active non-firing chambers, or inactive non-firing chambers for that actuation cycle.
The method according to any one of the preceding claims, wherein for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if either: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid.
The method according to any one of claims 1 to 5, wherein for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; or either of the two adjacent chambers was or will be assigned as a firing chamber or an active non-firing chamber in the preceding or current actuation cycle; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid.
The method according to any one of claims 1 to 5, wherein for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; or either of the two adjacent chambers was or will be assigned as a firing chamber or an active non-firing chamber in the preceding or current actuation cycle; or either of the adjacent chambers is to be assigned as a firing chamber in the successive actuation cycle; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid.
9. The method according to any preceding claim, wherein the chambers of the array are arranged in three sub-arrays, such that any three adjacent chambers includes at least one chamber which is a member of each sub-array, and wherein the chambers of the array are arranged in a repeating pattern of chambers belonging to each of three sub-arrays arranged in the same order to from the repeating pattern; wherein each actuation cycle includes the actuation of three actuation sub-cycles, each actuation sub-cycle including: the assignment of the chambers of a different one of the sub-arrays as firing chamber, active non-firing chambers, or inactive non-firing chambers; the actuation of at least one wall of each firing chamber of that sub-array to release at least one droplet; the actuation of at least one wall of each active non-firing chamber without releasing a droplet; and wherein each actuation cycle includes the actuation of the three actuation sub-cycles in order of a first actuation sub-cycle, followed by a second actuation sub-cycle, followed by a third actuation sub-cycle.
10. The method according to claim 9, wherein for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; or either of the two adjacent chambers was or will be assigned as a firing chamber or an active non-firing chamber in the preceding or current actuation cycle; or that chamber is part of a second or third sub-array, configured to be actuatable as part of the second or third actuation sub-cycles respectively and the adjacent chamber forming part of the preceding sub-cycle will be assigned as a firing chamber or an active non-firing chamber in the successive actuation cycle; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid, including where that chamber is part of a first sub-array, configured to be actuable as part of the first actuation sub-cycle.
11. The method according to claim 9, wherein for each chamber in the array in a given actuation cycle, the method includes: if that chamber is assigned as a firing chamber, actuating at least one of the chamber walls to deposit a droplet of fluid onto the medium; and otherwise: that chamber is assigned as a non-firing chamber, and wherein if: that chamber is to be assigned as a non-firing chamber in the successive actuation cycle; or that chamber was assigned as a firing chamber or an active non-firing chamber in the two preceding actuation cycles; or either of the two adjacent chambers was or will be assigned as a firing chamber or an active non-firing chamber in the preceding or current actuation cycle; or that chamber is part of a second or third sub-array, configured to be actuatable as part of the second or third actuation sub-cycles respectively and the adjacent chamber forming part of the preceding sub-cycle will be assigned as a firing chamber or an active non-firing chamber in the successive actuation cycle; then that chamber is assigned as an inactive non-firing chamber and the chamber walls are not actuated; if that chamber is assigned as a non-firing chamber and will be assigned as a firing chamber in the successive actuation cycle and neither that chamber was assigned as a firing chamber or an active non-firing chamber nor were or will either of the adjacent chambers assigned as firing chambers or active non-firing chambers in the preceding or current actuation cycle and that chamber is part of the second or third sub-array, configured to be actuatable as part of the second or third actuation sub-cycles respectively and the adjacent chamber forming part of the preceding sub-cycle will not be assigned as a firing chamber or an active non-firing chamber in the successive actuation cycle; then the adjacent chamber forming part of the preceding sub-cycle is assigned as an active non-firing chamber in the successive actuation cycle; otherwise that chamber is assigned as an active non-firing chamber and at least one of the chamber walls is actuated without depositing a droplet of fluid, including where that chamber is part of a first sub-array, configured to be actuable as part of the first actuation sub-cycle.
12. The method according to any one of the claims 1 to 5, wherein the input data is processed to assign, for each actuation cycle, each chamber in the array as a firing chamber, an active non-firing chamber or an inactive non-firing chamber based on whether that chamber has been assigned as a firing chamber or an active non-firing chamber in the preceding two actuation cycles or whether either adjacent chamber is or will be assigned as a firing chamber or an active non-firing chamber in the current or preceding actuation cycle.
13. The method according to claim 12, wherein, for each pixel in the grid, the assignment as a firing chamber, an active non-firing chamber or an inactive non-firing chamber is performed using a pixel-wise AND operation on a bitmap version of the input data and a vicinity mask formed by considering the previous two actuation cycles of each chamber and the preceding and current actuation cycles of each adjacent chamber.
14. The method according to claim 13, wherein the chambers of the array are arranged in three sub-arrays, such that any three adjacent chambers includes at least one chamber which is a member of each sub-array, and wherein the chambers of the array are arranged in a repeating pattern of chambers belonging to each of three sub-arrays arranged in the same order to from the repeating pattern; wherein each actuation cycle includes the actuation of three actuation sub-cycles, each actuation sub-cycle including: the assignment of the chambers of a different one of the sub-arrays as firing chamber, active non-firing chambers, or inactive non-firing chambers; the actuation of at least one of the walls of the firing chambers of that sub-array to release at least one droplet; the actuation of at least one of the walls of the active non-firing chambers without releasing a droplet; and wherein each actuation cycle includes the actuation of the three actuation sub-cycles in order of a first actuation sub-cycle, followed by a second actuation sub-cycle, followed by a third actuation sub-cycle; and where the pixel is in the second or third actuation cycle, the vicinity mask includes a consideration of the adjacent chamber in the preceding actuation sub-cycle in the next actuation cycle.
15. The method according to claim 13 or claim 14, wherein the bitmap version of the input data is formed by a pixel-wise OR combination of a first bitmap representation of whether a chamber was assigned as a firing chamber in preceding actuation cycles with a second bitmap representation of whether a chamber was assigned as an active non-firing chamber in preceding actuation cycles.
16. The method according to any one of claims 13 to 15, wherein the bitmap version of the input data and/or the vicinity mask is a 3 x 4 grid with the dimension having extent 3 representing the current pixel and both adjacent pixels and the dimension having extent 4 representing the current actuation cycle, the next actuation cycle and two preceding actuation cycles.
17. The method according to any one of the preceding claims, wherein prior to a first actuation cycle, an initiation cycle is executed, the initiation cycle comprising actuating at least one of the walls of one or more chambers without depositing a droplet of fluid onto the medium.
18. The method according to claim 17, wherein the initiation cycle includes actuating at least one of the walls of all of the chambers in the array without depositing a droplet of fluid onto the medium.
19. The method according to claim 17 or claim 18, wherein the initiation cycle includes actuating at least one of the walls of each chamber within a subset of the chambers, wherein the subset comprises only those chambers which are to be assigned as firing chambers in the first three actuation cycles.
20. The method according to any one of the preceding claims, wherein walls of the inactive non-firing chambers do not move during the actuation cycle.
21. The method of any one of the preceding claims, wherein each chamber has a single electrode disposed across both chamber walls.
22. A droplet deposition apparatus comprising one or more fluid chambers, the apparatus being configured to carry out a method according to any preceding claim.
23. The droplet deposition apparatus according to claim 22, further comprising a computer in data communication with said one or more fluid chambers, wherein the computer is programmed to carry out the assigning step based on the input data.
24. The droplet deposition apparatus according to claim 23, wherein the computer is further programmed to send instructions to the one or more fluid chambers, so as to cause them to carry out the steps of actuating the walls.
25. A computer program comprising instructions to cause the droplet deposition apparatus of any of claims 22 to 24 to execute the method of any of claims 1 to 21.