Technique for printing ink droplet measurement and control of fluid deposition within tight tolerances

Abstract

To provide a suitable technique for printing ink droplet measurement and control of fluid deposition within tight tolerances.SOLUTION: An ink printing process plans droplet combinations to reach a specific total ink fill per target area and adopts droplet volume per nozzle measurement and processing software that ensures compliance with minimum and maximum ink fills set by specifications. In various embodiments, different droplet combinations are generated through different printhead / substrate scan offsets, offsets between print heads, use of different nozzle drive waveforms, and / or other techniques. These combinations can be based on repeated high-speed drop measurements that develop an understanding of each nozzle of expected drop volume, velocity, trajectory averaging and diffusion, along with the planned drop combinations based on these statistical parameters. Optionally, random fill variations may be introduced to reduce the uneven effect in a completed display device. The disclosed technique has many possible uses.SELECTED DRAWING: Figure 8A

Description

[Technical field]

[0001] (CROSS-REFERENCE TO RELATED APPLICATIONS) This application claims priority to U.S. Provisional Patent Application No. 61 / 950,820, entitled "Techniques For Print Ink Droplet Volume Measurement And Control Over Deposited Fluids Within Precise Tolerances," filed March 10, 2014 (first inventor Nahid Harjee). This application is also a continuation-in-part of U.S. Provisional Patent Application No. 14 / 162,525, entitled "Techniques for Print Ink Volume Control To Deposit Fluids Within Precise Tolerances," filed January 23, 2014 (first inventor Nahid Harjee). U.S. Patent Application No. 14 / 162,525 claims priority to Taiwan Patent Application No. 102148330, entitled "Techniques for Print Ink Volume Control To Deposit Fluids Within Precise Tolerances," filed on December 26, 2013 (first inventor Nahid Harjee), and is a continuation of PCT Patent Application No. PCT / US2013 / 077720, entitled "Techniques for Print Ink Volume Control To Deposit Fluids Within Precise Tolerances," filed on December 24, 2013 (first inventor Nahid Harjee).PCT Patent Application No. PCT / US2013 / 077720 is a joint venture of U.S. Provisional Patent Application No. 61 / 746,545, filed December 27, 2012 (first inventor Conor Francis Madigan) and entitled "Smart Mixing," U.S. Provisional Patent Application No. 61 / 822855, filed May 13, 2013 (first inventor Nahid Harjee) and entitled "Systems and Methods Providing Uniform Printing of OLED Panels," U.S. Provisional Patent Application No. 61 / 842351, filed July 2, 2013 (first inventor Nahid Harjee) and entitled "Systems and Methods Providing Uniform Printing of OLED Panels," U.S. Provisional Patent Application No. 61 / 842351, filed July 23, 2013 (first inventor Nahid Harjee) and entitled "Systems and Methods Providing Uniform Printing of OLED Panels," U.S. Provisional Patent Application No. No. 61 / 857,298 (first inventor Nahid Harjee), entitled “Systems and Methods Providing Uniform Printing of OLED Panels,” filed November 1, 2013 (first inventor Nahid Harjee), and U.S. Provisional Patent Application No. 61 / 898,769 (first inventor Nahid Harjee), entitled “Techniques for Print Ink Volume Cont.,” filed December 24, 2013 (first inventor Nahid Harjee). No. 61 / 920,715 (first inventor Nahid Harjee), entitled "OLED Printing Systems and Methods Using Laser Light Scattering for Measuring Ink Drop Size, Velocity and Trajectory," filed April 26, 2014 (first inventor Alexander Sou-Kang Ko), and U.S. Provisional Patent Application No. 61 / 816696 (first inventor Alexander Sou-Kang Ko), entitled "OLED Printing Systems and Methods Using Laser Light Scattering for Measuring Ink Drop Size, Velocity and Trajectory," filed August 14, 2013 (first inventor Alexander Sou-Kang Ko), each of which is incorporated herein by reference.

[0002] The present disclosure relates to techniques for measuring, with a high degree of statistical accuracy, inkjet drop volumes used in organic light-emitting diode ("OLED") device fabrication, the use of printing processes to transfer droplets of fluid ink in precise total quantities to target areas of a substrate, and related methods, devices, improvements, and systems. In one non-limiting application, the techniques provided by the present disclosure can be applied to manufacturing processes for OLED display panels. [Background technology]

[0003] (background) In printing processes where a printhead has multiple nozzles, not all nozzles respond in the same way to a standard drive waveform, i.e., each nozzle may produce a slightly different volume of droplets. In situations where nozzles are relied upon to deposit fluid droplets into respective fluid deposition areas ("target areas"), lack of consistency can lead to problems. This is particularly true in manufacturing applications where the ink transports materials that will become permanent thin film structures in electronic devices. One exemplary application where this problem arises is in manufacturing processes applied to the fabrication of displays, such as organic light-emitting diode ("OLED") displays, as used in small and large electronic devices (e.g., in portable devices, large high-definition television panels, and other devices). When a printing process is used to deposit inks that carry the light-emitting materials of such displays, volume differences across rows or columns of pixels contribute to visible lighting or color defects in the displayed image. "Ink" as used herein refers to any fluid applied to a substrate by the nozzles of a print head, regardless of color characteristics; for example, in the described OLED display processing application, the ink is typically deposited in place and then treated, dried, or cured to directly form a permanent material layer; it should be noted that this process may be repeated with the same ink or different inks to form several such layers.

[0004] FIG. 1A is used to introduce this nozzle-droplet inconsistency problem with an exemplary schematic generally referenced using the numeral 101. In FIG. 1A, a printhead 103 can be seen to have five ink nozzles, each depicted using a small triangle at the base of the printhead, numbered (1)-(5), respectively. Note that in a typical manufacturing application, there may be more than five nozzles, e.g., 24-10,000 nozzles, depending on the application, and in the case of FIG. 1A, five nozzles are referenced simply for ease of understanding. In an exemplary application, it is desired to deposit 50 picoliters (50.00 pL) of fluid into each of five specific target areas of an array of such areas, and further assume that each of the five nozzles of the printhead ejects 10 picoliters (10.00 pl) of fluid into each of the various target areas with each relative movement ("pass" or "scan") between the printhead and the substrate. The target regions may be any surface region of the substrate, including adjacent unseparated regions (e.g., such that the deposited fluid ink partially diffuses and mixes together between the regions), or respective fluidically isolated regions. These regions are generally represented in FIG. 1A using ovals 104-108, respectively. Thus, it may be assumed that exactly five passes of the print head are required as depicted to fill each of the five specific target regions. However, print head nozzles will, in practice, have some minor variations in structure or operation, such that a given drive waveform applied to each nozzle transducer will result in a slightly different drop volume for each nozzle. For example, as depicted in FIG. 1A, firing of nozzle (1) produces a drop volume of 9.80 picoliters (pL) on each pass, with five 9.80 pL drops depicted within oval 104. Note that while each of the drops is represented in the diagram by a specific location within target region 104, in practice, the locations of each of the drops may be identical or may overlap. Nozzles (2)-(5), in contrast, produce slightly different discrete drop volumes of 10.01 pL, 9.89 pL, 9.96 pL, and 10.03 pL.With five passes between the print head and the substrate, with each nozzle mutually exclusively depositing fluid into the target regions 104-108, this deposition would result in a total deposited ink volume variation of 1.15 pL across the five target regions. This may be unacceptable for many applications. For example, in some applications, a difference of just 1 percent (or even less) in the deposited fluid may cause problems. In the case of OLED display processing, such variation could potentially result in image artifacts observable in the finished display.

[0005] Thus, manufacturers of televisions and other forms of displays would effectively specify a precise volume range that must be observed with a high degree of precision, e.g., 50.00 pL ± 0.25 pL, for the resulting product to be deemed acceptable. Note that in this exemplary case, the specified tolerance must be within 0.5 percent of the 50.00 pL target. In an application where each nozzle represented by FIG. 1A was to deposit into pixels within a respective horizontal line of a high definition television ("HDTV") screen, the depicted variation of 49.02 pL to 50.17 pL would thus represent approximately ± 1.2% (instead of the desired maximum tolerance of, e.g., ± 0.5% variation) and thus result in unacceptable quality. Although display technology is cited as an example, it should be understood that nozzle-droplet inconsistency problems can occur in other contexts.

[0006] In FIG. 1A, the nozzles are specifically aligned with the target regions (e.g., wells) such that a particular nozzle prints into a particular target region. In FIG. 1B, an alternative case 151 is shown in which the nozzles are not specifically aligned, but the nozzle density is high relative to the target region density, in which case any nozzles that happen to cross certain target regions during a scan or pass are used to print into those target regions, potentially with several nozzles crossing each target region on each pass. In the example shown, the print head 153 is seen to have five ink nozzles, and the substrate is seen to have two target regions 154-155, each positioned such that nozzles (1) and (2) cross target region 154, nozzles (4) and (5) cross target region 155, and nozzle (3) does not cross any target region. As shown, on each pass, one or two droplets are deposited into each well as depicted. Again, droplets can be deposited in an overlapping manner or at discrete points within each target area; note that the particular illustration of FIG. 1B is merely illustrative; similar to the example presented in FIG. 1A, it is again assumed that it is desired to deposit 50 picoliters (50.00 pL) of fluid into each of target areas 154-155, with each nozzle having a nominal droplet volume of approximately 10.00 pL. Utilizing the same droplet volume variation per nozzle as observed in connection with the example of FIG. 1A, and assuming that each nozzle overlapping a target area in a given pass will deliver droplets into that target area until a total of five droplets have been delivered, it is observed that the target area is filled in three passes, with a total deposited ink volume variation of 0.58 pL from a target of 50.00 pL across the two target areas, a difference outside the specified tolerance, which again may be unacceptable for many applications.

[0007] With respect to the above examples, it is noted that the problem of droplet inconsistency is further exacerbated by the problem that droplet volume may vary statistically even for a given nozzle and a given drive waveform. Thus, in the examples discussed above, it was assumed that nozzle (1) of the printhead from Figures 1A and 1B would produce a droplet volume of 9.80 pL in response to a given drive waveform, but in practical real cases, the droplet volume may be assumed to vary slightly depending on various factors, such as process, voltage, temperature, printhead age, and many other factors, such that the actual droplet volume may not be accurately known.

[0008] Techniques have been proposed to address the problem of droplet inconsistency, but generally speaking, these techniques either still do not reliably provide a fill volume that remains within the desired tolerance range or they significantly increase manufacturing time and cost, i.e., they are inconsistent with the goal of having high quality along with a low consumer price point, which may be important for applications involving commodity products such as HDTVs.

[0009] Therefore, what is needed is a technique useful in depositing fluid into target areas of a substrate using a printhead with a nozzle. More specifically, what is needed is a technique for precisely controlling the amount of fluid deposited in each target area of ​​a substrate despite variations in nozzle and droplet ejection volume, ideally on a cost-effective basis that enables high speed fluid deposition operations and thus improves the speed of device fabrication. The techniques described below meet these needs and provide further related advantages. Summary of the Invention [Means for solving the problem]

[0010] The present disclosure relates to the use of printing processes for transferring layer material to a substrate, techniques for drop measurement with a high degree of accuracy, and related methods, improvements, devices, and systems.

[0011] The nozzle consistency problem introduced above can be addressed by measuring the drop volume per nozzle (or the drop volume variation across nozzles) of a printhead for a given nozzle firing waveform. This allows planning of printhead firing patterns and / or motions to deposit a precise total loading of ink in each target area. With an understanding of how the drop volume varies across nozzles, the printhead / substrate position offset and / or drop firing pattern can be planned in a manner that accommodates the drop volume differences but optimizes simultaneous printing in adjacent target areas with each pass or scan. From a different perspective, rather than normalizing or averaging the nozzle-to-nozzle variation in drop volume, the specific drop volume characteristics of each nozzle are measured and used in a planned manner to simultaneously achieve a total volume within a specific range for each of multiple target areas of the substrate. In many embodiments, this planning is done using a process that reduces the number of scans or printhead passes depending on one or more optimization criteria.

[0012] Several different embodiments that contribute to achieving these results will be presented below, and it is expressly contemplated that each embodiment can be used in isolation and that features of any embodiment can be mixed and matched at will with features of the different embodiments.

[0013] An embodiment presents a system and technique that provides individualized drop measurements across very large printhead assemblies (e.g., with hundreds to thousands of nozzles or more). Using sub-deposition measurement techniques (i.e., by redirecting light away from the vicinity of the printhead beyond the relative distance where the substrate would normally be positioned for deposition), the logistical difficulties associated with positioning the optics are resolved, for example, using an optical assembly that can be articulated in up to three dimensions so that a large printhead assembly (e.g., in a confined space) can be parked at will (e.g., in a printer service station) and the drop measurement device can be precisely articulated relative to the large printhead assembly. The precise placement of the sub-deposition optical assembly allows for drop volume measurements of a packed nozzle array at the required distance from the nozzle plate despite the confined space (printhead assemblies typically operate at about 1 millimeter from the substrate surface). In one optional embodiment, the optical system employs shadowgraphy and repeated measurements of droplets originating from a particular nozzle (and optionally, a variety of nozzle driving waveforms) to increase statistical confidence in the expected droplet volume. In another optional embodiment, the optical system employs interferometry and repeated measurements of droplets originating from a particular nozzle (and optionally, a variety of nozzle driving waveforms) to increase statistical confidence in the expected droplet volume.

[0014] It should be noted that in a production line, it is typically desirable to have as little downtime in production as possible to maximize productivity and minimize manufacturing costs. In another optional embodiment, therefore, the drop measurement time is "hidden" or "stacked" behind other line processes. For example, in an optional flat panel display processing production line, as each new substrate is loaded or otherwise handled, processed, or transferred, the printer's printhead assembly is analyzed using drop measurements to facilitate an accurate statistical understanding of the drop volume per nozzle (and / or per nozzle, per drive waveform). For a printhead assembly with tens of thousands of nozzles, repeated drop measurements (e.g., tens of drop measurements per nozzle, per drive waveform if multiple drive waveforms are used) can take a significant amount of time. Thus, an optional system control process and associated software can optionally perform drop measurements on a dynamic incremental basis. For example, if a virtual loading / unloading process requires, say, 30 seconds and each print takes 90 seconds, the printhead assembly can be measured during the loading / unloading process in two minute cycles, and the droplet measurements are updated to obtain a droplet volume average and confidence interval per nozzle using a sliding window of nozzles / droplets analyzed during the loading / unloading process associated with each two minute cycle. It should be noted that many other processes are possible, and a continuous dynamic process is not required for all embodiments. However, in practice, it is believed that not only will the droplet volume for a given nozzle and driving waveform vary relative to other nozzles and driving waveforms, but also that typical values ​​will change over time due to subtle variations in ink properties, nozzle age and degradation, and other factors. Thus, a process of periodically updating measurements, for example every few hours to days, can advantageously further improve reliability.

[0015] In yet another optional embodiment, the drop measurement system uses interferometry and non-imaging techniques to obtain very fast drop measurements, e.g., measurements per drop in microseconds and repeat drop measurements across a printhead assembly with thousands of nozzles in less than 30 minutes. In contrast to imaging techniques (which use cameras and captured image pixel processing techniques to derive volume measurements), interferometric techniques can provide accurate drop volume measurements by using multiple optical sensors to detect interference pattern spacing representative of the drop shape and by correlating this spacing with the drop volume. In one implementation, the laser source and / or associated optics and / or sensors are mechanically mounted for sub-surface measurements and effective articulation to a large printhead assembly. The very fast measurements obtainable with such systems make interferometric techniques particularly useful in such dynamic incremental measurement embodiments, where tens to hundreds of nozzles can be subjected to repeat drop measurements (e.g., 30 drop measurements per nozzle) to achieve high statistical confidence around each expected drop volume with each print cycle.

[0016] In yet another optional embodiment, many droplet measurements are taken per nozzle, and per nozzle drive waveform (for embodiments using multiple nozzle drive waveforms). As the number of measurements increases, the mean and standard deviation for each nozzle-waveform combination becomes more robust (assuming normal random distribution). Using mathematical processes implemented by the software, a statistical model of each droplet can be created and accurately combined to generate a statistical model of the composite ink fill per target area. To provide an example, many measurements are taken for each nozzle for each drive waveform. If a given single measurement of droplet volume is expected to be accurate to a standard deviation of 2 percent, then by taking many measurements, a statistically accurate average with reduced variance or standard deviation is obtained; that is, again assuming normal random distribution, the standard deviation is reduced by the number of measurements n according to σ / (n)½, such that four measurements of droplet volume reduces the standard deviation by half, etc. Thus, in one embodiment, software is used to achieve a much higher confidence interval around the expected droplet volume through specifically designed repeated measurements that help to substantially reduce measurement error. While many different statistical measures can be used, for example, for an embodiment in which the composite fill is expected to fall within ±x% (e.g., ±0.5% of the target fill), then for each nozzle and for each different drive waveform, drop measurements can be made to ensure that a 3σ (99.73%) confidence interval is obtained around the expected drop volume within the same range (e.g., ±0.5%) of the mean drop volume. Perhaps stated differently, with accurate statistical models constructed for each different drop, known techniques can be used to plan drop combinations based on mathematical combinations of related statistical models to yield a higher degree of precision around the total ink fill per target area (despite nozzle-to-nozzle and waveform-to-wavelength drop volume variations). Note that while a normal random distribution is used in selected embodiments, any statistical model can be used (e.g., Poisson, Student's T, etc.) in which the individual distributions can be combined (e.g., by software) to obtain an aggregate distribution representing the different drop combinations.Also, note that while in some embodiments a 3σ (99.73%) scale is used, other contemplated embodiments use other types of statistical scales, such as 4σ, 5σ, or 6σ, or scales that are not specifically associated with a random distribution.

[0017] It should be noted that similar techniques can be applied to generate models of droplet velocity and flight trajectory for each nozzle-waveform combination. These variables can also be applied in other optional embodiments.

[0018] Any permutation or subset of the techniques and embodiments described above can be applied to precisely plan the total ink fill in a target area, i.e., in a manner that plans a specific composite volume based on drop volume variations per nozzle, i.e., rather than trying to average volume differences across nozzles, these differences are understood and specifically used in the print control process to combine different drops (e.g., from different nozzles or using different drive waveforms) to obtain a very precise ink fill.

[0019] In one optional embodiment, the print head and / or substrate are "staged" in a variable amount to vary the nozzle or nozzles used for each target area on various passes to eject a specific desired drop volume, as appropriate. For example, the print head may be "staged" in a variable amount to the substrate, as appropriate, to vary the nozzle or nozzles used for each target area on various passes to eject a specific desired drop volume. By selectively offsetting the head or printhead assembly, droplets from one nozzle (e.g., with an average droplet volume of 9.95 pL) can be combined with droplets from a second nozzle (e.g., with an average droplet volume of 10.05 pL to obtain a total combined volume of 20.00 pL). Multiple passes are planned so that each target area receives a specific total fill that matches the desired target fill. That is, each target area (e.g., each well in a row of wells that will form a pixelated component of a display) receives a planned combination of one or more droplet volumes to achieve a total volume within a specific tolerance range using different geometric steps of the printhead relative to the substrate. In a more detailed feature of this embodiment, a Pareto-optimal solution can be calculated and applied such that, given the positional relationship of the nozzles relative to each other, a tolerance of volume variation in each target area is allowed within specifications, while at the same time, the printhead / substrate movement is planned to maximize the average simultaneous use of the nozzles for each target deposition area. The statistical techniques discussed above can be used to ensure that the statistical model of the composite (i.e., multiple drop) ink fill falls within any desired tolerance range. In one optional refinement, features are applied to reduce and even minimize the number of printhead / substrate passes required for printing to achieve these objectives. Briefly contemplating these various features, the printing of layers of material on a substrate can be accomplished quickly and efficiently, thus significantly reducing processing costs.

[0020] Note that in a typical application, the target areas that will receive the ink may be aligned, or laid out, in rows or columns, and the swaths represented by the relative printhead / substrate motion will deposit ink in a manner that is within some of all of the rows (of the target areas of the array), but covers all of the columns of the array in one pass. Also, the number of rows, columns, and printhead nozzles may be extremely large, e.g., with hundreds or thousands of rows, columns, and / or printhead nozzles.

[0021] Another optional embodiment addresses the nozzle consistency issue in a slightly different manner. A set of multiple pre-arranged alternative nozzle firing waveforms with known (and different) drop volume characteristics are made available to each nozzle; for example, four, eight, or another number of alternative waveform sets can be hardwired or otherwise predefined to provide a corresponding set of slightly different drop volumes that can be selected. The per-nozzle volume data (or difference data) and any associated statistical models are then used to plan the simultaneous deposition of multiple target areas by determining a set of nozzle-waveform combinations for each target area of ​​the substrate. Again, the specific volume characteristics, confidence intervals, etc. of each nozzle (in this case, each nozzle-waveform combination) and associated distributions are relied upon to achieve a particular fill volume with a high degree of confidence; that is, rather than trying to compensate for the per-nozzle volume variation, the variation is used specifically in combination to obtain a particular fill volume within a well-understood statistical range. Note that typically, to meet these objectives, there will be a large number of alternative combinations that can be used to deposit droplets to reach the desired range in each target area of ​​the substrate. In more detailed embodiments, a "general set" of nozzle waveforms can be shared across several (or even all) nozzles of a printhead, and drop volumes per nozzle are stored and available for mixing and matching different drop volumes to achieve a particular fill. As a further option, a calibration stage can be used to select different waveforms in an offline process (e.g., the dynamic incremental measurement process introduced above), and a particular set of nozzle firing waveforms is selected based on the calibration to achieve each uniquely desired set of volume characteristics. Again, in more detailed embodiments, optimization can be performed to plan printing in a way that improves print time, for example, by minimizing the number of scans or printhead passes, by maximizing simultaneous nozzle usage, or by optimizing some other criteria.

[0022] Yet another embodiment relies on the use of multiple printheads (or equivalently, a printing structure with multiple rows of nozzles that can each be offset relative to one another) in a printhead assembly, where each printhead and its nozzles can be offset relative to one another. Using such deliberate offsets, volume variations per nozzle can be intelligently combined with each pass or scan across the printhead (or row of nozzles). Again, there are typically many alternative combinations that can be used to deposit droplets to reach the desired coverage in each target area of ​​the substrate, and in detailed embodiments, optimization is performed to plan the use of offsets in a way that improves printing time, for example, by minimizing the number of scans or printhead passes, or by maximizing simultaneous nozzle usage, etc.

[0023] It should be noted that one benefit of the techniques described above is that by accepting droplet volume variations but combining them to achieve a specific, pre-determined target area fill volume, a high degree of control can be achieved over the precise volumes and intentionally controlled (or injected) variations in such volumes, as well as the ability to meet desired fill tolerance ranges. The presence of geometric patterns from the deposition process that may result in unevenness or observable patterns can be mitigated through some of the techniques presented herein. That is, even slight differences in target fill volume at low spatial frequencies can be visible to the human eye and thus introduce undesirable, unintentional geometric artifacts. Thus, in some embodiments, it is desired to intentionally, but randomly, vary the specific combination of droplets used to achieve the composite fill volume or composite fill of each target area while still being within specifications. Rather than simply arbitrarily ensuring that all target area fill is at a value within this tolerance range using an exemplary tolerance of 49.75 pL to 50.25 pL, such applications may desire to introduce intentional variations within this range, for example, so that any patterns of variation or difference are not observable to the human eye as patterns in a finished, working display. When applied to color displays, one exemplary embodiment intentionally adds such fill volume variation in a statistically independent manner across at least one of: (a) the x dimension (e.g., along the direction of the rows of target areas), (b) the y dimension (e.g., along the direction of the columns of target areas), and / or (c) across one or more color dimensions (e.g., independently for red vs. blue, blue vs. green, red vs. green target areas). In one embodiment, the variations are statistically independent across each of these dimensions. It is believed that such variations render any fill volume variations imperceptible to the human eye, and thus contribute to the high image quality of such displays.For embodiments that use planned combinations of droplets from different nozzles, generated through a repeatable set of "geometric steps" or offsets in the scan path, it is noted that the use of subtle but deliberate droplet volume variation for each nozzle (i.e., generated through the use of multiple alternative firing waveforms for each nozzle) provides a powerful technique for suppressing the possibility of mura without the need to vary the scan path. In one contemplated embodiment, for example, each nozzle is assigned a set of alternative waveforms that produce a respective average volume within ±10.0% of the ideal volume. The combination of droplets from different nozzles can then be planned according to a precise average (i.e., to achieve a precise intended fill), with mura being suppressed through the use of injected variation in the droplet pattern (either through planned combinations of droplet volumes from different nozzle-waveform pairings, or through waveform variation injected after selection / planning of nozzle-drop combinations to achieve a particular fill). In other embodiments, all with the same intent, intentionally different composite drop volumes can be pre-arranged for each target area to produce a total fill, or different nozzle-droplet combinations can be applied along the scan path, or a non-linear scan path can be used. Other variations are possible.

[0024] Also, traditional droplet measurement techniques take hours or hours and therefore require long measurement times. While possible variations in droplet characteristics during printing can lead to errors in the printing process, the use of high-speed techniques such as interferometric techniques and related structures (introduced above) can facilitate a more up-to-date and therefore more accurate dynamic understanding of nozzle-to-nozzle and droplet volume-to-droplet volume variations, allowing the use of planned combinations as previously described with a high degree of confidence. For example, traditional droplet measurement techniques can take many hours to perform, but through the use of non-imaging techniques (such as interferometry), the droplet measurements can be kept continuously up-to-date and therefore accurately tracked for process, voltage and temperature (PVT variations), printhead nozzle degradation, ink replacement, and other dynamic processes that may affect the accuracy of the measurements. For example, through the use of a rolling measurement process, which hides incremental droplet measurements in substrate loading and unloading times as described above, it is expected that the droplet measurements can be re-performed and updated nearly continuously (e.g., every 3-4 hours or less for each nozzle), thus allowing accurate models to be presented that allow composite fill planning as previously described. In one embodiment, droplets produced by all nozzles or nozzle-waveform pairings are re-measured (e.g., from the beginning) periodically, e.g., once every 2-24 hour period, preferably at a shorter time interval such as 2 hours. Note that the rolling process is not required for all embodiments, i.e., in one embodiment, measurements can be made (or re-made) for all nozzles during a dedicated calibration process during which printing is interrupted. To provide an example, in one possible embodiment, a printhead assembly having 6,000 nozzles and 24,000 nozzle-waveform combinations can be measured over 15 seconds during the substrate loading and unloading phases for each 90 second print cycle as a successive affair, with each iteration examining a different rolling subset of the 24,000 nozzle-waveform combinations until all nozzle-waveform combinations have been processed, and then returning to repeat the process on a cyclical basis.Alternatively, in embodiments that use a dedicated calibration process (e.g., every three hours), such a printhead assembly can be parked for a period of time (e.g., 30 minutes) to generate statistical models for all nozzle-waveform combinations before returning to active printing.

[0025] Again, it should be noted that each of the optional techniques and embodiments introduced above are considered to be mutually optional, and conversely, various embodiments contemplate that such techniques may be optionally combined in any possible permutation or combination. As an example, measurements of droplet velocity and / or flight angle per nozzle / drive waveform may be used to disqualify "errant" droplets for a given nozzle-waveform combination based on a determination that the particular nozzle-waveform combination produces an abnormal droplet "average" or based on a determination that the particular nozzle-waveform combination produces a statistical spread of droplets that exceeds a threshold. To provide another non-limiting example, interferometry or other non-imaging techniques may be used to dynamically update the velocity and / or flight angle behavior by incrementally and dynamically making such measurements at intermittent intervals, i.e., at various windows of nozzle-waveform combinations as the printhead assembly is "parked" during substrate loading and / or unloading. Clearly, many combinations and permutations are possible based on the permutations introduced above. [Brief description of the drawings]

[0026] [Figure 1A] FIG. 1A is a schematic diagram presenting a hypothetical problem of depositing ink in target areas of a substrate in which a printhead with five nozzles is used to deposit a target fill of 50.00 pL in each of five specific target areas. [Figure 1B]FIG. 1B is another schematic diagram presenting a hypothetical problem of depositing ink in target areas of a substrate in which a printhead with five nozzles is used to deposit a target fill of 50.00 pL in each of two specific target areas. [Figure 2A] FIG. 2A is an exemplary diagram showing a drop measurement system capable of measuring drop volume for each nozzle of a large printhead assembly. [Figure 2B] FIG. 2B is a schematic diagram of a method showing various processes and options associated with measuring drop volume for each nozzle. [Figure 2C] FIG. 2C is a schematic diagram of a method showing various process options associated with measuring the drop volume for each nozzle to achieve a reliable understanding of the expected drop volume. [Figure 2D] FIG. 2D is a schematic diagram showing the layout of various components used in one embodiment to perform droplet measurements. [Figure 2E] FIG. 2E is a schematic diagram showing the layout of various components used in another embodiment to perform droplet measurements. [Figure 3A] FIG. 3A provides an illustrative diagram showing a series of optional steps, products, or services that can each independently embody the previously introduced techniques. [Figure 3B] FIG. 3B is an exemplary diagram showing a hypothetical arrangement of a printer and substrates in an application where the substrates will ultimately form a display panel having pixels. [Figure 3C] FIG. 3C is a cross-sectional close-up view of the printhead and substrate of FIG. 3B taken from the perspective of line CC from FIG. 3B. [Figure 4A]4A is a schematic diagram similar to FIG. 1A, but illustrating the use of a combination of drop volumes to reliably generate ink fill for each target area within a predetermined tolerance range. In one optional embodiment, the combination of different drop volumes is generated from a set of predetermined nozzle firing waveforms, and in another optional embodiment, the combination of different drop volumes is generated from each nozzle of the printhead using relative motion (405) between the printhead and the substrate. [Figure 4B] FIG. 4B is a schematic diagram used to illustrate relative printhead / substrate motion and the ejection of combinations of different drop volumes into respective target areas of the substrate. [Figure 4C] FIG. 4C is a diagram used to illustrate the use of different nozzle driving waveforms at each nozzle to generate a combination of different droplet volumes into each target area of ​​the substrate. [Figure 4D] 4D is a schematic diagram similar to FIG. 1B, but illustrating the use of a combination of drop volumes to reliably generate ink fill for each target area within a predetermined tolerance range. In one optional embodiment, the combination of different drop volumes is generated from a set of predetermined nozzle firing waveforms, and in another optional embodiment, the combination of different drop volumes is generated from each nozzle of the printhead using relative motion (472) between the printhead and the substrate. [Diagram 5] FIG. 5 provides a block diagram showing a method for planning droplet combinations for each target area of ​​a substrate, which method can be applied to any of the optional embodiments introduced by FIGS. 4A-D. [Figure 6A] FIG. 6A provides a block diagram for selecting a particular set of allowed droplet combinations for each target area of ​​a substrate, which can be used, for example, with any of the previously introduced embodiments. [Figure 6B] FIG. 6B provides a block diagram for iteratively planning printhead / substrate motion and nozzle usage based on drop combinations for each print area. [Figure 6C]FIG. 6C specifically provides a block diagram illustrating further optimization of print head / substrate motion and nozzle usage to sequence scans in a manner that allows printing to occur as efficiently as possible. [Figure 6D] 6D is a virtual plan view of a substrate that will ultimately produce multiple flat panel display devices, such as 683. As represented by region 687, the print head / substrate motion can be optimized for a specific region of a single flat panel display device, and the optimization is repeated or cyclically used across each display device (such as the four depicted flat panel display devices). [Figure 7] FIG. 7 provides a block diagram for intentionally varying the fill amount within an acceptable tolerance to reduce visual artifacts on a display device. [Figure 8A] FIG. 8A provides a block diagram showing how drop measurements can be used to accommodate statistical variations in drop volume per nozzle and per drive waveform, yet still allow for precise total ink filling within a given target area. [Figure 8B] FIG. 8B provides a block diagram showing how drop measurements can be designed to accommodate statistical variations in drop volume per nozzle and per drive waveform, while still allowing for precise total ink filling within a given target area. [Figure 9A] FIG. 9A provides a graph showing the variation in target area fill volume without adjustment for nozzle-to-nozzle drop volume variation of the printhead. [Figure 9B] FIG. 9B provides a graph showing the variation in target area fill volume where different nozzles are randomly used to statistically compensate for nozzle-to-nozzle drop volume variation in the printhead. [Figure 9C] FIG. 9C provides a graph illustrating variation in target area fill, where one or more droplets of different volumes are used to achieve target area fill within precise tolerances on a planned basis. [Figure 10A]FIG. 10A provides a graph showing the variation in target area fill volume without adjustment for nozzle-to-nozzle drop volume variation of the printhead. [Figure 10B] FIG. 10B provides a graph illustrating the variation in target area fill volume where different nozzles are randomly used to statistically compensate for nozzle-to-nozzle drop volume variation in the printhead. [Figure 10C] FIG. 10C provides a graph illustrating variation in target area fill, where one or more droplets of different volumes are used to achieve target area fill within precise tolerances on a planned basis. [Figure 11] Figure 11 shows a plan view of a printer used as part of a processing apparatus. The printer may be in a gas enclosure, allowing printing to occur in a controlled atmosphere. [Figure 12] 12 provides a block diagram of a printer. Such a printer can optionally be employed in, for example, the converting apparatus depicted in FIG. [Figure 13A] FIG. 13A shows an embodiment in which multiple printheads (each with a nozzle) are used to deposit ink on a substrate. [Figure 13B] FIG. 13B illustrates rotation of multiple printheads to better align the nozzles of each printhead with the substrate. [Figure 13C] FIG. 13C illustrates offsetting of individual printheads of a plurality of printheads in conjunction with intelligent scanning to purposefully produce specific drop volume combinations. [Figure 13D] FIG. 13D shows a cross-sectional view of a substrate including layers that can be used in an organic light emitting diode (OLED) display. [Figure 14A] FIG. 14A illustrates several different ways to customize or vary the nozzle firing waveform. [Figure 14B] FIG. 14B illustrates a method for defining a waveform according to discrete waveform sections. [Figure 15A]FIG. 15A illustrates an embodiment in which different combinations of predetermined nozzle firing waveforms can be used to achieve different drop volume combinations. [Figure 15B] Figure 15B shows circuitry associated with generating and applying programmed waveforms to the nozzles of a print head at programmed times (or locations), providing, for example, one possible implementation of each of circuits 1523 / 1531, 1524 / 1532, and 1525 / 1533 from Figure 15A. [Figure 15C] FIG. 15C illustrates a flow diagram of an embodiment using a different nozzle firing waveform. [Figure 15D] FIG. 15D shows a flow diagram associated with nozzle or nozzle waveform qualification. [Figure 16] FIG. 16 shows a perspective view of an industrial printer. [Figure 17] FIG. 17 shows another perspective view of the industrial printer. [Figure 18A] FIG. 18A presents a schematic diagram showing the layout of components in an embodiment of a shadowgraphy-based droplet measurement system. [Figure 18B] FIG. 18B presents a schematic diagram showing the layout of components in an embodiment of an interferometry-based droplet measurement system. [Figure 19] FIG. 19 shows a flow diagram associated with one exemplary process of integrating a drop measurement system with an industrial printer, optionally for use in OLED device fabrication. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0027] The subject matter defined by the recited claims may be better understood by reference to the following detailed description, which should be read in conjunction with the accompanying drawings. This description of one or more specific embodiments, which are designed below to enable construction and use of various implementations of the technology described by the claims, is not intended to limit the recited claims, but rather to illustrate their applications. Without limiting the foregoing, the present disclosure provides several different examples of techniques used to fabricate material layers by planning printhead movement to maintain deposited ink volume within a given tolerance, while not excessively increasing the number of printhead passes (and thus the time required to complete a deposited layer). In conjunction with these techniques, precise drop measurements can be made to accurately plan composite ink fill in any target area, and the measurements are highly integrated with production printing. The various techniques can be embodied as software for performing these techniques, in the form of a computer, printer, or other device that executes such software, in the form of control data (e.g., a print image) for forming material layers, as a deposition mechanism, or in the form of an electronic or other device (e.g., a flat panel device or other consumer end product) that is fabricated as a result of these techniques. Although specific examples are presented, the principles described herein may be applied to other methods, devices, and systems.

[0028] (Detailed Description) An example will help introduce some concepts regarding intelligent planning of fill volume per target area. The per-nozzle volume data (or differential data) for a given nozzle firing waveform can be used to plan the simultaneous deposition of multiple target areas by determining a set of possible nozzle-droplet volumes for each target area of ​​a substrate. Typically, there will be many possible combinations of nozzles and / or drive waveforms that can deposit ink droplets in multiple passes to fill each target area to the desired fill volume within a narrow tolerance range that meets the specifications. Returning briefly to the hypothesis introduced using Figure 1A, if the allowable fill volume according to specification is between 49.75 pL and 50.25 pL (i.e., within 0.5% of the target), then, without limitation, (a) five passes through nozzle 2 (10.01 pL) for a total of 50.05 pL, (b) one pass through nozzle 1 (9.80 pL) and four passes through nozzle 5 (10.03 pL) for a total of 49.92 pL, (c) one pass through nozzle 3 (9.89 pL) and four passes through nozzle 5 (10.03 pL) for a total of 50.01 pL, and (d) one pass through nozzle 6 (9.89 pL) and four passes through nozzle 7 (10.03 pL) for a total of 50.01 pL. Acceptable fill volumes can also be achieved using many different sets of nozzles / passes, including (d) four passes through nozzle 3 (9.89 pL), three passes through nozzle 4 (9.96 pL), and one pass through nozzle 5 (10.03 pL) for a total of 49.80 pL, and (e) one pass through nozzle 2 (10.01 pL), two passes through nozzle 4 (9.96 pL), and two passes through nozzle 5 (10.03 pL) for a total of 49.99 pL. Other combinations are clearly possible as well. The droplet measurement techniques introduced above can be used to obtain these expected (e.g., average) droplet volumes despite the relatively large statistical tolerances (e.g., ±2% of volume) associated with single droplet measurements. Thus, even if only one choice of nozzle driving waveforms were available for each nozzle (or all nozzles), it would be possible to apply as many nozzles as possible during each scan to deposit droplets (e.g., in different target areas), but with a series of planned offsets or "multiple" waveforms that combine the deposited droplets for each target area in a specifically intended manner. The first embodiment introduced above can be used to offset the printhead relative to the substrate in "geometric steps." That is, many combinations of nozzles and droplet volumes in this assumption can be used to achieve a desired fill volume within a well-understood range of statistical variance consistent with specification tolerances, and the specific embodiment effectively selects a particular one of the allowable droplet combinations for each target region (i.e., a particular set for each region) through its selection of scanning motion and / or nozzle driving waveforms to facilitate simultaneous filling of different rows and / or columns of the target region using each nozzle. By selecting a pattern of relative printhead / substrate motion in a manner that minimizes the time over which printing occurs, this first embodiment provides substantially enhanced manufacturing throughput. Note that this enhancement can be embodied, optionally, in a form that minimizes the number of printhead / substrate scans or "passes," in a manner that minimizes the raw distance of relative printhead / substrate movement, or in a manner that otherwise minimizes overall printing time. That is, print head / substrate movement (e.g., scanning) can be planned in advance and used to fill a target area in a manner that meets predefined criteria, such as minimum print head / substrate passes or scans, minimum print head and / or substrate movement in one or more defined dimensions, printing in a minimum amount of time, or other criteria.

[0029] All of the same approaches apply equally to the hypothetical of FIG. 1B, where the nozzles are not specifically aligned with each target area. Again, if the allowable fill volume according to specifications is 49.75 pL to 50.25 pL (i.e., within 0.5% on either side of the target), the allowable fill volume can also be achieved with many different sets of nozzles / passes, including, but not limited to, all of the examples described above for FIG. 1A, as well as additional examples specific to the hypothetical of FIG. 1B, where two adjacent nozzles are used in one pass to fill a particular target area, e.g., two passes of nozzle (4) (9.96 pL) and nozzle (5) (10.03 pL) for a total of 49.99 pL, and one pass of nozzle (2) (10.01 pL). Again, each such volume can be equated to a statistical average based on many droplet measurements. For example, if nozzles (4), (5), and (2) in this example were associated with a statistical model featuring a stated average and a 3σ value of 0.5% or less of the stated average, the total fill would also generally have a 3σ value equal to or less than ±0.5% of 49.99 pL, meeting the specified tolerance with a high degree of statistical precision. Note that for high resolution OLED displays (i.e., involving millions of pixels), a 3σ (99.73%) value that closely matches the fill tolerance may be insufficient, e.g., statistically indicating that potentially thousands of pixels may still be outside the desired tolerance, and for this reason, in many embodiments, a larger spread measure (e.g., 6σ) is matched to the composite fill tolerance, effectively ensuring that virtually every pixel of the high resolution display meets the manufacturer's specifications.

[0030] These same principles also apply to embodiments of multiple drive waveforms per nozzle. For example, in the hypothetical example presented by FIG. 1A, each of the nozzles can be driven by five different firing waveforms, identified as firing waveforms AE, where the resulting volume characteristics of the different nozzles for the different firing waveforms are illustrated by Table 1A below. Given a single target area 104 and a single nozzle (1), it would be possible to deposit a 50.00 pL target in five passes, for example, with a first printhead pass using predefined firing waveform D (to generate 9.96 pL drops from nozzle (1)), and four subsequent passes using predefined firing waveform E (to generate 10.01 pL drops from nozzle (1)), all without any offset of the scan path. Similarly, different combinations of firing waveforms can be used simultaneously in each pass for each nozzle to generate volumes within each of the target areas close to the target values, without any offset of the scan path. [Table 1A]

[0031] All of these same principles apply equally to the hypothetical of Figure 1B. For example, considering only one target area 154 and nozzles (1) and (2) (i.e., two nozzles that overlap target area 154 during scanning), it is possible to achieve 50.00 pL in three passes, for example, with a first printhead pass using nozzle (1) and predefined waveform B (for a drop volume of 9.70 pL) and nozzle (2) and predefined waveform C (for a drop volume of 10.10 pL), a second printhead pass using nozzle (1) and predefined waveform E (for a drop volume of 10.01 pL) and nozzle (2) and predefined waveform D (for a drop volume of 10.18 pL), and a third printhead pass using nozzle (1) and predefined waveform E (for a drop volume of 10.01 pL).

[0032] Note that in both the hypothetical of FIG. 1A and the hypothetical of FIG. 1B, it is possible to deposit each target volume in a single row of target areas in one pass. For example, it would be possible to rotate the print head 90 degrees and deposit exactly 50.00 pL with a single drop from each nozzle to each target area in a row using, for example, waveform (E) on nozzle (1), waveform (A) on nozzles (2), (4), and (5), and waveform (C) on nozzle (3) (10.01 pL+10.01 pL+9.99 pL+9.96 pL+10.03 pL=50.00 pL). It may also be possible to deposit all of the droplets required to achieve the target volume in one pass without rotating the print head. For example, nozzle (1) may be able to dispense one droplet using waveform D and four droplets from waveform E into region 104 in a single pass.

[0033] These same principles also apply to the printhead offset embodiment introduced above. For example, for the hypothetical example presented by FIG. 1A, the volume characteristics can reflect the nozzles for a first printhead (e.g., “printhead A”) that is integrated with four additional printheads (e.g., printheads “B”-“E”), each driven by a single firing waveform and having a respective drop volume characteristic per nozzle. As the printheads perform a scan pass, each of the nozzles identified as nozzles (1) for the printhead is aligned to print into a target area (e.g., target area 104 from FIG. 1A), each of the nozzles identified as nozzles (2) from the various printheads is aligned to print into a second target area (e.g., target area 105 from FIG. 1A), and so on, with the volume characteristics of the different nozzles for the different printheads being collectively organized as described by Table 1B below. Optionally, each printhead can be offset from one another, for example, using a motor to adjust the spacing between scans. Considering only one target area 104 and nozzle (1) on each printhead, it would be possible to deposit 50.00 pL in four passes, for example, with a first printhead pass in which both printhead D and printhead E fire droplets into the target area, and three subsequent passes in which only printhead E fires droplets into the target area. For example, it would be possible to deposit 50.00 pL in four passes, close to the 50.00 pL target, within the range of 49.75 pL to 50.25 pL. Other combinations are possible using fewer passes that can still produce a larger volume in the target area. Again, considering only one target area 104 and nozzles (1) on each printhead, it would be possible to deposit 49.83 pL in two passes, for example, with a first printhead pass in which printheads C, D, and E all fire drops into the target area, and a second printhead pass in which printheads D and E both fire drops into the target area. Similarly, different combinations of nozzles from different printheads can be used simultaneously in each pass to produce volumes in each of the target areas that are close to the target values, without any offset in the scan path. Thus, using multiple passes in this manner would be advantageous for embodiments in which it is desired to simultaneously deposit drops in different target areas (i.e., within different rows of pixels, for example). Again, statistical accuracy can be ensured by planning the drop measurements in a manner calculated to obtain the desired statistical characteristics associated with the drop volume per nozzle and / or per drive waveform and associated averages. [Table 1B]

[0034] All of the same approaches apply equally to the assumptions of FIG. 1B. Again, considering only one target area 154, and nozzles (1) and (2) on each printhead (i.e., the nozzles that overlap target area 154 during the scan), it is possible to deposit 50.00 pL in two passes, for example, using a first printhead pass in which printheads C and E fire nozzle (1) and printheads B and C fire nozzle (2), and a second printhead pass in which printhead C fires nozzle (2). It is also possible to deposit 49.99 pL in one pass (clearly within the exemplary target range of 49.75 pL to 50.25 pL with high statistical precision) using a printhead pass in which printheads C, D, and E fire nozzle (1) and printheads B and E fire nozzle (2).

[0035] It should also be apparent that the use of alternative nozzle firing waveforms, optionally combined with scan path offsets, significantly increases the number of drop volume combinations that can be achieved for a given printhead, and these options are further increased by multiple printheads (or equivalently multiple rows of nozzles) as described above. For example, in the hypothetical example conveyed by the discussion of FIG. 1 above, the combination of five nozzles with their respective unique ejection characteristics (e.g., drop volumes) and eight alternative waveforms could provide literally thousands of different sets of possible drop volume combinations. Optimizing the set of nozzle-waveform combinations and selecting a specific set of nozzle-waveform combinations for each target area (or for each row of print wells in the array) allows further optimization of printing according to desired criteria. In embodiments using multiple printheads (or rows of printhead nozzles), the ability to selectively offset these printheads / rows also further enhances the number of combinations that can be applied per printhead / substrate scan. Again, for these embodiments, considering that multiple sets of nozzle-waveform combinations (one or more) may alternatively be used to achieve a particular fill volume, the present embodiments select a particular set of "allowable" sets for each target area, and this selection of a particular set across a target area generally involves multiple That is, by varying parameters to minimize the time over which printing occurs, each of these embodiments facilitates increasing manufacturing throughput, minimizing the number of printhead / substrate scans or "passes" required, the raw distance of relative printhead / substrate movement along a particular dimension, or helping to meet some other criteria.

[0036] Many other processes can be used or combined with the various techniques introduced above, for example, it is possible to "tune" the nozzle drive waveform for each nozzle (e.g., shaping the drive pulse by changing the drive voltage, rise or fall slope, pulse width, decay time, number of pulses used per drop and their respective levels, etc.) to reduce the variation in drop volume per nozzle.

[0037] Although certain applications discussed herein refer to fill volumes in discrete fluid containers or "wells," it is also possible to use the described techniques to deposit "blanket coatings" with large topography onto other structures of a substrate (e.g., onto transistors, vias, diodes, and other electronic components, etc.). In such situations, the fluid ink carrying the layer material (which will, for example, be cured, dried, or hardened in situ to form a permanent device layer) will spread to some extent but will still retain certain properties relative to other target deposition areas of the substrate (taking into account ink viscosity and other factors). In this situation, it is possible to use the techniques herein to deposit blanket layers, such as encapsulation or other layers, with specific localized control of the ink fill volume for each target area, for example. The techniques discussed herein are not limited by the specifically presented applications or embodiments.

[0038] Other variations, advantages, and applications from the techniques introduced above will be readily apparent to those skilled in the art. That is, these techniques can be applied in many different fields and are not limited to the fabrication of display devices or pixelated devices. A printing "well" as used herein refers to any container in a substrate that is adapted to receive deposited ink and therefore has chemical or structural properties adapted to contain the flow of that ink. As exemplified for OLED printing below, this can include situations where each fluid container is adapted to receive a respective amount of ink and / or a respective type of ink, for example, in a display application where the described techniques are used to deposit light-emitting materials of different colors, a successive printing process can be performed for each color using a respective print head and a respective ink, where each process can deposit "every other well" in an array (for example, for all "blue" color components), or equivalently, all wells in a third array (interspersing wells with overlapping arrays for other color components). Each printing well is an example of one possible type of target area. Other variations are possible. Also, note that "row" and "column" are used in this disclosure without implying any absolute direction. For example, a "row" of printing wells can extend the length or width of a substrate, or in another manner (linear or nonlinear), and generally speaking, "row" and "column" will be used herein to refer to directions that each represent at least one independent dimension, although this need not be true in all embodiments. Also, note that since modern printers can use relative substrate / printhead motion with multiple dimensions, the relative movement does not need to be linear in path or velocity, i.e., the printhead / substrate relative motion does not need to follow a straight or even continuous path, or a constant velocity. Thus, a "pass" or "scan" of a printhead relative to a substrate simply refers to the repeated deposition of droplets using multiple nozzles over multiple target areas with relative printhead / substrate motion.However, in many embodiments described below for the OLED printing process, each pass or scan may be a substantially continuous linear motion, with each subsequent pass or scan offset by a geometric step relative to one another and parallel to the next pass or scan. This offset or geometric step may be a difference in the pass or scan start position, average position, end position, or some other type of position offset, and does not necessarily imply parallel scan paths. It should also be noted that various embodiments discussed herein speak of the "simultaneous" use of different nozzles to deposit in different target areas (e.g., different rows of target areas), and that the term "simultaneous" does not require simultaneous droplet ejection, but rather merely refers to the concept that different nozzles or groups of nozzles may be used to fire ink into respective target areas in a mutually exclusive manner during any scan or pass. For example, a first group of one or more nozzles may be fired during a given scan to deposit a first droplet in a first row of fluid wells, while a second group of one or more nozzles may be fired during this same scan to deposit a second droplet into a second row of fluid wells. The term "printhead" refers to an integral or modular device having one or more nozzles used to print (eject) ink towards a substrate. A "printhead assembly," in contrast, refers to an assembly or modular element that supports one or more printheads as a group for common positioning relative to a substrate. Thus, a printhead assembly can, in some embodiments, include only a single printhead, while in other embodiments, such an assembly includes six or more printheads. In some implementations, individual printheads can be offset relative to one another within such an assembly.It is noted that in typical embodiments used in large scale manufacturing processes (e.g., television flat panel displays), the printhead assemblies can be quite large, containing thousands of printing nozzles; depending on the implementation, such assemblies can be large, and the drop measurement mechanisms discussed herein are designed to articulate around such assemblies to obtain per drop measurements. For example, with a printhead assembly having six printheads and about 10,000 or more printing nozzles, the printhead assembly can be parked within the printer in an off-axis (printing) service station for various support operations, including drop measurements.

[0039] With the major components of several different embodiments laid out in this manner, the present disclosure will be organized approximately as follows: Figures 2A-2E will be used to introduce certain drop measurement configurations for imaging large-scale printhead assemblies. These configurations can optionally be integrated with a printer, e.g., a flat panel display fabrication device that prints ink materials that will form permanent thin film layers on a flat panel device substrate. In an optional implementation, these configurations can use three-dimensional articulation of some or all of the optics associated with drop measurement to articulate around a printhead assembly with multiple printheads and thousands of inkjet nozzles, e.g., parked in a printer's service station. Figures 3A-4D will be used to introduce some general principles regarding the nozzle consistency problem, OLED printing / fabrication, and how embodiments address the nozzle consistency problem. These techniques can optionally be used with the drop measurement configurations described. Figures 5-7 will be used to illustrate a software process that can be used to plan the combination of drops for each target area of ​​a substrate. 8A-B are used to illustrate the principles associated with building statistical models of drop volume for each nozzle / waveform combination, and using these models to generate statistical models of total ink fill for each target area. These principles can be used, optionally in conjunction with drop measurements (i.e., through the use of planned drop combinations) to ensure that composite ink fills that meet a specified tolerance range with quantifiable certainty (e.g., 99% or better confidence for a target area) despite nozzle consistency issues. FIGs. 9A-10C are used to present some empirical data that demonstrate the effectiveness of the planned drop combination technique described above in improving the consistency of target area fill. FIGs. 11-12 will be used to discuss an example application to OLED panel processing and associated printing and control mechanisms.Figures 13A-13C will be used to discuss print head offsets that can be used to vary the combination of drops that can be deposited in each scan. Figures 14A-15D will be used to further discuss different alternative nozzle firing waveforms that can be applied to provide different drop volumes or combinations. Figures 16-17 will provide additional details about the structure and configuration of an industrial printer, including a drop measurement device. Figures 18A and 18B will be used to discuss one detailed embodiment of a drop measurement system, for example, integrated with such an industrial printer. Finally, Figure 19 will be used to discuss techniques for hiding drop measurement time behind other system processes to maximize production time.

[0040] 2A-2E are generally used to introduce techniques for droplet per nozzle measurement.

[0041] More specifically, FIG. 2A provides an illustrative diagram depicting an optical system 201 and a relatively large printhead assembly 203 having multiple printheads (205A / 205B) each with a large number of individual nozzles (e.g., 207) of which there are hundreds to thousands of nozzles. An ink supply (not shown) is fluidly connected to each nozzle (e.g., nozzle 207), and a piezoelectric transducer (also not shown) is used to eject droplets of ink under control of an electronic control signal per nozzle. The nozzle design maintains a slight negative pressure of ink at each nozzle (e.g., nozzle 207) to avoid flooding the nozzle plate, and an electronic signal for a given nozzle is used to activate a corresponding piezoelectric transducer, pressurizing the ink for the given nozzle and thereby ejecting a droplet from the given nozzle. In one embodiment, the control signal for each nozzle is normally 0 volts, and a positive pulse or signal level at a given voltage is used for a particular nozzle to eject a droplet for that nozzle (one per pulse). In another embodiment, different tuned pulses (or other more complex waveforms) can be used for each nozzle. However, in relation to the example provided by FIG. 2A, assume that it is desired to measure the volume of droplets produced by a particular nozzle (e.g., nozzle 207) that are ejected downward from the printhead (i.e., in a direction "h", representing the z-axis height relative to a three-dimensional coordinate system 208) to be collected by a drain basin 209. In a typical application, the dimension of "h" is typically on the order of 1 millimeter or less, and it is noted that there are thousands of nozzles (e.g., 10,000 nozzles) in an operating printer that cause each droplet to be individually measured in this manner. Thus, in order to optically measure each droplet precisely (i.e., within a measurement window of about 1 millimeter as described, droplets originating from a particular nozzle among thousands of nozzles in a large printhead assembly environment), a technique is used in the disclosed embodiments to precisely position the optical assembly 201, the printhead assembly 203, or both elements relative to each other for optical measurement.

[0042] In one embodiment, these techniques utilize a combination of (a) xy motion control (211A) of at least a portion of the optical system (e.g., in the three-dimensional surface 213) to precisely position the measurement area 215 directly adjacent to any nozzle that is to generate droplets for optical calibration / measurement, and (b) sub-surface optical recovery (211B) (e.g., thereby allowing easy placement of the measurement area next to any nozzle despite the large printhead surface area). Thus, in an exemplary embodiment having about 10,000 or more printing nozzles, the motion system controls the motion of each nozzle of the printhead assembly. It is possible to position at least a portion of the optical system at as many as 10,000 discrete locations (for example) in close proximity to the ejection path. As discussed below, two considered optical measurement techniques include shadowgraphy and interferometry. With each, the optics are typically adjusted in position so that a precise focus is maintained on the measurement area to capture the droplet in flight (e.g., in the case of shadowgraphy, to effectively image the shadow of the droplet). Note that since a typical droplet can be on the order of a few microns in diameter, the optical arrangement is typically extremely precise, presenting challenges with respect to the relative positioning of the print head assembly and the measurement optics / measurement area. In some embodiments, to aid in this positioning, optics (mirrors, prisms, etc.) are used to orient the light capture path for sensing below the dimensional surface 213 arising from the measurement area 215, so that the measurement optics can be positioned near the measurement area without interfering with the relative positioning of the optical system and the print head. This allows for effective position control in a manner that is not limited by the millimeter-order pile height h within which the droplet is imaged, or the large x and y width occupied by the printhead being monitored. With interferometry-based droplet measurement techniques, separate light beams incident on the droplet from different angles create interference patterns that are detectable from a perspective approximately perpendicular to the optical path, and thus the optics in such systems capture light from angles approximately 90 degrees off the path of the source beam, but in a manner that utilizes sub-plane optical recovery to measure droplet parameters. Other optical measurement techniques can also be used. In yet another variant of these systems, the motion system 211A is made to be an xyz motion system, which optionally and advantageously allows for selective engagement and disengagement of the droplet measurement system without moving the printhead assembly during droplet measurement.Briefly, in an industrial processing device having one or more large printhead assemblies, it is contemplated that in order to maximize manufacturing uptime, each printhead assembly will be "parked" at times at a service station to perform one or more maintenance functions, and given the sheer size of the printhead and the number of nozzles, it may be desirable to perform multiple maintenance functions at once on different parts of the printhead. To this effect, in such an embodiment, it may be advantageous to move the measurement / calibration device around the printhead rather than the other way around. [This then allows for the engagement of other non-optical maintenance processes, e.g., on different nozzles if desired.] To facilitate these operations, the printhead assembly may optionally be "parked" with a system that identifies a particular nozzle or set of nozzles that are subject to optical calibration. Once the printhead assembly or a given printhead is stationary, the motion system 211A is engaged to move at least a portion of the optical system relative to the "parked" printhead assembly to precisely position the measurement area 215 at a location suitable for detecting droplets ejected from a particular nozzle, the use of the z-axis of motion allowing selective engagement of the light recovery optics from well below the face of the printhead to facilitate other maintenance operations instead of or in addition to optical calibration. Perhaps described differently, the use of the xyz motion system allows selective engagement of the drop measurement system independent of other tests or test devices used in the service station environment. This structure is not required for all embodiments, and it should be noted, for example, in connection with Figures 16-17 below, that a mechanism will be described that allows for motion of both the measurement assembly and the printhead assembly, e.g., z-axis motion of the printhead assembly relative to the measurement assembly having xy motion for purposes of droplet measurement. Other alternatives are possible where only the printhead assembly moves and the measurement assembly is stationary, or where parked printhead assembly is not required.

[0043] Generally speaking, the optics used to measure the droplets will include a light source 217, an optional set of light delivery optics 219 (which direct light from the light source 217 to the measurement area 215 as needed), one or more light sensors 221, and a set of recovery optics 223 that directs light used to measure the droplets from the measurement area 215 to the one or more light sensors 221. The stem 211A moves any one or more of these elements along with the drainage fountain 209 in a manner that allows for directing light post-droplet measurement from the measurement area 215 around the drainage fountain 209 to a sub-surface location, while optionally also providing a container (e.g., drainage fountain 209) to collect the ejected ink. In one embodiment, the light delivery optics 219 and / or light recovery optics 223 use mirrors that direct light to / from the measurement area 215 along a vertical dimension parallel to the drop travel, and a motion system moves each of the elements 217, 219, 221, 223, and drainage fountain 209 as an integral unit during drop measurement. This setup offers the advantage that the focus does not need to be recalibrated to the measurement area 215. As described by numeral 211C, light delivery optics are also optionally used to provide source light from a location below the dimensional surface 213 of the measurement area, with both light source 217 and light sensor 221 directing light on either side of the spittoon 209 for measurement purposes, for example as generally shown. As described by numerals 225 and 227, the optical system may optionally include a lens for focusing purposes, as well as a photodetector (e.g., for non-imaging techniques that do not rely on the processing of multi-pixelated "photos"). Again, note that the optional use of z-motion control for the optical assembly and spittoon allows for optional engagement and disengagement of the optical system, as well as precise movement of the measurement area 215 proximate to any nozzle at any time while the printhead assembly is "parked." Such parking of the printhead assembly 203 and xyz movement of the optical system 201 are not required for all embodiments. For example, in one embodiment, laser interferometry is used to measure droplet properties, and the printhead assembly (and / or optical system) is moved either in or parallel to the deposition plane (e.g., in or parallel to plane 213) to image droplets from the various nozzles. Other combinations and permutations are possible.

[0044] FIG. 2B provides a process flow associated with drop measurement for some embodiments. This process flow is generally designated using numeral 231 in FIG. 2B. More specifically, in this particular process, as used by reference numeral 233, the printhead assembly is first parked, for example, at a printer or deposition apparatus service station (not shown). A drop measurement device is then engaged with the printhead assembly (235), for example, by selective engagement of a portion or all of the optical system, through movement from below the deposition surface to a position where the optical system is able to measure individual drops. This relative movement of one or more optical system components with respect to the parked printhead, as per numeral 237, can optionally occur in the x, y, and z dimensions.

[0045] As previously alluded to, even a single nozzle and associated nozzle firing drive waveform (i.e., pulse or signal levels used to eject a droplet) can produce droplet volumes, trajectories, and velocities that vary slightly from droplet to droplet. In accordance with the teachings herein, in one embodiment, a droplet measurement system, as indicated by numeral 239, obtains n measurements of a desired parameter per droplet to derive a statistical confidence level regarding the expected nature of that parameter. In one implementation, the measured parameter may be a quantity, while for other implementations, the measured parameter may be flight speed, flight trajectory, or another parameter, or a combination of multiple such parameters. In one implementation, "n" may be different for each nozzle, while in another implementation, "n" may be a fixed number of measurements made for each nozzle (e.g., "24"), and in yet another implementation, "n" refers to a minimum number of measurements such that additional measurements can be made to dynamically adjust the measured statistical nature of the parameter or to refine the confidence level. Clearly, many variations are possible. For the example provided by FIG. 2B, assume that the drop volume has been measured to obtain an accurate average that represents the expected drop volume from a given nozzle and a tight confidence interval. This allows for optional planning of drop combinations (using multiple nozzles and / or driving waveforms) while ensuring that the composite ink fill is distributed in the target area around the expected target (i.e., relative to the composite volume of the drop average). As described by optional process boxes 241 and 243, ideally an optical measurement process is considered for interferometry or shadowgraphy that allows instantaneous or near-instantaneous measurement and calculation of volume (or other desired parameters); such rapid measurements allow for frequent and dynamic updates of volume measurements, for example, to account for changes in ink properties (including viscosity and materials of construction), temperature, power supply variations, and other factors over time.With this in mind, while shadowgraphy typically features the capture of an image of a droplet, for example using a high-resolution CCD camera as the light sensor mechanism, and can accurately image a droplet in a single image capture frame at multiple locations (e.g., using a strobe light source), image processing software typically involves a finite amount of time to calculate the droplet volume, such that imaging a sufficient droplet population from a large printhead assembly (e.g., with thousands of nozzles) can take several hours. Interferometry, which relies on multiple dual-photodetectors and the detection of interference pattern spacing based on the output of such detectors, is a non-imaging technique (i.e., does not require image analysis), and therefore produces droplet volume measurements many times faster (e.g., 50 times faster) than shadowgraphy or other techniques. For example, with a 10,000 nozzle printhead assembly, a large measurement population for each of thousands of nozzles can be obtained in minutes, which is expected to make frequent and dynamic droplet measurements feasible. As previously mentioned, in one optional embodiment, drop measurements (or measurements of other parameters such as trajectory and / or velocity) can be performed as a periodic intermittent process, with the drop measurement system engaged according to a schedule, or stacked between substrates (e.g., as substrates are loaded or unloaded), or for other assembly and / or other printhead maintenance processes. Note that for embodiments that allow alternative nozzle drive waveforms to be used in a manner specific to each nozzle, a high speed measurement system (e.g., an interferometer system) can readily enable the generation of statistical populations for each nozzle, and for each alternative drive waveform for that nozzle, thereby facilitating planned drop combinations of drops generated by various nozzle-waveform pairings, as previously suggested. Numbers 245 and 247 allow for very precise droplet combinations per target deposition area to be planned by measuring the expected droplet volume for each nozzle (and / or for each nozzle-waveform pairing) to an accuracy of better than 0.01 pL; composite fills can also be planned to 0.01 pL resolution, keeping the target volume within a specified error (e.g., tolerance) of 0.5% of the target volume or better.As indicated by numeral 247, a measurement ensemble for each nozzle or each nozzle-waveform pairing is planned, in one embodiment, to generate a confidence distribution model for each such nozzle or nozzle-waveform pairing, i.e., with a 3σ confidence (or other statistical measure such as 4σ, 5σ, 6σ, etc.) less than the specified maximum fill error. Once enough measurements have been made for various drops, the fill with these drop combinations can be evaluated and used to plan the printing (248) in the most efficient manner possible. As indicated by separation line 249, the drop measurements can be made by intermittently switching back and forth between the active printing process and the measurement and calibration process.

[0046] FIG. 2C illustrates one possible process flow 251 associated with planning the droplet measurements per nozzle (per nozzle-waveform pairing) and / or initializing statistical data with which to model the behavior of each nozzle. As indicated by numeral 253, in this process, data is first received that specifies a desired tolerance range, which may be established, for example, according to a manufacturer's specifications. In one embodiment, for example, this tolerance or tolerance range may be specified as ±5.0% of a given target, while in another embodiment, another range may be used, such as ±2.5%, ±2.0%, ±1.0%, ±0.6%, or ±0.5% of the desired target droplet size. It is also possible to specify a range or set of tolerance values ​​in alternative manners. Depending on the desired tolerance and the droplet system measurement error, Regardless of the method of specification, a threshold number of measurements is then identified (255). As indicated above, it should be noted that this number can be selected to achieve several objectives, namely, (a) obtaining a large enough population of drop measurements to provide a reliable measure of the expected drop parameters (e.g., average volume, velocity, or trajectory); (b) obtaining a large enough population of drop measurements to model the variation of the drop parameters (e.g., standard deviation or σ for a given parameter); and / or (c) obtaining sufficient data to identify nozzles or nozzle-waveform pairings with larger than expected errors for the purpose of disqualifying the particular nozzle / nozzle-waveform pairing from use during the printing process. With any planned number of drop measurements, or a desired metric or associated minimum thus defined, measurements are then taken (257) using a drop measurement system 259 (e.g., using optical techniques as discussed herein). Measurements are then taken for each nozzle (or nozzle-waveform) by process decision block 261 until a particular criterion is met. If the number of measurements meets the planned criteria, the method ends via process block 269. If additional measurements need to be taken, the measurement process loops until enough measurements are obtained, as shown in FIG.

[0047] FIG. 2C illustrates some exemplary process variations. First, as indicated by numeral 263, the measurement process is optionally applied to all nozzles (and / or all possible nozzle / waveform combinations) of the printhead assembly. This need not be true for all embodiments. For example, in one embodiment (see discussion of FIGS. 14A-15C below), a potentially infinite number of drive waveform variations can be used to affect the parameters of the ejected droplets for a given nozzle. Instead of testing each possible waveform exclusively, the droplet measurement process can experiment with a predefined set of waveforms, representing a wide distribution of possible waveforms, with an iterative interpolation search process used to select a small number of waveforms (e.g., likely to produce average droplet volumes that span a range of ±10% of the desired droplet size). In another embodiment, if a given nozzle is deemed defective (e.g., droplet volumes with a distribution of more than 20% from the desired average) based on initial measurements, that nozzle (or nozzle-waveform pairing) can be optionally eliminated from further consideration. In yet another example, in practice, when a print scan is planned that does not use certain nozzles, it may be advantageous to perform dynamic, additional drop measurements only on nozzles that are actively used in the planned scan, at least until some type of error or variance criterion is reached. Again, many possibilities exist, and function block 263 simply indicates that the applied process does not have to involve all nozzles (or nozzle-waveform pairings). Secondly, numeral 265 indicates that in one embodiment, the minimum criterion may involve a minimum threshold, which may be different for each nozzle or nozzle-waveform pairing. To cite a few examples of this functionality, in one embodiment, drop measurements are performed on a given nozzle or nozzle-waveform pairing, a distribution spread measure (e.g., variance, standard deviation, or another measure) is calculated, and measurements are made above the raw threshold until the spread measure meets a predetermined criterion.As should be understood, if the minimum is, for example, 10 droplet measurements per nozzle, and if 10 droplet measurements for a particular nozzle result in a larger than expected variance, additional measurements can be made uniquely for a given nozzle until a desired spread is achieved (e.g., 3σ≦1.0% of the mean amount) or until some maximum number of measurements have been made. Such an embodiment can result, for example, in a different number of measurements per nozzle, i.e., measurement iterations are scheduled to achieve some minimum criteria (e.g., a minimum number of measurements and a spread measure below a threshold in this example). Third, as indicated by numeral 267, it is also possible to use dead reckoning in the droplet measurement schedule, for example, to obtain "exactly 24" droplet measurements per nozzle (or nozzle-waveform), or to obtain x measurements per hour, etc. Finally, regardless of the measurement control technique, measurements can be applied to qualify or disqualify a nozzle or nozzle-waveform combination. Again, to cite possible implementation options, following the performance of a threshold number of measurements, numeral 270 may qualify or disqualify a nozzle or nozzle waveform based on the measured data. For example, if the ideal drop volume is 10.00 pL for an application, then any nozzle / nozzle waveform pairing that does not produce an average drop volume between 9.90 pL and 10.10 pL may be immediately disqualified. The same approach may be taken for statistical spread, e.g., following a minimum number of measurements, any nozzle / nozzle waveform pairing that produces a drop spread (e.g., variance, standard deviation, etc.) greater than 0.5% may be immediately disqualified. Again, many implementation examples exist.

[0048] FIG. 2D is a schematic diagram of one implementation of a drop measurement system foreseen with optical techniques, generally referenced by numeral 271. More specifically, a printhead 273 is illustrated in cross section as having five enumerated print nozzles arranged in a row of nozzles that will eject fluid ink downward in the z-direction (as indicated by reference legend 274). A light source 275A is arranged on the side of the printhead to illuminate a measurement area 278 through which droplets will pass for measurement. In the case of FIG. 2D, this measurement area (and part or all of the optical system) is arranged to measure droplets originating from nozzle (3) of the printhead. The light source 275A is depicted external to one side of the printhead 273 to generate an optical path 277 that will direct light into the optical measurement area (i.e., within a height on the order of millimeters represented by the variable h) to illuminate any of the multiple nozzles without interfering with the printhead 273. As represented by numeral 275B, in one embodiment, the light source may also instead be advantageously mounted below the deposition surface 289 (and the upper periphery of the drainage basin 286) to provide relatively easy fixed distance positioning of the optics relative to the droplet path from any nozzle. Again, while five nozzles are depicted in FIG. 2D, in one embodiment, there are hundreds to thousands of nozzles or more. Below-deposition surface light generation, with the optics used to direct illumination to the droplet measurement region 278, facilitates easy positioning of the optical system relative to any nozzle of the depicted printhead 273, as well as for selective engagement and disengagement of the droplet measurement system (e.g., to an optional service station as previously described). In the depicted embodiment, a mirror 285A is used to redirect light from the light source 275B to be incident on droplets in the measurement region 278 traveling from the printhead 273 towards the drainage basin 286. Other means of positioning the optical path relative to the light source 275B may also be used, such as, by way of non-limiting example, a prism, fiber optic cable, etc. For implementations in which imaging measurement techniques are used (eg, shadowgraphy), light sources 275A / 275B may be strobed thermal sources or monochromatic sources.2D also shows light from a third origin location directed along path 275C, where the light source is outside the drawing page and directs light into or out of the drawing page with or without the aid of light path routing optics (e.g., along the y dimension depicted by reference legend 274); note that such a positioning framework can be used when interferometry is relied upon, for example, with detection of interference patterns arising from directions perpendicular to the illumination path (or at another angle thereto). Note that regardless of the relative alignment of the illumination sources, light is directed along light path 277 and illumination surface 290 that are intermediate to the positions of print head 273 and deposition surface 290, and measurement light (i.e., from the measured droplet) is routed by light path routing optics 285B from the imaging surface to a photodetector mounted below deposition surface 289. Again, this allows for narrow directing and focusing of light despite the large print head size and relatively small height h. Also, mirrors, prisms, optical fibers, or other light redirecting devices and techniques can be used to achieve this sub-deposition surface routing of light recovery, similar to the light routing optics 285A. As seen in FIG. 2D, the measurement light is directed to the focusing optics 279 (e.g., a lens) and onto the photodetector 280. The distance of the light path between the focusing optics and the measurement area is identified by a distance f, which represents the focal length of the optical system. As previously suggested, the droplet measurement (depending on the optical technology) provides the precise focusing required to properly image the droplets, and to this effect, for the system depicted by FIG. 2D, it is desired that the light routing optics 285B, focusing optics 279, droplet measurement area 278, and spittoon 286 lens are all moved as an integral unit to measure droplets from different nozzles, as represented by the depicted connections to a common chassis 283. Light sources 275A / 275B and light source directing optics may also optionally be coupled to the chassis, depending on the embodiment.

[0049] Note that in an interferometry-based system, also represented diagrammatically by FIG. 2D, the light source 275A / 275B (or generating optical path 275C) may be a laser, where the beam is split at some point along the optical path into two or more different components that are used to generate an interference pattern. Additional details regarding these optics and the use of multiple beams to create interference patterns will be discussed further below in connection with FIG. 18B. For the time being, it will be assumed that laser sources (including light sources for interferometry) are encompassed by the references 275A / 275B / 275C.

[0050] FIG. 2E illustrates another schematic diagram of an implementation of a droplet measurement system foreseen with optical techniques, generally referred to by numeral 291. More specifically, the implementation seen in FIG. 2E relies on interferometry to measure droplet parameters (such as volume). As before, this configuration relies on a print head 273, a measurement area 278, a chassis 283, and a spittoon 286. However, in this embodiment, a laser is specifically used as a light source 292 to generate a light beam that is directed to the measurement area via an illumination path 293. It should be noted that typically, more than one beam is directed in this manner, as will be further explained below. An interference pattern is generated in the droplet in the measurement area 278, and this interference pattern is observed from a direction substantially perpendicular to the illumination path 293, as represented by numeral 297. This same relationship (measurement from a direction not parallel to the illumination path) was also represented by FIG. 2D (e.g., using path 275C), but in FIG. 2E the divergent measurement angle is such that the measurement light is naturally directed downward below the plane of the measurement area 278. Note that the light detector 295 is non-imaging in the sense that the use of a camera is not required (although typically multiple light detectors are used) and the use of image processing to identify the droplet outline in a pixelated image is not required, substantially increasing the speed of detection and measurement; that is, the interferometric approach simply measures the change in the interference pattern as the droplet passes through a field of simultaneous light beams, and the droplet volume can be derived from the result obtained. The use of more than two light beams (or an increased number of detectors) facilitates the measurement of droplet trajectory and velocity as well as other parameters. As before, the light source 292, the spittoon 286, and the light detector 295 can be moved as one (i.e., using a common chassis 283), facilitating the preservation of precise optical path parameters. In one implementation, movement of the optical system is performed in three dimensions relative to the “parked” printhead assembly, again to selectively engage and disengage the drop measurement device while the printhead assembly is in the service station, as well as to easily and precisely position the drop measurement device to measure any of the thousands of nozzles in the large scale printhead.

[0051] As discussed above, with a suitable configuration of a drop measurement device or system, an industrial printer (e.g., used in OLED device fabrication) can have the nozzles and their resulting drops repeatedly calibrated, allowing for very precise planning of drop combinations in any target area. That is, the measurement device can be used to rapidly generate accurate tightly grouped statistical distributions of volumes for each nozzle and each waveform used for the nozzle, allowing for precise planning of drop combinations used to achieve composite fill. In other embodiments, these same techniques are used to build models for drop velocity and flight angle so that these parameters can be applied in the printing process.

[0052] It should be noted that any of these various techniques (as well as any of the printing or composite filling techniques introduced in this disclosure) can be manifested in different products and / or different manufacturing stages. For example, FIG. 3A depicts several different implementation stages, collectively designated by reference numeral 301, each of which represents a possible discrete implementation of the techniques introduced above. First, the techniques introduced above can be embodied as instructions stored on a non-transitory machine-readable medium, as represented by graphic 303 (e.g., software for controlling a computer or printer). Second, by computer icon 305, the techniques can be implemented as part of a computer or network, for example, within a company that designs or manufactures components for sale or use in other products. For example, the techniques introduced above can be implemented as design software by a company that advises or designs for manufacturers of high definition televisions (HDTVs). Alternatively, the techniques can be used directly by such manufacturers to create televisions (or display screens). Third, as previously introduced and illustrated using storage medium graphic 307, the techniques introduced above can take the form of printer instructions, e.g., as stored instructions or data that, when acted upon, will cause a printer to process one or more layers of a component that are dependent upon the use of a projected droplet coalescence technique, per the above discussion. Fourth, as represented by processing device icon 309, the techniques introduced above can be implemented as part of a processing apparatus or machine, or in the form of a printer within such an apparatus or machine. For example, a processing machine can be sold or customized in a manner where droplet measurements and conversion of externally supplied "layer data" are automatically converted by the machine (e.g., through the use of software) into printer instructions that will print using the techniques described herein to transparently optimize / accelerate the printing process.Such data can also be calculated offline and then reproducibly reapplied in a scalable pipeline manufacturing process that produces many units. Note that the particular depiction of the processing device icon 309 represents one exemplary printer device, which will be discussed below (e.g., with reference to Figs. 11-12). The techniques introduced above can also be embodied as an assembly, such as an array 311 of multiple components that will be sold separately, for example, in Fig. 3, several such components are depicted in the form of an array of semi-finished flat panel devices that will be later separated and sold for incorporation into a final consumer product. The depicted device may have one or more layers (e.g., color component layers, semiconductor layers, encapsulation layers, or other materials), deposited, for example, depending on the methods introduced above. The techniques introduced above can also be embodied in the form of a final consumer product as referenced, for example, in the form of a display screen for a portable digital device 313 (e.g., an electronic pad or smartphone, etc.), as a television display screen 315 (e.g., HDTV), or other types of devices. For example, Figure 3A uses a solar panel graphic 317 to illustrate that the processes introduced above can be applied to other forms of electronic devices, such as depositing structures per target area (such as one or more layers of individual cells that make up an aggregate device) or blanket layers (such as an encapsulation layer for a television or solar panel). Clearly, many implementations are possible.

[0053] The techniques introduced above can be applied to, but are not limited to, any of the steps or components illustrated in FIG. 3A. For example, one embodiment of the techniques disclosed herein is an end consumer device, and a second embodiment of the techniques disclosed herein is an apparatus with data that controls the processing of layers using a specific nozzle volume combination to obtain a specific fill per target area. The nozzle volumes can be determined in advance or measured and applied in situ. Yet another embodiment is a deposition machine, for example using a printer to print one or more inks using the techniques introduced above. These techniques can be applied to one machine or more than one machine, for example using different nozzle volumes where different steps are applied. The present invention may be implemented on a single machine, network of machines, or a series of machines. All such embodiments, and others, may utilize the techniques introduced by this disclosure, either individually or collectively.

[0054] As represented by FIG. 3B, in certain applications, a printing process can be used to deposit one or more layers of material onto a substrate. The techniques discussed above can be used to generate printer control instructions (e.g., electronic control files that can be transferred to a printer) for later use in fabricating a device. In certain applications, these instructions can be geared for an inkjet printing process useful in printing layers of low-cost, scalable organic light-emitting diode ("OLED") displays. More specifically, the described techniques can be applied to deposit one or more emissive or other layers of such OLED devices, such as the "red," "green," and "blue" (or other) pixelated color components or other emissive layers or components of such devices. This exemplary application is non-limiting, and the described techniques can be applied to the fabrication of many other types of layers and / or devices, whether or not these layers are emissive and whether or not the device is a display device. In this exemplary application, various conventional design constraints of inkjet printheads provide challenges to process efficiency and film coating uniformity of the various layers of an OLED stack that can be printed using various inkjet printing systems. These challenges can be addressed through the teachings herein.

[0055] More specifically, FIG. 3B is a plan view of one embodiment of a printer 321. The printer includes a printhead assembly 323 that is used to deposit fluid ink onto a substrate 325. Unlike printer applications that print text and graphics, the printer 321 in this example is used in a manufacturing process to deposit fluid ink that will have a desired thickness. That is, in a typical manufacturing application, the ink carries materials that will be used to form a permanent layer of a finished device, with the layer having a specific desired thickness. The thickness of the layer produced by the deposition of the fluid ink depends on the amount of ink applied. The ink typically features one or more materials that will form part of the finished layer, formed as a monomer, polymer, or material carried by a solvent or other transport medium. In one embodiment, these materials are organic. Following deposition of the ink, the ink is dried, cured, or hardened to form a permanent layer; for example, some applications use an ultraviolet (UV) curing process to convert the liquid monomer to a solid polymer, while other processes dry the ink to remove the solvent and leave the material transported to a permanent location. Other processes are possible. Note that there are many other variations that distinguish the depicted printing process from traditional graphics and text applications, for example, in some embodiments, deposition of the desired material layer is performed in an environment that is controlled to either regulate the surrounding atmosphere to be other than air or otherwise exclude unwanted particulate matter. For example, as described further below, one contemplated application uses a processing mechanism that encloses the printer 321 in a gas chamber so that printing can be performed in the presence of a controlled atmosphere, such as an inert environment, including, but not limited to, nitrogen, any of the noble gases, and any combination thereof.

[0056] As further seen in Figure 3B, printhead assembly 323 includes a number of nozzles, such as nozzle 327. Note that in Figure 3B, for ease of illustration, printhead assembly 323 and nozzles are depicted as opening out from the top of the page, but in reality, these nozzles face downward toward the substrate and are hidden from view from the perspective of Figure 3B (i.e., Figure 3B actually shows what is a cutaway view of printhead assembly 323). The nozzles are arranged in rows and columns (example row 3 It can be seen that the nozzles are arranged in rows (e.g., 328 and 329) although this is not required for all embodiments; i.e., some implementations use only one row of nozzles (e.g., row 328). In addition, the rows of nozzles can be arranged on respective printheads, and each printhead can (optionally) be individually offset relative to each other as introduced above. In applications where the printer is used to fabricate materials for parts of a display device, for example each of the red, green, and blue color components of a display device, the printer typically uses dedicated printhead components for each different ink or material, and the techniques discussed herein can be applied separately to each corresponding printhead or printhead assembly.

[0057] FIG. 3B illustrates the printhead assembly 323 (i.e., one or more individual printheads are not depicted separately). The printer 321 includes two different motion mechanisms that can be used to position the printhead assembly 323 relative to the substrate 325 in this example. First, a traveler or carriage 331 can be used to mount the printhead assembly 323 and enable relative motion as represented by arrow 333. This motion mechanism can also optionally transport the printhead assembly 323 to a service station, if present, such a service station being represented by numeral 334 in FIG. 3B. Second, however, a substrate transport mechanism can be used to move the substrate relative to the traveler along one or more dimensions. For example, as represented by arrow 335, the substrate transport mechanism can enable motion in each of two orthogonal directions, such as according to x and y Cartesian dimensions (337), and can optionally assist with substrate rotation. In one embodiment, the substrate transport mechanism comprises a gas floating table that is used to selectively secure and enable substrate motion on gas bearings. Further, it should be noted that the printer optionally allows for rotation of the printhead assembly 323 relative to the traveler 331, as represented by rotation graphic 338. Such rotation allows the apparent spacing and relative configuration of the nozzles 327 to be changed relative to the substrate, e.g., where each target area of ​​the substrate is defined to be of a particular area or to have a spacing relative to another target area, rotation of the printhead assembly and / or substrate can change the relative separation of the nozzles along or perpendicular to the scan direction. In an embodiment, the elevation of the printhead assembly 323 relative to the substrate 325 can also be changed, e.g., along the z Cartesian dimension, which is in or out of the orientation of the diagram in FIG. 3B.

[0058] The two scan paths are illustrated by directional arrows 339 and 340, respectively, in FIG. 3B. Briefly, the substrate motion mechanism moves the substrate back and forth in the directions of arrows 339 and 340 as the print head moves in a geometric step or offset in the direction of arrow 333. Using a combination of these movements, the nozzles of the print head assembly can reach any desired area of ​​the substrate to deposit ink. As previously referenced, ink is deposited into discrete target areas of the substrate on a controlled basis. These target areas can optionally be aligned, i.e., arranged, in rows and columns, such as along the depicted y and x dimensions, respectively. Note that the rows of nozzles (such as row 328) are viewed in this illustration perpendicular to the rows and columns of target areas, i.e., such that a row of nozzles sweeps with each scan along the direction of the rows of target areas and traverses each of the columns of target areas of the substrate (e.g., along direction 339). This need not be true for all embodiments. For efficiency of motion, a subsequent scan or pass then reverses the direction of motion and addresses the rows of target regions in reverse order, ie, along direction 340.

[0059] The array of target areas in this example is depicted by the highlighted area 341, seen in enlarged view on the right side of the figure. That is, two rows of pixels, each with red, green, and blue color components. Each pixel is represented by numeral 343, while the column of pixels perpendicular to the scan direction (339 / 340) is represented by numeral 345, respectively. In the top left most pixel, it can be seen that the red, green, and blue color components occupy specific target areas 347, 349, and 351 as part of a respective overlapping array of areas. Each color component within each pixel can also have associated electronics, as represented, for example, by numeral 353. If the device being fabricated is a backlit display (e.g., as part of a conventional LCD television), these electronics can control the selective masking of light filtered by the red, green, and blue color regions. If the device being fabricated is a newer kind of display, i.e., where the red, green, and blue color regions directly generate unique colors with corresponding color characteristics, these electronics 353 can include patterned electrodes and other material layers that contribute to the desired light generation and light characteristics.

[0060] FIG. 3C provides a close-up cross-sectional view of printhead 373 and substrate 375 taken for the printhead assembly of FIG. 3B from the perspective of line CC. More specifically, numeral 371 generally represents the printer, while numeral 378 represents a row of print nozzles 377. Each nozzle is designated using a bracketed number, e.g., (1), (2), (3), etc. A typical printhead typically has a number of such nozzles, e.g., 64, 128, or another number, and in one embodiment there may be 1,000 to 10,000 or more nozzles arranged in one or more rows. As previously mentioned, the printhead in this embodiment is moved relative to the substrate to achieve a geometric step or offset between scans in the direction referenced by arrow 385. Depending on the substrate motion mechanism, the substrate may also be moved perpendicular to this direction (e.g., into and out of the page relative to the view of FIG. 3C) and, in some embodiments, in the direction represented by arrow 385. Note that FIG. 3C also shows a row 383 of respective target areas 379 of the substrate, arranged in this case as "wells" that will receive the deposited ink and hold the deposited ink within the structural constraints of each well. For purposes of FIG. 3C, it will be assumed that only one ink is represented (e.g., each depicted well 379 represents only one color of the display, such as a red component, with other color components and associated wells not shown). Note that the drawing is not to scale, e.g., the nozzles can be seen to be numbered from (1) to (16), while the wells are lettered from (A) to (ZZ), representing 702 wells. In some embodiments, the nozzles will be aligned with the respective wells such that the depicted printhead with 16 nozzles uses scanning relative printhead / substrate motion into and out of the page from the perspective of FIG. 3C to deposit ink in as many as 16 wells simultaneously in the direction of arrow 381.In other embodiments, such as those previously described (e.g., with reference to FIG. 1B ), the nozzle density is even greater than the target area density, and with any given scan or pass, a portion of the nozzles (e.g., a group of one to many nozzles, depending on which nozzles traverse each target area) will be used for deposition into each target area. For example, again using the illustrative example of 16 nozzles, nozzles (1)-(3) could be used to deposit ink in a first target area, and nozzles (7)-(10) could be used simultaneously to deposit ink in a second target area, mutually exclusive of each other, for a given pass.

[0061] Conventionally, a printer might be operated to use 16 nozzles to simultaneously deposit ink in as many as 16 rows of wells, moving back and forth as necessary in successive scans until, say, five drops have been deposited in each well, with the printhead being advanced as necessary using fixed steps that are an integer multiple of the width of the swath of locations traversed by the scan. However, the techniques provided by this disclosure allow for the use of multiple nozzles in combinations calculated to produce a particular fill volume for each well. The geometric step takes advantage of the inherent variation in droplet volume that is applied to each well. Different embodiments rely on different techniques to achieve these combinations. In one embodiment, the geometric step is varied to achieve different combinations and can be free to be other than an integer multiple of the width represented by the printhead swath location. For example, if appropriate for depositing a set of selected droplet combinations in each well 379 of FIG. 3C, the geometric step may actually be 1 / 160th of the printhead swath location, which in this example represents a relative displacement between the printhead and the substrate of a tenth of a row of wells. The next offset or geometric step may be appropriately different for the particular combination of droplets desired in each well, for example a virtual offset of 5 / 16th of the printhead swath location, corresponding to an integer spacing of wells. This variation may continue with both positive and negative steps as needed to deposit ink to obtain the desired fill volume. Note that many different types or sizes of offsets are possible and the step size need not be fixed between scans or a particular percentage of the well spacing. However, in many manufacturing applications, it is desirable to minimize print time in order to maximize production rate and minimize manufacturing cost per unit as much as possible. To achieve this goal, in a specific embodiment, printhead motion is planned and sequenced in a manner that minimizes the total number of scans, the total number of geometric steps, the offset or size of the geometric steps, and the cumulative distance traversed by the geometric steps. These or other measures may be used individually, together, or in any desired combination to minimize the total print time.In embodiments where independently offsettable rows of nozzles (e.g., multiple printheads) are used, the geometric step can be represented in part by an offset between printheads or nozzle rows, combined with an overall offset of the printhead components (e.g., a fixed step of the printhead assembly); such an offset can be used to achieve variable sized geometric steps and thus deposit a combination of droplets into each well. In embodiments where variation of the nozzle driving waveform is used alone, a conventional fixed step can be used, with droplet volume variation being achieved using multiple printheads and / or multiple printhead passes. As described below, in one embodiment, the nozzle driving waveform can be programmed for each nozzle between droplets, thus enabling each nozzle to generate and provide a respective droplet volume per well in a row of wells.

[0062] 4A-4D are used to provide additional detail regarding the reliance on specific drop volumes in achieving a desired loading volume.

[0063] 4A presents an illustrative view 401 of a printhead 404 and two associated schematics seen below the printhead 401. Because the printhead is used in an embodiment that optionally provides a non-fixed geometric step of the printhead relative to the substrate, numerals 405 are used to represent offsets that align particular printhead nozzles (e.g., 16 nozzles total with nozzles (1)-(5) depicted in the figure) with different target areas (five in this example: 413, 414, 415, 416, and 417). Recalling the example of FIG. 1A , if nozzles (1)-(16) produce fluid ink drop volumes (e.g., average drop volumes) of 9.80, 10.01, 9.89, 9.96, 10.03, 9.99, 10.08, 10.00, 10.09, 10.07, 9.99, 9.92, 9.97, 9.81, 10.04, and 9.95 pL, respectively, and if it is desired to deposit 50.00 pL per target area plus or minus 0.5 percent of this value, the print head can be used to deposit the drops in 5 passes or scans using geometric steps of 0, −1, −1, −2, and −4, respectively, resulting in (expected average) total fill values ​​per area of ​​49.82, 49.92, 49.95, 49.90, and 50.16 pL, as depicted in the figure. This is clearly within the desired tolerance range of 49.75 pL to 50.25 pL for each of the depicted target areas. All steps in this example are expressed incrementally relative to the previous position, but other scales can also be used. Depending on the expected per drop volume variance, it is still possible to virtually guarantee that the fill will match the desired tolerance range. For example, by making many drop measurements as referenced above (e.g., 20-30 or more drop measurements per nozzle), the expected variance of each drop volume can be made extremely small, allowing a high degree of confidence in the variance of the expected combined volume. Thus, as shown, a combination of drops can be used in a deliberate manner that depends on the respective drop volume and the desired fill for each target area to achieve a precisely regulated fill with a high degree of confidence.

[0064] Note that this same diagram can be used to represent nozzle drive waveform variations and / or the use of multiple printheads. For example, if nozzle references (1)-(16) refer to the drop volume for a single nozzle generated by 16 different drive waveforms (i.e., using waveforms 1-16), one can obtain a theoretical per-area fill volume simply by using different drive waveforms, e.g., waveform numbers 1, 2, 3, 5, and 9, in the target area 413. In practice, since process variations may result in different per-nozzle characteristics, the system would measure the drop volume for each nozzle for each waveform and intelligently plan the drop combinations on this basis. In embodiments where nozzle references (1)-(15) refer to multiple printheads (e.g., references (1)-(5) refer to a first printhead, references (6)-(10) refer to a second printhead, and references (11)-(15) refer to a third printhead), an offset between the printheads can be used to reduce the number of passes or scans. For example, the right-most target area 417 can have three droplets deposited in one pass, including drop volumes of 10.03, 10.09, and 9.97 pL (print head (1), 0 offset; print head (2), +1 offset; and print head (3), +2 offset). It should be apparent that the combination of these various techniques facilitates many possible combinations of droplets of a particular volume to achieve a particular fill volume within the tolerance range. Note in FIG. 4A that the variance in total ink fill volume between the target areas is small and within tolerance, i.e., within the range of 49.82 pL to 50.16 pL.

[0065] FIG. 4B shows an illustrative diagram 421 of a series of printhead scans, each scan perpendicular to the direction of arrow 422, with nozzles represented by different rectangles or bars as referenced by numerals 423-430. In relation to this diagram, assume that the printhead / substrate relative motion is advanced in a series of variable-sized geometric steps. Note again that typically each step will designate a scan that sweeps multiple columns of target areas (e.g., pixels) beyond one column of five regions represented on the plane of the drawing page (and represented by numerals 413-417). The scans are shown in a top-to-bottom order, including a first scan 423 in which the printhead is seen displaced to the right relative to the substrate such that only nozzles (1) and (2) are aligned with target areas 416 and 417, respectively. Within each print scan depiction (such as box 423), a circle represents each nozzle either solid black to represent that the nozzle is to be fired when it covers the specifically depicted target area during the scan, or "hollow" or filled white to represent that the nozzle is not to be fired at the relevant time (but may be fired for other target areas encountered in the scan). Note that in this embodiment, each nozzle is fired on a binary basis, i.e., each nozzle is either fired or not fired according to any adjustable parameters, e.g., to deposit a predetermined drop volume for each target area encountered during the scan. This "binary" firing scheme may optionally be implemented in any of the embodiments described herein (i.e., where multiple firing waveforms are used, e.g., where waveform parameters are adjusted between drops). In one embodiment, a second pass 423 may be employed (in a second embodiment). It can be seen that in the first pass 423, nozzle (1) is fired to deposit a 9.80 pL drop into the second rightmost target area, while nozzle (2) is fired to deposit a 10.01 pL drop into the rightmost target area 417. The scan continues to sweep other columns of target areas (e.g., other rows of pixel wells), depositing ink drops accordingly. After the first pass 423 is completed, the print head is advanced by a geometric step of -3, which moves the print head left relative to the substrate, such that nozzle (1) traverses target area 413 during a second scan 424 in the opposite direction to the first scan. During this second scan 424, nozzles (2), (3), (4), and (5) will also traverse areas 414, 415, 416, and 417, respectively. The filled in circles indicate that nozzles (1), (2), (3), and (5) will be fired at the appropriate times to deposit drop volumes of 9.80 pL, 10.01 pL, 9.89 pL, and 10.03 pL, corresponding to the inherent characteristics of nozzles (1), (2), (3), and (5), respectively. It should also be noted that in any one pass, the nozzles in a row of nozzles used to deposit ink will deposit into their respective target regions in a mutually exclusive manner; for example, for pass 424, nozzle (1) is used to deposit ink into target region 413 (and not into any of target regions 414-417), nozzle (2) is used to deposit ink into target region 414 (and not into any of regions 413 or 415-417), nozzle (3) is used to deposit ink into target region 415 (and not into any of regions 413-414 or 416-417), and nozzle (5) is used to deposit ink into target region 417 (and not into any of regions 413-416).The third scan, represented using the numeral 425, effectively advances the printhead by one row of the target area (-1 geometric step) such that nozzles (2), (3), (4), (5), and (6) traverse areas 413, 414, 415, 416, and 417, respectively, during the scan, and the filled-in nozzle graphic represents that during this pass, each of nozzles (2)-(6) will be actuated to fire a droplet to produce expected droplet volumes of 10.01, 9.89, 9.96, 10.03, and 9.99 pL, respectively.

[0066] If the printing process were stopped at this point, region 417 would have a fill of 30.03 pL (10.01 pL+10.03 pL+9.99 pL), corresponding to, for example, three drops, while region 413 would have a fill of 19.81 pL (9.80 pL+10.01 pL), corresponding to two drops. Note that the scan pattern, in one embodiment, follows the back and forth pattern represented by arrows 339 and 340 in FIG. 3B. Subsequent passes 426-430 of these target regions (or scans of multiple rows of multiple such regions) result in (a) 10.01 pL, 0.00 pL, 0.00 pL, 10.08 pL, and 10.09 pL droplets in region 413, corresponding to passes by nozzles (2), (3), (4), (7), and (9) in successive scans, respectively; (b) 0.00 pL, 0.00 pL, 10.03 pL, 10.00 pL, and 10.07 pL droplets in region 414, corresponding to passes by nozzles (3), (4), (5), (8), and (10), respectively, in successive scans; (c) 0.00 pL, 0.00 pL, 10.03 pL, 10.00 pL, and 10.07 pL droplets in region 415, corresponding to passes by nozzles (4), (5), (6), (7), (8), and (9) in successive scans, respectively; (d) 0.00 pL, 9.99 pL, 10.08 pL, 10.07 pL, and 0.00 pL droplets in region 416 corresponding to passes by nozzles (5), (6), (7), (10), and (12) in successive scans; and (e) 9.99 pL, 0.00 pL, 10.00 pL, 0.00 pL, and 0.00 pL droplets in region 417 corresponding to passes by nozzles (6), (7), (8), (11), and (13) in successive scans. Again, note that the nozzles in this embodiment are used on a binary basis and with only a single firing waveform (i.e., so that their droplet volume characteristics do not change with scan), so that, for example, in the fifth scan 427, nozzle (7) is not fired and produces no droplets (0.00 pL) for region 417, while in the next scan it is fired and produces a 10.08 pL droplet for region 416.

[0067] As can be seen in the graph at the bottom of the page, this virtual scan process easily produces expected total fills of 49.99 pL, 50.00 pL, 49.96 pL, 49.99 pL, and 50.02 pL, within the desired range of the target value (50.00 pL) ±0.5 percent (49.75 pL to 50.25 pL). Note that in this example, the nozzles were used to deposit ink into multiple target areas substantially simultaneously for each scan, and that the particular combination of drop volumes for each depicted area (i.e., as graphically identified by numerals 413-417) was planned to allow multiple passes to deposit multiple drops in each target area. Although the eight depicted passes together correlate with a particular set (or combination) of drop volumes that produce a fill volume within a particular tolerance range (e.g., for region 413, a combination of drops from nozzles (1), (2), (2), (7), and (9)), other sets of possible drops could conceivably have been used. For example, for region 413, it would have been possible to alternatively use five drops from nozzle (2) (5×10.01 pL=50.05 pL). However, this alternative would have been inefficient because nozzle (3) (9.89 pL) could not have been used extensively at the same time during this time (i.e., the result from five drops from this nozzle would have been 5×9.89=49.45 pL, outside the desired tolerance range), which would have required additional scans. In the example conveyed by Figure 4B, the particular scans and their order were selected to use less printing time, fewer passes, smaller geometric steps, and potentially smaller total geometric step distance, or according to some other criteria. Note that the depicted example is for narrative discussion only, and it may be possible to further reduce the number of scans using the presented drop volumes to fewer than eight scans to obtain the target fill.In some embodiments, the scanning process is planned in a manner that avoids worst-case scenarios with the number of scans required (e.g., one scan per line of target area with the printhead rotated 90 degrees.) In other embodiments, this optimization is applied to an extent based on one or more maximum or minimum values, for example, planning the scans in a manner that results in the fewest possible scans, considering all possible drop combinations for each target area for a given ink.

[0068] FIG. 4C presents a schematic diagram similar to FIG. 4B, but corresponding to the use of different nozzle driving waveforms for each nozzle. As should be understood, in an inkjet printhead, ink is typically ejected using a piezoelectric actuator that expands and contracts a fluid reservoir to eject ink from each printing nozzle. The ink is typically maintained in the reservoir under a slight negative pressure to avoid flooding the nozzle plate, with a voltage pulse applied to the actuator to eject droplets with properties that depend on the size and shape of the voltage pulse. Thus, different pulse characteristics can result in different volumes, velocities, and other characteristics of the ejected droplets. In FIG. 4C, assume that different pre-planned voltage pulse waveforms have been determined to generate a series of different droplet volumes (and associated droplet volume probability distributions). The scans are generally referenced by numeral 441, with each of the scans 443-447 occurring in a direction perpendicular to the bars 443-447, and within each scan bar (e.g., box 443), the numerical designation represents a particular printhead nozzle and the letter designation represents a different waveform for a particular nozzle. For example, reference "1-A" represents a first drive waveform "A" used for the actuator for nozzle (1), while reference "1-C" represents a third drive waveform "C" used for the actuator for nozzle (1). It should be noted that any desired number of waveforms can be tested during the calibration procedure to select a waveform that produces an expected drop volume (or multiple drop volumes) that matches the ideal target drop volume. In Figure 4C, for example, testing of multiple waveforms for nozzle (1) may result in two particular waveforms (e.g., "A" and "C") producing expected drop volumes close to the desired 10.00 pL average, e.g., 9.94 pL and 10.01 pL average, respectively. That is, if testing is unable to produce an expected average that exactly matches the ideal drop volume (e.g., 10.00 pL), then two or more waveforms can be selected that bracket the desired ideal volume, e.g., 9.94 pL / 10.01 pL, 9.99 pL / 10.01 pL, 10.03 pL / 9.95 pL, and 9.95 / 10.04 pL as depicted for nozzles (1), (3), (4), and (5).Similar to the above examples, different droplets can then be combined using different nozzle drive waveforms to specifically plan the total fill for each target area that is within the desired tolerance. Note that for the example of FIG. 4C, it is not necessary to offset the printhead assembly between scans to achieve these combinations, but in many embodiments, the use of multiple nozzle waveforms can be combined with partial swath location width offsets to produce many possible droplet combinations that can be used to generate the target fill using a minimum number of scans (and therefore a minimum print time per substrate). In FIG. 4C, it can be seen that the depicted process produces a very tightly grouped virtual fill, e.g., an expected fill volume of 49.99 pL to 50.02 pL.

[0069] 4D presents an illustrative view 471 of a printhead 474, similar to FIG. 4A, but now with nozzles not specifically aligned to specific wells, and two associated schematics seen below the printhead 474. Because the printhead is used in an embodiment that optionally provides for non-fixed geometric stepping of the printhead relative to the substrate, numeral 472 is used to represent the offset that aligns a particular printhead nozzle (e.g., 16 nozzles total with nozzles (1)-(5) depicted in the figure) with a different target area (in this example, two: 474 and 475). Again following the example of FIG. 4A , if nozzles (1)-(16) produce expected drop volumes of fluid ink of 9.80, 10.01, 9.89, 9.96, 10.03, 9.99, 10.08, 10.00, 10.09, 10.07, 9.99, 9.92, 9.97, 9.81, 10.04, and 9.95 pL, respectively, and if it is desired to deposit 50.00 pL per target area plus or minus 0.5 percent of this value, then the printhead can be used to deposit the drops in three passes or scans, using geometric steps of 0, −1, and −3, respectively, and firing one or two drops into each target area per scan. This would result in a total fill value of 49.93 or 50.10 per area as depicted in the figure, again clearly within the desired tolerance range of 49.75 pL to 50.25 pL for each of the depicted target areas. Thus, as shown, the same approach would equally apply in the case of nozzles not aligned to the wells, and a combination of droplets in a deliberate manner depending on the respective droplet volume and desired fill for each target area can be used to achieve a precisely regulated fill. Furthermore, as explained above for the hypothetical FIG. 4A, this same diagram can be used to represent nozzle drive waveform variations and / or the use of multiple printheads. For example, if the nozzle references (1)-(16) refer to the droplet volume for a single nozzle generated by 16 different drive waveforms (i.e., using waveforms 1-16), one can simply obtain the theoretical fill volume per area by using different drive waveforms.Those skilled in the art will appreciate that the same approach as described above with reference to Figures 4B-4C also applies equally to the case of nozzles that are not specifically aligned to the wells, i.e., a group of one or more nozzles is used for simultaneous droplet deposition into each well. Finally, it should be noted that Figures 4A-D also represent a relatively simple example, and that in a typical application there may be hundreds to thousands of nozzles and millions of target areas. For example, in an application where the disclosed technique is applied in the processing of each pixel color component of a current high-definition television screen (e.g., where the pixels each have red, green, and blue wells, and the pixels are arranged with 1080 horizontal lines of vertical resolution and 1920 vertical lines of horizontal resolution), there are approximately 6 million wells (i.e., three overlapping arrays of 2 million wells each) that can receive ink. Next-generation televisions are expected to increase this resolution by a factor of four or more. In such a process, to improve printing speed, the print head may use thousands of nozzles for printing, and there would typically be an overwhelming number of possible printing process permutations, for example. The simplified examples presented above are used to introduce the concepts, but it should be noted that given the overwhelming number of typical combinations presented, the permutations represented by real-world television applications are extremely complex, and print optimizations are typically applied by software and use complex mathematical operations. Figures 5-7 are used to provide non-limiting examples of how these operations can be applied.

[0070] An exemplary process for planning printing is introduced by Figure 5. This process and related methods and devices are generally referred to using the numeral 501.

[0071] More specifically, the drop volume for each nozzle (and for each nozzle for each waveform if multiple drive waveforms are applied) is specifically determined (503). Such measurements can be made using a variety of techniques, including, for example and without limitation, optical or laser imaging or non-imaging devices built into the printer (or factory-resident machine) that measure the drops in flight (e.g., during a calibration print operation or a live print operation) and calculate the volume with precision based on the drop shape, velocity, trajectory, and / or other factors. In specific embodiments, as described, each measurement is only approximately accurate, since drop volumes from even a single nozzle generated using a single drive waveform may vary from drop to drop. To this effect, drop measurement techniques can be used to generate statistical models for drops from each nozzle and for each nozzle-waveform combination, with each particular drop volume being represented as the average expected drop volume from a given nozzle and a given nozzle drive waveform. Other measurement techniques can also be used, including printing the ink and then using post-print imaging or other techniques to calculate individual drop volumes based on pattern recognition. Alternatively, the identification can be based on data provided by the printer or printhead manufacturer, e.g., based on measurements made at the factory well in advance of the fabrication process and provided to the machine (or online). In some applications, the drop volume characteristics may change over time, e.g., depending on ink viscosity or type, temperature, nozzle clogging, or other degradation, or due to other factors. Thus, in one embodiment, the drop volume measurements can be made dynamically in situ, e.g., upon power-up (or other type of power cycle event), with each new printing of the substrate, upon expiration of a predefined time, or on another calendar or non-calendar basis. In one embodiment, such measurements are made intermittently and continuously, as previously referenced, by making measurements for a moving window of printing nozzles and nozzle-waveform combinations each time a new flat panel substrate is loaded or unloaded to obtain dynamic updates. As represented by numeral 504, this (measured or provided) data is stored for use in the optimization process.

[0072] In addition to the drop volume data per nozzle (and optionally per drive waveform), information regarding the desired fill volume for each target area (505) is also received. This data may be a single target fill value applied to all target areas, respective target fill values ​​applied to individual target areas, rows of target areas, or columns of target areas, or values ​​resolved in some other manner. For example, as applied to fabricating a single "blanket" layer of a large material for an individual electronic device structure (such as a transistor or via), such data may consist of a single thickness to be applied to the entire layer (e.g., which software then converts to a desired ink fill volume per target area based on predetermined conversion data that is specific to the associated ink). In such a case, the data may be a single value for each "print cell" (in this case, each target In another example, the data can represent a specific value (e.g., 50.00 pL) for one or more wells, where range data is either provided or understood based on the context. As should be understood from these examples, the desired fill can be specified in many different forms, including, but not limited to, thickness data or volume data. Additional filtering or processing criteria can also be optionally provided to or implemented by the receiving device. For example, as previously referenced, random variations in fill volume can be injected by the receiving device into one or more provided thickness or volume parameters, such that line effects are not visible to the human eye in the finished display. Such variations can be done in advance (provided as a fill per respective target area that varies by area) or can be derived transparently and independently from the receiving device (e.g., by a downstream computer or printer).

[0073] Based on the target fill volume for each region and the individual drop volume measurements (i.e., per print head nozzle and driving waveforms per nozzle), the process then continues to calculate various drop combinations that optionally sum to a fill volume within a desired tolerance range (i.e., by process block 506). As described, this range can be provided with target fill data or can be "figured out" based on the situation. In one embodiment, the range is understood to be ±1 percent of the provided fill value. In another embodiment, the range is understood to be ±0.5 percent of the provided fill value. Clearly, many other possibilities exist for the tolerance range, whether larger or smaller than these exemplary ranges.

[0074] At this point, an example will help convey one possible method for calculating the set of possible droplet combinations. Returning to the simplified example previously described, assume that there are five nozzles, each with a hypothetical average droplet volume of 9.80 pL, 10.01 pL, 9.89 pL, 9.96 pL, and 10.03 pL, respectively, and that it is desired to deposit a target volume of 50.00 pL ±0.5 percent (49.75 pL to 50.25 pL) in the five wells. The method begins by determining how many droplets can be combined to reach but not exceed a tolerance range, and for each nozzle, the minimum and maximum number of droplets from that nozzle that can be used in any allowable permutation. For example, in this hypothesis, given the minimum and maximum droplet volumes of the nozzles under consideration, one would expect that only one droplet from nozzle (1), two droplets from nozzle (3), and four droplets from nozzle (4) would be usable in any combination. This step limits the number of combinations that need to be considered. With such constraints on the set consideration, the method then takes each nozzle in turn and considers combinations of the required number of droplets (five in this example). For example, the method first starts with nozzle (1), understanding that the only allowable combinations involving this nozzle feature one or fewer droplets from this nozzle. Having considered combinations involving a single droplet from this nozzle, the method then considers the minimum and maximum droplet volumes of the other nozzle-waveform combinations under consideration. For example, considering that nozzle (1) is determined to produce a droplet volume of 9.80 pL for a given drive waveform, as little as one droplet from nozzle (3) or two droplets from nozzle (4) may be used in combination with droplets from nozzle (1) to reach the desired tolerance range. The method then considers combinations of droplets from nozzle (1) and combinations of four droplets from other nozzles, such as four droplets from nozzles (2) or (5), three droplets from nozzle (2) and one droplet from nozzle (4), etc.Considering combinations involving only nozzle (1), to simplify the discussion, any of the following different combinations involving the first nozzle could potentially be used within the tolerance range: {1(1),4(2)}, {1(1),3(2),1(4)}, {1(1),3(2),1(5)}, {1(1),2(2),1(4),1(5)}, {1(1),1(2),1(3),2(5)}, {1(1),1(2),1(4),2(5)}, {1(1),1(2),3(5)}, {1(1),1(3),3(5)}, {1(1),2(4),2(5)}, {1(1),1(4),3(5)}, and {1(1),4(5)}. In the formulas described above, the use of square brackets represents a set of 5 drops representing a combination of drop volumes from one or more nozzles, with each parenthesis within these square brackets identifying a particular nozzle. For example, the formula {1(1),4(2)} represents 1 drop from nozzle (1) and 4 drops from nozzle (2), i.e., 9.80 pL + (4 x 10.01 pL) = 49.84 pL, that is expected to produce a composite fill within a particular tolerance range. In practice, the method in this example considers the largest number of drops from nozzle (1) that can be used to produce the desired tolerance based on various averages, evaluates the combination with this largest number, reduces the number by 1, and repeats the process of consideration. In one embodiment, this process is repeated to determine all possible sets of non-redundant drop combinations that can be used. When combinations involving nozzle (1) have been exhaustively explored, the method tests the composite average of each possible nozzle combination to determine if the desired tolerance range can be achieved, moving on to combinations involving nozzle (2) rather than nozzle (1), repeating the process, etc. For example, in this embodiment, the method determines that a combination of more than two drops from nozzle (1) cannot be used, so begins by considering combinations involving one drop from nozzle (1) and four drops from the other nozzles, in various combinations.The method actually evaluates whether four droplets from nozzle (2) can be used, determining that it could be {1(1), 4(2)}, then reduces this number by one (three droplets from nozzle 2) and determines that this number can be used in combination with a single droplet from nozzles (4) or (5), resulting in the allowed sets of {1(1), 3(2), 1(4)}, {1(1), 3(2), 1(5)}. The method then further reduces the number of allowed droplets from nozzle (2) and evaluates combinations of {1(1), 2(2)....}, then {1(1), 1(2)....}, etc. Once combinations involving nozzle (2) have been considered in combination with droplets from nozzle (1), the method then takes the next nozzle, i.e. nozzle (3), and considers combinations of nozzle (1) with this nozzle rather than nozzle (2), determining that the only allowed combination is given by {1(1), 1(3), 3(5)}. Once all combinations involving drops from nozzle (1) have been considered, the method then considers five drop combinations that involve drops from nozzle (2) but not nozzle (1), such as {5(2)}, {4(2),1(3)}, {4(2),1(4)}, {4(2),1(5)}, {3(2),2(3)}, {3(2),1(3),1(4)}, etc.

[0075] It should also be noted that the same approach applies equally in cases where a nozzle can be driven by multiple firing waveforms (each producing a different drop volume). These additional nozzle-waveform combinations simply provide additional drop volume averages for use in selecting a set of drop combinations that are within the tolerance of the target volume. The use of multiple firing waveforms can also improve the efficiency of the printing process by making available a larger number of allowable drop combinations, thereby increasing the likelihood of simultaneously firing drops from a majority of the nozzles on each pass. In cases where a nozzle has multiple driving waveforms and geometric steps are also used, the selection of the set of drop combinations will incorporate both the geometric offset used in a given scan and the nozzle waveform that will be used for each nozzle.

[0076] It should be noted that for purposes of illustration, a brute force approach is described, and an overwhelming number of possible combinations, e.g., where the number of nozzles and target areas is large (e.g., greater than 128 each), would typically be presented in practice. However, such a scheme The calculations are well within the capabilities of a fast processor with appropriate software. It should also be noted that there are various mathematical shortcuts that can be applied to reduce the calculations. For example, in a given embodiment, the method can eliminate from consideration any combination that would correspond to the use of less than half of the available nozzles in any one pass (or alternatively, can limit consideration to combinations that minimize the volume variance across the target region (TR) in any one pass). In one embodiment, the method determines only certain sets of drop combinations that will produce acceptable composite fill values, and in a second embodiment, the method exhaustively calculates all possible sets of drop combinations that will produce acceptable composite fill values. It is also possible to use an iterative approach in which print scans are made in multiple iterations and the amount of ink that is still not deposited to reach the desired tolerance range is considered with the aim of optimizing the next subsequent scan. Other approaches are possible.

[0077] Also note that, as an initial operation, if the same fill value (and tolerance) is applied to each target area, it is sufficient to calculate the combinations once (e.g., for one target area) and store these possible droplet combinations for initial use with each target area. This is not necessarily true for all set calculation methods and all applications (e.g., in some embodiments, the allowable fill range may vary for all target areas).

[0078] In yet another embodiment, the method uses mathematical shortcuts, such as approximations, matrix mathematics, random selection, or other techniques, to determine the set of allowable droplet combinations for each target region.

[0079] As represented by process block 507, once a set of allowable combinations is determined for each target area, the method then effectively plans scans in a manner that correlates with the specific set (or drop combination) for each target area. This specific set selection is done in a manner in which the specific set (one for each target area) represents process savings through the use of at least one scan to simultaneously deposit drop volumes in multiple target areas. That is, in the ideal case, the method selects one specific set for each target area, the specific set representing a specific drop volume combination in a manner in which the printhead can simultaneously print into multiple rows of the target area at one time. The selection of specific drops in the selected combination represents a printing process that meets predetermined criteria, such as minimum printing time, minimum number of scans, minimum size of geometric steps, minimum total geometric step distance, or other criteria. These criteria are represented by numeral 508 in FIG. 5. In one embodiment, the optimization is Pareto optimal, where the specific set is selected in a manner that minimizes each of the number of scans, total geometric step distance, and geometric step size, in that order. Again, the selection of this particular set can be done in any desired manner, some non-limiting examples of which are discussed further below.

[0080] In one embodiment, the method selects a droplet from each set for each target region that corresponds to a particular geometric step or waveform that is applied to all regions under consideration, and then subtracts this droplet from the available set to determine a residual. For example, the selection of available sets may initially be {1(1),4(2)}, {1(1),3(2),1(4)}, {1(1),3(2),1(5)}, {1(1),2(2),1(4),1(5)}, {1(1),1(2),1(3),2(5)}, {1(1),1(2),1(4),2(5)}, {1(1),1(2),3(5)}, {1(1),1(3),3(5)}, {1(1),2(4),2(5)} for each of the five target regions. , {1(1),1(4),3(5)}, and {1(1),4(5)}, the present embodiment would subtract one droplet (1) from this first set to obtain a residual specific to the first of the five target regions, one droplet (2) from the first set to obtain a residual specific to the second of the five target regions, one droplet (3) from the first set to obtain a residual specific to the third of the target regions, and so on. This evaluation would represent a geometric step of "0". The method would then evaluate the residual and repeat the process for other possible geometric steps. For example, if a geometric step of "-1" was applied, the method would then evaluate the residual by subtracting one droplet (2) from the first set for the first of the five target regions, one droplet (3) from the first set for the second of the target regions, and so on.

[0081] In selecting a particular geometric step (and nozzle firing) as part of the printing plan, the method analyzes the various residuals according to a scoring and priority function and selects the geometric step with the best score. In one embodiment, the scoring is applied to weight more heavily the step that (a) maximizes the number of nozzles used simultaneously and (b) maximizes the minimum number of combinations remaining for the affected target area. For example, a scan using drops from four nozzles during the scan would be greatly favored over one using drops from only two nozzles. Similarly, if the subtraction process discussed above in considering the different steps resulted in a remaining combination of 1, 2, 2, 4, and 5 for each target area for one possible step, and a remaining combination of 2, 2, 2, 3, and 4 for each target area for a second possible step, the method would weight the latter more heavily (i.e., the maximum minimum number is "2"). In practice, suitable weighting factors can be generated empirically. Clearly, other algorithms can be applied, and other forms of analysis or algorithmic shortcuts can be applied. For example, matrix mathematics can be used (e.g., using eigenvector analysis) to determine particular drop combinations and associated scan parameters that meet a given criteria. In another variation, other formulas can be used that take into account, for example, the use of planned random packing variations to mitigate line effects.

[0082] Once a particular set and / or scan path has been selected by numeral 507, numeral 509 orders the printer actions. It is noted that, for example, if total fill volume was the only consideration, the sets of droplets could typically be deposited in an arbitrary order. If printing is planned to minimize the number of scans or passes, the order of geometric steps can also be selected to minimize printhead / substrate motion. For example, if allowable scans in the hypothetical example involve relative geometric steps of {0, +3, -2, +6, and -4}, these scans can be reordered to minimize printhead / substrate motion and thus further improve printing speed, e.g., ordering the scans as a series of steps of {0, +1, +2, 0, and +4}. Compared to the first series of geometric steps {0, +3, -2, +6, and -4} with a total step increment distance of 15, the second series of geometric steps {0, +1, +2, +0, and +4} with a total step increment distance of 7, promoting faster printer response.

[0083] For applications involving multiple rows of target areas that are to receive the same target fill, as represented by numeral 510, the particular solution may then be represented as a repeatable pattern that is replicated across the subset areas of the substrate. For example, if an application had 128 nozzles arranged in a single row, and a target area with 1024 rows, it is expected that an optimal scan pattern could be determined for a subset area of ​​the target area with 255 rows or fewer, and thus, in this example, the same print pattern could be applied to four or more subset areas of the substrate. Thus, some embodiments utilize a repeatable pattern as represented by optional process block 510.

[0084] Note the use of non-transitory machine-readable medium icon 511. This icon represents that the methods described above are optionally implemented as instructions for controlling one or more machines (e.g., software or firmware for controlling one or more processors). The non-transitory medium can include any machine-readable physical medium, such as a flash drive, a floppy disk, a tape, a server storage or mass storage, a dynamic random access memory (DRAM), a compact disk (CD), or other local or remote storage. This storage can be embodied as part of a larger machine (e.g., resident memory in a desktop computer or printer) or on an isolated basis (e.g., a flash drive or a standalone storage device that will later transfer files to another computer or printer). Each of the functions described with reference to FIG. 5 can be implemented as part of a composite program or as standalone modules stored either together on a single media representation (e.g., a single floppy disk) or on multiple separate storage devices.

[0085] As represented by numeral 513 in FIG. 5, once the planning process is complete, data will be generated that effectively represents a set of printer instructions, comprising nozzle firing data for the printhead and instructions for relative movement between the printhead and the substrate to support the firing pattern. This data, effectively representing the scan path, scan order, and other data, is an electronic file (513) that can either be stored for later use (e.g., as depicted by non-transient machine-readable medium icon 515) or immediately applied to control the printer (517) to deposit ink representing a selected combination (a particular set of nozzles per target area). For example, the method can be applied on a stand-alone computer, with the instruction data stored in RAM for later use or for downloading to another machine. Alternatively, the method can be implemented to automatically plan the scan depending on the printer parameters (nozzle and drop volume data, etc.) and dynamically applied by the printer to the "incoming" data. Many other alternatives are possible.

[0086] 6A-6D generally provide a schematic diagram for the nozzle selection and scan planning process. Note again that the scan does not have to be continuous or linear in direction or speed of movement, and does not have to proceed from one side of the substrate to another.

[0087] The first block diagram is represented by numeral 601 in FIG. 6A. This diagram represents many of the exemplary processes discussed in the preceding narrative. The method begins by first retrieving from memory, by numeral 603, a set of acceptable drop volume combinations for each target region. These sets could be calculated dynamically or could have been pre-calculated, for example, using software on a different machine. Note the use of a database icon 605, which represents either a locally stored database (e.g., stored in local RAM) or a remote database. The method then effectively selects a particular one of the acceptable sets for each target region (607). This selection is indirect in many embodiments, meaning that the method processes the acceptable combinations to select particular scans (e.g., using the techniques referenced above), and it is these scans that actually define the particular sets. Nevertheless, by planning the scans, the method selects a particular set of combinations for each target region. This data is then used to sequence the scans to organize the motion and firing patterns (609), as referenced above.

[0088] The center and right of FIG. 6A illustrate several process options for planning scan paths and nozzle firing patterns, and in effect, selecting specific drop combinations for each target area in a manner that represents print optimization. As represented by the numeral 608, the illustrated The technique represents only one possible methodology for performing this task. Per numeral 611, the analysis can involve determining the minimum and maximum usage of each nozzle (or nozzle-waveform combination, in those cases where a nozzle is driven by more than one firing waveform) in an allowed combination. If a particular nozzle is faulty (e.g., does not fire or fires on an unacceptable trajectory), then optionally, that nozzle can be excluded for use (and consideration). Second, if a nozzle has either a very low or a very high expected drop volume, this can limit the number of drops that can be used from that nozzle in an allowed combination. Number 611 represents a pre-processing that reduces the number of combinations that will be considered. As represented by numeral 612, a process / shortcut can be used to limit the number of sets of drop combinations that will be evaluated. For example, instead of considering "all" possible drop combinations for each nozzle, the method can be optionally configured to exclude combinations that involve fewer than half the nozzles (or another quantity of nozzles, such as 1 / 4), combinations in which more than half the drops come from any particular nozzle waveform, or combinations that represent a high variance in drop volume or a large variance in the simultaneous drop volumes applied across the target area. Other metrics can also be used.

[0089] Subject to any limitations on the number of sets calculated / considered, the method then proceeds to calculate and consider permissible drop combinations, per numeral 613. Various processes can be used to plan scans and / or otherwise effectively select a particular set of drop volumes per target region (TR), as referenced by numerals 614 and 615. For example, as introduced above, one method assumes a scan path (e.g., a particular geometric step selection) and then considers the maximum of the fewest remaining set selections across all TRs being considered. The method can favorably weight those scan paths (alternative geometric steps) that maximize the ability of the next scan to cover multiple target regions at once. Alternatively, or in addition, the method can favorably weight those geometric steps that maximize the number of nozzles used at once. Returning to the simplified five nozzle discussion above, a scan that would apply five nozzles to a target region can be weighted more significantly than a scan or pass that would fire only three nozzles in one pass. Thus, in one embodiment, the following algorithm can be applied by the software:

number

[0090] FIG. 6A also shows some further options. For example, consideration of droplet sets in some implementations is done according to an equation / algorithm by numeral 617. The comparison metric can be expressed as a score that can be calculated for each possible alternative geometric step to select a step or offset. For example, another possible algorithmic approach involves an equation with three terms as shown below:

number

[0091] Number 619 in FIG. 6A represents that in one embodiment, the selection of droplet combinations can be performed using matrix mathematics, for example through the use of mathematical techniques that consider all droplet volume combinations simultaneously and use a form of eigenvector analysis to select scan paths.

[0092] An iterative process can be applied to reduce the number of droplet combinations considered, as represented by numeral 621. That is, for example, the geometric steps can be calculated one at a time, as represented by the previous description of one possible processing technique. Each time a particular scan path is planned, the method determines the incremental amount still required in each target area under consideration, and then continues to determine the scan or geometric offset that is most suitable to produce a total amount or fill amount per target area that is within the desired tolerance. This process can then be repeated for each iteration until all scan paths and nozzle firing patterns have been planned.

[0093] Numeral 622 also allows for the use of a hybrid process. For example, in one embodiment, a first set of one or more scans or geometric steps can be selected for use, for example, based on minimized variation in drop volume per nozzle and maximum efficiency (e.g., nozzles used per scan). Once a certain number of scans, e.g., 1, 2, 3 or more scans, have been applied, a different algorithm can be invoked that, for example, maximizes nozzles used per scan (e.g., regardless of the variation in drop volume applied). Any of the specific equations or techniques discussed above (or other techniques) can optionally be one of the applied algorithms in such a hybrid process, and other variations will naturally occur to those skilled in the art.

[0094] It is noted that, as previously referenced, in an exemplary display manufacturing process, the fill volume per target area can have a planned randomization (623) that is purposely injected to mitigate line effects. In one embodiment, a generator function (625) is optionally applied to purposely vary the target fill volume (or to make the total volume generated for the drop combination for each target area asymmetric) in a manner that achieves this planned randomization or other effect. As previously mentioned, in a different embodiment, it is also possible to apply an algorithmic approach as previously shown such that such variation is included in the target fill volume and tolerance, and to meet, for example, the fill requirements per target area, i.e., even before the drop combination is analyzed. As discussed below in connection with FIG. 8B, it is also possible to consider the randomization as a probability distribution, and to plan the drop measurements (and occurrences per nozzle, per waveform distribution) that depend on such randomization in a manner that is calculated to meet the composite fill tolerance. For example, if the planned fill randomization typically varies between ±0.2% of the target composite fill and the specified tolerance is ±0.5% of the target composite fill, then drop measurements for each nozzle and each nozzle-waveform combination can be planned to produce a 3σ value for each nozzle / nozzle-waveform that is within 0.3% of the target (0.2% + 0.3% = 0.5%).

[0095] FIG. 6B and numeral 631 refer to a more detailed block diagram related to the iterative droplet combination selection process referenced above. As indicated by numerals 633 and 635, again possible droplet combinations are first identified, stored, and retrieved for evaluation by the software as appropriate. For each possible scan path (or geometric step), by numeral 637, the method stores the footprint (639) identifying the scan path and the nozzle applied, and subtracts the firings per nozzle from the set per target area (641) to determine the remaining combinations for each target area (643). These are also stored. Then, by numeral 645, the method evaluates the stored data according to predefined criteria. For example, as indicated by optional (dashed) block 647, a method that seeks to maximize the minimum number of droplet combinations over all relevant target areas can assign a score indicating whether the just stored combination is better or worse than previously considered alternatives. If certain criteria are met (645), a particular scan or geometric step may be selected, and the remaining combinations are stored or otherwise flagged for use in consideration of another printhead / substrate scan or pass, as represented by numerals 649 and 651. If the criteria are not met (or consideration is incomplete), another step may be considered, and / or the method may adjust consideration of the geometric step under consideration (or a previously selected step), per numeral 653. Again, many variations are possible.

[0096] It was previously noted that the order in which the scans are performed or the droplets are deposited is not important to the final composite fill value of each target area. While this is true, to maximize printing speed and throughput, the scans are preferably ordered to result in the fastest or most efficient printing possible. Thus, if not previously included in the geometric step analysis, then reordering and / or ordering of the scans or steps can be done. This process is represented by FIG. 6C.

[0097] Specifically, numeral 661 is used to generally designate the method of FIG. 6C. For example, software running on a suitable machine causes a processor to retrieve (663) a selected geometric step, specific set, or other data that identifies a selected scan path (and optionally a nozzle firing pattern, which in those embodiments where a nozzle may be driven by more than one firing waveform, may further include data identifying which of multiple firing waveforms is to be used for each droplet). These steps or scans are then reordered or ordered in a manner that minimizes the incremental step distance. For example, again referring to the hypothetical example introduced earlier, if the selected steps / scan paths were {0, +3, -2, +6, and -4}, these may be reordered to minimize each incremental step, and to minimize the overall (total) distance traversed by the motion system between scans. For example, without reordering, the incremental distances between these offsets would be equivalent to 3, 2, 6, and 4 (so that the total distance traversed would be "15" in this example). If the scans (e.g., scans "a", "b", "c", "d", and "e") were reordered in the manner described (e.g., in the order "a", "c", "b", "e", and "d"), the incremental distances would be +1, +2, 0, and +4 (so that the total distance traversed would be "7"). As represented by numeral 667, at this point the method can assign motion to the print head motion system and / or the substrate motion system, and the order of nozzle firings can be reversed (e.g., if alternating back-and-forth scan path directions are used, per numerals 339 and 340 in FIG. 3B). As previously described and represented by optional process block 669, in some embodiments, planning and / or optimization can be performed for a portion of the target area, and the solution is then applied on a spatially iterative basis across the large substrate.

[0098] This iteration is represented in part by FIG. 6D. As suggested by FIG. 6D, assume for this description that it is desired to fabricate an array of flat panel devices. A common substrate is represented by numeral 681, and a set of dashed boxes, such as box 683, represent the geometry of each flat panel device. Preferably, fiducials 685 with two-dimensional characteristics are formed on the substrate and used to locate and align the various fabrication processes. Following final completion of these processes, each panel 683 will be separated from the common substrate using a cutting or similar process. If the array of panels represents respective OLED displays, the common substrate 681 will typically be glass, with the glass deposited thereon, followed by one or more encapsulation layers. Each panel will then be flipped over so that the glass substrate forms the light-emitting surface of the display. For some applications, other substrate materials can be used, e.g., transparent or opaque flexible materials. As described, many other types of devices can be manufactured according to the described techniques. For a particular portion 687 of the flat panel 683, a solution can be calculated. This solution can then be repeated for other similarly sized portions 689 of the flat panel 683, and then the entire solution set can be repeated for each panel formed from a given substrate.

[0099] Considering the various techniques and considerations introduced above, manufacturing processes can be made to mass-produce products quickly and at low per-unit cost. When applied to display device manufacturing, e.g., flat panel displays, these techniques enable a high-speed per-panel printing process where multiple panels are produced from a common substrate. By providing high-speed repeatable printing techniques (e.g., using common inks and printheads from panel to panel), it is believed that printing can be substantially improved, e.g., reducing printing time per layer to a fraction of the time that would otherwise be required while ensuring that the fill amount per target area is within specifications. Returning again to the example of a large HD television display, it is believed that each color component layer can be accurately and reliably printed for a large substrate (e.g., a 8.5 generation substrate that is approximately 220 cm x 250 cm) in 180 seconds or less, or even 90 seconds or less, which represents a significant process improvement. Improving the efficiency and quality of printing paves the way for a significant reduction in the cost of producing large HD television displays, and therefore a lower end-consumer cost. As mentioned above, display manufacturing (specifically OLED manufacturing) is one application of the techniques introduced herein, however, these techniques can be applied to a wide variety of processes, computers, printers, software, manufacturing equipment, and end devices and are not limited to display panels.

[0100] One benefit of the ability to deposit precise target area volumes (e.g., well volumes) within tolerances is the ability to inject intentional variations within the tolerances, as described. These techniques facilitate significant quality improvements in displays, as they provide the ability to hide the pixelation artifacts of the display and make such "line effects" imperceptible to the human eye. FIG. 7 provides a block diagram 701 associated with one method for injecting this variation. As with the various methods and block diagrams discussed above, block diagram 701 and associated methods can optionally be implemented as software, either on a stand-alone medium or as part of a larger machine.

[0101] The variation can be made dependent on certain criteria, as represented by numeral 703. For example, it is generally understood that the sensitivity of the human eye to contrast variations is a function of luminance, expected viewing distance, display resolution, color, and other factors. Considering the typical human eye sensitivity to spatial variations in contrast between colors at different luminance levels as part of the specified criteria, measures are used to ensure that such variations will be smoothed in a manner that is not perceptible to the human eye, e.g., varied in a manner that does not provide a human-observable pattern (a) in one or more arbitrary directions, or (b) between color components given the expected viewing conditions. This can optionally be achieved using a planned randomization function, as previously referenced. Once the minimum criteria are specified, the target fill amount for each color component and each pixel can be purposefully varied in a manner calculated to hide any visual artifacts from the human eye, as represented by numeral 705. Note that the right side of FIG. 7 represents various process options, e.g., the variations can be made independent across color components (707), with tests for perceptible patterns applied in an algorithmic criterion to ensure that the fill variations do not result in perceptible patterns. As described by numeral 707, for any given color component (e.g., any given ink), the variation can also be made independent in each of multiple spatial dimensions, e.g., the x and y dimensions (709). Again, in one embodiment, not only is the variation smoothed for each dimension / color component so that it is not perceptible, but any patterns of differences between each of these dimensions are also suppressed from being visible. Per numeral 711, one or more generator functions can be applied to ensure that these criteria are met, for example, by optionally using any desired criteria to assign minor target fill variations to the fill of each target area prior to drop volume analysis. As represented by numeral 713, in one embodiment, the variation can be optionally made random.

[0102] The selection of a particular drop combination for each target area, per numeral 715, is thus weighted in favor of the selected variation criterion. This can be done via target fill variation, as described, or at drop (e.g., scan path, nozzle-waveform combination, or both) selection time. There are other ways to impart this variation as well. For example, in one contemplated implementation, per numeral 717, the scan path is varied non-linearly, effectively varying the drop volume across the average scan path direction. Per numeral 719, the nozzle firing pattern can also be varied, for example, by adjusting the firing pulse rise time, fall time, voltage, pulse width, or by using multiple signal levels per pulse (or other forms of pulse shaping techniques) to provide minor drop volume variations. In one embodiment, these variations can be pre-calculated, and in a different embodiment, only waveform variations that produce very minor volume variations are used, with other measures employed to ensure that the total fill remains within a specified tolerance range. In one embodiment, for each target area, multiple drop combinations that fall within a particular tolerance range are calculated, and for each target area, the selection of which drop combination is used for that target area is varied (e.g., randomly or based on a mathematical function), or the particular waveform (i.e., used to generate a given volume of drops) is varied for one nozzle that contributes to the selected combination, e.g., to provide slight volume variation, thereby effectively varying the drop volume across the target area but not the planned scan path. Such variation can be implemented along the scan path direction, across a row of target areas, across a column of target areas, or both.

[0103] 8A-8B are used to explain a method for generating a statistical model that is used to evaluate the droplets produced by each nozzle or nozzle-waveform combination, and, optionally, to plan combinations of multiple droplets according to a statistical average determined from measurements. In the example of FIGS. 8A-8B, a statistical model is constructed for the droplet volume that can be expected from a given nozzle-driving waveform pairing, and in alternative embodiments, a similar statistical model may be used. It should be noted that a theoretical model can be constructed for droplet velocity, droplet flight trajectory (eg, relative to normal), or for some other parameter.

[0104] The method depicted by FIG. 8 is generally designated by the numeral 801. Per function block 803, the method in this embodiment begins with the establishment of a specification range, e.g., a maximum and minimum fill, for a given target area that will receive ink. In the previously presented examples, this specification range can be expressed as an average ± a specified value (e.g., 50.00 pL ± 0.5%), although almost any range or expression of tolerance can be used. In one contemplated implementation, the specified tolerance for the target is ± 0.5%, although other values ​​such as, but not limited to, 1.0% or 2.0% can also be used. Consistent with the previous examples, for this example, it will be assumed that the target is 50.00 pL and the tolerance is ± 0.5% (such that the tolerance range is 49.75 pL to 50.25 pL), although almost any range or tolerance criteria can be used.

[0105] Per numeral 805, one or more candidate waveforms are selected for each nozzle of the printhead or printhead assembly. In embodiments that use only a single drive waveform (e.g., a square voltage pulse of fixed voltage), no selection needs to be made. In embodiments that allow for customized waveform definition (see, e.g., the discussion below in conjunction with FIG. 14B and FIG. 15A-B), it is typically desired to evaluate several alternative waveforms, representing a range of values ​​(e.g., one can interpolate between to ultimately identify multiple acceptable waveforms for each nozzle under consideration). This selection can be made according to a manual design process (807) (i.e., with waveforms selected by the designer and preprogrammed into the system), or, per numeral 809, the selection process can be automated.

[0106] Once one or more waveforms are defined for each nozzle, droplet measurements are planned for different droplet ejections for a given nozzle-waveform pairing. For example, in one embodiment, several droplets (e.g., “24”) may be required for each nozzle to provide a basis for evaluation of the measured statistical distribution for the various droplets. A droplet measurement device (e.g., imaging or non-imaging type) may be used for this purpose, as discussed herein. The 24 (or other number) measurements may be planned for an immediate measurement, or to be performed for each or multiple measurement cycles or iterations. Furthermore, in one embodiment, a threshold number of measurements may be planned for initialization, and then the system increases the measurement data set over time such that a strong confidence level regarding the measured statistical distribution is generated, and in alternative embodiments, each measurement may be planned for a moving time window (e.g., remeasurements may be planned “every 3 hours”, or measurement data may be kept only for some limited time interval to be used for analysis), and thus, in one embodiment, each measurement is stored with a timestamp to indicate its validity and expiration during evaluation. Whatever measurement and / or measurement retention criteria are used, a number of measurements can be scheduled for each nozzle-waveform pairing for purposes of statistical analysis (811). Advantageously, the respective measurements for the droplets resulting from each nozzle-waveform pairing are grouped as a set and scheduled in a manner that encourages the generation of a known common distribution form using well-understood rules for mathematical processing (including aggregation). For example, normal, Student's T, and Poisson distributions all have associated parameters that can be combined according to known mathematical processes to predict the total or composite distribution of fill that would result from the combination of individual droplets (for each nozzle-waveform pairing). Thus, measurement scheduling can be performed according to the techniques described herein to generate a droplet data set that allows for the statistical combination of droplets associated with potentially different nozzle-waveform pairings to achieve precision fill within a specified tolerance with a very high degree of confidence (e.g., typically greater than 99% confidence, per figure 813). Thus, one of the techniques described can be used to generate a droplet data set that allows for the statistical combination of droplets associated with potentially different nozzle-waveform pairings to achieve precision fill within a specified tolerance with a very high degree of confidence (e.g., typically greater than 99% confidence, per figure 813). In one implementation, the drop measurements for each nozzle-waveform combination are planned to satisfy a set of parameters that represent a known type of probability distribution (e.g., the number of measurements or components n, statistical mean μ, and standard deviation σ for a normal distribution), and the measurement data is stored (once acquired) for all possible nozzles and nozzle-waveform pairings under consideration. In one embodiment, the planning and measuring can be iterative, i.e., repeated until some desired criterion is reached, such as a minimum number of raw measurements (n), a minimum number of measurements that meet some criterion, a minimum statistical spread (e.g., a 3σ value that meets some criterion or desired confidence interval), or another. Whatever planning criteria is applied (e.g., by software), the system, including the drop measurement device and printhead assembly under consideration, then undergoes drop measurements (815), which are applied individually to each nozzle (and each drive waveform for a given nozzle) to generate a statistically significant number of drop measurements. As described by numerals 817 and 819, such measurements are optionally made in situ (e.g., optionally in a printer or OLED device processing equipment, under a controlled atmosphere) and in a manner sufficient to generate statistical confidence. The collected data can then be stored (821) as an aggregate probability distribution and / or optionally in a manner that retains individual measurement data (including, for example, any timestamps used to indicate measurements per nozzle).

[0107] As previously discussed, in one embodiment, droplets from potentially different nozzles and / or nozzle driving waveforms are intelligently combined to obtain a precise fill within a high degree of statistical confidence. Once a common form probability distribution is constructed for each nozzle, this combination (and associated planning) is accomplished by combining statistical parameters for each droplet to obtain a precise fill (and a well-understood probability distribution for each fill). This is represented in FIG. 8A by numbers 823, 825, and 827. More specifically, droplet averages are combined (e.g., corresponding to associated normal distributions) in one embodiment to obtain a predicted total fill for a target area. As an example, if for a given first and second nozzle-waveform pairing, the average droplet volume is measured as 9.98 pL and 10.03 pL, respectively, then the average total fill based on one droplet associated with each pairing is expected to be 20.01 pL (μc=μ1+μ2 when normal distributions are involved). In this same hypothetical example, if the standard deviations are 0.032 pL (σ1) and 0.035 pL (σ2) for each droplet, then the expected standard deviation of the tally would be 0.0474 pL (i.e., based on σ2c=σ21+σ22), and the 3σ value of the tally would be about 0.142 pL (note that 1σ equals about a 68.27% confidence interval, while 3σ equals about a 99.73% confidence interval). Similar techniques can be applied to any common distribution type through treating the droplet measurements for each nozzle-waveform pairing as independent random variables. Thus, the technique employed herein uses the droplet measurement techniques to build a statistical model for each nozzle-waveform pairing to plan various droplet combinations based on the analysis of the tabulated random variables, as represented by box 825 (for normal distribution). Nearly any type of probability distribution can be used, provided that the type of probability distribution is amenable to random variable tabulation. As indicated by function block 827, given a desired specification range (e.g., 0.5% for the target), the proposed combinations are analyzed (e.g., by software) to ensure that they meet the desired range with a high degree of statistical confidence.For example, in one embodiment, as described, a desired confidence criterion (e.g., 3σ representing a 99.73% confidence interval) is tested to ensure that it falls within the desired tolerance range. As an example, if the desired tolerance is 49.75-50.25 pL per the example provided above, and the possible droplet combinations are expressed as an average of 49.89 pL with a 3σ value equal to 0.07 pL, this translates to 99% confidence that the total fill will be 49.82 pL-49.96 pL, well within the desired tolerance range, and the particular combination will be deemed an acceptable combination (per the droplet combination analysis function described above). Again, any desired statistical criteria or goodness of fit data may be used. In another embodiment, a 4σ value (99.993666%) or other value is analyzed for a desired tolerance range. Once an acceptable drop combination is determined for each print well, a specific combination of drops for each well (representing simultaneous deposition by multiple nozzles of a printhead assembly) can then be planned (see FIGS. 5-7), with the next print (829) following the pre-planned drop combination for each well.

[0108] FIG. 8B provides another method 851 for accommodating intentional target area filling variations according to a desired criteria, and optionally performing a variable number of droplet measurements per nozzle (or per nozzle waveform). More specifically, the method may again be implemented as instructions stored on a non-transitory machine storage medium that controls at least one processor to perform a set of functions dictated by the instructions. A desired tolerance range is received as a first operand "x" by number 853, and may specify, for example, that the target area (e.g., pixel well) filling should be within a given percentage of the target volume, e.g., 50.00 pL ±0.5%. This tolerance range may be dictated by a customer or industry specification, as indicated by function 855. If it is desired to plan for intentional variations in the composite volume (e.g., random variations within a small range to avoid line effects or other noticeable artifacts in the finished display), that range is received as a second operand "y" by function 857. Based on these two operands, the method calculates the effective allowable maximum variation, standard deviation, or other measure, per block 859. In one embodiment, y is subtracted from x as depicted in the figure, equal to the effective allowed fill variation. For example, if the specification requires a fill within ±0.5% per the above example, and an intentional random variation of ±0.1% is injected into the planned composite mean of well fill (e.g., 49.95 pL to 50.05 pL), then again, using the target example of 50.00 pL ±0.5%, the allowed variation (prior to random variation) can be limited to 49.80 pL to 50.20 pL. Note that other techniques are possible, e.g., instead of simply subtracting these measures, a different set of boundary criteria can be used, e.g., based on mathematics associated with a statistical combination of standard deviations or variances for independent random variables, and many other criteria can be applied depending on the embodiment.Per block 859, the remaining range (e.g., ±0.4% of the target) can then be equated with a desired confidence interval (e.g., a 3σ interval or other statistical measure) and used to assess whether possible droplet combinations are acceptable or removed from consideration, per the examples given above.

[0109] Alternatively, the remaining ranges and associated confidence intervals can be applied as criteria to govern the drop measurements to build a desired statistical model for each drop, as indicated by function blocks 861 and 863. For example, once the desired confidence interval is defined (e.g., 3σ<=0.4% of target), as represented by block 861, a desired variance or maximum allowed variance can be identified, effectively defining a baseline number n of drop measurements that need to be made for each nozzle-waveform combination in a manner calculated to generate a statistical model that meets the desired statistical criteria. For example, the desired valid tolerance range can be used to identify a number of measurements (e.g., 24, 50, or another number) that is calculated to generate a statistical distribution that will be tight, regardless of whether or not the fill is intentionally varied, and thus lead to a large number of possible drop combinations that can be used in a print plan. This calculation can be applied in a number of ways, for example, (a) identifying a threshold number of measurements to be applied to each nozzle-waveform combination (e.g., 24 drop measurements for each), or (b) identifying a threshold statistical criterion that must be met for each nozzle-waveform combination (e.g., a potentially variable number of measurements are made per nozzle or nozzle waveform up to a threshold criterion, e.g., variance, standard deviation, etc.). Various exemplary relationships represented by this test are then described in function box 865. A drop test function is applied using the drop measurement device to make measurements using a number (863). For example, ni drops may be measured for each nozzle (or nozzle-waveform pairing) "i" as shown in box 865. For each measurement, software controlling the drop measurement device may make incremental drop volume measurements (867) and store the data in memory (869). Following each measurement (or after a threshold number of measurements), the collective measurements for a given nozzle-waveform combination may be tallied to calculate (871) statistical parameters for the particular nozzle-waveform combination (e.g., mean and standard deviation, μ and σ, for a type of normal distribution). These values ​​may then be stored in memory (873). Optionally, these same or different measurement techniques may be applied to store one or more drop measurements for velocity v, and x- and y-dimensional trajectories (α and β), via function box 874. As reflected by numeral 875, a decision criterion can then be applied to determine whether sufficient measurements have been made for a given parameter (e.g., amount) for a particular nozzle-waveform combination (i) or whether additional measurements are desired. If additional measurements are required, the method loops via flow arrow 877 to acquire such additional measurements, i.e., so that a statistical model that meets the desired robustness criteria can be constructed for the particular nozzle-waveform combination. If additional measurements are not required, the method can then proceed to the next nozzle 879, looping accordingly via flow arrow 881 until all nozzles and / or nozzle-waveform combinations have been processed. It should be noted that this order is not required for all embodiments, and for example, loops 877 and 881 can be reordered, e.g., as droplet measurements are made for each nozzle in succession, the process is repeated until sufficiently robust data is obtained, and that such a process provides certain advantages for embodiments in which droplet measurements are made incrementally in a manner that is stacked against other system processes (see, e.g., the discussion of FIG. 19 below).Once all nozzles or nozzle-waveform combinations have been fully tested, the method ends, or temporarily stops if performed intermittently, as indicated at 883. For the droplet tests described, the acquired data, including measured data and / or calculated statistical parameters, is stored in machine-readable memory 885, for use in planning droplet combinations, such as those discussed above. The acquired data can also be optionally used in other manners, instead of or in addition to intelligent blending of different droplet volumes. In one embodiment, as described, the stored data can represent any desired droplet parameters, including one or more of droplet volume, droplet volume, and / or droplet trajectory, again in the form of individual measurements and / or statistical parameters.

[0110] 9A-10C are used to provide simulation data for the techniques discussed herein. While FIG. 9A-9C represents the expected composite fill volume based on 5 drops, FIG. 10A-10C represents the expected composite fill volume based on 10 drops. For each of these figures, the letter designation "A" (e.g., FIG. 9A and 10A) represents a situation where a nozzle is used to deposit drops without consideration regarding volume difference. In contrast, the letter designation "B" (e.g., FIG. 9B and 10B) represents a situation where a random combination of drops (5 or 10) is selected to "average" the expected volume difference between nozzles. Finally, the letter designation "C" (e.g., FIG. 9C and 10C) represents a situation where scanning and nozzle firing are dependent on a specific total ink volume per target area that attempts to minimize the total fill distribution across the target area. In these various figures, the variation per nozzle is assumed to match the variation observed in an actual device, with each vertical axis representing the total fill volume in pL and each horizontal axis representing the number of target areas, e.g., pixel wells or pixel color components. Note that the emphasis in these figures is to show the variation in the total fill volume assuming randomly distributed drop variation about an assumed average. For Figures 9A-9C, the average volume per nozzle is assumed to be slightly below 10.00 pL per nozzle, and for Figures 10A-10C, the average drop volume per nozzle is assumed to be slightly above 10.00 pL per nozzle.

[0111] The first graph 901 depicted in FIG. 9A shows the volume variations per well assuming nozzle drop volume differences without attempting to mitigate these differences. Note that these variations can be extreme (e.g., by peak 903), with a range of total fill volume of approximately ±2.61%. As noted, the average of the five drops is just under 50.00 pL, and FIG. 9A shows two sets of sample tolerance ranges around this average, including a first range 905 representing a ±1.00% range around this value, and a second range 907 representing a ±0.50% range around this value. As seen by the numerical peaks and troughs (e.g., peak 903) that fall outside either range, such a printing process would result in a large number of wells that would fail to meet specifications (e.g., either one or the other of these ranges).

[0112] A second graph 911 depicted in FIG. 9B shows volume variation per well using a randomized set of five nozzles per well in an attempt to statistically average out the effects of droplet volume variation. Note also that such techniques do not allow for precise production of a specific volume of ink in any particular well, nor does such a process guarantee a total volume within range. For example, while the percentage of fill volumes that fall outside of specification represents a much better case than depicted by FIG. 9A, there are still situations where individual wells (as identified by trough 913) fall outside of specification, e.g., the ±1.00% and ±0.50% variations represented by numbers 905 and 907, respectively. In such cases, the min / max error is ±1.01%, reflecting an improvement using random mixing over the data depicted in FIG. 9A.

[0113] FIG. 9C represents a third example using a specific combination of droplets per nozzle according to the techniques described above. Specifically, graph 921 shows that the variation is entirely within the ±1.00% range, and very close to meeting the ±0.50% range for all represented target areas. Again, these ranges are represented by numbers 905 and 907, respectively. In this example, five specifically selected droplet volumes are used to fill the wells in each scan line, with printhead / substrate shifts as appropriate, for each pass or scan. The min / max error is ±0.595%, reflecting further improvement with this form of "smart mixing." Note that improvements and data observations will be consistent for any form of intelligent droplet volume combination to achieve a specific fill or tolerance range, for example, where offsets between nozzle rows (or multiple printheads) are used, or multiple preselected drive waveforms are used to enable specifically selected droplet volume combinations.

[0114] As described, Figures 10A-10C present similar data, but assume a combination of 10 droplets per well with an average droplet volume of approximately 10.30 pL per nozzle. Specifically, graph 1001 in Figure 10A represents a case where no attention is paid to mitigating droplet volume differences, graph 1011 in Figure 10B represents a case where droplets are randomly applied in an attempt to statistically "average" the volume differences, and graph 1021 in Figure 10C represents a case of planned mixing of specific droplets (to achieve the average fill volume of Figures 10A / 10B, i.e., approximately 103.10 pL). These various figures show tolerance ranges of ±1.00% and ±0.50% variation about this average, represented using arrows 1005 and 1007, respectively. Each of the figures further shows respective peaks 1003, 1013, and 1023, represented by the variation. However, note that FIG. 10A represents a ±2.27% variation around the target, FIG. 10B represents a ±0.707% variation around the target, and FIG. 10C represents a ±0.447% variation around the target. From this, it can be seen that the "random droplet" solution of Figure 10B achieves a ±1.00% tolerance range on the average, rather than a ±0.50% range. In contrast, the solution depicted by Figure 10C is seen to meet both tolerance ranges, demonstrating that it is possible to constrain the variation to be within specification, while still allowing for variation in droplet combinations between wells.

[0115] One optional embodiment of the techniques described in this disclosure is as follows: For a printing process in which nozzles with a drop volume standard deviation of x% are used to deposit a total fill volume with a maximum variation of ±y%, conventionally there are some means of ensuring that the total fill volume will vary by ±y%. This presents a potential problem. Droplet averaging techniques (e.g., as represented by the data seen in Figures 9B and 10B) statistically reduce the standard deviation of the total volume across a target area to x% / (n)1 / 2, where n is the average number of drops required per target area to achieve the desired fill volume. However, even with such a statistical approach, there is no mechanism for ensuring with certainty that the actual target area fill volume will in fact be within the maximum error limits of ±y%, especially when y and n are small. The techniques discussed herein provide a mechanism for providing such confidence by guaranteeing a known percentage of the target area to achieve a composite fill within ±y%. Thus, one optional embodiment provides a method for generating control data or controlling a printer, and related devices, systems, software, and improvements, where the standard deviation of the amount across the target area is better than x% / (n)1 / 2 (e.g., significantly better than x% / (n)1 / 2). In a specific implementation, this condition is met in a situation where the print head nozzles are used simultaneously to deposit droplets in each row (e.g., each pixel well) of the target area in each scan.

[0116] Using a set of basic techniques for combining droplets such that the sum of their volumes is specifically selected to meet the specific targets thus described, reference is now made to a more detailed description of specific devices and applications that can benefit from these principles. This discussion is intended to be non-limiting, i.e., to describe a small number of specifically contemplated implementations for practicing the methods introduced above.

[0117] As seen in FIG. 11, the multi-chamber processing apparatus 1101 includes several general modules or subsystems, including a transfer module 1103, a printing module 1105, and a processing module 1107. Each module maintains a controlled environment, such that, for example, printing can be performed by the printing module 1105 in a first controlled atmosphere, and other processing, e.g., another deposition process such as an inorganic encapsulation layer deposition or a curing process (e.g., for the printed material), can be performed in a second controlled atmosphere. The apparatus 1101 uses one or more mechanical handlers to move substrates between modules without exposing the substrates to an uncontrolled atmosphere. Other substrate handling systems and / or specific devices and control systems can be used within any given module that are adapted to the processing performed for that module.

[0118] Various embodiments of the transfer module 1103 can include an input load lock 1109 (i.e., a chamber that provides buffering between different environments while maintaining a controlled atmosphere), a transfer chamber 1111 (also having a handler for transporting the substrate), and an atmospheric buffer chamber 1113. Other substrate handling mechanisms can be used within the print module 1105, such as a floating table for stable support of the substrate during the printing process. In addition, an xyz motion system, such as a split axis or gantry motion system, can be used for precise positioning of at least one print head relative to the substrate, as well as providing a y-axis transport system for transport of the substrate through the print module 1105. It is also possible to use multiple inks for printing within the print chamber, e.g., using respective print head assemblies, so that, e.g., two different types of deposition processes can be performed within the print module in a controlled atmosphere. The printing module 1105 can include a gas enclosure 1115 that houses the inkjet printing system along with means for introducing an inert atmosphere (e.g., nitrogen, a noble gas, another similar gas, or a combination thereof) and otherwise controlling the atmosphere for environmental conditioning (e.g., temperature and pressure), gas constituents, and the presence of particulate matter.

[0119] The processing module 1107 can include, for example, a transfer chamber 1116, which also has a handler for transporting the substrate. In addition, the processing module can also include an output load lock 1117, a nitrogen stack buffer 1119, and a curing chamber 1121. In some applications, the curing chamber can be used to cure the monomer film into a uniform polymer film, for example, using a thermal or ultraviolet radiation curing process.

[0120] In one application, the apparatus 1101 is adapted for mass production of liquid crystal display screens or OLED display screens in bulk, for example fabricating an array of eight screens at a time on a single large substrate. These screens can be used as display screens for televisions and other forms of electronic devices. In a second application, the apparatus can be used in a more similar manner for mass production of solar panels.

[0121] When applied to the drop volume combination techniques described above, the print module 1105 can be advantageously used in display panel manufacturing to deposit one or more layers, such as light filtering layers, emissive layers, barrier layers, conductive layers, organic or inorganic layers, encapsulation layers, and other types of materials. For example, the depicted apparatus 1101 can be loaded with a substrate and controlled to move the substrate back and forth between various chambers to deposit and / or cure or harden one or more printed layers, all in an uninterrupted manner with intervening exposure to an uncontrolled atmosphere. Optionally, ink drop measurements (if used in connection with the depicted system) can be taken as the substrate is moved or processed in any chamber. For example, a first substrate can be loaded via the input load lock 1109, and during this process, a print head assembly in the print module 1105 can be engaged with a drop measurement device to take drop measurements on some of the printing nozzles. In embodiments with many print nozzles, the drop measurements can be periodic and intermittent, such that during various print cycles, different nozzles, representing a cyclical and incremental portion of all nozzles of the print assembly, are calibrated and associated drops are measured to generate statistical models for each of the drop volume, ejection angle (relative to normal), and velocity. A handler located in the transfer module 1103 moves the first substrate from the input load lock 1109 to the print module 1105, at which time the drop measurements are disengaged and the print head assembly is moved to a position for active printing. Following completion of the printing process, the first substrate can be moved to the processing module 1107 for curing. Again, a new cycle of drop measurements can be performed and, optionally, a second substrate can be loaded into the input load lock 1109 (if supported by the system). Many other alternatives and process combinations are possible. By repeated deposition of subsequent layers, for example by moving the first substrate back and forth for repeated iterations of printing and curing, a controlled amount per target area, each of the cohesive layer properties can be built up to suit any desired application.In an alternative embodiment, an output load lock 1117 can be used to transfer the first substrate to a second printer (e.g., for continuous pipeline printing of new layers, e.g., new OLED material layers, or encapsulation or other layers). Again, it is noted that the techniques described above are not limited to display panel manufacturing processes, and many different types of tools can be used. For example, the configuration of the apparatus 1101 can be varied to place the various modules 1103, 1105, and 1107 in different juxtapositions, and additional or fewer modules can also be used. As represented by numerals 1121 and 1123, a computing device (e.g., a processor) running suitable software can be used to control the various processes and to perform optional drop measurements as described above in conjunction with other processes, i.e., to minimize downtime of the apparatus, to keep the drop measurements as up to date as possible while maintaining robust statistical models, and to stack the drop measurement process as much as possible to overlap other system processes.

[0122] While Figure 11 provides one example of a set of linked chambers or processing components, clearly many other possibilities exist. The ink drop measurement and deposition techniques introduced above can be used to control processing processes performed with the device depicted in Figure 11, or indeed by any other type of deposition equipment.

[0123] FIG. 12 provides a block diagram showing various subsystems of one apparatus that can be used to fabricate a device having one or more layers as specified herein. Coordination across the various subsystems is provided by a processor 1203, acting under instructions provided by software (not shown in FIG. 12). During the fabrication process, the processor provides data to the print head 1205 to cause the print head to eject various amounts of ink in response to nozzle firing commands. The print head 1205 typically has multiple inkjet nozzles arranged in a row (or rows of an array) and associated reservoirs that allow the ejection of ink in response to activation of a piezoelectric or other transducer per nozzle, such transducer causing the nozzle to eject a controlled amount of ink in an amount governed by an electronic nozzle drive waveform signal applied to the corresponding piezoelectric transducer. If there are multiple print heads, there may be a processor for each print head, or one processor may control the entire print head assembly. Other firing mechanisms may also be used. Each print head applies ink to the substrate 1207 at different xy locations that correspond to grid coordinates within the various print cells as represented by the halftone print image. Positional variation is accomplished by both the print head motion system 1209 and the substrate handling system 1211 (e.g., causing the print to represent one or more swaths of locations across the substrate). In one embodiment, the print head motion system 1209 moves the print head back and forth along a traveler while the substrate handling system provides stable substrate support and "y" dimension transport of the substrate to enable "split-axis" printing of any portion of the substrate. The substrate handling system provides relatively fast y dimension transport while the print head motion system 1209 provides relatively slow x dimension transport. In another embodiment, the substrate handling system 1211 can provide both x and y dimension transport. In yet another embodiment, the primary transport can be provided entirely by the substrate handling system 1211. An image capture device 1213 can be used to assist in locating, aligning and / or error detection of any fiducials.

[0124] The apparatus also includes an ink delivery system 1215 and a printhead maintenance system 1217 that supports the printing operations. The printheads may be periodically calibrated or undergo a maintenance process, and to this end, during a maintenance sequence, the printhead maintenance system 1217 is used to perform appropriate priming, ink or gas purging, testing and calibration, and other operations as appropriate for the particular process.

[0125] As previously introduced, the printing process can be performed in a controlled environment, i.e., in a manner that presents a reduced risk of contaminants that may degrade the effectiveness of the deposited layer. To this effect, the apparatus includes a chamber control subsystem 1219 that controls the atmosphere within the chamber, as represented by functional block 1221. Optional process variations are described. The method may include ejecting the deposition material in the presence of an ambient nitrogen gas atmosphere so that the deposition material can be ejected in a uniform manner.

[0126] As mentioned above, in the embodiments disclosed herein, the individual drop volumes are combined to achieve a specific fill volume per target area, which is selected depending on the target fill volume. A specific fill volume can be planned for each target area, and the fill value varies about the target value within an acceptable tolerance range. For such embodiments, the drop volume is specifically measured in a manner that depends on the ink, the nozzle, the driving waveform, and other factors. To this end, reference numeral 1223 represents an optional drop volume measurement system in which the drop volume 1225 is measured for each nozzle and for each driving waveform, and then stored in memory 1227. Such a drop measurement system can be an optical strobe camera or a laser scanning device (or other volume measurement tool) that is incorporated into a commercially available printing device, as mentioned above. In one embodiment, such a device uses a non-imaging technique (e.g., using a simple optical detector instead of image processing software acting on pixels) to achieve real-time or near real-time measurement of individual drop volumes, deposition flight angles or trajectories, and drop velocities. This data is provided to the processor 1203 either during printing or during a one-time intermittent or periodic calibration operation. Optionally, a pre-arranged set of firing waveforms can also be associated with each nozzle for later use in generating a combination of droplets per particular target area, as indicated by numeral 1229. When such a set of waveforms is used in an embodiment, drop volume measurements are advantageously calculated for each waveform during calibration using a drop measurement system 1223 for each nozzle. Providing a real-time or near real-time drop volume measurement system greatly enhances reliability in providing target area volume fill within a desired tolerance range, since measurements can be taken as needed and processed (e.g., averaged) to minimize statistical volume measurement errors.

[0127] Numeral 1231 refers to the use of print optimization software running on processor 1203. More specifically, this software uses this information to plan the print in a way that combines drop volumes appropriately to obtain a particular fill volume per target area based on a statistical model of the drop volumes 1225 (measured in situ or otherwise provided). In one embodiment, the above example allows the total volume to be planned down to a resolution of 0.01 pL or better within a certain error tolerance, despite the fact that the drop measurement device may have a lower accuracy associated with individual drop measurements, i.e., by using the techniques described herein to build a statistical model of the drop volume per nozzle and per nozzle / waveform combination, a degree of statistical accuracy can be estimated from that represented by the accuracy of the drop measurement system. Once the print is planned, the processor calculates printing parameters such as the number and order of scans, drop diameters, relative drop firing times, and similar information, and builds a print image that is used to determine nozzle firing for each scan. In one embodiment, the print image is a half-toe-in image. In another embodiment, the printhead has multiple nozzles, as many as 10,000. As described below, each drop can be represented according to a time value and a firing value (e.g., data representing a firing waveform or data indicating whether the drop will be fired "digitally"). In embodiments where geometric steps and binary nozzle firing decisions are relied upon to vary the drop volume per well, each drop can be defined by a bit of data, a step value (or scan number), and a position value indicating where the drop is to be placed. In implementations where the scan represents continuous motion, a time value can be used as the equivalent of a position value. Whether based on time / distance or absolute position, the value represents a position relative to a reference (e.g., a synchronization mark, position, or pulse) that precisely identifies where and when the nozzle should be fired. In some embodiments, multiple values ​​can be used.For example, in one specifically contemplated embodiment, a synchronization pulse is generated for each nozzle in a manner corresponding to each micron of relative printhead / substrate motion during scanning. For each synchronization pulse, each nozzle is programmed with (a) an offset value representing an integer clock cycle delay before the nozzle is fired, (b) a 4-bit waveform select signal representing one of 15 waveform selections preprogrammed into memory dedicated to the particular nozzle driver (i.e., one of 16 possible values ​​specifying the "off" or non-firing state of the nozzle), and (c) a repeatability value specifying whether the nozzle should be fired only once, once every sync pulse, or once every n sync pulses. In such a case, the waveform selection and address of each nozzle is associated by the processor 1203 with specific drop volume data stored in memory 1227, and the firing of a particular waveform from a particular nozzle represents a planned determination that a particular corresponding drop volume is to be used to deliver total ink to a particular target area of ​​the substrate.

[0128] Figures 13A-15D will be used to introduce other techniques that can be used to combine different drop volumes to obtain precise tolerance fill volumes for each target area. In the first technique, rows of nozzles can be selectively offset relative to one another in the print assembly during printing (e.g., during scanning). This technique is introduced with reference to Figures 13A-13B. In the second technique, nozzle drive waveforms can be used to adjust the piezoelectric transducer firing and therefore the properties of each ejected drop (including volume). Figures 14A-14B are used to discuss some options. Finally, in one embodiment, a set of multiple alternative drop firing waveforms is pre-calculated and made available for use with each print nozzle. This technique and associated circuitry are discussed with reference to Figures 15A-B.

[0129] FIG. 13A provides a plan view 1301 of a print head 1303 traversing a substrate 1305 in a scanning direction indicated by arrow 1307. The substrate can now be seen to consist of a number of pixels 1309, each pixel having wells 1309-R, 1309-G, and 1309-B associated with a respective color component. Again, it should be noted that this depiction is merely an example, i.e., techniques as used herein can be applied to any layer of a display (e.g., not limited to individual color components, and not limited to color-imparting layers), and these techniques can also be used to create other than display devices. In this case, assuming that the print head deposits one ink at a time and that the ink is color component specific, it is intended that a separate printing process would be performed for each well of the display, one for each color component. Thus, if a first process is being used to deposit an ink specific to red light production, then only the first well of each pixel, such as well 1309-R of pixel 1309 and the similar well of pixel 1311, will receive ink in the first printing process. In the second printing process, only the second well (1309-G) of pixel 1309 and the similar well of pixel 1311 will receive the second ink, etc. The various wells are thus considered to be three different overlapping arrays (in this case fluid containers or wells) of the target area.

[0130] Print head 1303 includes several nozzles, as represented using numbers 1313, 1315, and 1317. In this case, each of the numbers refers to a separate row of nozzles, where the row extends along the column axis 1318 of the substrate. It can be seen that nozzles 1313, 1315, and 1317 form a first row of nozzles with respect to substrate 1305, and nozzle 1329 represents a second row of nozzles. As depicted by FIG. 13A, the nozzles are not aligned with the pixels, and as the print head scans across the substrate, some nozzles will pass over the target area while others will not. Furthermore, the figure shows that print nozzles 1313, 1315, and 1317 are precisely aligned with the center of the row of pixels starting from pixel 1309, while print nozzle 1329 will also pass over the row of pixels starting from pixel 1311, but the alignment of print nozzle 1329 is not aligned with the center of pixel 1311 and beyond. 1B and 4D , it may be desirable to have a group of nozzles generally aligned with each row. This alignment / misalignment of the column of nozzles with the row of wells is depicted by lines 1325 and 1327, which respectively represent the centers of the printing wells that are to receive ink. In many applications, the precise location where the droplets are deposited within the target area is not important and such misalignment is acceptable (e.g., as discussed in connection with FIGS. 1B and 4D , it may be desirable to have a group of nozzles generally aligned with each row).

[0131] FIG. 13B provides a second view 1331 in which all three rows of nozzles (or individual printheads) can be seen to be rotated approximately 30 degrees relative to axis 1318. This optional capability was previously referenced by numeral 338 in FIG. 3B. More specifically, the rotation changes the spacing of the nozzles along the column axis 1318, with each column of nozzles aligned with well centers 1325 and 1327 or otherwise adjusted to increase the apparent density of nozzles per target print target area during scanning. Note, however, that with such rotation and scanning motion 1307, nozzles from each column of nozzles will traverse the column of pixels (e.g., 1309 and 1311) at different relative times and therefore potentially have different position firing data (e.g., different timing for firing droplets). Methods for adjusting the firing data of each nozzle will be discussed below in connection with FIGS. 15A-B.

[0132] As represented by FIG. 13C, in one embodiment, a printhead assembly optionally given multiple printheads or rows of nozzles can have such rows selectively offset from one another. That is, FIG. 13C provides another plan view in which each of printheads (or nozzle rows) 1319, 1321, and 1323 are offset relative to one another, as represented by offset arrows 1353 and 1355. These arrows represent the use of optional motion mechanisms, one for each row of nozzles, that allow selective offsetting of the corresponding row relative to the printhead assembly. This provides different combinations of nozzles (and associated specific drop volumes) with each scan (e.g., by numeral 1307), and thus different specific drop combinations. For example, in such an embodiment, as depicted by FIG. 13C, such offsets allow both nozzles 1313 and 1357 to be aligned with centerline 1325 and thus have their respective drop volumes combined in a single pass. It should be noted that this embodiment is considered a particular case of an embodiment that varies the geometric step, e.g., even if the geometric step size between successive scans of the printhead assembly 1303 relative to the substrate 1305 is fixed, each such scan movement of a given row of nozzles is effectively positioned at a variable offset or step using the movement mechanism relative to the position of the given row in other scans. Also, or alternatively, such offsets can be performed to adjust the effective print grid to provide various spacing between deposited drops. Consistent with the principles previously introduced, the use of optional offsets allows the drop volume per individual nozzle to be tallied with a particular combination (or drop set) for each well, but with a reduced number of scans or passes. For example, in the embodiment depicted in FIG. 13C, three drops can be deposited in each target area (e.g., wells for red color component) in each scan, and the offsets allow for planned variation of the drop volume and / or spatial combination.

[0133] Figure 13D illustrates a cross-section of the completed display for one well (e.g., well 1309-R from Figure 13A) taken in the direction of the scan. Specifically, this figure shows a substrate 1352 of a flat panel display, specifically an OLED display. The depicted cross-section shows active area 1353 and conductive ends 1355, which receive electronic signals that control the display (including the color of each pixel). A small oval region 1361 in the figure extends from the right side of the figure to illustrate the layers in the active area above the substrate 1352. 13. These layers include an anode layer 1369, a hole injection layer ("HIL") 1371, a hole transport layer ("HTL") 1373, an emissive or light-emitting layer ("EML") 1375, an electron transport layer ("ETL") 1377, and a cathode layer 1378, respectively. Additional layers such as polarizers, barrier layers, primers, and other materials may also be included. In some cases, an OLED device may include only some of these layers. When the depicted stack is finally operated following fabrication, an electric current causes recombination of electrons and "holes" in the EML, resulting in the emission of light. The anode layer 1369 may comprise one or more transparent electrodes common to several color components and / or pixels; for example, the anode may be formed from indium tin oxide (ITO). The anode layer 1369 may also be reflective or opaque, and other materials may be used. The cathode layer 1378 typically consists of a patterned electrode to provide selective control over each color component for each pixel. The cathode layer can comprise a reflective metal layer, such as aluminum. The cathode layer can also comprise an opaque or transparent layer, such as a thin layer of metal combined with a layer of ITO. Together, the cathode and anode serve to source and collect electrons and holes that enter and / or pass through the OLED stack. The HIL 1371 typically functions to transport holes from the anode into the HTL. The HTL 1373 typically functions to transport holes from the HIL into the EML while impeding the transport of electrons from the EML into the HTL. The ETL 1377 typically functions to transport electrons from the cathode into the EML while impeding the transport of electrons from the EML into the ETL. These layers thereby act together to supply electrons and holes into the EML 1375 and confine them therein so that they can recombine to generate light. Typically, the EML consists of separately controlled active materials for each of the three primary colors, red, green, and blue, for each pixel of the display, represented in this case by the red light generating material as described.

[0134] The layers in this active area can be degraded through exposure to oxygen and / or moisture. It is therefore desirable to enhance OLED lifetime by encapsulating these layers on both the face and sides (1362 / 1363) of the layers opposite the substrate, as well as on the outer edges. The purpose of the encapsulation is to provide an oxygen-resistant and / or moisture barrier. Such encapsulation can be formed in whole or in part through the deposition of one or more thin film layers.

[0135] The techniques discussed herein can be used to deposit any of these layers, as well as combinations of such layers. Thus, in one contemplated application, the techniques discussed herein provide ink volumes for EML layers for each of the three primary colors. In another application, the techniques discussed herein are used to provide ink volumes for HIL layers, and so on. In yet another application, the techniques discussed herein are used to provide ink volumes for one or more OLED encapsulation layers. The printing techniques discussed herein can be used to deposit organic or inorganic layers, as appropriate for the process technology, and layers for other types of display and non-display devices.

[0136] FIG. 14A is used to introduce nozzle drive waveform adjustments and the use of alternative nozzle drive waveforms to provide different ejected drop volumes from each nozzle of a printhead. A first waveform 1403 is considered a single pulse consisting of a quiet period 1405 (0 volts), an upslope 1413 associated with the decision to fire the nozzle at time t2, a voltage pulse or signal level 1407, and a downslope 1411 at time t3. The effective pulse width, represented by number 1409, is of duration approximately equal to t3-t2, depending on the difference between the upslope and downslope of the pulse. In one embodiment, any of these parameters (e.g., upslope, voltage, downslope, pulse duration) can be adjusted to affect the ejection drop volume of a given nozzle. The second waveform 1423 is similar to the first waveform 1403, except that it represents a larger drive voltage 1425 relative to the signal level 1407 of the first waveform 1403. Due to the larger pulse voltage and finite rising slope 1427, it takes longer to reach this higher voltage, and similarly the falling slope 1429 is typically delayed relative to the similar slope 1411 from the first waveform. The third waveform 1433 is also similar to the first waveform 1403, except that in this case a different rising slope 1435 and / or a different falling slope 1437 can be used in place of the slopes 1413 and 1411 (e.g., through adjustment of the nozzle firing path impedance). The different slopes can be either steeper or shallower (steeper in the depicted case). In contrast, in the fourth waveform 1443, the pulse is made longer, for example using a delay circuit (e.g., a voltage controlled delay line) to both increase the time of the pulse at a given signal level (as represented by numeral 1445) and to delay the falling edge of the pulse as represented by numeral 1447. Finally, the fifth waveform 1453 represents the use of multiple discrete signal levels as also providing a means of pulse shaping. For example, this waveform can be seen to include a time at the first described signal level 1407, then a ramp up to a second signal level 1455 applied partway between times t3 and t2. It can be seen that the falling section of this waveform 1457 is delayed after the falling edge 1311 due to the larger voltage.

[0137] Any of these techniques can be used in combination with any of the embodiments discussed herein. For example, the drive waveform adjustment technique can be used to vary the droplet volume within a small range, optionally after the scanning motion and nozzle firing have already been planned, to mitigate line effects. Designing the waveform variation in such a manner that the second tolerance matches the specification promotes deposition of a high quality layer with planned non-random or planned random variation. For example, returning to the previously introduced assumption that a TV manufacturer specifies a fill volume of 50.00 pL ± 0.50%, a fill volume per area can be calculated within a first range of 50.00 pL ± 0.25% (49.785 pL to 50.125 pL) with a non-random or random technique applied to the waveform variation, where the variation statistically provides a volume variation of only ± 0.025 pL per droplet (considering the 5 droplets required to reach the total fill volume). Alternatively or additionally, the drive waveform variation can be used to affect the velocity or trajectory (flight angle) of the ejected droplets. For example, in one process, droplets may be required to meet a predetermined set of criteria regarding volume and / or velocity and / or trajectory, and if the droplets fall outside of accepted norms, the nozzle driving waveform may be adjusted until compliance is achieved. Alternatively, a set of predetermined waveforms may be measured, with some of these waveforms selected based on their conformance to desired norms. Clearly, many variations exist.

[0138] As mentioned above, multiple signal levels can be used to shape the pulse in one embodiment, represented by the fifth waveform 1453 from Figure 14A. This technique is further discussed with reference to Figure 14B.

[0139] That is, in one embodiment, the waveforms can be predefined as a series of discrete signal levels, e.g., defined by digital data, and the drive waveforms are generated by a digital-to-analog converter (DAC). Number 1451 in FIG. 14B refers to waveform 1453, having discrete signal levels 1455, 1457, 1459, 1461, 1463, 1465, and 1467. In this embodiment, each nozzle driver includes circuitry for receiving and storing up to 16 different signal waveforms, each waveform defined by a series of up to 16 signal levels, each represented as a multi-bit voltage and duration. That is, in such an embodiment, by defining different durations for one or more signal levels, the pulse width can be effectively varied, and the drive voltage can be waveform-shaped in a manner selected to provide subtle droplet size variation, e.g., droplet volumes are metered to provide specific volume increments, such as in 0.10 pL increments. Thus, in such an embodiment, the corrugation provides the ability to adjust the drop volume to be closer to the target drop volume value, and when combined with other specific drop volumes, such as using the techniques exemplified above, these techniques facilitate precise fill volumes per target area. In addition, however, these corrugation techniques also facilitate measures to reduce or eliminate line effects, for example, in one optional embodiment, drops of specific volumes are combined as discussed above, but the final drop (or drops) are selected in a manner that provides variation to the boundaries of a desired tolerance range. In another embodiment, a predetermined waveform can be applied, with optional further corrugation or timing applied as appropriate to adjust the drop volume, velocity, and / or trajectory. In yet another example, the use of nozzle drive waveform alternatives provides a mechanism to plan volumes such that further corrugation is not necessary.

[0140] Typically, the effect of different drive waveforms and the resulting drop volume is measured in advance. For each nozzle, up to 16 different drive waveforms are then stored in 1k synchronous random access memory (SRAM) per nozzle for later selective use in providing discrete volume variations as selected by software. With the different drive waveforms at hand, each nozzle is then instructed on a drop-by-drop basis as to which waveform to apply via programming of data that achieves the particular drive waveform.

[0141] FIG. 15A illustrates such an embodiment, generally designated by numeral 1501. Specifically, a processor 1503 is used to receive data defining the intended fill volume per target area for a particular layer of material to be printed. As represented by numeral 1505, this data may be a layout file or a bitmap file defining the drop volume per grid point or position address. A series of piezoelectric transducers 1507, 1508, and 1509 generate associated ejected drop volumes 1511, 1512, and 1513, respectively, that depend on many factors, including nozzle drive waveforms and printhead-to-printhead manufacturing variations. During a calibration operation, each set of variables is tested for its effect on drop volume, including nozzle-to-nozzle variations and the use of different drive waveforms, given the particular ink that will be used. If desired, this calibration operation can be made dynamic, for example to respond to changes in temperature, nozzle clogging, or other parameters. This calibration is represented by a drop measurement device 1515, which provides measurement data to the processor 1503 for use in managing the print plan and the next print. In one embodiment, this measurement data is calculated on the fly, taking literally minutes (e.g., for thousands of printhead nozzles and potentially a large number of possible nozzle firing waveforms), e.g., only 30 minutes for thousands of nozzles, and preferably even less. In another embodiment, as described, such measurements can be made iteratively, that is, updating different portions of the nozzles at different times. Non-imaging (e.g., interferometric) techniques can optionally be used as previously described, potentially resulting in tens of drop measurements per nozzle, covering tens to hundreds of nozzles per second. This data and any associated statistical models (and averages) can be stored in memory 1517 for use in processing the layout or bitmap data 1505 as it is received. In one implementation, the processor 1503 is part of a computer that is remote from the actual printer, while in a second implementation, the processor 1503 is integrated with either the processing mechanism for the product (e.g., a system for processing a display) or the printer.

[0142] To effect the ejection of droplets, a set of one or more timing or synchronization signals 1519 are received for use as a reference and these are passed through a clock tree 1521 for delivery to each nozzle driver 1523, 1524, and 1525 to generate drive waveforms for a particular nozzle (1527, 1528, and 1529, respectively). Each nozzle driver has one or more registers 1531, 1532, and 1533, respectively, which receive multi-bit programming data and timing information from the processor 1503. Each nozzle driver and its associated registers receive one or more dedicated write enable signals (wen) for purposes of programming the registers 1531, 1532, and 1533, respectively. In one embodiment, each of the registers comprises a significant amount of memory, including a 1k SRAM that stores a number of predefined waveforms, and programmable registers to select between these waveforms and otherwise control the waveform generation. Although the data and timing information from the processor is depicted as multi-bit information, this information can be provided via either serial or parallel bit connections to each nozzle (as seen in FIG. 15B, discussed below, in one embodiment this connection is serial as opposed to the parallel signal representation seen in FIG. 15A).

[0143] For a given deposit, printhead, or ink, the processor selects a set of 16 drive waveforms for each nozzle that can be selectively applied to generate droplets. Note that this number is arbitrary, e.g., one design may use 4 waveforms while another may use 4000 waveforms. These waveforms are advantageously selected to provide the desired variation in output drop volume for each nozzle, e.g., to have each nozzle have at least one waveform selection that produces a near-ideal drop volume (e.g., an average drop volume of 10.00 pL), and to provide a range of deliberate volume variations from each nozzle. In various embodiments, the same set of 16 drive waveforms is used for all of the nozzles, but in the depicted embodiment, there are 16 separate, possibly unique, waveforms predefined for each nozzle, each waveform providing a respective drop volume characteristic.

[0144] During printing, data selecting one of the predefined waveforms is then programmed into each nozzle's respective register 1531, 1532, or 1533 on a nozzle-by-nozzle basis to control the deposition of each drop. For example, considering a target drop volume of 10.00 pL, the nozzle driver 1523 can be configured to set one of 16 waveforms, corresponding to one of 16 different drop volumes, through writing data to the register 1531. Once the drop volumes and associated distributions per nozzle (and per waveform) have been registered by the processor 1503 and stored in memory in aid of producing the desired target fill, the volume produced by each nozzle will have been measured by the drop measurement device 1515. By programming the register 1531, the processor can define whether it wants a particular nozzle driver 1523 to output a processor-selected one of the 16 waveforms. Additionally, the processor can program registers to have a per-nozzle delay or offset to the firing of the nozzles for a given scan line (e.g., to align each nozzle with a grid traversed by the printhead, to correct for errors including velocity or trajectory errors, and for other purposes), the offset being accomplished by a counter that makes a particular nozzle (or firing waveform) asymmetric by a programmable number of timing pulses for each scan. To provide an example, if the results of the drop measurement indicate that one particular drop tends to have a lower than expected velocity, the corresponding nozzle waveform can be triggered earlier (e.g., advanced in time by shortening the dead time before the active signal level used for piezoelectric actuation), and conversely, if the results of the drop measurement indicate that one particular drop has a relatively high velocity, the waveform can be triggered later, etc. Other examples are clearly possible, for example, in some embodiments, slow drop velocities can be prevented by increasing the drive strength (i.e., the signal level and associated voltage used to drive the piezoelectric actuator of a given nozzle).In one embodiment, the synchronization signal delivered to all nozzles occurs at a defined time interval (e.g., 1 microsecond) for synchronization purposes, while in another embodiment, the synchronization signal is adjusted to the printer motion and substrate topography, for example to fire incremental relative motion between the print head and the substrate every micron. A high speed clock (φhs) is run thousands of times faster than the synchronization signal, for example at 100 megahertz, 33 megahertz, etc., and in one embodiment, multiple different clocks or other timing signals (e.g., strobe signals) can be used in combination. The processor also programs values ​​that define the grid spacing, and in one implementation, the grid spacing is common across the set of available nozzles, although this need not be true for each implementation. For example, in some cases, a regular grid can be defined, where all nozzles fire "every 5 microns." This grid can be specific to the printing system, the substrate, or both. Thus, in one optional embodiment, the grid can be defined for a particular printer, with the synchronization frequency or nozzle firing pattern used to effectively transform the grid to match the substrate topography, which is unknown a priori. In another contemplated embodiment, a memory is shared across all nozzles that allows the processor to pre-store several different grid intervals (e.g., 16) that are then read out from all nozzles (e.g., to define an irregular grid) so that the processor can select (on demand) a new grid interval. For example, in an implementation where a nozzle is firing for all color component wells of an OLED (e.g., to deposit a non-color specific layer), three or more different grid intervals can be applied sequentially by the processor in a round-robin fashion. Clearly, many design alternatives are possible. Note that the processor 1503 can also dynamically reprogram the registers of each nozzle during operation, i.e., a sync pulse is applied as a trigger to activate any programmed waveform pulses set in that register, and if new data is received asynchronously before the next sync pulse, the new data will be applied with the next sync pulse.The processor 1503 also controls the start and speed of the scan (1535), in addition to setting parameters for sync pulse generation (1536). In addition, the processor controls the rotation of the print head (1537) for various purposes described above. In this manner, each nozzle can be fired in unison (or simultaneously) using any one of 16 different waveforms for each nozzle at any time (i.e., with any "next" sync pulse), and the selected firing waveform can be interleaved with any other of the 16 different waveforms dynamically between fires during a single scan.

[0145] FIG. 15B shows additional details of a circuit (1541) used in such an embodiment to generate an output nozzle drive waveform for each nozzle, the output waveform being represented in FIG. 15B as "nzzl-drv.wvfm". More specifically, the circuit 1541 receives inputs of a synchronization signal, a single bit line carrying serial data ("data"), a dedicated write enable signal (we), and a high speed clock (φhs). A register file 1543 provides data for at least three registers that convey an initial offset, a lattice definition value, and a drive waveform ID, respectively. The initial offset is a programmable value that adjusts each nozzle to align with the start of the lattice, as described. For example, considering implementation variables such as multiple printheads, multiple rows of nozzles, different printhead rotations, nozzle firing speeds and patterns, and other factors, the initial offset can be used to align the drop pattern of each nozzle with the start of the lattice to account for delays and other factors. The offset can be applied differently across multiple nozzles, for example to rotate a grid or halftone pattern relative to the substrate topography or to correct for substrate misalignment. Similarly, as described, the offset can also be used to correct for abnormal speed or other effects. The grid definition value is a number representing the number of sync pulses to be "counted" before the programmed waveform is triggered, and in the case of an implementation printing a flat panel display (e.g., an OLED panel), the printed target area will likely have one or more regular intervals for different print head nozzles, corresponding to a regular (constant interval) or irregular (multiple interval) grid. As mentioned above, in one implementation, the processor keeps its own 16-entry SRAM to define up to 16 different grid intervals that can be read into a register circuit for all nozzles on demand. Thus, if the grid interval value is set to 2 (e.g., every 2 microns), each nozzle will fire at this interval. The drive waveform ID represents the selection of one of the pre-stored drive waveforms for each nozzle, and can be programmed and stored in many ways, depending on the embodiment.In one embodiment, the drive waveform ID is a 4-bit selection value and each nozzle has its own dedicated 1k byte SRAM to store up to 16 predefined nozzle drive waveforms stored as 16x16x4B entries. Briefly, each of the 16 entries for each waveform contains 4 bytes representing programmable signal levels, which in turn represent 2-byte resolution voltage levels and 2-byte programmable durations used to count the number of pulses of a high speed clock. Thus, each programmable waveform can consist of anything from 0-1 discrete pulses up to 16 discrete pulses (e.g., of duration equal to 1-255 pulses of a 33 megahertz clock), each of programmable voltage and duration.

[0146] Numerals 1545, 1546, and 1547 designate one embodiment of a circuit showing how a specific waveform can be generated for a given nozzle. A first counter 1545 receives a sync pulse to begin counting down an initial offset, triggered by the start of a new line scan. The first counter 1545 counts down in micron increments, and when it reaches zero, a trigger signal is output from the first counter 1545 to a second counter 1546. This trigger signal essentially starts the firing process of each nozzle for each scan line. The second counter 1546 then implements a programmable grid spacing in micron increments. While the first counter 1545 is reset in conjunction with a new scan line, the second counter 1546 is reset using the next edge of the high speed clock following its output trigger. The second counter 1546, when triggered, activates a waveform circuit generator 1547, which generates a selected drive waveform shape for a particular nozzle. As represented by dashed boxes 1548-1550, seen below the generator circuit, this latter circuit is based on a high speed digital to analog converter 1548, a counter 1549, and a high voltage amplifier 1550, timed according to a high speed clock (φhs). When a trigger from the second counter 1546 is received, the waveform generator circuit takes the number pair (signal level and duration) represented by the drive waveform ID value and generates a given analog output voltage according to the signal level value, and the counter 1549 is effective to hold a DAC output of the duration according to the counter. The associated output voltage level is then applied to the high voltage amplifier 1550 and output as the nozzle drive waveform. The next number pair is then latched from the register 1543 to define the next signal level value / duration, etc.

[0147] The depicted circuit provides an effective means of defining any desired waveform according to data provided by the processor 1503. The duration and / or voltage level associated with any particular signal level (e.g., a first "0" signal level that defines an offset to synchronization) can be adjusted if necessary to conform to grid geometry or to mitigate nozzles with anomalous speeds or flight angles. As described, in one embodiment, the processor predetermines a set of waveforms (e.g., 16 possible waveforms per nozzle) and then writes the definitions of each of these selected waveforms into the SRAM for each nozzle's driver circuit, and then the "fire time" determination of the programmable waveforms is accomplished by writing a 4-bit drive waveform ID into each nozzle register.

[0148] FIG. 15C illustrates how different waveforms and different configuration options can be used per nozzle. FIG. 15 provides a flowchart 1551 that discusses the process of determining whether a nozzle is to be driven by a printer. As represented by 1553, the system (e.g., one or more processors acting under instructions from suitable software) selects a set of predetermined nozzle driving waveforms. For each waveform and for each nozzle (1555), the drop volume is specifically measured, e.g., using a laser measurement device or a CCD camera, and a statistical model is constructed. These volumes are stored in a memory accessible to the processor, such as memory 1557. Again, the measured parameters may vary depending on the ink selection and many other factors. Thus, depending on these factors and the planned deposition activity, a calibration is performed. For example, in one embodiment 1561, the calibration is performed at the factory that manufactures the printhead or printer, and this data can be pre-programmed into the sales device (e.g., printer) or made available for download. Alternatively, for printers that possess an optional drop measurement device or system, these volume measurements can be performed on first use (1562), e.g., during initial device configuration. In yet another embodiment, measurements are made at each power or substrate cycle (1563), for example, whenever the printer is turned "on" or powered up from a low power state or otherwise transitioned to a state ready to print. As previously mentioned, for embodiments in which ejected drop volume is affected by temperature or other dynamic factors, calibrations can be made intermittently or periodically (1564), for example, after expiration of a defined time interval, when an error is detected, at each new substrate operation state (e.g., substrate loading and / or during loading), daily, or on some other basis. Other calibration techniques and schedules can also be used (1565).

[0149] Calibration techniques can optionally be performed in an offline process or during a calibration mode, as represented by process separation line 1566. As described, in one embodiment, such a process is completed in less than 30 minutes for potentially thousands of print nozzles and one or more associated nozzle firing waveforms. During online operation (or during a print mode), represented below this process separation line 1566, the measured drop volumes are used in selecting a set of drops per target area based on the specific measured drop volumes, such that the drop volumes for each set sum to a specific total volume within a defined tolerance range, as represented by numeral 1567. The volume per area can be selected based on a layout file, bitmap data, some other representation, as represented by numeral 1568. Based on these drop volumes and the allowable combinations of drop volumes for each target area, a firing pattern and / or scan path is selected, which in effect represents a specific combination of drops for each target area (i.e., one of the allowable set of combinations) that will be used for the deposition process, as represented by numeral 1569. As part of this selection or planning process 1569, an optimization function 1570 can be optionally employed to, for example, reduce the number of scans or passes to a number less than the product of the average number of droplets per target area multiplied by the number of rows (or columns) in the target area (e.g., less than what would be required for a row of nozzles rotated 90 degrees, stepping one row at a time, and depositing droplets in multiple passes for each row of the target area so that all nozzles in the row can be used in each scan for each affected target area). For each scan, the printhead can be moved and the nozzles can be programmed with waveform data per nozzle to achieve droplet deposition instructions according to a bitmap or layout file, these functions are variously represented by numerals 1571, 1573, and 1575 in FIG. 15C. After each scan, the process is repeated for the next scan, by numeral 1577.Optionally, these techniques and their implementations may be embodied in printer control files 1579 that are created for subsequent or repeatable use in controlling the emission of ink at specific times.

[0150] Again, note that several different implementations are described above that are optional with respect to each other. First, in one embodiment, the drive waveform is not varied, but each nozzle , remains constant for . Drop volume combinations are generated by using variable geometric steps representing printhead / substrate offsets to overlap different nozzles with different rows of the target area, as needed. Using the measured drop volumes per nozzle, this process allows for combinations of specific drop volume averages to achieve very specific fill volumes per target area (e.g., to 0.01 pL resolution) with a high degree of confidence that any drop volume variations can be accommodated within a desired tolerance. This process can be planned such that multiple nozzles are used to deposit ink into different rows of the target area with each pass. In one embodiment, the printing solution is optimized to produce as few scans as possible, and as fast a print time as possible. Second, in another embodiment, different drive waveforms can be used for each nozzle, again using the specific measured drop volumes. The printing process controls these waveforms such that the specific drop volumes are tallied in specific combinations. Again, using the measured drop volumes per nozzle, this process allows for combinations of specific drop volume averages to achieve very specific fill volumes per target area (e.g., to 0.01 pL resolution). The process can be designed such that multiple nozzles are used to deposit ink in different rows of the target area on each pass. In both of these embodiments, a single row of nozzles can be used, or multiple rows of nozzles arranged as one or more printheads of a printhead assembly can be used. For example, in one contemplated implementation, 30 printheads can be used, each printhead having one row of nozzles, with each row having 256 nozzles. The printheads can be further organized into various different groups, for example, these printheads can be organized into a printhead assembly having five printheads mechanically mounted together, and these resulting six groups can be separately mounted into the printing system. In yet another embodiment, a collective printhead assembly is used with multiple rows of nozzles that can be further positionally offset from one another.This embodiment is similar to the first embodiment above in that different droplet volumes can be combined using a variable effective position offset or geometric step. Again, using the measured droplet volume per nozzle, this process allows for the combination of specific droplet volume averages to achieve a very specific fill volume per target area (e.g., to 0.05 pL, or even 0.01 pL resolution). This does not necessarily imply that the measurement is free of statistical uncertainties such as measurement errors. In one embodiment, such errors are small and are included in the target area fill planning. For example, if the droplet volume measurement error is ±a%, then the fill volume variation across the target area can be planned within a tolerance range of the target fill ±(b-an1 / 2)%, where ±(b)% represents the specification tolerance range and ±(n1 / 2) represents the square root of the average number of droplets per target area or well. Perhaps stated differently, a tolerance range less than the specification can be planned for such that when expected measurement errors are included, the resulting total fill for the target area can be expected to fall within the tolerance range of the specification, e.g., as described above in connection with Figures 8A-8B. Of course, the techniques described herein can optionally be combined with other statistical processes.

[0151] Droplet deposition can optionally be planned such that multiple nozzles are used to deposit ink in different rows of the target area on each pass, and the printing solution is optionally optimized to produce the fewest possible scans and the fastest possible print times. As mentioned above, any combination of these techniques with each other and / or with other techniques can also be employed. For example, in one specifically considered scenario, a variable geometric step is used with drive waveform variation per nozzle and a quantity measurement per drive waveform per nozzle to achieve a very specific quantity combination planned for the target area. For example, in another specifically considered scenario, a fixed geometric step is used with drive waveform variation per nozzle and a quantity measurement per drive waveform per nozzle to achieve a very specific quantity combination planned for the target area.

[0152] By maximizing the number of nozzles that can be used simultaneously during each scan, and by planning the drop volume combinations to ensure they meet specifications, these embodiments promise high quality displays. Also, by reducing print times, these embodiments foster very low per unit printing costs, thus helping to reduce price points for the end consumer.

[0153] FIG. 15D shows a flow diagram for nozzle qualification. In one embodiment, droplet measurements are made to produce statistical models (e.g., distributions and averages) for either and / or each of droplet volume, velocity, and trajectory for each nozzle, and for each waveform applied to any given nozzle. Thus, for example, if there are two choices of waveform for each of 12 nozzles, there are up to 24 waveform-nozzle combinations or pairings. In one embodiment, measurements of each parameter (e.g., volume) are made for each nozzle or waveform-nozzle pairing enough to generate a robust statistical model. Note that despite planning, it is conceptually possible that a given nozzle or nozzle-waveform pairing may produce an exceptionally wide distribution, or average that is sufficiently unusual that it should be treated specially. Such special treatment applied is conceptually represented in one embodiment by FIG. 15D.

[0154] More specifically, a general method is represented using reference numeral 1581. Data generated by the droplet measurement device is stored in memory 1585 for later use. During application of method 1581, this data is retrieved from memory and data for each nozzle or nozzle-waveform pairing is extracted and processed individually (1583). In one embodiment, a normal random distribution is constructed for each variable to be qualified, as represented by the mean, standard deviation, and the number of droplets measured (n), or using an equivalent measure, as described. Again, note that other distribution forms (e.g., Student's T, Poisson, etc.) can be used. The measured parameters are compared to one or more ranges (1587) to determine whether the associated droplets can be used in practice. In one embodiment, at least one range is applied to disqualify the droplets from use (e.g., if the droplets have a sufficiently high or low volume relative to the desired target, the nozzle or nozzle-waveform pairing can be excluded from short-term use). To provide an example, if a 10.00 pL droplet is desired, then, for example, nozzles or nozzle waveforms associated with droplet averages that are more than 1.5% away from this target (e.g., <9.85 pL or >10.15 pL) can be excluded from use. Also, or instead, range, standard deviation, variance, or another measure of spread can be used. For example, if it is desired to have a statistical model of droplets with a narrow distribution (e.g., 3σ<1.005% of the mean), then droplets with measurements that do not meet this criterion can be excluded. It is also possible to use an elaborate / complex set of criteria that considers multiple factors. For example, an anomalous mean combined with a very narrow spread can be accepted, e.g., the associated droplets can be used if the spread (e.g., 3σ) away from the measured (e.g., anomalous) mean μ is within 1.005%. For example, if it is desired to use droplets with a 3σ volume within 10.00 pL ± 0.1 pL, a nozzle-waveform pairing that produces a 9.96 pL average with a 3σ value of ± 0.8 pL may be excluded, while a nozzle-waveform pairing that produces a 9.93 pL average with a 3σ value of ± 0.3 pL may be acceptable.Clearly, many possibilities are possible, according to any desired rejection / anomaly criteria (1589). Note that the same type of processing can be applied to the flight angle and velocity per droplet, i.e., flight angle and velocity per nozzle-waveform pairing will exhibit a statistical distribution, and it is expected that some droplets can be excluded depending on the measurements and statistical models derived from the droplet measurement device. For example, droplets with mean velocity or flight trajectory that is outside of 5% of normal, or variance in velocity outside a particular target, can be hypothetically excluded from use. Different ranges and / or evaluation criteria may be applied to each droplet parameter measured and provided by storage device 1585.

[0155] It should be noted that depending on the rejection / abnormal criteria 1589, the droplets (and nozzle-waveform combinations) can be processed and / or treated differently. For example, certain droplets that do not meet the desired criteria can be rejected (1591) as described. Alternatively, additional measurements can be selectively made for the next measurement iteration of a particular nozzle-waveform pairing, and as an example, additional measurements can be made specifically for a particular nozzle-waveform pairing to improve the tightness of the statistical distribution through additional measurements if the statistical distribution is too broad (e.g., the variance and standard deviation depend on the number of measured data points). It is also possible to adjust the nozzle drive waveforms, per numeral 1593, for example, to use higher or lower voltage levels (e.g., to provide a higher or lower speed, or a more consistent flight angle), or to shape the waveform to generate an adjusted nozzle-waveform pairing that meets a particular criterion. The timing of the waveforms can also be adjusted, per numeral 1594, for example, to compensate for an abnormal average speed associated with a particular nozzle-waveform pairing. As an example (previously suggested), slower droplets can be fired at earlier times relative to other nozzles, and faster droplets can be fired at later times to compensate for the faster flight times. Many such alternatives are possible. Finally, any adjusted parameters (e.g., firing times, waveform voltage levels or shapes) can be stored, per numeral 1595, and optionally, if desired, the adjusted parameters can be applied to remeasure one or more associated droplets. After each nozzle-waveform pairing (modified or otherwise) has been qualified (passed or rejected), the method then proceeds to the next nozzle-waveform pairing, per numeral 1597.

[0156] As should be appreciated, the nozzle drive structure described provides the versatility of printing droplets of different sizes. The use of precise fill volume, drop volume, drop velocity, and drop trajectory per target area allows the use of advanced techniques to vary fill volume and plan nozzle / waveform and / or droplet usage according to defined criteria (within specifications). This provides further quality improvements over conventional methods.

[0157] Figures 16-18B will now be used to provide further details as predicted for two considered drop measurement devices (or systems), namely shadowgraphy and interferometry, respectively. Figures 16-17 will be used to illustrate one embodiment of a printer having a drop measurement system, while Figures 18A and 18B will be used to discuss shadowgraphy and interferometry, respectively.

[0158] As previously mentioned, the present teachings disclose various embodiments of an industrial inkjet thin film printing system that includes a drop measurement device integrated into the printing system. Various embodiments of the inkjet thin film printing system of the present teachings can utilize imaging techniques such as shadowgraphy, or non-imaging techniques such as phase Doppler analysis (PDA) (a technique based on interferometry), which can provide significant advantages for high speed measurement of multiple nozzles of an inkjet printhead, and various embodiments of a printhead assembly used in a thin film inkjet printing system according to the present teachings can have multiple printheads. Such high speed measurements can be performed in situ at any time during the printing process and can provide data that can include volume, velocity, and trajectory for each drop from each nozzle of each printhead. Collective data obtained from the drop measurement device integrated into the inkjet thin film printing system can be utilized to provide uniformity of the ink volume delivered to each of the millions of pixels on an OLED panel display.

[0159] When depositing films in the manufacture of OLED panels, the deposited film material Since thickness often affects panel performance and good display uniformity is an important attribute of a good OLED panel, it is often desirable to deposit a film material with a uniform thickness across the panel. When using an inkjet printing method to deposit a film, droplets of ink are ejected from a printing device onto the panel substrate, and the thickness of the deposited film in each area of ​​the panel is typically related to the amount of ink dispensed across that area of ​​the panel, which is further related to the amount and placement of the droplets on the panel surface. Thus, it is often desirable to dispense the amount of ink uniformly, both in terms of the amount and location of the dispensed droplets across the OLED panel display.

[0160] As mentioned above, an inkjet printing system can have at least one printhead, typically with multiple inkjet nozzles, each nozzle capable of dispensing droplets of ink onto the panel surface. Typically, there is variation across the multiple nozzles of the printhead with respect to the volume, trajectory, and velocity of the dispensed droplets. Such variation can result from a variety of origins, including, but not limited to, variations in nozzle operating conditions, variations in inherent nozzle actuator behavior, including age of piezoelectric nozzle drivers, variations in the ink, and variations in inherent nozzle size and shape. The effects of such variations can result in non-uniformity of volume loading across the panel. For example, variations in droplet volume can directly lead to variations in the deposited volume, while variations in droplet velocity and trajectory can indirectly lead to variations in the deposited volume of ink by causing variations in the placement of the droplets on the OLED panel surface. In theory, these variations can be avoided by using only a single nozzle when printing, but printing with a single nozzle is too slow to be practical in real-world manufacturing applications. In light of such variations in ink droplets dispensed from different nozzles, and the practical need to use multiple nozzles to obtain reasonable processing speeds when using inkjet printing for manufacturing applications, it is desirable to have a method and associated apparatus that provides uniform dispensing of an ink amount across an OLED panel area, despite such nozzle-to-nozzle droplet variations.

[0161] A measurement device integrated into a thin film inkjet printing system according to the present teachings can be used to provide actual measurements of volume, velocity, and trajectory for each nozzle of an inkjet printhead at any time during or intermittently during the execution of a printing process. Such measurements can provide for mitigation of nozzle-to-nozzle drop variation to achieve a more uniform deposition of film material using inkjet methods. In some embodiments, such measurements can be used to adjust printhead performance by adapting drive waveforms to each individual nozzle to directly reduce nozzle-to-nozzle drop variation. In some embodiments, such measurements can be used as input to a print pattern optimization system that can reduce nozzle-to-nozzle drop variation by adjusting nozzle selection for drop deposition to average out nozzle-to-nozzle drop variation in the deposited film. Various embodiments of a measurement device integrated into a thin film inkjet printing system of the present teachings can utilize various imaging techniques, such as shadowgraphy, or non-imaging techniques, such as PDA. PDA can provide a significant advantage of rapidly analyzing multiple nozzles of an inkjet printhead, which is particularly useful for systems with many nozzles and / or printheads.

[0162] In this regard, an inkjet thin film printing system according to various embodiments of the present teachings can be comprised of several devices and apparatuses that enable reliable placement of ink droplets at specific locations on a substrate. These devices and apparatuses can include, by way of non-limiting examples, a printhead assembly, an ink delivery system, a motion system, a substrate support apparatus such as a floating table or chuck, a substrate loading and unloading system, a printhead maintenance system, and a printhead measurement apparatus. In addition, the inkjet thin film printing system can be mounted on a stable support assembly that can include, for example, a granite or metal base. The printhead assembly can be comprised of at least one inkjet printhead with at least one orifice capable of ejecting droplets of ink at a controlled rate, such ejected droplets being further characterized by their volume, velocity, and trajectory.

[0163] Because printing requires relative motion between the printhead assembly and the substrate, the printing system can include a motion system such as a gantry or split-axis XYZ system. Either the printhead assembly can move over a stationary substrate (gantry-style) or both the printhead and substrate can move, for example, in a split-axis configuration. In another embodiment, the printing station can be fixed and the substrate can move in the X and Y axes relative to the printhead, with Z-axis motion being provided by either the substrate or the printhead. As the printhead moves relative to the substrate, droplets of ink are ejected at the correct time to be deposited at the desired location on the substrate. Substrates are inserted and removed from the printer using a substrate loading and unloading system. Depending on the printer configuration, this can be accomplished using a mechanical conveyor, a substrate floating table, or a robot with an end effector. The printhead measurement and maintenance system can consist of several subsystems that enable measurements such as verifying drop volume, drop volume, velocity, and trajectory measurements, as well as printhead maintenance procedures such as wiping the inkjet nozzle surface and preparing it for ejection into a waste bin. Considering the various components that may comprise an inkjet thin film printing system, various embodiments of the inkjet thin film printing system according to various embodiments of the present teachings can have different footprints and form factors.

[0164] As a non-limiting example, Figure 16 depicts an inkjet thin film printing system according to various embodiments that can be used to print substrates such as, but not limited to, OLED panels. In Figure 16, the inkjet thin film printing system 1600 utilizes a split-axis motion system. The inkjet thin film printing system 1600 can be mounted on a printing system support assembly 1610, which can include a pan 1612 carried by a support frame 1614. A base 1616 is mounted above the pan, which can optionally be constructed from granite or metal. The inkjet thin film printing system can include a motion system 1620, such as a split-axis motion system as shown.

[0165] The motion system 1620 can be seen to include a bridge 1622 that supports an X-axis carriage 1624, which in turn mounts a Z-axis mounting plate 1626. The Z-axis mounting plate in turn supports a printhead mounting and clamping assembly 1628, which is used to mount a replaceable printhead assembly 1640. For the split-axis motion system 1620, a Y-axis track 1623 can be mounted on the base 1616 to provide support for a Y-axis carriage 1625, which in turn carries a substrate support assembly 1630, and these various components provide Y-axis movement of a substrate mounted on the substrate support assembly 1630. As shown in FIG. 16, for various embodiments of the thin film printing system, the substrate support assembly 1630 can be a chuck. The substrate support assembly can be provided by a floating table, for example, as described in detail in U.S. Pat. No. 8,383,202, which is incorporated herein by reference. Inkjet thin film printing system 1600 can utilize a system assembly that supports one or more modular inkjet printhead assemblies, such as the various printhead assemblies shown mounted in tool carousel 1645. By providing selective interchange of the various printhead assemblies, an end user can have the versatility to efficiently sequentially print different inks of different formulations on a substrate during a printing process, such as during printing of an OLED panel substrate. It should be noted that this is not required for all embodiments, i.e., other embodiments may feature a single printhead assembly that is not swapped between different printing processes. For example, one contemplated embodiment features an assembly line of multiple printers, each performing a respective printing process (e.g., using a respective ink). The techniques described herein can be applied to each such printer.

[0166] The printhead assembly can include a fluid system having, for example, an ink reservoir in fluid communication with at least one inkjet printhead for delivering OLED film-forming materials onto a substrate. In that regard, as shown in FIG. 16, the printhead assembly 1640 can include at least one printhead 1642. In various embodiments, the printhead assembly can optionally include fluid and electronic connections to each printhead. Each printhead can have multiple nozzles and orifices capable of, in turn, ejecting ink droplets at a controlled rate with measurable droplet volumes, velocities, and trajectories. Various embodiments of the printhead assembly 1640 can have from about 1 to about 30 printheads per printhead assembly. The printhead 1642 can have from about 16 to about 2048 nozzles, each capable of ejecting droplet volumes of from about 0.10 pL to about 200.00 pL.

[0167] Measuring the performance of each nozzle of a given printhead can include checking nozzle firing, as well as measuring drop volume, velocity, and trajectory. As previously mentioned, having such measurement data can provide either adjusting the head before printing to provide more uniform performance for each nozzle, or using the measurement data to provide printing algorithms that can compensate for differences during printing, or a combination of such approaches. Having a clear, reliable, and up-to-date data set of measurement data can provide various approaches that can use the measurement data to compensate for nozzle-to-nozzle drop volume variations, and can enable planned printing processes that combine different volume drops (from the same nozzle or from each nozzle using different drive waveforms). As previously mentioned, measurement data is advantageously collected to generate a population of measurements that represent the distribution of each nozzle, so that expectations of average drop volume, trajectory, and velocity can be generated, and the expected variations of each such drop parameter can be fully understood and used in the printing plan.

[0168] In that regard, the depicted inkjet thin film printing system can include a drop measurement device or system 1650, which can be mounted on a support 1655. It is contemplated that various embodiments of the drop measurement system 1650 can be based on imaging or non-imaging techniques as described, such as shadowgraphy or interferometry-based methods. An embodiment utilizing a non-imaging PDA technique can provide a significant advantage of rapidly analyzing about 16 to about 2048 nozzles of each printhead, such as printhead 1642 (e.g., about 50 times faster than typical imaging techniques). Recalling that a printhead assembly can include, for example, 30 printheads (i.e., with printing systems using more than 10,000 nozzles), this allows for rapid in-printer, in-situ dynamic measurement of all nozzles (and all alternative driving waveforms, if relevant to the embodiment), with drop recalibration every 2 to 24 hours, or more frequently. Also, various embodiments of the systems and methods according to the present teachings can utilize a PDA measurement device integrated into a gas enclosure assembly and system that can house the printing apparatus. Such systems and methods utilizing a PDA measurement device integrated into a gas enclosure assembly and system housing the printing apparatus can provide high speed in-situ measurement of multiple nozzles in a printhead, which is particularly useful for ensuring uniform deposition rate and reducing any mura effects across a large substrate, for example, having one or more OLED devices.

[0169] Numeral 1617 is used to designate the area of ​​the inkjet printing device that is associated with drop measurement system 1650. This area is shown in enlarged detail in FIG.

[0170] As shown in FIG. 17, the printhead assembly 1740 can be held during printing by a printhead mounting and clamping assembly 1728, itself again carried by a Z-axis mount 1726 of the motion system. In this regard, the motion system is used to position the printhead assembly 1740 for measurement in proximity to a drop measurement system 1750, for example, in a service area or service station. As previously mentioned, the drop measurement system 1750 can be designed for selective engagement and disengagement while the printhead assembly 1740 is in this position. With large printhead assemblies (e.g., having thousands of nozzles), such a configuration allows the drop measurement system to perform testing while the printhead assembly 1740 is "parked," with other testing being performed simultaneously by other testing or calibration equipment or processes (not shown). For example, the printhead nozzles can be purged, cleaned, or otherwise managed through the use of concurrent processes applied to help minimize any downtime of the overall inkjet printing system, which helps maximize manufacturing productivity. As mentioned above (and explained below with respect to FIG. 19 ), drop measurements (and other inspections) can be performed while the substrate is being transported, dried, cured, loaded, or unloaded, further minimizing any system downtime by stacking the drop measurements against other unavoidable tasks associated with the printing / manufacturing operation. Each nozzle of the printhead 1742 of the printhead assembly 1740 can be adapted with a measurement area 1756 for measurement of the drops emitted from each nozzle using the drop measurement system 1750. In this embodiment, individual printheads 1742 can be moved relative to the other printheads for analysis, but again, note that this is not required for all embodiments. For example, each printhead can also be statically mounted during measurement, with the drop measurement system advanced to each printhead location and each nozzle location within a given printhead.As previously mentioned, this allows for the simultaneous processing or "stacking" of multiple inspection operations while the printhead assembly is parked. It is also possible to use multiple drop measurement systems, for example, to independently measure different nozzles of different spatially separated printheads.

[0171] For purposes of illustration, assume that the droplet measurement system is a PDA device (i.e., an interferometry-based device) having a light source, such as a laser source and light transmission optics, a beam splitter, and a transmission lens. In addition, such a PDA device may also have a receiving optics, including a receiving lens and multiple photodetectors. For example, a first optical side 1752 of the droplet measurement system 1750 may provide one or more light beams for measurement and focus the light onto a measurement region 1756, as shown by the diagonal lines, while a second optical side 1754 may pass measurement light that is dispersed from a droplet in the measurement region 1756 to the receiving optics and one or more photodetectors.

[0172] Droplet measurement system 1750 can be interfaced, either directly or remotely, to a computer or computing device (not shown). Such a computing device can be configured to receive signals representing the measured drop volume, velocity, and trajectory of each droplet produced by a nozzle (or nozzle-waveform combination) from each printhead 1742 of printhead assembly 1740. Again, multiple measurements of many droplets from each nozzle / nozzle-waveform pairing are advantageously made to generate a statistical population representing a variety of reproducible droplets.

[0173] As discussed above in connection with Figures 11 and 12, various embodiments of the printing system can be housed in a gas enclosure that provides an inert, low-particle environment, and the drop measurements preferably occur in such an environment. In one embodiment, the drop measurements are performed in the common atmosphere used for printing, e.g., in the same common (enclosed) chamber. In a second embodiment, a separate, fluidly isolated chamber is used for the measurements, e.g., as part of a service station area.

[0174] FIG. 18A shows the layout of a droplet measurement system 1801 specifically configured to use the shadowgraphy technique. Specifically, print head 1803 is seen in a position where it will eject droplet 1805 into a drainage basin (not shown in FIG. 18A). During the flight of droplet 1805, the droplet traverses a measurement area where the droplet is illuminated by a light source, seen in FIG. 18A to consist of, for example, a strobe light 1807 and optional light source optics 1809 employed to direct light from strobe light 1807 to the measurement area (e.g., from below the measurement surface as previously illustrated in connection with FIGS. 2A-E or FIG. 16-17). The optics direct the light to illuminate a relatively wide area represented by focusing or redirecting paths 1811 to repeatedly expose the droplet in rapid succession at different positions for capture in a single image frame. Thus, FIG. 18A shows three different positions of the same droplet, representing different flashes of the strobe that are collectively imaged together. Thus, for example, the image frame being analyzed will show what appears to be multiple droplets at different locations (i.e., multiple instances of number 1805), but these are in fact the same droplet at different locations along the flight trajectory. The second set of optics 1813 provides light collection and focusing such that the captured image clearly depicts both the droplet's outline and a variable amount of shadow representing the droplet diameter that is used by the image processing software to calculate the droplet volume. As should be appreciated, by imaging the same droplet during its flight at multiple locations, the droplet measurement system can use one image to calculate the droplet volume, velocity, and trajectory; the shadow parameters are used to calculate the droplet mass, and therefore the volume, and the relative position of the droplet is used to calculate both the velocity and trajectory. For example, droplets that increase in apparent diameter at the "down position" in the captured image frame are moving towards the receiving optics 1813, and conversely, those that decrease in diameter are moving away. The receiving optics 1813 in turn transmit the captured light to a camera 1815 , for example a high resolution CCD camera, which images the droplet outline and shadow as depicted by graphic 1817 .The droplet measurement system optionally provides control over the zoom / focus (1819) and / or XY position (1821) of the receiving optics, all under the control of a controlling computer system 1823 (and instructions stored on a non-transitory mechanical storage medium used by one or more processors of the computer system for such control). In one embodiment, as described, the receiving optics and the light source are mounted and transported together on a common chassis, providing a fixed focus path, although this need not be the case for each embodiment. While the depicted system captures the movement of each droplet in a few microseconds, image processing application software 1825 executed by the computer system 1823 then calculates the droplet parameters. By way of example, the computer can provide a display and visualization (1827) of the droplets and / or the measured parameters, and can calculate values ​​for various parameters such as volume, velocity, and trajectory (1829, 1831, and 1833), or other parameters. It should be noted that computer system 1823 may be part of the inkjet printing system or may be remote (e.g., connected by a local area network "LAN" or wide area network "WAN", e.g., the Internet, to collect data remotely) and similarly, display and visualization 1827 may also be provided at a location remote from computer system 1823, similarly via a LAN or WAN. The computer system 1823 compiles the measured parameters to form a statistical ensemble of measurements for a given nozzle that produced a drop (and for a given nozzle-waveform pairing, if alternative drive waveforms are used by a particular embodiment of the printing system). The computer system 1823 optionally stores the individual measurements themselves and / or a statistical summary (e.g., mean and standard deviation or variance in the case of a normal distribution, and equivalent metrics if other distribution types are supported) in a database 1837. Once a sufficiently robust ensemble has been measured, the database can then be applied in planning and / or optimizing the printing process as described above, for example using a particular combination of droplet averages to obtain a composite fill per target area, where the composite fill can be based on different drop volumes (e.g., from different nozzles and / or drive waveforms).

[0175] FIG. 18B shows the layout of a drop measurement system 1851 specifically configured to use PDA (interferometry) techniques. A print head is shown in position for measurement, as referenced by numeral 1853. The print head will eject a drop from a particular nozzle (e.g., using a particular drive waveform) downward into the drop measurement region, as indicated by numeral 1855. As with the previous embodiment, the drop measurement system can optionally be designed for three-dimensional transport relative to the parked print head, such that the drop measurement region is effectively "carried" into the drop flight of the particular nozzle. A light source, in this case a laser 1857, generates a light beam 1859, which is directed to be incident on a beam splitter 1861. The beam splitter generates two or more light beams 1863 and 1864 (only two are shown in FIG. 18B), which optics 1865 then redirects in a converging manner, i.e., so that the light beams intersect at a location coincident with the droplet in flight, as represented by numerals 1866 and 1867. Note that optics 1865 optionally provides a laser 1857 mounted below the measurement surface (see discussion of FIGS. 2D and 2E above), and optionally redirects light paths 1859 or 1863 / 1864 to reach the measurement region (e.g., by redirecting one or more of the light paths around the drainage basin). Note that numeral 1869 is used to represent the general continuous dimensions of the illumination optics (e.g., light paths 1866 and 1867, etc.). As previously described, using interference-based techniques, the diffraction pattern is captured from a direction that is angularly offset to this continuous dimension 1869, as represented by angle 1873. This angular deviation is typically 90 degrees, although other directions of capture can be used. Thus, measurement light 1871 is received from the incident light at this angular deviation by a second set of optics 1875 (labeled "Optics 2") and redirected by a non-imaging detector 1877 for sub-deposition surface measurements.These detectors generate data representing a diffraction pattern, as illustrated by graphic 1879, and as should be understood (e.g., by contrasting this graphic 1879 with graphic 1817 from FIG. 18A), the spacing of the lines in the diffraction pattern provides a measure of the droplet volume, which is processed much more rapidly than the imaging method represented by FIG. 18A to measure droplet volume. Note that while FIG. 18B illustrates the use of one light source 1857 and two incident beams 1866 and 1867, other embodiments use more than one light source and more than two incident beams, for example to capture droplet velocity, trajectory, and other parameters. As with the embodiment of FIG. 18A, in FIG. 18B a computer (1881) optionally provides zoom / focus (1883) and XY transport of the measurement optics, runs suitable application software (1887) to calculate various droplet parameters, and provides display and visualization (1889). Just as before, these various elements can be integrated with the printer or manufacturing device, or can be distributed across a WAN or LAN controlled by multiple separate processors in respective computers or servers. As before, the measured parameters can include drop volume (1891), velocity (1893), and trajectory (1895), with data representing statistical populations (1897) stored in a database (1899) for scan planning purposes. The scan plan can again combine droplet parameters from different nozzles and / or waveforms to provide precision filling of a target area purposefully based on multiple different droplet volumes.

[0176] It was previously described that the drop parameters may change over time, for example, according to system parameters, ambient conditions or ink characteristics. Thus, industrial printing systems advantageously update drop measurements of not only single drops, but of the statistical population of each drop (as well as the expected average volume / velocity and trajectory of each drop) relatively frequently, which helps ensure precise drop data that is always accurate and up to date, allowing planned drop combinations that reliably match the maximum tolerances of the composite ink fill. Drop parameters have been found in practice to change somewhat slowly, for example, with detectable fluctuations every 2-12 hours. It should be noted that the use of in situ drop measurements allows for iterative dynamic measurement and construction of new statistical populations of the measured parameters within this time range, which with this conventional method would take many hours to measure a large printhead or printhead assembly, and through the use of high-speed techniques such as PDAs as discussed above, it becomes possible to update all statistical measurements very rapidly, for example, with lead times of 30 minutes or less, even when thousands of print nozzles are involved. Thus, a system utilizing some or all of the techniques discussed above would facilitate and enable industrial printers with recalibrated drop measurement parameters based on statistical distribution within the described 2-12 hour time frame, thus facilitating more accurate printing within maximum tolerance of target area fill variation.

[0177] As mentioned above, in one contemplated embodiment, the printer is controlled, intermittently or continuously, to perform drop parameter measurements whenever the printer is not actively printing. This helps maximize the operational uptime of the manufacturing line. As described, in one embodiment, the printer's printhead assembly can be diverted for drop parameter measurements whenever it is not in use. For example, whenever a substrate is loaded or unloaded, advanced between chambers, or dried, cured, or otherwise processed, the print carriage can transport the printhead assembly to a service station for drop measurement and / or other service operations. Such operations, as described, further help provide frequent dynamic updates of the drop statistical population for each nozzle, and optionally, when employed in conjunction with a PDA-based drop measurement device (e.g., an interferometry-based technique), such a control scheme can make the drop measurement task transparent to any desired printing operations. It is noted that in the contemplated system, this control is implemented by control software running on at least one processor that manages the printing process. Additionally, it should be noted that this software may reside on the printer, on one or more computers or servers, or both.

[0178] 19 shows one embodiment of a flow for such a control process 1901. As described, this process may optionally be implemented by instructions stored on a non-transitory machine storage medium that, when executed, cause at least one processor to perform / provide the recited steps.

[0179] FIG. 19 is divided into three general regions 1903, 1905, and 1907, which respectively represent start-up and offline initialization processes, online printing, and offline special operations. When the system is powered on, it typically undergoes an initialization process 1909, in which new measurements are made for each nozzle to generate a statistical population, as described above. At the same time, an iterative process, as previously described, is performed to select 16 waveforms that will produce drop volumes within ±10.0% of the target drop volume. A calibration process (not shown) may also be invoked to select a number of nozzle firing waveforms for each nozzle (using a calibration parameter). Thus, a statistical ensemble is generated for each such waveform and / or nozzl...

Claims

1. A printing device, comprising: a print head having a nozzle; a substrate floating table for placing a substrate on a deposition surface at a printing position for printing; a print head transport mechanism for transporting the print head in an X direction and a Z direction; a droplet measurement system disposed adjacent to the substrate floating table, the droplet measurement system comprising: a light source; a plurality of non-imaging photodetectors; an interference optic; a collector disposed between the light source and at least one of the photodetectors; and an optical path extending from the light source to the photodetectors, wherein the light source, the photodetectors, the collector, the interference optic, and the optical path are mounted in a fixed positional relationship with respect to one another, the light source, the photodetectors, the interference optic, and the collector are mounted below the deposition surface, and the optical path is configured such that a portion of the optical path extends above the deposition surface; a motion system for moving the drop measurement system relative to the printhead in three independent directions; a controller configured to control the printhead, the printhead transport mechanism, the drop measurement system, and the motion system to measure the drops from the plurality of nozzles by applying a plurality of waveforms to at least one of the nozzles, and to exclude a portion of the measurements based on at least one criterion; A printing device comprising:

2. 2. The printing apparatus of claim 1, wherein the light path is configured to pass through an illumination area adjacent to one nozzle of the print head.

3. The printing apparatus of claim 2 , wherein the three independent directions include a direction extending between an intersection of the light and the droplet and one of the nozzles.

4. The printing apparatus of claim 1 , wherein the drop measurement system outputs a signal representative of one of a trajectory or a velocity of a drop ejected from the printhead.

5. the nozzle defines a first plane; The droplet measurement system further comprises a mirror; the motion system moves the segment of the optical path along a second plane, the second plane being parallel to the first plane; The printing apparatus of claim 1 , wherein the light source and the light detector are disposed outside the second plane.

6. The printing apparatus of claim 5 , wherein the second plane is above the deposition surface.

7. The printing apparatus of claim 6 , further comprising at least one processor that calculates values ​​of the droplet characteristics based on an electronic output of the drop measurement system.

8. the nozzles include at least 1000 nozzles; 8. The printing apparatus of claim 7, wherein the motion system moves the segment while the print head is in the second position to cause the segment to intersect with a flight trajectory of a droplet of liquid ejected from any one of the at least 1000 nozzles.

9. The printing apparatus of claim 8 , wherein the three independent directions include directions between one of the nozzles and an intersection of the optical path and the flight path.

10. the nozzles include at least 1000 nozzles; the motion system positions the optical path above the deposition surface to intersect a flight path of a droplet ejected from any of the at least 1000 nozzles; The printing apparatus of claim 1 , further comprising a non-transitory memory device that stores measurements of the drop measurement system.

11. The printing apparatus of claim 9 , wherein the interference optics includes an optical fiber.

12. 10. The printing apparatus of claim 1, further comprising a gas enclosure for maintaining a controlled atmosphere, the print head, the substrate floating table, and the drop measurement system being within the gas enclosure.

13. 13. The printing apparatus of claim 1, further comprising an electronic control system that automatically controls the print head transport mechanism, the drop measurement system, and the motion system to automatically measure characteristics of droplets ejected from at least one of the nozzles during processing of successive substrates.

14. The printing apparatus of claim 1 , wherein the print head transport mechanism includes a bridge that extends over the substrate floating table.

15. The printing apparatus of claim 14 , wherein the drop measurement system is disposed adjacent an end of the bridge.

16. An apparatus, comprising: a print head having a nozzle; a substrate floating table for placing a substrate on a deposition surface at a printing position for printing; a print head transport mechanism for transporting the print head in an X-direction and a Z-direction between a first position and a second position adjacent the printing position; a droplet measurement system disposed adjacent to the substrate floating table, the droplet measurement system comprising a light source, a plurality of non-imaging photodetectors, an interference optics system, a collector disposed between the light source and at least one photodetector, and an optical path extending from the light source to the photodetector, wherein the light source, the photodetector, the collector, the interference optics, and the optical path are mounted in a fixed positional relationship with respect to one another, the light source, the photodetector, the interference optics, and the collector are mounted below the deposition surface, and the optical path is configured such that a portion of the optical path extends above the deposition surface; a motion system for moving the drop measurement system in three independent directions relative to the printhead to position a portion of the optical path at a measurement area adjacent one of the nozzles of the printhead; a controller configured to control the printhead, the printhead transport mechanism, the drop measurement system, and the motion system to measure the drops from the plurality of nozzles by applying a plurality of waveforms to at least one of the nozzles, form a statistical representation of the measurements, and exclude a portion of the measurements based on at least one criterion; The droplet measurement system is a device that detects interference patterns and images droplets.

17. The apparatus of claim 16 , further comprising a processor that determines droplet characteristics from an electronic output of the droplet measurement system.

18. The apparatus of claim 17 , wherein the characteristic is a volume, a trajectory, or a velocity of a droplet.

19. 20. The apparatus of claim 18, wherein the three independent directions include directions extending from one of the intersections of the light path and the droplet to one of the nozzles.

20. 17. The apparatus of claim 16, wherein the nozzle defines a first plane, the motion system moves a segment of the light path that interacts with the droplets in flight in a second plane above the deposition plane, and each of the light source and the detector are positioned outside the second plane.

Images