Multiplexed power generator output with channel offsets for pulsed driving of multiple loads

The power supply assembly with a multiplexing unit efficiently distributes pulses to multiple loads in PECVD processes, addressing inefficiencies in conventional systems by avoiding overlap and maintaining desired duty cycles, thereby reducing power supply size and cost.

KR103047303B1Active Publication Date: 2026-11-11AES GLOBAL HLDG PTE LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
KR1020257037852
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Priority Date
2019-07-29
Filing Date
2020-07-29
Publication Date
2026-11-11
Estimated Expiration
2040-07-29

AI Technical Summary

Technical Problem

Conventional Plasma Enhanced Chemical Vapor Deposition (PECVD) processes require separate power supplies for each chamber, leading to inefficient power usage as each supply is limited to less than about 10% of the time due to low-duty-cycle pulse signals with high power delivery, and this design does not account for varying pulse frequencies and duty cycles across different loads.

Method used

A power supply assembly with a multiplexing unit that distributes pulses to multiple loads, adjusting pulse timing to avoid overlap and maintain desired duty cycles, allowing a single power supply to efficiently power multiple loads with different recipes by applying phase shifts and scheduling adjustments.

Benefits of technology

Enables efficient power delivery to multiple loads with reduced power supply size and cost, ensuring continuous achievement of desired film characteristics while preventing excessive power draw and damage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure R1020257037852_ABST
    Figure R1020257037852_ABST
Patent Text Reader

Abstract

The present disclosure describes a system, method, and apparatus for a pulse power supply assembly that distributes pulsed power to two or more loads using a single pulse power supply. These loads may include one or more electrodes of a plasma processing chamber. The pulse power supply of the assembly may phase-shift pulses for different loads to ensure that there is no overlap between pulses at the outputs, even if the target frequency and / or duty cycle for the different loads otherwise requires such pulse overlap. Variations applied by the pulse power supply may be limited by an attempt to maintain the average parameter of the pulse trains provided to the different loads within a predetermined variation.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] This patent application claims priority to Provisional Application No. 62 / 880,074 filed on July 29, 2019, with the title of invention "MULTIPLEXED POWER GENERATOR OUTPUT WITH CHANNEL OFFSETS FOR PULSED DRIVING OF MULTIPLE LOADS", which is assigned to the assignee of this application and is thereby expressly incorporated by reference into this specification.

[0002] The present disclosure generally relates to pulse power supplies. In particular, but without limitation, the present disclosure relates to systems, methods, and apparatuses for driving a plurality of loads using a single pulse generator. Background Technology

[0003] Conventional Plasma Enhanced Chemical Vapor Deposition (PECVD) processes use separate power supplies for each of the multiple chambers. Each chamber may contain a separate wafer "boat" and is considered a distinct power load. Each generator typically provides a low-duty-cycle pulse signal with high power delivery during each pulse. Each load can use a different recipe, and each recipe can vary the pulse frequency, duty cycle, and power over time. One disadvantage of this design is that each power supply is typically limited to providing power for less than about 10% of the time. Therefore, there is a need to use these power supplies more efficiently. means of solving the problem

[0004] The following provides a simplified overview of one or more aspects and / or embodiments disclosed herein. As such, the following overview should not be construed as a broad overview of all aspects and / or embodiments considered, nor should it be construed as identifying important or decisive elements of all aspects and / or embodiments considered, or describing the scope associated with any particular aspect and / or embodiment. Accordingly, the following overview is intended solely to present specific concepts related to one or more aspects and / or embodiments of the mechanisms disclosed herein in a simplified form to precede the detailed description set forth below.

[0005] To overcome the problem of supplying power to multiple loads with multiple power supplies, in one aspect of the present disclosure, a power supply assembly is disclosed that provides maximum power to multiple loads through multiplexing, wherein pulses for each of the multiple loads may be phased to avoid overlap between different channels. In one embodiment, a multiplexing unit attached to the output of the power supply may multiplex or distribute pulses from the power supply into a plurality of channels (or distributed pulse trains), each channel being provided to a distinct one of the plurality of loads. In some embodiments, these loads may include one or more electrodes of a plasma processing chamber. Since recipes for different loads may require different frequencies or pulsing, or may change over time to display different frequencies of different loads simultaneously, beat frequency and channel overlap may occur. Channel overlap may lead to excessive power draw and poor power delivery or damage to the power supply. To avoid this, a pulse power supply or multiplexing unit can maintain the desired average duty cycle for each channel while adjusting or pacing the timing of the pulses when channel overlap occurs. In this way, a single power supply can provide power to multiple pulse loads (e.g., to list just three: PECVD, PVD, CVD) while continuously achieving the desired film characteristics.

[0006] Some embodiments of the present disclosure may be characterized as a pulse power supply assembly. The assembly may include a multiplexing unit and a pulse power supply. The multiplexing unit may include two or more pulse outputs, each configured to be coupled to one of two or more loads. The multiplexing unit may further include a power distribution switching assembly coupled to the input of the multiplexing unit and to the two or more pulse outputs. The power distribution switching assembly may be configured to divide pulses of a pulse train received at the input into pulses distributed from the two or more pulse outputs. The pulse power supply may be configured to provide a pulse train to the input of the multiplexing unit. The pulse power supply may include a periodic pulse generator, a phase shift circuit, a memory, and a processing portion. The phase shift circuit may be connected to the output of the periodic pulse generator and may be connected to the output of the pulse power supply. The memory may store target parameters (e.g., frequency, duty cycle, conduction time, on time, off time, start time, etc.) for each of the two or more loads. The processing section may be connected to memory and configured to read processor-executable instructions in memory that, when executed, cause the processing section to access target parameters in memory and generate a schedule of distributed pulses for the multiplexing unit. These also cause the processing section to predict instances of overlap between the distributed pulses in the schedule. They also cause the processing section to generate a new schedule that avoids overlap and to instruct the phase shift circuit to modify the periodic pulse train from the periodic pulse generator to generate a pulse train such that the pulse train is aligned with the new schedule.

[0007] Another aspect of the present disclosure may be characterized as a pulse power supply assembly configured to be coupled to and drive two or more loads. The assembly may include a multiplexing unit and a pulse power supply. The multiplexing unit may include two or more pulse outputs, each configured to be coupled to one of two or more loads. It may also include a power distribution switching assembly coupled to the input of the multiplexing unit and to two or more pulse outputs, configured to divide pulses from a pulse train received at the input into pulses distributed from two or more pulse outputs. The pulse power supply may be configured to provide a pulse train to the input of the multiplexing unit. The pulse power supply may include means for generating a pulse train and a memory configured to store target parameters for two or more loads. The pulse power supply may also include a processing portion, which may be configured to access the target parameters and predict the schedule of distributed pulses for two or more loads. The processing portion may also be configured to identify one or more overlaps between pulses in different pulses among the distributed pulses. The processing part may also be configured to command a means for generating a pulse train to apply a phase shift to one or more pulses of the pulse train so that no overlap occurs when the power distribution switching assembly divides the pulse train into distributed pulses.

[0008] Other aspects of the present disclosure may be characterized as a non-transient type of computer-readable storage medium encoded with processor-readable instructions for performing a method to drive two or more distinct loads using a single pulse power supply and to avoid overlapping pulses for two or more distinct loads. The method may include the steps of accessing a target parameter for each of two or more distinct loads and predicting instances of overlap between distributed pulses configured to be provided to two or more distinct loads when the target parameter is followed without variance. The distributed pulses may be formed in a single pulse train. The method may further include the steps of scheduling a first variation in the schedule of distributed pulses to avoid instances of overlap and instructing a power distribution switching assembly to distribute pulses of the pulse train to two or more distinct loads according to the schedule.

[0009] Some embodiments of the present disclosure may be characterized as pulse power supplies driving two or more distinct loads. The pulse power supply may include two or more pulse outputs, a power distribution switching assembly, a processing section, a memory, and a multiplexing module stored in the memory and executable in the processing section. The memory may be coupled to the processing section and may store a target pulse frequency and a target duty cycle for each of the two or more pulse outputs and for a certain time period (e.g., the target frequency and target duty cycle may change over time). At least two of the two or more pulse outputs may differ in the target pulse frequency or target pulse duty cycle during the time period. The multiplexing module may be executed in the processing section to access the target pulse frequency and target duty cycle in the memory for each of the two or more pulse outputs during the time period. The multiplexing module may be executed in the processing section to access the target pulse frequency and target duty cycle in the memory for each of the two or more pulse outputs during the time period. A multiplexing module may be executed in the processing part to predict instances of overlap between pulses of two or more pulse outputs during that time period. A multiplexing module may be executed in the processing part to command a power distribution switching assembly to schedule the timing of distributed pulses to avoid overlap during that time period.

[0010] Another aspect of the present disclosure may be characterized as a pulse power supply assembly configured to be coupled to and driven by two or more loads. The assembly may include a pulse power supply and a multiplexing unit. The pulse power supply may provide a first pulse train at a first output. The multiplexing unit may include a first input, two or more pulse outputs, a power distribution switching assembly, and a memory. The first input may be connected to the first output. Each of the two or more pulse outputs may be configured to provide a pulse train distributed to a corresponding one of the two or more loads. The power distribution switching assembly may be configured to distribute pulses from the first pulse train to the two or more pulse outputs and to schedule the timing of the distributed pulses. The memory may store a target pulse frequency and a target duty cycle for each of the two or more loads for a certain time period. During that time period, the target pulse frequencies for at least two of the two or more loads are different. The processing part may be connected to memory and may read processor executable instructions from memory, and when the instructions are executed, the processing part may: access the target pulse frequency and target duty cycle in memory for each of two or more pulse outputs; predict instances of overlap between the pulses of two or more pulse outputs; and instruct the power distribution switching assembly to schedule the timing of the distributed pulses to avoid overlap.

[0011] Some embodiments of the present disclosure may be characterized as a multiplexing unit connected to a pulse power supply to drive two or more distinct loads. The pulse power supply may provide a first pulse train. The multiplexing unit may include two or more pulse outputs, a power distribution switching assembly, a processing part, and memory. The processing part may control the power distribution switching assembly. The memory may be connected to the processing part and may store a target pulse frequency and a target duty cycle for each of the two or more pulse outputs for a certain time period. The pulse frequencies of at least two of the two or more pulse outputs may differ during that time period. The multiplexing module may be stored in the memory and may be executable by the processing part to access the target pulse frequency and target duty cycle in the memory for each of the two or more pulse outputs for a certain time period; predict instances of overlap between pulses of the two or more pulse outputs; and instruct the power distribution switching assembly to schedule the timing of pulses reaching at least two pulse outputs to avoid overlap.

[0012] Another aspect of the present disclosure may be characterized as a pulse power supply assembly configured to be coupled to and driven by two or more loads. The assembly may include a pulse power supply having a multiplexing unit and providing a first pulse train to the multiplexing unit. The multiplexing unit may include two or more pulse outputs, a power distribution switching assembly, and a memory. Each of the two or more pulse outputs may be configured to provide a pulse train to a corresponding one of the two or more loads. The power distribution switching assembly may be configured to distribute pulses from the first pulse train to the two or more pulse outputs and to schedule the timing of the distributed pulses. The memory may store a target pulse frequency and a target duty cycle for each of the two or more loads during a time period, wherein the target pulse frequencies for at least two of the two or more loads during the time period are different. The processing part may be connected to memory and may read processor executable instructions from memory, and when the instructions are executed, the processing part may: access the target pulse frequency and target duty cycle in memory for each of two or more pulse outputs; predict instances of overlap between the pulses of two or more pulse outputs; and instruct the power distribution switching assembly to schedule the timing of the distributed pulses to avoid overlap.

[0013] Other aspects of the present disclosure may be characterized as a non-transient type of computer-readable storage medium encoded with processor-readable instructions for performing a method to drive two or more distinct loads using a single pulse power supply and to avoid overlapping pulses for two or more distinct loads. The method may include the step of accessing a target pulse frequency for each of two or more distinct loads. The method may also include the step of accessing a target pulse duty cycle for each of two or more distinct loads. The method may also include the step of predicting instances of overlap between distributed pulses configured for supply to two or more distinct loads when the target pulse frequency and the target pulse duty cycle are followed without variation and the distributed pulses are formed from a single pulse train. The method may also include the step of instructing a power distribution switching assembly to schedule a first variation in the schedule of distributed pulses to avoid overlap. The method may also include the step of monitoring the difference between the target frequency and the average frequency of each distributed pulse, and reducing this difference by applying additional variations to future distributed pulses. Brief explanation of the drawing

[0014] A more complete understanding of the various problems and advantages of the present disclosure is made obvious and more easily recognized by referring to the following detailed description and appended claims, together with the accompanying drawings. FIG. 1 illustrates an embodiment of a power delivery system for supplying pulsed power to two or more loads using a single pulse power supply. FIG. 2 illustrates an alternative embodiment to FIG. 1. FIG. 3 illustrates a detailed embodiment of a pulse power supply. FIG. 4 shows a specific embodiment of a multiplexing unit. FIG. 5 is the same as that shown in FIG. 1 to 4, but shows a timing chart representing exemplary timing that can be caused by a system having N = 5. Figure 6 illustrates another embodiment of a timing chart for an N = 5 system. FIG. 7 illustrates an exemplary scheduling for a five-channel system that results in overlap in all five channels during the second pulse cycle. FIG. 8 illustrates a solution or new schedule for the second cycle of pulses shown in FIG. 7. Figure 9 illustrates a timing chart that helps show how the analysis of the first cycle of the pulse occurs. FIG. 10 illustrates a series of steps for removing overlap between the first channel and the second channel. FIG. 11 illustrates a series of steps for removing overlap between the first channel and the second and third channels. FIG. 12 illustrates a series of steps to eliminate overlap between the second channel and the third channel. FIG. 13 illustrates a method for driving two or more separate loads using a single pulse power supply and avoiding overlapping pulses for two or more separate loads. FIG. 14 illustrates another method for driving two or more individual loads using a single pulse power supply and avoiding overlapping pulses for two or more individual loads. Figure 15 illustrates a timing diagram that helps show how the processing part of the pulse power supply checks the overlap of distributed pulses for an exemplary 5-channel system. FIG. 16 illustrates an embodiment of a power delivery system for supplying pulsed power to two or more loads using a single pulse power supply. FIG. 17 illustrates an alternative embodiment to FIG. 16. FIG. 18 shows a specific embodiment of a multiplexing unit. FIG. 19 illustrates a method for driving two or more separate loads using a single pulse power supply and avoiding overlapping pulses for two or more separate loads. FIG. 20 illustrates a timing chart for a 3-channel system in which a multiplexing unit applies a phase shift to distributed pulses separated from a periodic pulse train from a pulse power supply. FIG. 21 illustrates a timing chart for a 3-channel system in which a pulse power supply changes the phases of its pulses and a multiplexing unit separates these pulses into distributed pulses without applying a phase shift in the multiplexing unit. FIG. 22 illustrates a block diagram showing physical components that can be used to realize the multiplexing units described throughout the present disclosure. Specific details for implementing the invention

[0015] The present disclosure generally relates to pulse power supplies. In particular, but without limitation, the present disclosure relates to systems, methods, and apparatuses for driving a plurality of loads using a single pulse generator.

[0016] To use a pulse power supply more efficiently for powering electrodes running different recipes in parallel, the present disclosure describes a single pulse power supply coupled to a multiplexing unit that distributes pulse power from a single pulse power supply to multiple loads, each having distinct recipes or pulse frequency and phase requirements. In other words, the multiplexing unit may receive a series of fast pulses from the single pulse power supply and output multiple pulse trains (or distributed pulses) to multiple loads, wherein the pulse trains may have different frequencies and / or duty cycles. In some embodiments, the multiplexing unit may be a component or subsystem within the single pulse power supply. Each output pulse train may have a frequency lower than the input frequency, and each output pulse train may have different frequencies. Phase shifts may be applied to the pulse power supply or the multiplexing unit to prevent overlap between pulses provided to different loads. In other words, since each input pulse can be supplied to only one of a plurality of loads, a portion of the input pulses may be phase-shifted in a pulse power supply or multiplexing unit to avoid overlap between pulse trains for the plurality of loads. In one embodiment, the present disclosure covers any variation of nD < 100%, where n is the number of channels or loads driven and D is the average duty cycle of all pulse trains. That is, as the number of channels increases, the average duty cycle per channel decreases. Alternatively, a higher duty cycle may be achieved when fewer channels are used.

[0017] In addition to improving the reliability of the power supply, this disclosure enables the use of smaller power supplies. Traditionally, when multiple loads can be powered simultaneously (i.e., when redundant output pulses are allowed), power supplies are often sized to handle multiple simultaneous pulses. This disclosure allows for power supplies customized to provide no more power than required for a single pulse by eliminating the possibility of pulse overlap. Thus, smaller and less expensive power supplies can be used in the implementation of this disclosure. In some embodiments, the loads may include one or more electrodes of a plasma processing chamber.

[0018] The word “exemplary” is used herein to mean “that which acts as an example, case, or example.” Any embodiment described herein as “exemplary” must not be interpreted as being preferred or advantageous over other embodiments.

[0019] Preliminary Note: The flowcharts and block diagrams in the following figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, some blocks in these flowcharts or block diagrams may represent a module, segment, or part of code containing one or more executable instructions for implementing specific logical function(s). It should be noted that in some alternative implementations, the functions mentioned in the blocks may occur out of the order mentioned in the figures. For example, two consecutive blocks may actually be executed substantially simultaneously, or the blocks may be executed in reverse order depending on the functionality involved. Each block in the flowchart figures and / or block diagrams, and combinations of blocks in the flowchart figures and / or block diagrams, may be implemented by special-purpose hardware-based systems that perform specific functions or actions, or combinations of special-purpose hardware and computer instructions.

[0020] FIG. 1 illustrates an embodiment of a power delivery system for supplying pulsed power to two or more loads using a single pulse power supply. The pulse power supply (102) may be connected to a multiplexing unit (104) that converts a single pulse train from the pulse power supply (102) into a plurality of pulse trains referred to as distributed pulse trains, each of which is provided to an individual pulse output of the multiplexing unit (104), and each of these outputs is connected to a separate load. In the illustrated embodiment, loads 1 (110) through N (112) are illustrated, where N can be any number. The pulse power supply (102) ensures that pulses for the loads do not overlap. This may include scheduling the pulse trains in the pulse power supply (102), which may include applying a phase shift (e.g., phase delay) in the pulse power supply (102), changing the frequency, and / or changing the duty cycle. In some embodiments, and as illustrated in FIGS. 16 through 18, generating and / or coordinating the schedule of pulses may alternatively be performed in the multiplexing unit (104). Either way, the pulse power supply provides pulse power to the multiplexing unit (104), and the pulse power supply (102) also communicates with the multiplexing unit (104). The processing part (108) of the pulse power supply (102) may access target parameters for 1 to N loads of memory (106) and generate a schedule for the pulses distributed from the multiplexing unit (104). This analysis and schedule generation may be performed in the processing part (108) which executes processor-executable code stored in a processor-executable medium of the same type as memory (106).The generation of the schedule may involve predicting when distributed pulses for the load (110, 112) should occur when the target parameters (e.g., frequency, duty cycle, conduction time) are tracked. This predicted schedule may also indicate where overlap between distributed pulses will occur, and the processing unit (108) may adjust the schedule to avoid such overlap (e.g., applying a phase delay to specific pulses). In the process, the processing unit (108) may also change the schedule of the distributed pulses (e.g., perform such phase shifting) for the purpose of maintaining the average duty cycle and frequency of each pulse train of the distributed pulses as close as possible to the target parameters (e.g., processing recipe) (some variation from the target parameters is often allowed, but the amount of allowable variation depends on the recipe). Since a specific pulse train among the pulse trains may be given priority in this regard, the processing part (108) must choose between phase-shifting the first pulse train and the second pulse train, and if the second pulse train has priority, the processing part (108) may choose to phase-shift one or more pulses from the first pulse train.

[0021] Although the multiplexing unit (104) is depicted as a separate component or system from the single pulse power supply (102), in other embodiments, the multiplexing unit (104) may be a sub-component or sub-system of the single pulse power supply (102). For example, FIG. 2 illustrates one such embodiment. In this embodiment, the multiplexing unit (204) may be part of the pulse train generation of the pulse power supply and may operate to distribute pulses to a number of loads (e.g., loads 1 through N). FIG. 2 illustrates one alternative embodiment of the multiplexing unit and its arrangement in or for the pulse power supply, and many other non-limiting examples are possible.

[0022] FIG. 3 illustrates a detailed embodiment of a pulse power supply (302) (e.g., 102 of FIG. 1). In some embodiments, a single pulse power supply may provide a pulse train to a pulse output (314) configured to be coupled to a multiplexing unit (104), and the multiplexing unit (104) may distribute the pulse train to a plurality of pulse outputs. The pulse train may have one or more phase shifts or other scheduling changes applied thereto to achieve a pulse train for loads 1 through N that avoids overlap between distributed pulses while satisfying target parameters for each load. For example, it may be desirable to apply pulses of a first frequency to load 1 and pulses of a second frequency to load N. A memory (306) may store target parameters for each load (e.g., recipe information), such as pulse frequency, duty cycle, and conduction time for various loads. The pulse frequency and duty cycle for each load may be associated with a time period (t1), while different pulse frequencies and duty cycles may be stored during a second time period (t2). That is, as the recipe changes over time, these target parameters may change. These are referred to as "target" parameters rather than strict and fast requirements, and are associated with acceptable variation, so that a phase delay may be applied to the pulse train of each load within the limits of acceptable variation. This initiation is particularly useful when there is a frequency difference between pulses within a time period (t1), because if this frequency difference is not adjusted, it can cause overlapping pulses in a multiplexed system.

[0023] The memory (306) may also include processor-executable instructions that, when executed on the processing unit (308), cause the processing unit (308) to control the phase shift circuit (324) to schedule a pulse train at the output (314). Additionally, the processing unit (308) may control the phase shift circuit (324) to phase shift one or more pulses in the pulse train of the pulse power supply (302) to ensure that there is no overlap between the output pulse trains. More specifically, the processing unit (308) may instruct the phase shift circuit (324) to generate a scheduled pulse train that can be divided into a plurality of distributed pulse trains without any overlap between the distributed pulse trains by the multiplexing unit (104).

[0024] In one embodiment, the phase shift circuit (324) may receive a periodic pulse train from the periodic pulse generator (330) of the pulse power supply (302) and apply a phase shift to generate a pulse train provided to the output (314). In this embodiment, the periodic pulse generator (330) may include any number of different topologies configured to generate periodic pulse outputs. For example, the periodic pulse generator (330) may include a DC voltage supply coupled to one or more pairs of half-bridge or full-bridge switching circuits configured to convert a DC voltage into a pulse output. However, other switching power supplies may also be implemented to perform the periodic pulse generator (330). In another embodiment, the phase shift circuit (324) and the periodic pulse generator (330) may be, for example, an integrated circuit (not shown), wherein the DC voltage supply provides a DC voltage to one or more switching converters whose switching is controlled by a phase controller that receives a command / indication from the processing part (308). The phase controller can control the timing of switching in one or more switching converters so that the output includes a pulse train containing pulses that are phase-shifted from a periodic pulse train to satisfy the schedule required to supply to the multiplexing unit (104) and to avoid pulse overlap between distributed pulse trains.

[0025] In any of these topologies, the plot for the single pulse supply output at the bottom of FIG. 5 shows one exemplary but non-limiting pulse train that includes a phase delay to avoid overlap in the distributed pulse trains (see channels 1-5) that the multiplexing unit (104) splits from a single input pulse train and provides to the loads (1 to N).

[0026] At the same time, the processing unit (308) adjusts the phase shift to maintain the average duty cycle, frequency, or conduction time for the distributed pulses of the multiplexing unit (104) that are aligned with the required target parameters (e.g., recipe) for the frequency, duty cycle, or conduction time. In other words, the processing unit (308) selects the phase shift to maintain the recipe or to etch within the stated variation parameters. This may include changing future pulses in a relationship opposite to the first variation to achieve the target average duty cycle, frequency, or conduction time, or to maintain within the recipe variation (or variation threshold). For example, if a pulse of a channel is delayed at time (t1) to avoid overlap, the same channel may have a subsequent pulse accelerated at a future time (t2) where overlap is not an issue to result in an average conduction that matches the target. Accordingly, the processing unit (308) attempts to generate a pulse train in the output (314) that achieves a desired average frequency and duty cycle for each channel over a "long" time period, such as 1 second, while avoiding the overlap of pulses in different distributed pulse trains when multiplexed by the multiplexing unit (104).

[0027] Additionally, since certain pulse trains among the distributed pulse trains may be given priority in this regard, the processing part (308) must choose between phase-shifting the first pulse train and the second pulse train, and if the second pulse train has priority, the processing part (308) may choose to phase-shift one or more pulses from the first pulse train.

[0028] After the processing unit (308) generates a schedule for the pulse train at the output (314) and a schedule for the distributed pulses at the multiplexing unit (104), it transmits some or all of this information to the multiplexing unit (104) via the communication output (316) through the communication link. The multiplexing unit (104) can use this information to control the distribution of pulses to 1 to N loads. In particular, the processing unit (308) can transmit the schedule of the distributed pulses to the multiplexing unit (104) so ​​that the multiplexing unit (104) can know which channel (or load) to distribute each pulse of the pulse train from the pulse power supply (302) to. In one embodiment, the processing unit (308) can transmit a channel indication to the communication output (316) in time with the pulses at the power output (314). For example, FIG. 6 illustrates a timing chart for a single pulse train (604) from a pulse power supply (302), as well as a timing chart (602) for five channels of a multiplexing unit (104) and a channel indicator (606) corresponding to each of the pulses (604). The multiplexing unit (104) can receive this channel indicator (606) from the pulse power supply (302) and send its output to the corresponding channel. In this way, the multiplexing unit (104) operates very similarly to a "slave" device for the pulse power supply (102).

[0029] FIG. 4 illustrates a multiplexing unit (404) according to one embodiment of the present disclosure. The illustrated multiplexing unit (404) may include a power distribution switching assembly connected to a communication input (416) and a power input (414). The power distribution switching assembly (410) may also be coupled to a plurality of outputs (418, 420), one for each of 1 to N loads. A pulse train from a pulse power supply (102) may arrive through the power input (414), and a communication informing the power distribution switching assembly (410) where to send each pulse of the pulse train arrives through the communication input (416). For example, a channel indicator may arrive at the communication input (416) at the same time as or slightly before the pulse of the pulse train arrives at the power input (414). Communication, which may include a channel indicator, may allow or command the power distribution switching assembly (410) to send each pulse of the pulse train to a corresponding pulse output among the pulse outputs (418, 420). The power distribution switching assembly (410) may divide the pulse train into a plurality of distributed pulse trains and provide these distributed pulse trains to the pulse outputs (418, 420). The pulse outputs (418, 420) are configured to be coupled to 1 to N loads. Later embodiments will discuss how the power distribution switching assembly (410) may apply a phase shift to the pulse train, but in this embodiment, the power distribution switching assembly (410) only serves to distribute the pulses to different outputs (418, 420), while the timing of the pulses is handled by the pulse power supply (102).

[0030] Various topologies may be implemented to perform the power distribution switching assembly (410). For example, a single input connected to the power input (414) may be connected to 1 to N outputs through a switch for each output. A switch driver may control which of the switches is opened at any given time to control the distribution or multiplexing of an incoming pulse train to the various outputs (418, 420).

[0031] Returning to FIG. 3, the memory (306) of the pulse power supply (302) may contain a minimum off period for each possible channel. The minimum off period may be used to ensure that all channels can be turned on before any one channel starts its next cycle. The minimum off period for a channel may vary depending on the number of channels, the on time and dead time of each channel. For example, assuming a 5-channel system and referring to FIG. 6, the minimum off period for channel 1 is five times the sum of the on times of the other channels 2-5 plus the dead time.

[0032] FIG. 5 illustrates an exemplary timing chart that may be associated with the embodiments of FIGS. 1 through 4, where N = 5. Each of channels 1 through 5 provides a pulse train distributed to a distinct load among the five loads. The pulse power supply can access recipe requirements for each load and acceptable variations therefrom from memory, and thus can predict the timing of pulses of each channel and thereby predict overlaps between channels. For example, the pulse power supply can recognize that channels 2 and 3 have higher frequencies than channel 1, which leads to predicted pulse overlap in the second pulse between channels 1 and 2 and predicted pulse overlap in the third pulse between channels 1 and 3. The pulse power supply can phase-shift channels 2 and 3 to avoid these overlaps. However, if channel 2 is shifted, a new predicted overlap appears between channels 2 and 3. Consequently, the pulse power supply can phase-shift channel 3 to avoid these overlaps. The example shows how such phase shifting can be continued throughout the recipe and for each of the various channels. For example, the fourth pulse period of Channel 1 exhibits a phase shift to avoid overlap with the phase-shifted Channel 5. Once the pulse power supply determines a schedule for a channel that does not include pulse overlap between distributed pulses, the pulse power supply can command the multiplexing unit to follow this schedule and generate a pulse train that supports this schedule (e.g., by delivering the channel indicator in time along with all pulses of the pulse train). That is, instead of sending a periodic pulse train, the pulse power supply applies a phase shift or other variation to the pulses of its pulse train so that the pulse train is delivered to the multiplexing unit, which already has the timing of the pulses to be multiplexed built into that pulse train.For example, the bottom plot of Fig. 5 shows the resulting pulse train that can be generated by the pulse power supply and supplied to the multiplexing unit after determining the pulse schedule for the five channels shown above a single pulse train.

[0033] FIG. 6 illustrates another exemplary timing chart that may be associated with the embodiments of FIGS. 1 through 4, where N = 5 and shows possible configuration parameters for the processing portion (308). Although the present disclosure has discussed frequency, duty cycle, and conduction time as possible target parameters, in other embodiments, channel on and off times and channel start times may be target parameters and, in some cases, the same results as specifying frequency, duty cycle, and / or conduction time may be achieved. In particular, each channel may be assigned a target cycle, channel on time, channel off time, and channel start time.

[0034] In one embodiment, the multiplexing unit may switch channels before a certain time period before the arrival of each corresponding pulse from the pulse power supply. In one embodiment, the pulse power supply may send a channel indicator before sending the corresponding pulse (e.g., the pulse is delayed by 500 μs from the channel indicator). For example, plot (606) may show the channel indicator from the pulse power supply and plot (608) may show the arrival of the pulse train, which is delayed by 500 μs from the pulse train (604) so ​​that the channel indicator and the pulse train are perfectly aligned. This delay helps ensure that each pulse reaches approximately the middle of the channel selection indicator, as shown in plot (606). The delay may be calculated as half a dead time period.

[0035] FIGS. 7 through 11 illustrate aspects of timing charts showing phase delays applied to a 5-channel system. In the illustrated embodiment, an algorithm for predicting pulse overlap and applying phase adjustment to avoid overlap (in a pulse power supply or multiplexing unit) begins by finding overlap in a schedule of distributed pulses where the scheduling is based on target parameters. In this example, the configuration is: ON time = 1 ms for all channels; Channel 1 off time = 57 ms; Channel 2 off time = 52 ms; Channel 3 off time = 48 ms; Channel 4 off time = 42 ms; Channel 5 off time = 36 ms. This example uses target off times, but other target parameters such as frequency, duty cycle, and conduction time may be used, to give a few non-limiting examples. Given these target parameters, the processing part may evaluate overlap periodically (e.g., every 10 μs). If overlap is predicted in the schedule, the processing part may instruct that a delay be applied to one or more channels. To avoid delays, multiple delays may be applied to multiple channels, and in some cases, it may be necessary to return to previously delayed channels and apply one or more additional delays. In other words, the process of changing the schedule can be repeated until all overlaps are removed from the schedule.

[0036] FIG. 8 illustrates an overview of the steps that the processing unit may take to eliminate the overlap shown in FIG. 7. The result (804) of applying a variation to the schedule of the second cycle of pulses generates a timing chart (804) without overlap. In other words, after the variation (e.g., phase delay) is applied, the timing of the distributed pulses may appear on the far right (804). The time points (806) of times 1 through 5d are each associated with the phase delay applied by the processing unit as shown in FIG. 9 through 12.

[0037] Figure 9 illustrates the analysis of the first cycle of pulses, where it can be seen that no overlap is predicted and therefore no variation needs to be applied.

[0038] FIG. 10 illustrates the analysis of pulses in the second cycle in which overlap between channels 1 and 2 is identified. In response, the processing portion delays channel 2 by 4 ms.

[0039] FIG. 11 illustrates the identification of additional overlaps in the second cycle of pulses and the application of a variation to eliminate these overlaps. Here, the processing part identifies overlaps between channels 1 and 3 and between channels 2 and 3. Then, delays of 2ms and 4ms are applied to channels 2 and 3, respectively. However, this still leaves overlaps between channels 2 and 3.

[0040] FIG. 12 illustrates the identification of additional overlaps in the second cycle of pulses and the application of a variation to eliminate these overlaps. Here, the processing part identifies the overlap between channels 2 and 3 and applies a delay of 4 ms to channel 2 so that none of the five channels overlap any further.

[0041] The distributed pulses of this scheduled set for the second cycle of pulses can then be executed by the pulse power supply, and the new channel sequence and timing can be transmitted to the multiplexing unit to control the pulse column distribution when it reaches the multiplexing unit.

[0042] FIG. 13 illustrates a method of distributing two or more pulse trains to respective loads, wherein the power supply for such distribution is a single pulse power supply. The method (1300) includes accessing a recipe (or other target parameters) for two or more loads (Block 1302). Such recipes may include duty cycles, frequencies, and / or conduction times for one or more time periods of a recipe associated with a load, such as a plasma processing load. For example, the duty cycle of the pulse power delivered to the plasma processing chamber may be changed as the recipe progresses to achieve different thin film deposition or etching characteristics. Different pulse duty cycles and / or frequencies may also achieve different etching characteristics, such as sidewall profiles or aspect ratios for wafer through-vias, to give just two examples. Since different chambers may use different recipes or recipes may be started at different times, pulse overlap may occur if power is supplied independently to each load. The recipes may include acceptable variations for parameters such as pulse frequency, pulse duty cycle, and conduction. While this method (1300) and most of the present disclosure describe a “recipe” that requires different pulse parameters for each of two or more loads, applications other than plasma processing can also benefit from the present disclosure’s ability to provide different sequences of power pulses to multiple loads while avoiding pulse overlap.

[0043] The method (1300) can predict the timing for each pulse for two or more channels (e.g., the output channel of a pulse power supply having a multiplexing function or the output channel of a multiplexing unit separated from the pulse power supply) (Block 1304). For example, the first two cycles of pulses shown in FIG. 8 may be predicted pulse timings for a 5-channel system, where it can be seen that there is a predicted overlap in the second of the two pulse cycles. The pulse power supply or multiplexing unit may phase-shift these future pulses to avoid overlap (see Phase-shifted schedule of the second cycle (804)). In other words, the pulse power supply or multiplexing unit may set a new schedule for future pulses and implement this schedule through internal phase shifting or, to two non-limiting examples, adjustments to other parameters such as frequency and duty cycle. The method (1300) can identify pulse overlap between two or more channels in the prediction (Block 1306), and can apply a phase delay to the lower priority channel to avoid the predicted overlap between two or more channels given optional priorities (Block 1308). When the phase delay is applied, it is typically applied for only a single cycle, and the period of subsequent cycles is kept the same so that the next cycle after the delay is shorter than the average, which helps to shift the average back to the target parameter. However, optionally, the method (1300) identifies future pulses for the channel to be viewed for phase delay, where additional phase delay or "phase hastening" (the opposite of phase delay) can be applied to pull the average duty cycle and / or average frequency back toward the target value (Block 1310).For example, if a phase delay of 4ms is applied to the first channel in the first time period and the method (1300) identifies a later pulse for the first channel, a 4ms phase acceleration can be applied without creating pulse overlap, and then the 4ms phase acceleration can be applied to the first channel to bring the average of the first channel back to the target.

[0044] The method (1300) can be applied to embodiments (Figs. 1 to 4) in which the processing part and schedule generation are in a pulse power supply, as well as embodiments (e.g., Figs. 16 to 18) in which the processing part and schedule generation are in a multiplexing unit.

[0045] FIG. 14 illustrates a method for driving two or more distinct loads using a single pulse power supply and avoiding overlapping pulses for two or more distinct loads. The method (1400) may include accessing target parameters (e.g., frequency, duty cycle, conduction time, etc.) for each of two or more distinct loads (Block 1402). This may include switching to the memory of the pulse power supply (e.g., the memory (106 or 306) of FIG. 1 and FIG. 3) and accessing parameters stored for each of a plurality of loads, including at least a target pulse frequency and / or a target pulse duty cycle, wherein the target pulse frequency and the target pulse duty cycle may be scheduled to change over time. Allowable variations in these target parameters may also be accessed from the memory. The method (1400) may further include the step of predicting instances of overlap between distributed pulses configured to be provided to two or more distinct loads when the target parameters are followed without variation (Block 1404). That is, the pulse power supply can monitor target parameters for each load and predict when a pulse for each load will occur if the target parameters are accurately followed (Block 1406). If one or more distributed pulses for different loads overlap in time (i.e., are scheduled to occur at least partially simultaneously), the method (1400) may instruct the phase shift circuit of the pulse power supply (e.g., 324) to change the schedule of the pulses in the pulse train to avoid the overlap(s) (Block 1408). This may include applying a phase shift or changing the frequency and / or duty cycle.

[0046] Once a new schedule is established and a variation is applied to achieve the schedule, the method (1400) transmits the channel identification to the multiplexing unit to indicate which channel each pulse of the pulse train is to be delivered to (Block 1410). Additionally, the method (1400) may look for future opportunities to apply an opposing variation or phase shift to shift the average frequency and / or target pulse duty cycle back to the target pulse frequency and / or target pulse duty cycle for each channel (Block 1412). In particular, the method (1400) may monitor (1) the difference between the target frequency and the average frequency of each distributed pulse train and (2) the difference between the target duty cycle and the average duty cycle of each distributed pulse train. With these differences, the method (1400) may look for future scheduling opportunities to apply one or more additional variations to the distributed pulse schedule to reduce these differences (Block 1412). For example, if the first channel exhibits a phase delay to avoid a first overlap with the second channel, future pulses for that channel may be accelerated or phase-shifted forward in time to shift the average frequency back to the target frequency. Such future variations may be applied in a pulse power supply. The method (1400) may then be repeated for subsequent additional pulse overlaps.

[0047] FIG. 15 illustrates a timing diagram that helps show how the processing part of a pulse power supply checks for overlaps of distributed pulses for an exemplary 5-channel system. The cycles shown in FIG. 15 correspond to pulse cycles. During phases 1 through 5 (P1 through P5), the processing part can sequentially check for any overlap with a specific channel. For example, P1 checks for an overlap with channel 1, and P2 checks for an overlap with channel 2. When each analysis occurs, the delay can also be based on a configured priority. For example, when comparing two channels, a channel may be delayed only if there is an overlap, and the channel to be delayed is scheduled to send a pulse after the other channel.

[0048] FIG. 16 illustrates an embodiment of a power delivery system for supplying pulsed power to two or more loads using a single pulse power supply. A pulse power supply (1602) may be connected to a multiplexing unit (1604) that converts a single pulse train from the pulse power supply (1602) into a plurality of pulse trains called distributed pulse trains, each of which is provided to an individual pulse output of the multiplexing unit (1604), and each of these outputs is connected to a separate load. In the illustrated embodiment, loads 1 (1610) through N (1612) are illustrated, where N can be any number. The multiplexing unit (1604) ensures that the pulses for the loads do not overlap. This may involve a multiplexing unit that performs a phase shift on the pulses, or may involve a multiplexing unit that instructs the pulse power supply (1602) to change the frequency and / or duty cycle of a specific pulse among its pulses or to phase shift its pulses. In either case, the pulse power supply provides pulse power to the multiplexing unit (1604), and the pulse power supply (1602) also communicates with the multiplexing unit (1604). The pulse power supply (1602) may display the pulse frequency and duty cycle to the multiplexing unit, and the multiplexing unit may provide this information to a processing unit (1608) that predicts output pulse trains for loads (1610, 1612) based on a recipe (e.g., duty cycle and frequency) for each of the loads stored in memory (1606). If the processing unit (1608) predicts an overlap of pulses between different pulse trains, the multiplexing unit (1604) may phase-shift one or more pulses from one or more pulse trains to avoid the overlap.The processing unit (1608) may also perform this phase shifting for the purpose of maintaining the average duty cycle and frequency of each pulse train as close as possible to the target parameter (e.g., processing recipe) (some variation from the target parameter is often allowed, but the amount of allowable variation depends on the recipe). Since certain pulse trains among the pulse trains may be given priority in this regard, the processing unit (1608) must choose between phase shifting the first pulse train and the second pulse train, and if the second pulse train has priority, the processing unit (1608) may choose to phase shift one or more pulses from the first pulse train.

[0049] Although the multiplexing unit (1604) is depicted as a separate component or system from the single pulse power supply (1602), in other embodiments, the multiplexing unit (1604) may be a sub-component or sub-system of the single pulse power supply (1602). For example, FIG. 17 illustrates one such embodiment. In this embodiment, the multiplexing unit (1704) may be part of the pulse train generation of the pulse power supply and may operate to distribute pulses to a number of loads (e.g., loads 1 through N). For example, the multiplexing unit (1704) may instruct the switching circuit of the pulse power supply to generate a single pulse train that aligns with a desired single pulse train that already includes phase adjustment so that the multiplexing unit (1704) only needs to distribute pulses to different loads without further phase changes. FIG. 17 illustrates one alternative embodiment of a multiplexing unit and its arrangement in or for a pulse power supply, and many other non-limiting examples are possible.

[0050] FIG. 18 illustrates a detailed embodiment of a multiplexing unit (e.g., 1604 of FIG. 16 or 1704 of FIG. 17). In some embodiments, a single pulse power supply (e.g., 1602 of FIG. 16) may provide a series of periodic pulses from a power input (1814) to the multiplexing unit (1804), and the multiplexing unit (1804) may distribute power signals to a plurality of pulse outputs (1818, 1820) with one or more different phase shifts applied to achieve desired pulse series to multiple loads. For example, it may be desirable to apply pulses of a first frequency to load 1 and pulses of a second frequency to load N. One or more of these frequencies may be higher or lower than the frequency of the pulses received at the power input (1814). In particular, a memory (1806) may store recipe information, such as pulse frequencies and duty cycles for various loads. The pulse frequency and duty cycle for each load may be associated with a time period (t1), while different pulse frequencies and duty cycles may be stored during a second time period (t2). That is, as the recipe changes over time, these parameters may change. These may be target parameters rather than strict and fast requirements, and may be associated with acceptable variation, so that a phase delay may be applied to the pulse train of each load within the limits of acceptable variation. This initiation is particularly useful when there is a frequency difference between pulses within a time period (t1), because if this frequency difference is not adjusted, it can cause overlapping pulses in a multiplexed system.

[0051] The memory (1806) may also include processor-executable instructions that, when executed on the processing part (1808), cause the processing part (1808) to control the power distribution switching assembly (1810) to distribute a single pulse from the power input (1814) into a plurality of pulse outputs (1818, 1820) (or distributed pulses). Additionally, the processing part (1808) may control the power distribution switching assembly (1810) to phase-shift one or more pulses from one or more of the distributed pulse trains to ensure that there is no overlap between the output pulse trains. More specifically, the power splitting circuit (1822) may split an input pulse train from the power input (1814) into two or more distributed pulse trains at the pulse outputs (1818, 1820). A phase shift circuit (1824) for each of the N outputs can delay or speed up (i.e., phase shift) the distributed pulses to avoid overlap between pulse trains (this can be seen as a temporary change in duty cycle or frequency).

[0052] At the same time, the processing unit (1808) adjusts the phase shift to maintain the average duty cycle, frequency, or conduction time for each of the pulse outputs (1818, 1820) that are aligned with the required target parameters (e.g., recipe) for the frequency, duty cycle, or conduction time. In other words, the processing unit (1808) selects the phase shift to maintain the recipe or to etch within the stated variation parameters. This may involve changing future pulses in a relationship opposite to the first variation to achieve the target average duty cycle, frequency, or conduction time or to maintain within the recipe variation. For example, if a pulse of a channel is delayed at time (t1) to avoid overlap, the same channel may have a subsequent pulse accelerated at a future time (t2) where overlap is not an issue to result in an average conduction that matches the target. Accordingly, the processing part (1808) seeks to generate a pulse train at the output (1818) that achieves the desired average frequency and duty cycle for each channel over a "long" time period, such as 1 second, while avoiding overlapping pulses. In some embodiments, the processing part (1808) may instruct the pulse power supply to change its pulse train to support the schedule, in which case the phase shift circuit (1824) may not be required.

[0053] Additionally, since certain pulse trains among the distributed pulse trains may be given priority in this regard, the processing part (1808) must choose between phase-shifting the first pulse train and the second pulse train, and if the second pulse train has priority, the processing part (1808) may choose to phase-shift one or more pulses from the first pulse train.

[0054] In another embodiment, the processing unit (1808) may determine the frequency, phase, and duty cycle of the input pulse train in 1814 by communicating with the single pulse power supply through the communication input / output (I / O) (1816) or by measuring the pulse train. The processing unit (1808) may instruct the power supply to modify the phase, duty cycle, and frequency of its output to help support the schedule set by the multiplexing unit (1804). Alternatively, the processing unit (1808) may instruct the power distribution switching assembly (1810) without causing adjustment to the pulse power supply.

[0055] Memory (1806) may include a minimum off period for each possible channel. The minimum off period may be used to ensure that all channels can be turned on before any one channel starts its next cycle. The minimum off period for a channel may vary depending on the number of channels, the on time and dead time of each channel. For example, assuming a 5-channel system and referring to FIG. 6, the minimum off period for channel 1 is five times the sum of the on times of the other channels 2-5 plus the dead time.

[0056] In FIGS. 16 and 18, it will be seen that communication between the pulse power supply and the multiplexing unit is not always necessary. For example, if the processing part (1608) determines the phase, frequency, and / or duty cycle at the input (1614) and controls only the phase shift within the power distribution switching assembly (1610), communication with the pulse power supply (1602) may not be used.

[0057] Referring again to FIG. 5, the exemplary timing chart can be associated with the embodiment of FIG. 16, where N = 5. Each of channels 1 through 5 provides pulse power to a distinct load among the five loads. The multiplexing unit can access recipe requirements for each load and acceptable variations therefrom from memory, and thus can predict the timing of pulses of each channel and thereby predict overlaps between channels. For example, the multiplexing unit can recognize that channels 2 and 3 have higher frequencies than channel 1, which leads to predicted pulse overlap in the second pulse between channels 1 and 2 and predicted pulse overlap in the third pulse between channels 1 and 3. The multiplexing unit can phase-shift channels 2 and 3 to avoid these overlaps. However, if channel 2 is shifted, a new predicted overlap appears between channels 2 and 3. Consequently, the multiplexing unit can phase-shift channel 3 to avoid these overlaps. The example illustrates how such phase shifting can be continued throughout the recipe and for each of the various channels. For example, the fourth pulse period of Channel 1 exhibits a phase shift to avoid overlap with the phase-shifted Channel 5. Once the multiplexing unit determines a schedule for channels that does not involve pulse overlap between distributed pulses, the multiplexing unit may instruct the pulse power supply to apply a phase shift to the pulse train from the pulse power supply to follow this schedule, or to apply a variation to a single pulse train to generate a pulse train that supports that schedule. For example, the bottom plot of FIG. 5 shows the resulting pulse train that the pulse power supply can generate and provide to the multiplexing unit.

[0058] In one embodiment, a forward-looking desired pulse pattern at the input (1814) can be communicated back to a single pulse power supply via communication I / O (1816), and the single pulse power supply can modify its pulse pattern to generate a pulse pattern visible at the bottom of the chart—this is an ideal input pulse pattern that is multiplexed into a distributed pulse pattern visible at channels 1-5. In other words, the single pulse power supply can use this information (or a command from the multiplexing unit) to generate a pulse train that supports a pulse schedule determined by the multiplexing unit.

[0059] As previously discussed, the multiplexing unit can switch the channel for a certain period of time before each pulse from the pulse power supply arrives, as illustrated in FIG. 6. This can also be implemented even if the multiplexing unit includes a processing part rather than a pulse power supply.

[0060] FIGS. 8 to 12 can also be applied to embodiments of FIGS. 16 to 18, where the processing part is in a multiplexing unit.

[0061] FIG. 19 illustrates a method for driving two or more distinct loads using a single pulse power supply and avoiding overlapping pulses for two or more distinct loads. The method (1900) may include accessing target parameters (e.g., frequency, duty cycle, conduction time, etc.) for each of two or more distinct loads (Block 1902). This may involve switching to the memory of a multiplexing unit (e.g., memory (1806) of FIG. 18) and accessing parameters stored for each of a plurality of loads, including at least a target pulse frequency and / or a target pulse duty cycle, wherein the target pulse frequency and the target pulse duty cycle may be scheduled to change over time. Allowable variations in these target parameters may also be accessed from memory. The method (1900) may further include the step of predicting instances of overlap between distributed pulses configured to be provided to two or more distinct loads when the target parameters are followed without variation (Block 1904). That is, the multiplexing unit can observe the target parameter for each load and predict when a pulse for each load will occur if the target parameter is accurately followed (Block 1906). If one or more distributed pulses for different loads overlap in time (i.e., are scheduled to occur at least partially at the same time), the method (1900) may instruct (1) a power distribution switching assembly (e.g., the power distribution switching assembly (1810) of FIG. 18) to apply a phase shift to one or more distributed pulses to achieve a new distributed pulse schedule that avoids the overlap(s) (Block 1908) or (2) a pulse power supply to apply a phase shift to one or more pulses in its single pulse train to avoid the overlap(s) (Block 1910). This may include the application of a phase shift or a change in frequency and / or duty cycle.These two alternatives (1908, 1910) allow the pulse power supply or multiplexing unit to handle phase shifting or other pulse train changes to avoid overlapping of distributed pulses.

[0062] For example, in alternative A, the multiplexing unit may take a periodic and unmodified pulse train from a pulse power supply and distribute the pulses as pulses distributed to two or more individual loads, as well as apply an internal phase shift to these pulses so that the distributed pulses do not overlap. This alternative is illustrated by the timing chart in FIG. 20, where the pulse power supply provides a periodic pulse train that does not change its pulse train, but instead the multiplexing unit distributes and applies a variation or phase shift to achieve non-overlapping distributed pulses as seen in channels 1-3 (the pulses in dashed lines represent the original scheduled pulses to which no variation was applied). As illustrated, the variation may be temporally forward or backward. However, in an alternative embodiment, it may be preferable to apply only a delay (not shown), and this can also be applied in FIG. 20 by simply using a different algorithm to determine the variation.

[0063] In Alternative B, the multiplexing unit transmits a command to the pulse power supply to modify its single pulse train and then generates a modified pulse train that supports the distributed pulse schedule (e.g., via a power distribution switching assembly). For example, FIG. 21 shows a timing chart for a multiplexing unit (channels 1-3) supplying power to three separate loads (channels 1, 2, and 3) and a pulse power supply at the bottom of the four plots. The timing of the pulse train from the pulse power supply may be periodic before the variation is applied (see dashed pulse). However, with a schedule planned by the multiplexing unit for the three channels as illustrated, where pulse overlap does not occur, each channel is affected by different target parameters for the distributed pulse train. To this end, the pulse power supply is instructed to apply a variation to the pulse train to obtain a new pulse train (see solid pulse) that supports the scheduling of channels 1-3. When this modified pulse train reaches the multiplexing unit, the multiplexing unit distributes pulses to each load without applying any internal phase shifting.

[0064] Once a new schedule is established and a variation is applied to achieve that schedule, the method (1900) may find a future opportunity to apply an opposing variation or phase shift to shift the average frequency and / or target pulse duty cycle back to the target pulse frequency and / or target pulse duty cycle for each channel (Block 1912). In particular, the method (1900) may monitor (1) the difference between the target frequency and the average frequency of each distributed pulse train and (2) the difference between the target duty cycle and the average duty cycle of each distributed pulse train. With these differences, the method (1900) may find a future scheduling opportunity to apply one or more additional variations to the distributed pulse schedule to reduce these differences. For example, if the first channel exhibits a phase delay to avoid a first overlap with the second channel, future pulses for that channel may be accelerated or phase-shifted forward in time to shift the average frequency back to the target frequency. These future variations may be applied in a pulse power supply or a power distribution switching assembly. The method (1900) can be repeated for the next additional pulse superposition.

[0065] The methods described in connection with the embodiments disclosed herein may be implemented directly in hardware, in processor-executable code encoded in a non-transient type of processor-readable storage medium, or in a combination thereof. For example, referring to FIG. 22, a block diagram illustrating physical components that may be utilized to realize a multiplexing unit (104, 204, 304, 404, 1604, 1704, and 1804) according to an exemplary embodiment is shown. As illustrated, in this embodiment, a display portion (2212) and a non-volatile memory (2220) are coupled to a bus (2222) which is also coupled to a random access memory ("RAM") (2224), a processing portion (including N processing components) (2226), an optional field-programmable gate array (FPGA) (2227), and a transceiver component (2228) including N transceivers. The components illustrated in FIG. 22 represent physical components, but FIG. 22 is not intended to be a detailed hardware diagram; therefore, many of the components illustrated in FIG. 22 may be realized by common configurations or distributed among additional physical components. Furthermore, it is considered that other existing and as yet undeveloped physical components and architectures may be utilized to implement the functional components described with reference to FIG. 22.

[0066] This display unit (2212) generally operates to provide a user interface to a user, and in several implementations, the display is realized by a touchscreen display. Generally, the non-volatile memory (2220) is a non-transient memory that functions to store (e.g., continuously) data and processor executable code (including executable code associated with carrying out the methods described herein). In some embodiments, for example, the non-volatile memory (2220) includes bootloader code, operating system code, file system code and non-transient processor executable code to facilitate the execution of the method described with reference to FIG. 13 and FIG. 14, which are further described herein.

[0067] In many implementations, non-volatile memory (2220) is realized by flash memory (e.g., NAND or ONENAND memory), but it is considered that other memory types may also be utilized. While it may be possible to execute code from non-volatile memory (2220), executable code in non-volatile memory is typically loaded into RAM (2224) and executed by one or more of N processing components in the processing section (2226).

[0068] In relation to RAM (2224), N processing components generally operate to execute instructions stored in non-volatile memory (2220) to enable the distribution of pulses from a single pulse power supply to multiple loads. For example, non-transient, processor-executable code for carrying out the methods described with reference to FIG. 13 or FIG. 14 may be continuously stored in non-volatile memory (2220) and executed by N processing components associated with RAM (2224). As will be known to those skilled in the art, the processing unit (2226) may include a video processor, a digital signal processor (DSP), a microcontroller, a graphics processing unit (GPU), or other hardware processing components or a combination of hardware and software processing components (e.g., an FPGA or an FPGA including digital logic processing units).

[0069] Additionally or alternatively, the processing part (2226) may be configured to implement one or more aspects of the methodologies described herein (e.g., the method described with reference to FIG. 13 or FIG. 14). For example, non-transient processor-readable instructions may be stored in non-volatile memory (2220) or RAM (2224) and, when executed in the processing part (2226), may cause the processing part (2226) to perform multiplexing of pulse power to supply to a number of loads without overlapping of pulses for a number of loads and while attempting to maintain an average duty cycle and / or frequency for each load. Alternatively, non-transient FPGA configuration instructions may be persistently stored in non-volatile memory (2220) and accessed by the processing part (2226) to configure the hardware configurable parts of the processing part (2226) to perform the functions of the multiplexing units (104, 204, 304, 404, 1604, 1704 and 1804).

[0070] The input component (2230) operates to receive a signal (e.g., a communication signal at the output (316)) representing one or more aspects of a set of pulses from the single pulse power supply (102). The signal received at the input component may include, for example, a channel identifier or indications of target parameters. The output component operates to provide one or more analog or digital signals to implement an operational aspect of a multiplexing unit (104, 204, 304, 404, 1604, 1704, 1804), such as a control signal transmitted to the power distribution switching assembly (310) and / or the single pulse power supply (102). For example, the output portion (2232) may provide a command to the power distribution switching assembly (310) to perform a desired distribution of pulses to multiple loads or to command the single pulse power supply (102) to change its pulse pattern.

[0071] The illustrated transceiver component (1128) includes N transceiver chains, which may be used to communicate with external devices via wireless or wired networks. Each of the N transceiver chains may represent a transceiver associated with a specific communication method (e.g., WiFi, Ethernet, Profibus, etc.).

[0072] Some parts are presented in terms of algorithms or symbolic representations of operations on data bits or binary digital signals stored within computing system memory, such as computer memory. These algorithmic descriptions or representations are examples of techniques used by those skilled in the art in data processing fields to convey the content of their work to others skilled in the art. An algorithm is a self-consistent sequence of operations or similar processing that produces a desired result. In this context, operations or processing involve the physical manipulation of physical quantities. Typically, although not essential, such quantities may take the form of electrical or magnetic signals that can be stored, transmitted, combined, compared, or otherwise manipulated. In principle, for reasons of general use, it has sometimes proven convenient to refer to such signals as bits, data, values, elements, symbols, characters, terms, numbers, numerical values, etc. However, it must be understood that all these and similar terms are associated with the appropriate physical quantities and are merely convenient labels. Unless otherwise specifically stated, it will be understood that discussions throughout this specification utilizing terms such as “processing,” “computing,” “calculating,” “determining,” and “identifying” refer to actions or processes of computing devices, such as one or more computers or similar electronic computing devices, that manipulate or convert data represented as physical electronic or magnetic quantities within memories, registers, or other information storage devices, transmitting devices, or display devices of a computing platform.

[0073] As will be understood by those skilled in the art, embodiments of the invention may be embodied as a system, a method, or a computer program product. Accordingly, embodiments of the invention may take the form of an entire hardware embodiment, an entire software embodiment (including firmware, resident software, microcode, etc.), or an embodiment combining software and hardware embodiments, which may all be generally referred to herein as a “circuit,” “module,” or “system.” Additionally, embodiments of the invention may take the form of a computer program product embodied in one or more computer-readable media(s) on which computer-readable program code is implemented.

[0074] As used herein, the description “at least one of A, B, and C” is intended to mean “any combination of A, B, C, or any combination of A, B, and C.” The prior description of the disclosed embodiments is provided to enable those skilled in the art to practice or use the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the spirit or scope of the present disclosure. Accordingly, the present disclosure is not intended to be limited to the embodiments shown herein, but should be given the broadest scope suitable for the principles and novel features disclosed herein.

Claims

Claim 1 A system comprising: a pulse power supply configured to provide a pulse train; and a multiplexing unit comprising: a first output configured to be connected to a first load; and a second output configured to be connected to a second load, wherein the multiplexing unit divides the pulse train into a first pulse train and a second pulse train, provides the first pulse train to the first output and provides the second pulse train to the second output, and is configured to prevent overlap between the pulses of the first pulse train and the pulses of the second pulse train. Claim 2 A system according to claim 1, wherein the multiplexing unit comprises a phase shift circuit for generating the pulse train capable of dividing the pulse train into the first pulse train and the second pulse train. Claim 3 A system according to claim 1, comprising: a memory storing target parameters for the first load and the second load; and a processing part connected to the memory and configured to read processor-executable instructions within the memory, wherein the processor-executable instructions, when executed, cause the processing part to command the pulse power supply to change the pulse train to support a schedule that ensures there is no overlap between the pulses of the first pulse train and the pulses of the second pulse train. Claim 4 In claim 3, the system, wherein the schedule is adjusted to remain within the acceptable variation of the target parameters. Claim 5 In claim 3, the system, wherein the target parameters include the frequencies of the pulses for each of the first pulse train and the second pulse train. Claim 6 In claim 3, the system, wherein the target parameters include the on-time and off-time of pulses for each of the first pulse train and the second pulse train. Claim 7 In claim 1, the system wherein the multiplexing unit is part of the pulse power supply. Claim 8 A method comprising: accessing target parameters for a first load and a second load; predicting instances of overlap between pulses distributed to the first load and the second load when the target parameters are followed without variation, wherein the distributed pulses are formed from a single pulse train; scheduling variations into a schedule of the distributed pulses to avoid instances of overlap; and distributing pulses of the single pulse train to the first load and the second load according to the schedule. Claim 9 In claim 8, the step of distributing the pulses comprises the step of distributing the pulses to a power distribution switching assembly. Claim 10 A method according to claim 8, comprising the step of applying a phase shift while maintaining average parameters for the pulses within variations stored in memory. Claim 11 In claim 8, the method wherein the target parameters include the frequency of the pulses for each of the distributed pulses. Claim 12 In claim 8, the method comprises the target parameters including the on-time and off-time of the pulses for each of the distributed pulses. Claim 13 In claim 8, the distributing step comprises the step of distributing the pulses to a multiplexing unit. Claim 14 A non-transient type computer-readable storage medium encoded with processor-readable instructions for driving a first load and a second load using a single-pulse power supply and performing a method for avoiding overlapping pulses for the first load and the second load, wherein the method comprises: accessing target parameters for each of the first load and the second load; predicting instances of overlap between distributed pulses configured for provision to the first load and the second load when the target parameters are followed without variation, wherein the distributed pulses are formed from a single pulse train; scheduling variations into a schedule of the distributed pulses to avoid instances of overlap; and distributing pulses of the single pulse train to the first load and the second load according to the schedule. Claim 15 A computer-readable storage medium of a non-transient type, wherein the step of distributing the pulses in claim 14 includes the step of distributing the pulses to a power distribution switching assembly. Claim 16 In claim 15, the target parameters include the on-time and off-time of the pulses for each of the distributed pulses, a non-transient type of computer-readable storage medium. Claim 17 In claim 15, the target parameters include the frequency of the pulses for each of the distributed pulses, a non-transient type of computer-readable storage medium. Claim 18 In claim 14, the variation is a non-transient type of computer-readable storage medium that is below a variation threshold from the target parameters. Claim 19 In claim 14, the scheduling is a non-transient type of computer-readable storage medium adjusted to maintain the average parameters of the distributed pulses within the variation of the target parameters.

Citation Information

Patent Citations

  • Bipolar pulse power source and power source device formed by a plurality of bipolar pulse power sources

    KR101190138B1

  • Pulse width control signal generation circuit, power conversion control circuit, and power conversion control LSI

    KR1020090009786A

  • Channel adaptation to compensate for interference from neighbor powerline communication networks

    KR1020160013923A

  • Plasma processing apparatus

    US20180374687A1