US20040159354A1

US20040159354A1 - Applications for closed-loop motor control

Info

Publication number: US20040159354A1
Application number: US10/652,720
Authority: US
Inventors: Per Cederstav; Emmanuel Vyers; William Ballard; Sean Mallory
Original assignee: Nor Cal Products Inc
Current assignee: Nor Cal Products Inc
Priority date: 2000-12-15
Filing date: 2003-08-28
Publication date: 2004-08-19
Also published as: US6814096B2; EP1419427A2; US6612331B2; US20020109115A1; AU2002229013A1; US20020117212A1; WO2002048813A3; WO2002048813A2

Abstract

A closed loop pressure controller system that sets, measures and controls the process pressure within a semiconductor process is shown. The system is commonly composed of a pressure sensor to collect the pressure information, a controller box that hosts the control electronics, and a valve to physically affect the conductivity of the inlet or outlet gas line and accordingly the process pressure. The present invention differs from the prior art by using closed-loop motor control of the valve, rather than the method of the prior art, where the valve position is controlled by a stepper motor actuator driven in an open loop fashion. It is demonstrated that the utility of such prior art open-loop configurations is limited by the fact that the achievable precision of the valve position is hindered by static friction in the valve system, and the non-linear character of the torque versus shaft-angle of the motor (among other error components). The method of the present invention more accurately positions the valve, and accordingly enhances the overall precision and allowable loop-gain of the pressure control system by providing the valve drive with feedback as to the actual angular position of the valve in extremely high resolution. The closed-loop motor control approach of the present invention has also been adapted to other semiconductor and vacuum processing applications, including linear and rotary feedthrough devices and impedance matching networks.

Description

This application is a continuation of application Ser. No. 09/738,194, filed Dec. 13, 2000, and application Ser. No. 10/052,75 7, filed Jan. 18, 2002, both applications now pending.[0001]

BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates generally to automated motor control applications and, more specifically, to Applications for Closed-loop Motor Control

2. Description of Related Art

The term, “semiconductor processing equipment,” refers to a seemingly infinite variety of large, highly expensive pieces of machinery that are used to conduct a variety of different processes that ultimately result in a completed semiconductor device. What is a common design aspect for many pieces of semiconductor processing equipment is the need for accurate, fast and reliable pressure control of the vacuum within the chamber where the process is taking place. If we look at FIG. 1, we can review how a conventional semiconductor

processing tool system

10 is arranged today.

FIG. 1 is a depiction of a conventional semiconductor

processing tool system

10. As shown in FIG. 1, the processing tool 12 is typical supplied by gas that is transmitted from a gas supply 14 (such as the bottle shown) through a gas supply line 16 until it gets to the vicinity of (or inside of) the semiconductor processing tool 12, where the actual flow of the gas to the chamber is controlled by a mass flow controller 18. In this way, the tool 12 can regulate when and how much gas to inject into the processing chamber 20.

There is generally a

chamber pressure sensor

22 that provides an external signal via the pressure signal conduit 34. This external pressure signal typically can be either analog or digital in form, and represents the pressure conditions within the chamber 20. The signals are carried by a pressure signal conduit 34 to a conventional pressure control means 30. Within the pressure control means 30, the pressure signal is generally summed with a host tool logic signal later referred to as host tool pressure setpoint. The host tool pressure setpoint is generally generated by the tool logic controller 32, with its content being an analog or digital pressure setpoint value. These tool logic 215 are transmitted to the pressure control means 30 by a tool logic signal conduit 36.

If we refer back to the

tool

12, we can also see that another important feature that is many times found within the tool 12 is a plasma generator unit 23. This feature is important since plasma generators create sudden and sometimes large pressure deviations. Plasma generators essentially energize the gas molecules which splits them into ionized atoms and species. These ionized species are much more reactive than their molecular “parents” thus greatly speeding up and increasing the selectivity of processes such as etch and deposition. The instant the plasma is turned on, a fraction of the gas molecules split in to pieces thereby producing instant undesirable increases in chamber pressure. Similarly, the supply lines 16 (and the gas they transmit) also have an effect on the pressure within the chamber 20. The chamber 20 is generally kept in a vacuum state in order to prevent impurities from contaminating the semiconductor process. The conventional arrangement for maintaining the vacuum condition in the chamber 20 is via a vacuum source 24, such as the vacuum pump 24 shows The vacuum pump 24 simply pumps to an exhaust 25 while drawing a vacuum on a vacuum transmission line 26. Between the vacuum source 24 and the vacuum transmission line 26 is found a valve 28. It is by actuation of this valve 28 that the pressure can be raised and lowered (usually in the sub-atmospheric range) within the chamber 20.

Once the pressure signal and tool logic signal are summed in the pressure control means 30, the resulting signal is sent to a motor driver circuit 42 via an external valve command conduit 38. This conduit 38 is either hard wired via conventional cable, printed circuit board trace, or wire, however, it could also be wireless. The motor driver circuit 42 is actually a sub-component of a valve control assembly 40. The other components of the valve control assembly 40 are an internal valve command conduit 44 and a motor/valve drive assembly means 46 for actuating the valve 28. As should be appreciated, the signals generated by the pressure control means 30 are acted upon by the valve control assembly 40 to open and close the valve 28 such that the pressure in the chamber 20 is regulated As described above, the pressure control system is influenced by external factors called states of the process, in particular, the turning on and off of gas inputs to the chamber and the initiation of RF events to create plasma are primary contributing factors. The pressure control algorithm (executed by the Pressure Control Means 30) constantly works at maintaining the pressure regulated at the required value by actuating the valve in order to compensate and balance the pressure responsive to the changing states of the process. It is clear that the pressure regulation task can be performed only as well as the individual elements comprising the closed loop system permit. As such the valve control assembly (40) is an essential component in terms of its accuracy and speed of response to maintain quality and/or stability of the control system If we now turn to FIG. 2, we can look more closely at the valve control assembly 40 of the conventional system

FIG. 2 depicts a conventional

valve control assembly

40. As can be seen, the resultant signal of the summed commands from the pressure control means 30 in FIG. 1 arrive at the motor driver circuit 42 via an external valve command conduit 38. As discussed above, this is typically a cable that is rum for whatever length necessary to extend between the pressure control means 30 and the motor driver circuit 42. Between the motor driver circuit 42 and the motor/valve drive assembly means 46 is an internal valve command conduit 44. In the conventional system, this conduit, too, is an external cable running between the motor driver circuit 42 and the motor/valve drive assembly means 46. The motor/valve drive assembly means 46 conventionally comprises a motor drive 48 such as a conventional stepper motor, which in turn drives a required reduction gear, or other means of mechanical advantage 52 via a motor shaft 50. In other forms, the motor drive 48 is connected to a valve stem 54 via belts and pulleys. In any case, it is conventional in the art that there not be a direct connection or coupling between the motor drive 48 and the valve means 28 without some method of mechanical advantage or reduction gearing having the effect of increasing the number of revolutions of the motor drive 48 needed to create a full open to close cycle of the valve means 28. This mechanical advantage typically also has the beneficial effect of increasing the step resolution as many folds as the reduction factor of the mechanical reducer means. However, it also represents an actuation speed penalty of the same magnitude, as the motor has to travel farther for the same valve displacement. Additionally, the increased resolution is partially absorbed and degraded by the inherent nonlinearity (backlash) introduced by the mechanical reducer means. That actuation speed handicap has proved to be more detrimental to the quality of the pressure control dynamic characteristics and transient response performance than initially expected. A further note is that within the conventional internal valve command conduit 44, there is typically one single unidirectional path that extends from the motor driver circuit 42 to the motor drive 48 with the exception of two limit switches that are normally used within the motor valve drive assembly to reference the open and closed valve positions. These switches return a binary logic signal that cannot resolve position continuously across the stroke of the valve but only at two discrete locations—in order to distinguish these limit-switch-generated signals from signals to be discussed later on in connection with FIG. 4, we shall refer to these signals as “stroke reference feedback signals” We will refer to this path as the command leg 56. The command leg 56, again, is unidirectional (excluding the stroke reference feedback sis), and only extends from the motor driver circuit 42 to the motor drive 48, and not vice versa. If we now turn to FIG. 3, we can examine how the conventional chamber pressure control process 300 operates.

We will start with the host tool

pressure setpoint signal

302 arriving at the pressure control means 30. The pressure control means further comprises summing junction means 31 for the pressure sensor signal 314 to be compared with the host tool pressure setpoint signal 302 and generate a pressure error signal 304. That error signal is operated on by a pressure control algorithm 303 to produce a pressure control signal 306 that represents the desired change in valve position intended to correct said pressure error. If the system incorporates a conventional step motor drive, the pressure control signal 306 is transmitted from pressure control means 30 to the motor driver circuit 42 where it is converted to a position control signal 310. This signal 310 is then transmitted to the motor drive assembly means 46. Valve motion 312 is generated by actuating the valve stem 54. The valve stem 54 accordingly opens or closes the valve means 28 which, in turn, reduces or increases the conductance of the vacuum transmission line 26. This will respectively result in an increase or decrease in pressure within the processing chamber 20—a quantity that is continuously monitored by the pressure sensor 22. The monitored pressure is used to generate a pressure sensor signal 314 which is fed back to and again compared with the host tool pressure setpoint 302 by the summing junction 31. This above defined closed loop will herein be referred to as the pressure control loop. In practice the implementation of the pressure control loop is executed with electronics incorporating both discrete and continuous signals and is repeated in an iterative fashion.

As can be seen here, the

vacuum transmission line

26, the processing chamber 20 and the chamber pressure signal 316 are all depicted in dashed lines; this is to highlight the fact that the position of the valve is not the only condition to affect the chamber pressure. Because of numerous external factors such as the turning on and off of gas inputs to the chamber and the initiation of RF events, the stability of the process is often challenged or disturbed. The efficiency with which these disturbances can be handled or rejected is substantially dependent on the accuracy with which the valve drive means can be rapidly and efficiently operated. In that context the remaining portion of this application will be devoted to illustrating the advantage of a system that provides nested closed-loop position control of the motor drive assembly means 46 by the motor drive circuit means 42. This is implemented specifically to minimize the chamber pressure sensitivity to process variations and better exploit the pressure feedback information thus enhancing the pressure control performance.

In another realm related to semiconductor processing as well as vacuum processing, a device known as a “feedthrough” device is used. A feedthrough device is a piece of equipment that enables a user to manipulate an article that is located within a processing chamber while that chamber is in a processing condition (i.e. under vacuum, perhaps high temperature, very high cleanliness). Feedthrough devices are conventionally available that are manual, meaning that the user manipulates levers, dials and knobs outside of the processing chamber, which cause a sample head located within the processing or testing chamber to move in a desired fashion The simplest form of feedthrough device is the “linear” feedthrough device, which permits the user to move the sample head only linearly, in one, two or three axes (the number of axes of movement depends upon the particular feedthrough device design and purpose). The more complex feedthrough device is the “rotary” feedthrough device. A rotary feedthrough not only enables the sample head to be manipulated in linear fashion, but also enables it to be rotated.

Any of these feedthrough devices can be automated, which typically means that instead of having levers, knobs and dials (or in addition to these), the sample head is manipulated by electric motors. It is these automated feedthrough devices that is the subject of the present invention; tuning to FIG. 11, we can review the conventional apparatus.

FIG. 11 is a side view of a conventional

rotary feedthrough device

100. This device 100 has been depicted as a manual unit for simplicity of this discussion, and is very unlikely to be found as a component in a semiconductor processing line, but it should be apparent that it could easily function as an automated feedthrough device if the drives were replaced with electric motors. In fact, in the semiconductor processing environment, devices that function in a mechanically similar way to automated feedthrough devices are referred to as robotic arms or “wafer lifts;” wherever the term feedthrough device is used herein, it intended to apply to robotic arms, wafer lifts and other devices intended to manipulate articles inside the process chamber from outside the process chamber. The device 100 is divided between an atmospheric portion 102 (outside the chamber), and a chamber portion 104 by a base flange 112 (which is generally how the device 100 attaches to the chamber/tool.

The

chamber portion

104 is defined by a sample head 106 that comprises sample holder 108 for holding onto the silicon wafer sample (or other articles inside the chamber) as welt as other accessories (e.g. sample heater(s), electrical supply, etc.). The sample head 106 extends from the shaft 110 that passes through the base flange to the atmospheric portion 102 for control of its positioning.

The

atmospheric portion

102 comprises an X-axis drive 114 for causing the sample head 106 to move linearly along the X-axis, a Y-axis drive 116 for causing the sample head 106 to move linearly along the Y-axis, and a Z-axis drive 118 for causing the sample head 106 to move linearly along the Z-axis There is also at least one primary rotary drive 120 for causing the sample head 106 to rotate. In this example, there is also a secondary rotary drive 122 for finer adjustments of the rotary motion of the sample head 106 around the primary axis 124.

The problem with the conventional automated rotary feedthrough device design is related to the accuracy and responsiveness of the motor-actuated drive& FIG. 12 discusses the elements involved in this control

FIG. 12 is a block diagram depicting the functional description of the pertinent elements of the conventional

rotary feedthrough

100 device of FIG. 11. The device 100 has a user or tool interface 126 which provides electrical signals for directing the re-positioning of the sample head 106. These signals are received by the sample head position control means 30A which takes the position request signal and breaks it into X, Y, Z axis components and a rotational component (if appropriate) by a position control algorithm executed by the sample head position control means 30A These discrete, or axis-specific position commands are each transmitted to their respective motor drive means 42A within the sample head drive means 42A. The sample head drive means 42A translates the incoming discrete position commands into discrete (i.e. axis-specific) command signals to the respective axis-specific motors within the sample head motor means 46A (all of the motors are collectively the sample head motor means 46A). It should be understood that some embodiments employ single motors or motor combinations for making movements of the sample head in more than one axis (i.e. a combined X- and Y- axis motor means). Collectively, the sample head drive means 42A and the sample head motor means 46A are refereed to as the sample head drive assembly means 40A. The motors 46A then cause the sample head 106 to move via movement of the shaft 110. The process followed by the aforementioned conventional system is discussed in FIG. 13.

FIG. 13 is a block diagram depicting the conventional sample head

positioning control process

500. The process 500 commences with the user/tool position signal 502 being received by the sample head position control means 30A. Here, the discrete X, Y, Z and Rotational command signals 506 are generated 504 by the sample head position control means 30A The si(s) 506 are transmitted to the sample head drive means 42A, where the discrete X, Y, Z and Rotational position control signals 510 are generated 508. The position control signals 510 are transmitted to the sample head motor means 46A, wherein sample head movement is generated 512 (in the motor means') and transmitted to physical movement of the sample head 106 by the shaft 110. Because the control process 500 and system 100 is open loop conventionally, there is little or no control of error due to normal variation, wear, friction, calibration errors and other effects common with motor-driven systems The result of these effects is an potentially unstable (and therefore inaccurate) control system with unacceptable error introduced into the sample head positioning process.

Another control application that is related to semiconductor processing is that of plasma processing. FIG. 16 depicts a schematic side view of a conventional

plasma processing tool

140. The pertinent elements of the processing tool 140 are the process chamber 142, which has a substrate holder 144 and a RF electrode 146 located either therein or in an inlet manifold opening into the chamber 142. The chamber 142 is generally connected to a vacuum source 148 to maintain a stable, sub-atmospheric pressure within the chamber 142.

Plasma assisted (or enhanced) processing is an electrical phenomenon where a gas (molecular—not atomic) within the processing chamber is broken up in to its constituents (either atoms with a charge, or sub-molecular structures also with a charge). These “particles” (called radicals, because they are electrically unstable) are drawn towards, or even accelerated (e.g. reactive ion etching) toward the substrate (resting on the substrate holder 144). However, once they hit the substrate—or anything else for that matter—they recombine with the substrate yielding a neutrally charged final element. A positively charged substrate holder 144 will attract negatively charged radicals in a direction orthogonal to its surface. So, if the radicals' ultimate goal is to coat (i.e. deposition process) they can selectively be used to coat the tops of peaks, or bottoms of valleys, without clogging up the trenches (the line widths). If the process is etch, then the active radicals will again attack surfaces parallel to the wafer substrate without eating away at the wall structures.

In plasma processing, moreover, the generation of process gas plasmas is by means of electrical/magnetic field manipulation, namely the charge difference between the

RF electrode

146 and the substrate holder 144. It is the appropriate and timely adjustment of the electric/magnetic field's strength that determines how well the plasma is generated. That, in turn, affects how uniform and repeatable the process is.

The system for controlling the electrical field of the

RF electrode

146 is known as the impedance matching network 150. The impedance matching network 150, through detection and continuous control of critical impedances within the network 150, controls the energy level of the plasma, which in turn determines amount of material being deposited or etched by the RF electrode. In its automated form, the impedence matching network can generally be described as a motor-controlled variable capacitor, of which one example is shown below in FIG. 17.

FIG. 17 is a partial perspective view of the pertinent elements of a conventional

variable capacitor

154 used in a conventional automatic impedance matching device of FIG. 16. A variable capacitor 154 allows for a range of capacitance. Variable capacitors are designed so that capacitance can be changed through a mechanical means. Variable capacitors are commonly used when the application requires an adjustment of capacitance such as in a radio tuner.

The

capacitor

154 has two sets of plates. One set is called the rotor plates 156 and the other the stator plates 158. The rotor plates 156 are mounted on a rotor shaft 160 and the stator plates 158 are mounted on a stator shaft 162. A set of rotor plates 156 mounted to a rotor shaft 160 is collectively referred to as the rotor assembly 161; a set of stator plates 158 mounted to a stator shaft 162 is collectively referred to as the stator assembly 163.

The

rotor shaft

160 is connected to the knob outside the capacitor (manual variable capacitor) or to motor means 164 for rotating the rotor shaft 160. The two sets of

plates

156 and 158 are alternatingly spaced along their

respective shafts

160 and 162, close together but not touching. Air is the dielectric in a variable capacitor. As the capacitor 154 is adjusted, the sets of plates become more or less meshed (due to the non-circular shape of the plates 156 and 158), increasing or decreasing the amount of surface area overlapping between the

plates

156 and 158. As the

plates

156 and 158 become more meshed, the overlap area increases and capacitance increases. Conversely, as the

plates

156 and 158 become less meshed, capacitance decreases.

In another non-depicted conventional embodiment of a variable capacitor, rather than employing a rotor-stator pair having a rotating rotor, the rotor and stator could be simply large parallel plates. The stator plate would be stationary and the rotor plate would be in spaced relation to the stator plate, but would be movable both parallel and perpendicular relative to the stator plate. Just as with the rotating rotor version, with the “translating” version of rotor, movement of the rotor plate that results in a change in the overlap area between the rotor and stator plates or the distance between the plates will result in a change in capacitance. In order to be consistent, a “rotor linkage means” shall herein refer to the previously-discussed rotor shaft (for the rotating rotor capacitor version) as well as referring to the mechanical linkage or assembly necessary for mechanically ling the motor means 164 to the rotor plate in the translating variable capacitor. In FIG. 18, wherever the same process applies for the rotating capacitor version as the translating capacitor version, however, the rotor linkage means is the assembly or device that takes the place of the rotor shaft. FIGS. 18A and 18B provide additional background on the operation of conventional impedance matching networks.

FIG. 18A is an electrical schematic of the pertinent components of a conventional impedance matching network (series arthitecture) 150A and its association with the plasma processing chamber. FIG. 18B is an electrical schematic of the pertinent components of a conventional impedance matching network (parallel architecture) 150B and its association plasma processing chamber. The high frequency power source is provided by the frequency generator 170; its power is increased as desired by the amplifier 172. The variable capacitor 154 is adjusted so that the impedance matches the conditions of the plasma being generated by the RF electrode 146.

The control or target signal is “reflected RF”

energy

608 sampled from within the impedance matching network 150. For optimum efficiency, this value is preferably 0. The goal of the impedance matching network is to adjust the capacitor 154, which acts similar to a resistor in this environment, until the target reflected RF level 608 is met. As any of a multitude of conditions change within the process chamber, the capacitor 154 must be continually adjusted. Conventional control of the variable capacitor 154 is discussed below in FIG. 19.

FIG. 19 is a block diagram depicting the conventional plasma generating

control process

600. This example is for a impedance matching network having a series architecture; the pertinent steps will be the same for parallel architecture.

The

process

600 commences with a desired impedance signal 602 being received from the tool logic systems by the impedance control means 604. The impedence control means 604 sums the impedance matching setpoint signal 602 (which is usually 0) with a signal 608 received from the circuitry that represents the reflected RF energy. The summed signals are then manipulated by the impedance matching control algorithm 606 to create a step command signal 610 that is transmitted to the rotor shaft drive means 612. The rotor shaft drive means 612 responsively generates a position control signal 614 that is transmitted 616 to motor means 164, which is connected either directly or through reduction gear means to the rotor shaft assembly 161 and the rotor plates 156.

The motor means 164 (usually a stepper motor) generates motion of the rotor assembly 618. The problem with this system is the same as those other prior systems already described, namely, that it's motor control is open-loop. As such, it cannot distinguish between plasma influences and external electromechanical influences on the mechanical control linkages and system& Because of this, the control accuracy and responsiveness of the system 600 is not at its optimum. What is needed, then, is a plasma generating system having closed-loop motor control of the motor means 164.

SUMMARY OF THE INVENTION

In light of the aforementioned issues and fundamental shortcomings associated with the prior systems and methods, it is an object that the present invention provide a method that allows for greater quality and accuracy of control resulting in both faster times to setpoint and better steady state pressure stability. The preferred invention will rely on an enhanced valve control scheme that integrates a valve position servo control system nested within the conventional pressure control loop. In other words, it is a further object that the pressure control function be accomplished by generating a pressure control signal in terms of valve position. That control signal would in turn be transformed into an actual valve position by a valve/motor drive feedback system. In contrast with prior art systems that make use of open loop motor control, closed loop motor control brings an overwhelming advantage to the pressure control function. One further object is to utilize the higher-resolution addressability of motion that allows for a conventional motor to be directly linked to the valve stem without a geared reducer thus enabling the valve to operate at a faster speed, and to further provide the improved positional precision that is achievable by closed loop operation. It is a still further object that the improved system relieve the pressure control function of the design constraints of low valve speed and limited accuracy of valve positioning. It is also an object to adapt the closed-loop motor control approach to other semiconductor applications, including linear and rotary feedthrough devices and impedance matching networks.

BRIEF DESCRIPTION OF THE DRAWINGS

The objects and features of the present invention, which are believed to be novel, are set forth with particularity in the appended claims The present invention, both as to its organization and manner of operation, together with further objects and advantages, may best be understood by reference to the following description, taken in connection with the accompanying drawings, of which: [0035]
FIG. 1 is a depiction of a conventional semiconductor processing tool system; [0036]
FIG. 2 depicts a conventional valve control assembly; [0037]
FIG. 3 is a flow chart depicting a conventional chamber pressure control process in which only a closed-loop pressure control system is used; [0038]
FIG. 4 is the improved chamber pressure control process of the present invention in which both a closed-loop pressure control system and a closed-loop position control system are used; [0039]
FIG. 5 depicts the improved valve control assembly of the present invention; [0040]
FIG. 6 depicts a semiconductor processing tool system having the embodiment of the present invention of FIGS. 4 and 5 incorporated within it; [0041]
FIG. 7 is a partial schematic of the improved valve control assembly of FIGS. 4 through 6; [0042]
FIGS. 8A and 8B are alternate embodiments of the improved valve control assembly of the present invention; [0043]
FIG. 9 is a graph showing the improved performance demonstrated by the system of the present invention over the prior art; [0044]
FIG. 10 is a graph showing the valve conductance curves for three different species of valves; [0045]
FIG. 11 is a side view of a conventional rotary feedthrough device; [0046]
FIG. 12 is a block diagram depicting the functional description of the pertinent elements of the conventional rotary feedthrough device of FIG. 11; [0047]
FIG. 13 is a block diagram depicting the conventional sample head positioning control process; [0048]
FIG. 14 is a block diagram depicting a closed-loop sample head positioning control process of the present invention; [0049]
FIG. 15 is a block diagram depicting the functional description of the pertinent elements of the closed-loop-controlled rotary feedthrough device of the present invention; [0050]
FIG. 16 is a schematic side view of a conventional plasma processing tool; [0051]
FIG. 17 is a partial perspective view of the conventional variable capacitor used in a conventional automatic impedance matching device; [0052]
FIGS. 18A and 18B are electrical schematics of the pertinent portions of conventional series and parallel impedance matching networks; [0053]
FIG. 19 is a block diagram depicting the conventional plasma generating control process; [0054]
FIG. 20 is a block diagram depicting the closed-loop-controlled plasma generating process of the present invention; and [0055]
FIG. 21 is a block diagram depicting the function description of the pertinent elements of the closed-loop-controlled plasma generating system of the present invention. [0056]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The following description is provided to enable any person skilled in the art to make and use the invention and sets forth the best modes contemplated by the inventors of carrying out their invention. Various modifications, however, will remain readily apparent to those skilled in the art, since the generic principles of the present invention have been defined herein specifically to provide Applications for Closed-loop Motor Control The present invention can best be understood by initial consideration of FIG. 4. [0057]
FIG. 4 is a depiction of the improved chamber [0058] pressure control process 400 of the present invention. Similar to the system displayed in FIG. 3 a position setpoint signal is generated by comparing host tool pressure setpoint si 302 and pressure sensor signal 314 within pressure control means 30. Said position setpoint signal 306 is then transmitted to an improved closed loop motor drive means 58. Therein, summing junction means 59 then sums the position setpoint signal 306 with a motor position feedback signal 406 to generate a position error signal 404. That error signal is operated on by the position control algorithm 402 to produce a motor control signal 310 intended to correct said position error. The signal 310 is then transmitted to the improved motor drive assembly means 60. The improved motor drive assembly means 60 then generates both a valve motion action which is transmitted by the valve stem 54, and a motor position feedback signal 406. The feedback signal 406 is then generated and transmitted by the motor position feedback sensor means 61 to the summing junction means 59 within the improved closed loop valve drive means 58. It should be appreciated that by permitting the valve drive 58 to have direct feedback regarding the position of the valve means 28, there is a substantial improvement in the ability to apply closed loop pressure control methods to drive the assembly means 60. To be more specific, compared with the prior art, the invention is eliminating the effects of coulomb friction, hysteresis and external torques on position accuracy by the proper design of the position control algorithm 402. Next, the effect of backlash such as is characteristic in gear-driven systems are also compensated for. Therefore, the improved valve drive is by its enhanced accuracy of response enabling the design and implementation of a more effective pressure control algorithm 303. Essentially, valve position errors that would previously filter out in the pressure control loop are corrected at the source by the position control loop.
An example of the substantial benefits of this system is provided below in FIG. 9. If we now turn to FIG. 5, we can examine more detail about the improved [0059] valve control assembly 62 of the present invention.
FIG. 5 depicts the improved [0060] valve control assembly 62 of the present invention One substantial distinction is that the improved internal valve command conduit 64 not only comprises a command leg 56 for signals being transmitted from the improved closed-loop motor driver circuit 58 to the improved motor drive 66, but it further includes a feedback leg 57 going in the opposite direction. Furthermore, within the improved motor drive 66 there is found a feedback signal generator means 61 for transmitting these position feedback signals via the feedback leg 57 to summing junction means 59 within the improved closed-loop motor driver circuit 58. The combination of the summing junction means 59, the feedback leg 57 and the feedback signal generator means 61 is referred to as a valve/motor drive feedback system 68. In this example, the improved drive assembly means 60 is shown as having reduction gear means 70 incorporated with it. It should be understood, however, that since the system of the present invention really makes possible positive addressable position of the valve means 28 with a high degree of accuracy, a conventional stepper motor can then be used to directly drive the valve stem 54, without the need for the reduction gear 70. It should further be understood that when we discuss feedback signals emanating from the feedback signal generator means 61, we refer to them as “valve position feedback signals,” to be contrasted with the earlier-described “stroke reference feedback signals;” the difference being that the stroke reference feedback signals are simply endpoint reference signals, whereas the valve position feedback signals of the present invention are signals that indicate the actual positioning of the improved motor drive 66 over the entire range of stroke of the valve.
Extending from the reduction gear means [0061] 70 (if included, as here) is the conventional valve stem 54 to operate the valve means 28 in response to the improved direction/speed signals received by the improved motor drive 66. If we now turn to FIG. 6, we can examine how the system of the present invention would operate as a part of the conventional semiconductor processing tool system
FIG. 6 depicts a semiconductor [0062] processing tool system 10 having the of the present invention of FIGS. 4 and 5 incorporated within it. As shown in FIG. 6, we can see how the improved valve control assembly 62 essentially fits within the system 10 without any modification In fact, since the valve control assembly 62 has an internal motor closed-loop control system, it has been demonstrated that the assembly 62 can be installed in-situ on a valve means 28 that it was not originally designed to operate. If we now turn to FIG. 7, we can see just how this feedback signal is created at its elemental level FIG. 7 is a partial schematic of the improved valve control assembly 62 of FIGS. 4 through 6. Feedback signals are those back EMF pulses that are generated when the rotor 72 of a motor is moved. In this improved control assembly 62, the conventional two-phase stepper motor is slightly modified so that one coil each of the phase A coils 74A and the phase B coils 74B is used to drive the rotor 72 while at the sane time the second coil in the phase A coils 74A and the phase B coils 74B feeds a feedback leg 57A and 57B, respectively. As such, rather than power being applied to the feedback legs 57A and 57B, power is actually drawn off or generated by the movements of the rotor 72. The operation of this back EMF is well explained in U.S. Pat. Nos. 5,134,349, 5,202,613 and 5,321,342. The difference between these prior patents and present invention is that the conventional back EMF motion control has here been used to control a valve stem for pressure control, an application where it has never before been used, and from which unexpected performance results are obtained.
Continuing to describe FIG. 6, the [0063] motor drive 66 then interfaces with the internal valve command conduit 64 and the valve/dive feedback system 68 in order to get inputs from and provide feedback to the closed-loop motor driver circuit 58. It should be understood that the closed-loop motor driver circuit 58 could be provided by the combination of specialty integrated circuit devices and processors, or in its preferred form, it will be incorporated within a digital signal processing device (“DSP”) wherein all of the control and feedback is handled by software. In this way, the internal valve command conduit is actually incorporated within the same housing as the motor driver circuit 58 and the motor drive 66. If we now tun to FIGS. 8A and 8B, we can see how these alternative embodiments might look
FIGS. 8A and 8B are alternate embodiments of the improved [0064] valve control assembly 62 of the present invention. As shown in FIG. 8A, this embodiment of the valve control assembly 62A has a processor device 78 and the ASIC 76 incorporated within a single housing as the closed-loop motor driver circuit58A These are then connected by the cable-type internal valve command conduit 64A to the drive assembly means 60A. In contrast, and as shown in FIG. 8B, this alternative embodiment of the valve control assembly 62B has the closed-loop motor driver circuit 58B and the drive assembly mans 60B incorporated within a single housing 90. It should be understood that the housing 90 might actually be two separate enclosures that are immediately adjacent to one another such that the internal valve command conduit 64B is essentially eliminated. The benefit of eliminating the external cable is that all EMI effects (which are typically prevalent within a conventional semiconductor processing facility) are eliminated in the control scheme of the valve. This further improves the performance of the pressure control system If we now turn to FIG. 9, we can see just how beneficial the results are as compared to the conventional valve control systems.

FIG. 9 is a graph showing an example of improved performance demonstrated by the system of the present invention over the prior art. As can be seen by the valve angle shown at the top half of the chart, the valve with the improved valve control assembly of the present invention demonstrates the steepest response curve in response to a signal It is believed that this is principally related to improvements in valve speed of operation and valve position angular accuracy brought about by the invention. As can be seen from the chamber pressure curve, none of the conventional valve arrangements come as close to the set-point pressure as the valve with the improved valve control assembly of the present invention. In fact, and as shown below in Table I, in this series of experiments the valve with the improved control assembly of the present invention is nearly 11 seconds faster (approximately 15%) than its closest conventional competitor.

TABLE I


RESPONSE TIME COMPARISON

		Valve with
Step #	Setpoint	IVCA	Valve	1	Valve 2

1	9.0 sec	10.7	Setpoint not	13.7
			reached
2	27.2 sec	27.9	31.6	30.5
3	41.1 sec	42.8	Setpoint not	44.4
			reached
4	59.4 sec	62.1	66.3	65.5

Σ[Time(Valve) − Setpoint] =	6.8 sec	Non-	17.4 sec
		computable

Finally, turning to FIG. 10, we can examine a substantial benefit provided by the present invention FIG. 10 depicts the pressure response curves of three conventional species of valves. Each valve species has a different profile for its pressure response to valve movement. In this case, Valve ([0066] 1), a conventional small-size throttling butterfly valve, has a fairy gradual slope over much of its position setting& Since the slope is so gradual, the effective control range extends from nearly zero percent up to approximately fifty percent. This wide of an effective control range is fairly simple for even a conventional motor drive controller. When we look at the steeper response curves of Valve (2) (a conventional medium-sized throttling butterfly valve) and Valve (3) (a conventional large-size throttling butterfly valve or any size sealing throttling valve such as poppet, gate or pendulum types), we can see that the effective control ranges are much smaller than for Valve (1). These narrow control ranges mean that the highest resolution valve positioning is necessary; if there is not enough granularity in the valve positioning system, the motor drive will simply not be able to control at a setpoint, but will instead oscillate above and below the desired pressure. In the closed-loop valve control assembly of the present invention, an effective resolution ranging from 100,000 to 8,000,000 motor positions (from 0% to 100% valve position) has been demonstrated; this is sufficient to provide good pressure control performance even in the steepest valve response curves. In contrast, the conventional open-loop valve control assemblies cannot actually tell where the valve is positioned, but only where it should be positioned. As a result of the effects of friction, backlash, and other previously-described effects, the resulting valve positioning error makes using high resolution control ineffective (since the small angular steps many times will be inadequate to overcome the positioning error). Consequently, the conventional valve control assembly will typically only provide in the range of 1,600 to 12,000 motor steps between 0% and 100% valve position Since there is such a low resolution, these prior open-loop valve control assemblies may not even be capable of effectively operating a valve having the profile of Valve (3).
It should further be understood that while all of the previous examples provided herein have involved the operation of a valve located downstream of the process chamber to control the pressure in the process chamber (“downstream pressure control”), that other configurations are certainly included within the present method and system Namely, the use of a closed-loop valve control assembly located upstream of the process chamber to control the pressure within the chamber (“upstream pressure control”). Furthermore, the method and system of the present invention could be applied in combination with a valve and the signal from a fluid flow meter (in contrast to the s from a pressure sensor) in order to regulate fluid flow (i.e. gas or liquid) in a conduit; again, the same improved results are expected It should also be understood that improvements to valve position control speed and accuracy can also be realized by the use of feedback mechanisms and methods other than back EMF pulses Examples of such methods may include, but are not limited to, the use of potentiometers and motor encoders. The degree to which these alternative methods are effective for improving valve actuation performance may depend on the resolution with which these feedback mechanisms can be employed [0067]
Other desirable applications for closed-loop motor control are automated feedthrough devices and impedance matching networks FIG. 14 provides additional detail in this regard. [0068]
FIG. 14 is a block diagram depicting a closed-loop sample head [0069] positioning control process 514 of the present invention The user/tool position signal 502 is received by the sample head position control means 30A, wherein discrete (i.e. X-axis, Y-axis, Z-axis and Rotational as appropriate) command signals 506 are generated 504. The command signals 506 are transmitted or passed to the closed-loop sample head drive means 128 for controlling the feedback-generating sample head motor means 130. The position control algorithm 509 is applied to the discrete position command signals 506 and the discrete position feedback signals 516; the result is discrete (again, X-Y-Z- and R-specific) direction/speed command signals 510. Each discrete 510 then drives a distinct sample head motor means (see FIG. 16); the group of motor means' referred to as feedback-generating sample head motor means 130. Attached to, incorporated within or otherwise associated with the motor means 130 is the feedback signal generator means 132 for generating a signal indicating the rotational position of the motor means; this is the discrete (X-, Y-, Z- and R-specific) position feedback signal 516. As discussed in the original parent (valve control) application, this may be “back EMF” feedback, wherein a portion of the coils of the electric motor are dedicated to actually generating a voltage; the characteristics of the voltage signal very clearly constitute a feedback signal on the physical position of the electric motor. Again, other approaches may be used to provide this motor feedback signal 516, including potentiometers and motor encoders. The discrete position feedback signals 516 are summed with the discrete command position command system 506 through the position control algorithm 509 to provide error-correction to the discrete direction/speed command 510. Adding closed loop motor control to what is otherwise an open-loop system without feedback is expected to substantially improve responsiveness and accuracy over the conventional rotary or linear feedthrough device. FIG. 15 provides information regarding the structure of this embodiment.
FIG. 15 is a block diagram depicting the functional description of the pertinent elements of the closed-loop-controlled [0070] rotary feedthrough device 136 of the present invention. As a preliminary matter, it should be understood that this invention simply applies the previously-described motor-operated valve control to the linear or rotary feedthrough application. The functional relationships between the operating components, as it applies to the control of the motor (and sample head or valve) position is the same.
The sample head position control means [0071] 30A receives its input signals for position from the user interface or the tool itself (i.e. tool logic generated). The sample head position control means 30A applies the discrete position control algorithm(s) to the incoming signal and transmits them to the sample head motor and feedback system 131 (this is analogous to the improved closed loop valve drive means 58). The system 131 comprises discrete pairs of closed-loop motor drive means having summing junction means 134 incorporated therein and feedback generating motor means having feedback signal generator means 132 incorporated therein. The group of discrete motor drive means' is collectively referred to as the closed-loop sample head drive means 128; the group of discrete feedback-generating motor means is collectively referred to as the feedback generating sample head motor means 130.
The output of the motor means [0072] 130 is movement of the sample head 106 via operation of the shaft 110.
Regarding an embodiment of the present invention related to impedance matching networks, FIG. 20 is a block diagram depicting the closed-loop-controlled [0073] plasma generating process 603 of the present invention. The impedance matching setpoint s 602 is received from the tool logic; it is summed with the actual plasma impedance 608 within the impedance control algorithm 606 of the impedance control means 604, thereby generating the step command signal 610. Thus far, the process is identical to the prior art, however, the conventional open-loop motor-controlled variable capacitor(s) have now been replaced with closed-loop rotor linkage drive means 620 and feedback-generating rotor assembly motor means 623 operating the variable capacitor(s) 154. Within the drive means 620, the step command signal 610 is summed with the position feedback sign 626 and operated on by the position control algorithm 622 to create the direction/speed command signal 616 for operation by the feedback-generating motor means 623 (which may include a conventional motor means 164). The feedback-generating motor means 623 then generates the motion within the rotor assembly 618 which is connected to the rotor plates 156 by the rotor linkage means 161. When moving, the feedback signal generator means 624 transmits position feedback signal(s) 626 to the drive means 620 for providing error correction to the direction/speed command signal 616.
FIG. 21 depicts the functional description of the pertinent elements of this closed-loop-controlled [0074] plasma generating system 180. The arrangement of the elements is generically identical to the previous two application its. Here, the closed loop motor control system is the rotor assembly motor drive and feedback system 650, which comprises the rotor linkage drive means 620 and the feedback-generating rotor assembly motor means 623.
Those skilled in the art will appreciate that various adaptations and modifications of the above-described preferred embodiment can be configured without departing from the scope and spirit of the invention Therefore, it is to be understood that, within the scope of the appended claims, the invention may be practiced other than as specifically described herein. [0075]

Claims

What is claimed is:

1. A feedthrough device, comprising:

a sample head;

a sample head motor means operatively connected to position said sample head, said sample head motor means further comprising a feedback signal generator for generating motor position feedback signals; and

a sample head drive means in communication with said sample head motor means, said drive means configured to generate discrete position commands for said motor means and said sample head drive means further configured to receive said motor position feedback signals.

2. The device of claim 1, further comprising an internal sample head command conduit for carrying said commands and said signals between said sample head motor means and said sample head drive means.

3. The device of claim 2, wherein said sample head drive means further comprises a summing junction for summing said position feedback signals and said position commands and responsively generating resultant position commands.

4. The device of claim 3, wherein said summing junction, said internal sample head command conduit and said feedback signal generator comprise a sample head motor drive and feedback system.

5. The device of claim 4, wherein said sample head motor means comprises a X-motor means, said X-motor means operatively connected to position said sample head along an X-axis, said X-motor means further comprising an X-axis feedback signal generator for generating X-motor means position feedback signals.

6. The device of claim 4, wherein said sample head motor means comprises a Y-motor means, said Y-motor means operatively connected to position said sample head along a Y-axis, said Y-motor means further comprising a Y-axis feedback signal generator for generating Y-motor means position feedback signals.

7. The device of claim 4, wherein said sample head motor means comprises a Z-motor means, said Z-motor means operatively connected to position said sample head along an Z-axis, said Z-motor means further comprising a Z-axis feedback signal generator for generating Z-motor means position feedback signals.

8. The device of claim 4, wherein said sample head motor means comprises an R-motor means, said R-motor means operatively connected to rotationally position said sample head around a primary axis, said R-motor means further comprising an R-axis feedback signal generator for generating R-motor means position feedback signals.

9. The device of claim 5, wherein said sample head motor means comprises a Y-motor means, said Y-motor means operatively connected to position said sample head along a Y-axis, said Y-motor means further comprising a Y-axis feedback signal generator for generating Y-motor means position feedback signals.

10. The device of claim 9, wherein said sample head motor means comprises a Z-motor means, said Z-motor means operatively connected to position said sample head along an Z-axis, said Z-motor means further comprising a Z-axis feedback signal generator for generating Z-motor means position feedback signals.

11. The device of claim 10, wherein said sample head motor means comprises an R-motor means, said R-motor means operatively connected to rotationally position said sample head around a primary axis, said R-motor means further comprising an R-axis feedback signal generator for generating R-motor means position feedback signals.

12. An impedance matching network for controlling generated RF power, said power usedby an RF electrode to generate and control a plasma field in a process chamber, said network comprising a rotor linkage means for actuating a rotor, the impedance matching network comprising:

a feedback generating rotor assembly motor means for actuating said rotor linkage means, said motor means further comprising a position feedback signal generator means;

a rotor shaft drive means in communication with said motor means, said rotor shaft drive means configured to send position command signals to said motor means and further to receive signals generated by said position feedback signal generator means; and

a motor means/drive means feedback system interconnecting said motor means and said rotor shaft drive means.

13. The network of claim 12, wherein said motor means further comprises a motor drive attached to said rotor linkage means.

14. The network of claim 13 wherein said motor means further comprises reduction gear means operatively attached between said motor drive and said rotor linkage means.