DETAILED DESCRIPTION OF THE INVENTION

[0061] This application contains subject matter that is related to U.S. patent application Ser. No. 10/131,603, entitled SYSTEM AND METHOD FOR A MASS FLOW CONTROLLER, filed Apr. 24, 2002, which is herein incorporated by reference in its entirety.

[0062] Typically a fluid flow path exists in a pressure environment. The pressure environment may include the pressure at the inlet side of the flow path (referred to as inlet pressure), and pressure at the outlet side of the valve (referred to as outlet pressure), and other pressures within the environment. For example, the pressure environment of the flow path may also include pressure differentials such as, for example, the pressure drop across a bypass or across a valve. The pressure environment may also include various pressure transients including pulses introduced by a regulator, turbulence caused by the geometry of a flow sensor, or various other pressure perturbations. However, the pressure environment is not often monitored. As such, performance of mass flow controller may be vulnerable to pressure transients in a flow path to which the controller is coupled for the purpose of controlling the fluid flow.

[0063] According to one aspect of the present invention, Applicants have recognized that measurements of the pressure environment of a flow path may be used to reduce or eliminate performance degradations, instabilities, and/or inaccuracies of a mass flow controller caused by changes in the pressure environment. As such, Applicants have developed various methods for compensating for pressure transients in the pressure environment of a flow path and mass flow controller.

[0064] As discussed in the foregoing, a mass flow controller typically includes a flow meter that senses the fluid flow in a fluid flow path. The flow rate sensed by the flow meter is often part of a feedback control loop that controls the flow rate of a fluid being provided to a process (e.g., a semiconductor fabrication process) at the outlet side of the flow path.

[0065] In many cases, the actual flow rate provided to the process must be accurately controlled. However, the pressure transients may cause local fluctuations in the fluid flow that are sensed by the flow meter. These local fluctuations may not be an accurate indication of the actual flow rate being provided to the process. This false flow information is then provided to the control loop of a mass flow controller. The controller may then adjust the flow rate provided to the process in response to the false flow information. As such, the controller may momentarily lose control of the process and/or provide undesired flow rates to the process.

[0066] As used herein, the term false flow refers to fluid flow that does not correspond to the actual flow provided to a process. For example, local variations or fluctuations in fluid flow that are not substantially experienced at the outlet side of a flow path is considered false flow. As such, false flow information generally describes an indication of flow that does not correspond to the flow rate being provided to a process.

[0067] According to one embodiment of the invention, it is appreciated that may be advantageous to measure the pressure of the flow path (e.g., the inlet pressure) and provide a control system that incorporates this information. More particularly, to reduce the performance degradation due to pressure transients, it may be desirable to measure the pressure in the flow path and adjust control parameters of a mass flow controller in response to the changes in pressure.

[0068] One embodiment of the present invention includes measuring the inlet pressure of a flow path and providing the inlet pressure measurement to the mass flow controller. For instance, a pressure transducer may be coupled to the flow path to provide a pressure signal indicative of the inlet pressure of the flow path.

[0069] Applicants have recognized and appreciated that by providing a pressure signal to a mass flow controller, various deficiencies in the conventional operation of the mass flow controller can be addressed. Accordingly, applicants have identified various methods of utilizing a pressure signal to improve the performance and accuracy of a mass flow controller. One method according to one embodiment of the present invention includes compensating for spurious flow signals that may occur due to pressure transients in a fluid path coupled to a mass flow controller.

[0070] One problem associated with pressure transients in a flow path that may have deleterious effects on a mass flow controller is described below. When the pressure in a flow path changes, fluid accelerates down the pressure gradient in order to fill the volume, referred to as dead volume, created by the change in pressure. A sensor of a flow meter may register an increase in fluid flow due to this local acceleration of the fluid into the dead volume. However, this is considered a false flow of fluid because this flow is not indicative of the flow being provided to the process. As such, the sensor output signal from the sensor carries false flow information that is propagated to the control loop of the mass flow controller. As discussed above, this false flow information may have undesirable consequences with respect to the accuracy and performance of the mass flow controller.

[0071] FIG. 12A illustrates a case in which a pressure transient in the shape of a pressure pulse is introduced at an inlet side of a flow path. Graph 1200 a shows a plot of a pressure pulse 1210 as a function of time. Pressure pulse 1210 is introduced to the inlet side of the flow path. As a result, the flow sensor responds with sensor output signal 1220 as shown in graph 1200 b of FIG. 12B . Assuming that the actual flow (i.e., the flow being delivered to the process) of the flow path has not changed, the spike in the sensor output signal contains a large false flow component with respect to the actual flow. As such, the mass flow controller reacts to the flow spike accordingly and may momentarily lose control of the process.

[0072] According to one embodiment of the present invention, applicants have recognized that pressure measurements in the flow path may be utilized to anticipate false flow indications and compensate for the negative impact they may incur on a mass flow controller. One embodiment of the present invention includes a method for controlling flow including measuring the pressure in a fluid flow path and adjusting an output signal provided by a flow sensor coupled to the flow path based on the pressure measurements.

[0073] By analyzing a flow sensor's response fluid flow fluctuations caused by pressure transients, Applicants have developed methods for generating a false flow signal that recreates the false flow component of a flow signal provided by a flow sensor in response to a pressure transient. This generated false flow signal may be used by a system to compensate for the spike in flow sensor output. For instance, this signal may be used by a compensator to reduce induced value drive motion resulting from the spike in output.

[0074] FIGS. 11 A-D illustrate one embodiment of the invention for generating a false flow signal from a pressure signal indicative of the pressure measured in a fluid flow path. The false flow signal can then be subtracted from the flow signal provided by the flow sensor to produce an indicated flow signal that does not include the false flow information. As such, the false flow information is prevented from causing the controller to respond erroneously (e.g., by providing unwanted flow to the process).

[0075] Graph 1100 a shown in FIG. 11A illustrates a pressure transient, and in particular, a pressure pulse 1110 , that a flow path may experience during operation. Graph 1100 b of FIG. 11B shows a pressure signal 1120 resulting from pressure pulse 1110 . The pressure signal may be measured by a pressure measurement device (e.g., a pressure transducer) coupled to the flow path and adapted to measure the pressure at some portion of the flow path.

[0076] FIG. 11C shows a system 1100 c having a flow path 200 with a flow sensor 1140 being coupled to flow path 200 to sense fluid flow in flow path 200 . As shown, pressure pulse 1100 is introduced to the flow path 200 (e.g. by non-ideal performance of an upstream regulator) at the inlet of flow path 200 . Pressure pulse 1100 may cause a local fluctuation in the fluid flow sensed by flow sensor 1140 . Flow sensor 1140 , in turn, produces a sensor output signal 1150 that is corrupted with false flow information.

[0077] According to one aspect of the present invention, a compensation filter is provided that compensates for the false flow information. In one embodiment of the invention, as shown in FIG. 11D, a compensation filter 1180 is provided that receives the pressure signal 1120 produced by transducer 295 and produces a false flow signal 1160 . Because filter 1180 receives pressure information indicative of the pressure in a portion of the flow path, compensation filter 1180 can predict the response the flow sensor will have to the fluid flow fluctuations resulting from the pressure transient.

[0078] As such, filter 1180 may construct a false flow signal that closely resembles the false flow information produced by the sensor. More particularly, filter 1180 recreates the false flow information produced by flow sensor 1140 and provides this information as false flow signal 1160 . False flow signal 1160 can then be subtracted from the sensor output signal 1150 to effectively remove the effects of pressure pulse 1100 . In one embodiment, a false flow signal includes a transfer function that emulates the behavior of the flow sensor in response to flow fluctuations caused by pressure transients.

[0079] According to one embodiment of the present invention, a compensation filter 800 is provided that emulates the behavior of the flow sensor. More particularly, FIG. 8 shows a compensation filter 800 that includes a time delay block 810 , a differentiator 820 , a series-connected bank of 2 ^nd -order filters 830 a - f (collectively, item 830 ), and an adder 840 . Compensation filter 800 receives pressure signal 1120 and provides it to time delay block 810 . Time delay block 810 delays the pressure signal such that the delayed output signal is aligned in time with a sensor output signal (not shown). In particular, some finite amount of time elapses between a pressure transient and when a flow sensor responds to the pressure transient (i.e., there is a delay between a pressure pulse and when the false flow information appears on the sensor output signal). As such, the pressure signal may be delayed such that the generated false flow signal is subtracted off the proper portion of the sensor output signal.

[0080] Delay block 810 provides a delayed pressure signal 815 to differentiator 820 which calculates a derivative of the delayed pressure signal 815 and provides a derivative signal 825 to a series of second-order filters 830 . The derivative of the delayed pressure signal is calculated because the false flow resulting from a pressure transient is proportional to the pressure gradient resulting from a pressure transient. In addition, the derivative of the delayed pressure signal ensures that a constant pressure results in a zero false flow signal. That is, when the pressure signal is constant, the compensation filter has no effect on the sensor output signal.

[0081] Derivative signal 825 is provided to the first second order filter 830 a in the series of filters 830 . The output of each second order filter is provided as the input to the next second order filter in series 830 . In addition, the output from each second-order filter is tapped off and provided to a respective gain block 850 a - 850 f that scales the respective output of each filter by a respective constant gain factor K _n .

[0082] Each of the scaled outputs from the individual 2 ^nd -order filters contributes to the construction of false flow signal 1160 . Adder 840 sums the contributions of the scaled outputs and provides the false flow signal 1160 . In one embodiment, false flow signal 1160 is a recreation of the false flow information provided by a flow sensor in response to pressure transients. As such, false flow signal 1160 may be subtracted from the sensor output to compensate for this false flow information.

[0083] It should be appreciated that the number of filters and type of filters illustrated in FIG. 8 is not limiting. Indeed, any filter configuration of any order and arrangement may be used to provide a false flow signal. The configuration illustrated in FIG. 8 has been shown to provide sufficient control over characteristics of the false flow signal that Applicants have found useful such as dead time, rise time, overshoot and parabolic attributes such that a false flow signal closely resembling the false flow information superimposed on the sensor output signal in response to a pressure transient may be recreated. However, other filter designs and arrangements that will occur to those skilled in the art may be applicable and are considered to be within the scope of the present invention. For instance, the order of several components may be different (e.g., delay block 810 , differentiator 820 ), and/or one or more of these blocks may be eliminated altogether.

[0084] In one example, design of one embodiment of the filters shown in FIG. 8 are described in more detail below. A generalized second order transfer function of the second order filters can be represented as:

K /( s ² +2ξω _n s+ω _n ² ) (1)

[0085] Where:

[0086] K=Gain

[0087] s=Laplace Operator

[0088] ω _n =Natural Frequency

[0089] ξ=Damping Factor

[0090] Scaling factors may be added such that each filter can be tailored independent of each other. As such, the filter bank 830 may be optimized to provide a different response in terms of “height” (gain), “width” (frequency response) and over/undershoot (damping) such that the shape of the constructed false flow signal can be “dialed” and by changing the scaling factors ξ, ω, and δ.

[0091] One exemplary specific transfer function can be represented as:

Kδω _n ² /( s ² +ξδω _n s+δ ² ω _n ² ) (2)

[0092] The K term in the transfer function is illustrated as a constant gain factor K _n . As such, the output from each second order filter is multiplied by K _n and provided to adder 840 . Adder 840 sums the contributions from each filter to provide false flow signal 1160 . False flow signal 1160 is subtracted from the sensor output signal to provide an indicated flow signal. As such, the false flow information superimposed on the flow signal due to pressure transients is subtracted off by the constructed false flow signal leaving a flow signal indicative of the actual flow supplied to the process at the outlet side of the flow path.

[0093] Mass flow controllers are often vulnerable to instability due to factors including non-linearities in the various components of the mass flow controller dependencies on various operating conditions of a mass flow controller, or other factors. The term operating condition applies generally to any of various conditions that can be controlled and that may influence the operation of a mass flow controller. In particular, operating conditions apply to various external conditions that can be controlled independent of a particular mass flow controller. Exemplary operating conditions include, but are not limited to, fluid species, set point or flow rate, inlet and/or outlet pressure, temperature, etc.

[0094] However, it should be appreciated that other internal conditions may be present during the operation of a mass flow controller such as signal characteristics, system noise, or perturbations that cannot be controlled independent of a particular flow controller. In particular, various signals employed by the mass flow controller may have frequency components containing many different frequencies. However, the frequency composition of a signal is inherent to the signal and is not considered to be controllable independent of a particular mass flow controller. Accordingly, such conditions, unless specifically stated otherwise, are not considered to be encompassed within the term operating conditions as used herein.

[0095] The term mass flow rate, fluid flow, and flow rate is used interchangeably herein to describe the amount of fluid flowing through a unit volume of a flow path (e.g. flow path 103 of FIG. 1 ), or a portion of the flow path, per unit time (i.e., fluid mass flux).

[0096] The term species applies generally to the properties of a specific instance of a fluid. A change in species applies to a change in at least one property of a fluid that may change or affect the performance of a mass flow controller. For example, a change in species may include a change in fluid type (e.g., from nitrogen to hydrogen), a change in the composition of a fluid (e.g., if the fluid is a combination of gases or liquids, etc.), and/or a change in the state of the fluid or combination of fluids. The term species information applies generally to any number of properties that define a particular fluid species. For example, species information may include, but is not limited to, fluid type (e.g. hydrogen, nitrogen, etc.), fluid composition (e.g., hydrogen and nitrogen), molecular weight, specific heat, state (e.g., liquid, gas, etc.), viscosity, etc.

[0097] Often a mass flow controller comprises several different components (i.e., a flow sensor, feedback controller, valve etc.) coupled together in a control loop. Each component that is part of the control loop may have an associated gain. In general, the term gain refers to the relationship between an input and an output of a particular component or group of components. For instance, a gain may represent a ratio of a change in output to a change in input. A gain may be a function of one or more variables, for example, one or more operating conditions and/or characteristics of a mass flow controller (e.g., flow rate, inlet and/or outlet pressure, temperature, valve displacement, etc.) In general, such a gain function is referred to herein as a gain term. A gain term, and more particularly, the representation of a gain term may be a curve, a sample of a function, discrete data points, point pairs, a constant, etc.

[0098] Each of the various components or group of components of a mass flow controller may have an associated gain term. A component having no appreciable gain term can be considered as having a unity gain term. Relationships between gain terms associated with the various components of a mass flow controller is often complex. For example, the different gain terms may be functions of different variables (i.e., operating conditions and/or characteristics of the components), may be in part non-linear, and may be disproportionate with respect to one another.

[0099] Accordingly, the contributions of each gain term associated with the components around a control loop of a mass flow controller is itself a gain term. This composite gain term may itself be a function of one or more variables and may contribute, at least in part, to the sensitivity of the mass flow controller with respect to change in operating conditions and/or characteristics of the various components of the mass flow controller. According to one embodiment of the present invention, a mass flow controller is provided having a control loop with a constant loop gain. According to one embodiment, the constant loop gain is provided by determining a reciprocal gain term by forming the reciprocal of the product of the gain terms associated with one or more components in the control loop of the mass flow controller and applying the reciprocal gain term to the control loop. According to one embodiment, the pressure signal is used to adjust the gain in the mass flow controller (e.g., in a GLL controller associated with the mass flow controller) to provide a constant gain.

[0100] A constant loop gain as used herein describes a gain of a control loop of a mass flow controller that remains substantially constant with respect to one or more operating conditions of the mass flow controller. In particular, a constant loop gain does not vary as a function of specific operating conditions associated with a mass flow controller, or as a function of the individual gain terms associated with the control loop. It should be appreciated that a constant loop gain may not be precisely constant. Imprecision in measurements, computation and calculations may cause the constant loop gain to vary. However, such variation should be considered encompassed by the definition of a constant loop gain as used herein. Further, a constant loop gain may not necessarily be constant over all operating ranges or conditions. However, one benefit of having a constant loop gain over operating conditions includes the mass flow controller being able to operate (and be tuned and calibrated) for one fluid and not need to be tuned and/or calibrated for other fluids and/or operating conditions.

[0101] It should further be appreciated that the gain of certain components of the mass flow control may vary with operating frequency, and that signals of the mass flow controller may have frequency components at many different frequencies. However, frequency is not considered an operating condition, and as such, is not considered as a condition over which a constant loop gain remains constant.

[0102] Following below are more detailed descriptions of various concepts related to, and embodiments of, methods and apparatus according to the present invention for control and configuration of a mass flow controller. Such a flow controller with which various aspects may be implemented is described with particularity in U.S. patent application Ser. No. 10/131,603, entitled SYSTEM AND METHOD FOR A MASS FLOW CONTROLLER, filed Apr. 24, 2002, incorporated by reference herein in its entirety. Although various aspects of the present invention may be implemented in the mass flow controller described therein, it should be appreciated that any mass flow controller may be used, and the invention is not limited to being implemented in any particular mass flow controller.

[0103] It should also be appreciated that various aspects of the invention, as discussed above and outlined further below, may be implemented in any of numerous ways, as the invention is not limited to any particular implementation. Examples of specific implementations are provided for illustrative purposes only.

[0104] In the following description, various aspects and features of the present invention will be described. The various aspects and features are discussed separately for clarity. One skilled in the art will appreciate that the features may be selectively combined in a mass flow controller depending on the particular application.

[0105] A. Control of a Mass Flow Controller

[0106] FIG. 1 illustrates a schematic block diagram of a mass flow controller according to one embodiment of the present invention. The mass flow controller illustrated in FIG. 1 includes a flow meter 110 , a Gain/Lead/Lag (GLL) controller 150 , a valve actuator 160 , and a valve 170 .

[0107] The flow meter 110 is coupled to a flow path 103 . The flow meter 110 senses the flow rate of a fluid in the flow path, or portion of the flow path, and provides a flow signal FS 2 indicative of the sensed flow rate. The flow signal FS 2 is provided to a first input of GLL controller 150 .

[0108] In addition, GLL controller 150 includes a second input to receive a set point signal S 12 . A set point refers to an indication of the desired fluid flow to be provided by the mass flow controller 100 . As shown in FIG. 1 , the set point signal SI 2 may first be passed through a slew rate limiter or filter 130 prior to being provided to the GLL controller 150 . Filter 130 serves to limit instantaneous changes in the set point in signal SI 1 from being provided directly to the GLL controller 150 , such that changes in the flow take place over a specified period of time. It should be appreciated that the use of a slew rate limiter or filter 130 is not necessary to practice the invention, and may be omitted in certain embodiments of the present invention, and that any of a variety of signals capable of providing indication of the desired fluid flow is considered a suitable set point signal. The term set point, without reference to a particular signal, describes a value that represents a desired fluid flow. Based in part on the flow signal FS 2 and the set point signal SI 2 , the GLL controller 150 provides a drive signal DS to the valve actuator 160 that controls the valve 170 . The valve 170 is typically positioned downstream from the flow meter 110 and permits a certain mass flow rate depending, at least in part, upon the displacement of a controlled portion of the valve. The controlled portion of the valve may be a moveable plunger placed across a cross-section of the flow path, as described in more detail with respect to FIG. 16 . The valve controls the flow rate in the fluid path by increasing or decreasing the area of an opening in the cross section where fluid is permitted to flow. Typically, mass flow rate is controlled by mechanically displacing the controlled portion of the valve by a desired amount. The term displacement is used generally to describe the variable of a valve on which mass flow rate is, at least in part, dependent. As such, the area of the opening in the cross section is related to the displacement of the controlled portion, referred to generally as valve displacement.

[0109] The displacement of the valve is often controlled by a valve actuator, such as a solenoid actuator, a piezoelectric actuator, a stepper actuator etc. In FIG. 1 , valve actuator 160 is a solenoid type actuator, however, the present invention is not so limited, as other alternative types of valve actuators may be used. The valve actuator 160 receives drive signal DS from the controller and converts the signal DS into a mechanical displacement of the controlled portion of the valve. Ideally, valve displacement is purely a function of the drive signal. However, in practice, there may be other variables that affect the position of the controlled portion of the valve.

[0110] For example, in the valve illustrated in FIG. 10, a pressure differential between the backside of the plunger 1026 and the face of the plunger 1025 , over the jet orifice 1040 and plateau 1050 attempts to force the plunger towards the jet. The plunger face over the orifice experiences a pressure substantially equal to the outlet pressure of the flow path. From the edge of the orifice 1045 to the outer edge of the plateau 1055 , the plunger face experiences a pressure gradient, with pressure at the outer edge of the plateau substantially equal to the inlet pressure less any pressure drop through the sensor bypass. The remainder of the plunger, including the backside, experiences a pressure substantially equal to the inlet pressure less any pressure drop through the sensor bypass. Accordingly, the plunger 1020 will experience a pressure dependent force that can be expressed as: Force=(P ₁ −P ₀ )*A, where, P ₁ is equal to the inlet pressure, P ₀ is equal to the outlet pressure, and A is equal to the effective area of the plunger. The effective area of the plunger may change from valve to valve is typically within the range of the area of the orifice and the area of the orifice plus the plateau.

[0111] As such, when the valve experiences a pressure transient, this force changes and the plunger may undergo undesirable displacement. That is the plunger may be displaced by some amount different than the valve displacement that is desired by the control loop. This undesirable displacement may provide a fluid flow to the process having a component that is unintended. In addition, this undesired displacement may cause the control loop to oscillate as described below.

[0112] However, if pressure transients that may cause undesirable movement of the controlled portion of the valve can be detected, then the drive signal applied to the valve actuator can be adjusted to compensate for this undesired valve displacement. Stated differently, the drive signal may be adjusted such that it has a component indicative of the drive level necessary to keep the plunger stationary under a detected pressure transient.

[0113] Accordingly, one embodiment according to the present invention includes determining a displacement compensation signal from a pressure measurement, wherein the displacement compensation signal is the drive level necessary to prevent the plunger from moving due to pressure transients. The displacement compensation signal is then added to the valve drive signal. As such, the valve drive signal applied to the valve has a component indicating the valve displacement desired by the control loop of the mass flow controller and a component indicating the drive level necessary to hold the plunger steady in the pressure environment recorded by the pressure measurements.

[0114] The term pressure environment refers generally to various pressures that a valve experiences. As the different portions of the valve may “see” different pressures and at different times, the term pressure environment is meant to refer to the entire set of pressures that may affect a force on the valve. Similarly, a valve environment refers to the set of forces that act on the valve and may include pressures, magnetic forces, spring forces, mechanical forces etc., as described in further detail below.

[0115] One embodiment according to the present invention involves using a force model of the valve to predict the pressure induced valve displacement from a pressure signal indicative of at least one pressure measurement in the valve environment.

[0116] FIG. 9 illustrates one method of pressure induced valve displacement compensation. FIG. 9 illustrates the outlet side of a flow path 200 . Valve 170 is coupled to the flow path to control the fluid flow through the outlet to a process. Valve actuator 160 controls the displacement of the valve depending on the drive level indicated by drive signal DS′. For example, valve and valve actuator pair 170 and 160 may be the same as that described in connection with FIG. 1 .

[0117] In addition, a pressure transducer 295 ′ is coupled to the flow path. The pressure transducer measures at least one pressure in the valve environment. The pressure transducer 295 ′ provides at least one pressure signal indicative of at least one pressure in the valve environment (e.g., inlet pressure, outlet pressure, etc.). For the purpose of this example, the pressure transducer measures the inlet pressure of the flow path and provides pressure signal 270 ″ indicating the inlet pressure. While the pressure transducer is illustrated as being upstream from the valve, it should be appreciated that it may be placed downstream of the valve. In addition, more than one pressure transducer may be disposed along the flow path in order to measure any desirable pressure in the valve environment and to output an associated pressure signal indicative of the pressure measurement.

[0118] Pressure signal 270 ″ is provided to displacement compensation block 920 . displacement compensation determines a drive level sufficient to substantially counter-act a pressure induced displacement effected on the valve by the pressure environment indicated by pressure signal 270 ″. Displacement compensation block 920 provides displacement compensation signal to summation block 950 . At summation block 950 , the displacement compensation signal is added to the drive signal DS issued from a controller 150 . For example, controller 150 may be a GLL controller as illustrated in FIG. 1 . The summed drive signal DS′ is then provided to the valve actuator which mechanically displaces the controlled portion of the valve according to drive signal DS′.

[0119] As such, drive signal DS′ has a component that effectively zeroes out the force effect the pressure environment has on the valve displacement and a component provided by the control loop. As such, the net valve displacement resulting from the valve environment is the displacement desired by the control loop of the mass flow controller.

[0120] In one embodiment of displacement compensation, a force model of a valve is used in order to determined the pressure induced displacement of the valve in a pressure environment. FIG. 13 is similar to FIG. 9 , however, the displacement compensation 920 ′ includes a force model 1300 that models the forces in the valve environment. On suitable force model for a valve operating with a free floating plunger is described in the section E. entitled “Force Valve Model.”

[0121] Many different force models may be formulated to predict pressure induced valve displacement in a pressure environment. Force models may vary with respect to the type of valve and conditions under which the valve is intended to operate. The invention is not limited to any particular force model.

[0122] As discussed above, the various components of the mass flow controller may have a gain term associated with the operation thereof. For example, FIG. 1 illustrates gain terms A, B, C and D associated with the flow meter 110 , the GLL controller 150 , the valve actuator 160 , and valve 170 , respectively. These components and their associated input and output signals, in particular, flow signal FS 2 , drive signal DS, valve signal AD, and the fluid flowing in the flow path 103 , form a control loop of the mass flow controller. The gains A, B, C, and D, in turn, are associated with the relationship between said inputs and outputs. It should be appreciated that the gain terms around this control loop contribute to a composite control loop gain.

[0123] Typically, this control loop gain term is the product of the gain terms around the control loop (i.e., the control loop gain term is equal to the product A*B*C*D). As used herein, a composite gain term describes any gain term comprising the contributions of a plurality of individual gain terms. The notation for a composite gain term used herein will be appear as the concatenation of the symbols used to represent the individual gain terms contributing to the composite gain term. For example, the control loop gain term describe above will be represented as gain term ABCD. Unless otherwise noted, the notation described above for a composite gain term is assumed to be the product of its constituent gain terms.

[0124] The individual gain terms associated with a control loop of a mass flow controller may have differing characteristics and dependencies resulting in a composite gain term that may have multiple dependencies. These dependencies or variables may include set point or flow rate, fluid species, temperature, inlet and/or outlet pressure, valve displacement, etc. Applicants have recognized and appreciated that a mass flow controller having an arbitrary control loop gain term may be vulnerable to instability and may be sensitive to changes in some or all of the dependencies mentioned above. Below is a description of each of the exemplary gain terms illustrated in FIG. 1 .

[0125] Gain term A is associated with the flow meter and represents the relationship between the actual fluid flow through the mass flow controller and the indicated flow (e.g., FS 2 ) of the flow meter (e.g., change in indicated flow divided by change in actual fluid flow). Gain term A is calibrated to be a constant function of at least flow rate. However, this constant may depend at least upon the fluid species with which the mass flow controller operates.

[0126] Gain term B is associated with the GLL controller and represents the relationship between the indicated flow signal FS 2 received from the flow meter and the drive signal DS provided to the valve actuator. Gain term B is related to the various gains and constants used in the feedback control of the GLL controller.

[0127] Gain term C is associated with the valve actuator and represents the relationship between a drive signal and the displacement of the valve. Gain C may include the combination of two separate gains including the gain associated with the conversion of a drive signal to an electrical current or voltage control signal, and the gain associated with the control signal and the mechanical displacement of the controlled portion of the valve.

[0128] Gain term D is associated with the valve and represents the relationship between a flow rate of the mass flow controller and valve displacement (e.g., a change in flow rate divided by a change in valve displacement.) Gain term D may be dependent on a variety of operating conditions including fluid species, inlet and outlet pressure, temperature, valve displacement, etc. According to one aspect of the present invention described in more detail below, a physical model of a valve is provided that facilitates the determination of a gain term associated with the valve with arbitrary fluids and operating conditions.

[0129] Gain term G is a reciprocal gain term formed from the reciprocal of the product of gain terms A, C, and D. As will be appreciated further from the discussion herein, gain term G permits the mass flow controller to operate in a consistent manner irrespective of operating conditions by providing to a control loop of the mass flow controller a constant loop gain.

[0130] According to one aspect of the present invention, a system gain term is determined for a particular mass flow controller by determining the composite gain term of various components around the control loop of the mass flow controller. A reciprocal gain term is formed by taking the reciprocal of the system gain term. This reciprocal gain term is then applied to the control loop such that the control loop operates with a constant loop gain. Thus, as the various gain terms around the control loop vary, the reciprocal gain term may be varied in order to maintain a constant loop gain.

[0131] Because the loop gain of the mass flow controller is held constant irrespective of the type of fluid used with the mass flow controller, and irrespective of the operating conditions with which the mass flow controller is operated, the response of the mass flow controller with different fluids and/or operating conditions can be made stable and to exhibit the same behavior as that observed during production of the mass flow controller on a test fluid and test operating conditions.

[0132] Unless otherwise noted, the system gain term is the composite of gain terms around the control loop associated with various components of the mass flow controller that inherently vary as a function of one or more operating conditions. For example, the system gain term in FIG. 1 is composite gain term ACD.

[0133] In block 140 of FIG. 1, a reciprocal gain term G is formed by taking the reciprocal of system gain term ACD and applying it as one of the inputs to the GLL controller. It should be appreciated that the reciprocal gain term may be the reciprocal of fewer than all of the gain terms associated with the various components around the control loop of the mass flow controller. For example, improvements in control and stability may be achieved by forming the reciprocal of composite gain terms AC, AD, CD etc. However, in preferred embodiments, gain term G is formed such that the loop gain remains a constant (i.e., gain G is the reciprocal of the system gain term).

[0134] According to one aspect of the invention, pressure may be sensed at the inlet, and a pressure signal (e.g., pressure signal 190 ) may be produced that can be used in association with a mass flow controller. For example, a pressure signal may be produced that can be used in a flow sensor portion of the mass flow controller to compensate for spurious indications due to pressure transients. Further, the pressure signal may be used for feed forward control of the valve. Also, the pressure signal may be used to adjust the gain in a GLL controller.

[0135] FIG. 2 illustrates a more detailed schematic block diagram of the flow meter 110 . A flow meter refers generally to any of various components that sense flow rate through a flow path, or a portion of a flow path, and provide a signal indicative of the flow rate. The flow meter 110 of FIG. 2 includes a bypass 210 , a sensor and sensor electronics 230 , a normalization circuit 240 to receive the sensor signal FS 1 from the sensor and sensor electronics 230 , a response compensation circuit 250 coupled to the normalization circuit 240 , and a linearization circuit 260 coupled to the response compensation circuit 250 . The output of linearization 260 is the flow signal FS 2 as illustrated in the mass flow controller of FIG. 1 .

[0136] Although not shown in FIG. 2 , in some embodiments, the sensor signal FS 1 may be converted to a digital signal with the use of an analog to digital (A/D) converter so that all further signal processing of the mass flow controller 100 may be performed by a digital computer or digital signal processor (DSP). Although in one preferred embodiment, all signal processing performed by the mass flow controller 100 is performed digitally, the present invention is not so limited, as analog processing techniques may alternatively be used.

[0137] In FIG. 2, a sensor conduit 220 diverts some portion of the fluid flowing through the flow path, with the remainder and majority of the fluid flowing through the bypass. Sensor and sensor electronics 230 are coupled to the sensor conduit and measure the flow rate through the conduit. A pressure transducer 295 is coupled to flow path 200 upstream of the bypass to measure the inlet pressure at the inlet side of the flow path 200 . Pressure transducer 295 provides a pressure signal 270 indicative of the inlet pressure.

[0138] As discussed in the foregoing, pressure transients may cause local fluctuations in the fluid flow that is sensed by sensor and sensor electronics 230 . However, this is considered false flow information as it is not indicative of the flow rate provided to the process at the outlet side of the flow path. As such, flow signal FS 0 may be corrupted with false flow information resulting from transients in the inlet pressure. For example, flow signal FS 0 may contain false flow information resulting from local fluid flow fluctuation caused by fluid rushing to fill a dead volume caused by a pressure pulse or other pressure transient.

[0139] In order to mitigate the effects of the false flow information, compensation filter 280 receives pressure signal 270 from pressure transducer 295 and constructs false flow signal 290 . False flow signal 290 is constructed to model the erroneous response of sensor and sensor electronics 230 due to fluid flow fluctuations caused by pressure transients. That is to say, false flow signal 290 is constructed to equal or closely approximate the false flow information superimposed on the flow signal as a result of pressure transients. One suitable compensation filter was described in detail with respect to FIGS. 8 and 12 . The false flow signal 290 is then subtracted off flow signal FS 0 (e.g., by subtractor 297 ) to provide sensor signal FS 1 having the false flow information effectively removed.

[0140] Sensor signal FS 1 is then further processed in order to provide indicated flow signal FS 2 . In particular, the amount of fluid flowing through the conduit is proportional to the fluid flowing in the bypass. However, within the range of flow rates with which a mass flow controller is intended to operate, the relationship between the flow rate in the conduit and the flow rate in the bypass may not be linear.

[0141] In addition, thermal sensors measure flow rate by detecting temperature changes across an interval of the conduit. Accordingly, in some embodiments, particularly those that implement thermal sensors, there may exist temperature dependencies, particularly at the two extremes of the range of flow rates with which a mass flow controller operates (referred to herein as zero flow and full scale flow, respectively).

[0142] Normalization circuit 240 receives the sensor signal FS 1 and corrects for potential temperature dependence at zero flow and at full scale flow. In particular, when no fluid is flowing through the conduit and/or bypass (i.e., zero flow), the sensor may produce a non-zero sensor signal. Furthermore, this spurious indication of flow may depend on temperature. Similarly, the sensor signal FS 1 may experience fluctuation that is dependent on temperature at full-scale flow. Correction for temperature dependent variation in the signal FS 1 at zero flow may be performed by measuring the value of the sensor signal FS 1 at zero flow at a number of different temperatures, and then applying a correction factor to the signal FS 1 based upon the temperature of the sensor. Corrections for temperature dependent variation of sensor signal FS 1 at full-scale flow may be performed in a similar manner based upon measurement